Organic Reactions VOLUME I EDITORIAL BOARD ROGER ADAMS, Editor-in-Chief WERNER E . BACHMANN JOHN R . JOHNSON LOUIS F. FIESER H. R. SNYDER ASSOCIATE EDITORS A. H . BLATT CHARLES R . HAUSER F . F. BLICKE MARLIN T . LEFFLER NATHAN L . DRAKE ELMORE L . MARTIN REYNOLD C . FUSON RALPH L. SHRINER LEE IRVIN SMITH NEW YORK OHN W ILEY & SON S, INC. LONDON: CHAPMAN & HALL, LIMITED 1942
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
In the course of nearly every program of research in organic chemistrythe investigator finds it necessary to use several of the better-knownsynthetic reaction s. To discover the optimum conditions for the appli-cation of even the most familiar one to a compound not previously sub-jected to the reaction often requires an extensive search of the litera-
ture; even then a series of experiments may be necessary. W hen theresults of the investigation are published, the synthesis, which may haverequired mon ths of work, is usually described without comm ent. Thebackground of knowledge and experience gained in the literature searchand experimentation is thus lost to those who subsequently have occa-sion to apply the general method. The stud en t of preparative organicchemistry faces similar difficulties. The textbooks and laborato ry man-uals furnish numerous examples of the application of various syntheses,but only rarely do they convey an accurate conception of the scope and
usefulness of the processes.For many years American organic chemists have discussed these prob-
lems. The plan of compiling critical discussions of the more importantreactions th us was evolved. Volume I of Organic Reactions is a collec-tion of twelve chapters, each devoted to a single reaction, or a definitephase of a reaction, of wide applicability. The authors have had ex-perience with the processes surveyed. The subjects are presented fromthe preparative viewpoint, and particular attention is given to limita-
tions, interfering influences, effects of structure, and the selection ofexperimental techn iques. Ea ch chapter includes several detailed pro-cedures illustratin g th e significant modifications of the method. M ostof these procedures have been found satisfactory by the author or oneof the editors, but unlike those in Organic Syntheses they have not beensubjected to careful testing in two or more labora tories . When allknown examples of the reaction are not mentioned in the text, tablesare given to list compounds which have been prepared by or subjected
to the reaction. Ev ery effort has been made to include in the tablesall such compounds and references; however, because of the very natureof the reactions discussed and their frequent use as one of the severalsteps of syntheses in which not all of the intermediates have been iso-lated, some instances may well have been missed. N evertheless, the
investigator will be able to use the tables and their accompanying bibli-ographies in place of most or all of the literature search so often re-quired.
Because of the systematic arrangement of the material in the chap-ters and the entries in the tables, users of the book will be able to findinformation desired by reference to the table of contents of the appro-priate chapter. In the interes t of economy the entries in the indexhave been kept to a minimum, and, in particular, the compounds listedin the tables are not repeated in the index.
The success of this publication, which will appear periodically involumes of about twelve chapters, depends upon the cooperation oforganic chemists and their willingness to devote time and effort to thepreparation of the chapters. They have manifested the ir interest al-
ready by the almost unanimous acceptance of invitations to contributeto the work. The editors will welcome their continued interest andtheir suggestions for improvements in Organic Reactions.
If this addi t ion complex i s s table , then the product obta ined by hydroly-sis of the reaction mixture is a /3-ketoester.
XZnO O
— C — C — C O 2R + H X - > — C — C — C O 2R + R'O H + ZnX 2
1 I I0R '
IV V
If the addit ion product decomposes spontaneously, the /3-ketoester (V)may again be the f inal product , or i f the keto group in this ketoester is
react ive and an excess of the organozinc halide (I) is present furtherreac t ion may take place as in equat ions 2 and 3 above .
OZnX 0 x
—G C—CO 2R —» — C— C— CO 2R + Zn
N A P '
O R ' 0 R
Evidence for the existence of the organozinc halide (I) as an inter-media te was provided by G. Dain , 3 who isolated and analyzed the follow-ing compounds .
ZnBr ZnBr
(C H 3)2C C O 2C 2H 6 (C H 3)2C H C H C O 2C 2H 6
Three addi t ion products corresponding to the complex I I were a lsoobta ined .
OZnBr OZnBr OZnBr
C 6H 6C H C H C O 2C 2H 6 C 6H 6C H C H C O 2C 2H 6 C 6H 6C H — C — C O 2C 2H 6
I I / \C H 3 C H ( C H 3)2 C H 3 C H 3
These complexes, therefore, parallel the intermediates formed in thewell-known reac t ions involving the Grignard reagent or s imi lar organo-
meta l lic ha l ides an d carbonyl compo unds. Inde ed, mag nesium m ay beused in place of zinc (p. 16), and apparently the intermediate complexesare analogous. G rignard reage nts cann ot be prepa red f rom a-haloestersand magnesium alone; hence the Reformatsky reac t ion offers a pro-
cedure by which the equivalen t of a Grignard reag ent from an a-haloesteris available for synthe tic work. In th e subsequent discussion theseintermediates will not always be written and only the reactants andmain products will be shown. I t is to be understood, however, th at thesteps shown above are always involved.
Relative Reactivities of R eagents. The order of reactivity of carbonylcompounds in the Reformatsky reaction is RCHO > R 2CO > RCCVC2H5. The order of reactivity of the haloacetates is ICH2CO2C2H5 >BrCH 2C02C 2H5 > C1CH2CO2C2H 5. The a-chloroacetic esters oftenreact slowly or not at all, and the a-iodoesters are not readily available.Consequently, most Reformatsky reactions have been carried out withthe a-bromoesters. Es ters containing a secondary or ter tiar y a-chlorineatom are much more reactive tha n the corresponding primary derivatives
and in some cases are repo rted to give good yields. The three types of a -bromoesters appear to react equally well.
Side Reactions. Various side reactions m ay be expected whenever theReforma tsky reaction is carried out. The intermediate organozinchalide may add to the carbonyl group of the a-haloester used as thereagent; for example, Hann and Lapworth4a have reported that zincand ethyl bromoacetate react to produce ethyl 7-bromoacetoacetate.
OZnBr
2BrCH2CO2C2H6 + Zn -> BrCH2C—CH2CO2C2H6
OC2H6
iB r C H 2C 0 C H 2C O 2C 2H 6 + C2H 6OZnBr
Since a ldehydes and ke tones possess fa r grea ter carbonyl reac t ivi ty than
the es ter group, this s ide reac t ion i s not important when a ldehydes andketo nes are used. M oreo ver, i ts significance m ay be minim ized by using
an excess of the brom oester an d ad ding th e la t te r in successive por t ions .A common side react ion is the coupling of the haloester by the zinc.
C H 2C O 2C 2H 6
2BrCH 2C O 2C 2H B + Zn —> ZnB r2 + |C H 2CO 2C 2Hs
When a l iphat ic a ldehydes or a l iphat ic or a l icycl ic ke tones are used,these may undergo aldolizat ion under the influence of the zinc salts .
*"• H a n n a n d L a p w o r t h , Proc. Chem. Soc., 19, 189 (1903).
Lengthening the Carbon Chain. Lengthening the Carbon Chain of an
Aldehyde without Branching the Chain.
R C H OZn
• R C HOHC H 2C O 2C 2H 6
Dehydration. *
R C H = C H C O 2 C 2 H 5
Reduction (catalytic)
RCH 2CH 2COC1
R C H 2C H 2C O N H 2
Hydrogenolysis
xvO H 2C H 2C H 2O H
Oxidation
R C H 2C H 2C H O
R C H 2C H 2C N
The process may be repea ted , leading to R(CH 2 ) 4 C H 0 .
Lengthening the Carbon Chain with Branching on the a-Carbon Atom.
R '
R C HO + B rC HC O 2C 2H B
R '
R C H 2C H C O 2C 2H 6
R 'I
R C H O H C H C O 2C 2H 6
^ R '
R C H = C C O 2 C 2 H 6
Use of the sequence of react ions outlined under the f i rst example to
conver t the es ter group in to an a ldehyde group leads to the synthesis
of branched-chain esters of the following type.
R '
R [ — C H 2— C H — ] „— C O 2C 2H 6
* The dehydration of /3-hydroxyesters frequently produces a mixture of a,(3- and /3,y-unsaturated esters (see p. 12). Bo th may be reduced catalytically to the satu rated ester.
The nature of the R and R' groups is determined by the ketone andth at of the R " group by the haloester.
Lengthening the Carbon Chain with D ouble Branching on the a-Carbon
Atom.
R '
R C H O -
R '
R—C=O •
B r C C O 2 C 2 H s
R "
R ' "
B r C C O 2 C 2 H 6
R "
Zn
Zn
R '
R C H — C — C O 2C 2H 6
O H R "
R ' R ' "
R — C — C — C O 2C 2H 6
O H R "
Occasionally, hydroxyesters of this type may be dehydrated to /3,7-unsaturated esters which can then be reduced to the saturated esters.However, conversion of these a,a-disubstituted-|3-hydroxyesters to thesatura ted esters is usually bes t effected by refluxing with phosphorus andhydriodic acid.
These five general types of reactions therefore constitute methods forsynthesizing straight-chain and branched-chain hydroxyesters and un-saturated and saturated esters and acids.-
Whether or not the Reformatsky reaction is the best method forlengthening a given carbon chain depends on a num ber of factors. Fo rexample, cinnamic acid may be prepare d by a ny of the following reactions.
Perkin reaction
C 6H 6C H O
Claisen condensation
(C H 3C O )2O(C 6H 6N)
C H 3C O 2C 2H 6
Reformatsky reaction
CeHaCHO + BrCH 2C O 2C 2H 6
Knoevenagel condensation
C E U C H O + C H 2(C O 2H )2
Zn
Yield, % *
C 6 H 6 C H = C H C O 2 H 80
• C 6H 6C H = = C H C O 2C 2H 6 74
iC 6H 6C H = C H C O 2H 72
O H
• C eH sC H — C H 2C O 2C 2H 6 64
2i
C eH 6 C H = C H C O 2C 2H s 57
IC 6 H 6 C H = C H C O 2 H 55
• C6H 6C H = C ( C O 2H )2
I - c o 2
C 6 H B C H = C H C O 2H 80
*• These figures represen t the over-all yields of the p roducts shown, based on ben zaldehyde.
On the basis of yields alone, the Knoevenagel or Perkin condensationwould be preferred for preparing cinnamic acid. From an economicpoint of view, the reaction chosen would depend on the relative cost ofthe reagents and the time involved in the prepara tion. The Reformatsky
reaction would not be selected.However, in the synthesis of an unsaturated acid with branching on
the /3-carbon atom (C 6H 5C = C H C O 2H) from the ketone (C 6H 5COR)
the Reformatsky is the only method of these four which will give goodyields; the Perkin reaction fails to take place, the Claisen condensationleads to an en tirely different product (a 1,3-diketone), and th e K noevena-gel condensation gives low yields for small R groups and fails if R is large.Branching of the chain on both a- and /3-carbon atoms can be accom-plished only by the Reformatsky method.
Synthesis of Arylacetic Acids. The Reformatsky reaction is alsoparticularly well adapted to the synthesis of arylacetic acids or theiresters. Thus, ketones such as 1-tetralone or 1-ketotetrahydrophenan-threne5 give hydroxyesters which are readily dehyd rated to dihydroaryl-acetic esters. The lat ter may be easily dehydrogenated to th e aroma ticcompounds.
C H 2C O 2C 2H 6
-O HC H 2C O 2C 2H 6 C H 2C O 2C 2H 6
C H 2C O 2C H 8 C H 2C O 2C H 3 C H 2C O 2C H a
Synthesis of /3-Ketoesters. Very few applications of the Reformatskyreaction to the synthesis of /3-ketoesters by reactions involving the car-bonyl group of an ester are recorded. E thy l 7-bromo acetoacetate isformed by the action of zinc or magnesium on ethyl bromoacetate. 40
Hamel6
reported 56% yields of ethyl y-chloroacetoacetate by the actionof amalgamated magnesium on ethyl chloroacetate. E thy l y-ethoxyace-toacetate has been prepared in 10 to 33% yields from ethyl ethoxy-
6 B a c h m a n n , J. Org. Chem., 3, 434 (1938).6 H a m e l , Bull. soc. Mm ., [4] 29, 390 (192 1); Stolle, Ber., 41 , 954 (1908) .
a-bromopropionate is used, the a-methyl derivative is produced.8
0
C 2H 6O C H 2C — O C 2H B + B r C H 2C O 2C 2H 6
OZnBr
C 2H 6O C H 2C — C H 2C O 2C 2H 6
O C 2H 6
O
I IC 2H 6O C H 2C — C H 2C O 2C 2H 5
Ethy l 3 ,4 -d ike toad ipa te 9has been ob ta ined f rom e thy l oxa la te , e thy l
ch lo roace ta te , andzinc .
CO 2C 2H B Z C O C H 2C O 2C 2H 6
| + 2C1CH 2CO 2C 2H 6 > • |
C O 2C 2H B COCH2CO2C2H6
O n the o t h e r h a n d , e t h y l a -b ro m o i s o b u t y ra t e is r e p o r t e d to reac t wi th
e thy l oxa la te to form ethyl a ;a - d i m e t h y l m a l a t e .1 0It is e v i d e n t t h a t
reduc t ion t akes p lace du r ing th i s reac t ion .Br
C O 2C 2H 6 I Zn H O C H C O 2 C 2 H 6
I + (C H 3 ) 2 C — C O 2 C 2 H 6 —^± IC O 2 C 2 H 6 (C H 3 ) 2 C — C O 2 C 2 H 6
T h e c h i e f p r o d u c t f r o m e t h y l f o r m a t e , e t h y l c h l o r o a c e t a t e , andz inc is
ethyl trimesate.11
Ethyl formate undergoes the normal Reformatsky
reaction to produce the aldehydoester which then trimerizes.
O OZnCl
C2HBO—C—H + C1CH2CO2C2H6 -^-> C2H6O—C—CH2CO2C2HB
H
O=C—CH2CO2C2H6
|
H
' S o m m e l e t , Bull. soc chim., [4] 29, 553 (1921); Compt. rend., 164, 706 (1912).8 Johnson, J. Am. Chem. Soc, 35, 582 (1913); Johnson andChernoff, / . Am. Chem. Soc,
35, 585 (1913); 36, 1742 (1914).9 Fi t t ig and Daimler , Ber., 20, 202 (1887).
10 Rassow and Ba ue r , Ber., 41, 963 (1908).11 Reformatsky, / . Russ. Phys. Chem. Soc, 30, 280 (1898); J. prakt. Chem., 54, 477
With ethyl a-bromopropionate, the presence of the a-methyl groupin the intermediate aldehydoester preven ts the trime rization. Hence asecond Reformatsky reaction occurs leading to ethyl 2,4-dimethyl-3-hydroxyglutarate.12 Ethyl a-bromoisobutyrate, ethyl formate, and zinc
react in a similar fashion to produce ethyl 2,2,4,4-tetramethyl-3-hydroxy-glutarate.13
Oxidation* of the /3-hydroxyesters, obtained by the Reformatsky re-action on aldehydes, by means of the calculated amount of chromicacid in glacial acetic acid as the solvent, produces ^-ketoesters in lowyields (30-50%).
RCHOHCH2CO2CH3 -^% RCOCH2CO2CH3
Thus, /3-ketoesters with no a-subs tituents m ay be obtained. This isuseful since the Claisen condensation of esters (other than ethyl acetate)yields a-substituted /J-ketoesters (see Chapter 9).
D E H Y D R A T I O N O F T H E / 3 -H Y D R O X Y E S T E R S
If the temperature of the reaction mixture is high it occasionally
happens that the product from the Reformatsky reaction is the unsat-urated ester. However, if the reaction is run in the usual solvents, suchas ether or benzene (p. 15), the chief constituent of the reaction mixtureis the hydroxyester. Because of their tendency to lose water during dis-tillation or saponification,14 the /?-hydroxyesters and their derivatives cansometimes be isolated in the pure state only with difficulty and in pooryields, whereas dehydration of the crude reaction mixtures leads tohigher yields of the u ns atu rate d products.
Dehydration may be accomplished by heating the ^-hydroxyesterwith acetic anhydride, acetic anhydride and acetyl chloride,15 fusedpotassium acid sulfate,16 8 5% formic acid,17 anhydro us formic acid,5' 18 '19
zinc chloride in acetic acid,20 or sulfuric acid 21 of various strengths (20 to
*See p. 22, reference 48.n Reformatsky , Ber., 28, 3262 (1895)." B l a i s e , Compt. rend., 126, 1808 (1898).14 Schroeter, Ber., 37, 1090 (1904); 40, 1589 (1907)." S t o e r m e r a n d F r e d e r i c i , Ber., 41 , 324 (1908).
" W a l l a c h , Ann., 365, 255 (1909)." Rupe, Ann., 369, 321 (1909).18 Cook, J. Chem. Soc, 2524 (1931); Bac hma nn and E dger to n , J. Am. Chem. Soc, 62,
2971 (1940).19 Bergmann and Bograchov, / . Am. Chem. Soc, 62, 3017 (1940)." W a l l a c h , Ann., 314, 147 (1901); Tetry, Bull. soc. chirn., [3] 27 , 600 (190 2).21 Jaworsky and Reformatsky , Ber., 35, 3633 (1902).
The proportion of the two isomeric esters depends on the reagent usedand on the structure of the compound. The dehydration of a number of/3-hydroxyesters by means of four dehydrating agents has been studiedby Kon andNargund.24
The total yield of the mixture of a,/3- and f},y-
unsaturated esters was 80-95%. In Table I is shown the percentageof the total pro duct which was the a,/3-unsaturated ester.
T A B L E I
D E H Y D R A T I O N OF /S-HYDROXYESTERS
0-Hydroxyester
C H 3
C 2H 6— C — C H 2C O 2C 2H 6
O H
C2H6
C 2H 6— C — C H 2C O 2C 2H 6
O H
C S H 7
C H 7 — C — C H 2C O 2C 2H 6
O H
C 2 H 6 CH3
C 2H 6"*—O OHCO 2C 2H6
A H
o <0 H
^ — ' C H 2C O 2C 2H S
*— / C H 2C O 2C 2H 6
[Percentage of <x,0-Unsaturated Ester
P 2O 6
39
23
24
28
19
30
POCI3
62
68
51
43
43
58
SOC12
53
50
31
33
62
50
(fused)KH8O4
57
63
51
28
54
38
I t is occasionally possible to obtain either one of the isomeric dehy-dration products by proper choice of the experimental conditions. Forexample, dehydration of ethyl 1-hydroxycyclohexylacetate with aceticanhydride followed by saponification gives A'-cyclohexenylacetic acid;
if the ester is first saponified and the 1-hydroxycyclohexylacetic acid isdehydrated with acetic anhydride the chief product is cyclohexylideneacetic acid.16
In syntheses of saturated esters or acids it is unnecessaryto separate the a,/3- and/3,7-esters or acids before reduction.
Sometimes cleavage occurs as a side reaction in dehydration of /3-
hydroxyacids. T hus hea t causes the decomposition of a-(l-hydroxy-3-methylcyclohexyl) propionic acid.16
OH
I j +C H 3C H 2C O 2H
C H 3 \ ^
Sulfuric acid causes the cleavage of a,a-dialkyl-j3-hydroxy acids.11
R
CH3CH—C—CO2HH2S°4> CH3CHO + R2CHCO2H
OH R
"Hot concentrated alkalies may also cause cleavage of the molecule incertain instances.
CO2CH3
CH2CO2CH3
OHCH3O
In order to obtain theunsaturated compound andavoid this cleavage itis essential to dehydrate before hydrolyzing.
28'29
SELECTION OF EXPERIMENTAL CONDITIONS. PROCEDURES
In the earlier experiments,2'30the a-haloester, carbonyl compound,
and zinc dust were mixed at room temperature and cooled in order tomoderate the initial reaction which may cause a considerable temperaturerise (60° to 120°). Themixture wasallowed to stand at room tempera-ture for periods ranging from two days to three mo nths. After a finalwarming to 60-70° for two to three hours the mixture was decomposedwith dilute acid. The ester was separated or extracted by a solvent,dried, anddistilled in vacuum.
28 Baohmann , Cole , and Wilds , J. Am. Ckem. Soc, 62, 824 (1940).2 9 B a o h m a n n and Wilds , J. Am. Chem. Soc, 62, 2086 (1940).30 R e f o r m a t s k y andPlesconossoff, Ber., 28, 2838 (1895).
Control of the initial exothermic reaction may be accomplished byaddition of the zinc dust in portions to the other re acta nts or by the useof a solvent. In most of the recent applications of the R eformatsky reac -tion a solvent has been employed. This permits be tter control of the
temperature and facilitates stirring. I t is essential th a t the surface ofthe zinc be kep t clean. The formation of an oily product which coats thezinc may stop the reaction . By the proper selection of the solventmixture it is often possible to keep the addition product in solution or tocause it to crystallize so th a t it is more readily shaken from the metal bythe stirrer. The zinc ma y be suspended in a copper ba sk et 3 1 in order tofacilitate removal of the addition compounds.
By raising the temperature to the boiling point of the solution the
condensation can be effected in a much shorter time (usually one-half tothree hours). A prolonged reaction time 32"'32 b even at a low tem pera turereduces the yield of /3-hydroxyester and increases the amount of high-boiling by-products. The solvents used have been ethyl ether, buty lether, benzene, toluene , and xylene. A mixture of equal amounts ofbenzene and toluene,27 which permits refluxing at temperatures between90° and 105°, is especially advantageous when the carbonyl reagent is aketone. Somewhat lower tem pera tures (70-80°) are be tte r when an ali-
phatic aldehyde is employed. However, where paraformaldehyde isintroduced into the reaction mixture as a source of formaldehyde, thetemperature must be high enough (80-100°) to cause depolymerization.
The reagents should be pure and dry. Th e appa ratus should also beclean and dry and protected from the moisture of the air. The observ-ance of strictly anhydrous conditions not only improves the yield butalso reduces the induction period so tha t the reaction usually s tart s im-mediately. If difficulty is experienced, the add ition of a few crysta ls ofiodine, a little amalgamated zinc, or a very little methylmagnesiumiodide may help in initia ting the reaction . Th e copper complex ofethyl acetoacetate has been used as a catalyst.33 Once started, the reac-tion is quite vigorous. Fo r this reason, only a small portion of thereactants should be used a t th e s ta rt and the bulk of the ma terials shouldbe added gradually. Since a-haloesters are lachrymators and skin irri-tants, precautions should be taken to avoid contact with them.
Zinc dust, zinc foil, granulated zinc, and mossy zinc have been used.Variations in the quality of the zinc are responsible for differences of
opinion concerning yields, cata lysts, and purification procedures. I t is8 1K o h l e r a n d G i l m a n , / . Am. Chem. Soc, 41 , 683 (1919) .82 0 Nieuwland and Da ly , / . Am. Chem. Soc, 53, 1842 (1931).S2 b Lipkin and S tewart , ibid., 61 , 3295 (1939).3 8Kohle r , Her i t age , and Mac leod , Am. Chem. J., 46, 221 (1911).
desirable that the zinc be as pure as possible and have a fresh cleansurface. Any of the forms of zinc may be purified by washing rapidlywith 2% hydrochloric or hydrobromic acid, then with water, alcohol,acetone, and absolute ether. The zinc is then w armed in a vacuum ovenat 100° for a short time and used immediately. A very active metal has
been obtained by immersing 30-mesh zinc in hot (100°) concentratedsulfuric acid containing a few drops of nitric acid.34 After about fifteenminutes the surface becomes bright and the acid is diluted with a largevolume of water. Th e zinc is washed with water and acetone and the ndried. Zinc foil may be cleaned with sandpaper and cut into small strip s.
In certain instances amalgamated zinc and a mixture of zinc dust andcopper powder 32° have been used to effect the condensation. Cadm iumpowder and mixed cadmium-cogper powder are ineffective.32" Mag-
nesium has also been employed in place of zinc but usually results inlower yields. Fo r example, Zelinsky and G ut t3 B used magnesium toeffect the reaction between cyclic ketones anda-bro mo -and a-iod o-es ters.The yields ranged from 20 to 50%, whereas other investigators reportthat when zinc was employed the yields were 56 to 70% for the samereactants. Kon and N arg un d 2 i obtained yields of 48 % in th e condensa-tion of aliphatic ketones with a-chloroesters and magnesium.
Many different experimental conditions have been described in con-
nection with the Reformatsky reaction, and inspection of the literaturereveals tha t the re is no uniformity as regards the procedures. Hencethe yields shown in Tables II, III, and IV of the succeeding part donot necessarily represent the highest attainable.
Four procedures have been chosen to illustrate th e best m ethods avail-able at the present tim e. These procedures not only illustrate the use ofdifferent forms of zinc but also bring out other experimental variations.One of the first three procedures should be selected when the reactantsare easily available. Procedure 1 illustrates the Reformatsky reaction onan aldehyde, and procedures 2 and 3 on ketones. If the carbonyl com-pound is one which does not readily undergo self-condensation in thepresence of zinc salts, then higher yields can be obtained by treating itrepeatedly with zinc and the a-haloester as illustrated by procedure 4.This method is especially advantageous when the ketone is available inonly small amounts.
Ethyl /8-Phenyl-/3-hydroxypropionate.36 In a clean, dry 500-cc. three-necked flask fitted with a mechanical stirre r, a 250-cc. separatory funnel,
34 Fieser and Johnson, / . Am. Chem. Soc, 62, 575 (1940).36 Zelinsky and Gutt, Ber., 35, 2140 (1902); W illatatter and H at t, Ann., 418, 148
(1919).36 Hauser and Breslow, Org. Syntheses,21 , 51 (1941).
of a 2-1. three-necked flask fitted with a mechanical stir rer and a refluxcondenser.
In the flask is placed 70 g. of zinc dust (which has been cleaned with5% hydrobromic acid, washed with water, alcohol, and acetone, anddried). About 50 cc. of th e m ixture is added to th e zinc dust, the stirreris started, and the mixture is heated by means of a steam bath until thereaction sta rts . The remainder of the solution is added at such a ra teth a t gen tle refluxing takes place. After the addition is complete, th estirring and refluxing are continued for one hour. The mixture is thencooled to room tem peratu re and hydrolyzed b y th e addition of 400 cc. ofice-cold 20% sulfuric acid. The benzene layer is separated and theaqueous layer extrac ted with two 50-cc. portions of benzene. The com-bined benzene extracts are washed with a 50-cc. portion of cold 5% sul-
furic acid, then with 25 cc. of 10% aqueous sodium carbonate, andfinally with two 25-cc. portions of water. The benzene solution is driedwith about 25 g. of anhydrous magnesium sulfate and the solvent re-moved by distillation from a steam ba th. The residual oil is distilled invacuum. The ester is a colorless oil boiling a t 134-135 °/9 mm. T heyield ranges from 150 to 161 g. (75-81% ).
Dimethyl Ester of 7-Methoxy-2-methyl-2-carboxy-l-hydroxy-l ,2,3,4-
tetrahydrophenanthrene-1-acetic Acid.28 To 2.5 g. of granulated zinc
(20-mesh, previously washed with dilute hydrochloric acid, water, ace-tone, and dried) and 0.07 g. of iodine in a mixture of 25 cc. of dry ben-zene (thiophene-free) and 25 cc. of anhydrous ether, are added 1.5 g. of7-methoxy-2-methyl-2-carbomethoxy-l-keto-l,2,3,4-tetrahydrophenan -thren e and 0.75 cc. of methyl brom oacetate. As the mixture is refluxedon a water bath, the iodine color fades and the solution becomes cloudy.After five to ten minutes a colorless addition p roduct is deposited. F iveadditions of 2.5 g. of zinc and a trace of iodine are made at forty-five-
minute intervals and an additional 0.75 cc. of methyl bromoacetate isintroduced after one and one-half hours . The mixture is refluxed fora total of four hours, with occasional vigorous shaking to keep the zincfree from adhering crystals.
The addition product is dissolved by adding a little acetic acid andmethanol, and the solution is decanted from the zinc into water. Th emixture is acidified with acetic acid. The ether-benzene layer is sep-arated, the aqueous solution is extracted with benzene, and the com-bined extracts are washed with water and then with dilute aqueousammonia until no more color is removed. Th e residue obtained byevaporation of the ether-benzene solution crystallizes readily frommethano l. The yield is 1.5-1.6 g. The produc t is recrystallized frommethanol containing a few drops of acetone; colorless leaflets, m.p.
125-125.5° are obta ined . B y reworking the m other liquors a tota lyield of 85-90% may be obtained.
E X A M P L E S O F T H E R E FO R M A T S K Y R E A C T I O N
In the tables which follow, a n umber of examples of th e Reformatsky,reaction have been collected to indicate its applicability in synthesis.The tables are undoubtedly incomplete because the reaction frequentlyhas been used as merely one step in a synthesis and hence may not beindexed as a Reform atsky process. As pointed ou t previously (p. 16),because of the wide variations in the experimental conditions employedby different investigators, the yields given are not necessarily the best
obtainable. Fo r the same reason comparisons of yields reported bydifferent authors and often referred to different standards of purity arenot significant.
Aldehydes (Table II). Aliphatic and aromatic aldehydes, saturatedand unsa turated aldehydes undergo the reaction easily. The reactionhas been reported to fail with phenolic aldehydes,38" bu t recent work byConnor m indicates that a reaction does take place.
8 M Reformatsky , J. prakt. Chem., 54, 469, 477 (1896).88 6 Ralph Connor , p r iva te communica t ion .
Ketones (Table III). Aliphat ic , a romat ic , cycl ic , sa tura ted, and
unsa tura ted ke tones have been found to undergo the reac t ion smoothly.In the case of a ketoester , i t is the keto group which reacts with thehaloester . T he reac t ion follows an abn orm al course wi th ha logenated
aliphatic keto nes an d fai ls with pheno lic keto nes . M os t a , /?-unsaturatedketones undergo the normal Reformatsky reac t ion wi th a-haloesters ofmonobasic ac ids . Ho wever , i t has been observed by Ko hler , He r i tage ,a n d Ma c le o d 3 3 tha t methyl bromozincmalonate adds 1 ,4 to benzalace-
OZnBr
C 6H 6C H = C H C O C 6H 6 + BrCH(CO 2C H 3)2 - ^ » C 6 H 6 C H — C H = C C 6 H 6
I CH(CO 2C H 3)2
C 6H 6C H — C H 2C O C 6H 8
CH(CO 2C H 3)2
tophenone . E th yl a-bro mo isob utyra te a lso add s 1,4 to benzalace tophe-n o n e 3 1 in the presence of z inc. W hen ace tone is t rea ted wi th meth yl
bromomalonate in the presence of z inc , the only product i sola ted i s tha tcorresponding to 1,4-addition of the haloester to mesi tyl oxide.71 6 E v i -dently, mesi tyl oxide is formed by the condensat ion of acetone inducedby the RZnX complex .
2CH3COCH3 - » (C H 3 )2C = C H C O C H 3
OZnBr
I
(C H 3)2C = C H C O C H 3 + BrCH(CO 2C H 3)2 - ^ - > ( C H 3 )2C — C H = C — C H 3
9H !O sO O - laO !i (sH O )9H 5O i! O O IaH O sH 5O
8HO ZOOIOHO 9H ZOsH zO sO O - i a H O sH O
' H ^ O ' O O ' H O I9H ZOSOO SHOIO
ja'jsaofBjj-n
' H O O O ' H O ' H O — K ^ S
8H O SHO
!! ( i: H O ) H O zH O H O = H O O O 8H OH SH O ) H O 5H O H O = H O O O 8H O5 ( £ H O ) O = H O H O = H O O O 8 H OH 8 H O ) O = H O H O = H O O O 8 H O
S H O !( H O = H O ) O O E H Oz ( 8 H 0 ) O = H 0 O 0 8 H 0£H O H O = H O O O 8 H O
(« )1H 8O O O iH sO(« )(« ) iH sO O O iH sO ( « )
SH
5O OO
9H
ZO
(« )iH
8O O O
8H O
8H O O O
8H O
SH O O O
£HO
eH O O O £H OSH O O OSH O8H O O O 8H O8H O O O 8H O8H O O O 8H O
76Barbier and Bouveault, Compt. rend., 122, 393 (1896).
77Karrer, Salomon, Morf, and Walker, Helv. Chim. Acta, 15, 878
(1932).78
K u h n and M o r r i s , Ber., 70 , 853 ( 193 7) ." H e i l b r o n , J o n e s , L o w e , and W r i g h t , / . Chem. Soc, 561 (1936) .80 K a r r e r and M o r f , Helv. Chim. Acta, 16, 625 ( 193 3 ) .81 P h a l n i k a r and N a r g u n d , J. Univ. Bombay, 8, Pt . Il l, 184 ( 193 9) .82 W a l l a c h , Ann., 347,328 ( 1 9 0 6 ) .83 W a l l a c h , Ann., 360,26 ( 1 9 0 8 ) .84 G r e e n l e e , P h . D . t he s i s , U n i v . of 111., 1939.86 C l e m o and O r m s t o n , J. Chem. Soc, 1778( 1 9 3 2 ) .86 M y e r s and L i n d w a U , J. Am. Chem. Soc, 60 , 644 ( 1 9 3 8 ) .
87 L u k e s , Collection Czechoslov. Chem. Commun., 4, 81 ( 1 9 3 2 ) . ^88 F r i e d , R u b i n , P a i s t , and E lde r f i e ld , Science, 91 , 435 ( 1 9 4 0 ) . #89 P e r k i n and T h o r p e , J. Chem. Soc, 71, 1169 ( 1897) . ^90 L a w r e n c e , J. Chem. Soc, 71, 45 7 ( 1897) . O91 H a b e r l a n d , Ber., 69, 1380 ( 193 6) . 392 N e w m a n , / . Am. Chem. Soc, 62, 2295 ( 1940) . 293
Bradfield, Hedge, Rao, Simonsen, and Gillam, J. Chem. Soc, 667
(1936).94Adamson, Marlow, and Simonsen, J. Chem. Soc, 774 (1938).
USE OF COMPOUNDS OTHER THAN CARBONYL DERIVATIVES 3 7
When cer ta in oxides a re used in place of carbonyl compounds, rear-rangem ents m ay occur . Fo r example , a -pinene oxide gives the samehydroxyester as the a ldehyde formed by rearrangement . 6 2 I t has beenshown tha t z inc bromide causes rearrangement of the oxide to the a lde-
hyde which then reac t s wi th e thyl bromo ace ta t e to produce the hydroxy -ester . C am phen e oxide ,10 9 norpinene oxide,10 9 and d-A 3-carene oxide 84
react similarly.
C H 3 OH
Cyclohexene oxide produces ethyl /3-cyclopentyl-/3-hydroxypropionatewhen t rea ted- wi th e thyl brom oac eta te an d z inc . T he same ester i sobtained from cyclopentanealdehyde,1 0 8 demons t ra t ing tha t r ing con-
t rac tion has tak en place dur ing the condensa t ion wi th cyclohexene oxide .
:oB r C H 2 C O 2 C 2 H 6
V O H
V-CHCH2CO2C2HB
> - C HO
9,10-Octahydronaphthalene oxide 8S reac t s wi th e thyl bromoace ta t e
and z inc to form a ke tospi ran and a hydroxyester whose s t ruc ture i s
uncertain.
"9Arbuzov, J. Gen. Chem . (U.S.8.R.), 9, 255 (1939).
Preparation of Amides 51p-Homoanisamide 51Anthraquinone-2-acetanilide 522-Hydroxy-3-naphthylacetanilide 52
Preparation of Esters 52Ethyl a-Naphthylacetate 52Dimethyl Ester of 7-Methoxy-2-methyl-2-earboxy-l,2)3,4-tetrahydrophe-
nanthrene-l-/3-propionic Acid 53
SURVEY OP THE ARNDT-EISTERT SYNTHESIS 53
Table of Products and Yields 55
INTRODUCTION
The Arndt-Eistert synthesis is a procedure for converting an acid toits next higher homolog or to a derivative of the homologous acid, suchas an ester or amide. The synthesis, which is applicable to both ali-
phatic and aromatic acids, involves the following three operations.1. Formation of the acid chloride.
2. Reaction of the acid chloride with diazomethane to yield a diazo-ketone.
R—COC1 + 2CH 2N2 -> R—C—CHN2 + CH 3C1 + N 2
I IO
3. Rearrangement of the diazoketone, with loss of nitrogen, in thepresence of suitable reagents and a catalyst (colloidal silver, platinum,copper). An acid is formed in the presence of water, an ester is producedin an alcohol, and an amide results when ammonia or an amine is used.
R—C—CHN2 + HOH
I I0
R—C—CHN2 + R'OH
O
R—C—CHN2 + NH 3
I I0
R — C — C H N 2+R'NH 2
I IO
A g
A g
A g
A g
R—CH2CO 2H + N 2
R—CH2CO2R' + N 2
R—CH2CONH2 + N 2
R—CH2CONHR' + N 2
The discovery that diazoketones can be converted into derivatives ofan acid was made by Wolff
1 and this phase of the synthesis is known asthe Wolff rearrangement. Wolff found, for example, th a t the tre atm en tof w-diazoacetophenone, C 6H 5— C O— C HN 2, with ethanolic ammoniaand silver oxide gave phenylacetamide, CeH 5— CH 2— C ONH 2, in goodyield. This reaction had no synthetic value at the time, for Wolff pre-pared the diazoketones through a complex series of reactions.
The practical application of the Wolff rearrangement as part of a pre-parative procedure awaited the discovery of a method of obtainmg thediazoketones convenien tly. This discovery arose from a study of thereaction between acid chlorides and diazomethane. N ierenstein and hiscollaborators2 made an extensive study of the reaction between aromatic
acid halides and diazomethane, but, strangely enough, they neverobserved the formation of diazoketones but always obtained w-halo-
1Wolff, Ann., 394, 25 (1912).
2 Clibbens and Nierens te in , J. Chem. Soc, 107, 1491 (1915); Lewis , Nierens te in , andRich, J. Am. Chem. Soc, 47, 1728 (1925); Malkin and Nierens te in , ibid., 52, 1504 (1930).
m e t h y l k e t o n e s , R C 0 C H 2 X (X = ha logen) . How ever , A rnd t an d
co-workers,3 ' 4> 8 and shor t ly thereaf te r Robinson and Brad ley , 6 showed
tha t d iazoke tones were ob ta ined in near ly quan t i ta t ive y ie ld when the
acid chloride was added slowly to a cold solution of an excess of diazo-
m etha ne . Th is p rocedure var ied from th a t of N ie ren s te in , who usua l ly
added one mole of d iazomethane to the acid chlor ide , sometimes a ts l ightly e levate d tem pe rat ure s (35°). Ac cording to A rn dt and co-
worker s and Brad ley and Schwarzenbach ,7 the fo l lowing react ions take
place when an acid chlor ide is added to diazomethane.
(a) RCOC1 + C H 2N 2 - > R C 0 C H N 2 + HC1
(6) HC 1 + C H 2N 2 -» CH 3C1 + N 2
(c) RCO CH N2 + HC1 - » RCOCH2CI + N 2
The in i t ia l react ion is the formation of the diazoketone with l iberat ion
of hydrog en chlor ide (o) . T he hydrog en chlor ide th en reac ts w ith a
second molecule of d iazomethane to form methyl chlor ide (b). I f any
of th e h ydro gen chlor ide is no t des troy ed in th is react ion , i t wil l
reac t with th e diazok etone to y ie ld th e co-chloromethylketone (c). In
general , where there is a lways an excess of d iazomethane, react ion (c)
takes place to a very l imited exten t , because the excess d iazomethane
reacts with the hyd roge n chlor ide a lmost as fas t as th e hydroge n chlor ide
is formed. H owe ver , when th e react ion is run so th a t ther e is a lways
an excess of acid chlor ide (by adding the diazomethane s lowly to the
acid chlor ide) , some chloromethyl ketone is formed, especial ly a t h igher
tempera tu res , a l though the h igh y ie lds o f th i s p roduc t ob ta ined by
Nierens te in have no t been duplica ted by o ther inve s t iga to r s .7
With the d iazoke tones r ead i ly ava i lab le , Arnd t and Eis te r t 8 mad e a
s tudy of the Wolf f rear rangement and showed that i t was of qui te gen-era l app l ica t ion . T hey po in ted ou t th a t a com bina t ion of the two reac-
t ions , the formation of the diazoketone f rom acid chlor ides and the
Wolff rear rangement , const i tu ted a new method of lengthening a carbon
cha in by one methylene g roup .
The diazoketones are bel ieved to decompose by way of in termedi-
a tes s imilar to those involved in the Cur t ius rear rangement of acid
3 Arndt, Eistert, and Partale, Ber., 60, 1364 (1927).4 Arndt and Amende, Ber., 61, 1122 (1928).6 Arndt, Eistert, and Amende, Ber., 61, 1949 (1928).6 Robinson and Bradley, J. Chem . Soc, 1310 (1928).7 Bradley and Schwarzenbach, J. Chem . Soc, 2904 (1928).8 Arndt and Eistert, Ber., 68, 200 (1935).
uThe nitrogen is eliminated, and a short-lived radical is
produced which rearranges to the corresponding ketene.
RCOCHNa -> N2 + [RCOCH=] -> RCH=C=O
RCON3 -» N2 + [RCON=] ->• RN=C=O
In several cases the intermediate ketenes have been isolated,12
but
ordinarily they are converted to the acids, esters, or amides by the
water, alcohol, ammonia, or amine present in the reaction mixture.
RCH=C=O + HOH -> RCH2CO2H
+ R'OH -» RCH2CO2R'
+ NH3 -> RCH2CONH2
+ R'NH2 -+ RCH2CONHR'
The rearrangement of optically active diazoketones, in which the carbon
atom attached to the carbonyl group was asymmetric, resulted in the
formation of optically active products except in one or two instances.10
'u
This result is similar to that observed in the rearrangement of optically
active acid azides.
It is considered that the metal catalyst which is usually required forthe reaction accelerates the decomposition of the diazoketone to the
ketene, since in the absence of such a catalyst no rearrangement takes
place and the product formed is a derivative of the ketone. Thus, if
diazoacetophenone is heated with water at 70-80°, benzoylcarbinol is
obtained.1'
4
C6H5COCHN2 + H2O -»• C6H6COCH2OH + N2
If silver is present, rearrangement takes place and phenylacetic acid is
formed. Wolff l found that the addition of powdered silver did notcatalyze the decomposition of diazoacetone in the presence of ammonia,
but that the reaction was rapid if either silver oxide or silver nitrate was
added. Thus, it appears that the catalyst, if it is metallic silver, must be
colloidally dispersed. Arndt and Eistert8
found that even with highly
purified diazoketone there was always a small amount of reduction of
the silver salts which could account for the production of the necessary
catalyst. Powdered copper and platinum have also been used as catalysts
in the rearrangement but much less frequently.9 Eis te r t , Ber., 68, 208 (1935).
10 Lane, Wil lenz , Weissberger , and Wall is , J. Org. Chem., 5, 276 (1940).11 L a n e and Wall is , J. Org. Chem., 6, 443 (1941).12 Schroeter , Ber., 42, 2346 (1909); 49, 2704 (1916); S taudinger and Hirzel , Ber., 49,
I t i s apparen t tha t by the Arnd t -Eis te r t syn thes i s an ac id can be con-
ver te d to i ts nex t h igher homolog by a three -s tep process . T he over-all
y ie ld is o rdinar i ly between 50 an d 80% . O ther well -known metho ds for
accomplishing the same result include the following processes, which arepresented in outl ine form.
[1] RCO 2H -> RCOC1 -> RCH O -> RCH 2O H - » RCH 2Br ->
R C H 2CN (or RCH 2MgBr) -> RCH 2CO 2H
[2] RCO 2H - > RCO 2C 2H 6 - > RCH 2O H - • R C H 2Br -> RCH 2C N
(or RCH 2MgBr) -* RCH 2CO 2H
[3] RCO 2H -> RCOC1 -> R COC N -> RC OCO 2H -> R C H 2C O 2H
The choice of the method to be used depends on several factors, such
as the amount of the acid desired, the type of acid, and the over-all
y ie lds poss ible . M eth od s 1 an d 2 , which consist of mo re s teps tha n th e
Arndt-Eis ter t react ion , of ten g ive lower over-al l y ie lds and require a
longer working t im e. M et ho d 3 general ly g ives poor y ie lds of the prod -
uc t . Th e A rnd t -E is te r t r eac t ion can be ca r r ied th roug h rap id ly , one da yusually being suff icient for the complete synthesis , and i t is thus an ideal
me thod when only smal l am ou nts of the final prod uct are des ired. I t i s
o f in te res t tha t E is te r t 1 3 and Burger and Avakian u have worked suc-
cessfully with amounts of diazoketone as large as 100 g.
Each of the three methods out l ined above involves a more or less
dras t ic reduct ion which may in ter fere with i ts appl icat ion to a com-
pound co ntain ing a n i t ro , quinon e, keto , lac tone , es ter , o r o the r reducible
g roup . T he Arnd t -Eis te r t r eac t ion invo lves no such s tep and can beused for the preparation of molecules which are sensit ive to reducing
agen ts . Fo r example , th e n i t roph enylac et ic acids can be prepa red eas i ly
and in good yields from the nitrobenzoyl chlorides. 8 ' 16
COC1 COCHN 2 C H 2C O 2H
N O 2 N O 2 N O 2
" E i s t e r t , Ber., 69, 1074 (1936).14 Burger and Avakian , J. Org. Chem., 5, 606 (1940).16 Bachmann and Holmes , unpubl i shed re su l t s .
d-Homopilopic acid can be prepared from d-pilopic acid, the lactone
ring remaining intact throughout the synthesis.16
'16a
C2H6CH CHCO2H C2H6CH CHCOCHN2 C2H6CH CHCH2CO2H
O=C CH2 O=C CH2 O=C CH2
d-Pilopio acid rf-Homopilopic acid
An illustration of the conversion of a 0,7-unsaturated acid to its homolog
is the preparation of /3-(2-methylcyclohexenyl)-propionic acid from 2-
methylcyclohexenylacetic acid.17
C H 2 C O O H
A d i c a r b o x y l i c a c i d can be c o n v e r t e d , t h r o u g h its a c i d e s t e r , to its
n e x t h i g h e r h o m o l o g , a p r o c e s s w h i c h w o u l d be diff icul t to a c c o m p l i s h
b y o t h e r m e t h o d s . T h u s , g l u t a r i c a c i d has b e e n c o n v e r t e d to a d i p i c
a c id t h r o u g h the i n t e r m e d i a t e e s t e r c h l o r i d e .
1 8
CH 2COC1 CH 2C O — C H N * C H 2C H 2C O 2C H S C H 2C H 2C O 2H
C H 2C H 2C O 2C 2H 6 C H 2C H 2C O 2C 2Hii CH 2C H 2C O 2C 2H 6 C H 2C H 2C O 2H
T h e A r n d t - E i s t e r t r e a c ti o n is i d e a l for use on c o m p l e x m o l e c u l e s . The
r e a c t i o n is c a r r i e d out at m o d e r a t e l y low t e m p e r a t u r e s so t h a t the
c h a n c e s of d e c o m p o s i n g the m o l e c u l e are not as g r e a t as in s o m e of the
o t h e r s y n t h e s e s . An i n t e r e s t i n g e x a m p l e in the s y n t h e s i s of the sex
h o r m o n e e q u i l e n i n is the c o n v e r s i o n of one of the i n t e r m e d i a t e s to its
n e x t h i g h e r h o m o l o g in g o o d y i e ld ( 8 0 - 8 4 % ) . 1 9
C H 8 CH3
16 Preobrashenski , Po l jakowa , and Preobrashenski , Ber., 68, 850 (1935) .16(1 Pol jakowa , P reo brashenski , and Preobrashenski , Ber., 69, 1314 (1936) ." P l e n t l and Boger t , J. Org. Chem., 6, 669 (1941).18 B a c h m a n n and Sheehan, unpubl ished resul ts ." B a c h m a n n , C o l e , and Wilds , J. Am. Chem. Soc, 62, 824 (1940).
Although diazoketones have been prepared successfully from 2-,23 3-,24
and 4-pyridinecarboxylic acid 24 and from 4-quinolinecarboxylic acid,24"the Wolff rearrangement on the diazoketones has no t been reported. Thecomplete synthesis has been carried out on A^-methylpyrrole-2-carbox-ylic acid.20a
An ingenious application of the synthesis has been m ade in a syn thesisof papa verine. The diazoketon e prepa red from the acid chloride ofveratric acid and diazomethane was allowed to react with homovera-trylamine to give the s ubstitute d amide of homo veratric acid, which wasthen cyclized and dehydrogenated to papaverine.25
C H 2
CH3O
C H 3 ' N H 2 + N 2C H C O
A g2o C H 3 O> C H 3 O
0 C H 3
20 Blicke and M. F . Zien ty , J. Am. Chem . Soc, 63 , 2945 (1941).2 0 a Arndt and Eis te r t , Ger . pa t . , 650,706 [C. A., 32, 595 (1938)].2 1 Crook and Davies , J. Chem. Soc, 1697 (1937).22 Titoff, Miiller , and Eeiohstein, Helv. Chim. Ada, 20, 883 (1937).2 3 Winterfeld and Cosel , Arch. Pharm., 289, 70 (1940).2 4 B a u m g a r t e n a n d D o r n o w , Ber., 73, 44 (1940) ; D ornow, Ber., 73, 185 (1940).2 4 0 Ki ng a nd W or k , J. Chem. Soc, 1307 (1940).26 Ei s t e r t , Angew. Chem., 54, 124 (1941).
Since the pro duc t obta ine d in the A rnd t-E is te r t syn thesis is an ac id, or
a derivat ive which ca n be hydro lyzed t o an ac id, i t is possible to con tinuethe chain- lengthening process . Th e m etho d is par t ic ular ly ada pte d forthe pre pa rat io n of a hom ologous series of acids. In a nu m be r of cases
two methylene groups have been added to the chain of an ac id by carry-ing out two successive Arndt-Eis te r t syntheses .1 7 '26> 27 T wo m e t hy le n egroups hav e been in t rod uced in to dicarboxylic ac ids in o ne opera t ion(bishomologat ion) by th e A rnd t-Eis te r t me thod . T hu s, adipic ac id ha s
been converted to suberic acid, and sebacic acid to decane-l ,10-dicarbox-ylic ac id thro ugh th e in term edia te bisdiazok etone s. 2 8 '2 9
CH 2CH 2COC1 CH2CH2CO—CHN 2 C H 2C H 2C H 2C O 2H
CH 2CH 2COC1 CH 2C H 2C Q — C H N 2 CH2CH 2C H 2CO 2HAdipyl chloride 1,4-bisdiazo- Suberic acid
acetylbutane
Lit t le work has been done on the use of diazo compounds other than
diazomethane in the A rnd t -E i s t e r t syn thes i s . I t has been repor t ed t h a tthe diazoketone obta ined f rom p-ni t robenzoyl chlor ide and diazoethaneyie lded p-ni t rophenylmethylace tani l ide when rearranged in ani l ine .3 0
p-NO2C6H4COCl C H "C H N 2> p-NOsCH^CO—C(CH,)N a
p-NO2C«H 4CO—C(CH3)N2,+ C6H 6NH 2 -» p-NO2C6H4CH(CH3)CONHC6H5 + N ,
Severa l ca rboa lkoxydiazoke tones , RCO—CN 2 C O 2 R' , formed by in ter-action of acid chlorides and diazoacetic ester, 3 1 have been submi t t ed torearrangement .12 The diazoketone prepared f rom a-furoyl bromide andmethyl diazoaceta te yie lded dimethyl a-furylmalonate when rearranged
in methanol in the presence of plat inum. 3 2
L-COBr
CH3OH
— C O— C N 2CO 2CH 3
—CH(CO 2CH 3)2
The chlorides of two hindered acids ha ve been found to resist the act io n
of diazomethane; these are the chloride of the acid ester of homocam-
phor ic acid in which the acid chlor ide group is a t tached to a ter t iary
carbon a tom 33 and mesi toyl chlor ide .18
COC1
C H 2
C H 2
CHa-
C H 3
- C — C 0 C 1i
- C — C H 3
|-CH—CH2 CO2C2 HB
C H 3
Un like th e acid chlorides of carboxy lic a cids, sulfonyl chlorides fail to
reac t wi th d iazomethane .3 4
Funct iona l g roups as pheno l ic hydroxyl , a ldehyde , ac t ive methy lene ,
and a , |3-unsaturated carbonyl groups , which are capable of react ing with
d iazomethane , migh t be expec ted to in te r f e re in the Arnd t -Eis te r t
synthe s is . On ly a few acid chlor ides con tain ing such group s hav e been
studie d. Fr om 4-f luorenonecarboxyl ic acid chlor ide , th e m eth yl es ter
of 4-f luorenoneac etic acid was ob tain ed in 8 4 % yield,3 5 a l though the
parent ketone, f luorenone, reacts with diazomethane. 3 6 Likewise ,
2 -hydroxy-3-naphthoyl ch lo r ide y ie lds the d iazoke tone wi thou t methy l -
a t ion of the hydroxyl group. 3 7 However , i t i s no t ce r ta in tha t the ac id
chlor ide group in o ther compounds contain ing react ive groups wil l react
p refe ren t ia l ly wi th the d iazomethane .One of the s ide react ions that occurs in the preparat ion of the diazo-
keton es is th e forma tion of w-halomethyl keton es . As has a l re ady bee n
poin ted out , th is react ion is no t s ign if icant i f the react ion is car r ied out
a t low tem pe ra tur e in th e presence of an excess of d iaz om etha ne . I f the
diazoketone is t reated with halogen acids , the w-halomethyl ketone can
be obtained in excel len t y ie ld , and th is react ion has been used recent ly
for preparat ive purposes .24"128> 29f 38
R — C — C H N 2 + HC1 -» R— C— CH 2C1 + N 2
I I II0 0
Other side reactions apparently accompany the formation of some diazo-ketones, since the latter are sometimes contaminated with impuritiesas yet unidentified. In view of the reaction between acid chlorides anddiazoacetic ester,31 '32 there is a possibility that the diazoketone formed
83
Li tvan and Robinson , J. Chem. Soc, 1997 (1938).34 Arndt and Scholz, Ber., 66, 1012 (1933).3 6 B a c h m a n n a n d S h e e h a n , J. A m. Chem. Soc, 62, 2687 (1940).38 Sohul tz , Schul tz , and Cochran , J. Am. Chem . Soc, 62, 2902 (1940).37 Krzika l la and Ei s t e r t , J. prakt. Chem., 143, 50 (193S).3 8 H a b e r l a n d , Ber., 72, 1215 (1939).
initially may react with a second molecule of the acid chloride, but thishas not been established.
EXPERIMENTAL CONDITIONS A N D P R O C E D U R E S
The acid chloride used in the first step of the Arndt-Eistert reactionmay be prepared by any of the usual methods, but it should be care-fully purified, by distillation whenever possible. The solven ts and appa-ratus must be scrupulously dry, especially when aliphatic chloridesare employed, in order to avoid hydrolysis. Any free acid formed byhydrolysis will be converted to the methyl ester by the diazomethane,thus contaminating the product and decreasing the yield.
Diazomethane must be prepared with care. I t is extremely toxic, andrepeated exposure to even very low concentrations causes increasedsensitivity to the substance. An account of a case of acute diazomethanepoisoning has been published.39 A good hood with a forced draft isstrongly recommended for work with diazom ethane. D iazomethaneis explosive in the gaseous state, and, although the ethereal solutions,which are generally used, are safe to handle at room temperature orlower, a certain amo unt of care must be exercised. Fo rtunate ly, anether solution of diazomethane can be prepared at 0° from JV-nitroso-
methylurea,40 ' 41 and the solution can be used without purification by dis-tillation. An inexpensive me thod for preparing iV-nitrosomethylureafrom urea and methylamine hydrochloride has been described.42 I tshould be mentioned that A/-nitrosomethylurea has been known toexplode when kept at room temperature, but when stored in a cold placethe compound remains unchanged for months.
The preparation of diazomethane from iV-nitrosomethylurethan a byvon Pechmann's method is convenient for small amounts, although the
diazomethane usually requires purification by distillation. Th e methodconsists in decomposing the urethan by means of a sodium alcholate.Higher alcohols, such as propanol19 and ethylene glycol,44 have beenused to make the aleoholate in order to minimize contamination of thediazomethane.
Diazomethane has been prepared also from hydrazine, chloroform, andpotassium hydroxide.45 A new method which appears attractive, but
39Sunderman, Connor, and Fields, Am. J. Med. Sci., 195, 46 9 (1938).
which has not yet found extensive use, consists in the treatment ofnitrosomethylaminomesityl oxide with sodium isopropoxide.46 Therequisite intermediate is obtained readily from mesityl oxide, methyl-amine, and nitrous acid.
The concentration of diazomethane in a solution is estimated best bytitration with benzoic acid according to the procedure of Marshall andAcree.47
From a consideration of the reactions which occur on interaction of anacid chloride and diazomethane, it is evident that the acid chlorideshould be added to an excess of diazomethane, for in this manner the s idereaction leading to the formation of the co-halomethyl ketone is sup-pressed. A solution (or suspension) of the acid chloride (1 mole) inethe r or benzene is added slowly to a cold (0-5°) solution of diazom ethane
(3 moles) in ether or benzene with swirling or mechanical stirring of themix ture. Generally a brisk evolution of nitrogen takes place. W ithreactive acid chlorides, such as most alipha tic acid chlorides, the reactionappears to be complete as soon as addition has been made, but usuallythe mixture is allowed to stand at 20-25° for an hour or two. W itharomatic and other less reactive acid chlorides, two hours and more(sometimes twelve to twenty-four hours) is generally allowed.
Some diazoketones crystallize from the solution as they are formed, or
when th e solution is cooled to —10° or lower. Usually they a re isolatedby evaporating the solvent under reduced pressure from a water bathheld at 20-30°. As a rule, th e residual diazoketone is satisfactory forrearrangem ent withou t further purification. If it is crystalline, th ediazoketone may be purified by trituration with a small volume of coldsolvent in order to dissolve oily impurities, and many diazoketones havebeen recrystallized. Purification by distillation is no t recommended.Diazoacetone explodes when distilled at atmospheric pressure (113-115°), but it has been distilled without decomposition under reducedpressure.1 While most diazoketones appear to be stable under ordinaryconditions, and some even in cold methanolic potassium hydroxide solu-tion,48 the crystalline diazoketone obtained from cinnamoyl chloride anddiazomethane is unstable and decomposes on stan ding.7
For the rearrangement of the diazoketones to yield acids, esters,amides, and substituted amides, silver oxide is frequently employed.Freshly prepared silver oxide and commercial silver oxide have beenused with equal success. The silver oxide may be prepared by adding a
dilute solution of sodium hydroxide to a solution of silver nitrate (10%)
46Adamson and Kenner, J. Chem. Soc, 286 (1935).
47Marshall and Acree, Ber., 43 , 2323 (1910).
48Reichstein and v. Euw, Helv. Chim. Ada, 22, 1209 (1939); 23, 136 (1940).
until precipitation is just complete, an excess of alkali being avoided.The silver oxide is washed several times with distilled water bydecantation and then filtered by suction and washed well withwater.
In order to prepare an acid, a dioxane solution of the diazoketone isadded slowly to a warm (60-70°) aqueous solution of silver nitrate andsodium thiosulfate or to a suspension of silver oxide in a dilute solutionof sodium thiosulfate. If the conversion to the acid fails to give goodresults, it may be advisable to employ the procedures for making theester or amide, which are obtained generally in higher yields than theacids, and obtain the free acid by hydrolysis of the derivative.
Esters of the homologous acids are prepared by adding silver oxide to
a hot solution or suspension of the diazoketone in an anhydrous alcohol.Methanol, ethanol, and propanol have been used, methanol most fre-quently. The silver oxide is added generally in the form of a slurry inthe alcohol, best results being obtained if it is added in portions over aperiod of an hour or two rather than in one lot.49 The silver oxide isreduced by hot methanol to metallic silver, which usually deposits as amirror on the sides of the flask.
There is an appreciable difference in the rates with which variousdiazoketones rearran ge and form esters. Sometimes the reaction iscomplete in an hour; however, as much as twelve hours may be neces-sary for completion of the reac tion. The presence of unre acted diazo-ketone may be detected by the evolution of nitrogen which takes placewhen a sample of the solution is treated with a drop or two of concen-trated hydrochloric acid.13 If the reaction is slow, it may be advisableto continue the addition of more silver oxide. In a few res istan t cases,the solution was filtered from the sludge of silver and silver oxide and thefiltrate was treate d with fresh silver oxide. In one prepara tion,35 best
results were obta ined by refluxing a suspension of silver oxide in m ethano luntil a thin silver mirror was formed (about fifteen minutes), then add-ing the diazoketone and continuing the refluxing.
The conversion of a diazoketone to an acid amide has been accom-plished by passing ammonia into a cold solution of the diazoketone inethanol con taining a small am ount of silver oxide. The procedure hasbeen reversed also and the diazoketone added to an ethanolic solutionof ammonia, followed by the a ddition of silver oxide or silver nitr ate. A
more widely used scheme consists in treating a warm solution of thediazoketone in dioxane with a 10-28% aqueous solution of amm onia con-taining a small amount of silver nitrate, after which the mixture isheated for some tim e. 8 '1 4 It would appear desirable to take precautions
(use of shield) when heating mixtures containing ammoniacal silvernitrate.
A number of procedures have been employed to prepare anilides fromthe diazoketones. Some have been prepared by the gradual addition of
the diazoketone to boiling aniline ; *•
37
after each addition one waits un tilthe evolution of nitrogen has ceased before making ano ther add ition. Abetter procedure consists in warming a solution of the diazoketone andaniline in ethanol or dioxane containing a small amo unt of aqueous silvernitrate.8
Occasionally the product obtained in a reaction may contain tracesof colloidal silver or silver sa lts . These may be removed by filtering asolution of the compound through alumina, after first making the solutionalkaline if an acid is the product.
Preparation of Diazomethane
Prom iV-Nitrosomethylurea.40 A mixture of 150 cc. of ordinary etherand 45 cc. of 40 % aqueous potassium hydroxide is cooled to 5°. To thisis added, with continuous cooling and efficient stirring or swirling, 15 g.of finely powdered JV-nitrosomethylurea in small portions as rapidly asthe c rystals dissolve (a few minute s). The deep yellow ethereal solution
of diazomethane can be separated from the aqueous layer by decanta-tion or by means of a separa tory funnel. The solution, which containsabout 4.2 g. of diazomethane, is dried for several hours over pellets ofpure potassium hydroxide or over soda-lime. A solution of diazo-methane in benzene may be prepared in the same way. Larger runsmay be made by varying the amounts of material accordingly.
From 7V-Nitrosomethylurethan.19 A solution of 2.8 g. of powderedpotassium hydroxide (85%) in 10 cc. of warm propanol is prepared in a125-cc. Claisen flask; 60 cc. of anhydrous ether is added to the solution,
and the flask is attached to a dry condenser, which is connected to areceiver (a suction flask fitted with a drying tube) containing about10 cc. of anhydrous ether. Th e end of the condenser dips below th esurface of th e ether in the receiver. Through a dropping funnel a solu-tion of 4.5 cc. of nitrosomethylurethan in 10 cc. of anhydrous ether isdropped into the alkaline mixture; the diazomethane is distilled fromthe mixture as it is formed. The ethereal solution contain s between0.72 and 0.9 g. of diazomethane and is suitable for reaction without
drying.Preparation of Acids
Conversion of a-Naphthoic Acid to a-Naphthylacetic Acid.8'26 A solu-
tion of 19 g. of a-n aphtho yl chloride in 50 cc. of absolute ether is added
at 5-10° to a solution of diazomethane prepared from 35 g. of nitroso-methylurea in 500 cc. of ether. After several hours at 20-25°, the etheris removed under reduced pressure, finally at 30°. The crystallineyellow residue of a-naphthoyldiazome tha ne (m.p. of a sample afterrecrystallization from benzene, 54-55°) weighs 18 g. (92%).
A solution of 15 g. of the diazoketone in 100 cc. of dioxane is addeddropwise with stirring to a mixture of 2 g. of silver oxide, 5 g. of an-hydrous sodium carbonate, and 3 g. of sodium thiosulfate in 200 cc. ofwater at 50-60°. Stirring is continued for one hour after addition iscomplete, and the temperature of the mixture is raised finally to 90-100°. The solution is cooled, diluted with water, and acidified withdilute nitric acid. The a-naphthylace tic acid, which precip itates, is fil-
tered from the mixture and recrystallized from water; yield, 10-12 g.(79-88%); m.p. 130°.
Bishomologation. Sebacic Acid to Decane-l,10-dicarboxylic Acid(p. 45). An ethereal solution of sebacyl chloride prepared from 20 g. ofsebacic acid is added slowly to an ethereal solution of diazomethane(prepared from 50 g. of nitrosomethylurea), and the mixture is allowedto stand overn ight. The ether and excess of diazomethane are removedunder reduced pressure, and the residual crysta lline 1,8-bisdiazoacetyloc-
tane is collected; yield, 19.3 g. (77% , based on th e acid ); m.p. 91°, afterrecrystallization from benzene.29
A solution of 6.8 g. of the diazoketone in 100 cc. of warm dioxane isadded with stirring to a suspension of 7 g. of freshly precipitated silveroxide in 250 cc. of an aqueous solution contain ing 11 g. of sodium thiosul-fate at 75°. A brisk evolution of nitrogen occurs. After one and one-half hours at 75°, the black silver residue is removed by filtration, theclear, almost colorless filtrate is acidified w ith nit ric acid, and the decane-
1,10-dicarboxylic acid is extrac ted with ether . From the ether extract,4.5 g. (72%) of crude acid is obtained. After rec rystallization from 20%aqueous acetic acid, it melts at 127-1280.28
Preparation of Amides
Conversion of ^-Anisoyl Chloride to the Amide of /-HomoanisicAcid.
14 To an ethereal solution of diazomethane obtained from 380 g. ofnitrosomethylurea 150 g. of p-anisoyl chloride is added, and the solutionis allowed to stand o vernight. The solvent is removed by distillation,and the crystalline diazoketone is recrystallized from benzene, fromwhich it separates as transpa ren t, hexagonal prisms ; m.p. 90-9 1°; yield,109 g. (70.3%).
A solution of 20 g. of the diazoketone in 100 cc. of dioxane is treatedwith 150 cc. of aqueous amm onia (sp. gr. 0.9) and 30 cc. of 10% aqueoussilver n itr ate solution a t 60-70 °. The mixture is boiled under reflux fortwo hours, cooled, and the p-homoanisamide is precipitated by the a ddi-tion of wa ter. Recrysta llization from ethan ol yields 15 g. (81%) of thepure am ide; m .p. 188-189°.
Preparation of Anthraquinone-2-acetanilide.8 To a solution ofdiazomethane in dioxane (prepared from 35 g. of nitrosomethylurea) isadded 27 g. of anthraquinone-2-carboxylic acid chloride. W hen the re -action is complete, a few cubic centimeters of water are added and then30 cc. of aniline and 30 cc. of 10% aqueous silver nit ra te so lution. Arenewed evolution of gas occurs. The reaction is completed by hea tingon a steam bath. Th e produc t begins to separate while the mixture is
still warm; after cooling, the product is filtered, dried, and recrystallizedfrom xylene, from which the an ilide is obtained as sm all, colorless needles;m.p. 267-268°.
Preparation of 2-Hydroxy-3-naphthylacetanilide.37 To an e therea l
solution of diazomethane prepared from 35 g. of nitrosomethylurea isadded 25 g. of 2-acetoxy-3-naphthoyl chloride. After one-quarter hourat room temperature, the mixture is cooled for one hour at —15°, andthe precipitated diazoketone [23 g. (90% ); m .p. 122-123°, dec ] is
filtered from the mixture.Ten grams of the diazoketone is added in portions to 30 g. of boiling
aniline; after each addition the reaction is allowed to run to completionbefore the next portion is added . The mixture is boiled for a short timeafter all the diazoketone has been added, cooled, and poured into dilutehydrochloric acid. The anilide is filtered from the mixture and recrystal-lized from ethanol or acetic acid ; m .p. 215-216°; yield, 7.1 g. (58% ).
Preparation of Esters
Preparation of the Ethyl Ester of a-Naphthylacetic Acid.26
-8 T h e
diazoketone is prepared from the acid chloride of a-naphthoic acid inthe manner described (p. 50). To a solution of 10 g. of the diazoketonein 150 cc. of ethanol at 55-60° is added a small amount of a slurry ofsilver oxide, prepared from 10 cc. of 10% aqueous silver nitrate andstirred with 30 cc. of ethano l. As soon as the evolution of nitrogen sub-sides, more of the silver oxide is introduced, and this process is continued
un til all the slurry has been added. The mixture is then refluxed for ashort time, trea ted with charcoal, filtered, and eva porated. D istilla-tion yields 8-9 g. (73-82%) of ethyl a-naphthylacetate, boiling at 175-178711 mm.
Preparation of the Dimethyl E ster of 7-M ethoxy-2-methyl-2-carboxy-
l,2,3,4-tetrahydrophenanthrene-l-/3-propionic Acid 1 9 (p. 43 ) . To 4 cc.of ice-cold dry benzene in a 125-cc. filter flask fitted with a drying tubeare added 2 drops of pyridine and then 1.5 cc. of pure thionyl chloride.
To the cold solution is added 1.71 g. of 7-methoxy-2-methyl-2-carbom eth-oxy-l,2,3,4-tetrahydrophenanthrene-l-acetic acid (p. 43) in powderedform. After standing at room tem perature for one-half hour, themixture is warmed to about 40° for ten m inutes. The orange-yellowsolution, containing some pyridine hydrochloride in suspension, isevaporated under reduced pressure; 2 cc. of benzene is added, and thesolution is evaporated again in order to remove traces of thionyl chlo-ride. The crystalline acid chloride is dissolved in 16 cc. of w arm benzene;
the solution is cooled somewhat and decan ted carefully (through a smallplug of cotton in the side arm of the flask) drop by drop into a cold (5°)solution of diazom ethane in ether (prepared from 4.5 cc. of nitrosom eth-ylurethane); during the addition the diazomethane solution is swirledconstantly.
After fifteen to thirty minutes, the ether and excess of diazomethaneare removed under reduced pressure at room tem pera ture. To thecrystalline diazoketone is added 35 cc. of anhydrous methanol, and to
the warm (50°) mixture is added one-half of the silver oxide which hasbeen prepared from 3.6 cc. of 10% aqueous silver nitrate solution andmade into a slurry with me thanol. Th e mixture is warmed on a waterbath at about 60° with frequent swirling. N itrogen is evolved, andafter fifteen to tw en ty m inutes all of the r ath er insoluble diazoketone hasgone into solution. At this time a small amoun t of silver oxide is addedand the heating is continued; further additions of silver oxide are madeevery five minutes, so that after six additions all of it has been added.Then the mixture is refluxed for fifteen minutes, treated with Norit,
filtered, and concentrated to a small volume. On cooling, the produc tcrystallizes; yield, 1.48-1.56 g. (80-84% ); m.p. 97-101°. If the crystalsdarken on exposure to light, a benzene solution of the product is passedthrough a short column of alumina in order to remove traces of silvercompounds present.
SURVEY OF THE ARNDT-EISTERT SYNTHESIS
In the following table are given nearly all the examples of the syn-thesis which had been reported prior to November, 1941. The firstcolumn gives the name or formula, or both, of the acid used as the start-ing ma terial. The acids are listed in the following ord er: aliphatic,cycloalkyl, arylalkyl, arom atic, and heterocyclic acids. Fre quently an
ester or amide of the homologous acid was prepared in the synthesis, andthe derivative was then hydrolyzed to the free acid, the weight of whichwas recorded. Th e second column shows th e product (acid, ester, oramide) which was prepared initially, and the third column indicates the
compound which was isolated. The yields, which are repor ted in thefourth column, represent th e conversion of the startin g acid to the com-pound which was isolated and are the over-all yields for the three steps:preparation of the acid chloride, formation of the diazoketone, andrearrangement of the diazoketone.
Decane-l,10-dicar boxy ie acid amideC2H5CH C H C H 2C O 2H
O C C H 2
d-Homopilopic acidrac-Homoisopilopic acidrac-Homoisopilopamiderac-Ethyl homoisopilopate3-(2-Methylcyclohexenyl)-propionic acidAmide of the above acidEthyl ester of the above acid•y-(2-Methylcyclohexenyl)-butyric acid
diazoethane)o-Bromophenylacetic acidp-Homoanisamide3,4-Dimethoxyphenylacetic acid3,4,5-Trimethoxyphenylacetic acid3,5-Dibenzyloxy-4-methoxyphenylaceticacid2-Biphenylacetic acido-(2-Methyl-6-nitrophenyl)-a-toluanih'dea-Naphthylacetic acido-NaphthylacetamideEt hy l a-naphthylacetate
60 Bachmann, unpublished results.61 Kloetzel, J. Am. Chem. Soc, 62, 1708 (1940).62 Bachmann an d Edgerton, J. Am. Chem. Soc, 62, 2219 (1940).63 Bachmann an d Struve, J. Org. Chem., 5, 416 (1940).54
Bachmann an d Chemerda, / . Org. Chem., 6, 36 (1941).65 Bachmann an d Edgerton, J. Am. Chem. Soc, 62 , 2550 (1940).66 Bachmann an d Sheehan, J. Am. Chem. Soc, 63, 204 (1941).57 Bachmann an d Sheehan, J. Am. Chem. Soc, 63 , 2598 (1941).68 Bachmann an d Carmack, J. Am. Chem. Soc, 63, 2494 (1941).69 Bachmann an d Thomas, / . Am . Chem. Soc, 63, 598 (1941).60 Bachmann an d Thomas, / . Am . Chem. Soc, 64, 94 (1942).61 Bachmann an d Wilds, / . Am . Chem. Soc, 62 , 2084 (1940).
62 Bachmann and Holmes, J. Am. Chem. Soc, 63, 595 (1941).63 Bachmann and Holmes, J. Am. Chem. Soc, 63 , 2592 (1941).64 Bachmann and Holmes, J. Am. Chem. Soc, 62 , 2750 (1940).65 Bachmann an d Ness, unpublished results.66
Marker and Rohrmann, J. Am. Chem. Soc, 62, 900 (1940).67 Fieser an d Kilmer, / . Am. Chem. Soc, 62, 1354 (1940).68 Slotta and Muller, Z. physiol. Chem., 238, 16 (1936).69 Schopf and Winterhalder, Ann., 544, 62 (1940).70 Schonberg an d Warren, J. Chem. Soc, 1840 (1939).71 Gilman, Parker, Bailie, an d Brown, J. Am. Chem. Soc, 61 ,
2844 (1939).72 Gilman and Cheney, J. Am. Chem. Soc, 61 , 3149 (1939).
The replacement of a hydrogen atom by a chloromethyl group in asingle operation has come to be known as chloromethylation. Theprocess may be illustrated by the earliest example, a synthesis of benzylchloride carried out by Grassi and Maselli* in 1898. These authorsused benzene, hydrogen chloride, paraformaldehyde, and zinc chloride.
C6H6 + CH2O + HC1 -» C6H6CH2C1 + H2O
Chloromethylation is of value in synthetic work inasmuch as the— CH2C1 group can be converted to other groups such as — CH 2OH,—CHO, C H 2C N , and— C H 3.
The present review has been limited to nuclear chloromethylation of
aromatic compounds. Typical procedures aregiven, and anattempt hasbeen made to indicate the scope and limitations of the reaction. Thereactions are listed in tabular form.
THE SCOPE AND LIMITATIONS OF THE REACTION
Chloromethylation is generally applicable to aromatic hydrocarbons.Benzene, naphthalene, anthracene, phenanthrene, biphenyl, and many
of their derivatives have been converted to chloromethyl derivatives.Terphenyl, however, resists chloromethylation altogether.2 Monoalkylbenzene derivatives yield para chloromethyl compounds frequentlyaccompanied by lesser amounts of the ortho isomers. A second chloro-methyl group usually can be introduced, andsometimes excellent yieldsof dichloromethyl derivatives are obtained. Examples are the dichloro-methyl derivatives of m-xylene 3
andmesitylene.4
C H S CH3
, C H 2 C 1 I I ' C H 2 C 1
CH
The presence of a halogen atom on the ring causes the reaction to bemore difficult to effect. A lthough such compounds as bromo- andchloro-benzene, bromo- and chlorotoluenes, and p-dichlorobenzene can be
chloromethylated, the yields are frequently low. More highly halo-1 Grass i and Masel l i , Gazz. chim. Hal.,28, II , 477 (1898).2 v . Braun , I rmish , and Nelles , Ber., 66, 1471 (1933) .s
v. B r a u n and Nelles, Ber., 6T, 1094 (1934).4 N a u t a and Dienske , Rec. trav. chim., 55, 1000 (1936) .
genated derivatives generally fail to undergo chloromethylation. Asmight be expected, however, halogen derivatives of polymethylbenzenessometimes react readily to give high yields of chloromethyl compounds.Bromomesitylene is an example.5
CH
+ CH 2O + HC1 - *
N itro groups tend to inhibit the reaction. Nitrobenzene,6' 7 o-nitro-toluene,6 p-nitrotoluene,6 nitromesitylene,7 and 1-nitronaphthalene havebeen found to give chloromethyl derivatives, but usually in low yields.m-Dinitrobenzene and 1,3,5-trinitrobenzene, as well as 0- and p-chloro-nitrobenzene, fail to react.6
Ketones are generally unreac tive. Acetophenone appears to react,7
but benzophenone8 and anthraquinone6 are recovered unchanged.However, chloromethylation is successful with ketones such as acetome-sitylene, acetoisodurene, and 2,4,6-triethylacetophenone.8
Phenols, as mig ht be expected, react so readily th a t the reaction gen-erally goes too far, yielding polymeric materia ls. The presence of anitro group counteracts this tendency; satisfactory yields from nitro-phenols have been reported.9 '10 ' u A suitable device for getting aroundthe difficulty with phenols is to convert them to esters by treatm en t w ithethyl chlorocarbonate; the ethyl aryl carbonates can be chloromethylatedsuccessfully.12'13> u
The most important side reaction is that leading to the formation of
the corresponding diarylmethan e derivative. Highly reactive com-pounds of many sorts—naphthalene, anisole, phenols, polymethylben-zenes, etc.— tend to yield this typ e of produc t, and i t is often difficult orimpossible to isolate the interm ediate chloromethyl deriva tive. Ex am -ples are a- and /3-naphthol.16
5 Fuson , Kne is l ey , L indsey , Rabjohn , and Spera t i , unpubl i shed work .8 Stephen , Shor t , and Gladding , J. Chem. Soc, 117, 510 (1920).'Vavon, Bol le- , and Cal in , Bull. soc. chim., (5) 6, 1025 (1939).8 F u s o n a n d M c K e e v e r , J. Am. Chem. Soc, 62, 784 (1940).9
Stoe rmer and Behn , Ber., 34, 2455 (1901).10 Buehler , Kirchner , and Diebel , Org. Syntheses, 20, 59 (1940).11 Ger. pat., 132,475 (1900) [Chem. Zentr., 73, II, 81 (1902)].n Sommelet , Bull. soc. chim., [4] 53, 853 (1933).13 Sommele t and Marszak , Compt. rend., 198, 2256 (1934)." S o m m e l e t , Compt. rend., 197, 256 (1933).16 Castiglioni, Gazz. chim. Hal., 67, 324 (1937).
Aromatic amines react very readily, but it has not been possible toisolate their simple chloromethyl derivatives.16 These could hardly beexpected to be stab le, since the highly reac tive chloromethyl group wouldundoubtedly condense with any amino group that might be present inthe molecule.
In a study of the effect of substituents on the ease of chloromethyla-tion of benzene by chloromethyl ethe r in the absence of a catalys t, Vavon,Bolle, and Calin 7 have found that the rate is increased by —CH 3,— C 2H 5, — C 3H 7, —OCH3, and — OC 3H 7, and diminished by — Cl, —Br,— I, — CH2C1, — CO 2H, and — NO 2. These effects a re illustrated by thefollowing relative rates of reaction.
The procedure for chloromethylation has been modified in numerousways. The formaldehyde may be added as formalin, or it may be gen-erated in the reaction mixture by depolymerization of paraformaldehyde(trioxymethylene). (The terms paraformaldehyde and trioxymethylene,used interchangeably in the literature, refer to the polyoxymethylenes—polymers having the structure HOCH 2O(CH 2O )nC H 2OH . The trimer
(CH2O)3, melting at 62-63°, is called aZpfta-trioxymethylene.
17
It isanhydrous, whereas paraformaldehyde generally contains from 2 to 5%of water.) Ins tead of formaldehyde an d hydrochloric acid, diethyl ordimethyl formal and hydrochloric acid may be used. W hen chloro-methyl ethers or dichloromethyl ether are employed, the reaction usuallycan be effected without hydrochloric acid.
Ca talysts may or ma y no t be required. Among the catalysts whichhave been found to be especially useful are zinc chloride, sulfuric acid,and acetic acid. Yields with p-bromotoluene are increased abou t thre e-fold by mixing a little aluminum chloride with the fused zinc chloride. 18
16 W a g n e r , J. Am . Chem. Soc, 55, 724 (1933)." P r a t e s i , Gazz. chim. ital., 14, 139 (1884).18 Fieser and Seligman, J. Am. Chem. Soc, 67, 942 (1935).
introduced the —CH2C1 group into aromatic hydrocarbons by
means of a mixture of formalin or parafoimaldehyde and hydrochloric
acid in the presence of zinc chloride. Darzens and Levy,20
in their syn-
theses of derivatives of naphthalene, employed paraformaldehyde and
hydrochloric acid in acetic acid solution. Quelet and his co-workers,21
"28
who have carried out numerous syntheses starting with aryl ethers,
employed formalin and hydrochloric acid, with or without a catalyst,
and modified the technique according to the sensitiveness of the chloro-
methylation product which was expected. Vavon, Bolle, and Calin,7
as has already been stated, developed a technique permitting them to
follow the course of the reaction and to study the influence of substitu-
ents on the ease of introduction of the —CH2C1 group. They used
chloromethyl ether, without a catalyst, usually in acetic acid solution.
The most successful method for the chloromethylation of aromatic
hydrocarbons is that of Blanc.19
It has been modified in various ways.
The preparation of benzyl chloride illustrates one of these variations.
Chloromethylation of Benzene 19
(Method of Blanc)
+ CH2O + HC1
A mixture of 600 g. (7.7 moles) of benzene, 60 g. (2 moles) of para-
formaldehyde,* and 60 g. of pulverized zinc chloride f is heated to 60°
with stirring. While this temperature is maintained, a rapid stream of
hydrogen chloride is passed into the reaction mixture until no more gas
is absorbed (about twenty minutes). The organic layer is removed,
* It is possible to use 40% formalin in place of para formaldehyde . In this case morezinc chloride is requi red . The fol lowing proport ions are mos t s a t i s fac to ry : 400 g. of ben-
zene, 75 g. of 40% formal in , and 100 g. of pulverized zinc chloride. The reac t ion is carr iedo u t as described; if allowed to run twelve hours , a70 % yie ld of d i p h e n y l m e t h a n e is obta ined .
t If the propor t ion of zinc chloride is increased, the yield of dichloromethyl de r iva t ive is
correspondingly greater ; if less zinc chloride is used, a lmost no dichloromethyl compound is
produced but the yield of benzyl chloride is dimin i shed .19 Blanc , Bull. soc. chim., [4] 33, 313 (1923).20 Darzens and Levy, Compt. rend., 202, 73 (1936).2 1 Quele t , Compt. rend., 198, 102 (1934) .22
Quele t and Anglade , Bull. soc. chim., [5] 3, 2200 (1936).23 Quele t and Allard, Bull. soc. chim., [5] 4, 620 (1937).24
Q u e l e t , Bull. soc. chim., [4] 5 3 , 510( 1 9 3 3 ) .26
Q u e l e t , Compt. rend., 1 9 6 , 1 4 1 1 ( 1 9 3 3 ) .26 Quele t , Bull. soc. chim., [5] 1, 539 (1934).27 Quele t , Bull. soc. chim., [5] 1, 904 (1934).28 Quele t and Allard, Compt. rend., 205, 238 (1937).
washed with water and the n w ith dilute sodium bicarbo nate,* dried overcalcium chloride, and fractionally distilled. After the excess benzenehas been removed there is obtained 200 g. (79%) of benzyl chloride; b.p.70° (15 mm.).
There are also produced about 12 g. of p-xylylene dichloride, m.p.100°, and a small amount of diphenylmethane.Although the reaction usually is carried out with zinc chloride as the
catalyst, sulfuric acid and aluminum chloride have been used also.These catalysts are sometimes objectionable because they tend to favorthe formation of diphenylmethane derivatives. Fo r the chloromethyla-tion of compounds which do not react readily, stannic chloride has some-times been found to be a superior ca taly st. 6 '29 The use of stannic chlor-ide as the ca talys t is exemplified by the prepa ration of 2,4,6-triisopropyl-
benzyl chloride. This method is interesting also because chloromethylether is used in place of formaldehyde or paraformaldehyde.
Chloromethylation of 1,3,5-Triisopropylbenzene6
C H ( C H 3 )2
=\
(C H 3)2CH— C \ + CH 3OCH 2C1 S n C U >
CH(CH 3)2
=\(C H 3)2CH— i }—CH 2C1 + CH 30H
The chloromethyl ether is prepared by the method of Reyschuler.30
Three hundred grams of paraformaldehyde and 200 cc. of methanol aremixed and cooled. A rapid stream of hydrogen chloride is passed throug hthe mass until two layers form and all the paraformaldehyde has dis-appeared. I t is necessary to keep the mixture cool to preve nt th e forma-tion of methylal. A bout 300-400 g. of hydrogen chloride is required.The upper layer is separated, dried over calcium chloride, and fraction-ated several times. Th e product boils at 57-59° and is abou t 90% pure.By washing with concentrated hydrochloric acid, it is possible to obtain
a product which is 95 % chloromethyl e ther.* I t is absolutely necessary to remove all the zinc salt by the washings. W ithout this
precaution the product almost completely resinifies during the distillation period.29
S o m m e l e t , Compt. rend., 1 5 7 , 1 4 4 3 ( 1 9 1 3 ) .30
R e y s c h u l e r , Bull. soc. chim., [4] 1 , 11 95 (190 7) .
A mixtu re of 300 g. (1.47 moles) of 1,3,5-triisopropylbenzene* an d200 g. (2.5 moles) of chloromethyl ether is diluted with 600 cc. of carbondisulfide an d cooled to 0°. To th is solution is added, over a period of onehour, 120 g. (0.46 mole) of stan nic chloride. The reaction m ixture isstirred during the addition an d for one hour afterward. I t is poured onice, and the organic layer is separated and dried over calcium chloride.Removal of the solvent and distillation of the residue in vacuum givesthe benzyl chloride in yields of 300-315 g. (81-85%); b.p. 129-130°(4 m m.).
The chloromethylation of highly alkylated benzenes generally can beaccomplished witho ut an y ca talyst. I t is sufficient to tr ea t the hydro-carbon with a mixture of formaldehyde and concentrated hydrochloric
acid.3'4 Th e chlorom ethylation of p-xylene, for example, is conducted inthe following manner.
Chloromethylation of jp-Xylene8
CH2C1
C H 3
C1H 2 C
v/
= (J
C H
<—C
2C1
' H 3 ,
C1H 2(
C H 3 -
C\
<
CH2C1
II I I I
One mole of the hydrocarbon is mixed with an equal weight of 37%formalin (1.3 moles of formaldehyde) and five times its weight of concen-tra ted hydrochloric acid. The mixture is stirred a t 60-70° for sevenhours, during which time a stream of hydrogen chloride is introduced.The resulting oil is tak en u p in ether, an d the solution is dried. Distilla-tion gives 106 g. of a fraction which is chiefly 2,5-dimethylben zyl chloride(I); b.p. 103° (12 mm .). A second fraction, am ounting to abo ut 10 g.,consists mainly of a
1,a
4-dichlorodurene (I I) ; m .p. 133°. A very small
amount of a2
,a '-dichloroprehnitene (I II ) (m.p. 68-70°) also can beisolated.
The chloromethylation of naphthalene has received much attention.Although, by the use of petroleum ether in the Blanc method, the reac-
* The Dow Chemical Company product, Alkazene-13, was used.
tion gives yields of 30% of the theoretical amount,19 other methods havebeen found to be more useful. Darzens and Le vy 20 and, more recently,Ruggli and Burckhardt,31 Jones,32 Fieser and Novello,33 Fieser andGates,34 and Cambron 35 have obtained the chloromethyl derivative byusing a large amount of glacial acetic acid as a solvent for the hydro-carbon. Cole and D odds 36 preferred to carry out the reaction in anaqueous mixture w ith sulfuric acid as the catalyst.
The procedure of Cambron is as follows.
Chloromethylation of Naphthalene3 S
A m ixture of 288 g. (2.25 moles) of the hydrocarbon, 90 g. (3 moles) ofparaformaldehyde, 250 g. of glacial acetic acid, 280 cc. of concentrated
hydrochloric acid, and 135 cc. of syrupy phosphoric acid is heated, withefficient stirr ing , at 98-100° for four and one-half hours . The reactionmixture is the n poured into 2 1. of cold water. The aqueous layer isdecanted from the heavy oily layer, which is washed two or three timeswith 2-1. portions of water. After each washing the wa ter is removed bydec antation . The oil is filtered to remove the small am ount of solidmaterial and distilled under reduced pressure. The yield of a-chloro-methylnaph thalene is 223 g.; b.p . 145-160° (6-8 m m.). This is 56 .5%
of the theoretical yield based on the amount of naphthalene used.Phenols and their ethers, as has been indicated, react much more
readily tha n the hydrocarbons. Fo r anisole and the methyl cresyl ethers,monochloromethylation with 35-40% formalin and hydrochloric acid ismost successful if conducted at 0-15° and witho ut a ca talys t. H ighertemperatures and the presence of zinc chloride favor the formation ofdiphenylmethane derivatives and also dichloromethylation products.Phenyl esters, hydroxy aldehydes, ethers of hydroxy aldehydes, nitro-
An interesting illustration is the synthesis of 2-hydroxy-5-nitrobenzylchloride by chloromethylation of p-nitrophenol. M ethylal is used as thesource of formaldehyde, and a little sulfuric acid is added to acceleratethe reaction.
31 R u g g l i a n d B u r c k h a r d t , Helv. Chim. Ada, 23 , 441 (1940).32 Jones, U. S. pat. , 2,212,099 [C. A., 35, 462 (1941)].33 Fieser and Novel lo , J. Am. Chem. Soc, 62, 1855 (1940).
34 Fiese r and Ga tes , J. Am. Chem. Soc, 62, 2335 (1940).36 C a m b r o n , Can. J. Research, 17B, 10 (1939).86 Coles and Dodds , J. Am. Chem. Soc, 60, 853 (1938).
A mixture of 50 g. (0.36 mole) of p-nitrophenol, 650 cc . of concent ra tedhydrochloric acid, 5 cc. of concentrated sulfuric acid, and 76 g. (1 mole)of m ethyla l is st i rred for four to five ho urs at 70 -72°. D urin g this period
hydrogen chlor ide i s passed in to the reac t ion mixture . A bout an ho urafter the react ion is begun the 2-hydroxy-5-ni trobenzyl chloride beginsto separa te . I t is remove d by f il t ra t ion af te r the reac t ion mixture ha sbeen chilled. T he yield is 46 g. (6 9% ).
Ketones having mesi tyl , duryl , i soduryl , or o ther highly a lkyla tedaryl radica ls undergo chloromethyla t ion in yie lds of 25 to 88%. 8 T h eprocedure employs paraformaldehyde and concent ra ted hydrochlor icac id, bu t no ca ta ly st . T he chloromethyla t ion of ace tom esi tylene gives
very sat isfactory results.
Chloromethylation of Acetomesitylene8
C HCOCH3
C H 3
COCH3C H 3
CH 2O + HC1
C H 3 C H 3
A mixture of 40 g. (0.25 mole) of acetomesitylene, 9 g. (0.3 mole) ofparaformaldehyde, and 150 cc. of concentrated hydrochloric acid is
shaken on a mechan ica l shaker ove rn ight a t room tem pera ture . T hea 3-chloroacetoisodurene precipi tates from the react ion mixture in clus-te rs of a lmost w hi te needles. The se are remo ved by ni t ra t io n an d washed
with wa ter . T he y are recrystall ized from low-boiling petr ole um e ther,
then from me than ol . Th ere i s obta in ed 40 g . (77%) of pu re ma ter ia l ;m.p . 74.5-75.5° .
The expectation that condensations analogous to chloromethylationwould take place if other aldehydes or other halogen acids were em-ployed has been realized in a number of instances.
Brom omethylation. B y the use of hydrogen brom ide in place ofhydrogen chloride it has been possible to prepare bromomethyl deriva-tives.37 a-Bromomethylnaphthalene,20 benzyl bromide,6 p-chlorobenzylbromide,6 and a
l,a
4-dibromo-p-xylene 6 have been made in this way.Ethyl anisate,7 salicylaldehyde,38 salicylic acid,39 and phenyl ether 40
also undergo bromo methylation. D arzens 41 states that the methodis general but that the yields are lower than in chloromethyla-tion.
Iodomethylation. Iodomethylation has been reported by Sandin andFieser ** who converted 9-methyl-l,2-benzanthracene (I) to 9,10-dimethyl-l,2-benzanthracene (III) through the intermediate iodomethylderivative (I I). Th e iodomethylation was carried out by treating thehydrocarbon with chloromethyl ether or paraformaldehyde in glacial
i n
acetic acid solution and then adding hydriodic acid (sp. gr. 1.7). Thebright yellow iodomethyl compound formed in yields of 90%.
This preparation is especially interesting in the light of the failure ofBadger and Cook to isolate the corresponding chloromethylation prod-uct.43
Chloroethylation. By the use of paraldehyde in place of formaldehydeit has been possible to effect chloroethylation . Anisole and its homologs,when treated with paraldehyde and hydrochloric acid, give the corre-
87T s c h u n k u r a n d E i c h l e r , G e r . p a t . , 5 0 9 , 14 9 [C . A . , 2 5 , 7 1 1 ( 1 9 3 1 ) ; Chem. Zentr., 1 0 2 ,
I , 3 6 0 (193 1) ] .38
G e r . p a t . , 114,194 (1900) [Chem. Zentr., 7 1 , I I , 9 2 8 (1900)] .39 F . Bayer and Company, Ger. pat . , 113,723 (1900).40
Brunner, Ger. pat . , 569,570 [Chem.Zentr., II, 609 (1933)].4 1
Darzens, Compt. rend., 208, 818 (1939).42
Sandin and Fieser, J. Am. Chem. Soc, 62, 3098 (1940).43
sponding chloroethyl derivat ives in yields of 40-60%. 2 7 ' 44 ' 46> 46' 47 T h e
synthesis of 4-methoxy-a-chloroethylbenzene is an example.
CHCICH3
C H 3CH O + HC1 -> + H 2 0
OCH3 OCH;
Xylene a lso has been chloroe thyla ted.4 6 The chloroe thyl der iva t ivesreadily lose hydrogen chloride, yielding the corresponding vinyl deriva-t ives . Anisole gives a 90% yield of p-vinylanisole accompanied by a
10% yield of the ortho isomer.46
Ch loroac etaldehy de can be used also ; with anisole i t gives a , |S-di-chloroethylanisole.48
O C H 8 OCH 3
Chloropropylation. Chloropropylat ion of anisole followed by dehydro-
chlorinat ion furnishes a synthesis of anethole. 4 6
OCH3 OCH3 OCH
CHC1CH 2CH 3 C H = C H C H 8
Anethole
Chlorobutylation. Chlorobutylat ion of anisole has also been re-
por ted.4 6 ' 46 ' 49 By us ing butyra ldehyde , Ducasse 49 obta ined 2-methoxy-
5-methyl-a-chlorobutylbenzene in a yie ld of 3 0% . Ch loroisob utyla t ion
of anisole has likewise been effected.46
" Q u e l e t , Compt. rend., 199, 150 (1934).« Sommelet and Marszak, Fr . pa t . , 787,655 [C. A., 30, 1185 (1936)].46 Quele t , Bull. soc. chim., [5] 7 , 196 (1940).47 Quele t , Bull. soc. chim., [5] 7, 20 5 (1940).48 Quele t and Al lard, Bull. soc. chim., [5] 7, 215 (1940 ).49 Ducasse , Bull. soc. chim., [5] 3, 2202 (1936).
7 4 C H L O R O M E T H Y L A T IO N O F A R O M A T IC C O M P O U N D S
TABLES OF DATA ON CHLOROMETHYLATION
The following tables list compounds which have been chloromethyl-ated, together with the reaction products. References have been given
to pertinent litera ture sources. W here available, the per cent yield isindicated in parentheses following the reference number.The compounds have been arranged, according to the nature of the
parent substances, in five groups: hydrocarbons (Table I), halogen andnitro derivatives of hydrocarbons (Table II), phenols and phenyl esters(Table III), ethers and thioethers (Table IV), and aldehydes andketones (Table V).
60 V a v o n andBolle , Compt. rend., 204, 1826 (1937) .61 Sommelet , Bull. soc. chim., [4] 15, 107 (1914).62 B e r t , Compt. rend., 186, 373 (1928).6 3 H o c h , Compt. rend., 192, 1464 (1931) ." B a k e r and N a t h a n , / . Chem. Soc, 1840 (1935).66 Smi th and M o M u l l e n , J. Am. Chem. Soc, 58, 629 (1936).
«• Pinke rvil le, U. S. pat., 2,219,873 (1940)." B e r g , Rocmiki Chem., 14, 1249 (1934).68
R e i c h s t e i n a n d O p p e n a u e r , Helv. Chim. Acta, 16,1 3 7 3 (193 3 ) .69
A r n o l d , J. Am. Chem. Soc, 6 1 , 1 4 0 5 (193 9) .69 0
A n d e r s o n a n d S h o r t , J. Chem. Soc, 4 8 5 (193 3 ) .60
R e d d e l i e n andL a n g e , G e r . p a t . , 508,890 (1929) [C.A . ,2 5 , 716(193 1) ; Chem. Zenir.,
102, I , 1 8 3 0 (1931)] .
" B u e h l e r , B r o w n , H o l b e r t , F u l m e r , andP a r k e r , J. Org. Chem., 6, 902 (1941) .82
M a n s k e andL e d i n g h a m , Can. J. Research, 1 7 B , 14 (193 9) .63 Robl in andHeohenble ikner , U. S. pat. , 2,166,554 (1939) [C . A., 33,8628 (1939); Chem.
Zentr., 110, II , 4354 (1939)].64 I. G. F a r b e n i n d u s t r i e , Fr. pat., 695,095 (1930) [Chem. Zentr., 102, I, 2396 (1931)].66 B r u n n e r and G r e u n e , V. S. pat., 1,910,462 (1933) [C. A., 27, 4092 (1933); Chem.
Zentr., 104,II , 4355 (1933)].66 Reddel ien andL a n g e , Ger. pat., 519,807 (1929) [C. A., 25, 3363 (1931); Chem. Zentr.,
102, II, 124 (1931)]." L a n g e , Ger. pat., 533,132 (1930) [C. A., 26, 4064 (1932); Chem. Zentr., 102, II, 2659
(1931)].68 Reddel ien and L a n g e , XJ. S. pat., 1,853,083 (1932) [Chem. Zentr., 1 0 3 ,1 , 3894 (1932)].6B B r u n n e r and G r e u n e , Ger. pat., 533,850 (1929) [C. A., 26, 734 (1932); Chem. Zentr.,
103, II , 3159 (1932)].70 Cook, Dans i , Hewet t , Iba l l , Mayneord , and Roe,J. Chem. Soc, 1319 (1935).71 v. B r a u n , Ber., 70, 979 (1937).72
E. Kamp, Dissertation, Frankfurt, 1936.73 W o o d andFieser , J. Am. Chem. Soc, 62, 2674 (1940).7 3 a Smi th andH o r n e r , / . Am. Chem. Soc, 62, 1349 (1940).74 Miller, Ph.D. thes i s , Unive rs i ty of Il l inois , 1940.76 B r a c e and K a h n , J. Am. Chem. Soc, 60, 1017 (1938).7 6
1 . G. Farben indus t r i e , Br i t , pat., 473,522 (1937) [C. A., 32, 1946 (1938)]." I . G. F a r b e n i n d u s t r i e , Fr. pat., 695,602 (1930); Brit , pat., 347,887 (1931) [Chem.
Zentr., 103, I, 2997 (1932)].78
I. G. F a r b e n i n d u s t r i e , G e r . p a t . , 494,803 (1930) [Chem. Zentr., 1 0 1 , I I , 466(1930)] .79 Buehle r , Bass , Dar l ing , and L u b s , J. Am. Chem. Soc, 62, 890 (1940).79(1 Buehler , Deebel , and E v a n s , J. Org. Chem., 6, 216 (1941).796 Smi th and Opie , / . Am. Chem. Soc, 63, 937 (1941).79c Smi th andCar l in , J. Am. Chem. Soc, 64, 524 (1942).79<* Q ue le t and A n g l a d e , Compt. rend., 203, 262 (1936).80
B r u n n e r , G e r . p a t . , 567,753 (1928) [C.A . ,2 7 , 2694 (1933)] .81
B r u n n e r , U . S. p a t . , 1,887,396 (1933) [C. A . , 2 7, 1 3 5 9 (193 3 ) ] .82 Quele t and Allard, Bull. soc.chim., [5] 3, 1794 (1936).83 Quele t , Bull, soc chim., [4]5 3 , 851 (1933).S 4Ofner , Helv. Chim. Acta, 18, 951 (1935).S 6
1 . G. F a r b e n i n d u s t r i e , Fr. pat. , 695,477; Bri t , pat., 347,892 (1930) [Chem . Zentr., 103,
I, 2997 (1932)].8 6 1 . G. Farben indus t r i e , Br i t , pat., 347,892 (1930) [C. A., 26, 2750 (1932)].s 8 a D u c a s s e , Bull. soc.chim., [5] 2, 1283(1935).87
Q u e l e t , Compt. rend., 1 9 5 , 1 5 5 ( 1 9 3 2 ) .88
Q u e l e t , Compt. rend., 1 9 8 , 2 1 0 7 ( 1 9 3 4 ) .89
Q u e l e t a n d G e r m a i n , Compt. rend., 2 0 2 , 1 4 4 2 (193 6) .90
B r u n n e r , G e r . p a t . , 569,569 (1933) [C.A . , 2 7 , 3723 (1933)] .
SUMMARY OF AMINATIONS OF HETEROCYCLIC BASES (TABLE) 102
INTRODUCTION
Heterocyclic bases such as pyridine and quinoline and their derivativesreact with metal amides to yield amino derivatives. For example,pyridine is converted to 2-aminopyridine by the action of sodium amide;an intermediate metal derivative is formed, and this is hydrolyzed to thefree amine. (This reaction was discovered by Chichibabin 1 in 1914.)
N a N H 2 -» H2 + || I - ^ ^ || "I + NaOH
1 Chichibabin and Seide, J. Russ. Phys. Chem. Soc, 46, 1216 (1914).
92 AMINATION OP HETEROCYCLIC BASES BY ALKALI AMIDES
It has been suggested 2> 3 ' 4 that the initial step in the reaction is theaddition of the metal amide to the — C H = N — group ; the resultingproduct is then transformed to the metal derivative of the amine, eitherthrough intramolecular rearrangement or through decomposition to the
ammo compound and sodium hydride which interact to give the metalderivative.
O+ NaNH2 -> (| J /NH, - (| ] + NaH -> ( V
Na
This mechanism accounts for the formation of small amounts of 4-
aminopyridine (by 1,4-addition) and for the lack of formation of th e 3-isomer. Evidence of the formation of an unstab le additio n product hasactually been obtained for quinoline.5
THE SCOPE AND LIMITATIONS OF THE REACTION
The study of the amination of molecules containing the — C H = N —group has been confined almost entirely to the heterocyclic compounds.
The few Schiff's bases (which also contain the — C H = N — group) whichhave been aminated in this way have given yields of 20% or less 3 | 4
and the products are more readily synthesized by other methods. Of theheterocyclic bases only pyridine and quinoline and their derivatives givesatisfactory results; amino derivatives of other heterocyclic bases such aspyrazines, pyrimidines, and thiazoles are not obtained readily by thisreaction (see tab le). Th e amino derivatives of pyridines and quino-lines, which are very difficultly available by other me thods, are obtained
directly in yields ranging from 50 to 100% by the use of alkaliamides.
The more common methods of preparing a romatic amines, such as thereduction of nitro compounds, are generally of little value because ofthe difficulty in obtain ing the desired interm ediate s. Fo r example,nitration of pyridine with nitric acid is unsuccessful, and nitration withnitrogen peroxide (NO 2) gives a 10% yield of 3-nitropyridine.6 Othermethods of synthesis of aminopyridines and aminoquinolines are illus-trated in the following scheme.
2 Ziegler and Zeiser, Ber., 63, 1848 (1930).3 Kirsanov and Ivaschenko , Bull. soc. chim., [5] 2, 2109 (193 5).4 K i r s a n o v a n d P o l y a k o v a , Bull. soc. chim., [5] 3, 1600 (193 6).6 Bergs t rom, J. Org. Chew,., 2, 411 (1937).6S h o r u i g i n a n d T o p c h i e v , Ber., 69, 1874 (1936).
The synthesis of 2-aminopyridine from the hydroxy derivative,7
as
indicated above, results in over-all yields of less than 50%, and both this
procedure and that involving the Hofmann degradation8
are long and
tedious. The latter method is useful, however, for the preparation of
3-aminopyridines, which cannot be obtained by direct amination. The
synthesis of 2-aminoquinoline derivatives from the alkali sulfonates is a
convenient method when the corresponding 2-chloro derivatives are
available.9
By contrast with these methods, the direct amination
process is a convenient and economical one.
The ease with which a substituted base undergoes amination is
affected by the nature of the substitutents. When 2-alkylpyridines are
treated with alkali amides in liquid ammonia, the only reaction observed
is the formation of the salt of the enamic modification,10
but in hydro-
carbon solvents at higher temperatures the 2-alkyl-6-aminopyridines are
produced.11
+ MNH2
Low temperature
J 'High temperature
CH2—N H 3
IM
H 2
If both the 2- and 6-positions are occupied by alkyl groups, the amino
group is forced into the 4-position. Thus, 2,6-dimethylpyridine and
sodium amide in boiling xylene form 4-amino-2,6-dimethylpyridine.12
7 Fargher and Furness , J. Chem. Soc, 107,690 (1915); Rath, Ger. pat., 510, 432 (1930).8 Camps, Arch. Pharm., 240, 347 (1902)."Zerweck and K u n z e , U. S. pat., 2,086,691 (1937); Ger. pat., 615,184 (1935).
10 Bergatrom, J. Am. Chem. Soc, 53, 4065 (1931).11 Seide, J. Russ. Phys. Chem. Soc, 50, 534 (1920).12 Chichibabin, J. Buss. Phys. Chem. Soc, 47, 835 (1915); Chichibabin and Vidonova,
94 AMINATION OF HETEROCYCLIC BASES BY ALKALI AMIDES
NHM
+ MNH2 -> _ II ] „ „ + H 2CHsk-TM^CHs CHsk - N T ^ C H
A study h as been made of the effect of various su bstitu en ts on the courseof animation of the quinoline nucleus in liquid ammonia.13 In thissolvent good yields of aminoquinolines are generally obtained, but, if analkyl group is present in either positions 2 or 4, then salt formationoccurs unless more vigorous conditions are employed. Fo r example, 4-methylquinoline is converted to 2-amino-4-methylquinoline when thereaction is carried out in dimethylaniline at 120°,u but none of theproduct is obtained when the reaction is attempted in liquid ammoniaat 2O0.16 It might be expected that other salt-forming groups, such asamido, amino (aromatic), carboxyl, ethynyl, hydroxyl, imino, isonitroso,and active methylene groups, would exert the same effect on amination.This is not always tru e. Th us, a carboxyl group in the 2- or 4-positionactually increases the rate of the reaction and improves the yield.13 2-Aminoquinoline-4-carboxylic acid and 4-aminoquinoline-2-carboxylicacid are obtained in yields of 70 and 8 1 % respectively from the corre-sponding acids, potassium amide, and potassium ni tra te in liquid am -monia; under the same conditions, 2-aminoquinoline is obtained from
quinoline in only 50% yield.13 On the other hand, an amino group inposition 2 of quinoline prevents the amination, as does also a hydroxylgroup in either position 2 or 8.13
W hen a sulfonic acid or methoxyl group is presen t in the 2-position ofquinoline, it is replaced by an amino group by the action of potassiumamide in liquid ammonia.13
JN H 2
JOCH3
Ordinarily the amination of pyridine and its derivatives can becontrolled so th a t only one amino group is introduced. For example,from the reaction of pyridine with sodium amide in dimethylanilineat temperatures below 120°, 2-aminopyridine is obtained in yields ofabout 7 5 % ;1 6 a small am oun t of the 4-isomer may or may not be formed,
13 Bergs t rom, J. Org. Chem., 3, 233 (1938).14 Leffler, unpublished observations.15 B e r g s tr o m , / . Am. Chem. Soc, 53, 3027 (1931).16 Schering A.-G. , Ger . pa t . , 663,891 (1938) .
depending on the quality and quantity of the sodium amideused.14' 17' 18' 19 By increasing the amount of the amide and operatingat tem peratures near 170°, either in dimethylaniline 16 or in the absenceof a solvent,17 2,6-diaminopyridine is obtained as the major product; a
small amount of 4-aminopyridine is formed at the same time, but 2,4-diaminopyridine has no t been isolated. 2,4,6-Triaminopyridine isformed only at high temperature and in the presence of a large excess ofthe metal amide.16
W hen quinoline is trea ted with potassium amide in liquid amm onia, 2-and 4-aminoquinolines are formed in the ra tio 5: 1. Subs titution ofbarium amide for the potassium amide prevents the formation of the4-isomer.6 It is probable that a similar result is not to be expected if the
reaction is carried out in solvents other than liquid ammonia.Secondary reactions, in which the alkali salt of the aminoheterocyclicbase acts as an an ima ting age nt, are sometimes observed. Thus, dipy-ridylamine has been isolated as a by-product in the preparation of 2-aminopyridine.x8
M
'N '
Quinoxal ine i s conver ted to f luorubin by potassium amide . 2 0 '2 1
K
2 K N H 2 ' ^
The only recorded attempt to produce a secondary amine by the reac-
tion of sub stitute d alkali amides with heterocyclic bases is the reac tion ofsodium phenylamide and pyridine; a small amount of 2-phenylamino-pyridine was obtained.1
Another side reaction that takes place in the amination reaction iscoupling. B ipyridyls are always produced in the prep aratio n of amino-pyridines. Thus , 2,2'-bipyridyl, 4,4'-bipyridyl and also dihydro-4,4'-bipyridyl have been isolated as by-products in the amination of pyri-dine.17' 18> 22 These products are often formed in significant quantities
17 Shreve , R ieehe rs , Rubenkoen ig , and Goodman, Ind. Eng. Chem., 32, 173 (1940),18 W i b a u t a n d D i n g e m a n s e , Rec. trav. chim., 42, 240 (1923).19 Chichibabin and Seide , J. Russ. Phys. C hem. Soc, 50, 522 (1920).20 Bergs trom and Ogg, J. Am. Chem. Soc, 53, 245 (1931) .21 Bergs trom and Fernel ius , Chem. Rev. 12, 162 (1933).22 Bergs trom and Fernel ius , Chem. Rev., 12, 156 (1933).
96 AMINATION OF HETEROCYCLIC BASES BY ALKALI AMIDES
when hydrocarbon solvents are employed but their formation is sup-
pressed when the reaction is carried out in dialkylanilines.
The coupled products may undergo animation if the conditions of reac-
tion are sufficiently strenuous. For example, 2,2'-bipyridyl is only
slightly attacked by sodium amide in boiling toluene but undergoesappreciable reaction in boiling xylene.
23The 4,4'-isomer is more readily
aminated.24
Similar coupling products are formed from other heterocyclic bases25
and are often the major products of the reaction between metal amides
and complex heterocyclic substances.26
EXPERIMENTAL CONDITIONS
Direct amination is usually effected by treating the heterocyclic base
with an alkali amide in the presence of a solvent. Potassium nitrate is
often used to accelerate the amination of quinoline and its derivatives
(see p. 100). The exact manner in which it functions is unknown but
appears to be related to the oxidizing capacity of the nitrate ion.
The Alkali Amides. Many patents have been granted and much has
been written about the preparation and properties of various metal
amides, particularly of sodium amide.27
In the selection of the proper
amide for any amination, the character of the compound to be aminated
and the type of solvent to be used must be considered. On a manu-
facturing scale, the fact that sodium amide is much less expensive than
other metal amides may be the determining factor.
Certain precautions must be rigorously observed in the handling of
any metal amide. Most of the knowledge of this class of compound has
been gained from the study of sodium amide, because of its wide use. It
is especially important that the alkali amide be freshly prepared for each
reaction. This is necessary, not only from the standpoint of the repro-ducibility of the experimental results, but also for reasons of safety. It
has been shown a number of times27
that alkali amides react with the
oxygen, carbon dioxide, and water of the air to give dangerously explo-
sive mixtures containing the hydroxides, carbonates, and nitrites. A
patent has been granted to Ziegler28
for the preparation of a homo-
geneous paste by grinding an alkali amide with several times its weight
23 Tjeen Will ink, Jr., and W i b a u t , Rec. Irav. Mm., 54, 281 (1935) .2 4 H o r s t e r s and D o h r n , Ger. pat., 398,204 (1924).26 Chichibabin and Zatzepina , J. Russ. Phys. Chem. Soc, 50, 553 (1920).26 Chichibabin and Shchukina , J. Russ. Phys. Chem. Soc, 62, 1189 (1930).27
For a review, "The C h e m i s t r y of Alka li Am ides , " see (a) Bergs t rom and Ferne l ius ,Chem. Rev., 12, 43 (1933); (6) Bergs t rom and Perne l ius , ibid., 20, 413 (1937).
of an ine rt liquid such as benzene. It is reported th a t such a paste can behandled and transported with safety. Even when stored under a dryhydrocarbon, an alkali amide should be carefully protected from the airand samples which develop a yellow or green or darker color should be
discarded.Sodium amide is employed in most animations except those in which
liquid ammonia is used as the solvent. Because of its insolubility inliquid ammonia, it is inferior to potassium or barium amide, both ofwhich are soluble. W ibau t and Dingem anse 18 found that an especiallypure sodium amide 29 failed to react with pyridine under conditionswhich were very satisfactory when a commercial grade of sodium amidewas used. This and other reports indicate tha t the am ination is influ-
enced by impurities, probably the substances used as catalysts in thepreparation of the amide (p. 99).The Solvent. Various hydrocarbons (such as benzene, toluene , xylene,
cumene, mesitylene, and petroleum fractions), dimethylaniline, diethyl-aniline, and liquid amm onia have been used as solvents. The aminationof pyridine in the absence of a solvent is also successful.17 With quino-lines and isoquinolines good yields are obtained in liquid ammonia solu-tion,6' 13 but, since the reactions must be carried out at room tempera-ture or above, special apparatus must be used to prevent the develop-ment of dangerous pressures due to the hydrogen evolved. The yieldof 2-aminopyridine obtained in reactions employing liquid ammonia as asolvent is less than 30 % . By the use of hydrocarbon solvents such astoluene, yields as high as 80% of this product have been reported; 30
however, it has been the general experience of several workers 14 '1 8 tha tthe pure mate rial is usually obtained in yields of 5 0% or less.
The introduction of dialkylanilines 1 6 '3 1 as solvents has greatly in-creased the practical value of amination of pyridine and its homologs.
For example, 2-aminopyridine is obtained in 70-80% yields from pyri-dine and sodium amide in dimethylaniline at 90-115°, and 2,6-diamino-pyridine in yields of 80 -90% a t 15O-18O0.16 I t is probable tha t the va lueof dimethylaniline and diethylaniline u depends on their solvent actionon sodium amide and on th e sodium amide-pyridine addition compounds.Unfortunately, the investigation of these solvents in animations ofheterocyclic bases other than pyridines has been very limited.
Temperature. An amination should be carried out at the lowest
temperature which will prom ote the desired reaction. In m onoamina-tions this is usually the temperature at which a steady evolution of
29 Ti the r ley , / . Chem. Soc, 65, 504 (1894).30 Vieweg, Ger . pa t . , 476,458 (1929) .31 Oetromis lensky, J. Am. Chem. Soc, 56, 1713 (1934).
Procedure.34 A 500-cc. three-necked flask is equipped with a gas-tight mechanical stirrer, a bubbling tube, and an outlet tube attachedto a wide-bore soda-lime tub e. Approximately 250 cc. of liquid am moniafrom a tank is collected in the flask, and 0.15 g. of ferric nitrate (anhy-
drous or hyd rate d) is added. About 0.5 g. of clean sodium is then added,and after it has dissolved the solution is stirred and dry air is slowlybubbled in until the blue color has disappeared. The oxide so formedacts as a catalyst in the subsequent reaction. The bubbling tube isremoved and 11.5 g. (0.5 atom) of clean sodium is added to the stirredsolution in portions sufficiently small to prevent vigorous reaction . Themixture is stirred for fifteen to twenty minutes after the addition of thesodium is complete.
If the amide is to be used in a solvent other than ammonia, theammonia is allowed to evaporate while the new solvent is slowly addedfrom a dropping funnel. If the dry amide is desired, the produc t may befreed from amm onia by evacua tion at 100°. In an y event, sodiumamide prepared by this method must be used immediately. Becauseof its finely divided condition and the presence of oxides, it rapidlychanges to explosive substances.
Preparat ion o f 2 -Aminopyr id ine
Mi 16> 31
The flask con taining the suspension of sodium amide in liquid ammonia(preceding paragraph) is fitted with a small dropping funnel, and 45 cc.of dry dimethylaniline is added cautiously, the ammonia being allowedto escape throu gh th e soda-lime tube . After all the ammonia has beendriven out, the soda-lime tube is removed and a dry vertical condenser,protected by a calcium chloride tube , is atta ched . Th e mixture is stirredand 31.6 g. (0.4 mole) of dry pyridine is added through the dropping
funnel. Th e funnel is then replaced by a thermom eter which dips intothe reaction mixture. The flask is heated in an oil bath, the tem pera tureof the reaction mixture being maintained at 105-110° until the evolutionof hydrogen has ceased. Hydrogen is produced rapidly a t first, as shownby the continuous stream of bubbles observed when a rubber tube con-nected to the calcium chloride tube is dipped under water. After e ightto ten hours the formation of hydrogen is negligible. Near th e end of thisperiod it may be necessary to discontinue the stirring because of theformation of a solid cake in the reaction flask.
When the reaction is complete, the mixture is cooled and 5% aqueoussodium hydroxide solution (about 75 cc.) is gradually added until thevigorous decomposition has stopped. W ater (about 300 cc.) is then
100 AMINATION OF HETEROCYCLIC BASES BY ALKALI AM IDES
added to complete the hydrolysis of the sodium salt. The mixture isextracted with 75 cc. of petroleum ether (b.p. 30-60°) to remove thedimethylaniline; if necessary more water may be added to assist in theseparation of the layers. The aqueous solution is cooled to 15°, satu -rated with solid sodium hydroxide, and extracted several times with
benzene. The combined benzene extrac ts are dried over anhydroussodium sulfate, and the residue from the distillation of the solvent isdistilled under diminished pressu re. The pro duc t boiling at 117-120°/36mm . weighs 23-28.6 g. (66-7 6% ). The residue consists of 4,4'-bipyridyl,2,2'-dipyridylamine, and other unidentified products."
Preparation of 4-Amino-2-phenylquinoline 35
In leg A of the two-legged tube (Fig. 1) are placed 1.05 g. (0.27 atom)of potassium and 0.02 g. of ferric oxide. Tube C is closed with a stopper,and legs A and B are sealed off as indicated by the dotted lines while a
F I G . 1
stream of ammonia is passed in throug h the stopcock. Through tube C1.83 g. (0.0089 mole) of 2-phenylquinoline and 1.61 g. (0.016 mole) ofpotassium nitrate are introduced into leg B. Tube C is then sealed offas indicated. A t intervals amm onia is condensed in leg A, by cooling A
in a solid carbon dioxide-acetone bath, until the rapid conversion ofpotassium to potassium am ide is complete. H ydrogen is occasionallyvented during this operation. Ammonia is then condensed in the appa r-atus until 15-20 cc. is present and the contents of the tubes are mixedthoroughly by shaking. The appara tus is allowed to stand a t roomtemperature, with the stopcock closed, for four hours.
The ammonia is evaporated from the reaction mixture and the con-ten ts of the tube are rinsed out with ethanol and benzene. W ater is
added to the mixture, and the greater part of the organic solvents isremoved by distillation. The 4-amino-2-phenylquinoline which sep-arates is collected by filtration. The dry, nearly pure produc t weighs
1.96 g. (99.7% ). After rec rystallization from benzene or dilute ethanol,it melts at 164-165°.
Runs of larger size should not be a ttem pted in the appa ratu s described.Apparatus for larger runs has been devised.36
S U M M A R Y O F A M I N A T I O N S O F H E T E R O C Y C L IC B A S E S ( T A B LE )
In the table are summarized the aminations of heterocyclic basesreported prior to Janu ary 1, 1941. I t is possible th at many of the yieldsrecorded in the table, particularly in connection with preparations inwhich hydrocarbon solvents were used, might be improved by carryingout the reactions in dimethylaniline solution.
37 Reference 27a, p. 163.38 K abatchnik and K atze lsohn , Bull. soc. chim., [5] 2, 576 (1935).3 9
B e r g s t r o m , Ann., 515, 34 (1934).40 Chich ibab in and O par ina , J. Russ. Phys. Chem. Soc, 50, 543 (1920).41 Bergs t rom and Rodda, J. Am. Chem. Soc, 62, 3030 (1940).42
M o r g a n a n d W a l l s , J. Chem. Soc, 2 2 2 9 (1 9 32 ) .43
R e f e r e n c e 2 7 6 , p . 4 7 2 .44 Reference 27a, pp. 154-158; 56, p. 463.46 Phili pp, 17. S. pat . , 1,789,022 (1931).46 Plazek , Roczniki Chem., 16, 403 (1936).
47 Seide, J. Russ. Phys. Chem. Soc, 50, 534 (1920).48 Schne iderwir th, U. S. pat . , 2,062,680 (1936).49
61 Mensch ikov , G r igorov i t ch , and Oreehoff, Ber., 67, 289 (1934).62 Chich ibab in and K i r s anov , Ber., 57, 1163 (1924).63 O chia i and K ar i i , / . Pharm. Soc Japan, 59, 18 (1939).64 O gg and Bergs t rom , / . Am. Chem. Soc, 53, 1849 (1931).65 Reference 27a, p. 162.66 Ochiai , J. Pharm. Soc Japan, 58, 1040 (1938).
2-(4'-Hydroxyphenylamino)-8-naphthol-6-sulfonic Acid and 2-(4'-Hydroxy-phenylamino)-naphthalene-6-sulfonic Acid 121
TABLE OF COMPOUNDS PREPARED BY THE BUCHERER REACTION 122
INTRODUCTION
The Bucherer reaction is the reversible conversion of a naphthylamineto a naphthol in the presence of an aqueous sulfite or bisulfite. It hasproved to be of value in the synthesis of naphthalene derivatives, parti-cularly in the manufacture of dye intermediates. In certain instancesit is conveniently used in the preparation of naphthols from naphthyla-
mines; in others it is employed for the reverse transformation, the syn-thesis of naphthylamines from naphthols.
N H 2 O H
(1)
(2)
H 2 0
N H
NaHSO3
N H 3
H 2O
The second reaction has been extended to the synthesis of certain alkyl-and aryl-aminonaphthalenes by the use of alkyl- and aryl-amines and
sodium bisulfite, to the synthesis of naphthylhydrazines by the use ofhydrazine sulfite, and to the synthesis of carbazoles by the use ofphenylhydrazine and bisulfite.
Although Lepetit u 2 was the first to discover the amazingly easytransformation of naphthionic acid to l-naphthol-4-sulfonic acid (equa-tion 1), Hans T. Bucherer3 discovered the reaction independently,recognized its usefulness, and dem onstrated its reversibility. As a con-sequence, the name Bucherer has continued to be associated with thesetransformations.
M E C H A N I S M
Studies of the mechanism of the formation of a naphthylamine from anaphthol, sodium bisulfite, and ammonia 4 indicate that the reactioninvolves addition of the bisulfite to the keto form of the naphthol.
The reaction of the addition product with ammonia is similar to that ofthe sodium bisulfite addition product of formaldehyde, which yieldssodium aminomethanesulfonate.6 Compounds similar to the bisulfite
1 Lepet i t , pl i cachete No. 888, May 16, 1896; Bull, soc. ind. Mvlhouse, 326 (1903).2 Friedl i inder , Ber., 54, 620 (1921).3 Buchere r , J. prakt. Chem., [2] 69, 49 (1904).4 Fuchs and S t ix, Ber., 55, 658 (1922).6 Raschig, Ber., 59, 859 (1926).
addition complex and the amine and corresponds to the reaction typediscussed below.
Reactions of Secondary and Tertiary Amines. N-Mono- and N,N-dialkyl derivatives of naphthylamines can be converted to naphthols bytreatment with aqueous sodium bisulfite.12 These reactions frequentlytake place with greater ease tha n those of primary amines. In the caseof the N-monobenzyl derivatives of l-naphthylamine-4,7- and -4,8-disulfonic acids the yield of benzylamine varies from 60 to 7 7% . In thecase of N -monobenzyl-l-naphthylamine-4-sulfonic acid the yield of
benzylamine is smaller and the time for conversion longer;1S
one wouldanticipate a ready cleavage because of the activating effect of the sul-fonic acid group, but the sparing solubility of the compound hinders thereaction . The disulfonic acid, which is more soluble, reac ts more readily.Apparently the N,N-dibenzyl derivatives of the same compounds arenot cleaved at all under comparable conditions or even by heating in aclosed container at 125-150°.13
Conversion of Hydroxyl Compounds to AminesPreparation of Primary Amines. 1- and 2-Naphthols and their deriva-
tives can be converted into primary amines by treatment with ammoniaand ammonium sulfite or by the action of ammonia on their bisulfiteaddition products.13 The effect of substituents on ease of replacement isthe same as th at mentioned above. Hydroxyquinolines may be ami-nated similarly.14
Inasmuch as 2-nitronaphthalene cannot be obtained by direct nitra-
tion, the Bucherer process for preparing 2-naphthylamine and itsderivatives is of considerable importance. In th e preparatio n of 2-naphthylamine from 2-naphthol, reaction begins around 100° but pro-ceeds much more rapidly in an autoclave at about 150° .3 Yields givenrange from 88% 15 to "practically quantitative."3 Other references to2-naphthylamine will be found under C10H9N in the table of com poundsprepared by the Bucherer reaction on p. 122. An advantage of theBucherer method for the preparation of 2-naphthylamine is that theprocess can be carried o ut a t a temp erature such th at there is practically
12 Buchere r , J. prakt. Chem., [2] 70, 345 (1904).13 Bucherer and Seyde, J. prakt. Chem., [2] 75 , 249 (1907 ).14 Woroshtzow and Kogan , Ber., 65, 142 (1932)." B e z z u b e t z , / . Chem . Ind. (Moscow ), 7, 908 (1930) [C. A., 25, 4545 (1931)].
no formation of 2-2'-dinaphthylamine; 2-naphthylamine is filtered fromthe cooled reaction mixture, and the mother liquor can be used again.3
2,8-Dihydroxynaphthalene-6-sulfonic acid, "G acid," is converted to2-amino-8-naphthol-6-sulfonic acid in 80 % yield.3 Similarly, 2,5-dihy-droxynaphthalene-7-sulfonic acid yields 2-amino-5-naphthol-7-sulfonicacid, and l,5-dihydroxynaphthalene-7-sulfonic acid yields l-amino-5-naphthol-7-sulfonic acid. In these instances the hindering effect of th esulfonic acid group causes the reaction to take place in the other ring.
The behavior of 2-hydroxy-3-naphthoic acid in the Bucherer reactionis worthy of note. This acid undergoes decarboxylation below 100°when heated in the presence of aqueous sodium bisulfite, although theacid itself can be heated in water for eighteen hours at 125° without
change.16
When heated with ammonia and ammonium sulfite at 150-155° for nine hours it is converted into 2-naphthylamine (67%) and2,2'-dinaphthylamine (23% ). Th e bisulfite addition produc t of 2-hydroxy-3-naphthoic acid is related to a /3-keto ac id; the decarboxylationis therefore to be expected. The observed 13 stability of ethyl 2-hydroxy-3-naphthoate toward boiling sodium bisulfite solution is understandable;no loss of carbon dioxide would be expected even though the bisulfiteaddition product of the keto form were produced, for /3-keto esters are
quite stable. Bucherer's experiments do no t prove whether or not thebisulfite addition pro duct of the keto ester is formed, bu t they do demon-strate that replacement of the 2-hydroxyl group by an amino group doesnot occur. The hindering effect of the carbethoxy group is thus to becompared with the similar influence of the sulfonic acid group.
8-Hydroxyquinoline is converted "almost quantitatively" into 8-aminoquinoline by heating with ammonia and ammonium sulfite in aclosed vessel at 150-160° for six to seven hours.14 6-Hydroxyquinolineand 8-hydroxyquinoline-5-sulfonic acid are similarly converted to thecorresponding aminoquinolines.
Preparation of Secondary Amines. Conversions of naphthols toN-alkyl- or N,N-dialkyl-aminonaphthalenes require more vigorous con-ditions than are necessary for the production of primary amines bymeans of amm onia and ammonium sulfite. Fo r example, amination ofl-naphthol-4-sulfonic acid takes place smoothly at 90°, bu t the subs titu-tion of methylamine for ammonia necessitates carrying out the processat 150° in an autoc lave.12 I t is possible in such instances to he at together
one mole of naphthol, one mole of alkylamine sulfite, and one mole ofalkylamine in an autoclave at 125-150° until reaction is complete (testfor residual naphtholsulfonic acid), or to prepare the addition productfrom the naphthol and excess sodium bisulfite a nd , after acidification
" B u c h e r e r , Z. Farb. Text. Chem., 1, 477 (1903).
1-naphthols und er special conditions. Th e sa lt of an arylamine will reactat a temperature between 100° and 200° with a molecular equivalent ofthe isolated bisulfite addition product of a 1-naphthol; the product, anarylaminonaphthalene, is formed in good yield. Th e reaction may also
be carried out in aqueous solution; the bisulfite addition product is pre-pared in th e usual way in aqueous medium, excess bisulfite is neutralizedor removed by acid, the requisite amount of amine hydrochloride isadded, and the m ixture is heated in an autoclave. I t has been sug-gested 22a th a t th e intermed iate involved is a salt of the addition productand the amine.
HO SO3NH3C6H5
In many cases it is possible to isolate such saltlike addition products,which, on heating, yield the expected amine, sulfur dioxide, andwater.
The usefulness of any particular arylamine in the Bucherer process isdetermined not only by its own tendency to enter the reaction but also
by the reactivity of the bisulfite addition compound of the naphtholwith which it is being condensed. p-Toluidine in the presence of bisulfitedoes not react rapidly with 2-naphthol-6-sulfonic acid; however, theyield is practically quantitative when the isomeric 2,8-acid is used.11
Likewise, benzidine, which reacts with /3-naphthols only with extremedifficulty, reacts much more readily with 2-hydroxy-3-naphthoic acidand with 2,8-dihydroxy-3-carboxynaphthalene-6-sulfonic acid, both ofwhich are notable for the ease with which they undergo animation bythe Bucherer process.
I t is possible to use relatively complex amines in this process. Thusp-rosaniline reacts readily in the presence of sodium bisulfite with 2-naphthol-6-sulfonic acid to form substituted rosanilines.11 '21 The exactconstitution of the reaction products has not been established.
The effects of substituent sulfonic acid groups in the naphthol nucleusupon the ease of reaction w ith an a rylamine are identical w ith those men-tioned earlier (p. 108).
Preparation of Secondary A mines from P rimary A mines
2-Naphthylamines can be substituted for 2-naphthols in any of thereactions described on pp . 110-113; 1-naphthylamines can be substitutedfor 1-naphthols only in those processes involving alkylamination or dial-
kylamination. The 2-naphthylamines react more easily than the corre-sponding naphthols.11 Thus l-methylamino-7-naphthol-4-sulfonic acidcan be prepared from l-amino-7-naphthol-4-sulfonic acid by treatmentwith sodium bisulfite and methylamine.23 Likewise, 2-(4'-hydroxy-phenylamino)-naphthalene can be prepared from 2-naphthylamine and
p-aminophenol,24 and 2-phenylaminonaphthalene-6-sulfonic acid can beprepared from 2-aminonaphthalene-6-sulfonic acid.11 '21
2-Amino-8-naphthol-6-sulfonic acid, 2-aminonaphthalene-6,8-disul-fonic acid, and 2-amino-5-naphthol-7-sulforiic acid all react in the pres-ence of sodium bisulfite with p-rosaniline to form substituted rosani-lines.11
The process discussed above may be summarized as follows.
ArNH2 ^ 1 ^ ' > ArNHAr' or ArNR2 (R = alkyl or hydrogen)or
EjNH
I t should be noted t h a t no useful reversal of the B ucherer reaction takesplace when N-aryl-2-naphthylamines are heated with sodium bisulfitesolution.
Reactions Involving Hydrazines
Arylhydrazines are formed in the reaction of hydrazine sulfite andhydrazine with naphthols.2 5 '2 6 '2 7 Thus hydrazines can be preparedfrom 1- and 2-naphtho l, and 2,7-dihydroxynaphthalene yields 7-hydroxy-2-naphthylhydrazine (82%) with a very small amount of dihydrazineunder the conditions used.266 Both hydroxyl groups of 2,3-dihydroxy-naphthalene can be replaced by hydrazine residues;26a the yield of crudeproduct is abou t 57% . Similarly resorcinol yields m-phenylenedihy-drazine.26c The latter compound cannot be isolated as such but can be
obtained as its reaction product with benzaldehyde (yield 50% ). Py ro-catechol, hydroquinone, 3,4-diaminotoluene, and salicylic acid do notundergo the reaction.260 It is to be noted that both 1- and 2-naphtholsundergo this reaction, and that more than one hydrazine residue can beintroduced readily.
W hen phenylhydrazine, sodium bisulfite, and a naphthol (or naph thyl-amine) are heated together a rather complicated series of reactions takes
23 Ger. pat. , 676,856 [C. A., 33, 7319 (1939)].24 Bri t , pa t . , 479,447 [C. A., 32, 5003 (1938)].26 F r a n z e n , Habilitationsschrift, Heidelberg (1904) .26 (a ) F r a n z e n , J. prakt. Chem., 7 6 , 2 0 5 ( 1 9 0 7 ) ; (6 ) 7 8 , 1 4 3 ( 1 9 0 8 ) ; (c ) 7 8 , 1 5 7 ( 1 9 0 8 ) ; (d )
Ber., 3 8 , 2 6 6 ( 1 9 0 5 ) .27 Buchere r and Schmidt , J. prakt. Chem., [2] 79, 369 (1909).
When a naphthylhydrazine reacts with aqueous bisulfite the first reac-tion apparently is removal of the hydraz ine residue with the formation ofthe bisulfite addition compound of the parent naphthol27 which thencombines with unchanged naphthylhydrazine to form a compoundsimilar to I I I . If 1-naphthylhydrazine is used, this prod uct is apparently
stable 27 but is converted by treatment with hot mineral acids into1,2,7,8-dibenzocarbazole.
2-Naphthylhydrazine behaves somewhat differently in that the principalproduc ts are 3,4,5,6-dibenzocarbazole an d a compound of typ e V. Thissubstance loses its sulfonic acid group readily to form the correspondingcarbazole. Experiments 27 have shown that it is possible to prepare thetype V compound from 2-hydroxynaphthoic acid directly by treatmentwith 2-naph thylhydrazine in sodium bisulfite solution. 1-Naphthyl-hydrazine also condenses easily with 2-hydroxy-3-naphthoic acid; thecondensation product (type V) is formed in good yield and is readilytransformed into 1,2,5,6-dibenzocarbazole by the action of mineral acid.27
l-Naphthylamine-4-sulfonic acid and the corresponding naphtholsul-fonic acid reac t readily with phenylhydrazine in the presence of bisulfite.Apparently the reaction proceeds to a type III compound; evidence forthe structure of this compound is its conversion by oxidation in alkalinesolution into l-phenylazonaphthalene-4-sulfonic acid. Trea tmen t withhot concentrated hydrochloric acid converts the hydrazo compound inpart into 1,2-benzocarbazole.30
bisulfite and phenylhydrazine, but here again the reactions are complex
and the nature of the products is obscure.
In general the reaction of a hydrazine with a naphthol (or naphthyla-
mine) in the presence of bisulfite takes place more readily than the corre-sponding reaction involving an amine and a naphthol (or naphthyla-
mine). In this connection it is interesting to note that "R acid" (2-naph-
thol-3,6-disulfonic acid), which does not react with amines30
in the pres-
ence of bisulfite because of the hindering effect of the 3-sulfonic acid,
condenses readily with phenylhydrazine under similar conditions.
The Use of Bisulfite Addition Products in the
Preparation of Azo Compounds
Bisulfite addition products obtained from dihydroxy- or diamino-
naphthalenes can be employed in the preparation of azo dyes. Those
compounds containing a free amino group in the aromatic ring of the
addition complex can be converted into diazonium salts which couple
in the usual way. After the coupling the hydroxyl group can be regen-
erated by treatment with alkali or the addition product can be converted
into an amine. Obviously a bisulfite addition product can be coupled
with any diazonium salt provided that there is an activating group
(hydroxyl or amino) in the aromatic ring; coupling must take place in the
31 Konig and Haller, J. prakt. Chem., 101, 38 (1920).32 Bucherer and Z i m m e r m a n n , J. prakt. Chem., 103, 277 (1921).33 Bucherer and W a h l , J. prakt. Chem., 103, 253 (1921).
ring conta ining the f ree a romat ica l ly bound amino or hydroxyl group(directed coup ling). F or exam ple, diazon ium salts m igh t couple withl,8-dihydroxynaphthalene-4-sulfonic acid in ei ther the 2- or the 7-posi t ion. A ctua lly th e f irst mo le of diazo com pou nd couples almo stexclusively in the 2-position. When sodium bisulfite reacts with 1,8-
dihydroxynaphthalene-4-sulfonic acid, the r ing holding the sulfonic acidgro up is involved (ac t ivat in g influence of th e 4-sulfonic ac id) . T he reac-t ion product couples with a diazo compound to form a substance of thefollowing structure.
HO SO 3NaHO
SO 3Na
When this compound is warmed wi th a lka l i , i t i s reconver ted to a dihy-
droxynaphthalenesulfonic ac id.
OH OH
SO 3Na
Th us a di rec ted coupling has been accomplished. Th is azo compoundcould be again coupled with a different diazonium salt with formation ofa bis-a,zo d y e . 1 2
OH OH
SO 3Na
Sui tably loca ted amino groups can be diazot ized more c leanly in addi-t ion compounds because the hydroxyl-conta ining r ings are considerablyless react ive to w ard chance excess .of ni t r ou s acid th an those of th e p ar en taminonaphthols . 12 , Azo dyes related to a naphthol can be made suffi-cient ly water-soluble as addit ion compounds with bisulfi te , even thoughthey contain originally no sulfonic acid group, so that they can beapplied to th e f iber. T he co mbined bisulfite can be rem oved when thedye is on the fiber.12
SELECTION OF EXPERIMENTAL CONDITIONS
Ex per im enta l condi t ions necessar ily va ry over a wide ran ge . R eac -
t ion may take place a t a tempera ture as low as 90° , or i t may proceed
satisfactorily only in the neighborhood of 150°. If some of the reac tan tsare only sparingly soluble, intimate mixing of the phases is essential tothe success of the process. Aminations involving the use of amm onia andammonium sulfite are ordinarily conducted in closed vessels at tempera-
tures from 100-150°. Arylaminations will proceed slowly under refluxbut take place more rapidly in an autoclave at about 150°. Generaldirections for preparation of N-aryl-2-naphthylamine derivatives aregiven by Bucherer.21 The requisite 2-naphthol- or naphthylamine-sul-fonic acid is dissolved in a minimum of boiling water and then graduallymixed a t 80-90° with a warm solution of sodium bisulfite. If a sulfonicacid should be salted out by the mixing, the salt is brought back intosolution by warming on the water ba th. The arom atic amine is next
added either as such or as a mixture of the hydrochloride a nd an equiva-lent of aqueous sodium hydroxide.
The mixture is then heated under reflux until a titration with p-nitro-benzenediazonium chloride shows no more decrease in original naphtholand no increase in product. To carry out the test for complete reactiona small test portion of the mixture is made distinctly alkaline to phenol-phthalein and freed of the excess of the amine used as aminating agentby steam distillation. The mixture is then m ade acid to Congo red withsulfuric acid and boiled until all the sulfur dioxide has been expelled.Diazonium salt solution is then added dropwise from a calibrated pipetuntil a drop of the mixture on filter paper shows no color in the run-outeither with the diazonium salt solution or with Schaeffer's acid (2-hy-droxynaphthalene-6-sulfonic acid). As soon as this poin t is reached,the precipitated dye is filtered from the main test portion and washedwith a little saturated sodium chloride solution, the washings beingadded to the test portion . Sodium acetate is then added to the testsolution, and the solution is again titra ted with the same diazonium salt
solution. The ratio of the volume of diazonium sa lt solution employed inthe coupling in acid solution and the volume used in the coupling insodium acetate solution gives the proportion between the newly formedamine and the remaining naphthol.
If specific directions for the preparation of the desired compound arenot available, orientation experiments controlled as above, using rela-tively small quantities of material, are necessary in order to determineoptimum conditions of time, temperature, and proportions of reactants.
The following examples illustrate both simple amination and aryl-amination.
One hundred forty-four grams (1 mole) of 2-naphthol is placed in asuitable pressure vessel together with a solution of ammonium sulfiteprepared by passing sulfur dioxide into 400 cc. of cooled, concentratedamm onia (sp. gr. 0.90) un til 100 g. of gas has been absorbed. An appar-atus such as that employed for high-pressure hydrogenation will serve;it is essential tha t provision be made for shaking or stirring the reactionmixture. The autoclave is closed and heated a t 150° with continualshaking or stirring for eight hours and is then allowed to cool with shak-ing.
The reaction mixture is removed from the autoclave, which is rinsed
with abo ut 500 cc. of water. The produc t is filtered on a Biichner funnel,and the crude m ateria l is dissolved in a bo iling m ixture of 150 cc. of con-cen trated hydrochloric acid and 400 cc. of w ater and then diluted with 11.of water. Ten grams of N orit is added, and the mixture is boiled forfive minutes. After filtration (heated funnel) from a ny undissolveddinaphthylamine, the product is precipitated by pouring the hot solu-tion with stirring into a solution of 120 g. of sodium hydroxide in 500 cc.of water. The resulting slurry , which should be alkaline to phenol-
phtha lein, is cooled w ith stirring to 20°, filtered, and washed with 2 1. ofcold water.
The product is dried to constan t weight at 50°. I t is a light tanpowder and weighs 135-137 g. (94-96% of the amount theoreticallypossible). The produc t melts at 111-112°.
Preparation of 7-Methyl-l-naphthylamine34> 36
A m ixture of 50 g. of 7-m ethy l-l-naphtho l, 150 cc. of w ater, 75 cc. offreshly prepared ammonium sulfite solution (prepared from aqueousamm onia [sp.gr. 0.90] and sulfur dioxide), and 75 cc. of aqueous am moniasolution (sp.gr.0.90) is prepared in a 35-mm. Pyrex tube approximately400 mm . in leng th. The tube is carefully sealed and heated in an elec-trically hea ted furnace constructed of iron pipe and attac hed to a shakingmachine.
The furnace is heated to a temperature of 160-165° as recorded by athermometer under the resistance wire, and the furnace and its contentsare shaken at this tem pera ture for thir ty to thirty-five hours. The shaker
is then stopped and the furnace allowed to cool to room temperaturebefore it is opened.
34 Ruzicka and Morgel i , Helv. Chim. Ada, 19, 377 (1936).36 Howard , Ph .D. thes i s , Unive rs i ty o f Maryland , p . 25 (1938) .
The contents of the tube are extracted with three 250-cc. portions ofether; the extracts are combined and extracted with 10% hydrochloricacid until a small test portion of the last extract gives no precipitate ofamine when made alkaline with 10% aqueous sodium hydroxide. The
extracts are made alkaline with 10% aqueous sodium hydroxide, where-upon the amine precipitates and is filtered and dried in vacuum. Th eyield is 40-45 g. (80-90% ). Th e combined yields of several such runsare distilled from a sausage flask under 3 mm . pressure. The bulk of thematerial boils at 139-140°/3 mm .; the product melts at 5 8-59°. Ifdesired the amine can be crystallized from petroleum ether, from whichit separates in the form of fine needles.
Preparation of 2-^-Tolylamino-5-hydroxynaphthalene-7-sulfonic Acid
A mixture of 216 g. (2 moles) of distilled p-to luidine, 215 g. (0.9 mole)of 2-amino-5-hydroxynaphthalene-7-sulfonic acid ("J ac id"), 167 g. ofsodium bisulfite, and 500 cc. of water, in a 3-1. three-necked round-bot-tomed flask provided with a reflux condenser and a mechanical stirrer,is heated under reflux with stirring for thir ty hours. Sodium carbo nateis then added until the mixture is alkaline and the excess p-toluidine isremoved by steam distillation. The residual solution is cooled in arefrigerator until crystallization is complete, and the crystals are suckeddry on a Biichner funnel and washed with about 50 cc. of cold saturatedsodium chloride solution. The product is dissolved in about 700 cc. ofhot water to which enough hydrochloric acid is added to make themixture acid to Congo red. The mixture is allowed to stand in a refriger-ator until crystallization is complete; the crystalline acid is filtered andwashed on the filter with a little ice-cold hydrochloric acid and thentwice with small por tions of cold wa ter. The 2-p-tolylamino-5-hydroxy-7-sulfonic acid is dried at 100°; it weighs about 185 g. (65%).
Preparation of 2-(4'-Hydroxyphenylamino)-8-naphthol-6-sulfonic Acidand 2-(4'-Hydroxyphenylamino)-naphthalene-6-sulfonic Acid21
A mixture of 25 g. of "y acid" (2-amino-8-hydroxynaphthalene-6-sul-fonic acid), 50 cc. of water, 250 g. of sodium bisulfite solution (33%),20 g. of p-aminophenol hydrochloride, and 16 g. of sodium hydroxide isboiled under reflux for twenty hours. When the mix ture has cooled toroom temperature, it is acidified to Congo paper and the crude product
is filtered on a B iichner funnel. It is purified by solution in alkali andreprecipitation by acid. The pure produc t weighs abo ut 13 g. (37.5%).
Substitution of 25 g. of "Schaeffer's ac id" (2-hydroxynaphthalene-6-sulfonic acid) for "y acid " above results in a yield of 20 g. (61%) of2-(4'-hydroxyphenylamino)-naphthalene-6-sulfonic acid.
36Le v i , (Horn. chim. ind. applicata, 3 , 9 7 (1921) .
37B r i t , p a t . , 184 ,284 [C . A . , 1 7 , 1 1 0 (1923)] .
38V. S . pa t . , 1,880,701 [C . A., 27, 515 (1933)].
39 Ger. pat. , 109,102 [Frdl. 5, 164 (1897-1900)].40 Bucherer and TJhlmann, J. prakt. Chem., [2] 80, 201 (1909).41 Ger. pat . , 451,980 [C. A., 22, 4130 (1928)].
42 Har tung , Minn ick , and Koehle r , J. Am . C hem. Soc, 63, 507 (1941).43 Ger. pat . , 643,221 [C. A., 3 1, 4342 (19 37)]; B rit , p at. , 451,348 [C. A., 31,113 (1937)] .44 Fr. pa t . , 750,243 [C. A., 28, 779 (1934)].46
F r . p a t . , 8 0 7 , 7 6 5 [ C . A., 3 1 , 5 8 1 3 ( 1 9 3 7 ) ] ; c f . F r . p a t . , 6 4 5 , 1 5 0 , a n d G e r . p a t . , 6 4 2 , 5 4 9 .46
F r . p a t . , 7 8 9 , 5 8 9 [ C . A., 3 0 , 2019 ( 1 9 3 6 ) ] .47
B r i t , p a t . , 4 3 7 , 7 9 8 [ C . A., 3 0 , 2 2 0 3 ( 1 9 3 6 ) ] .48
G e r . p a t . , 1 1 7 , 4 7 1 [ F r d l . , 6, 19 0 ( 1 9 0 0 - 0 2 ) ] .49 Ger. pat. , 120,016 [Chem. Zentr., I, 1074 (1901)].60 Ger. pat. , 121,683 [Frdl., 6, 192 (1900-02)].61
G e r . p a t . , 1 2 6 , 1 3 6 [ F r d l , 6, 18 9 ( 1 9 0 0 - 0 2 ) ] .62
G e r . p a t . , 1 3 2 , 4 3 1 [ F r d l . , 6, 19 3 ( 1 9 0 0 - 0 2 ) ] .
63 Ger. pat . , 134,401 [Frdl., 6, 186 (1900-02)]." Ger. pat. , 254,510 [C. A., 7, 1617 (1913)].65 Brit , pat. , 11,427 [C. A., 6, 3023 (1912)].66 Ger. pat . 442,310 [Frdl., 15, 511 (1927-29)].67 U . S . p a t . 1,727,506 [cf. G e r . p a t . , 4 68 , 811 ( C . A . , 2 3 , 2723 [1929]) ; Frdl, 1 6 , 5 1 0
(192 7-2 9) ] .6 8
G e r . p a t . , 122 ,570 [Frdl, 6 , 1 9 4 (190 0 -0 2 ) ] .
General Reference, Bucherer Reaction
BtrcHERER, "Lehrbuch der Farbenchemie," 2nd ed., p. 200, Otto Spamer, Leipzig,1914.
Diaryl ketones having a methyl or methylene substituent adjacent tothe carbonyl group often suffer cyclodehydration when submitted topyrolysis and afford a certain amount of the corresponding anthracenederivative. Although an early instance of the production of a hydro-
carbon by this process was reported briefly by Behr and van Dorp,1 the1 Behr and van Dorp, Ber., 6, 753 (1873); 7, 16 (1874).
reaction is generally accredited to Elbs,2"8 for this investigator was thefirst to explore the gen erality and syn thetic uses of the reaction . Elbsand his co-workers studied the pyrolysis of various polyalkylbenzo-phenones and, finding that some of these substances failed to condensewhile others afforded anthracene homologs in no better than 20-25%yield (Table I), were inclined to discount the value of the method, par-ticularly where the hydrocarbon in question can be obtained by thephthalic anhydride synthesis of the anthraquinone, followed by reduc-tion. From the accumulated da ta now available, it has become apparen tthat, although the Elbs condensation in general is subject to many limi-tations and shortcomings, there are instances in which the reaction pro-ceeds smoothly and affords the best known means of obtaining impor-ta n t hydrocarbons. The reaction also has found significant use in th e
synthetic preparation of hydrocarbons not available by other knownmethods. A low yield in the pyrolysis is often offset by th e ready avail-ability of the required ketone.
The reaction usually is carried out by heating the ketone without cata -lyst or solvent at the reflux temperature, or at a temperature in therange 400-450°, un til water is no longer evolved. A t the high tempera-tur e required to effect ring closure considerable carbonization may occurand much ma terial ma y be lost as the result of cleavage of the ketone by
the water liberated, elimination or degradation of alkyl substituents,and molecular rearrangements. The main hydrocarbon reaction productmay not be that normally expected on the basis of the structure of thestarting material, and the product is frequently, if not always, accom-panied by related hydrocarbon s. W ith the exception of a few particu-larly favorable applications of the reaction, a product of the Elbs con-densation usually requires extensive purification, and the probablestructure as inferred from analogy should be investigated by independ-ent methods. Th e total weight of the crude hydrocarbon fractionobtained from the pyrolysis mixture by distillation and initial crystal-lization usually does not provide a reliable index of the true yield unlessthe melting point can be shown to be reasonably close to that of a single,fully purified product.
The mechanism of the condensation is no t known. Cook 9 suggested
2 Elbs and Laraen , Ber., 17, 2847 (1884).3 Claus and E lbs , Ber., 18, 1797 (1885).4 Elbs and Olberg, Ber., 19, 408 (1886).6 Elbs , J. prakt. Chem.6 E l b s , / . prakt. Chem.,7 Elbs , J. prakt. Chem.,8 Elbs , J. prakt. Chem.
that the ketone may undergo tautomerism to an enolic form having adiene system to which intramolecular addition of the attached arylgroup ma y occur, giving the dihydroanthran ol. Fieser and D ietz 10
H
suggested that the same intermediate, which at the pyrolysis tempera-ture certainly would undergo rapid dehydration to the hydrocarbon,
may result alte rna tely from a forced 1,4-addition of the methyl substitu-ent to the conjugated system comprising the carbonyl group and the arylnucleus. The re is no evidence bearing on either hypothesis, and a sug-gested analogy n to the formation of anthracene derivatives by thecyclization of o-benzylbenzaldehyden and o-benzyl diaryl ketones 12
does not appear applicable because these cyclizations proceed under theinfluence of an acid catalyst and at a low temperature and hence underconditions wholly unlike those required for the non-catalytic high-temperature pyrolysis.
E X A M P L E S O F T H E R E A C T I O N
Synthesis of Anthracene Homologs. Observations concerning thepyrolysis of mono-, di-, tri-, tetra-, and penta-methyl derivatives ofbenzophenone are included in Table I. In most instances the materialpyrolyzed was the total distilled product of the condensation of a hydro-carbon with an acid chloride or with phosgene, and the published data
on the pyrolysis tem pera tures and th e yields are no t very specific. Seerand co-workers13t u followed Elbs' practice of refluxing the ketonegently for a prolonged period bu t obtained only very low yields. M or-gan and Coulson 15> 16 found it expedient to shorten the time of reactionand to remove the hydrocarbon formed from time to time in order toprotect it from destruction. Although this technique appa rently repre-sented a marked improvement, the yields reported refer merely to mate-rials of unspecified purity and consequently are ambiguous. The d a ta
10
Fieser and Dietz , Ber., 62, 1827 (1929).11 E . B e r g m a n n , J. Org. Chem., 4, 1 (1939).15 Bradsher , J. Am. Chem. Soc, 62, 486, 1077 (1940).13 Seer and S tanka , Monatsh., 32, 143 (1911).14 Seer and Ehrenzweig, Monatsh., 33 , 33 (1912).16 Morgan and Coul son , J. Chem. Soc, 2203 (1929).16 Morgan and Coul son , J. Chem. Soc, 2S51 (1929).
for the series of homologs hardly warrant any general conclusion exceptthat some o-methylbenzophenones afford anthracene derivatives in lowyields while others, under the conditions investigated, gave only nega-tive results. A lthough 2,4-dimethylbenzophenone apparen tly failedto undergo cyclization under conditions adequate for the formation of acertain amount of 2-methylanthracene from the isomeric 2,5rdimethyl-benzophenone, a corresponding difference was not noted with the 2,4,4'-and 2,5,4'-trimethyl compounds.
Table I includes an instance of the elimination of an isopropyl groupin the course of the pyrolysis and an example of the formation of ananthron e derivative along with the corresponding hydrocarbon . Theanthrone may possibly arise by the dehydrogenation of the postulatedintermediate dihydroanthrano l. The last entry of the table shows th at
cyclization can occur in b oth of two possible directions, the one involvingan ortho methyl group and the other an ortho methylene substituent.
1,2,5,6-Dibenzanthracene Series (Table II). The Elbs reaction affordsby far the most rapid and economical method known for the synthesis of1,2,5,6-dibenzanthracene (I II ), a hydroc arbon widely used for theexperimental production of cancer in animals. A num ber of points ofgeneral interest have been discovered in the extensive studies of thisexample of the reaction. One is the occurrence of a rearrangem ent in th e
pyrolysis of a,a'-d ina phth yl ketones. Although 2'-methyl-2,l'-dinap h-thyl ketone (I) and 2-methyl-l,l '-dinaphthyl ketone (II) would beexpected to yield isomeric hydrocarbons,19 ' 10 Cook24 showed that theyboth afford I I I as the chief product. Cook suggested th a t the keton e
(II) which reacts abnormally may undergo rearrangement to the isomerI at the pyrolysis temperature, and indeed it has been shown10 that theabnormal pyrolysis of I I proceeds far more slowly th an the normal con-densation of I. T he several examples listed in the second section ofTable II dem onstrate the g enerality of the rearrangem ent.
The hydrocarbon prepared from either ketone retains a bright yellowcolor not altered by distillation or repeated crystallization,19 ' 10 butCook 9 found that pure 1,2,5,6-dibenzanthracene is colorless and that thecolor is due to the presence of a persistent chrysogen which Wintersteinand Schon 26 later identified as the isomeric 1,2,6,7-dibenzanthracene(IV). The chrysogen, which evidently arises from the ketone I by con-densation of the methyl group into the /3-position of the second naphthylnucleus, con stitutes about 10% of the hydrocarbon m ixture. Variousmethods have been reported for the removal of the yellow contaminantbased upon its greater affinity for chemical reagents or adsorbents.These include (a) preferential sulfonation of the mixture in xylene solu-tion 9> 24 (extensive losses), (b ) chromatographic adsorption 26 (10-20%recovery), (c) treatment with maleic anhydride in boiling xylene,29 andtreatment with lead tetraacetate in acetic acid solution 80 (70-83%recovery).
In this series there are several instances of the loss of methyl groupsin the course of the E lbs reaction. The pyrolysis of the ethyl-substituted
ketone V affords 1,2,5,6-dibenzanthracene in relatively high yield, the
methyl group which normally would appear at a reactive meso positionof the product being completely eliminated. l,l'-D in ap hthy l ketoneshaving methyl groups at the 4- or 4'-positions (VI) are prone to losethese substituents, and there appears to be a general tendency for theelimination of substituents from a-positions in the carbonyl-containingrings of the dinaphthyl ketones.31 Another change observed in thecourse of a pyrolysis is dehydro genation. The 5,6,7,8-tetrahydride of
29Cook, J. Chem. Soc, 3273 (1931); Cook, Hieger, Kennaway, and Mayneord, Proc.
Roy. Soc, Bill, 46 9 (1932).30
Fieser and Hershberg, J. Am. Chem. Soc, 60, 1893 (1938).31
Fieser and Peters, J. Am. Chem. Soc, 54 , 3742 (1932).
the ketone I affords the fully aromatized hydrocarbon III when heated at430-450°,21 and other instances studied by Clar 26 are listed in the thirdsection of Table II, which includes data on the synthesis of higher poly-nuclear hydrocarbons by elaboration of the general scheme already illus-
trated.1,2-Benzanthracene Series (Table III). The most noteworthy feature
of the data on the conversion of methylated benzoylnaphthalenes into1,2-benzanthracene derivatives is the striking contrast in the behaviorof the 2-methyl and 2'-methyl compounds, VII and VIII. The firstketone on being pyrolyzed for three hours affords 1,2-benzanthracene inas high as 6 1 % yield, whereas the isomer V II I loses water only in thecourse of twenty-six hours and gives the same hydrocarbon in 10% yield.
The difference is understandable in terms of the mechanism suggestedby Fieser and Dietz, for in the favorable case (VII) the methyl group
condenses into a naphthalene nucleus, while in VIII this group mustsubstitute into a less reactive benzene nucleus. Cook's postulate th a tthe Elbs reaction is dependent upon a process of enolization does notexplain the observed difference, since VIII should be more prone toenolize than Vtl .
The favorable feature of structure encountered in o-tolyl a-naphthylketone (VII) is met with also in the series of o-methyl 2,l'-dinaphthylketones listed in T able I I , for example ketone I (p. 134), and in this series
the yields again are on th e whole definitely be tter tha n with the methy-lated benzophenones (Table I ). Ketones of the typ e of l-benzoyl-2-methylnaphthalene (VIII) thus fall into the same unfavorable class asthe benzophenone derivatives, and it will be seen from the data of TableIII, which refer almost entirely to ketones of the type of VIII, that theyields are regularly poor. Unfortunately polysubstituted ketones hav-ing the o-methyl group in the naphthalene nucleus are more readilyaccessible than the more favorably constituted isomers and have beenused exclusively for the synthesis of 1,2-benzanthracene homologs. In
this series aroyl rearrangements occur in several instances, and there areexamples of the loss and degradation of alkyl groups. M ethyl sub stitu -ents have been found to be eliminated from positions 5 and 8 of theresulting 1,2-benzanthracene, but there are examples of the retention ofmethyl at these same positions as well as at positions 4, 6, 7, 2', and 3'.
a In this series of experiments Cook 3i heated the fceton e until water was no longer evolved an d boiling ceased; usually the pyrolysis waa conducted a t 440-450°for two hours, or at 410-420° for four hours or sometimes longer. The yield of crude material w as 20-2 5% ; the yield of produc t purified by distillation, crystalliza-tion, and (usually) through the picrate was 5-10%.
32Fieser and Hershberg, J. Am. Chem. Soc., 59 , 2502 (1937).
33Fieser and Cason, J. Am. Chem. Soc, 61, 1740 (1939).
34Cook, J. Chem. Soc, 456 (1932).
35Geyer an d Zuffanti, J. Am. Chem. Soc, 57, 1787 (1935).
36Cook, J. Chem. Soc, 2529 (1931).
37Fieser an d Hershberg, / . Am . Chem. Soc, 62, 1640 (1940).
An isopropyl group was in part retained at position 6 and in part de-graded to furnish a 6-methyl group.
Synthesis of Cholanthrenes (Table IV). The synthesis of a hydro-carbon of the cho lanthrene series by the Elbs reaction is illustrated by theformulas shown in Table IV . Although singly linked alkyl substitu en tsare often eliminated at the pyrolysis temperature from the 5- and 10-positions of the 1,2-benzanthracene nucleus, the ace- or dimethylenebridge atta ched at these points appears to be more stable, for no instanceof a rup ture of these linkages is on record. Furthe rmore, bo th the4-hydrindyl-a- and /3-naphthyl ketones present structures particularlyfavorable for the Elbs condensation. As with o-tolyl a-naph thyl ketone(VII), ring closure involves a substitution into a reactive naphthalenenucleus, and ano ther auspicious circumstance is th at the ortho methylene
group undoubtedly surpasses a corresponding ortho methyl group inreactivity. I t is thus understandable tha t the reaction resulting in theformation of a cholanthrene takes place with particular rapidity andprobably at a slightly lower critical pyrolysis temperature than in anyother known example.
The generally favorable situation is reflected in the fact that theimportant carcinogens cholanthrene and 20-rnethylcholanthrene can beprepared in quantity in a thoroughly purified condition in 40-50% yield
and that the yields on the whole are definitely better than in any of theother series studied. M ethyl groups at the 1-, 6-, and 7-positions of thehydrindene nucleus pass through the pyrolysis unscathed, and the sameis true of a methyl situated in the naphthalene nucleus at the 4'(a)-posi-tion, whereas in the synthesis of 1,2,5,6-dibenzanthracenes such a groupinvariably is eliminated. The only troublesome instan ce of me thylelimination encountered is in the pyrolysis of 2,7-dimethyl-4-(a-naph-thoyl)-hydrindene, when the alkyl group which should appear at thehighly reactive meso-methylene group (C15) was retained only in part,and was in part lost. Syn theses of the 20-ethyl, 20-isopropyl, and 20-<-butyl derivatives have been accomplished successfully, if in low yield.It has even been possible, at least in some instances, to carry methoxylgroups and halogen atoms through the synthesis. A methoxyl substitu-ent located at either the 6'- or 7'-position of the ketone is retainedadmirably, and the corresponding 3- and 2-methoxycholanthrenes areobtainable in excellent yield. A methoxyl at the vulnerable 4'(ex-position, however, is completely lost. W ith a chlorine atom at th e 4'-
position of the naphthoylhydrindene, extensive elimination of the sub-stituent also occurred, but careful fractionation of the reaction mixtureafforded a small am ount of 6-chloro-20-methylcholanthrene. The 3-chloro isomer was obta ined without difficulty.
The second section of Table IV lists a number of variations of thecholanthrene synthesis. Th e ketones IX and X afford the expectedl',9-dimethylene-l,2-benzanthracene and 15,16-benzdehydrocholan-threne in '32% and 60% yield, respectively. Th e particularly high yield
in the latter case probably is associated with the presence in X of adoubly activa ted methylene group. The ketone X I is convertible into4,10-ace-l,2-benzanthracene , an isomer of cholan threne. T ha t theyield is only 10% is attributable to the fact that the condensationinvolves substitution into the benzene rather than the naphthalenenucleus. ar-a-T etralyl a-naphthyl ketone (X II) would be expected toyield homocholanthrene, but it affords instead 1,12-trimethylenechrysene(XIII) , evidently as the result of a disproportionative isomerization toan aromatic structure of greater stability.
XIII
Another variat ion consists in the use of certain aryl quinolyl ketonesfor the synthesis of polynuclear aromatic substances containing a con-densed pyr idine r ing. T hu s 5-quinolyl 7-methyl-4-hydr indyl ke to ne(XIV) on pyrolysis affords 20-methyl-4-azacholanthrene (XV) in 12%
ArMgBr + Ar'COCl (50%)ArCN + Ar 'MgBr (91%)Mixt. from a- and /3-C9H9BrArMgBr + Ar 'COCl (50% «)ArCN + Ar 'MgBr (89%)ArMgBr + Ar 'CN (49%)ArCN + Ar 'MgBr (61%)ArCN + Ar 'MgBr (82%)ArMgBr + Ar 'CN (46%)
yield. Similarly, 5-quinolyl o-tolyl ketone (Tab le I I I) yields 4'-aza-l,2 -benzan thracene 03-anthraquinoline). W ith 8-quinolyl 7-methyl-4-hydrindyl ketone the sole reaction product (50%) contains an atom ofoxygen and presumably is of a stabilized anthranol type of structure.
Pyrolysis of Diketones (Table V). The Elbs reaction has been appliedrather extensively, particularly by Clar and co-workers, to the synthesisfrom suitable diketones of higher hydrocarbons having two separate ormerged anthracenoid groupings (see refs. 57-64 in Table V ). Oneexample is the pyrolysis of 4,6-dibenzoyl-l,3-xylene (X V I), which yieldsa hydrocarbon having the probable structure of mesodihydropentacene(X V II). Th e formation of the dihydride rath er than th e fully aromatic
H 3C
0
x v i XVII
hydrocarbon doubtless is a consequence of the great reactivity ofpentacene. Th e most extensive elaboration of the method yet accom-plished is th e synthesis of 2,3,8,9-di-(naphtho-l',2')-chrysen e (X IX )
from the diketone XV II I. Th e hydrocarbon, which melts at 500° w as
X I X
obtained in 52% yield. Other examples listed in the table involvediketones which are similar to X V II I b ut in which one or both naph thoylgroups are replaced by benzoyl radicals.
Summary of Side Reactions. Examples have been cited in the fore-going sections of the occurrence of aroyl migrations in the course of theElbs pyrolysis, of the elimination of alkyl, halo, and methoxy substitu-ents, of the degradation of isopropyl to methyl, and of processes ofhydrogenation, dehydrogenation, and intramolecular disproportiona-
tion. The formation of anthrones in three instances represents theproduction of substances of a stage of oxidation higher than that of theexpected hydrocarbon, and there is one instance of an apparent reduc-tion. As a by-product in the synthesis of me thylcholanthrene, the re wasisolated 43 a substance which is resistant to dehydrogenation and which
probably is formed by the reduction of the carbonyl group of the startingketone. Another side reaction leads to the formation of hydroca rbonfragments such as phenanthrene from a phenanthryl aryl ketone,10 or
anthracene from an anthryl aryl ketone.
69
Apparently the ketonesuffers some cleavage by the water evolved, perhaps with subsequentdecarboxylation of the acid fragment: ArCOAr' + H2O —> ArCOOH +Ar'H —> ArH + CO 2 + A r'H. Clar 63 reports the formation of benzoicacid and benzaldehyde in the pyrolysis of l,5-dibenzoyl-2,6-dimethyl-naphthalene.
EXPERIMENTAL PROCEDURES
Preparation of the Required Ketones. The ketone requi red for a
given Elbs synthesis is often obtained most readily by the Friedel andCrafts reaction, and in many of the experiments cited the practice hasbeen to distil the total ketone or ketone mixture and submit it as suchto pyrolysis. Since th e distillate almost inva riably consists of a mixtureof isomers, this practice introduces uncertainties concerning the natureof the reaction and the yield. Except for the routine preparation ofmaterials by known methods, it is definitely advantageous either topurify and characterize the products o btained by th e Friedel and Crafts
method or to employ a synthesis from a Grignard or lithium derivative.The principal variations of this general synthesis have been studied
carefully in a number of instances, as summarized in the second columnof Tables I-I V . The reaction ArM gX + ArCOCl has been employed in10 instances with yields ranging from 40 to 59% and with an averageyield of 49 % . Bruce 41 has found that considerable losses are associatedwith side reactions resulting in the formation of ArH and (ArCO)2O.The use of a nitrile in place of an acid chloride is definitely advan tageous,
for in 22 examples the reaction ArM gX -+• ArCN has g iven pure ketonesin an average yield of 70% . Some of these syntheses represent pa rticu-larly difficult cases, for example where a cyanoquinoline constitutes onecomponent, and in the more normal instances th e yields frequently are inthe range 80-9 0% , particularly when th e inheren t slowness of the nitrilereaction has been recognized and adequate time allowed. The use of anamide as the second component has been investigated in only one in-stance, bu t with marked success. The condensation of phenylmagnesiumbromide with 1-acenaphthamide was found to proceed slowly (72 hr.)but very smoothly, affording 1-benzoylacenaphthene in 95% yield.33
Lithium derivatives have not been employed at all extensively but,except in special cases, probably are less satisfactory than the Grignardreagents. The reaction ArLi + ArCOCl has given yields described as"very low," 41 48%,50 and 21%.
between 1-acenaphthyllithium and a-naphthoyl chloride, which wasfound at least more satisfactory than the attempted condensation of 1-acenaphthylmagnesium iodide with the acid chloride. Yields reportedfor the reaction ArLi + ArCN are 51%,
1 8 50%,61 and 17.5% « (cyano-quinoline), and the synthetic method thus appears less advantageous
than the condensation of a Grignard reagent with the nitrile.Selection of Conditions for the Pyrolysis. A ttemp ts to find a catalyst
for the Elbs reaction have met with little success. Elbs 5 tried sulfuricacid, potassium bisulfate, phosphorus pentoxide, and zinc chloride withnegative results. M organ and Coulson 16 found piperidine and aceticanhydride also without effect and noted that 2,4,4'-trimethylbenzophe-none is cleaved by sulfuric acid to p-toluic acid and m-xylene.
The pyrolysis frequently has been conducted in the presence of a small
amount of zinc dust, and indeed in the first instance of the reaction Behrand van Dorp 1 passed the vapor of o-tolyl phenyl ketone over zinc dust.It is still questionable that the use of zinc results in any material im-provem ent. In two sets of parallel experiments 33' 64 conducted with andwithout zinc dust no difference was observable in the results. In thesynthesis of the 2-B0 and 3-methoxy B2 derivatives of methylcholanthrenethe yields in small-scale experiments were 36 and 38% in the presenceof zinc and 40 and 3 2% in its absence. I t was observed by Hershberg 32
that o-tolyl a-naphthyl ketone can be pyrolyzed at 400-410° in thepresence of zinc dust to give 1,2-benzanthracene in 6 1 % yield, but th a twithout zinc the reaction proceeds only very slowly a t th e same tempera-ture . This is the only concrete indication that zinc has any effect, andthe effect may be merely to lower slightly the pyrolysis tem peratu re. Acomparison of the first three entries in Table III would seem to indicatetha t the use of zinc improves the yield in the synthesis of 1,2,5,6-dibenz-anthracene, but in view of the experiment cited below as an example ofthe procedure it is probable that the higher yield reported by Bach-
mann 20 is attrib utab le more to his use of homogeneous G rignard ketonein place of the mixture re sulting from the F riedel and Crafts reac-tion.
Although many of the earlier experiments were conducted by heatingthe ketone over a free flame at the boiling point without control ormeasurement of the temperature, most workers now consider it advis-able to use a heating bath and to conduct the pyrolysis at the lowesttemperature at which a Steady liberation of water is observed.34'43 As
the bath temperature is brought slowly to or above 400°, the criticalpyrolysis temperature usually is sharply denned by a brisk bubblingwhich is hardly noticeable at a temperature 5° lower.43
Certain claims concerning modifications in the procedure of conducting
sealed on about 11 cm. above the flask and carrying a 100-cc. receivingbulb. The flask is charged with 152 g. of the crude ketone and heated in anitra te-nitrite b ath (care!) at 430° ± 5° (bath). Th e pyrolysis must beattended constantly and the upper part of the flask warmed occasion-ally with a free flame to prevent water from condensing and dropping
back into the hot mix ture . Sweeping of the vessel with dry nitrogen orcarbon dioxide perhaps facilitates somewhat the removal of water but isunnecessary and offers no ma terial advan tage. The evolution of waterslackens noticeably within about three hours, and after three and one-half hours th e flask is removed from the ba th , some glass wool is pusheddown into the bulb to promote even boiling, and the mouth of the flaskis sealed off. The product is the n distilled at 2-3 mm . pressure, the low-boiling material which comes over in a fore-run being removed from the
receiver. The dibenzanthracene distils largely a t a ba th temp era ture of300-320°, and in pa rt on raising the bath to 400°. During distillationthe upper part of the flask is kept hot with a free flame. By using aflask with a high side arm and distilling carefully, a clean distillate canbe obtained and redistillation is unnecessary. The distillate is melted,poured (and rinsed) into a 4-1. flask , and dissolved in abou t 3 1. of boilingbenzene. The solution is concen trated until crystallization sets in(about 1800 c c ). The yellow dibenzanthracene sepa rating in the firstcrop and melting at 260-262° (cor.) amounts to 44 g. (31% ). The mate-rial recovered from the mother liquor when recrystallized melts at 253-258° and weighs 4 g.; to ta l yield of yellow product, 3 3 % . Almostidentical yields were obtained in 20-g. and 80-g. runs and in runs con-ducted with added zinc dust.
In one method for the preparation of colorless dibenzanthracene,30 awarm solution of 2 g. of lead te tra ac eta te in 500 cc. of acetic acid is addedin small portions to a warm solution of 10 g. of yellow hydrocarbon in500 cc. of benzene, and the solution is refluxed gently for one hour. Thesolvent is then distilled slowly until the solution has been reduced involume to 300-350 cc. On cooling, dibenzanthracen e separates in com-pletely colorless plates with a blue fluorescence in ultrav iolet light, m.p.265-266° (cor.) (purest sample, 266-266.5°). The recovery ordinarilyamounts to 70-83%. With particularly poor samples of crude hydro-carbon a second treatment with lead tetraacetate may be required; thiswas true of a sample melting at 250-255°, from which the recovery ofthoroughly purified material was 50%.
Example 2. 1,2-Benzanthracene.20 '32 o-Tolyl a-naphthyl ketone isprepared 20 by adding 23.4 g. of o-tolunitrile to the Grignard reagentfrom 50 g. of a-brom onaphthalene in 75 cc. of ether and 75 cc. of benzene.The mixture is refluxed for eight hours, cooled, and hydrolyzed with ice
yellow solid weighing 113.3 g. This is dissolved in 400 cc. of benzene andthe solution is cooled slightly and diluted w ith 1 1. of ether. The bulk ofthe methylcholanthrene separates in a nearly pure condition as fine yel-low needles (72 g.). This is dissolved in 500 cc. of benzene, and 300 cc.of ether is adde d; on cooling, the hydroca rbon separa tes as yellow needles
of high purity (63 g.), m.p. 178.5-179.5° (cor.) (purest sample, 179.5-180°). The mother liquor from this crystallization when concentratedand trea ted with 12 g. of picric acid affords 12.5 g. of methylcholan threnepicrate, m .p. 176-177°. The oily material recovered from the originalmother liquor is pyrolyzed again and the product distilled, crystallizedonce from benzene-ether, and converted to the picrate in benzene solu-tion . Recrystallization affords 14.5 g. of satisfactory p icrate , m .p.178-179°. Th e total yield of materia l collected as such or as the picrate
T h e C l e m m e n s e n R e d u c t i o n in t h e P r e s e n c e of S o l v e n t s of B o t h T y p e s
( M e t h o d I V ) 1 68
T h e C l e m m e n s e n R e d u c t i o n w i t h U n a m a l g a m a t e d Z i n c ( M e t h o d V ) . . . 1 68
T A B L E O PC O M P O U N D S R E D U C E D B Y T H E C L E M M E N S E N M E T H O D . . . . 1 6 9 - 2 0 0
I N T R O D U C T I O N
The replacement of the oxygen atom of the carbonyl group in analde-hyde or ketone by two hydrogen atoms through the use of amalgamatedzinc and hydrochloric acid was first employed in 1913 by Clemmensen 1
and is known as the Clemmensen method of reduction. The process hasbeen applied to a large number of aldehydes and ketones as a step in
the synthesis of polynuclear hydrocarbons and alkylated aromatic com-pounds, including those containing one or more phenolic hydroxylgroups. It has also played an important role in the elucidation of thestructures of highly complex natural products.
The formation of hydrocarbons from aldehydes and ketones by theClemmensen reaction can be illustrated by the following equations:
RC—H+ 4(H)
Z° ™' >
RCH 3+ H
2O
RC—R' + 4(H) ^ f * > RCH 2R' + H20
The method is of peculiar value because nearly all other reducing agentswhich have been employed convert aldehydes and ketones to the corre-sponding carbinols or pinacols, rather than to the hydrocarbons. Thechief alternative methods of accomplishing the same transformation are
catalytic hydrogenation and reduction with hydrazine and alkali (Wolff-Kishner method).The mechanism of the reduction by amalgamated zinc and hydro-
chloric acid is not clearly understood. If the carbinol is assumed to bethe intermediate, then these same reagents should be suitable for the re-placement of an alcoholic hydroxyl group by a hydrogen atom. How-ever, with few exceptions, alcohols are not affected by zinc amalgamand hydrochloric acid. Only act ivated alcoholic hydroxyl groups, suchas those in /3-hydroxy acids and benzyl alcohols, are removed by the
Clemmensen reagents.The wide use of this method of reduction has resulted in the develop-
ment of several modifications of the original procedure. These consist1
Several other /3-diketones have been reduced without rearrangement;some of the reactions have been interrupted to produce monoketones.6
Aliphatic ketones containing primary,1" secondary,6 or ter t iary 7 hy-droxyl groups undergo reduction of the carbonyl group without changeof the alcohol function. These observations, together with the fact
that alcohols have been used satisfactorily as solvents, indicate that al-coholic hydroxyl groups are not ordinarily reduced by amalgamatedzinc and hydrochloric acid. However, the direct replacement of an un -activated hydroxyl group has been observed in one case; 3-hydroxy-7,12-diketocholanic acid is reduced to cholanic acid.8 The reduction of 1,2-glycols, which has been observed with certain sterol derivatives, 9 maydepend on preliminary dehydration to ketones which then react in theusual way.
Aliphatic-Aromatic Ketones. Most aliphatic-aromatic ketones reactnormally, and numerous carbonyl compounds of this type, particularlyphenolic ketones, have been reduced by the Clemmensen me thod. Cy-clic ketones obtained by ring closure of 7-arylbutyric acids are alsoreadily converted t o hydrocarbons. K etones of very slight water solu-bility are best reduced by employing a hydrocarbon solvent and operat-ing in such a way that the amalgamated zinc is in contact with boththe aqueous acid and the hydrocarbon solution l0 (see p. 167).
The presence of a carboxyl group attached to the aromatic nucleusfrequently causes the reaction to proceed more rapidly and in excellentyields.11 It is probable that the carboxyl group assists in maintainingthe required concentration in the acid mixture by increasing the solu-bility of the carbonyl compound.
The reduction of aliphatic-aromatic ketones containing one, two, orthree hydroxyl or methoxyl groups on th e a rom atic ring proceeds excep-tionally well. Q uan titative yields are obtained w ith the lower members,and even with the higher homologs the yields are very good. The reac-
tion proceeds rapidly, and in some cases it is possible to employ thecolor produced with ferric chloride as a control test.12
6 (a ) Wie land and Mar tz , Ber., 59, 2352 (1926); (6) Q udra t- i -K hud a, J. Chem. Soc,206 (1930); (c) Russicka, Koolhaas, and Wind, Helv. Chim. Ada, 14, 1151 (1931); (d)Chuang , Ma , and T ien , Ber., 68, 1946 (1935); (e) Friedmann, J. prakt. Chem., 146, 65(1936); ( / ) Bardhan and Sengupta , J. Chem. Soc, 2520 (1932).
6 M a r k e r a n d L a w s o n , J. Am. Chem. Soc, 6 1, 852 (1939).7 Lutz and Smal l , J. Org. Chem., 4, 220 (1939).8 Borsche and Ha l lwass , Ber., 55, 3325 (1922).9
M a r k e r , K a m m , O a k w o o d , W i t t l e , a n d L a w s o n , / . Am. Chem. Soc, 60, 1067 (1938).10 Mikeska , Smi th , and L iebe r , J. O rg. Chem ., 2, 499 (1938).11 Cox, J. Am. Chem. Soc, 52, 352 (1930).12 (a) Dohme, Cox, and Mil ler , J. Am. Chem. Soc, 48, 1688 (1926); (6) Cox, J. Am.
Side reactions accompany the reduction of many aliphatic-aromaticketones, and in a few cases resinous products are formed in considerablequan tities. Styrene, styrene polymers, and the pinacolone of acetophe-none (formed by rearrangement of the pinacol) have been isolated as
by-products in the preparation of ethylbenzene from acetophenone.13 Inthe reaction of 2,6-dihydroxyvalerophenone with amalgamated zinc andaqueous hydrochloric acid, cleavage of the ketone has been observed,but in ethanolic solution the reduction is satisfactory.14 Although mostindandiones which have been studied react normally,16 the indan pro-duced from 2,2-diethyl-6,7,8,9-tetrahydro-l,3-a-naphthindandione bythe ordinary procedure is not completely free of oxygen compounds,and reaction over an extended period yields the 2-alkyl-5,6,7,8-tetra-
hydronaphthalene, formed by reductive opening of the indan ring.16 0
Aromatic Ketones. The reduction of benzophenone and its homologsby the original Clemmensen procedure is reported to be unsatisfactorybecause of the formation of resinous mater ials. On the other hand, p-hydroxybenzophenone l c is transformed to p-hydroxydiphenylmethanein qua ntit ativ e yield. 2,4-Dihydroxy 16° and 2,4,6-trihydroxybenzophe-nones 166 give the expected products in somewhat less satisfactory yields.o-Benzoylbenzoic acid is converted to o-benzylbenzoic acid, but reduc-tion w ith zinc dust an d alkali is more convenient and gives be tte r yields.17
Either benzil or benzoin is transformed to diphenylethane in good yieldsby the action of amalgamated zinc and aqueous hydrochloric acid, lc
but the reduction of benzoin in the presence of ethanol affords stilbenein good yield.18 2,4,6,2',4',6'-Hexamethylbenzil is unaffected by zincamalgam and concentrated hydrochloric acid. A nthraquinone la andcertain of its derivatives 19 are reduced to dihydroanthracenes.
Keto Acids
a-Keto Acids. Th e carbonyl group of a-keto acids is attacke d under theconditions of the Clemmensen reduction, but the products are the a-hy-droxy derivatives rath er th an the completely reduced acids. For ex-ample, phenylglyoxylic acid and its ethyl ester give mandelic acid and
13 Steinkopf and Wolfram, Ann., 430, 113 (1923)." A d a m s , C a i n , a n d B a k e r , J. Am. Chem. Soc, 62, 2201 (1940).16 (a ) v . Braun , Ki rsohbaum, and Schuhmann , Ber., 53, 1155 (1920); (6) Fleischer and
co-workers, Ber., 63, 1255 (1920); 56, 228 (1923); Ann., 422, 231, 272 (1921).16 (a ) Kla rmann , J. Am. Chem. Soc, 48, 791 (1926); (6) K larm an n a nd F igdor, ibid.,
48, 803 (1926)." M a r t i n , J. Am. Chem. Soc, 58, 1438 (1936).18 Bal l a rd and Dehn , J. Am. Chem. Soc, 54, 3969 (1932).19 Backer , S t ra t ing , and Huisman , Bee trav. chim., 58, 761 (1939).
ethyl mandelate, respectively,13 and ethyl 9-fluoreneglyoxylate yieldsthe corresponding hydroxy ester.20
|3-Keto A cids. The reduction of a few esters of /3-keto acids has beeninvestigated. E thy l acetoacetate is transformed to ethyl butyra te in30% yield, and ethyl benzoylacetate to ethyl hydrocinnamate in 59%yield.13 The reduction of a /?-keto ester of the bile acid series, the me thylester of 6-ketolithobilianic acid,21 and of two bicyclic di-(/3-keto) es-ters 22 is recorded.
7-Keto Acids. The most important acids of this type are those ob-tainable by the Friedel and Crafts reaction of succinic anhydride or itssubstitution products with aromatic compounds or by the action of anaryl Grignard reagent with such an anhydride. Th e reduction of theseketo acids by one of the Clemmensen procedures is satisfactory, al-
though in cer tain cases some resinification occurs. A bimolecular by-product, the dilactone of 7,7'-diphenyl-y,7'-dihydroxysuberic acid, hasbeen isolated from /3-benzoylpropionic acid.23
/3-Aroylpropionic acids with methoxyl groups attached to the aro-matic ring are best reduced in the presence of a solvent (toluene) immis-cible with the hydrochloric acid.17 /3-(4,8-Dimethoxy-l-naphthoyl)-pro-pionic acid yields 7-(4,8-dimethoxy-l-naphthyl)-butyric acid and an ab-normal product, 7-(4-methoxy-5-tetralyl)-butyric acid.24 The formation
of the latter compound involves the reduction of the ring carrying thecarbonyl group and the elimination of the methoxyl group from thatring. A side reac tion in the reduction of /3-(p-bromobenzoyl)-propionicacid results in the replacement of the bromine atom by a hydrogenatom.25 Esters of /3-aroylpropionic acids undergo simultaneous reduc-tion and hydrolysis to give 7-arylbutyric acids.26
The Clemmensen reduction of purely aliphatic 7-keto acids and theiresters has no t been studied extensively. E thyl levulinate 13 yields ethyl
valerate, but neither 7-ketopimelic acid nor its dimethyl ester
27
is re-duced.Other Keto Acids. 5-Keto acids and molecules in which the keto
group is still further removed from the carboxyl group react normallyin both aliphatic and aliphatic-aroma tic series. Th us, the reductions
20 Wisl icenus and Weitemeyer , Ann., 436, 1 (1924).2 1 W i n d a u s , Ann., 447, 233 (1926) .2 2 G u h a , Ber., 72, 1359 (1939).23 Overbaugh, Al l en , Mar t in , and F iese r , Org. Syntheses, 15, 64 (1935).24 Fiese r and Hershberg , J. Am. Chem. Soc, 58, 2382 (1936).26 Fieser and Sel igman, J. Am. Chem. Soc, 60, 170 (1938).26 (a) F ieser and Peters , J. Am. Chem. Soc, 54, 4373 (1932); (6) Haworth and Mavin,
J. Chem. Soc, 2720 (1932).27 J fomppa , Ann. Acad. Sci. Fennicae, A51, No. 3 (1938) [C. A., 34, 2335 (1940)]. .
of 7-(p-anisoyl)-butyric acid,28 of octacosan-14-one-l,28-dioic acid 29
[H O 2C(CH 2)i2CO(CH 2)13CO2H], and of 22-phenyldocosan-13-one-l-oicacid30 [C 6H5(CH 2)9CO(CH 2)UCO 2H] have been reported. W ith thelast two compounds extended periods are required for the completion of
the reaction.a,/3-Unsaturated Carbonyl Compounds
Little information is available concerning the Clemmensen reductionof cc,j3-unsaturated compounds. Both th e carbonyl group and the ethy l-enic linkage of unsaturated acids of the |8-aroylacrylic acid type 31 arereduced. Similarly, 2,3-diphenylcyclopentene-2-one-l is converted to2,3-diphenylcyclopentane.32 w-Butylbenzene is obtained in 50% yieldfrom benzalacetone, but the major product from benzalacetophenone isa bimolecular one, l,3,4,6-tetraphenylhexane-l,6-dione.33 Isolateddouble bonds apparently are not affected by amalgamated zinc and hy-drochloric acid.
Chromanones are converted to chromans by means of amalgamatedzinc and hydrochloric acid;84 e.g., 7-hydroxy-2,2-dimethylchromanone isreduced to 7-hydroxy-2,2-dimethylchroman.34a Acylated coumarins arereduced to alkyl coumarins by the method of Clemmensen,36 and it isreported that 6,8-diethyl-5-hydroxy-4-methylcoumarin is obtained by
the reduction of 6-acetyl-8-ethyl-5-hydroxy-4-methylcoumarin.35c
The Reduction of Other Functional Groups by Amalgamated Zincand Hydrochloric Acid
Compounds containing sensitive groups in addition to carbonyl some-times undergo reductions of more tha n one type . It was mentionedabove that an ethylenic link is reduced when it is conjugated with a car-
bonyl group . Th e double bond of a,/3-unsaturated acids, such as cinna-mic acid,13 is also satu rate d by zinc amalgam and acid. Pyrroles 36 and28
Sengupta, / . Indian Chem. Soc, 17 , 183 (1940).32
Burton and Shoppee, J. Chem. Soc, 567 (1939).33
D i p p y a n d L e w i s , Rec trav. chim., 5 6 , 1 00 0 (193 7) .34 (a ) Br i d g e , Cr o c k e r , Cu b i n , a n d R o b e r t s o n , / . Chem. Soc, 1 5 3 0 ( 1 9 3 7 ) ; (6 ) G e o r g e
a n d R o b e r t s o n , J. Chem. Soc, 1 5 3 5 (193 7) ; (c ) A n d e r s o n a n d M a r r i a n , J. Biol. Chem.,
127, 6 4 7 (193 9) .36 (a ) C h o w d h r y a n d D e s a i , Proc. Indian Acad. Sd., 8 A , 1 (1938) [C. A . , 3 2 , 9 0 6 5
(1938) ] ; (6) L i m a y e a n d L i m a y e , Rasayanam (Suppl.) (1938) [C. A . , 3 3 , 1 6 98 (1939)] ;
(c) Desai a n d E k h l a s , Proc. Indian Acad. Sd., 8 A , 5 6 7 (1938) [C . A . , 3 3 , 3356 (1939)] .36 (a ) W i b a u t a n d H a c k m a n n , Rec. trav. chim., 5 1 , 1 1 5 7 ( 1 9 3 2 ) ; (6) W i b a u t a n d O o s t e r -
reaction at room temperature, particularly when the carbonyl com-pound is sensitive to th e strong acid mixture. In such cases the reac-tants are allowed to stand at room temperature for one to two days andthe reduction is then completed by heating to reflux for a period ranging
from fifteen minutes to two hours. By this method 3,4-dihydroxytoluenehas been o btained from 3,4-dihydroxybenzaldehyde,44 and Y-(a-thienyl)-butyric acid is produced in excellent yield from jS-(a-thenoyl)-propionicacid.45
An improvement in yield frequently results if the substance to be re-duced is first converted to a derivative which has a lower melting pointand a greater solubility in the reaction m ixture. Although /3-3-acenaph-thoylpropionic acid26a and j3-(l-methyl-4-naphthoyl)-propionic acid 2fl!>
are not attacked by amalgamated zinc and hydrochloric acid, theirethyl esters are reduced in yields of about 40%.The use of mechanical stirring has been reported in the conversion of
4-acylresorcinols 12a to alkylresorcinols, but in most cases sufficient agi-tation is provided by the ebullition of the hot acid.
The physical form of the zinc appears not to be critical, since zincturnings, zinc wool, granulated zinc, zinc powder, and mossy zinc havegiven good results. M ossy zinc has been most commonly used. I t hasbeen reported2 that a very satisfactory zinc dust can be prepared bypulverizing pure zinc.
The zinc is ordinarily amalgamated by t reatm en t with 5 to 10% ofits weight of mercuric chloride in the form of a 5 to 10% aqueous solu-tion. Th e time required for amalgamation can be diminished by em-ploying a solution of mercuric chloride in very dilute hydrochloricacid.17 In order to obtain a homogeneous amalgam, it is advisable toshake or stir the mixture during the amalgam ation. Th e quality of theamalgam is said to be improved by three washings of the zinc with hot
hydrochloric acid 2 before amalgamation.
Preparation of Zinc Amalgam
A m ixture of 100 g. of mossy zinc, 5 to 10 g. of m ercuric chloride, 5 cc.of concentrated hydrochloric acid, and 100 to 150 cc. of water is stirredor shaken for five minutes. The aqueous solution is decanted, and theamalgamated zinc is covered with 75 cc. of water and 100 cc. of concen-
trat ed hydrochloric acid. The material to be reduced, usually 40 to50 g., is then added immediately and the reaction is started.
44 Anshul tz and Wenger , Ann., 482, 25 (1930).48 Fieser and Kennel ly, J. Am. Chem. Soc, 57, 1611 (1935).
The Clemmensen Reduction in the Absence of an Organic Solvent
(Method I)
Reduction of /3-(^-Toluyl)-propionic acid.17 A mixture of amalga matedzinc (prepared from 100 g. of mossy zinc and 5 g. of mercuric chloride
as described above), 75 cc. of water, 175 cc. of concentrated hydrochloricacid, and 50 g. of /3-(p-toluyl)-propionic acid is refiuxed vigorously forten hours in a 1-1. round-bottomed flask. A 50-cc. portion of concen-trated hydrochloric acid is added every three hours during the heatingperiod. After the reaction mixture has been cooled to room tempera-ture, the solid 7-(p-tolyl)-butyric acid is collected and washed withsmall am ounts of cold wate r. Th e filtrate and washings are combinedand extrac ted with three 75-cc. portions of ether. Th e solid product is
dissolved in the combined extracts and, after filtration from a smallamount of insoluble m ateria l, the solution is dried over calcium chloride.Th e solvent is then removed and the residue is distilled under diminishedpressure. The produc t, a colorless oil, crystallizes to a white solid melt-ing at 61-62°. The yield is 41 g. (88% ).
Reduction of 2,4-Dihydroxyacetophenone.I("41o> 46 A mixture ofamalgamated zinc (prepared from 200 g. of mossy zinc and 10 g. of mer-curic chloride as described on p. 163), 150 cc. of water, 150 cc. of con-
centrated hydrochloric acid, and 50 g. of 2,4-dihydroxyacetophenone(resacetophenone) is refiuxed in a 1-1. round-bottom ed flask un til a dropof the liquid in ethano l gives no color with aqueous ferric chloride. Aportion of about 10-15 cc. of concentrated hydrochloric acid is addedhourly. W hen the color test indicates the reaction to be complete (threeto four hours) the mixture is cooled and the solution is decanted fromany unchanged zinc amalgam. The solution is satu rate d with sodiumchloride and extracted w ith ether to remove the reaction product. Re-moval of the solvent yields a light yellow solid which crystallizes from
benzene or chloroform as thick white prisms, m.p . 97°. The yield is44 g. (97%).
The Clemmensen Reduction in the Presence of a Solvent Misciblewith Aqueous Hydrochloric Acid (Method II)
Certain carbonyl compounds which are not appreciably soluble in theacid mixture are reduced with difficulty b y M ethod I. In such cases
the reaction is often facilitated by the addition of a solvent, such asethanol, acetic acid, or dioxane, which is miscible with the aqueous hy-drochloric acid. Fo r example, bilianic acid is reduced by means of a
produc t is a colorless liquid b .p. 132-133 °/25 mm. Th e yield is 88.5 g.(95%).
Reduction of Y -Keto-7-(2-fluorene)-butyric Acid.49 A mixture of 90 g.
of y-keto-7-(2-nuorene)-butyric acid, 450 cc. of ethanol, 450 cc. of con-
centrated hydrochloric acid, and 180 g. of amalgamated zinc is refiuxedfor one hour. A second 450-cc. portion of concen trated hydrochloricacid is then added, and refluxing is continued for eight hours . Th e mix-ture is cooled, and the solid is collected and dissolved by boiling with1000 cc. of 5% aqueous sodium hydroxide. After n itra tion and acidifi-cation the 7-(2-fluorene)-butyric acid sepa rates . Th e yield of crudeproduc t is 85 g. I t is readily purified by recrys tallization from aceticacid followed by recrystallization from benzene-petroleum ether, yield-ing white plates, m.p. 151-151.5°.
The Clemmensen Reduction in the Presence of a Solvent Immisciblewith the Hydrochloric Acid (Method III)
A large number of carbonyl compounds have been reduced in pooryields by Methods I and II, and, especially in the cases of certain ketoacids, the difficulty has been ascribed to the formation of insoluble poly-molecular reduction products which coat the surface of the zinc.17 The
addition of a hydrocarbon solvent, such as toluene, which is immisciblewith the hydrochloric acid is beneficial in those cases because it keepsmost of the material out of contact with the zinc, and in the aqueouslayer the reduction occurs at such a high dilution that polymolecularreactions are largely inhibited.
The modification is particularly advantageous with keto acids whichcontain m ethoxyl groups. Such compounds may suffer hydrolysis ofmethoxyl groups during the reduction; consequently it is desirable to
treat an alkaline solution of the crude reaction product with methylsulfate, in the presence of a trace of sodium hydrosulfite if darkeningoccurs during methylation, to recover any demethylated material.
Certain extremely insoluble compounds cannot be reduced by thismethod unless both the aqueous layer and the hydrocarbon layer are incontact with the zinc.
Reduction of /3-Benzoylpropionic Acid.17 To 120 g. of mossy zinc,
amalgamated as described on p. 163, 75 cc. of water, 175 cc. of concen-
trated hydrochloric acid, and 100 cc. of toluene is added 50 g. of /J-ben-zoylpropionic acid. The mix ture is refiuxed briskly for twenty-four tothirty hours, during which time a 50-cc. portion of concentrated hydro-chloric acid is added every six hours. The solution is cooled to roomtem pera ture, the aqueous layer is separated and, after dilution with 200
cc. of wa ter, is extracted w ith three 75-cc. portions of ether. The com-bined ether and toluene solutions are washed with a little water anddried over calcium chloride. The solvents are removed by distillationunder diminished pressure, and the residue is distilled. 7-Phenylbutyric
acid, b.p. 178-181°/19 mm., is obtained as a colorless oil which solidifiesto white crystals, m.p. 46-48°. The yield is 41 g. (90%).
Reduction of /3-(/»-Anisoyl)-propionic Add.17 To 120 g. of mossy zincamalgamated as described on p. 163 are added, in the order given, thefollowing: 75 cc. of water, 175 cc. of concentrated hydrochloric acid,100 cc. of toluene, and 50 g. of /3-(p-anisoyl)-propionic acid. The mix-ture is refluxed briskly for forty-eight hours, during which time a 25-cc.portion of concentrated hydrochloric acid is added every six hours. The
solution is cooled to room temperature; the aqueous layer is separatedand, after dilution with 200 cc. of water, is extracted with three 75-cc.portions of ether. The toluene and ether extracts are added to 300 cc.of 5% aqueous sodium hydroxide, and the solvents are removed bysteam distillation.
The residual alkaline solution is cooled to 80°, and 5 to 10 cc. ofme thyl sulfate is added. If necessary, aqueous sodium hydroxide is in-troduced to keep the solution alkaline. After the mixture has been
shaken or stirred for thirty to forty-five minutes, the excess alkali isneutralized and the solution is trea ted w ith charcoal. The colorless oryellow filtrate is cooled to 10° and acidified by the slow addition of hy-drochloric acid. The mixture is kept in an ice ba th un til the precipita-tion of the product is complete. I t is then filtered and the solid iswashed with a little cold water. The crude ma terial, obtained in quan-tita tiv e yield, is sufficiently pure for most purposes. Fo r purification i tis dissolved in ether and the solution is filtered from any insoluble ma-terial. The solvent is removed and the residue is distilled under dimin-ished pressure. The yield of 7-(p-anisyl)-butyric acid, b.p . 182-186°/4mm ., m.p. 61-62°, is 43 g. (94% ). For further purification the acid maybe recrystallized from petroleum ether (b.p. 30-60°).
Reduction of Stearophenone.10 Mossy zinc is added to a weighed 2-1.
Erlenmeyer flask un til a layer about 8 cm. deep is formed. Th e weight ofthe zinc is determined, and the metal is amalgamated by treatment withthe appropriate amounts of mercuric chloride, water, and hydrochloricacid (p. 163). To the zinc amalgam is added sufficient concen trated
hydrochloric acid to cover about one half of it, followed by a solutionof 250 g. of stearophenone in 750 cc. of xylene. The mixture is heatedunder reflux for seven hours, during which time gaseous hydrogen chlo-ride is passed into the bottom of the flask to replace losses. The xylenelayer is separated, the solvent removed, and the product distilled, b.p.
* Q . yield repor ted as qu an t i ta t ive ; G, yield repor te d aa good; P, yield repor ted as poor . A daahindicates that the yield is not repor ted.
t R eference num bers refer to the bibl iography on pp . 201-209.t R.P., reduction product.
CuHgOBrCisHuOBrC uH uO B rC i 3 H I 7O 2C lCiaHi7O 3C lC i 3 H 1 6O N
C i 3 H 2 1O NC i , H i i O , NCI3H17O3N
C o m p o u n d
a,a-Dimethyl- /3-p- toluylpropionic ac idEthyl /3-o- toluylpropionate
E thyl 0-p- to luylpropiona te0-4-Isopropylbenzoylpropionic ac id6 ,7-Dimethoxy-2-methyl - l - t e t ra lone
6 ,7-Dimethoxy-3-methyl - l - t e t ra lone2 ,6-Dimethoxyphenyl i sobuty l ke tone2 ,5-Dimethoxyphenyl i sobuty l ke tone4-Hydroxy-3-methoxyphenyl n -amyl ke tone2 ,4-Dihydroxyphenyl n -hexyl ke tone2 ,4 ,6 -Tr ihydroxybenzophenone8-Ace ty l -7-methoxy-4-methylcoumar in
7-Hydroxy-4-methyl -8-propionylcoumar in5-Hydroxy-4-methyl -6-propionylcoumar in3-Carboxy-4-hydroxyphenyl i soamyl ke tone3-Carboxy-4-hydroxy-n-caprophenone(3-4-Methoxy-2,5-dimethylbenzoylpropionic
acida -Methyl - /3 -3-methoxy-2-methylbenzoylpro-
pionic ac id5-p-Anisoylvalerie acid5-Hydroxy-6 ,7-d imethoxy-2-methyl - l - t e t -
ra lone
2 ,4 ,6 -Tr ihydroxyphenyl n -hexyl ke tonea-Methyl- /3-3,4-dimethoxybenzoylpropionie
acid
/3-Methyl-/3-3,4-dimethoxybenzoylpropionic
acid7-3 ,4-Dimethylbenzoylbutyr ic ac ida-Methyl- /3-2-hydroxy-3,4-dimethoxybenzoyl-
propionic ac id2- , 3-, or 4-Bromofluorenone-9o-Bromobenzophenone
4-Bromo-5 ,6-benzo- l - indanone4-Bromo-7- ( -buty l - l - indanone7-Bromo-4- i -bu ty l - l - indanone5-Chloro-2-hydroxyphenyl n-hexyl ketone5-Chloro-2,4-dihydroxyphenyl n-hexyl ketonel -Ke to-5 ,6-benzo- l ,2 ,3 ,4 ,7 ,8 -hexahydro-
pyr idoco l inel -Ke to-5 ,6-benzododecahydropyr idoco l ine/3-2-Quinolylpropionic acidEthyl |8-(3,5-dimethylpy'ridoyl-2) propionate
M e t h o d
II I
IIIj
I
II
I III
I I
II I
II
I
I II
I
I
II
II
II I
I
II
I II I
II I
I
III
Yield *
8243
—8064
64
4348——50—
———88
92
—73
—
30
45—
—73
—Q
—
50—
7070——
37
52—91
Refer-ence t
30617 9
3613 5
134• 1 9Q
13 4140280
915 356
274
424340
8989
94
225183
437
457
12913 4
134448
43719 6
351
883 3
26426416 5167
27627624429 9
* Q, yield reported as quantitative; G, yield reported as good; P, yield reported aa poor. A daabindicates that the yield is not reported.
t Reference numbers refer to the bibliography on pp.201-209.t R.P., reduction product.
Phenacyl phenyl sulfideR . P J (a-Methylbenzyl)phenyl sulfide
Method
III
IIIIII
Yield *
163220
Refer-
ence f
155
15541 921 2
96
C u
CuHi2O
CuHuOCisHuO
CisHuO
C I S H H O
C I J H M O
C I S H U O
d6Hi4 O
C16H1 6O
CuH1 6O
C1 6H1 8O
CisHisO
CuHisO
C ,SH 2 2OC i 6H 2 2 OC15H24O
Cl6Hl4O 2
C i s H u O aCl6Hl6O 2
C15H16O2
CisH 2 2O 2
Ci6H 22O 2
Ci6H 22O 2
B e n z a l a c e t o p h e n o n e
R . P . t T e t r a p h e n y l h e x a d i o n e2-Ethy l -4 ,5 -be nz o- l - inda nonel - K e t o - 2 - m e t h y l - l , 2 , 3 , 4 - t e t r a h y d r o p h e -
n a n t h r e n el - K e t o - 4 - m e t h y l - l , 2 , 3 , 4 - t e t r a h y d r o p h e -
n a n t h r e n el - K e t o - 9 - m e t h y l - l , 2 , 3 , 4 - t e t r a h y d r o p h e -
n a n t h r e n e4 - K e t o - 3 - m e t h y l - l , 2 , 3 , 4 - t e t r a h y d r o p h e -
n a n t h r e n e4 - K e t o - 7 - m e t h y l - l , 2 , 3 , 4 - t e t r a h y d r o p h e -
n a n t h r e n e8 - E t h y l p e r i n a p h t h a n o n e - 7l -K e to -3 ,4 ,5 ,6 ,12 ,13-he xa hydro -pe r ibe nz o-
a o e n a p h t h e n ea - K e t o - o c t a h y d r o m e t h y l e n e p h e n a n t h r e n e6 ,7 -Cyc lope n te no- l -ke to -2 ,2 -d ime thy l - l , 2 ,3 ,4 -
t e t r a h y d r o n a p h t h a l e n e7 - M e t h y l - l - k e t o - l , 2 , 3 , 4 - t e t r a h y d r o n a p h t h a -
lene-2 ,2-spi rocyc lopentanel - K e t o - l ,2 , 3 , 4 - t e tr a n a p h t h a l e n e - 2 , 2 -s p i r o -
cyc lohexane
A c e ty ld ie thy lme s i ty l e neP h e n y l »-oc ty l k e t o n e4a ,5 ,6 ,7 ,8 .8a -H e xa hydro -3 -n -p ropy l -4 -e thy l -
1 (2 ) -na ph tha l e n one4 - H y d r o x y - 3 , 5 - d i m e t h y l b e n z o p h e n o n e
4 - H y d r o x y - 3 - p h e n y l p r o p i o p h e n o n e4 ,5 -Cyc lohe xe ny l -2 ,2 -d ime thy l inda n- l , 3 -d ione
l -H ydroxy-2-na ph thy l r a -bu ty l ke tone2-H ydroxy-3 ,5 -d i -w-propy lp rop iophe none2-Hydroxy-3 ,5-dimethylphenyl re -hexyl ke tone
4-H ydroxy-3 ,5 -d ime thy lphe ny l n -he xy l ke tone
I
I
I
IV
I
I
II
II
I
I
I
II
I
I
I V
III
II
I
—
—
9 4
—.
—6 4
—
7 0
—
—
—
4 55 1
5 4
4 0
7 4
7 0
—
5 0
7 1
6 0
5 3
2 8 8
3 3
1 2 1
3 9 7
132
1 2 1
1 2 53 3
6 5
6 5
4 3 3
4 1 0
2 4 8
4 7 44 7 1
2 5 9
106
3 9 8
2 6
2 2
176
1 4 1
106
106
* Qt yield reported as quantitative; G, yield reported as good; P, yield reported as poor. A daabindicates tha t the yield is not reported
t Reference numbers refer to the bibliography on pp.201-209.$ R.P.,reduction product.
Cyc looc tadeean- l ,10-d ione/3-2-Anthroylpropionic acid)3-2-Phenanthroylpropionic ac id/3-3-Phenanthroylpropionic ac id/3-9-Phenanthroylpropionic ac id/J -[2-(9,10-Dihydrophenanthroyl)]-propionie
acid
3-K eto-2,5-diphen ylcyclope ntane -l -car boxy l ie
acido-(6-Tetroyl)benzoic ac ida ,a -Spi rocyc lopen tane- | 3 - l -naphthoylpro-
pionic ac id
a , /3-Dimethyl-a-phenyl-$-benzoylpropionie
acid/3- [9-( l ,2 ,3 ,4-Te t rahydrophenan throyl ) -pro-
pionic ac id1,4-Di-p-anisylbutanone-l
P-6-J -Buty l -2-naphthoylpropion icac id/3-5,6 ,7 ,8-Tetramethyl-2-naphthoylpropionio
acidM e t h y l a,a-dimethyl-/3-4-methyl-l-naphthoyl-
p r o p i o n a t e/3-[9-( l ,2 ,3 ,4,5,6 ,7 ,8-octahydrophenanthroyI)]-
propionic ac id|3-[6-( l ,2 ,3 ,4,9 ,10,11,12-octahydrophenan-
throyl)]-propionic ac id^-(5 or 6) -Cyc lohexane- l - sp i rohydr indoylpro-
pionio acid2 ,4-Dihydroxyphenyl n -undecyl ke tone
Methyl |3-4-methoxy-4 ' -xenoyIpropionateMethyl ^ -4-methoxy-3-xenoylpropiona tel ,5-Di-n-caproyl-2,4-dihydroxy benzenea-Phenyl-(3-3,4-dimethoxybenzoylpropionic
acid2-PheAyl-5-benzoylpyridineD i h y d r o c o d e i n o n eD i h y d r o h y d r o x y c o d e i n o n e
R.P.t DihydrohydroxythebainoneDihydrothebainone
Method
IIVI II I
I I I
IVIV
II
II
IV
I VI I IIV
I I I
I
I
I
IIIIII
IIIII
I
Yield *
70505079
9285
—83
—
80
965378
—
44
—
—
———2 0
5 8
—
9 1
0
—
—
4 2
Refer-ence t
824451 3 8
1 3 8
365
369295
8 339
417241
401
3 71387465
377
3 4 3
2 7 0
3 3 1
3 3 1
5 3
542 1 6
2 1 6
5 4
1 8 7
4 0
2 9
71368
71
* Qf yield reported as quantitative; G, yield reported as good; P, yield reported as poor. A dashindicates that the yield is not reported.
t Reference numbers refer to the bibliography on pp. 201-209.
* Q» yield reported as q u a n t i t a t i v e ; G, yield reported as good; P, yield reported aa poor. A dashindicates that the yield is not repo r t ed .
t Reference numbers refer to the bibl iography on pp. 201-209.t R.P., reduct ion product .
propionio acidKetolactone from tigogenin3-Ketobisnorallocholanicacidl-Myristyl-3,4-dimethoxybenzene
l-(3',4'-Dimethoxyphenyl)-tetradecanone-3
l-Myristyl-2,5-dimethoxybenzenel,4-Di-(p-ethoxybenzoyl)-butaneKeto acid from sarsasapogenin acetatel-Keto-6,7-dimethoxy-2-(3',4'-dimethoxy-benzyl)-3-methyl-l,2,3,4-tetrahydronaphtha-
' C l e m m e n s e n , Ber., 47, 681 (1914).4 M a j i ma a nd N a ka mura , Ber., 46 , 4089 (1913).6 Johnson and Hodge , J. A m. Chern. Soc, 35, 1014 (1913).6 M a j i ma a nd N a ka mura , Ber., 48, 1597 (1915).7 Maj ima and Tahara , Ber., 48, 1606 (1915).8 v. Auwers and Borsche, Ber., 48, 1716 (1915).9 Freund and Fleischer , Ann., 411, 14 (1916).
10 Johnson and Kohmann, J. Am. Chem. Soc, 36 , 1259 (1914)."Freund, Fle ischer , and Goffer je , Ann., 414, 1 (1917).12 Freund, Fleischer, and Gofferje, Ann., 414, 12 (1917).13 Freund, Fleischer, and Gofferje, Ann., 414, 26 (1917).14
Freund, Fleischer, and Gofferje, Ann., 414, 37 (1917).16 Freund, Fle ischer , and Stemmer, Ann., 414, 44 (1917).16 Borsche and Rosenkranz, Ber., 52 , 342 (1919).17 Windaus and Rahlen , Z. physiol. Chem., 101, 223 (1918).18 Borsche, Ber., 52 , 1353 (1919).19 v. Auwers, Ann., 419, 92 (1919).20 Fleischer and Wolff, Ber., 53, 925 (1920).21 v. Braun, Ki rschbaum, and Schuhmann, Ber., 53, 1155 (1920).22
Fle i sche r a n d Siefert , Ber., 5 3 , 1 2 5 5 (1920) .23
W i e l a n d a n d B o r s c h e , Z. physiol. Chem., 1 0 6 , 1 8 1 (1919) .24
Lang lo i s , Ann., chim., 1 2 , 2 6 5 (1919) .26
Fle i sche r , Ann., 4 2 2 , 2 3 1 (1921) .26 Fleischer^and Siefert, Ann., 4 2 2 , 272 (1921).27 Borsche and Roth, Ber., 54, 174 (1921).28 Windaus and S taden , Ber., 54, 1059 (1921).29 Mannich and Lowenhe im, Arch. Pharm., 258 , 295 (1920).30 Maj ima , Ber., 55, 191 (1922 ). '31 M a u t h n e r , / . prakt. Chem., 103, 391 (1922).32 Borsche, Nachr. kgl. Ges. Wins. G ottingen, Math.-physik. Klasse, 11, 188 (1920) [C. A.,
KroUpfeiffer and Schafer, Ber., 56 , 620 (1923).36 M a ye r a nd S t a mm, Ber., 56 , 1424 (1923).37 Ruzicka and Mingazaini , Helv. Chim. Acta, 5, 710 (1922).38 Steinkopf and Wolfram, Ann., 430, 113 (1923).39 Scbroeter, Ber., 54, 2242 (1921).40 Benary and Psil le, Ber., 57, 828 (1924).41 v. Auwers and Wit t ig, Ber., 57, 1270 (1924).42 Windaus and Bohne , Ann., 433, 278 (1923).43 Wisl icenus and Wei temeyer , Ann., 436, 1 (1924).44 Schroeter, Ber., 57, 2025 (1924).46 Hucke l and Goth , Ann., 441, 34 (1925).46
Hess and Bapper t , Ann., 441, 151 (1925).47 KroUpfeiffer, Schultze, Schumbohm, and Sommermeyer, Ber., 58, 1654 (1925).48 v. Braun and Reut t e r , Ber., 59, 1922 (1926).49 Wieland and Jacobi , Ber., 59, 2064 (1926).60 Wieland and Mar tz , Ber., 59, 2352 (1926).61 Ruzicka, Helv. Chim. Acta, 9, 1008 (1926).62 Shriner and Adams, J. Am. Chem . Soc, 17, 2727 (1925).
221Fieser and Hershberg, J. Am. Chem. Soc, 58, 2382 (1936).
222Fieser and Seligman, J. Am. Chem. Soc, 58, 2482 (1936).
223Tschesche and Bohle, Ber., 69, 2497 (1936).
224Reichstein, Helv. Chim. Ada, 19, 979 (1936).
226Ruzicka, Hofmann, and SoheUenberg, Helv. Chim. Ada, 19, 1391 (1936).
226
Robertson and Subramaniam, J. Chem. Soc, 278 (1937).227Bridge, Heyes, and Robertson, J. Chem. Soc, 279 (1937).
228King and L'Ecuyer, J. Chem. Soc, 427 (1937).
229Francis, King, and Willis, / . Chem. Soc, 999 (1937).
230Ali, Desai, Hunter, and Muhammad, J. Chem. Soc, 1013 (1937).
231Adelson and Bogert, J. Am. Chem. Soc, 59, 399 (1937).
232Bruce and Fieser, J. Am. Chem. Soc, 59, 479 (1937).
233Fieser and Seligman, J. Am. Chem. Soc, 59, 883 (1937).
234Strating and Backer, Rec trav. chim., 55, 903 (1936).
236Guha and Nath, Ber., 70, 931 (1937).
236Stenzl and Fichter, Helv. Chim. Acta, 20, 846 (1937).
237
I w a s a k i , Z. physiol. Chem., 2 4 4 , 181 (1936) [C. A . , 31, 1033 (1937)] .238Fujii and Matsukawa, / . Pharm. Soc. Japan, 56, 642 (1936) [C. A., 31, 1033 (1937)].
239K u r o d a andW a d a , Proc. Imp. Acad. (Tokyo), 12 , 23 9 (1936) [C . A . , 31, 1794 (1937)] .
240L i m a y e andG h a t e , Rasayanam, I, 39 (1936) [C.A . , 31 , 2182 (1937)] .
241Sengupte, Current Sci., 5, 295 (1936) [C. A., 31, 2587 (1937)].
242Desai and Wali, / . Univ. Bombay, 5, 73 (1936) [C. A., 31, 3038 (1937)].
243Tsukamoto, Ueno, and Ota, J. Pharm. Soc. Japan, 56, 931 (1936) [C. A., 31, 3493
(1937)].244
Kondo and Watanabe, J. Pharm. Soc. Japan, 54, 905 (1934) [C. A., 31, 104 (1937)].246
Kuwada and Matsukawa, J. Pharm. Soc. Japan, 54, 461 (1934) [C. A., 31, 108
(1937)].246
Shah and Laiwalla, Current Sci., 5, 197 (1936) [C. A., 31, 6219 (1937)].247Kazuno, J. Biochem. (Japan), 25, 251 (1937) [C. A., 31, 6669 (1937)].
248Sengupta, Science and Culture, 3, 57 (1937) [C. A., 31, 7866 (1937)].
249Sengupta, Science and Culture, 3, 56 (1937) [C. A., 31, 7868 (1937)].
250Clemo and Metcalfe, J. Chem. Soc, 1518 (1937).
261Bridge, Crocker, Cubin, and Robertson, J. Chem. Soc, 1530 (1937).
262George and Robertson, J. Chem. Soc, 1535 (1937).
263Bell, Bridge, and Robertson, / . Chem. Soc, 1542 (1937).
264Hill, Short, and Stromberg, J. Chem. Soc, 1619 (1937).
266Hill, Short, and Stromberg, J. Chem. Soc, 937 (1937).
266France, Maitland, and Tucker, / . Chem. Soc, 1739 (1937).
27 6A t k i n s o n a n d H a w o r t h , J. Chem. Soc, 1 6 8 1 (1938) .
276 Clemo, Cook, and Raper , J. Chem. Soc, 1318 (1938).27 7 Bergel , Jacob, Todd, and Work, J. Chem. Soc, 1375 (1938).27 8 Shah and Laiwalla, J. Chem. Soc, 1828 (1938).279 Jones and Ram age , / . Chem. Soc, 1853 (1938).28 0 Cruickshank and Robinson, J. Chem. Soc, 2064 (1938).28 1
Kawasaki , J. Pharm . Soc. Japan, 6 7, 736 (1937) [C. A., 32 , 188 (1938)].28 2 Kawasaki , J. P harm. Soc. Japan, 57, 742 (1937) [C. A., 32 , 188 (1938)].28 3 Desai and Wal i , Proc Indian Acad. Sci., 6A, 135 (1937) [C. A., 32 , 508 (1938)].28 4 Desai and Wal i , Proc. Indian Acad. Sci., 6A, 144 (1937) [C. A., 32 , 509 (1938)].286 Auterinen, Suomen Kemistilehti, 10B, 22 (1937) [C. A., 32 , 509 (1938)].28 6 Guha and Subramanian , Ber., 70, 2228 (1937).28 7
Steinkopf, Poulsson, and Herdey, Ann., 536 , 128 (1938).88 8 Dippy and Lewis , Rec. trav. chim., 56 , 1000 (1937).289
L e u c h s , Ber., 7 0 , 2455 (1937) .29 0 Ruzicka and Hofmann, Helv. Chim. Acta, 2 0, 1155 (1937).29 1
L i m a y e a n d L i m a y e , Rasayanam, I , 109 (1937) [C . A . , 3 2 , 2695 (1938)] .292
L e v y , Ann. chim., 9 , 5 (1938) [C . A . , 3 2 , 2926 (1938)] .29 3R e j o n o w s k i a n d S u s z k o , Arch. Chem. Farm., 3 , 1 3 5 (1937) [C . A . , 3 2 , 2939 (1938)] .
29 4 A w e a n d U n g e r , Ber., 7 0 , 4 7 2 (1937) .29 5 Burger and Moset t ig, J. Am. Chem. Soc, 59, 1302 (1937).296 Bergmann and B lum-Bergmann, J. Am. Chem. Soc, 59, 1441 (1937).29 7 Bergmann and B lum-Bergmann, J. Am. Chem . Soc, 59, 1572 (1937).29 8 Haberland and Kleinert , Ber., 71, 470 (1938).29 9 Spa th and 'Ga l inovsky, Ber., 71, 721 (1938).30 0 Beyer, Ber., 71 , 915 (1938).30 1 Wahl , Compt. rend., 2 0 6 , 683 (1938).302 Steiger and Reichstein, Hek. Chim. Acta, 2 1 , 161 (1938).30 3 Steiger and Reichstein, Helv. Chim. Acta, 2 1 , 828 (1938).304 Wanag and Walbe , Ber., 71, 1448 (1938).306 Weidlich, Ber., 71 , 1601 (1938).80 6 Sengupta , J. prakt. Chem., 151, 82 (1938).307 N e w m a n , / . Am. Chem. Soc, 6 0, 2947 (1938).80 8 Fieser and Jacobsen, J. Am. Chem. Soc, 6 0, 2761 (1938).309 Fieser and Jacobsen, J. Am. Chem. Soc, 6 0, 2753 (1938).31 0 T omi t a , J. pharm. Soc. Japan, 68, 510 (1938) [C. A., 32 , 7467 (1938)].81 1 Chowdhry and Desa i , Proc. Indian Acad. Sci., 8A, 1 (1938) [C. A., 32 , 9065 (1938)].31 2 Y a na g i t a , Ber., 71 , 2269 (1938).31 3 Haberland, Kleinert , and Siegert , Ber., 71, 2623 (1938).31 4 John, Gi inther , and Schmei l , Ber., 71, 2637 (1938).31 6 Werder and Jung, Ber., 71, 2650 (1938).31 6 Kondo and Keimatsu , Ber., 71, 2553 (1938).81 7 Bruce and Todd, J. Am . C hem. Soc, 6 1, 157 (1939).318 Hurd and Fowler , J. Am. Chem. Soc, 6 1 , 249 (1939).819 Smith and Kiess , / . Am. Chem. Soc, 6 1 , 284 (1939).32 0 Gilman and Turck , J. Am. Chem. Soc, 6 1 , 478 (1939).32 1 Fieser and Fieser, J. Am. Chem . Soc, 6 1 , 596 (1939).322 Crawford, J. Am. Chem. Soc, 6 1 , 608 (1939).828 M a rke r a nd R ohrma nn , J. Am. Chem . Soc, 6 1 , 846 (1939).324 Marker and Lawson, J. Am. Chem. Soc, 6 1 , 852 (1939).326 Fieser and Kilmer, J. Am. Chem. Soc, 6 1 , 862 (1939).328 M a rke r a nd R ohrma nn , J. Am. Chem . Soc, 6 1 , 946 (1939).327 B a c hma nn , J. Org. Chem., 3, 434 (1938).32 8 Windaus and Raichle , Ann., 537, 157 (1938).329 Penfold and Simonsen, J. Chem. Soc, 87 (1939).330 Shah and Shah, J. Chem. Soc, 132 (1939).
38 7 Richardson and Reid, J. Am. Chem. Soc, 62, 413 (1940).388 Fernholz, Ansbacher, and MacPhillamy, J. Am. Chem. Soc, 62, 430 (1940).38 9 Marker and Rohrmann, / . Am. Chem. Soc, 62, 896 (1940).39 0 Russell, Frye, and Mauldin, J. Am. Chem. Soc, 62, 1441 (1940).3 9 1 Fieser and Cason, J. Am. Chem. Soc, 62, 1293 (1940).3 92 Todd, Harris, and Noller, J. Am. Chem. Soc, 62, 1624 (1940).3 9 3
Kloetzel, / . Am. Chem. Soc, 62, 1708 (1940).3 9 4 Fieser and Novello, / . Am. Chem. Soc, 62, 1855 (1940).3 9 6 Fieser and Bowen, J. Am. Chem. Soc, 62, 2103 (1940).3 9 6 Adams, Cain, and Baker, J. Am. Chem. Soc, 62, 2201 (1940).3 9 7 Bachmann and Edgerton, / . Am. Chem. Soc, 62, 2219 (1940).398 Harris and Pierce, J. Am. Chem. Soc, 62, 2223 (1940).3 9 9 Bachmann and Edgerton, J. Am. Chem. Soc, 62, 2550 (1940).4 0 0
Marvel, Pearson, and Patterson, / . Am. Chem. Soc, 62, 2659 (1940).4 0 1 Tarbell and Weaver, J. Am. Chem. Soc, 62, 2747 (1940).4 0 2 Bachmann and Holmes, J. Am. Chem. Soc, 62, 2750 (1940).4 0 3 Fieser, Gates, and Kilmer, J. Am. Chem. Soc, 62, 2966 (1940).4 0 4
Bachmann and Edgerton, J. Am. Chem. Soc, 62, 2970 (1940).4 0 5 Marker, Jones, Turner, and Rohrmann, J. Am. Chem. Soc, 62, 3006 (1940).40 6 Chakravarti and Chakravarty, J. Indian Chem. Soc, 16, 144 (1939) [C. A., 34, 401
(1940)].40 7
Kurauti and Kazuno, Z. physiol. Chem., 262, 53 (1939) [C. A., 34, 1327 (1940)].4 0 8 Chien and Yin, J. Chinese Chem. Soc, 7, 40 (1939) [C. A., 34, 1979 (1940)].4 09 Komppa, Ann. Acad. Sci. Fennicae, A51, No. 3 (1938) [C. A., 34, 2335 (1940)].4 1 0
Sengupta, / . Indian Chem. Soc, 16, 349 (1939) [C. A., 34, 3242 (1940)].4 1 1 Rochelmeyer, Arch. Pharm., 277, 340 (1939) [C. A., 34, 3276 (1940)].4 12 Banerjee, Science and Culture, 5, 566 (1940) [C. A., 34, 4383 (1940)].4 1 3 Bohme, Arch. Pharm., 278, 1 (1940) [C. A., 34, 5427 (1940)].4 1 4
Guha and Hazra, J. Indian Chem. Soc, 17, 107 (1940) [C. A., 34, 5427 (1940)].4 1 6 Desai and Kamal, Proc Indian Acad. Sci., 11A, 139 (1940) [C. A., 34, 5435 (1940)].4 1 6 Akram and Desai, Proc. Indian Acad. Sci., 11A, 149 (1940) [C. A., 34, 5436 (1940)].4 1 7 Sengupta, J. Indian Chem. Soc, 17, 101 (1940) [C. A., 34, 5439 (1940)].4 1 8 Tamamusi, J. Pharm. Soc Japan, 60, 189 (1940) [C. A., 34, 5446 (1940)].4 1 9 Lions and Willison, J. Proc. Roy. Soc N.S. Wales, 73, 240 (1940) [C. A., 34, 5841
(1940)].4 20 Hukui and Tikamori, J. Pharm. Soc. Japan, 59, 158 (1939) [C. A., 34, 5848 (1940)].4 2 1 Sengupta, J. Indian Chem. Soc, 17, 183 (1940) [C. A., 34, 6609 (1940)].4 2 2 Miyasaka and Nomura, / . Pharm. Soc. Japan, 60, 328 (1940) [C. A., 34, 7289 (1940)].4 2 3 Vanghelovici, Bull, soc chim. Romania, 19A, 35 (1937) [C. A., 33, 639 (1939)].4 2 4
Limaye and Limaye, Rasayanam Suppl. (1938) [C. A., 33, 1698 (1939)].4 2 6 Dziewonski and Marusinska, Bull, intern, acad. polon. sci., Classe sci. math, not.,
1938A, 316 [C. A., 33, 1712 (1939)].4 2 6 Kurado and Wada, Sci. Papers Inst. Phys. Chem. Research (Tokyo), 34, 1740 (1938)
[C. A., 33, 2511 (1939)].4 2 7 Desai and Ekhlas, Proc. Indian Acad. Sci., 8A, 567 (1938) [C. A., 33, 3356 (1939)].4 2 8
Nazarova, J. Gen. Chem. (U.S.S.R.), 8, 1336 (1938) [C. A., 33, 4214 (1939)].4 2 9 Chatterjee and Barpujari, J. Indian Chem. Soc, 15, 639 (1938) [C. A., 33, 4586
(1939)].4 3 0 Huzii and Tikamori, / . Pharm. Soc. Japan, 59, 116 (1939) [C. A., 33, 4592 (1939)].4 3 1 Huzii and Tikamori, J. Pharm. Soc. Japan, 59, 124 (1939) [C. A., 33, 4592 (1939)].43 2
Mitter and De, J. Indian Chem. Soc, 16, 35 (1939) [C. A., 33, 5838 (1939)].4 3 3 Sengupta, J. Indian Chem. Soc, 16, 89 (1939) [C. A., 33, 5842 (1939)].4 3 1 Ohla, Z. physiol. Chem., 259, 53 (1939) [C. A., 33, 6327 (1939)].43 6 Harradence and Lions, J. Proc. Roy. Soc. N.S. Wales, 72, 284 (1939) [C. A., 33,
438 Harradence, Hughes , and Lions , J. PTOC. R oy. Soc. N.S. Wales, 72 , 273 (1939) [C .A., 33, 6841 (1939)].
43 7 Mit te r and De , J. Indian Chem . Soc, 16, 199 (1939) [C. A., 33, 7761 (1939)].43 8 Chat ter jee , J. Indian Chem . Soc, 14, 259 (1937) [C. A., 32 , 123 (1938)].439 Boki l and Nargund, J. Univ. Bombay, 6, pt. II, 93 (1937) [C. A., 32 , 3759 (1938)].44 0 Pet rov and Lapteva , J. Gen. Chem. (U.S.S.B.), 8, 207 (1938) [C. A., 32 , 5386 (1938)].44 1 Katsne l ' son and Kondakova , Compt. rend. acad. sci. U.R.S.S., 17 , 367 (1937) [C.
A., 32, 7023 (1938)].442 Chat ter jee , J. Ind ian Chem . Soc, 15, 211 (1938) [C. A., 32 , 7447 (1938)].4 4 3 Heilbron, Kennedy, and Spring, J. Chem. Soc, 329 (1938).44 4 R a ma ge , J. Chem . Soc, 397 (1938).44 6 Cook and Robinson, J. Chem. Soc, 505 (1938).44 6 Err ington and Lins tead , J. Chem. Soc, 666 (1938).44 7 B l um-B e rgma nn , / . Chem. Soc, 723 (1938).44 8 Hawor th and Atkinson , J. Chem. Soc, 797 (1938).449 Hawor th and Woodcock, J. Chem. Soc, 809 (1938).46 0 Mikeska, Smith, and Lieber , J. Org. Chem., 2, 499 (1938).45 1 Rice and H arden , / . Am. Pharm. Assoc, 2 5, 7 (1936) [C. A., 30, 2559 (1936)].46 2 Mit te r and De , J. Indian Chem . Soc, 12 , 747 (1935) [C. A., 30, 2562 (1936)].46 3
S h i m i z u a n d K a z u n o , Z. physiol. Chem., 2 3 9 , 6 7 (1936) [C . A . , 3 0 , 2984 (1936)] .46 4 Shah and Mehta , J. Univ. Bom bay, 4, 109 (1935) [C. A., 30, 5196 (1936)].466 Shah and Mehta , J. Indian Chem . Soc, 13, 358 (1936) [C. A., 30, 8187 (1936)].46 6 K uw a da , / . Pharm. Soc. Japan, 56, 469 (1936) [C. A., 30, 8237 (1936)].46 7 H a r t a nd Woodruff, J. Am. Chem . Soc, 58, 1957 (1936).45 8 Fieser and Lothrop, J. Am. Chem . Soc, 58, 2050 (1936).469 Shimizu and Oda, Z. physiol. Chem., 2 2 7 , 74 (1934) [C. A., 2 9, 174 (1935)].46 0 Ochiai and Hakozaki , J. Pharm. Soc Japan, 50, 360 (1930) [C. A., 24, 3793 (1930)].46 1 K a z i r o , Z. physiol. Chem., 1 8 5 , 1 5 1 (1929) [C . A . , 2 4 , 8 5 9 (1930) ] .462
B o r s c h e , Ber., 5 2 , 1 3 6 3 (1919) .46 3
W i e l a n d a n d Sch l i ch t i ng , Z. physiol. Chem., 1 5 0 , 2 6 7 (1925) .46 4 Overbaugh, Al len, Mart in, and Fieser , Org. Syntheses, 15, 64 (1935).46 5 M a r t in , / . Am. Chem. Soc, 58, 1438 (1936).46 6 M a r t i n , Org. Syntheses, 17, 97 (1937).46 7 T homps on , J. Chem. Soc, 2314 (1932).46 8 Fieser and Kennelly, J. Am. Chem. Soc, 57, 1611 (1935).469 Wi nda us , Ber., 53, 488 (1920).47 0 Bardhan and Sengupta , J. Chem. Soc, 2520 (1932).47 1 Ju, Shen, and Wood, / . Inst. Pelr., 26, 514 (1940).47 2 Heath-Brown, Hei lbron, and Jones , J. Chem. Soc, 1482 (1940).47 3 Bi lham and Kon, / . Chem. Soc, 1469 (1940).4 7 4Phi l ippi and Rie , Monatsh., 42, 5 (1921).47 6 Fieser and Desreux, J. Am. Chem . Soc, 60, 2255 (1938).47 6 Fieser and Peters , J. Am . Chem . Soc, 54, 4373 (1932).
In 1868 W. H. Perkin 1 described a synthesis of coumarin by heatingthe sodium salt of salicylaldehyde with acetic anhydrid e. Fu rth er s tudy
of this reaction led to the discovery of a new method for preparingcinnamic acid and its analogs by means of a synthesis of very generalapplication, which became known as the Perk in reaction.2 This reactionis brought about by heating an aldehyde of aromatic type with theanhydride of an aliphatic acid of the general formula RCH2CO2H, inthe presence of th e sodium sa lt of the acid.
Since the resulting /3-arylacrylic acids can be subjected to a variety ofchemical transformations, the P erkin reaction gives access indirectly to anumber of other types of compounds such as arylethylenes and acety-lenes, arylacetaldehydes, arylethylamines, arylpropionic and propiolicacids, and their der ivativ es. Several modifications and extensions of th e
Perkin reaction, such as the paraconic acid synthesis of Fittig and theazlactone synthesis of Erlenmeyer, have served to broaden the scopeand usefulness of the original process.
In the course of an extensive study of unsaturated acids Fittig 3 andhis collaborators made several importan t con tributions to the mechanismof the Perk in reaction. He showed th at th e aldehyde condenses with thealpha methylene group of the acid component (salt or anhydride) andconcluded that the reaction is an addition process, like an aldol conden-
sation, involving an intermediate /J-hydroxy compound that loses waterto form the a,/3-unsaturated acid.
C 6H BCH=O + (CH3CH 2CO)2O + CH3CH2CO 2Na -»
rC6H6CHOHCHCO2H~| -* C6H 6CH=CCO 2H
L C H S J C H 3
Perkin 2 had assumed, without experimental proof, that the carbon atomfarthest removed from the carboxyl group was probably the one which
1 Perkin, J. Chem. Soc, 21, 53, 181 (1868).2 Perkin, J. Chem. Soc, 31 , 388 (1877),' Fittig, A n n . , 1 9 5 , 169 1 8 7 9 ) ; 2 1 6 , 97 ( 1 8 8 3 ) ; 2 2 T , 48 ( 1 8 8 5 ) ; B e r . , 1 4 , 1 8 2 4 ( 1 8 8 1 ) ;
condenses with the a ldehyde, but Fi t t ig and others , 4> 6 quickly dis-proved Perkin ' s tenta t ive hypothes is .*
The view that the Perkin react ion involves an intermediate addit ionpro du ct of th e aldol ty pe is generally accepted at the presen t t im e. I t issupported by the actual isolation of derivatives of the intermediate
addition products in certain cases where the normal elimination of waterdoes no t occur. Fo r exam ple, benzald ehyd e, sodium isob uty rate, andisobutyric anhydride (or acetic anhydride) on heating at 100° give riseto the isobutyryl derivative of ;3-phenyl-/3-hydroxypivalic acid (<x,a-
dimethyl-/3-hydroxy-/S-phenylpropionic acid) 6 and the mixed anhydrideof this acid with isobutyric acid.
C 6 H 6 CH =O + (CH3)2CHCOOCOCH(CH 3)2 + (CH 3)2CHCO 2Na ->
C 6H BCH—C(CH 3)2CO 2H and C 6H 6CH—C(CH 3)2COOCOC 3H 7
OCOC 3H 7 OCOC3H 7
The total yield, calculated as /3-phenyl-|S-hydroxypivalic acid, is about3 3 % of the theoretical.7 A t 150° the same reac tants give the unsa tu-
ra ted hydrocarbon, 2-methyl- l-phenylpropene,8 which is formed pre-sum ably from th e abov e interm ediates by loss of carb on dioxide a nd
isobutyr ic acid (or anhydride) .
C 6H BCH—C(CH 3)2CO 2H -> C 6 H 5 C H = C ( C H 3 ) 2 + CO 2 + C 3H 7CO 2H
OCOC3H7
Likewise, furfural on heating with isobutyric anhydride and sodiumisobuty rate gives only 2-methyl- l- furylpropene,9 even a t temperatures aslow as 100°.
In typical examples of the Perkin reaction, involving derivatives ofacetic acid or monosubstituted acetic acids, decarboxylation has beenobserved in a few instances, notably with isovaleric acid.10 This s ide
* For an interesting account of early work on the Perkin reaction see Lachmann, "TheSpirit of Organic Chem istry," The M acmillan Co., London (1899), pp . 12-20; also, Cohen,"Organic Chem istry for Advanced Students," fifth edition, Longmans, Green and Co., NewYork (1928), Part I , pp . 288-293. An excellent review of recent work is given by W atson,Ann. R epls. Chem. Soc. (Lond on), 36, 210 (1939).
4 Baeyer and Jackson, Ber., 13, 115 (1880).6 Conrad and Bisehoff, Ann., 204, 183 (1880).6 Fittig and Jayne, Ann., 216, 115 (1883); Fittig and Ott, Ann., 227, 119 (1885).7Hauser and Breslow, J. Am . Chem . Soc, 6 1, 793 (1939).8 Perkin, J. Chem . Soc, 35, 138 (1879).9 Baeyer and Tonnies, Ber., 10, 1364 (1877).
10 Schaarschmidt, Georgeacopol, and Herzenberg, Ber., 51, 1059 (1918).
reaction is generally negligible at the temperatures usually employed(140-175°) bu t m ay become im po rtan t a t higher tem pe ratu res . Th us ,anisaldehyde on heat ing a t 170° with propionic anhydride and sodiumpropionate yields mainly p-anisyl-a-methylacrylic acid,11 but a t 200°
anethole (p-propenylanisole) is obtained.12
Further evidence for the formation of an intermediate of the aldol
type is afforded by the reaction of benzaldehyde with succinic anhydride(or acetic anhyd ride) and sodium succinate . Fi t t i g and Jay ne 13 showedthat if the reaction is carried out at 100° the product is 7-phenylpara-
conic acid, formed by lactonization of the intermediate hydroxy acid.
C«H BCH—CHCO2H 100o C 6H 6CH
OH CH 2— C 0 2H O\
CO7-Phenylparaconic acid
C 6H B C H = C H C H 2C O 2H + CO2PhenyiiBocrotonic acid
On heating to 150°, 7-phenylparaconic acid loses carbon dioxide andgives the /3,7-unsaturated acid, phenylisocrotonic acid, which Perkin
had obtained directly by carrying out the original condensation at 150°.The relative significance of the acid anhydride and the sodium salt in
the intimate mechanism of the Perkin condensation has been the subjectof numerous investigations extending over a period of more than fiftyyears. Perkin 2 believed th a t the cinnamic acids are formed by condensa-tion between the aldehyde and the acid anhydride, with the sodium saltfunctioning as a cataly st. He found tha t cinnamic acid is formed alonewhen benzaldehyde and acetic anhydride are heated at 180° with sodium
acetate, butyrate or valerate, whereas benzaldehyde on heating withpropionic anhydride and sodium acetate gives only a-methylcinnamicacid. Fittig 3 then studied the reaction with several anhydride-saltcombinations, particularly at lower tem pera tures. He found tha tbenzaldehyde, acetic anhydride, and sodium acetate (in equimolecularamounts) do not react at 100° even on long-continued heating; whensodium n-butyrate was used in place of the acetate, reaction occurredslowly and only a-ethylcinnamic acid was formed, but at 150° a mixturecontaining one part of a-ethylcinnamic to two parts of cinnamic acidwas obtained, and at 180° the product contained only one par t of
11 Perkin , J. Chem. S oc , 31 , 415 (18 77); 32, 669 (1878).12 Moureu and Chauve t , Bull. soc. chim ., [3] 17, 412 (189 7); M ou reu , Ann. chim., [7]
15, 135 (1898).13 Fi t t ig and Jayne , Ann., 2 1 6 , 100 (1883).
a-ethylcinnamic to ten parts of cinnamic acid.14 From these results F ittigconcluded that at 100° the reaction occurs between the aldehyde and thesalt, and explained the formation of cinnamic acid at 150° and 180° byassuming that an anhydride-salt exchange occurs at the higher tempera-
tures (but not at 100°), giving rise to sodium acetate, which then reactswith the aldehyde. Fi ttig considered, therefore, tha t the observedbehavior of anhydride-salt combinations was compatible only with theview that the aldehyde always condenses with the salt.
Michael15 was led by theoretical considerations to doubt the validityof Fittig's views and brought forward strong experimental evidence infavor of Perkin's contention that the condensation occurs between thealdehyde and anhydride. Michael and Ha rtm an 16 showed that theanhydride-salt exchange postulated by Fittig occurs rapidly even at100° and the position of equilibrium is very far on the side of the higheranhydride and sodium aceta te. They found th a t acetic anhydride onheating with sodium butyrate or caproate for a short time at 100° gaveexcellent yields of butyr ic or caproic anhydride, while buty ric anhydrideand sodium acetate do not react appreciably under these conditions.
These results show unmistakably that, in Fittig's experiment withbenzaldehyde, acetic anhydride, and sodium butyrate at 100°, the reac-tion mixture must have contained considerable butyric anhydride andsodium acetate and but little acetic anhydride and sodium butyrate.Consequently, the formation of a-ethylcinnamic acid as the main productunder these conditions affords strong evidence that the reaction proceedsbetween the aldehyde and the anhydride. Recen tly Breslow andH a u s e r n found th a t the same relative quantities of cinnamic anda-ethylcinnamic acids were formed when benzaldehyde was condensedwith a mixture of either acetic anhydride and sodium butyrate, orbutyric anhydride and sodium acetate, which had in each case beenheated previously for several hours to establish equilibrium. * At 100°the product contained about 80% a-ethylcinnamic and 20% cinnamicacid; at 180° there is a larger proportion of acetic anhydride in the reac-
* In these experiments the original anhydride-salt mixtures contained one mole ofanhydride to two moles of the salt; in Fittig's and Michael's experiments the anhydrideand salt were used in equimolecular quantities, and in Perkin's less than one-half mole of
salt was used per mole of anhydride.14 Fittig and Slocum, Ann., 227, 53 (1885).15 Michael, J. prakt. Chem ., [2] 60, 364 (1899).16 Michael and Hartman, Ber., 34, 918 (1901); see also Michael, Am. Chem . J., 50,
tion mixture and the product is made up of about 30% of a-ethylcin-namic and 70% cinnamic acid.
Fittig's view that the salt condenses with the aldehyde appeared to bestrongly supported by Stuart's observation n th at benzaldehyde, sodium
malonate, and acetic anhydride react at room temperature with evolu-tion of carbon dioxide and formation of cinnamic acid. Fi ttig regardedthis as a convincing proof of his view since he believed that malonic acidwas incapable of forming an anhydride and the reaction must haveoccurred between the aldehyde and sodium malonate. Michael pointedout that this argument also is not valid since a mixed anhydride ofmalonic and acetic acid could be formed and, in any event, malonic acidis much more reactive in condensation reactions than the anhydrides or
salts of monobasic acids. This view is confirmed by recent work17
whichhas shown tha t sodium malonate does not react w ith benzaldehyde to anyappreciable extent unless glacial acetic acid is present.
In spite of Michael's objections, Fittig's interpretation was widelyaccepted for many years and still persists in several of the current text-books of organic chem istry. However, the results of a num ber of work-ers now provide substantial evidence in favor of Perkin's and Michael'sview that it is the anhydride and not the salt that undergoes condensa-
tion with the aldehyde. Kalnin
18
has shown that benzaldehyde con-denses readily with acetic anhydride in the presence of inorganic andorganic bases (potassium carbonate, triethylamine, etc.) but does notcondense with sodium acetate in the presence of these catalysts (or inthe presence of inorganic dehydrating agents 19). These and otherresults 20 '21 indicate that the Perkin reaction is essentially an aldol con-densation of the aldehyde and anhydride, in which the salt of the acidfunctions merely as a base and promotes enolization of the anhydride.In this connection it is of interest to note that ketene, which may beregarded as an intramolecular anhydride of acetic acid, reacts readily at25° with benzaldehyde in the presence of potassium acetate to give amixed anhydride of cinnamic and acetic acids, along with styrene.22
CH aCO2K
C6H6CHO + 2CH2=C=O - > C6H6CH=CHCOOCOCH3
C6HBCH0 + CH 2 = C = O > C6HBCH=CH2 + CO2
This reaction does not take place with tributylamine in place of potas-
sium acetate , and with small amounts of the latte r (0.1 mole per mole of1 8 K a l n i n , Heh. Chim. Acta, 1 1 , 9 7 7 ( 1 9 2 8 ) .19 B a k u n i n a n d P e o c e r i l l o , Oazz. chim. ilal., 6 5 , 1 1 4 5 ( 1 9 3 5 ) .2 0 K u h n a n d I s h i k aw a , Ber., 6 4 , 2 3 4 7 ( 1 9 3 1 ) .2 1 M i l l l e r , Ann., 4 9 1 , 2 5 1 ( 1 9 3 1 ) .22Hurd and Thomas, / . Am. Chem. Soc, 55, 275 (1933).
temperatu re. This reaction is analogous to the paraconic acid synthesesinvolving succinic anhydride, and leads eventually to a lactone-acid.
+ C6H6CHO (followed by H2O, and H+)
CHOHC6HB
Likewise, Hauser and Breslow 7 have shown that benzaldehyde reactsinstantly at room temperature with the sodium enolates of ethyl acetateand isobutyrate to form /3-phenyl-/3-hydroxy esters.
SCOPE OF THE REACTION
The Perkin reaction may be regarded essentially as the condensationof a carbonyl component A and an acid anhydride-salt combination B.In the resulting acrylic acids, substituents in the carbonyl componentappear in the /3-position and those in the acid component appear in thea-position.
R C H = O + ( C H3CO)2O -> RCH=CHCO2HA B /3-Substituted acrylic acids
RCH=O+ (R 'CH2CO)2O -> RCH=CCO2H
R'a,0-Disubstituted acrylic acids
The following discussion gives a survey of the types of carbonyl com-ponents and acid anhydride-salt combinations that can be used, and ofthe yields that can be obtained under favorable conditions.
Carbonyl Components
In general the usual Perkin reaction is limited for practical purposes toaldehydes of the aromatic series and closely related types. Table I givesa brief survey of the yields of /J-arylacrylic acids obtained from varioussubstituted benzaldehydes, with acetic anhydride and sodium acetate,
under similar conditions of reaction.27 '28 '29The yields given are
typical but do not always represent the maximum that can be securedwith a given aldehyde, as theoptimum conditions of reaction (tempera-ture, duration ofheating, catalytic effects, etc.) vary somewhat for differ-ent substituents.
TABLE I
Y I E L D S OF CINNAMIC ACIDS FBOM SUBSTITUTED BENZALDEHYDES "
Substituent
None "
2-Methyl «
3-Methyl "4-Methyl»
2,6-Dimethyl "
2-Iodo2»
2-Chloro "
3-Chloro «
4-Chloro "
2,6-Dichloro "
Yield
(per cent)
45-50 b
15
2333
0
85
71
63
52
82
Substituent
2-Methoxy»»
2,5-Dimethoxy 31
4-Methoxy *°4-Ethoxy
32
4-Hydroxy30
4-DimethylaininoM
2-Nitro »'
3-Nitro «
4-Nitro "
2,4-Dinitro »
Yield
(per cent)
55
56
3036
62
0
75
75
82
70°
° The conditions were very similar but not identical in all experiments. In general, 1 mole of the
aldehyde washeated for eight hours at 180°with about 2 moles of acetic anhydride and0.7 mole ofsodium acetate.& It has been reported s3 that the yield of cinnamic acid can be increased to 80-85% byadding a
little pyridine aB catalyst; this result could not be checked in the Cornell laboratory. The yield isincreased to 70-75% (without addition of pyridine) byheating for twenty-four hours.27
c This yield is obtained with eight hours' heating at 150°; with four hours' heating at 180° the yieldis about 20%, andlonger heating gives lower yields.
These results indicate that the activity of substituted benzaldehydesin thePerkin reaction is similar to thetrends observed inother reactionsinvolving the carbonyl group. Ahalogen ornitro group in anypositionincreases the rate of reaction and theyield; a methyl group in anyposi-tion decreases the rate and theyield, and this effect falls off in theorder:ortho > meta > para. A methoxyl group in the ortho position has asmall favorable influence, but in the para position it has a definitelyunfavorable effect on therate and theyield.
The behavior of ort/io-substituted benzaldehydes indicates that thereaction is notadversely affected unless the type of substituent isunfav-orable. Thus, 2,6-dichlorobenzaldehyde and 2,6-dinitrobenzaldehyde
2 8 Lock and Bayer , Ber., 72, 1064 (1939).29 Meyer and Beer, Monatsh., 34, 649 (1913).30
P o s n e r , J. prakt. Chem., [ 2] 8 2 , 4 2 5 (1910 ) .3 1 Kauffmann and Burr , Ber., 40, 2355 (1907).32 Stoermer, Ber., 61,2326 (1928).33 Baoharach and Brogan , J. Am. Chem. Soc, 50, 3333 (1928).
give excellent yields, but 2,6-dimethylbenzaldehyde and 2,4,6-trimeth-
ylbenzaldehyde do not react appreciably.27
Substituted benzaldehydes with hydroxyl groups in the meta or para
positions give satisfactory results. In the course of reaction the hydroxyl
group is acetylated and the product is the corresponding acetoxycinnamic
acid. The latter need not be isolated and can be saponified readily to
give the free hydroxy acid. Salicylaldehyde gives coumarin,1
the lactone
of the cis form of o-hydroxycinnamie acid (coumarinic acid), together
with the acetyl derivative of the trans form (coumaric acid).34
r r n pro oC H
°C
°3 N
(C H 3 C O ) 2 O
The action of alkalies on coumarin gives salts of coumarinic acid, but the
acid is unknown in the free state as it undergoes ring closure spontane-
ously to regenerate coumarin. Strong alkalies or alcoholic sodium
ethoxide convert coumarin into salts of coumaric acid, from which the
free acid can be obtained by acidification. Methylation of sodium
coumarate gives £rans-o-methoxycinnamic acid, which is identical with
the acid obtained from o-methoxybenzaldehyde in the Perkin reaction.The aminocinnamic acids are not prepared directly by the Perkin
reaction but are obtained by reduction of the corresponding nitrocin-
namic acids with ferrous sulfate and ammonia.36
The ordinary (stable)
form of o-nitrocinnamic acid, obtained from o-nitrobenzaldehyde in the
Perkin reaction, gives irans-o-aminocinnamic acid, which on long heat-
ing with hydrochloric acid is converted to carbostyril (the nitrogen
analog of coumarin).36
CHjjCH=CH—CO2H HCI ^ ^ y V
The aminocinnamic acids can be diazotized and subjected to the usual
diazonium replacement reactions; this method has served for the prepara-
tion of the chloro-, bromo-, and iodocinnamic acids,36
and 0- and p-
fiuorocinnamic acids.
37
M T i e ma nn and Herzfeld, Ber., 10, 285 (1877).36 Gabriel , Ber., 15, 2294 (1882); Gabriel and Herzberg , Ber., 16, 2038 (1883).86 Baeyer and Jackson, Ber., 13, 115 (1880); Tiemann, Ber., 13, 2069 (1880); Posner ,
Ann., 389, 45 (1912); Stoermer and H e y m a n n , Ber., 45, 3099 (1912)." G r i e s s , Ber., 18, 961 (1885); Kindler , Ann., 464, 278 (1928).
The Perkin reaction has been carried out with aldehydes of thebiphenyl38 and naph thalene series. 1-Naphthaldehyde 39 and 4-bromo-1-naphthaldehyde 40 react quite satisfactorily, but 2-naphthaldehyde39
gives only a small yield of /3-2-naphthylacrylic acid. 2-H ydroxy-l-naphthaldehyde gives a 30% yield of /3-naphthocoumarin.41
Furfural42 (and substituted 2-furanaldehydes) and 2-thiophenealde-hyde 43 take part readily in the Perkin reaction, but there appears to beno report of the use of aldehydes of the pyridine and quinoline series.The 2- and 4-pyridineacrylic acids, and the corresponding quinolinederivatives, are prepared conveniently by condensation of the 2- or 4-methyl derivative with chloral, followed by hydrolysis.44
The condensation of indole-3-aldehyde with hippuric acid in the presenceof acetic anhydride and sodium acetate (Erlenmeyer's azlactone syn-thesis) has been reported.45
Cinnamaldehyde, which is a vinylog of benzaldehyde, gives excellentyields of /3-styrylacrylic acids under the usual conditions of the Perkinreaction.2
C6H6CH=CHCHO + (CH3CO)2O 165°
C6H 6CH=CHCH=CHCO2H
On heating cinnamaldehyde with phenylacetic acid in the presence ofacetic anhydride and litharge, decarboxylation occurs and 1,4-diphenyl-butadiene is obtained in 30% yield.46
C6H6CH=CHCHO + C6H6CH 2CO 2H(C H 3CO )2O
C 6H 6CH=CHCH=CHC 6H 6
38 H e y , / . Chem. Soc, 2478 (1931); see also Vorlander, Ber., 68, 453 (1935), andreference 28, p. 1069.
Under the same conditions two moles of cinnamaldehyde react with one
of succinic acid to give 1,8-dip hen ylocta tetrene.46
Th e bifunct ional arom atic a ldehydes , phthala lde hyde ,4 7 isophthala lde-hyde ,4 8 and te rephtha la ldehyde ,4 9 ' 60 can be converted to the correspond-ing benzenediacrylic acids in 20, 80, an d 5 0 % yields, respectively.Under mild condit ions terephthaldehyde gives the monoacryl ic acid,
4-formylcinnamic acid;4 9 on prolonged heat ing a mixture of the mono-and di-acryl ic acids is obta ined (2 5% and 50 % yields, respect ively) .60
CHO
C 6H 41
CHO
CH=CHCO2H
—> Ce H 4 -
CHO
C H = C H C O 2 H
-> C 6H 4i
CH=CHCO 2H
2,2'-Biphenyldialdehyde gives an 8-9% yield of 2,2'-biphenyldiacrylicacid.61
4-Cyanobenzaldehyde 52 and 4-carboethoxybenzaldehyde 49 have beenconverted to the corresponding cinnamic acids, apparently in satis-factory yields. 2-Cyanocinnamic acid has been prepared throughCaro's modification of the Perkin reaction, by heating 2-cyanobenzal
chloride with acetic anhydride and sodium acetate.
63
Aliphatic aldehydes such as valeraldehyde and heptaldehyde givemainly condensation products when heated with acetic anhydride andsodium acetate, and only small amounts of the /3-alkylacrylic acids areformed.64 Acetaldehyde w ith propionic anhydride and sodium propion-ate (thirty hours at 120-130°) gives a small yield of tiglic acid, and iso-butyraldehyde with the same reagents (thirty hours at 190-200°) givesa 15-20% yield of isomeric 4-methylpentenoic acids.65 The reaction withsodium phenylacetate and acetic anhydride, sometimes called Oglialoro's
modification of the Perkin reaction,66 is somewhat more satisfactory;with these reagents paraldehyde gives a-phenylcrotonic acid (methyl-atropic acid).67
47 Thiele and Falk, Ann., 347, 117 (1906).48 Ruggl i and Staub, Helv. Chim. Ada, 17, 1523 (1934).48 Low, Ann., 231 , 375 (1885).6 0 E p h r a i m , Ber., 34, 2784 (1901)." Wei tzenbock, Monatsh., 34, 208 (1913).62 Moses , Ber., 33, 2625 (19 00); see also Shopp ee, / . Chem. Soc, 985 (1930).M
Drory , Ber., 2 4, 2574 (1891).64 Fit t ig and Schneegans , Ann., 2 2 7 , 79 (1885); Fit t ig and Hoffken, Ann., 304, 334(1899).
Although the Perkin reaction in its simplest form is quite unsatis-factory w ith aliphatic aldehydes, modifications involving the replacementof the monobasic acid components by succinic acid (Fittig's synthesisof paraconic acids and /3,7-unsaturated acids)13 ' 64 ' 68 and by m alonic acid(Doebner,69 Knoevenagel60) are useful preparative methods in the ali-phatic and aromatic series.
Simple aliphatic and aromatic ketones cannot be used as carbonylcomponents in the Perkin reaction, or in the paraconic acid synthesis.Acetone condenses with malonic acid in the presence of acetic anhy-dride,61 or ammonia,62 to give /3,/3-dimethylacrylic acid. The best resultsare obtained by Doebner's method using malonic acid and pyridine,which gives a 60% yield;63 under these conditions diethyl ketone gives/8,/3-diethylacrylic acid in 30% yield, but cyclohexanone gives less than
5% of the corresponding acrylic acid.a-Ketonic acids react with acetic anhydride and sodium acetate, with
loss of carbon dioxide, to give ^-substituted acrylic acids.64 Pyruvic acidgives crotonic acid, and arylglyoxylic acids give the correspondingcinnamic acids.
RC0C02H + (CH3CO)2OC H 3 C
°2 N a > RCH=CHCO2H + CO2 + CH3CO2H
Pyruvic acid reacts in a similar way with sodium succinate in the pres-ence of acetic anhydride, to form dimethylmaleic anhydride. Thesereactions have little preparative value as the same products can usuallybe obtained from more readily accessible reactants.
Michael and Gabriel made the remarkable discovery that phthalicanhydride may be used as the carbonyl component in a Perkin reaction.On heating phthalic anhydride with acetic anhydride and potassiumacetate, for ten minutes at 150-160°, phthalylacetic acid is formed in50% yield.66
HCC02H
^ ^CH3CO2K
(CH3CO)2O — i
68 Fi t t ig and Franke l , Ann., 255, 18 (1889); Fi t t ig a nd P ol i t io, Ann., 255, 293 (1889).M Doebner , Ber., 33, 2140 (1900); Ber., 35, 1137 (1902).60 Knoevenagel , Ber., 31 , 2598 (1898 ); Ger. pats . , 97,734, 156,560, 161,171 (Frdl. , 7,
736; 8, 1268).61
Massot , Ber., 27, 1225, 1574 (1894).62 Knoevenagel , Ger . pat . , 162,281 (Frdl. , 8, 1267).63 D u t t , J. Indian Chem . Soc, 1, 297 (1925); C. A., 19, 2475 (1925).64 Homolka , Ber., 18, 987 (1885); Claus and Wollner , Ber., 18, 1861 (1885).65 Gabriel and Michael, Ber., 10, 1554 (1877); Gabrie l and Neumann, Ber., 26, 952
This acid undergoes a number of interes t ing transformations ; on treat-m ent with sodium methox ide and subsequent w arming with hydrochlor ic
acid, carbon dioxide is evolved and 1,3-diketohydrindene is obtained.Cold aqueous alkalies open the lactone ring of phthalylacetic acid to
form 2-carboxybenzoylacetic acid, which loses carbon dioxide readily toyield 2-acetylbenzoic acid.
Phthalyl-acetic acid
OCH 3
>-CO2H+ C
°2
Phthal ic anhydride reacts with phenylacet ic acid and sodium aceta te ,
with evolution of carbon dioxide, to give benzalphthalide in 71-74%
yields.66
H C — C 6H 6
CO C
/ \ / \(~ \ TT r \ i f~ i T T /"ITT (~*r\ XT v Z"t XT f\
\ / \ /
CO CO
Benzalphthal ide is conver ted by sodium methoxide into l ,3-diketo-2-phenylhydrindene, 67 and by concentrated aqueous alkalies into 2-
phenac etylbenzoic acid. These transfo rm ations of th e ph thal ic an hy -dr ide condensat ion products are useful preparat ive methods .
Acid C omponents
Although the Perkin reaction is considered to occur with the acid
anhydride, there are numerous ins tances in which the resul t ing acryl icacid corresponds to the sa l t used and no t th e anhy dride. T hu s , sodiumphenylaceta te and acet ic anhydride react with benzaldehydes to produce
a-phenylcinnamic acids in excellent yields (Oglialoro's modification),86
and a-acylaminoacetic acids react with benzaldehydes in the presence ofacetic anh yd ride an d sodium ace tate at 100° to give deriv atives of
66 Gabrie l , Ber., 18, 3470 (1885); Weiss, Org. Syntheses, 13, 10 (1933).67 N a t ha ns on , Ber., 26, 2576 (1893); Eibner , Ber., 39, 2203 (1906).
a-acylaminocinnamic acids (azlactone synthes is).68 Owing to the exchangereactions that occur in mixtures of acids, salts, and anhydrides, even at100°, the product will depend primarily upon the relative active-methyl-ene reactivity of the various acid species present. For this reason it will
be convenient in the present discussion to refer merely to the acid com-ponent that undergoes reaction, without necessarily specifying whetherit is introduced as the free acid, salt, or anhydride.
The Perkin reaction is limited practically to acetic and monosubsti-tuted acetic acids, RCH 2CO 2H, as two a-hydrogen atoms must beeliminated to form the a,|8-unsaturation. Disubstituted acetic acidssuch as isobutyric acid give |8-hydroxy-a,a-dialkylpropionic acids, butthis reaction has little preparative significance as the Claisen or Reform-atsky reaction (p. 8) is usually more satisfactory for such compounds;
at higher temperatures the dialkylacetic acids yield dialkylstyrenes(p. 212). The presen t survey is restricted to monosubstituted aceticacids and related types, which are considered according to th e na ture ofthe substituent in the a-position.
Alkylacetic acids having a straight-chain alkyl substituent reactquite readily with benzaldehyde to give a-alkylcinnamic acids in satis-factory yields.2 Propionic, n-butyric,3 and n-caproic 16 anhydrides reactwith aromatic aldehydes at lower temperatures (100°) than acetic an-
hydride, and often give slightly higher yields. Palmitic anhydride andsodium palmitate are reported to give a 55% yield of a-n-tetradecylcin-namic acid.69
Isocaproic acid appears to react normally 16 to form a-isobutylcin-namic acid, but isovaleric acid gives very small yields of the a-isopropylderivatives.10 Even at temperatures as low as 70°, a mixture of valericanhydride, sodium valerate, and benzaldehyde evolves carbon dioxide,and isopropylstyrene is the m ain product; the same behavior occurs with
p-anisaldehyde and with furfural. The decarboxylation is believed tooccur at an intermediate stage since the a-isopropylacrylic acids, onceformed, are stable above 100°. Cycloalkylacetic acids apparen tly havenot been used in the Perkin reaction.
Crotonic anhydride , which is a vinylog of acetic anhydride, reacts withbenzaldehyde in the presence of triethylamine (but no t potassium cro-tonate) to give a-vinylcinnamic acid in 40% yield.20
C 6H 5CHO + (CH 3C H = C H C O )2O ^ > C eH 6C H = C C O 2H
C H = C H 2
68 Erlenmeyer , Ann., 27 1 , 164 (1892); 337, 265 (1904); see also Plochl, Ber., 16, 2815(1883).
This reaction can be carried out in 20-25 minutes a t 140°; the p roduct isl,4-diphenylbutadiene-2-carboxylic acid, which is structurally analogousto the acids obtained from crotonic and dimethylacrylic anhydrides(p. 224).
Malonic acid, owing to the powerful activating effect of two carboxylgroups on the same carbon atom, undergoes condensation with aliphaticand aromatic aldehydes under very mild conditions. In this particularcase it is likely th a t the acid itself (in the enol form) reacts with the alde-hyde, and the condensation can be effected satisfactorily under a widevariety of conditions. Undoubtedly the use of malonic acid is the bestand most general preparative method for /9-substituted acrylic acids.Until quite recently malonic acid has been a relatively expensive reagentand its use has been restricted largely to the less common arom atic alde-
hydes. It has also proved especially useful for aliphatic aldehydes andfor various arom atic aldehydes tha t give poor results in the simple Perkinreaction (alkyl-, alkoxy-, dimethylamino-benzaldehydes, etc.).
The condensation of malonic acid with various aliphatic aldehydes(paraldehyde,76 propionaldehyde,77 isobutyraldehyde,78 isovaleralde-hyde,79 etc.) was first effected in glacial acetic acid and a little aceticanhydride. Knoevenagel60 found that the reaction can be carried outwith much better results using ammonia or primary or secondary amines
(especially piperidine) as cataly sts. Unfortunately ne ither of these mod-ifications is a good preparative method in the aliphatic series as mixturesof a,0- and /3,7-unsaturated acids are obtained.80 The most satisfactorymethod in the aliphatic (and aromatic) series is Doebner's modification59
using pyridine, which has been studied by von Auwers.81 He found thatacetaldehyde gives exclusively crotonic acid (60% yield).
CH3CHO+ CH2(CO2H)2 °^'N> CH3CH=CHCO2H + CO2
Propionaldehyde gives almost pure a,/3-pentenoic acid, with only a traceof the /?,7-isomer; isobutyraldehyde, isovaleraldehyde, and n-heptalde-hyde give almost entirely a,|9-unsaturated acids. W ith n-heptaldehyde 80
the /3,7-unsaturated acid amounts to 5-10%; the latter can be removedby stirring with 85% sulfuric acid at 8O0,82 which converts it to the7-lactone (insoluble in sodium carbonate solution).
76Komnenos, Ann., 218, 149 (1883).
77Fittig, Ann., 283, 85 (1894).
78Braun, Monatsh., 17, 213 (1896).
79 Schryver, J. Chem. Soc, 63, 1331, 1334 (1893).80
Zaar, Ber. Schimmel & Co. Akt. Ges., Jubilee Number, 299 (1929); C. A., 2 4, 2107(1930).
81 von Auwers, Meisnner, Seydel, and Wisaebaoh, Ann., 432, 46 (1923).82
Shukow and Schestakow, J. R-uss. Phys. Chem. Soc, 40, 830 (1908); Chem. Zentr., I I ,1415 (1908).
that pyridine exerts a definite catalytic effect; substituted pyridine
bases differ quantitatively in their effectiveness.87
Bachmann and
Kloetzel88
have reported excellent yields of the corresponding acrylic
acids from o-chlorobenzaldehyde and from several phenanthraldehydes
(1-, 2-, 3-, and 10-), using malonic acid and a small amount of pyridine.
Fittig3 observed that sodium methylmalonate reacted with benzalde-
hyde in the presence of acetic anhydride to form a-methylcinnamic acid.
No doubt other alkyl- and aryl-malonic acids would react with benzalde-
hyde to give a-substituted cinnamic acids, but these reactions would be of
little preparative value.
Fittig discovered that aromatic13
and aliphaticM
aldehydes react
readily with sodium succinate and acetic anhydride at 100°, to give
7-phenyl- and 7-alkyl-paraconic acids (p. 213) in satisfactory yields.
These acids lose carbon dioxide on heating and form the /3,7-unsaturatedacids, together with a small amount of the 7-butyrolactone.
RCH|
0\ . /
CO
-CHCO2H
1CH 2
-> R C H = C H C H 2C O 2H + |0
\ /CO
CH 2|
C H 2
This reaction affords a useful extension of the Perkin synthesis and hasbeen used as a preparative method for a number of 7-substituted vinyl-
acetic acids.68
Methylsuccinic acid gives a mixture of the isomeric a,y-
and /3,7-disubstituted paraconic acids.89
Phenylsuccinic acid and benz-
aldehyde react at 125° to give /3,7-diphenylvinylacetic acid.90
(c) Anisalde hyd e: Proc. Indian Acad. Sci., 4A, 134 (1936); C. A., 30, 8149 (1936), Chem.Zentr., I, 2767 (1937).
(d) p-Hyd roxybenza ldehyde : Proc. Indian Acad. Sci., 4A, 140 (1936) ; C. A., 30, 8149
(1936), Chem. Zentr., I, 2768 (1937).
(e) m-Hydroxybenzaldehyde: Proc. Indian Acad. Sci., 4A, 144 (1936); C. A., 30, 8149
(1936), Chem. Zenlr., I, 2768 (1937).(/) o-Methoxy- and m-methoxybenza ldehyde : Proc. Indian Acad. Sci., 5A, 437 (1937);
C. A., 31, 7412 (1937), Chem. Zentr., II, 3313 (1937).(g) 2-Hyd roxy- l -naphthaIdehy de : Proc. Indian Acad. Sci., 6A, 181 (1937); C. A., 32,
1260 (1938), Chem. Zentr., I, 1356 (1938).(h) 2,4-Dihydroxybenzaldehyde: Proc. Indian Acad. Sci., 7A, 381 (1938); C. A., 32,
7435 (1938), Chem. Zentr., II, 2736 (1938).( i ) p-Tolualdehyde: Proc. Indian Acad. Sci., 9A, 508 (1939); C. A., 33, 8589 (1939).(/) 3,4-Dihydroxy-, 3-methoxy-4-hydroxy-, and 3,4-dimethoxybenzaldehyde: Proc.
Indian Acad. Sci., 9A, 511 (1939) ; C. A., 33, 8589 (1939), Brit. Chem. Abstracts, All, 478
(1939).88 B a c hma nn , J. Org. Chem., 3, 444 (1938); Bach man n and Kloetzel , J. Am. Chem. Soc,
59, 2209 (1937).89 Fit t ig, Ann., 255, 5, 7, 108, 126, 257 (1889).90 Fichte r and L a t zko , J. prakt. Chem., [2] 74, 330 (1906).
nylglutaric acid reacts more satisfactorily and gives 7,S-diphenyl-7,
5-pentenoic acid.93
The Perkin reaction is unsuitable for the direct preparation of a-halo-
genated cinnamic acids. When benzaldehyde is heated with sodiumchloroacetate and acetic anhydride only a trace of a-chlorocinnamic acid
is formed.94t 98
Sodium bromoacetate94
and fiuoroacetate96
under
similar conditions give none of the corresponding a-halogenated cin-
namic acids. a-Bromocinnamic acid (in a variety of crystalline forms)
can be obtained by the action of bases on the bromide of cinnamic acid
under carefully controlled conditions; aqueous sodium carbonate or ace-
tate converts the dibromide largely to /3-bromostyrene.
a-Phenoxy- and cresoxy-cinnamic acids can be prepared by heating
the sodium salts of aryloxyacetic acids with benzaldehyde and acetic
anhydride,97
but some cinnamic acid is formed also. The parent com-
pound, a-hydroxycinnamic acid, is the enol form of phenylpyruvic acid.
C6H6—CH=C—CO2H ?± C6HB—CH2—CO—CO2H
OH
Owing to this relationship certain derivatives of a-thiolcinnamic acid
(benzalrhodanine, etc.)98
and a-acylaminocinnamic acid (azlactones,etc.) can be hydrolyzed to give phenylpyruvic acid, and this forms an
elegant preparative method for arylpyruvic acids " and related com-
pounds.98
Several derivatives of a-thiolcinnamic acid can be obtained from the
corresponding a-thiolacetic acids. Sodium thiodiglycolate reacts with
91 Fichter and Gre ther , Ber., 36, 1407 (1903).92 Fit t ig, Ann., 282, 334 (1894)." F i c h t e r and Merkens , Ber., 34, 4177 (1901).M P15chl , Ber., 15, 1945 (1882).96 Michael, J. prakt. Chem., [2] 40, 64 (1889).9 6 S w a r t s , Bull. soc. chim., [4] 25, 325 (1919).97 Oglialoro, Gazz. chim. ital., 10, 483 (1880); 20, 505 (1890).98 Granacher , Helv. Chim. Ada, 5, 610 (1922).98 Buck and Ide, Org. Syntheses, 15, 33 (1935); Herbat and Shemin, ibid., 19, 77 (1939).
two molecules of benzaldehyde in the presence of acetic anhydride to
give a-thio-ta's-cinnamic acid, and no cinnamic acid is formed under these
conditions.100
C 6H 6CHO + S(CH2CO 2Ka)2 - ^ S /— C — C O 2 H \
\ CHC 6H 6 A
The most signif icant reaction of this type for preparative purposes is the
condensation of cyclic sulfur compounds, such as rhodanine (and related
heterocycl ic der ivat ive s) , wi th arom atic aldehy des. This condensat ion
can be effected readily under various conditions as the methylene group
of rhoda nine is qu ite a cti ve ; * excellent yields are obtain ed usin g a com -
bination of glacial acetic acid and sodium acetate. 10 1
RCHO + CH 2 CO C H ,C O a H RCH=
S N H CH*C02Na> S
cs cs
The resulting derivatives are useful intermediates for the preparationof arylthiopyruvic acids,98 /3-arylalanines,98 arylacetonitriles, arylaceticacids, /3-arylethylamines, etc.101 '102 These reactions have been par-ticularly useful in the furan series 98 '102 and for alkoxyphenyl com-
pounds.101 Furfural has been converted to 2-furanacetic acid 102 in anover-all yield of 73% by the following typical series of transform ations(five steps).
R C H = C CO
RCH 2CO 2HI I
N — O H93% yield 88% yield 96% yield
It is difficult to find another series of reactions that gives such uncom-monly good yields. I t is of interest to note that the process does notrequire strong mineral acids at any stage and consequently is well
adapted for use with acid-sensitive groups.* For a survey of earlier references to these condensations see Granacher, reference 98.100
Loeven, Ber., 18, 3242 (1885); see also Hinsberg, J. prakt. Chem., [2] 84, 192 (1911).10 1
The a-cyanocinnamic acids are prepared conveniently by using anaqueous solution of sodium cyanoacetate obtained from sodium cyanideand chloroacetic acid.112 The a-cyanocinnamic acids cannot be used asintermediates for preparing benzalmalonic or cinnamic acids since theyare resistant to hydrolysis by acids and are cleaved into benzaldehydeand malonic acid by strong alkalies.111 The addition of sodium cyanideto ethyl a-cyanocinnamate and subsequent hydrolysis with acids givesphenylsuccinic acid in 90-95% yields.113 The addition of ethyl malonateto ethyl a-cyanocinnamate leads in a similar way to a-phenylglutaricacid in 75-8 5% yields.114
The condensation of benzylcyanide with aromatic aldehydes leadsdirectly to the nitriles of a-arylcinnamic acids, C6H5CH=C(C 6H5)CN,115
which have limited application in synthetic work.
C omparison with O ther Synthetic M ethods
From the standpoint of its application in organic synthesis the Perkin
reaction is used most generally for the preparation of /?-arylacrylic anda-substituted-j3-arylacrylic acids. Two othe r methods of very gen-eral utility are available for the same purpose—the Claisen condensa-tion of aldehydes with esters and the Reformatsky reaction. For thepurpose of this discussion the condensations of malonic acid in the pres-ence of ammonia and primary or secondary amines will be designatedas the Knoevenagel modification * of the Perkin reaction, and the use ofmalonic acid in pyridine (usually with a little piperidine added) will bedesignated as the Doebner modification, f A general comparison ofthese reactions may be made for a simple example, such as the prepara-tion of cinnamic acid from benzaldehyde (see also p . 8).
* The term Knoevenagel react ion is used very broadly to include the condensat ion ofesters, nitri les, nitroparaffina, etc. , with a variety of carbonyl components in the presenceof ammonia or pr imary or secondary amines .
t The term Doebner reaction is often used for the synthesis of a-alkyl- and a-aryl-
cinchoninic acids from aromatic amines, aldehydes, and pyruvic acid.11 2 Lapwor th and McRae , J. Chem. Soc, 121 , 1699 (1922); Lapworth and Baker , Org.Syntheses Coll. Vol., I, 175 (1932).
11 3 Lapwor th and Baker , Org. Syntheses Coll. Vol., I, 440 (1932).114 M a ns ke , / . Am. Chem. Soc, 53, 1106 (1931).116 Fros t , Ann., 250, 157 (1889); Walther, J. prakt. Chem., [2] 53, 454 (18 96); Bra nd
Knoevenagel: Malonic acid; ammonia, piperidine, or diethylamine;alcohol as solvent; two to four hours' heating a t 100°; yield, 70-80% .
Doebner: Malonic acid; trace of piperidine; pyridine as solvent; oneto two hours' heating at 100°; yield, 80-90% .
C laisen:Ethyl acetate, absolute; metallic sodium and a trace ofalcohol; excess of ethyl acetate serves as solvent; two hours at 0-5°;
yield, 68-74% ."6
Reformatsky: * Ethy l brom oacetate; m etallic zinc; benzene as solvent;one to two hours at 100°, followed by heating and distillation (to dehy-drate intermediate /3-hydroxy ester); yield, 50-60%.
In the Claisen reaction the product is an ester, which can be saponi-fied readily to obtain the acid; in the Reformatsky reaction, the inter-mediate /3-hydroxy ester is subjected to dehydration and the resulting
cinnamic ester distilled and saponified.Na
C 6H 6CHO + CH3CO2C2HB ^ ^ C 6H 6CH=CHCO 2C2H6 + H2O(C laisen reaction)
The direct formation of an ester may be advantageous in many instances,as the purification of an ester by distillation is likely to be more conven-ient and less wasteful of material than recrystallization of the solid acid.Moreover, the esters are often desired in preference to the free acids foruse in subsequent transformations, such as conversion to amides,catalytic hydrogcnation, and formation of addition or substitution
products.The Perkin reaction is particularly well suited for reactions involvingnitrobenzaldehydes and halogenated benzaldehydes, since especiallyhigh yields are obtained with these compounds and these types of sub-stituents are unfavorable for the Claisen or Reformatsky reactions.Benzaldehydes containing a free phenolic group are likewise unsuited forthe Claisen or Reformatsky reaction but may be protected by alkylationor acetylation. In the Pe rk in're ac tion o-hydroxybenzaldehydes givecoumarins; the m- and p-hydroxy compounds yield the correspondingacetoxycinnamic acids, which can be hydrolyzed readily with alkalies.The Doebner modification is suitable for hydroxy compounds and gives
* See Cha pte r 1.116 Marve l and King , Org. Syntheses Coll. Vol., I, 246 (1932).
especially good results if the reaction is carried out by long standing atroom temperature.86
The Claisen reaction is definitely superior to the ordinary Perkinreaction for alkylbenzaldehydes, alkoxybenzaldehydes, and p-dimethyl-
aminobenzaldehyde. These types give 60-85% yields of the corre-sponding cinnamic esters in the Claisen reaction, and similar good yieldsin the Doebner modification of the Perkin reaction; the Knoevenagelmodification is satisfactory also for such preparations. 2,4,6-Trimethyl-cinnamic acid is obtained only in traces in the usual Perkin reaction,bu t the ethyl ester can be prepared in 70% yield by the Claisen method.27
The Doebner modification is rapid and convenient, and for large-scalepreparations is less hazardous than the Claisen reaction. A large quan-
tity of pyridine is required, and it must be anhydrous for maximumyields. A technical fraction of pyridine bases (b.p. 120-160°), afterredistillation and thorough drying, gives as good results as pure pyridine;recent studies indicate that the pyridine bases can be used in stoichio-metric quantities m and even in catalytic amounts.87 The Knoevenagelmodification is simpler from the standpoint of solvent required, asalcohol is a satisfactory medium. Neither the D oebner nor the Knoeven-agel modification is used for a-substituted cinnamic acids as the requisite
monosubstituted malonic acids are not readily accessible.A satisfactory synthesis of substituted cinnamic acids from the corre-sponding benzyl halides has been developed by von Braun and Nelles.118
The benzyl halide is converted to the corresponding malonic acid in theconventional way; the resulting /3-arylmalonic acid is then brominated,decarboxylated, and treated with alkali.
RCH2Br -> RCH2CH(CO2H)2 - ^ » R C H2CBr(CO2H)2
RCH2CHBrCO2HN a
°H
> RCH=CHCO2HThis method is not suitable for aliphatic or alicyclic compounds butgives good over-all yields with a variety of substituted benzyl halides.The advantage of this synthesis over the Perkin or Claisen reaction liesin the circumstance tha t the benzyl halides are often more readily access-ible than the corresponding benzaldehydes.
a-Arylcinnamic acids are prepared most readily by the Perkin reac-tion , bu t good yields of the esters can be obtained in the Claisen reaction .
a-Alkylcinnamic acids are obtained readily by the Perkin reaction butsometimes more conveniently by the Claisen or Reformatsky reaction.Another method of preparative value involves the condensation of ben-
11 7 Dala i and Dut t , J. Ind ian Chem . Soc, 9, 309 (1932); C. A., 27, 279 (1933).118 von Braun and Nel les , Ber., 66, 1464 (1933).
zaldehyde with alkyl derivatives of acetone, and oxidation of the result-ing benzalacetones with sodium hypochlorite.119
""""' C6H6CH=CCO2H
RThis method has been used for a-n-propyl-, n-butyl-, and n-amyl-cin-namic acids. Benzalacetone itself gives cinnamic acid by hypochloriteoxidation, and a limited num ber of ring-substituted cinnamic acids havebeen prepared by this method.
The Doebner modification appears to be the best general method forthe preparation of /3-alkylacrylic acids and can be used to a limitedextent for /3,/3-dialkylacrylic acids. The acids obtained in th is way are
less likely to be contaminated with the isomeric /3,7-unsaturated acids.The Reformatsky reaction is the only one of the reactions th a t is suited
for the direct preparation of /3,/3-diarylacrylic acids, as benzophenone andits derivatives will react with bromoacetic esters and zinc but will nottake part in the Perkin or Claisen reaction.
SELECTION OF EXPERIMENTAL CONDITIONS
A number of studies have been m ade of factors influencing the yields inthe Perkin reaction, but it is difficult to draw any broad generalizations.In many of the preparations described in the literature the proportionsof reactants and the general procedure have been essentially those usedby Perkin: a mixture of two parts of the benzaldehyde with one part(by weight) of freshly fused sodium acetate and three parts (by weight)of acetic anhydride is hea ted for about eight hours at 175-180°. Theseproportions correspond, in the case of benzaldehyde, to about 1.5 molesof acetic anhyd ride and 0.65 mole of sodium acetate. Meyer and Beer29
reported tha t 2.1 moles of acetic anhydride and 0.7 mole sodium acetateper mole of aldehyde gave the best results for a group of substitutedaldehydes.
Recent work27 indicates that a slightly larger proportion of sodiumacetate, about 1 mole instead of 0.65-0.7 mole, gives a small improve-ment in the yields (5-10% ). Further increases in the amount of sodiumacetate, up to 2 moles, have little effect, but beyond this point the yieldsfall off. There is generally but little advantage in using more than 1.5
moles of acetic anhydride per mole of aldehyde; the use of 2 moles ofanhydride increases the yield only a small amount (1-3% ), and a largeexcess is deleterious. The use of an indifferent solvent such as toluene
119 Ger. pat., 21,162 (1882) (Frdl. , 1, 28); see also reference 131 , p . 24 3.
or nitrobenzene causes a marked drop in the yield and can impede thereaction completely. The addition of a small am ount of pyridine (8drops for 0.2 mole benzaldehyde) raises the yield of cinnamic acid from50-60% to 80-85% .38
It is reported27 that the yield of cinnamic acid is increased (using theproportions of Meyer and Beer) by prolonged heating at 180°. Theyields were as follows: heating two hours, 6% ; four hours, 2 1 % ; sixhours, 35 % ; eight hours, 45 % ; ten hours, 52 % ; fourteen hours, 6 1 % ;twenty-four hours, 7 2% ; fifty hours, 76.5%; one hundred hours, 77%.Although the yields may be increased in this way with certain aldehydes,with others better yields are obtained by shorter periods of heating andat lower tem peratu res. In general a period of seven to eight hours' hea t-
ing at 170-180° is adequa te when sodium acetate is used. A period ofthree to five hours' heating at 140-160° may be sufficient if potassiumacetate is used, and bette r yields are secured in this way with some alde-hydes.
Meyer and Bee r27 studied the influence of a series of metallic acetateson the yields of various cinnamic acids and observed that potassium ace-tate gave a definite improvement over sodium acetate (64% yield asagainst 48 % , with benzaldehyde). W ith o-chlorobenzaldehyde and
various metallic acetates, using 2.1 moles anhydride and 0.7 moleaceta te (eight hours at 180°), the yields were: lithium, 5 8% ; sodium,7 1 % ; potassium, 7 8% ; rubidium, 8 2% ; magnesium, 0 % ; calcium, 8 % ;barium, 3 % ; copper, 3 % ; lead, 70% ; mercury, 37% .
Kalnin 18 carried out an extensive study of various factors influencingthe yields in the Perkin reaction.* He found that te rtia ry amines cata-lyze the formation of cinnamic acid from benzaldehyde and aceticanhydride, in the absence of metallic acetates, and that their activityincreases with the basic streng th of the amine. Likewise, there is anoptimum ratio of amine to acid anhydride; with triethylamine this isabout one-third mole, but for weaker bases a larger proportion is re-quired. Benzaldehyde, acetic anhydride (1 mole), and triethylamine(0.33 mole), hea ted a t 180° for eight hours, gave a 29 % yield of cinnamicacid; the same am ount of pyridine gave only 1% y ield. These amineswere slightly more effective with propionic than with acetic anhydride.A mixture of benzaldehyde (1 mole), phenylacetic anhydride (0.5 mole),acetic anhydride (4 moles), and pyridine (2 moles) gave a 95 % yield of
a-phenylcinnamic acid after five hours' heating at 150°.Kalnin 18 also found that metallic salts other than acetates can act as
cata lysts in the Perk in reaction. The following yields of cinnamic acidwere obtained using benzaldehyde (1 mole), acetic anhydride (1.5
* K alnin's paper also includes a survey and critical review of earlier work in the field.
moles), and various metallic salts (0.65 mole-equivalent), with eighthours' heating at 180°.
Potassium acetate 72%Potassium carbonate 59%
Sodium carbonate 40%Sodium acetate 39%
Trisodium phosphate 36%
Potass ium sulfiteTripotassium phosphate
Potassium sulfidePotassiumPotassium
The effect of the duration of heating (at 180°)
cyanideiodide
was studiedthese catalysts, and the following yields were obtained.
O N E - F O U R T H HOTJB
Potassium carbonate 34%
Sodium carbonate 3%
Sodium acetate 0%
i ONE H O U R
40%14%
2 %
F O U R H O U R S
52%27%
20%
32%20%
8%0%0%
for three of
E I G H T H O U R S
59 %40%
39%Kalnin's results indicate that potassium carbonate may be substitutedadvantageously for sodium acetate but that it is not quite so effective aspotassium acetate.
Chappell12 0 investigated the duration of heating when potassiumacetate is used, and compared three aldehydes under similar conditions(1.5 moles anhydride and 0.63 mole potassium acetate, at 180°). Thefollowing yields were obtained.
Two HOURS F OUR HOUR S SIX H O U R S E I G H T H O U R S
Benzaldehyde 52% 55% 58% 60%
Anisaldehyde 30% 35% 30% 20%
Furfural 56% 49% 40% 28%
A parallel series of experiments using sodium acetate showed that theyields increased steadily up to eight hours' heating. It is clear that the
optimum conditions with potassium acetate are likely to be quite dis-similar for different types of aldehydes. For furfural the most favorable
results were obtained with four to five hours' heating at 150°, or six toseven hours at 1 4 0 ° ;m when potassium acetate was used the addition of
pyridine did not improve the yield.In the presence of the most active acetates cinnamic acid is formed
slowly at 100°; the following yields were obtained by boiling for oneminute to dissolve the salt, *and then heating at 100° for sixteen hours :120
* The solubility of the metallic acetates in the reaction mixture is an important factor
and Kalnin attributes the results of Meyer andBeer, inpart, to the lowsolubility of certainof the salts, for example, lithium ac etate. Kalnin's rate studies with the alkali carbonatessuggest that these bases neutralize the acetic acid formed during the reaction and therebyoffset its retarding effect.
monium, 1 8% ; thallous, 14% ; lead, 0% . Rubidium and cesium salts aretoo rare to be used for preparative purposes, but these results suggestthat quaternary ammonium salts might be good catalysts under appro-priate conditions.
Michael observed that free acetic acid has a retarding effect on theformation of cinnamic acid. This is readily understandable in term s ofthe current theory that the reaction involves enolization of the anhy-dride, since acetic acid would suppress the enolization. Kalnin obtainedthe following yields of cinnamic acid when increasing amounts of glacialacetic acid were added to the usual reaction mixtures and the reactionswere carried out at 180° for eight hours.
ACETATE USE D MO LES OF ACETIC ACID ADDED AND YIELD S
The effect of acetic acid depends upon the degree of activity of thereacting components. o-Chlorobenzaldehyde reac ts readily with a mix-ture of potassium ace tate and glacial acetic acid to give o-chlorocinnamicacid in 70% yield; with a less reactive salt, sodium acetate, only half ofthe aldehyde undergoes reaction and the yield is only 24% . W ith com-
pounds having a very active methylene group (malonic acid, cyanoaceticacid, rhodanine,101 hydantoin,10 9 etc.), excellent yields of condensationproducts can be obtained in the presence of glacial acetic acid.
The unfavorable effect of acetic acid is reduced in the customaryprocedures for the Perkin reaction by using an air-cooled condenser, andat the temperatures employed the acetic acid distils out of the reactionmixture. This means of overcoming the retard ing effect of the aceticacid formed in the reaction is of considerable importance with benzalde-
hyde and less reactive aldehydes. I t is quite likely th a t discrepancies inyields reported in the literature are due in large measure to variations inthe extent of removal of acetic acid.
The effect of various factors on the reaction of phenylacetic acid (oranhydride) with o-nitrobenzaldehyde has been studied exhaustively byBakunin and Peccerillo.19 They obtained the following yields of a-o-nitrophenylcinnamic acid when a standard reaction mixture (1 molealdehyde, 1 mole phenylacetic acid, 3 moles acetic anhydride, 1 molemetallic salt or organic base) was heated for twelve hours at 90°.
Thus, with two very reactive components, the tertiary amines provedto be very effective catalysts, whereas Kalnin found that triethylaminegives only a 29% yield of cinnamic acid with benzaldehyde and aceticanhydride. In another series of experiments, using phenylacetic anhy-
dride, Bakunin and Peccerillo
19
obtained the following yields.Rubidium acetatePotassium acetateSodium acetateLithium acetateBarium acetateWithout any salt or
These workers found th a t acetic anhydride can be replaced by propionic,butyric, or valeric anhydride, but benzoic anhydride gave low yields ofo-nitrophenylcinnamic acid. Inorganic dehydrating agents such asphosphorus pentoxide and anhydrous calcium chloride were ineffective.Likewise, ethyl phenylacetate could not be substituted for phenylaceticacid (or anhydride).
US E OF THE PERKIN REACTION IN S YNTHE S I S
The Perkin reaction and related condensations afford a means of trans-forming an aromatic aldehyde group into a variety of side chains. Thecorresponding reactions can be used only to a limited extent with ali-phatic aldehydes (and a few ketones) bu t are nevertheless of some prepar-ative value in the aliphatic series. The types of compounds tha t willparticipate in these reactions have been reviewed in considering thescope of the reaction (pp. 217-233). The following brief summary indi-cates the types of compounds that can be obtained directly by means of
the Perkin reaction in its varied forms.
a,P-Unsaturated Acids
R C H = C H C O 2H . jS-Arylacrylic acids are prepared by the usualPerkin reaction, or by the Knoevenagel and Doebner modifications usingmalonic acid. 1-Naphthaldehydes, 2-furanaldehydes, and 2-thiophene-aldehyde may be used instead of benzaldehyde. /3-Alkylacrylic acidscan be prepared from aliphatic aldehydes and malonic acid, preferably
by the Doebner modification.R 2C = C H C O 2H . /3,/3-Diarylacrylic acids cannot be prepared by the
Perkin reaction; the /3,/3-dialkylacrylic acids can be obtained from dialkylketones and malonic acid, preferably by the Doebner modification.
R C H = C C O 2H . a-Alkyl- and a-aryl-cinnamic acids are prepared
Rreadily by the Perkin reaction from benzaldehydes and substituted
acetic acids. a-Vinylcinnamic acids may also be prepared (see below).
O ther Unsaturated Acids
R C H = C H C H 2C O 2H . 7-Alkyl and 7-aryl derivatives of vinylaceticacid can be obtained by using sodium succinate and acetic anhydride inthe Perkin reaction. Under mild conditions (120°) the interm ediateparaconic acids can be obtained (Fittig 's modification). /3,7-Disubsti-
tuted derivatives are obtained by using sodium methyl- or phenyl-suc-cinate. I t is reported th a t the Knoevenagel modification, using malonicacid and amines, often gives mainly /3,-y-unsaturated acids when aliphaticaldehydes are used.80
R C H = C H C H 2C H 2C O 2H . The reaction of sodium glutarate withbenzaldehyde gives very low yields of 7-benzalbutyric acid (R = C 6H 5).Sodium a-phenylglutarate reacts more satisfactorily and gives 7,5-di-phenyl-7-pentenoic acid.
RCH==CHCH==CHCO 2H . Butadiene-1-carboxylic acid (R = H)and sorbic acid (R = CH 3) can be prepared from acrolein and croton-aldehyde, respectively, using the Doebner modification. 4-Phenyl-butadiene-1-carboxylic acid (R = CgHs) can be obtained from cin-namaldehyde in the usual Perkin reaction, and also by the Knoevenagelor Doebner modification.
R C H = C H C H = C C O 2H . 1-Alkyl and 1-aryl derivatives of 4-phenyl-
R '
butadiene-1-carboxylic acid are prepared from cinnamaldehyde and sub-stituted acetic acids in the usual Perkin reaction.
R C H = C C H = C H 2 . l-Phenylbutadiene-2-carboxylic acid is obtained
C O 2 H
by condensation of benzaldehyde with crotonic anhydride in the presenceof triethylamine.20 The corresponding 3-methyl homolog is obtained byusing /3-methylcrotonic anhydride.70
R C H = C C H = C H R . l,4-Diarylbutadiene-2-carboxylic acids are ob-C O 2 H
tained by condensing benzaldehydes with |3-benzalpropionic acid.122
R C H = C H C = C H C H = C H R . A sma ll am ou nt of 1,6-diphenylhexa-
C O 2H
tr iene-3-carboxylic acid is formed by condensing cinnam aldeh yde with
sodium /3-benzalpropionate a nd acetic anhy drid e und er mild condi-
tions,123 bu t this does not ap pear to be a sat isfactory prep arat ive method.
Cyclic Compounds
RC H CH—CO2H -y-Alkyl- an d 7-ary l-paraco nic acids are obta ined
I I by warm ing a l iphat ic and arom atic a ldehydes
v /2 with sodium succinate and acet ic anhydride a t
C o 100-125° (Fitt ig 's syn thesis). A t higher tem -peratures, or on heating the paraconic acids,
/3,7-unsaturated acids and 7-butyrolactones are formed.CHR Phth alylacet ic acid (R = CO 2H) is prepared
from phthal ic anhydride, potass ium aceta te ,an d acetic an hy drid e. W ith phen ylacetic acidand others , at higher temperatures, decarboxyl-ation occurs an d ben zalph thalide (R = C 6H 5 ,etc.) is formed. D isu bst itu ted com poun ds can
be obtained f rom phthal ic anhydride and disubs t i tuted acet ic acids .
Coumarin and r ing-subst i tuted coumarins canbe prep ared b y hea ting salicylaldehydes withacetic anhy dride and sodium aceta te . a-Alkyland a-aryl coumarins are obtained from substi-tu te d acetic acids. Ce rtain a, /3-disubstituted
coumarins can be prepared from o-hydroxy aryl ketones.124 '12B
R C H = C CO 5-Benzalrhodanine and re la ted compounds can
I I be prep ared by reaction of benzaldeh yde and
s. / i ts deriva tives w ith rhod anin e. 3-Sub stitutedQg rhoda nines m ay also be used.
R C H = C CO 5-Benzal derivative s of 4-oxazolone (azlactones)
I I are pre pare d from benzaldeh ydes and hippuric
•N / acid (R' = CeH 5) . O ther acyl der ivativ es of
C glycine give similar com pou nds (R ' = CH 3 ,
I C H 2 C 2 H 5 ) etc . ) .R '
12 3 Knell, Dissertation, Munich (1902); reported by Smedley, J. Chem. Soc, 93, 373(1908), and by Kuhn and Winterstein, reference 46, p. 220.
R C H = C CO 5-Benzal der iva t ives of hyd anto in (R ' = H ) ,
, 1 I 2-th iohy dan toin, creatinine, an d a nu m ber of
v / similar compounds can be prepared from ben-
QO zaldehyde and the app ropria te der ivat ives ofglycine.
Indirect Syntheses
The products obtained directly in the Perkin reaction and its variousramifications often serve as intermediates for the preparation of othertypes of compounds. The following paragraphs are intended merely toindicate in a brief way the essential operations involved in typical
syntheses th a t have some prepara tive value. For convenience the typesare listed for aryl compounds (where the starting material would usuallybe benzaldehyde). In many instances the reactions used are applicablealso to compounds with other organic radicals (R = alkyl, vinyl orpropenyl, styryl, 2-furyl, etc.)
R C H = C H 2 . Styrene and /3-alkylstyrenes can be obtained by thethermal decarboxylation of the corresponding cinnamic acid.126 A verygeneral method that can be applied to alkyl and aryl derivatives ofacrylic acid consists in adding hydrobromic acid (or hydriodic acid) atlow temperature, and treating the resulting /3-haloacid with sodium car-bonate.127' 128
R C H = C H B r . /3-Bromostyrenes are obtained by heating the dibro-mide of the corresponding cinnamic acid with sodium carbonate solu-tion,129 or with potassium (or sodium) acetate.13 0
B r 2 Na2CO»
R C H = C H C O 2 H > RCHBrCHBrCO 2H > R C H = C H B r + C O2(NaOAo)
/3-Alkyl-/3-bromostyrenesm
can be prepared from the dibromides ofa-alkylcinnamic acids by use of alcoholic sodium acetate (75% yields).1-Bromoolefins can be obtained from alkylacrylic acids, preferably bydehydrohalogenation of the dibromides with pyridine.132
R C = C H . Arylacetylenes and alkylacetylenes may be prepared fromthe corresponding bromostyrenes or 1-bromodlefins, obtained as de-
12 6Abbott and Johnson, Org. Syntheses Coll. Vol., I, 430 (1932).
scribed above, by dehydrohalogenation with solid potassium hydroxide,133
alcoholic alkalies,129 or preferably with sodium amide.134 Alkyl deriva-tives of phenylacetylene may be prepared from the corresponding;6-alkyl-j6-bromostyrenes,131 or by alkylation of phenylacetylene with alkylsulfates or toluenesulfonates.136
RC H 2C H O . Arylacetaldehydes may be prepared by addition ofhypochlorous acid to cinnamic acid, and heating the a-chloro-/3-hydroxyacid with sodium hydroxide or carbonate solution.136
T T O C t T sT a O P T
RCH=CHCO 2H > RCHOHCHCICO2H > RCH 2CH=O
A more refined method consists in treating the acrylic amides withhypochlorite in the presence of methanol, and hydrolysis of the resultingvinyl urethane with dilute acid.
RCH=CHCONH 2N a
°C 1
> RCH=CHNHCO2CH3 "°°>CH3OH H2SO4
RCH 2CH=O + NH3, etc.
This procedure, due to Weerman,137 has permitted the synthesis of sev-eral difficultly accessible aldehydes.138
RC H 2C H = N O H . Substituted acetaldehydes may also be obtainedvia the acetaldoximes, which can be prepared in excellent yields from
benzalrhodanines, etc. (see p. 230).
98
'
102
RC H 2C O 2H . Substituted acetic acids may be obtained by peroxideoxidation of the substituted pyruvic acids, which are secured by way ofthe azlactone synthesis (see p . 230). They are also obtained in good yieldsfrom the substituted acetaldoximes, by dehydration to the nitriles,R C H 2 C=N, and subsequent hydrolysis (see p. 230).
RC H 2 C = N . These may be prepared in good yields by dehydrationof the corresponding substituted aldoximes (see preceding paragraph).
RC H 2C H 2NH 2. ^-Substituted ethylamines may be obtained by reduc-tion or by catalytic hydrogenation of RCH 2C H = N O H , R C H 2CN, orR C H = C H N H C O 2C H 3 (see under R C H 2CH0) .
RC H 2C H 2C O 2H . /3-Substituted propionic acids are prepared readilyfrom the corresponding acrylic acids by reduction with sodium amalgam,by electrolytic reduction,139 or by catalytic hydrogenation.
1 3 3Hessler , Org. Syntheses Coll. Vol., I, 428 (1932).134 Bourguel, Ann. chim., [10] 3, 225 (192 5); Org. Syntheses Coll. Vol., I, 185 (1932).136 Truche t , Ann. chim., [10] 16, 309 (1931); John son, Schw artz, a nd Jacobs, J. Am.
Chem. Soc, 60, 1882 (1938).13 6 E r l e n m e y e r a n d L i p p , Ann., 2 1 9 , 1 8 2 ( 1 8 8 3 ) ; F o r r e r , Ber., 1 1 , 9 8 2 ( 1 8 7 8 ) .13 7 W e e r m a n , Ann., 4 0 1 , 1 ( 1 9 1 3 ) ; Rec. trav. chim., 2 9 , 1 8 ( 1 9 1 0 ) ; 3 7 , 1 ( 1 9 1 7 ) .1 3 8 R i n k e s , Rec. trav. chim., 3 9 , 2 0 0 , 7 0 4 ( 1 9 2 0 ) ; 4 5 , 8 1 9 ( 1 9 2 6 ) ; 4 6 , 2 6 8 ( 1 9 2 7 ) ; 4 8 , 9 6 0
(1929).1 3 9 I n g e r s o l l , O r g . Syntheses Coll. Vol., I , 3 0 4 ( 1 9 3 2 ) .
RCH2CHCO2H. a,/3-Disubstituted propionic acids are obtained by
R'
reduction of the corresponding acrylic acids.
RCHCH2CO2H. /3,/3-Disubstituted propionic acids can be prepared
R 'by addition of aromatic hydrocarbons to cinnamic acids in thepresenceof sulfuric acid,140
orpreferably aluminum chloride.141 Grignard reagentsundergo 1,4-addition to a,/3-unsaturated esters to give derivatives of/3,/3-disubstituted propionic acids.142
RCH = CH CN . ^-Substituted acrylonitriles can beprepared bydecar-boxylation of the a-cyanoacrylic acids, obtained by condensation ofcyanoacetic acid with aldehydes. Thea-aryl derivatives of /3-arylacry-lonitriles can be obtained directly by condensation of benzyl cyanidewith aromatic aldehydes.
R C = C C O 2H . Substituted propiolic acids may beobtained by dehy-drohalogenation of the dibromides of the corresponding cinnamicesters.143
RCH=CHCO2C2H6 -^» RCHBrCHBrCO2C2H5 ^ ° H > RC=CCO 2HEtOH
The free acid is notused asthis would favor decarboxylation to form the/3-bromostyrene, which isformed as an accessory product even when theesters areused.
RCH2COCO 2H. /3-Substituted pyruvic acids can beprepared ingoodyields by hydrolysis of the corresponding azlactones " or a-acylaminoacids106
(see p. 253). The corresponding a-thiopyruvic acids can beobtained by hydrolysis of the condensation products formed from rho-danine andaromatic aldehydes (see p.230).
RCOCH2CO2C2H5 Benzoylacetic ester can be prepared by additionof bromine to ethylcinnamate, dehydrohalogenation under mild condi-tions to a-bromocinnamic acid, and treatment of thea-bromo ester withcold concentrated sulfuric acid.144
etc.
C6HBCOCH2CO2C2H5
The aroylacetic esters can also be obtained from the corresponding1 4 0Liebermann and Hartmann, Ber., 25, 960, 2124 (1892).
arylpropiolic ester by hydration with cold sulfuric acid,146 or by additionof a secondary amine and subsequent hydrolysis of the /3-dialkylamino-cinnamic ester.146
and u-substituted derivatives (R = CeHs, etc.) can be obtained byhydrolysis of phthalylacetic acids or benzalphthalides, produced byinteraction of phthalic anhydride and acetic anhydride, phenylaceticacid, etc. (see p. 223).
RCH2CHCO2H. Substituted alanines can be obtained from the cor-
N H 2
responding azlactones by reduction (or catalytic hydrogenation) andhydrolysis of the resulting saturated acylamino derivative (see p. 231).
The details of the procedure may be varied according to the nature ofthe groups present, and this series of transformations has been used fora variety of substituted alanines.
The condensation products from aldehydes and rhodanine may beused in a similar way to obtain substituted alanines.88
RCHCH2CO2H. Derivatives of /3-alanine may be obtained by the
NH 2
action of an excess of hydroxylamine on substituted acrylic acids, ortheir esters.147 If a large excess of ammonia or methylamine is used inthe Knoevenagel modification, /3-aryl-/3-aminopropionic acids may beformed in considerable amount along with the /3-arylacrylic acid.84
RCH2CH2CH2NH2. 7-Substituted propylamines can be obtained byreduction or catalytic hydrogenation of ^-substituted acrylonitriles(RCH=CHCN), obtained from aldehydes and cyanoacetic acid as out-lined above.
RCH2CH2CH2CO2H. 7-Substituted butyric acids can be obtained by
hydrogenation of the /3,7-unsaturated acids obtained by Fittig's para-conic acid synthesis.
Derivatives of cinnamic acid have been of greatvalue for the synthesis of a number of polycyclicsystems. In 1898 Pschorr14 8 developed a very gen-eral method for the synthesis of phenanthrene andits derivatives, and this has found wide application
' " P e r k i n , J. Chem. Soc, 45, 174 (1884).148
34, 3998 (1901); 35, 4400, 4412 (1902); 39 , 3106 (1906); Ann., 391, 40 (1912), and otherpap ers. Fo r an excellent surv ey of Pscho rr 's synth esis see Fieser, reference 149.
in studies of morphine derivatives, carcinogenic hydrocarbons, etc.149
The essential features of Pschorr's synthesis are illustrated by themethod used to prepare phenanthrene-9-carboxylic acid. o-Nitroben-zaldehyde was condensed with sodium phenylacetate to give a-phenyl-
2-nitrocinnamic acid; this was reduced to the corresponding amino acid(I , 77% yield), which was diazotized and treated with copper powder, ascatalyst, to effect ring closure to phenanthrene-9-carboxylic acid (II,93% yield). The latte r gave phenanthrene upon decarboxylation (64%yield).
9O 2H
A similar series of reactio ns star tin g from o-nitrob enza ldehy de an d
sodium a-naphthylacetate leads to chrysene-5-carboxylic acid (III) ,1 5 0
which yields chrysene on decarbox ylation. W he n o-nitroben zaldehy de
COOH COOH
i n I V CO2H
and sodium /3-naphthylacetate are used as s tarting materials , the sub-sequ ent r ing closure tak es place at the 1- or 3-position of the na ph tha len ering leading respectively to 3,4-benzo-4-phenan throic acid (IV, 40% )and l ,2-benz-4-anthroic acid (V, 60%).1 5 1 The first synthesis of 1,2,5,6"
dibenzanthracene was accomplished by means of the Pschorr synthes is
sta rtin g from the acid obtain ed b y a dou ble condensation of 1,4-benzene-diacetic acid with two moles of o-nitrobenzaldehyde.162
149Fieser, "The Chemistry of Natural Products Related to Phenanthrene," second
edition, Reinhold Publishing Corporation, New York (1937), pp. 28-31 , 96-98, 343.160 Weitzenboek and Lieb, Monatsh., 33, 557 (1912).16 1
Cook, / . Chem. Soc, 2524 (1931). Earlier workers mistook 1,2-benzanthracene for3,4-benzophenanthrene; see Weitzenbock an d Lieb, reference 150, and Mayer an d Oppen-
heimer, Ber., 51, 513 (1918).162
Weitzenbock and Klinger, Monatsh., 39, 315 (1918).
The bimolecular reduction of methyl cinnamate by means of amal-gamated aluminum leads to methyl /3,/3'-diphenyladipate (meso andracemic forms).
AI-HK C6H6—CH—CH2—CO2CH 32C 6HB—CH=CH—CO2CH3 => |
C6H6—CH—CH2—CO2CH3
Although low yields are obtained in this reduction, it has served as asource of /3,/3'-diphenyladipic acid, which has been used for the synthesisof chrysene derivatives and of chrysene itself.
163
LABORATORY PROCEDURES
Cinnamic AcidUsing Acetic Anhydride and Potassium Acetate.* A mixture of 21
g. (0.2 mole) of freshly distilled benzaldehyde, 30 g. (0.3 mole) of 95%acetic anhydride, and 12 g. (0.12 mole) of freshly fused potassium aceta teis refluxed in an oil bath at 170-175° continuously for five hours, usingan air-cooled condenser.
The hot reaction mixture is poured into about 1200 cc. of warm water,part of which is used to rinse the reaction flask, and unchanged ben-
zaldehyde is removed by steam distillation. The residual liquid is cooledslightly, 3-4 g. of decolorizing carbon is added, and the mixture is boiledgently for five to ten minutes. The liquid is filtered rapidly through afluted filter paper; the clear nitrate is heated to boiling, 12-14 cc. ofconcentrated hydrochloric acid is added carefully, and the hot solutionis cooled rapidly w ith good stirring . After the cinnamic acid has crystal-lized completely the crystals are filtered with suction, washed with sev-eral small portions of water, and dried. The acid obtained in this way
melts at 131.5-132° and is pure enough for most purposes. The yieldis 16-18 g. (55-60%).
Using M alonic Acid and a Pyridine Base.117 A mixture of 10.6 g.benzaldehyde (0.1 mole), 10.4 g. malonic acid (0.1 mole), and 9.3 g.a-picoline (0.1 mole) is heated for three to four hours in a water bath at70°. At the end of this period evolution of carbon dioxide has ceased,and the reaction m ixture is then treated with 500 cc. of water and 25 cc.of concentrated hydrochloric acid. Unchanged benzaldehyde is removedby steam distillation, and the cinnamic acid is isolated as described in
* The advantage of potassium acetate over sodium acetate is that a shorter period ofheating is required to obtain comparable yields.
163 von Braun and Irmiach, Ber., 64, 2461 (1931); see also Robinson and collaborators,J. Chem. Soc, 607 (1933); 1412, 1414 (1935).
common aldehydes, where a good yield is important, and also for alde-hydes that do not give good yields in the usual Perkin reaction.
The presence of water in the reagents causes a marked lowering of theyields. Pyrid ine should be dried thoroughly over solid caustic and
redistilled; higher-boiling pyridine bases (boiling up to 165°) give asgood yields as pyridine, when dried thoroughly and distilled. Whenhigher bases are used the reaction m ixture is heated for two hours on th esteam bath instead of one hour followed by twenty minutes' boiling.
This general procedure is essentially that described in the literaturefor several alkoxybenzaldehydes.16 4'15 6 By the directions given above,4-methoxy- and 3,4-dimethoxy-benzaIdehyde furnish p-anisyl- andveratryl-acrylic acids, respectively, in 80% yields. p-Dimethylamino-benzaldehyde is reported to give the corresponding cinnamic acid in80% yield by a similar procedure.74
3-Methoxy-4-H ydroxycinnamic Acid (Ferulic Acid)
A solution of 15.2 g. (0.1 mole) of vanillin, 23 g. (0.22 mole) of malonicacid, and 1 g. (1.2 c c , 0.012 mole) of piperidine in 50 cc. of dry pyridineis allowed to stand a t room temperature for three weeks. During thistime the reaction mixture is protected by a soda-lime tube but must
not be corked as carbon dioxide is evolved; a Bunsen valve may be used.The reaction mixture is poured with stirring into a m ixture of 60 cc. of
concentrated hydrochloric acid and 100 g. of chopped ice. The acidprecipitates at once, and after standing until separation is complete it isfiltered w ith suction. The product is washed with 10 cc. of 5% hydro -chloric acid, followed by two 10-cc. portions of water, and then dried.The yield of ferulic acid, m.p. 173° (cor.), is 14-17 g. (70-85% ).
This procedure is an ad aptation of the Doebner modification developed
by Vorsatz8a
and is particularly advantageous for preparing cinnamicacids having a free phenolic group. These compounds give low yields at100° in the Doebner procedure, presumably owing to the ease of decar-boxylation of the hydroxycinnamic acids.
The following yields were reported by Vorsatz with other substitutedbenzaldehydes, with the same proportions of aldehyde and malonic acid:2,4-dihydroxycinnamic acid (caffeic acid), using 1.4 g. aniline instead ofpiperidine, allowing to stand overnight, and then warming at 50-55°until evolution of carbon dioxide was essentially complete (about threehours), in 87% yield; 3,4-methylenedioxycinnamic acid (piperonylacry-lic acid), using piperidine and standing four weeks a t room temperature,
164 Cain, Simonsen, and Smith, J. Chem. Soc, 63, 1035 (1913).166 Hawor th , Perk in , and Rankin , J. Chem. Soc, 125, 1693 (1924).
in 83 % yield after recrystallization from 75 % alcohol; 3,4-dihydroxy-coumarin-a-carboxylic acid (daphnetin-3-carboxylic acid), using anilineor pyridine and warming for twenty hours at 37°, in 83% yield.
If a large excess of am monia (60 moles) or methylam ine is used in this
reaction a mixture of the /3-aminopropionic and acrylic acids is formed.84
a-Methylcinnamic Acid 2
A mixture of 21 g. (0.2 mole) of freshly distilled benzaldehyde, 32 g.(0.25 mole) of propionic anhydride, and 20 g. (0.2 mole) of fused sodiumpropionate is heated with occasional shaking for thirty hours in an oilbath a t 130-135°. The warm mixture is poured into about 500 cc. of
water, stirred thoroughly, and neutralized by the addition of sodiumcarbonate solution. After removal of unchanged benzaldehyde bysteam distillation (or ether extraction), the solution is warmed with 3-4g. of decolorizing carbon and filtered while ho t. The warm filtrate ispoured slowly, with stirring, into an excess of concentrated hydrochloricacid mixed with chopped ice. After the acid has crystallized completelyit is collected with suction, washed with several portions of water, anddried. The crude product, amounting to 21-25 g., is recrystallized fromligroin and gives 19-23 g. (60-70% yield) of purified material.
The acid obtained in this way may melt at 81° or 74°, as a-methyl-cinnamic acid exists in two different crystalline forms. Both forms havethe same configuration (trans C 6H 5: CO2H) and give the same ester.Occasionally a mixture of the two trans forms is obtained which melts at77-78°. The true geometrical isomer, aWo-a-methylcinnamic acid (cisC 6H 5: CO2H), melts at 91° and can be obtained by long exposure of theordinary acid to ultra-violet light.
Very little cinnamic acid is formed in this reaction when sodium ace-
ta te is used as cata lyst. Although some acetic anhydride is formed bythe anhydride-salt exchange, the concentration is low and its rate ofreaction a t 135° is much less than th a t of propionic anhydride. Athigher tem peratures more cinnamic acid is formed (p. 213).
The procedure given is essentially that of Edeleano; 156 a-methylcin-namic acid has also been prepared using propionic anhydride and sodiumpropionate,2 or acetic anhydride and sodium propionate at 100°,14 andby heating benzal chloride with sodium propionate at 150° (Erd-
mann166
)." • E d e l e a n o , Ber., 20, 617 (1887); Bee also Rupe and Buaolt , Ann., 369, 320 (1909);
In a 200-cc. round-bottom ed flask, 17.4 g. (0.10 mole) of dry potassiumphenylacetate, 5 g. of dry potassium carbonate (0.035 mole), 0.5 cc.pyridine, 10.6 g. (0.10 mcle) freshly distilled benzaldehyde, and 15.3 g.
(0.15 mole) freshly distilled acetic anhydride are mixed thoroughly undernitrogen. An air-cooled reflux condenser is attached , and th e flask iscarefully inserted in an oil ba th a t 180°. A vigorous bubbling takes placefor a few m inutes, after which the reaction proceeds quietly. Heating iscontinued a t 180-190° for two hours. The mixture is allowed to cool,and 300-400 cc. water is added with gentle heating to break up lumps.Potassium hydroxide solution (6 iV) is added un til the solution is basic(about 30 cc. is required), but care should be taken not to add a large
excess of base as the potassium salt of the acid is easily salted out ofsolution. The mixture is heated until all soluble material has dissolved;some oily materia l will remain undissolved. The flask is cooled underthe water tap and the solution extracted with 300-400 cc. of ether toremove unchanged benzaldehyde and a little stilbene (ca. 1 g.). Th ewater solution is acidified with 6 N hydrochloric acid (15-20 cc. isrequired), the precipitated acid filtered off, and the filtrate tested withmore acid for completeness of prec ipitation . The precipita te is con-veniently dried on a porous plate in a vacuum desiccator. The yield ofcrude acid, melting about 160°, is 13-15 g. (60-65% of the theoretical).It can be recrystallized by dissolving in 50 % ethanol a t boiling tem pera-ture and adding water until the solution is jus t cloudy. The solution iscooled very slowly, and long needles form gradually. The purified acidamounts to 11-12 g. (50-55% of theoretical) and melts at 168-170°(uncor.). The acid obtained in this way is the trans form.
p-n-H exylacrylic Acid (a,P-N oneno ic Acid)
In a large flask 114 g. (1.1 moles) of malonic acid is dissolved in 185 cc.of dry pyrid ine; the reaction is slightly exothermic. The solution iscooled in ice water, and 114 g. (1 mole) of freshly distilled n-heptaldehydeis added with stirring or good shaking. After a pa rt of the aldehyde hasbeen added the mixture rapidly sets to a mush of crystals, but moderatestirring is possible. Th e mixture is allowed to stand a t room tem peraturefor sixty hours with frequent shaking. During this time the mixturefroths owing to evolution of carbon dioxide, and at the end most of the
167 This procedure was furnished through the courtesy of Professor C. R. Hauser and M issMildred Patterso n, of Duke University. It is a modern version of th e Oglialoro modifica-tion incorporating results of Bakunin and collaborators, and the use of potassium carbon-ate and pyridine, as suggested by Kalnin's studies.
half hours in a reflux ap paratus fitted with ground-glass joints . To thehot (90°) solution is added 1-2 g. of charcoal, and the whole is boiled fora few minu tes. The boiling solution is then filtered and cooled to roomtem pera ture. M ost of the o-nitrophenylpyruvic acid separates as an oil,
bu t seeding or scratching causes crystallization to begin. After s tandingat room tem pera ture for two hours, the m ixture is cooled at 0° overnight.The light tan crystals of o-nitrophenylpyruvic acid are filtered, washedwith 5 cc. of cold water, and dried in a vacuum desiccator. The productweighs 4.3 g. and m elts at 117-120°. The aqueous mother liquor is con-centrated in vacuum to about 50 cc, and the oily product is seeded andworked up as before. The second crop of crystals weighs 1.7 g. andmelts at 119-120°. The total yield of product is 6.0 g. or 83 % of th etheoretical. This product is sufficiently pure for synthetic purposes: it
Malonic (1.3 m) + N H 3 (1 TO)Ac2O (7.5 g.) + NaO Ac (1 g.)Ac2O (1.4 g.) + NaO Ac (0.6 g.)Ac2O (7.5 g.) + NaOAc (1 g.)Ac2O (7.5 g.) + NaOA c (1 g.)Ac2O (1.6 g.) + NaO Ac (0.7 g.)Ac2O (8 g.) + NaOAc (1 g.)Ac2O (2.1 m) + NaOA c (0.7 m)Ac2O (exes) + NaOAc (1 m)
Ac 2O (2.1 TO) + NaO Ac (0.7 TO)Ac 2O (6.5 TO) + NaOA c (1.7 TO)Malonic (2 m) + C 6H 6N + C 6H uNMalonic (2.5 m) + C 6H 6N + C 5 H U NMalonic (2 TO) + C6HBN + C 6 H U NMalonic (2 TO) + C 6H 6N + C 6 H n NAc 2O + NaOAcMalonic (1 TO) + C 6HBN (0.15 TO)
Ac 2O (2.5 m) + NaOAc (1.2 TO)
Malonic (1.2 TO) + C5H5N + C 6 H U NAC2O (2.7 TO) + NaO Ac (1.2 TO)AC2O + NaO AcMalonic acid + C5H5NAc 2O (3 g.) + NaO Ac (1 g.)Malonic (1.3 TO) + CH3CO2HMalonic (2 TO) + CsHsN + C 6H5NH2
Malonic (1 TO) + C 6H 6N (0.15 TO)Malonic (2 m) + CsHsN + C5H11N
Malonic (1.5 m) + C 6H 6N +C 6H U N
Ac 2O (3 g.) + NaO Ac (1 g.)Malonic (1.5 TO) + N H 3 (1.2 m)
Malonic (3 TO) + CsHjN + C5H11NMalonic (2 TO) + C5HBN + C 6 H n N
Seelig, Ann., 237 , 151 (1887).16 6 Reich and Chaskelis , Bull. soc. chim., [4] 19, 289 (1916).16 6 Clark , Moore , and McArthur , Trans. Boy. Soc. Can., I l l , 2 8, 97 (1934); Chem.
Zentr., 11,45 (1935).' " G a b r i e l , Ber., 49, 1608 (1916).168 Tanasescu, Bull. soc. chim., [4] 41, 1075 (1927).1 6 9 T h a y e r , Org. Syntheses Coll. Vol., I, 390 (1932).170 Alway and Bonner , Am. Chem. J. 32 , 392 (1904).17 1 van der Lee, Bee. trav. chim., 45, 684 (1926).172 Eichengrun and Einhorn , Ann., 2 6 2 , 153 (1891).17 3 Einhorn and Gernsheim, Ann., 2 84 , 148 (1894).17 4
Re ic h , Ber., 4 5 , 8 0 8 ( 1 9 1 2 ) ; Bull. soc. chim., [4] 2 1 , 2 1 7 (1 9 1 7 ) .176H a n z l i k a n d B i a n c h i , Ber., 3 2 , 1289, 2285 (1899) .
176 Harding and Cohen, J. Am. Chem . Soc, 2 3, 603 (1901).177 Mundici , Gazz. chim. ital., 34, II, 117, 119 (1925).178 G a t t e rma nn , Ann., 347, 370, 373 (1906).179 Lock, Ber., 72 , 304 (1939).180 van de Bunt , Rec. trav. chim., 48, 125 (1929).18 1 Chakravar t i and o thers , J. Annamalai Univ., 2, 227 (1933); 5, 254 (1936); C. A., 2 8,
2008 (1934); 30, 4500 (1936).182 Fieser and Bowen, J. Am . C hem. Soc, 6 2 , 2106 (1940).18 3 Smith and Agre, J. Am. Chem. Soc, 60, 651 (1938).184
Maxwel l and Adams, J. Am. Chem. Soc, 52 , 2959 (1930).186 Slotta and Heller, Ber., 6 3 , 3029 (1930).186 Blum-Bergmann, J. Chem. Soc, 1030 (1935).187 Miller and Kinkelin, Ber., 2 2 , 1709 (1889).18 8 Schnell, Ber., 17, 1383 (1884).189 Perkin, J. Chem. Soc, 39, 413 (1881).190 Clayton , / . Chem. Soc, 97, 2109 (1918).19 1 Tiemann and Ludwig , Ber., 15, 2048 (1882); Reiche, Ber., 2 2 , 2356 (1889).192 Robinson and Walker , / . Chem. Soc, 194 (1936).19 3 Brandt and Horn , J. prakt. Chem., [2] 115, 374 (1927); Chakravart i , Haworth, and
Perkin, J. Chem. Soc, 2269 (1927).194
Chakravar t i , Ganapa t i , and Aravamudhachar i , J. Chem. Soc, 171 (1938).19 6 Werner , Ber., 2 8, 2001 (1895).196 von Konek and Pacsu, Ber., 51, 856 (1918); see also Eigel, Ber., 20, 2530 (1887);
Zincke and Leisse, Ann., 32 2 , 224 (1902); Sonn, Ber., 46 , 4052 (1913).197 M a nc ho t , Ann., 387, 281 (1912).198 Robinson and Shinoda, / . Chem. Soc, 127, 1977 (1925).199 Borsche and Wal ter , Ber., 60, 2112 (1927); see also Gryszkiewicz-Trochimowski,
Chem. Zentr., I, 872 (1938); C. A., 33, 7761 (1939).20 0 Paa l and Mohr , Ber., 29, 2306 (1896); Wheeler and Johns, Am. Chem. J., 43, 16
(1910).20 1 Johnson and Kohmann, J. Am . Chem. Soc, 37, 165 (1915); see also Einhorn and
Grabfeld, Ann., 2 43 , 367 (1888).20 2 Clemo, Haw or th , and W al ton , / . Chem. Soc, 2368 (1929).20 3
M a u t h n e r , J. prakt. Chem., [2] 1 5 2 , 2 3 (1 9 3 9 ) .20 4 v o n K r a n n i c h f e l d t , Ber., 4 6 , 4021 (1913) .20 6
H a w o r t h , / . Chem. Soc, 2282 (1927) .20 6
P e r k i n a n d T r i k o j u s , J. Chem. Soc, 2932 (1926) .
20 7 Rubens te in , / . Chem. Soc, 652 (1926).20 8 Tiemann and Lewy, Ber., 10 , 2216 (1877).20 9 Picte t and Finkels te in, Ber., 42 , 1985 (1909).21 0 Perkin and Schiess, J. Chem. Soc, 85, 164 (1904).211 Kaufmann and Burr , Ber., 40, 2355 (1907).21 2
Limaye , Proc. Indian Acad. Sci., 1A, 163 (1934); C. A., 29, 1796 (1935).21 3 Tiemann and Nagoi , Ber., 11, 647 (1878).21 4 Robinson and Sugasawa, / . Chem. Soc, 3169 (1931).216 Robinson and Shinoda, J. Chem. Soc, 127, 1979 (1925).21 6 Raiford and Lichty, / . Am. Chem. Soc, 52 , 4580 (1930); see also Raiford, Webster
and Pot t e r , Proc. Iowa Acad. Sci., 38, 171 (1931).21 7 Lorenz, Ber., 13, 757 (1880); Perkin, / . Chem. Soc, 59, 152 (1891).21 8 Pauly and Neukam, Ber., 40, 3494 (1917).21 9
Sonn, Mi l l le r , Bi l low, a n d M e y e r , Ber., 5 8 , 1 1 0 3 (1925) .22 0 Kanevskaja, Sohemiakin, and Schemiakina, Arch. Pharm., 2 7 2 , 774 (1934).221 Schlit ter, Ber., 6 6 , 992 (1933).22 2
Kuroda and Perk in , J. Chem. Soc, 123, 2110 (1923).22 3 M a ut hne r , J. prakt. Chem., [2] 110, 125 (1925).22 4 M a ut hne r , J. prakt. Chem., [2] 116, 319 (1927).22 6 M a ut hne r , J. prakt. Chem., [2] 150, 257 (1938).22 6 Jansen, R ec trav. chim., 50, 301 (1931); van Alphen, ibid., 47 , 176 (1928).2 2 7Herzig, Wenzel , and Gehringer , Monatsh., 2 4, 868 (1903).22 8 Roaenmund and Boehm, Ann., 437, 144 (1924).22 9 Shinoda, Kawagoe, and Sato, J. Pha rm. Soc. Japan, 51, 249 (1931).2 3 0 S p a t h , Monatsh., 41, 278 (1920).231 M a ut hne r , Ber., 41, 2531 (1908).23 2 Cohn and Springer , Monatsh., 2 4, 94 (1903).
T he acetoacetic ester conde nsa tion* consists in th e reaction , in th e
presence of certain bases, of an ester ha vin g hyd roge n on th e a-carbon
atom with a second molecule of the same ester or with another ester(which may or may not have hydrogen on the a-carbon atom) to forma /3-ketoester. T h e bases capab le of effecting such reactio ns includ e
sodium alkoxides , t r iphenylmethylsodium, sodium amide, and cer ta inGrignard reagents such as mesi tylmagnes ium bromide and isopropyl-magnesium bromide; also, metallic sodium effects certain condensations,
the sodium alkoxide which is formed in the reaction mixture probablyserving as the active condensing agent.1 The classical example of theacetoacetic ester reaction is the formation of acetoacetic ester itself bycondensation of ethyl acetate by means of sodium ethoxide, for which
the following reaction may be written.
CH 3CO2C 2HB + CH 3CO 2C 2H 6 + NaOC 2HB ->
CH 3C(ONa)==CHCO2C 2H 6 + 2C2H 6OH
Th e reaction prob ably involves an ionic m ech anis m ,2 '3 the first stepof which is an acid-base exchange; in th e presence of th e ethox ide ion th ehydrogen on the a-carbon atom is ionized as a proton to form the ester
anion (enolate anion), which is probably a resonance hybrid of the twos t ruc tu re s -CH 2 — C = O ( O C 2 H 5 ) and CH 2 = C — O - ( O C 2 H 5 ) .
(1) CH 3C O 2C 2H 6+ -OC 2H B <=> (C H 2 C O 2 C 2 H 6 ) -+ C 2H 6OH
The second step involves the condensation of the ester anion with the
carbonyl group of a molecule of unchanged ester , presumably formingan intermediate anion (with the charge on the oxygen) which, on releaseof the ethoxide ion, forms acetoacetic ester .
* This type of condensation is frequently called a Claisen reaction—a term that is usedalso for certain other types of condensation effected by bases, including ketone-ester con-densations to form 1,3-diketones and such aldol reactions as the condensations of ethylacetate with benzaldehyde to form ethyl cinnamate and of acetophenone with benzalde-hyde to form benzalacetophenone.
1 Snell and McElvain, J. Am. Chem . Soe., 53, 2310 (1931).2 Hauser and Renfrow, J. Am. Chem . Soc, 59, 1823 (1937).8 Hauser, J. Am. Chem. Soc, 60, 1957 (1938); Arndt and Eistert, Ber., 69, 2384 (1936).
#Kenyon and Young, / . Chem.Soc, 216 (1940).6 Brown and Eberly, J. Am. Chem.Soc, 62, 113 (1940).6 Schlenk, Hillemann, and Rodloff, Ann, 487, 135 (1931).7 Muller, Gawlick, and Kreutzmann, Ann., 515, 97 (1934).8 Hauser and Renfrow, Org. Syntheses,19, 43 (1939).8 Hudson and Hauser, J. Am. Chem.Soc, 62, 2457 (1940).
10 Renfrow and Hauser, J. Am. Chem.Soc, 60, 463 (1938).11 Hudson and Hauser, J. Am. Chem.Soc, 61,3568 (1939).
The reversibili ty of the acetoacetic ester condensation is well estab-
lished. C erta in /3-ketoesters, especially tho se ha vi ng one or tw o su b-
stituents on the a-carbon atom, are cleaved by alcoholic sodium ethoxide
to form es ters . Th us , a l though ethyl a-propio nylpro piona te is formed
by the self-condensation of ethyl propionate in the presence of sodium
ethoxide, when treated with alcoholic sodium ethoxide it reverts toe thyl propiona te ;1 2 similarly, ethyl diethylacetoacetate is cleaved by
alcoholic sodium ethoxide to form ethyl diethylacetate and ethyl
acetate.12
C H 3 C H 2 C O C H ( C H 3 ) C O 2 C 2 H S **™?> 2 C H 3 C H 2 C O 2 C 2 H 6
C2H5OH
C H 3 C O C ( C 2 H 6 ) 2 C O 2 C 2 H 6 ^ ° C ^ 5 > H C ( C 2 H 6 ) 2 C O 2 C 2 H B + C H 3 C O 2 C 2 H 6
An interesting example is the reversion of the product from ethyl iso-bu tyrate and ethyl benzoate. Although ethy l benzoyldimethylacetateis obtained by short treatment of these esters with triphenylmethyl-sodium,10 on standing in the presence of sodium ethoxide and triphenyl-methane (both of which are by-products of the condensation) it revertsto ethyl benzoate and ethyl isobutyrate, the latter undergoing self-condensation to form ethyl isobutyrylisobutyrate which is convertedinto its sodium derivative.13 These reactions can be followed from theionic equations represented above.
There seems little doubt that the acetoacetic ester condensation is in-fluenced in the first step by the acidic strength of th e ester B and by thebasic strength of the condensing agent,2 in the second step by the rateand position of equilibrium of the reaction of the ester anion with ester,14
and in the third step by the acidic strength of the /3-ketoester and thestreng th of the base. At least with triphenylmethylsodium the first andthird steps are relatively rapid and complete and the second step is
relatively slow. Apparen tly, the influence of stru ctu re on the overallreaction is most pronounced in the second step.14 In general, it may be
12 D i e c kma nn , Ber., 33, 2670 (1900).13 Hudson and Hauser , J. Am. Chem. Soc, 62, 62 (1940).14 Abramovi tch and Hauser , unpubl ished observat ions .
stated that the acetoacetic ester type of condensation will take placewhen a base is formed which is weaker than th a t used as the condensingagent. Th us, in the formation of ethy l acetoacetate from ethy l ace tateand sodium ethoxide, th e eno late' anion, (CH3COCHCO2C2H5) ~, isweaker than the ethoxide ion, and in the formation of ethyl benzoyldi-
methylacetate from ethyl isobutyrate and ethyl benzoate in the presenceof triphenylmethylsodium the ethoxide ion (a by-product of the con-densation) is weaker than the triphenylmethyl ion.
SCOPE AND LIMITATIONS
The acetoacetic ester type of reaction is used to prepare a variety of/3-ketoesters and certain other typ es of compounds. The self-condensa-
tion of esters having hydrogen on the a-carbon atom may be effectedreadily; this amounts to an acylation of the ester by another moleculeof the same ester.
The condensation between two different ethy l esters m ay be indicated
as follows.I I
R C O 2 C J H 5 + H—C—CO2C2HB -> RCOCCO2C2HB + C2H6OH
The first ester may be designated as the acylating ester. This condensa-tion is generally satisfactory only when one of the esters (the acylatingester) has no active hydrogen. The condensation of two esters each ofwhich has active hydrogen atoms may result in the formation of a mix-
ture of four different /3-ketoesters, the two self-condensation productsand the two mixed ester condensation products, although in certain casesone of the latter may be the principal product. Even the application ofthe special technique of first converting one of the esters largely into itssodium enolate by means of triphenylmethylsodium and then condens-ing the enolate with an ethyl ester has not been particularly successfulthus far, as mixtures of /3-ketoesters are still obtained.16 Certain acyla-tions by means of phenyl or diphenyl esters, however, have been suc-
cessful.
14
Three of the more common esters which have no active hydrogen andwhich have served satisfactorily as acylating esters are ethyl formate,
Apparently, not all esters having hydrogen on the a-carbon atomundergo the acetoacetic ester condensation to form /3-ketoesters. Thus,ethyl dichloroacetate when treated with alcoholic sodium ethoxide yieldsethyl oxalochloroacetate diethyl acetal and ethyl diethoxyacetate.17
Although methyl diphenylacetate is converted by triphenylmethylso-dium into its sodium enolate (which may be condensed with acidchlorides to form /3-ketoesters),7 the self-condensation of this esterapparently has no t been effected. The unsaturated esters, ethyl acry-late 18 and ethyl crotonate,19 when trea ted with sodium ethoxide, undergocondensations of the Michael type; however, as indicated above, ethylcrotonate undergoes the acetoacetic ester reaction with certain esters(ethyl oxalate and ethyl formate).
Phenyl ace tate fails to condense in the presence of sodium phenoxide.20
Although purely aliphatic alkyl acetates (in which the alkyl group ismethyl, ethyl, propyl, etc.), undergo the normal acetoacetic ester con-densation when treated with the corresponding sodium alkoxide (or withmetallic sodium), the phenyl-substituted alkyl acetates, benzyl andbenzohydryl ace tates (and also allyl and cinnamyl ace tates) undergo so-called abnorm al acetoacetic ester reactions. Thus, benzyl ace tate withsodium benzyloxide yields only traces of benzyl acetoacetate,21 and whenheated with metallic sodium this ester yields the "alkylated" product,
j3-phenylpropionic acid;21> 22 allyl acetate with sodium undergoes thesame typ e of reaction. Benzohydryl acetate and sodium yield stillanother "abnormal" product, tetraphenylethane;23 cinnamyl acetateundergoes the same ty pe of reaction.23 Benzohydryl acetate with sodiumbenzohydryloxide yields the "alkylated" product, /3,/?-diphenylpropionicacid, and other products.21 It should be pointed out that these so-calledabnormal acetoacetic ester reactions presumably require relatively hightemperatures (100-300°) and that at least benzyl acetate undergoes the
normal acetoacetic ester condensation when treated with triphenyl-methylsodium at room temperatures.24
Side Reactions
The most im por tan t type of side reaction tha t is encountered when theacetoacetic ester condensation is carried out involves the reaction of the
17 Cope, J. Am. Chen. Soe., 58, 570 (1936).18 P e c hma nn a nd R ohm, Ber., 34, 428 (1901).
19 P e e hma nn a nd R ohm, Ber., 33, 3324 (1900).20 Fisher and McElva in , J. Am. Chem . Soc, 56, 1766 (1934).21 Bacon, Am. Chem. J., 33, 68 (1905).22 Conrad and Hodgkinson , Ann., 193, 298 (1878).23 Tseou and Wang, J. Chinese Chem. Soc, 5, 224 (1937).24 Hudson and Hauser , unpubl ished observat ions .
Metallic sodium is capable of reacting with the carbonyl group ofesters to form acyloins
27(RCHOHCOR) and diketones
27(RCOCOR),
but in the presence of excess of ethyl acetate or ethyl propionate, metallic
sodium effects only the acetoacetic ester condensation.28
With ethyl
n-butyrate or ethyl isobutyrate and sodium, however, the acetoacetic
ester condensation does not take place even in the presence of an excess
of the ester; instead, acyloins, diketones, and higher-boiling products are
formed.28
In certain cases, side reactions involving the alcohol portion of the ester
are encountered. Thus, in the presence of potassium amide in liquid
26See Bergstrom and Fernel ius , Chem. Rev., 12, 142-150 (1933); ibid., 20, 459 (1937).
26 Schlenk and Oohs, Ber., 49, 610 (1916).27 Bouveault and Locquin, Bull. soc. chim., [3] 35, 629 (1906).28 Snell and M cE lv a in , J. Am. Chem. Soc, 53, 750 (1931).
274 T H E A C E T O A C E T I C E S T E R C O N D E N S A T I O N
ammonia, /3-phenylethyl acetate is converted partly to styrene.29 Theso-called abnormal acetoacetic ester reactions discussed above (p. 272)may also be regarded as other types of side reactions involving thealcohol portion of the ester.
C yclizations (Dieckmann reaction)
Certain esters having hydrogen on the 5- or e-carbon atom which isactivated (generally by a carbonyl group) undergo intramolecular cycli-zation. These reactions may be illustrated by the formation of 2-car-boethoxycyclopentanone from ethyl adipate.
C H 2
/
CH2
2
\
CH 2
CO2C2H5
2
\
CaO C2H6 CH 2 CO
CH 2 CHCO2C 2H5 + C2HBOH
Similarly, ethyl pimelate can be cyclized to a cyclohexanone derivative,but ethyl suberate affords 2-carboethoxycycloheptanone in low yield.The esters of glutaric, azelaic, and sebacic acids fail to cyclize intramolec-ularly in the presence of sodium ethoxide.
This cyclization has proved particularly useful in preparing polycyclic
comp ounds . Fo r example , the cyclic ketoe s ter which is an interme diatein the synthesis of the sex hormone equilenin can be obtained in practi-t ica l ly quanti ta t ive yie ld.30
>
CHCOj CH,
CH 3O+ CH3OH
Certain intramolecular cyclizations are accompanied by decarboxyl-
ation, i l lustrated as follows.
C2H5 C2H6
C H 2 — C — C O 2 C 2 H 6
N a O C 2H 6
H
C H 2— C — C O 2 C 2 H 6
CO2C2H5
C H 2 — 0 — G O 2 C 2 H 5
c=o
C H j — C — C O 2 C 2 H S
H29 Skell and Hauaer, unpublished observations.30Bachmann, Cole, and Wilds, J. Am . Chem. Soc, 62, 835 (1940).
Fi ve - a n d six-membered r ings m a y also b e formed b y in termolecular
condensa t ion a n d cycl izat ion, examples of which m a y be represented asfollows.
CO 2C 2H 5
C HCH 2CO 2C 2H 6 / \
2 | N a C H 2 COCH 2CO 2C 2H 6 > | | + 2C 2H 6OH
CO C H 2
\ /CH
CO2C2H6
CO2C2H6
CH 2CO 2C 2H 6 CH—CO/ CO2O2H5 | J Q Q XT /
R C H + I *-^ R C H\ CO 2C 2H 6 \
CH 2CO 2C 2H 6 CH—CO
( R = H , C H 3 ,C 6 H 6 ) CO 2C 2H B
CO 2C 2H 6
+ 2C2HBOH
i
CO 2C 2H 6
CO
CHCO 2C 2H 6 + 2C 2H 6OH
The Acylation of Esters with Acid Chlorides
Closely related to the acylation of esters with esters (as occurs in theacetoacetic ester reaction) is the acylation of esters with acid chloridesor anhydrides. For example, ethyl isobutyra te in the form of its sodiumenolate (prepared from the ester and triphenylmethylsodium) may beacylated not only with ethyl benzoate 10 or phenyl benzoate,31 but also
with benzoic anhydride
31
or benzoyl chloride;
31
the reactions with thelast three reagents (especially the one with the acid chloride), beingessentially irreversible, give the best yield of ethyl benzoyldimethylace-tate. These reactions may be represented by the following general equa-
81 Hudson, Dick, and Hauser, J. Am. Chem . Soc, 60, 1960 (1938).
The reactions of the sodium enolates of ethy l isobutyrate and other estersof disubstituted acetic acids with various acid chlorides are of particularvalue for the preparation of a,a-disubstituted |S-ketoesters 16 of the typeRCOCR2CO2C2HS. The acylation of the sodium enolate of ethyl ace-tate with acid chlorides does not stop with monoacylation but producesmainly the diacylated acetate 16 (RCO)2CHCO2C 2H 5.
EXPERIMENTAL PROCEDURES
Choice of Base
The base most commonly used for the acetoacetic ester condensationis the sodium alkoxide that corresponds to the alcohol portion of theeste r; for example, sodium ethoxide is used with ethyl esters. Thesebases are generally readily available and usually they produce no by-products except the corresponding alcohol, which is easily separated fromthe condensation produc t. Under the proper conditions, sodium alkox-
ides effect the condensation of acetates and most esters that have twohydrogens on th e a-carbon atom (in reactions of either two similar or twodifferent ester molecules); two such esters, however, ethyl isovalerate32
and e thyl £-butylacetate,32 as well as esters th a t have only one a-hydrogenatom (e.g., ethyl isobutyrate33) fail to condense in the presence ofsodium ethoxide.
The second most useful base is triphenylmethylsodium, which con-denses not only ethyl acetate u and presumably all esters that are con-
densed by sodium alkoxides, but also certain esters that cannot becondensed by means of the latte r bases. Thus, triphenylmethylsodiumeffects the.self-condensations of ethyl isovalerate 16 and ethyl isobuty-rate 2 and the mixed ester condensations between ethyl isobutyrate andesters with no a-hydrogen, for example, ethyl oxalate.16 Also, tri-phenylmethylsodium is the only base that has been found to be gener-ally satisfactory for the condensations of esters with acid chlorides.16
With the proper equipment, triphenylmethylsodium is readily prepared,and it generally produces no appreciable amounts of by-products except
triphenylmethane, which usually may be separated readily from thecondensation product.
32 Roberts and McElvain, /. Am. Chem. Soc, 59, 2007 (1937).33MoElvain, J. Am . Chem . Soc, 61, 3124 (1929).
Mesitylmagnesium bromide34 effects the self-condensation of ethylisovalerate and ethyl isobutyrate (and also ethyl Z-butylacetate), butthe yields of products are not so high as those obtained with triphenyl-methylsodium. Apparently, mixed ester condensations have not been
attempted with mesitylmagnesium bromide, but it seems likely that atleast certain of them might be effected; however, an attempt to condenseethyl isobutyrate with benzoyl chloride by means of mesitylmagnesiumbromide has been unsuccessful.16
Certain other bases have limited application. Although isopropyl-magnesium bromide is not satisfactory for the self-condensation of ethylacetate or ethyl isovalerate,34 this Grignard reagent does bring about theself-condensations of ethyl phenylacetate 35 (in which the a-hydrogen is
activated by the phenyl group) and of i-butyl acetate24
(in which thecarbonyl group is deactivated by the i-butyl group ). Also, potassiumamide effects the self-condensation of t-huty\ acetate,24 but sodium amidereacts with ethyl acetate to give only a low yield of acetoacetic ester 36
and with ethyl isobutyrate to give little or none of the /3-ketoester.37
Sodium amide is satisfactory, however, for the cyclization of ethyl adi-pate 38 (and especially for various ketone-ester Claisen condensations).26
Sodium n-amylacetylide, N aC = C — (CH 2)4CH 3, has been used for theself-condensation of certain esters.39
In general, the appropriate sodium alkoxide would be chosen for acondensation if it is capable of effecting the reaction; if not, triphenyl-methylsodium would be chosen unless the triphenylm ethane produced isdifficult to separate from the condensation product, and in that case,mesitylmagnesium bromide would be tried . In special cases, otherbases may be chosen; thus for the self-condensation of ethyl phenylace-tate, isopropylmagnesium bromide35 would be used instead of sodiumethoxide, 32 since a considerably better yield of condensation product is
obtained with the Grignard reagen t. For the self-condensation off-butyl acetate, triphenylmethylsodium,14 potassium amide,24 or iso-propylmagnesium bromide24 may be chosen instead of sodium i-butox-ide, since the first two bases give as good or bet ter yield of condensationproduct and this particular sodium alkoxide is rather difficult to pre-pare; 20 the yield of product with the Grignard reagent is slightly lowerthan yields obtained with the other bases.
34
Spielman and Schmidt , J. Am. Chem. Soc, 59, 2009 (1937).36 Conant and B la t t , J. Am. Chem . Soc, 51, 1227 (1929).36 Titherly, J. Chem. Soc, 81, 1520 (1902); Freund and Speyer , Ber., 35, 2321 (1902).37 Scheibler and Stein, J. prakt. Chem., 139, 107 (1934).38 Haller and Cornubert , Bull. soc. chim., [4] 39, 1626 (1926); Compt. rend., 179, 315
(1924).39 Moureu and De Lange , Bull, soc chim., [3] 27, 378 (1902).
Procedures for condensations using the two most generally applicablebases, sodium alkoxides and tripheny lmethylsodium , are described below.
S election of Experimental C onditions with Sodium Alkoxides
A variety of experimental conditions have been used in acetoaceticester condensations brought about by sodium alkoxides. In general, thebasic procedure involves the reaction of the ester or ester mixture withthe sodium alkoxide under a reflux condenser. Th e time and tempera-ture of reaction vary greatly with different esters, ranging from severalminutes to a few days at tem peratu res from 25° to 140°. The reactionmixture is generally neutralized in the cold with dilute acetic or sulfuricacid, and the condensation product isolated, dried, and distilled invacuum, or, if solid, recrystallized.
Anhydrous alcohol-free sodium alkoxides are to be preferred for mostcondensations, although in certain reactions the presence of a littlealcohol apparen tly does not decrease the yield appreciably. E thy lethoxalylacetate (sodium salt) is prepared commercially from ethyl ace-tate and ethyl oxalate using an alcoholic solution of sodium ethoxide.40
It is convenient to generate the sodium alkoxide in the reaction mixtureby means of metallic sodium, but this procedure apparently is satisfactoryonly for the condensations of ethyl acetate and ethyl succinate (and
possibly ethyl propionate) 41 with themselves or with certain other esters,and for certain cyclizations. Generally sodium is used in the form ofwire or powder.
With alcohol-free sodium alkoxides, esters should be pure and dry.When metallic sodium is used the ester should contain a little bu t not toomuch alcohol; except for the alcohol, the ester should be pure and dry.The apparatus should be dry and protected from moisture of the air bymeans of a calcium chloride tube or a soda-lime tube . When a stirrer is
used, it should be provided with a mercury seal. Ordinarily, no specialprecautions are taken to exclude atmospheric oxygen. M any condensa-tions (especially self-condensations) are carried out w ith no solvent othertha n the ester, which may be present in considerable excess. Other con-densations (especially cyclizations) are carried out in dry ether, benzene,or toluene.
Certain departures from the basic procedure have led to improvedresults. When the self-condensations of higher homologs of ethyl
acetate are carried out in the presence of sodium ethoxide, removal (bydistillation under reduced pressure) of the alcohol formed during the10 Private communication from W. L. Johnson, U. S. Industrial Chemicals, Inc.,
Baltimore, M d.11 See reference 28, p . 755, and reference 33, p . 3130.
is diluted with twice its volume of the inert solvent and added dropwiseto the vigorously stirred contents of the flask. When the initial reactionsubsides, nearly all the sodium has reac ted. The mixture is then refluxedwith continuous stirring until the sodium has completely disappeared.The solvent may then be distilled, the last traces being removed under
reduced pressure.Sodium m ethoxide and sodium ethoxide may be prepared in this m an-
ner, but the method is not satisfactory for sodium alkoxides higher thansodium ethoxide; methods of preparation of higher alkoxides are de-scribed in the literature.20
Ethanol-free sodium ethoxide may also be prepared by adding freshlycut sodium to an excess of absolute ethanol contained in a round-bot-tomed flask which is imm ediately connected to a condenser set downward
for distillation and to a source of dry nitrogen; a filter flask to which asoda-lime tube is attached is used as a receiver. When the reaction hasceased the excess ethanol is removed by distillation. Dry nitrogen isthen admitted and the flask is heated in an oil bath at 150°/20 mm. forone hour. Before use the w hite cake of sodium ethoxide should be pul-verized by stirring or shaking in an atmosphere of nitrogen .
S elf-C ondensation of Various Alkyl A cetates,2 0 Ethyl Propionate,32
and Ethyl Butyrate32 in the Presence of Sodium Alkoxides. T he self-
condensation of ethyl acetate by means of sodium is described in detailin Organic Syntheses.*8 The following procedure, involving sodium
alkoxides, may be applied to a variety of esters of acetic acid as well asto the ethyl esters of propionic and butyric acids.
In a 500-cc. three-necked flask, fitted with a stirrer, reflux condenser,and a thermom eter which dips below the surface of the reaction mixture,are placed 0.2 mole of the alcohol-free alkoxide and 1.2 moles of thecorresponding ester. The contents of the flask are heated with stirringto the tem pera ture and for the time indicated in Table I. At the end ofthe reaction time the flask is surrounded by ice and the reaction mixtureis cooled to 10°. The reflux condenser is replaced by a dropping funnel,and 36 g. of 33.3 % aqueous acetic acid is added dropwise to the mixtureat such a ra te th a t the tempera ture remains below 15°. When the solidmaterial has completely dissolved, the ester layer is separated and theaqueous layer is extracted w ith four 50-cc. portions of ether. The com-bined ester layer and ether extracts, after drying over anhydroussodium sulfate, is fractionally distilled. The conditions of reaction, the
maximum yields, and the boiling points of various /3-ketoesters obtainedare given in Table I.
« Inglis and Roberts.'Org. Syntheses, Co ll. Vol., 1, 230 (1932).
a A higher yield of the 0-ketoester is obtained by periodic distillations of portions of the ester togetherwith the alcohol that is formed during th e reaction (see following procedure), but considerable excessof the pure ester is required in the process.
6 Reaction mixture heated at 115° for four hours and then allowed to stand at room temperature fortwelve hours.
Forced S elf-C ondensation of Ethyl Esters of n-Valeric ** and H igher
Aliphatic Acids42 in the Presence of Sodium Ethoxide. In a 125-cc.modified Claisen flask with a fractionating side arm 35 cm. long areplaced 0.1 mole of th e purified ester and 0.05 mole of ethanol-f ree sodiumethoxide (prepared from absolute ethanol and powdered sodium underdry ether, p. 279). The reaction flask is attached to the receiving flask(which is not cooled), and this flask in turn is attached through a soda-lime tower and a safety bottle to a manometer and a water pum p. The
safety bottle contains a stopcock which can be opened to the air and bywhich the pressure in the system can be regulated. The reaction flask isthen heated carefully in an oil bath to a temperature and under a pres-sure that cause a moderate, but not too vigorous, evolution of ethanolvapor as shown by the ebullition of the reaction mixture. The requiredtemperature and pressure vary with the boiling point of the ester, themore volatile ones requiring lower reaction temperatures and higherpressures in order to avoid loss of ester. Consequently the time neces-
sary for the completion of the reaction in these cases is increased. Asummary of the conditions for the reaction of the various esters is givenin Table II . Column 2 shows the temperatures and column 3 the pres-sures which are most satisfactory at the beginning of each reaction toensure a moderate evolution of alcohol. After the reaction has pro-
ceeded for some time the temperature and pressure can be raised andlowered, respectively, without appreciable loss of ester. Column 4 givesthe time required for completion of the reaction; at the end of thisperiod the reaction mass ceases ebullition. The reaction produc t aftercooling is treated with the calculated quantity of 30% acetic acid and
shaken vigorously until the sodium sa lt is completely decomposed. Theketoester is then extracted with 25 cc. of benzene, and the resulting ben-zene solution, after washing with water, is dried over anhydrous sodiumsulfate. Th e benzene is removed by distillation. E thyl a-lauryllaurateand ethyl a-my ristylmyristate are recrystallized from absolute methanol.The liquid products are purified by distillation. This procedure is quitesatisfactory for all the ketoesters except ethyl a-pelargonylpelargonateand ethyl a-caprylcaprate, both of which suffer a small amount of
pyrolysis to the corresponding ketone, which appears as a low-boilingsolid fraction in the distilla te. The yields and boiling (or melting) pointsof the /3-ketoesters are shown in Table II (last column).
TABLE II
CONDITIONS AND TIME REQUIRED FOB FORMATION OP /3-KBTOBSTERS
C ondensation of Two Different Este rs 4 6 in the Presence of Sodium.
Preparation of Ethyl -yiV-Diethoxyacetoacetate43 and Ethyl Benzoyl-
acetate.40' 47 In a three-necked flask fitted with a stirrer, a dropping46 For a detailed procedure for the condensation of ethyl oxalate with ethyl propionate
in the presence of sodium ethoxide, see Cox and McElvain, Org. Syntheses, 17, 54 (1937).47 Yuoh Fong Chi and Yung Mao Lee, Trans. Science Soc. China, 8, 87-89 (1934).
funnel, and a reflux condenser carrying a calcium chloride tube , 66 g.(0.49 mole) of ethyl diethoxyacetate is heated to 85-90° and portions of2 g. of sodium wire and 9 cc. of ethyl acetate are added at half-hour inter-vals until 34 g. (1.5 atoms) of sodium and 130 g. (1.5 moles) of ethyl
aceta te have been introduced. The reaction is quite vigorous at first,but after it subsides the sodium and ethyl acetate can be added a littlemore rapidly; the seventeen additions can be made in about six hours.The brown, viscous reaction mixture is stirred continuously, and stirringand heating at 85-90° are continued for four hours after the last addi-tion of sodium and ethyl ace tate . Ethano l (30 cc.) is added to dissolvethe residual sodium, and then the oil, cooled somewhat but not allowedto become too viscous, is poured into a mixture of 130 cc. of concen-
tra ted hydrochloric acid and 130 g. of ice. The oily layer is imm ediatelyseparated, and the aqueous layer is extracted once with a small quantityof ether. The oily layer and ether extract are combined, washed withsodium carbonate solution, dried, and the ether and ethanol distilled on abath a t 100°. Fractional distillation of the residue gives 76 g. (71%) ofethyl 7,7-diethoxyacetoacetate boiling at 112°/4-6 mm. A considerableamount of ethyl acetoacetate passes over in the fore-run, along withsome ethyl diethoxyacetate.
By a similar procedure ethyl benzoylacetate is obtained in 55-77%yield from ethyl acetate, ethyl benzoate, and sodium.47 Ethyl benzoyl-acetate is prepared commercially essentially in this manner 40 in a yieldof 6 8% ; much e thyl acetoacetate is also obtained in the same reaction.40
By a similar procedure methyl benzoylacetate is obtained in 45-85%yield from methyl acetate, methyl benzoate, and sodium.40' 48 Themethod is not very satisfactory, however, for the acylation of ethylacetate with its purely aliphatic homologs.48
S elf-C ondensation F ollowed by C yclization.49 Preparation of Ethyl
Succinylsuccinate by the Use of Sodium Ethoxide m or Sodium.6 1 T o29 g. (0.43 mole) of ethanol-free sodium ethoxide covered with 140 cc. ofdry ether is added 38 g. (0.21 mole) of e thyl succinate. The m ixture isrefiuxed three or four days. The ether is then distilled and the residue isneutralized in the cold with dilute sulfuric acid. The crude crystallineester is collected and washed with water. I t is dissolved in 200 cc. of
49 For the eyclization of a number of esters of polyfunctional acids, see Dieokmann,Ann., 317, 51 (1901); for a detailed procedure for the cyclization of ethyl adipate, see(a) Pinkney, Org. Syntheses, 17, 30 (1937), and (6) Linstead and Meade, J. Chem. Soc,940 (1934).
"Piutti, Gazz. chim. ilal., 20, 167 (1890).
" Upenski and Turin, Chem. Zentr., I l l , 754 (1923).
95% ethanol, decolorized with 1 g. of charcoal, and allowed to crystal-lize. The yield of ethy l succinylsuccinate, m .p. 126-127°, is about 60% .
When sodium is used, the procedure involves the addition of a slightexcess of powdered 62 sodium (27 g., 1.17 atom) to ethyl succinate (75 g.,0.43 mole) containing a small am ount (4 cc.) of absolute ethanol. Afterthe initial reaction, which may require cooling to prevent flooding of thereflux condenser, the mixture is heated to 60° for five hours, then to100° for two hours, and finally to 110° for twenty-five hours. I t is thencooled, added cautiously to cold dilute sulfuric acid, and worked up asjust described; the yield is about 60% .
C ondensation of Two Different E sters Followed by C yclization.Preparation of SjS-Dicarboethoxycyclopentanedione-l^.
63 To 34 g.(0.5 mole) of ethanol-free sodium ethoxide, covered with 200 cc. of
anhydrous ether and contained in a flask fitted with a reflux condenser,is added 36.5 g. (0.25 mole) of ethy l oxa late. After mixing thoroughly,47 g. (0.25 mole) of ethyl glutarate is added over about fifteen minutesand the mixture is heated to refluxing. After approxim ately one hour,when solution is complete, the ether is distilled and the residue is heatedto 120-130° un til it changes to a yellow solid (about three hours). Th ereaction mixture is cooled and washed with ice-cold dilute sulfuric acid(10%), then with ice-water. After drying in the air (twenty-four hours)
the crude product, m .p. 90-104°, weighs 43 g. I t is recrystallized from80 cc. of 95% ethanol, and 30 g. (50%) of pure material (m.p. 115°) isobtained.
It is reported that ethyl /3-methylglutarate condenses with ethyl oxa-late to give an almost quantitative yieldB3 of 4-methyl-3,5-dicarbo-ethoxycyclopentanedione-1,2, melting at 108°, and that ethyl /8-phenyl-glutarate with ethyl oxalate gives an excellent yield 63 of the 4-phenylderivative, m .p. 160-161°. E thyl /S,/°-dimethylglutarate with ethy loxalate gives only a low yield 63 of the 4,4-dimethyl derivative by thisprocedure, but a considerably better yield is obtained using the corre-sponding methyl esters and sodium methoxide.64
S election of Ex perimental C onditions with T riphenylmethylsodium
The first step in procedures for carrying out self-condensations ofesters, mixed ester condensations, or ester acid chloride condensations bymeans of triphenylmethylsodium consists in converting the ester to be
acylated into its sodium enolate. This is done simply by adding the
less) and at an initial tempera ture of approximately 20° or less. W ithmore concentrated solutions or when the room temperature is high, thereaction mixture should be cooled by means of an ice bath.
Triphenylmethylsodium is conveniently prepared in almost quanti-tativ e yield (90% ) by shaking a solution of pure triphenylchloromethane(m.p. 112-113°) in dry ether with an excess of freshly prepared sodiumamalgam. Since the base reacts readily with active hydrogen compounds(water, ethano l, etc.) and with oxygen, the materials should be pure andthe base should be prepared and used in an atmosphere of dry nitrogen.The base is commonly prepared and used in approximately 0.15 molarconcentrations; however, concentrations up to 0.5 molar have beenemployed.
Procedures have been chosen to illustrate the preparation of tri-
phenylmethylsodium, the self-condensation of an ester, and a mixedester condensation. An ester acid-chloride condensation is described indetail in Organic Syntheses;
8 the reaction on a larger scale is described inthe literature.15
Procedures
Triphenylmethylsodium.55116 Nine hundred and fifty grams of 1.5%sodium amalgam is prepared in the following manner. In a 250-cc. Pyrex
Erlenmeyer flask 14 g. (0.61 atom) of freshly cut sodium is covered to adep th of 2 cm. with high-boiling mineral oil. The flask is heated untilthe sodium begins to melt. Then 935 g. of mercury, contained in a sepa-ratory funnel whose stem passes through a cardboard shield (8 cm.square), is added rapidly to the molten sodium (hood!). The flask isstoppered and shaken until no solid particles of amalgam remain. Whenthe flask has cooled to approximately 80°, or when the amalgam firstbegins to crystallize, the flask is cooled rapidly to room temperature byswirling in cold wa ter. The oil is decanted, and the am algam (950 g.) iswashed twice with dry benzene or ligroin.
To a mixture of 70 g. (0.25 mole) of triphenylchloromethane (m.p.112-113°) and 950 g. of freshly prepared 1.5% sodium amalgam in a 2-1.Pyrex glass-stoppered bottle , 1500 cc. of absolute ether is added. Theglass stopper is lubricated with a little Lubriseal and firmly inserted.The bottle is clamped securely in a mechanical shaker which makes a4- to 5-in. stroke and three to four strokes a second. Shaking is begun;if the temperature of the bottle rises above approximately 40°, shaking
is interrupted until the bottle cools somewhat. The charac teristic deepred color appears after five to fifteen minutes' shaking. After shakingfor three to six hours the bottle is cooled to room temperature and
66 Renfrow and Hauser, Org. Syntheses, 19, 83 (1939).
ing; little heat is generated after the appearance of the color. Shakingis continued until no pieces of solid amalgam remain, and then for twohours longer. The bo ttle is cooled, removed from the shaker, andallowed to stand, as described above. The solution is then analyzed, byremoving a 10-cc. aliquot and diluting with 25 cc. of ether before extrac-tion and titration.
Except when alkylations are to be carried ou t, it is frequently permissi-ble to use the solution of triphenylmethylsodium without separating itfrom the sludge of sodium chloride and amalgam . The to tal volumeof solution may be considered to be equal to the volume of the etheremployed plus 0.77 cc. per g. of triphenylchloromethane used. When thesolution is not separated from the amalgam and when the above volumecorrection is applied in the calculation of the quantity of base available,
yields of 85-93% of the theoretical amount are obtained.S elf-C ondensation. Ethyl a-Isovalerylisovalerate.16 To a solution of
0.21 mole of triphenylmethylsodium in approximately 1400 cc. of ethercontained in a 2-1. Erlenm eyer flask, is added 31.8 cc. (27.5 g., 0.21 mole)of ethyl isovalerate (b.p. 134-135°). The flask is stoppered well, shakento effect complete mixing, and allowed to stand at room temperature forsixty hours. The reaction mixture is then acidified by the addition, withshaking, of 15 cc. (approxim ately 0.25 mole) of glacial acetic acid. The
mixture is extracted with 100 cc. of water. The resulting ether solutionis washed with 50-cc. portions of 10% sodium carbonate solution untilfree from excess acid. The ether solution is dried by shaking withanhydrous sodium sulfate and allowing to stand over Drierite. Thesolution is filtered and the ether distilled on a water ba th. The residueis distilled in vacuum . The fraction boiling up to 170°/15 mm. is redis-tilled through a 6-in. Widmer column, and the fraction boiling at118-119°/15 mm. is collected. The yield of ethyl a-isovalerylisovalerateis 13.3 g. (63%).
M ixed E ster C ondensation. Ethyl a-Ethoxalylisobutyrate.16 To a solu-tion of 0.205 mole of triphenylmethylsodium in approximately 1400 cc.of ether, contained in a 2-1. Erlenm eyer flask , is added 27.3 cc. (23.8 g.,0.205 mole) of ethyl isobu tyra te (b.p. 111-112°). The flask is stoppered,shaken, and allowed to stand. After five minutes, 27.8 cc. (30 g., 0.205mole) of ethyl oxalate (b.p. 72-74°/10 mm.) is added slowly and withshaking. The reaction is vigorous, and the mixture may boil gently.After standing for ten minutes at room temperature, the reaction mix-
ture is acidified with 15 cc. of glacial acetic acid and extracted with100 cc. of water. The resulting ether solution is washed free from excessacid with 50-cc. portions of satura ted sodium bicarbonate solution. Theether solution is dried by shaking with anhydrous sodium sulfate and
allowing to stand over Drierite. The solution is filtered and the etherdistilled on a water bath . The residue is distilled in vacuum, and thefraction boiling up to 200°/50 mm. is fractionated through a 6-in.Widmer column. The yield of ethy l a-ethoxalylisobutyrate (b.p. 122-
123/15 mm.) is 27.2 g. (61%).
EXAMPLES OF THE ACETOACETIC ESTER TYPE OF CONDENSATION
In Table III are listed self-condensations of esters; in Table IV, con-densations between different esters; in Table V, intramolecular cycliza-tions; in Table VI, intermolecular condensations and cyclizations; inTable VII, ester-acid chloride condensations.
Ethyl /3-methyladipateEthyl pimelateMethyl ester of /?-7-methoxy-2-methyl 2-carboxy-
1,2,3,4-tetrahydrophenanthrene-l-propionic acidEthyl glutarateEthyl suberateEthyl azelateEthy l sebacateEthyl a-carboethoxy-a'-ethyladipate
Ethyl a-ethyl-a, a'-dicarboethoxyadipate
Ethy l a, a'-dicarboethoxyadipate
Methyl a-methyl-^-ethylacrylidenemalonate
Condensing Agent
Sodium
Sodium amideSodiumSodiumSodiumSodium methoxide
SodiumSodiumSodiumSodiumSodium ethoxide
Sodium ethoxide
Sodium ethoxide
Sodium hydroxide(ale. soln.)
Product
2-Carboethoxycyclopentanone
2-Carboethoxycyclopentanone2-Carboethoxy-5-methylcyclopentanone2-Carboethoxy-4-methylcyclopentanone2-CarboethoxycyclohexanoneMe thyl ether of d,i-16-carbomethoxy-
of acetonitrile by esters and subsequent alcoholysis of the ketonitrile;(3) theacylation of ethyl acetoacetate w ith acid chlorides or anhydridesand the subsequent ammonolysis or alcoholysis of the product; (4) thereaction of ethyl cyanoacetate w ith Grignard reagen ts; (5) the hydration
of a,/3-acetylenic acids and esterification; (6) the acylation of methyl ke-tones with ethyl carbonate; and (7) the oxidation of /3-hydroxyesters(p. 11).
The acylation of ethyl acetate by another ester (method 1) consists in
a mixed ester condensation, which, as already pointed out (p. 270), isin general satisfactory only when the acylating ester has no activehydrogen. The acylation of acetonitrile with esters (method 2)
80'88
appears to have had a somewhat more limited use than method 1. Thenitrile and ester are condensed by means of sodium ethoxide (or tri-
phenylmethylsodium) 89 and the resulting /3-ketonitrile alcoholized.80' 88
RCO2C2H5 + CH 3CN -> RCOCH3CN -> RCOCH2CO2C2H6
Method 3 may be represented as follows.
CH3COCH2COC2H5 >CH3COCHCO2C2H6 ^ RCOCH2CO2C2H6
RCO
The acylation of ethyl acetoacetate (in the form of its sodium enolate)is readily carried out with acid chlorides or anhydrides,90 and the
ammonolysis (or alcoholysis) of the acyl acetoacetic ester at least inseveral cases gives good yields of the desired acyl acetate.91 However,ethyl propionylacetoacetate on ammonolysis gives a mixture of ethylpropionylacetate and ethyl acetoacetate which is difficult to separate.91
The Grignard reagent may react not only with the cyanide group, but
also w ith the ester group and with theactive hydrogens, resulting in mix-
tures of products. It has been shown that 1mole of ethyl cyanoacetateis capable of reacting with 4 moles of Grignard reagent.93
It should be
pointed out, however, that the /3-ketoester is not contaminated withethyl acetoacetate as sometimes happens with methods 1 and 3.
88
Cox, Kroeker, and McElvain, / . Am. Chem. Soc, 56, 1172 (1934).89Abramovitch and Hauser, unpublished observations.
90 Bouveaul t and Bonger t , Bull. soc. chim., [3] 27, 1046 (1902).81 Bouveaul t and Bonger t , Bull. soc. chim., [3] 27, 1089 (1902)." B l a i s e , Compt. rend., 132, 978 (1901).93 Brekpot , Bull. soc. chim. Belg., 32, 386 (1923).
In general, the hydration of the acetylenic acids appears to give goodyields of /3-ketoacids,
94but the esterification of the latter may be diffi-
cult. The use of the method is somewhat limited by the fact that the
acetylenic acids or hydrocarbons are generally not readily available.
Method 696
may be represented as follows.
RCOCHa +NaOC2H6
RCOCH2CO2C2H6
This method consists in heating or digesting the ketone with sodium or
potassium ethoxide (orother alkoxide) in a large excess of ethyl carbonate
(or other alkyl carbonate). This direct method appears to be very satis-
factory for the synthesis of several of the higher acylacetates, but it is
not satisfactory for the synthesis of ethyl propionylacetate or ethyl
isobutyrylacetate.96
In Table VIII are collected the yields that have been reported in the
preparation of typical /3-ketoesters by these methods. The question
T AB L E VI I I
PERCENTAGE YIELDS OF ETHYL ACYLACETATES RCOCH2CO2C2H5 BY
VARIOUS METHODS
Acyl Group(RCO)
CH 3CH 2CO—
CH 3(C H 2)2CO—(C H 3)2CHCO—CH 3(C H 2)3CO—(CH 8)2CHCH 2CO—CH 3(C H 2)4CO—CH 3(CH 2)6CO—CH 3(CH 2)8CO—
C6H6CO—
Method1
11
19-22—
18
"Poor"—
22
—
5 5 - 7 0 d
Method2
_
1 7 "36——.———
42
Method3
10-12
75 "——
"Excellent" °"Excellent" "
——
49-58*
Method4
10-60
40—15—.———
—
Method5
?———
50-80?
76(crude)
Method6
?
—Poor—
60657 4 "
—
60
o Yields given are for the methyl acylacetates.b Over-all yield for both acylation andammonolysia.c Yield for ra-propyl acylacetate.d With sodium ethoxide as condensing agent theyield is 37%.S0
mark indicates th at the method was used bu t tha t no yield was reported.I t should be noted t ha t th e yields given under method 3, with the excep-tion of ethyl benzoylacetate, are for the ammonolysis reaction only anddo not include the yields obtained in the preparation of the acyl aceto-
acetic esters. The yields given under method 1 are those obtained under"special conditions" with sodium as condensing agent and are calculatedwithout taking into consideration the quantities of starting materialsrecovered unchanged.
It can be seen from Table VIII that only one of the /J-ketoesters listed,ethyl (methyl, in method 3) n-butyrylacetate, has been prepared by atleast five of the methods. This compound appears to be best preparedby method 3; however, the 75% yield does not include the acylation ofethyl acetoacetate.91 Ethyl isobutyrylacetate has been prepared in
fairly good yield 88 by method 2, while several of the higher aliphatic/3-ketoesters have been prepared satisfactorily by methods 3,91 5,94 or6.95 Ethyl benzoylacetate has been prepared satisfactorily by methodsI,47 2,80 3,96 and 6,95 the Organic Syntheses method 96 being basically thesame as 3, and the commercial method M basically the same as 1.
None of the m ethods described above appear to be satisfactory for thepreparation of ethyl propionylace tate. One investigator reported ayield of 55% 62 using method 4, but another obtained only a 10-12%
yield93 by this method. Although Fischer and Orth 97 record a y ield of60% by method 4, they point out that the preparation is inferior tomethod 3, in which the yield is only 10-12% . E thyl propionylacetatehas been obtained in fair yield (44%) by condensing the sodium enolateof ethyl acetate (prepared by means of triphenylmethylsodium) withp-diphenyl propionate.14 In a similar manner, n-amyl propionylacetatehas been obtained from the sodium enolate of n-amyl acetate and phenylpropionate.14 In both cases essentially pure products were obtained;
apparently the only disadvantage of the method is that a relativelylarge amount of triphenylmethylsodium is required. E thyl propionyl-acetate has been prepared also from the sodium enolate of ethyl acetateand a large excess of propionyl chloride, but the yield was only 15%, themain product (39% yield) being the dipropionylacetate.15 The latteron ammonolysis according to the second step of method 3 gave ethylpropionylacetate in a yield of 50%.16
( E ) R C O C H R C O 2C 2H 5. 1. Special Case: R CH2CO CHR CO2C2H 5,
in Which the T wo Groups, R , Are the Same.Most /3-ketoesters of thiskind are best prepared by the self-condensation of esters of the type
RCH2CO2C2H5 or by th e action of ethanol on the appropriate diketene.96 Shriner, Schmidt, and Roll , Org. Syntheses, 18, 33 (1938)."Fischer and Orth, "Die Chemie des Pyrol les ," Vol . I , p . 404 (1934), Leipzig.
compounds of the type RCOCR2CO2G2H5 consists in condensing theappropriate ester, in the form of its sodium enolate, with a suitable acidchloride.16 The yields are high (50-75%), and the products are of highpurity.
(G) Miscellaneous P-Ketoesters. Ethyl ethoxalylacetate and ethylformylacetate and their homologs are probably best prepared by mixedester condensations (see p . 271). Also a number of cyclic /3-ketoestersare probably best prepared by ester-ester condensations (see pp. 274,275).
The Use of Secondary Amines 307With Ketones 308With Aldehydes 309With Acids and Esters 310With Phenols 311
With Acetylenes 311With a-Picolines and Quinaldines 312
The Use of Primary Amines 312With Ketones 312With Aldehydes 313With Acids and Esters.' 313With Phenols and Acetylenes 314With a-Picolines and Quinaldines 314
The Use of Ammonia 315
With Ketones 315With Acids 316
RELATED REACTIONS 316
APPLICATION OF THE MANNICH REACTION IN SYNTHESIS 318
Unsaturated Compounds 318Preparation of Ethylenic Compounds 318Preparation of Pyrazolines 319Use of a Mannich Base as a Source of Unsaturated Ketone for Condensa-
tions with an Active Methylene Compound 320Conversion of a Ketone to Its Next Higher Homolog 322
Syntheses Dependent on the Active Methylene Group in the Aminoketone . 322
Syntheses Dependent on the Activity of the Dimethylamino Group inDimethylaminomethylphenols 323
The Mannich reaction consists in the condensation of ammonia or aprimary or secondary amine, usually as the hydrochloride, with formal-dehyde and a compound containing at least one hydrogen atom ofpronounced reactivity . The essential feature of the reaction is thereplacement of the active hydrogen atom by an aminomethyl or sub-stituted aminomethyl group. The product from acetophenone, formal-dehyde, and a secondary amine salt is an example. In the equation thereactive hydrogen atoms are underlined.
The product from a methyl ketone contains reactive hydrogen atoms,and in some cases it is possible to carry the reaction one step further,yielding a compound with two basic groups.
CeHsCOCHaNRa-HCl + CH2O + R2NH-HC1 ->
C6H6COCH(CH2NR2-HC1)2+ H2O
If the substance used in the condensation contains reactive hydrogenatoms on two or more different carbon atoms, then substituted amino-methyl groups may appear at different points in the molecule, leadingto a mixture of isomers.
If the condensation is effected with a primary amine or its salt, the
product is a secondary amine.
C 6H 6COCH 3 + CH2O + R N H 2 • HC1 -> C6H 6COCH 2CH 2N HR • HC1 + H2O
In many cases the resulting secondary amine reacts further to yielda ter t iary amine.
CaHsCOCHs + CH2O + C 6H 6C O C H 2CH 2NHR-HC1 -»
(C 6H B C O C H 2C H 2)2NR-HC1+ H2O
Frequently such products , derived from two molecules of ketone, twomolecules of formaldehyde, and one molecule of primary amine, are
unstable and readily undergo cycl ization. Th e comp ounds obtainedfrom acetone, formaldehyde, and methylamine are i l lus tra t ive . 1
CH 3
2CH 3COCH 3 + 2CH 2O + CH 3N H 2-HC1
0=0
CH 3COCH 2 C H 2
ICH 2
/N - H C 1
C H 3
HO CH 3
\ /
/ C \CH 3COCH CH 2
I I
C H 3 CH 3
I
CH 2
\
IC H 2
CH3COC CH 2 CH 3COCH CHI I a n d I ICH2 CH2 CH 2 CH2
/ \ /N - H C 1
IC H 3
/N - H C 1
IC H 3
/N-HC1
IC H 3
The product to be expected f rom a Mannich react ion involving an
amm onium sa l t is a pr im ary amine. In ma ny cases , the pr imary amine
so produced reacts further, as above, to form a secondary amine, a
tert iar y amine, or a cyclic subs tance . T he situatio n is further com-
plicated by the fact that methylamine, produced from the ammoniumsalt and formaldeh yde, also tak es p a rt in th e reaction. F or example,
the compounds shown above as products of acetone, formaldehyde, and1(a) Mannich and Ball, Arch. Pharm., 264, 65 (1926); (6) Mannich and Bitsert, ibid.,
methylamine hydrochloride are also obtained from acetone, formalde-hyde, and ammonium chloride.16
The first observation of a condensation of the type now known as theMannich reaction was made by Tollens,2 '3 who isolated the tertiaryamine from ammonium chloride, formaldehyde, and acetophenone.Later Petrenko-Kritschenko 4 and his students studied condensationsof this kind but failed to recognize the reaction as a general one. Thedetailed study by Mannich, begun in 1917, was initiated by the obser-vation that antipyrine salicylate, formaldehyde, and ammoniumchloride reacted to form a tertiary amine.6
3C,H»Ni
C H 3
N — C C H 3
2 311 + 3CH 2O + NH 4C1 •
C H 3
N CCH 3
C 6H 6N\
C H 2— N -H C 1
Since Aminopyrine (Pyramidon, 4-dimethylaminoantipyrine) failedto react, it was evident that the reaction involved the hydrogen atomof carbon 4 of antipyrine.
The mechanism of the Mannich reaction has not been established.The addition of the amine to formaldehyde has been considered as apossible primary step.
R
R
NH + CH20
R
R
NCH2OH
The fact that, in the case of antipyrine, the reaction of dimethyl-aminomethanol gives a poorer yield of condensation product tha n eitherformaldehyde and the amine or formaldehyde and the amine hydro-chloride indicates that this view is not correct.6 The possibility thatthe initial step is the formation of the methylol from the ketone hasbeen examined.
RCOCHs + CH 20 -» RCOCH2CH 2OH
2
van Marie and Tol lens , Ber., 36, 1351 (1903).3 Sohafer and Tollens, Ber., 39, 2181 (1906).4 Pet renko-Kr i t schenko and co-workers : (a) Ber., 39, 1358 (1906); (6) Ber., 41, 1692
The methylols of acetone and cyclohexanone do condense with dimethyl-amine to give the expected products. However, the methylol fromantipyrine does not react at all with dimethylamine.6 Apparentlyneither of these processes represents the primary step of the Mannich
reaction.THE SCOPE OF THE MANNICH REACTION
The Use of Secondary Amines
The secondary amines which have been used successfully are listedin Table I.
Dimethylamine is very reactive and usually leads to excellent yields.Diethylamine appears to be less reactive; it has been reported 7 thatthe typical condensation does not take place with ethyl methyl ketone,diethylamine, and formaldehyde. On the other hand, formaldehydeand this amine do give normal products with acetone,8 benzal-acetone,9 acetophenone,10 and several derivatives of the last.11'12 I thas been reported that 2-acetylfuran and formaldehyde react normallywith salts of dimethylamine, dipropylamine, di-n-butylamine, anddiethanolamine, but not with the salt of diethylamine.13 In other caseswhere dimethylamine, diethylamine, and dipropylamine have givengood results, di-n-butylamine and diethanolamine have failed to react.13
The cyclic secondary amines mentioned above generally react about aswell as dimethylam ine. However, dicyclohexylamine 14 and tetrahy-droquinoline Ui 1B are said not to take part in the reaction.
7 Kermack and M uir, / . Chem. Soc, 3089 (1931).8 du Feu, McQuillin, and Robinson, J. Chem.Soc, 53 (1937).9
Mannich and Schutz, Arch. Pharm., 265, 684 (1927).10 Blicke and Burckhalter, J. Am. Chem. Soc, 64, 451 (1942).11 Mannich and Lammering, Ber., 55, 3510 (1922).12 Mannich and Dannehl, Arch. Pharm., 276, 206 (1938).13 Lewy and Nisbet, J. Chem . Soc, 1053 (1938).14 Burger and Bryant, J. Am. Chem . Soc, 63, 1054 (1941).16 Burger and Mosettig, J. Am . Chem . Soc. 58, 1570 (1936).
With Ketones. Saturated ketones, cycloalkanones, a,/3-unsaturatedketones, aliphatic aromatic ketones, including those in which thearomatic ring is heterocyclic, and certain heterocyclic ketones con-taining a carbonyl group in the ring all undergo the Mannich reaction
with secondary amines, usually in good yields.In Table II are listed ketones which have been treated with formal-dehyde and salts of secondary amines with the successful formationof a /J-dialkylaminoketone. In the formulas the replaceable hydrogenatom is underlined. A detailed list of the Mannich reactions involvingthese ketones is given in Table V, p. 331.
The following ketones have proved to be unreactive: o-aminoaceto-phenone and itsacetyl andbenzoyl derivatives ;12 m-aminoacetophenone(the acetyl and benzoyl derivatives doreact in this case 12); p-acetoami-noacetophenone;11
and /3-tetralone.16 l-Phenyl-3-methylpyrazolone-5,17
do not react.With Aldehydes. Thebehavior of aldehydes in the Mannich reaction
is similar to that of ketones. The a-hydrogen atom of the aldehydeis substituted by a dialkylaminomethyl group. A secondary reaction
which sometimes occurs involves the simultaneous introduction of amethylol group on the a-carbon atom.18
16 Moset t ig and May,J. Org. Chem., 5, 528 (1940).17 Mannich and K a t he r , Arch. Pharm., 257, 18 (1919).18 Mannich, Leaaer, and Silten, Ber., 65, 378 (1932).
In the case of acetaldehyde the only product isolated is one of more
complicated nature in which two dimethylaminomethyl groups and one
methylol group have entered the molecule . '8
(CH 3)2N C H 2
CCHO
/I(CH3)2NCH2 CH 20H
The aldehydes have been less extensively studied than the ketonesand there are recorded merely the condensations of acetaldehyde,propionaldehyde, butyraldehyde, isobutyraldehyde, isovaleraldehyde,and hexahydrobenzaldehyde with dimethylamine or piperidine hydro-chloride. The products from the reactions are shown in Table V, p. 331 .
With Acids and Esters. A number of acids containing highly activehydrogen atoms in the a-position can be used instead of aldehydes orketones. When an acid is employed the free secondary amine, ratherthan its salt, is used. The acids which have given satisfactory resultsare listed in Table I I I . The replaceable hydrogen atoms are underlined.
the acid, as in the condensation of ethylacetoacetic acid with formal-
dehyde and dimethylamine20
CO2H
CH3CH2CH + CH2O + (CH3)2NH ->
COCH3
CH3CH2CHCH2N(CH3)2 + CO2 + H20
COCH3
In those cases where two dialkylamino groups enter the molecule,
carbon dioxide is invariably eliminated.
With Phenols. The 0- and p-hydrogens in phenols are sufficiently
active to enter into the Mannich reaction. Thus, products from
phenol,21
'22
'23
4-acetaminophenol,21
0- and p-cresol,22
m-cresol,23
3,5-
dimethylphenol,24
2-methyl-4-ethylphenol,22
2- and 4-methoxyphenol,25
i3-naphthol,25
and 8-hydroxyquinoline21
with formaldehyde and di-
methylamine or piperidine or morpholine, have been reported. From
p-cresol a mono- and a di-substitution product are obtained, and from
phenol and m-cresol, trisubstitution products.
OH
CH;
CH2N(CH3)2
Interaction of 2-methyl-6-ethylphenol, formaldehyde, and dimethyl-amine is reported to yield a mixture of methylenedi-(2-methyl-6-ethyl-
phenol) and l-(dimethylaminomethoxy)-2-methyl-6-ethylbenzene.22
With Acetylenes. Phenylacetylene and certain substituted phenyl-
acetylenes, such as the 2-nitro, 2-amino, and 4-methoxy derivatives,
react readily with formaldehyde and secondary amines.26
C6H6C=CH + CH20 + (C2HB)2NH -> C6H6C=CCH2N(C2H6)2
20 Mannich and Bau r o th , Ber., 57, 1108 (1924).21
Ger. pat., 92,309; Frdl., 4, 103 (1899).22
D S c o m b e , Compt. rend., 1 9 6 , 866 ( 1 9 3 3 ) .23 Bruson and M acM u l l en , J. Am. Chem. Soc, 63, 270 (1941).24 Caldwell and Thompson , / . Am. Chem. Soc, 61, 765 (1939).25 Dfcombe , Compt. rend., 197, 258 (1933).28 Mannich and Chang , Ber., 66, 418 (1933).
With a-Picolines and Quinaldines. Since an a-methyl group in apyridine or quinoline nucleus has hydrogens of about the same activityas those in th e me thyl group of a methyl ketone, the M annich reactionmigh t be expected to take place with such molecules. a-Picoline,27 2-methylquinoline 7l 27 '28 (quinaldine), 2-methyl-4-hydroxyquinoline,28 2-
methyl-8-nitroquinoline,28 and 2-ethoxy-4-methylquinoline 28 have beencondensed with dimethylamine, diethylamine, m ethyldiethylenediamine,piperidine, and methylaniline, either as the free amine or as the aminehydrochloride. Thus, a-picoline, formaldehyde, and diethylamineyield 2-(/3-diethylaminoethyl)-pyridine.27
N^CH
2CH
2N (C
2H
6)
2
The Use of Primary Amines
The primary amines listed in Table IV have been used successfullyin the Mannich condensation.
TABLE IV
PRIMARY AMINES IN THE MANNICH REACTION
Methylamine /3-Phenylethylamine
Ethylamine Ethylenediamine/S-Hydroxyethylamine Eth yl am inoacetate/3-Chloroethylamine w-AminoacetophenoneAllylamine Tetrahydro-j3-naphthylamineBenzy lamine Aniline *
3, 4-Methylene-dioxybenzylamine
Hydrazine 17 and guanidine,17 have failed to react.
* Reacts only in certain instances.
W ith K etones. When a primary amine or its salt is used in a Mannichreaction the first product is a secondary amine, but this often reactswith more of the reagents to give a tertia ry amine. Aliphatic ketonesand primary amines give rise to a number of products; for example,four substances have been isolated from the reaction of formaldehyde,diethylketone, and methylamine hydrochloride.29 The structures ofsome of them are still in doubt (see also the reaction of acetone, methy l-
The name "bispidin" has been suggested for the bicyclic r ing systemproduced in such react ions .3 3 '3 4
Th is reaction can be used to build up tr icyclic syste ms . Th us, th ehydrochloride of methyl tropanone-2,4-dicarboxylate reacts in thesame way as the pyr idone above.33
CO2CH3
CEUOH
> N C H 3
CH 2O
C H 2 C H
CHs N- HCl
C H 2 C H —
CHa
A similar react ion occurs when a te trahydropyrone 3 5 derivative is
used in place of th e pyrido ne. Fo r example, a bicyclic pro du ct isobtained f rom ethyl dimethylte trahydropyronedicarboxylate , formal-dehyde , and methylamine .
CO2C2H5
>NCH3
I t has been suggested that the bicyclic r ing system so formed be termedthe "pydin" nuc leus .
With Phenols and Acetylenes. No Mannich react ions involving pr i-mary amines and e i ther phenols or acetylenes have been reported.
With a-Picolines and Quinaldines. Of the compounds containing a
methyl group in the 2-position of a pyridine nucleus only 2-methyl-8-ni troquinoline has been treated with a pr imary amine and formalde-
34 Mannich and Mohs , Ber., 63, 608 (1930).36 Mannich and Muck, Ber., 63, 604 (1930).
hyde. The amine used was ethylamine, and the product was a tertia ry
a m i n e /
| C H CH2O + C
2H
6N H
2-HC1
NO 2
CH2CH2 / NC 2H B-HC1/2
N O ,
The Use of Ammonia
With Ketones. A primary amine is the f irs t product to be expectedfrom a Mannich react ion in which ammonia or an ammonium sal t andformaldehyde react with a compound containing an act ive hydrogen
atom . W ith the simple ketones subsequen t react ion of th e pr im aryamine so formed usually leads to the production of tertiary amines.Salts of certain of thes e prim ary an d secon dary am ines have beenisolated and found to be stable, but the free bases change to the tertiary
amines . Th e disproport ionat ion of the pr im ary and secondary aminesobtained f rom acetophenone, formaldehyde, and ammonia is an ex-
In some instances cyclic products are obtained from ketones, am-monia , and formaldehyde. Fro m acetophenone, am m on ium chloride ,
and formaldehyde there has been isolated a substance which is believedto be a substituted piperidine.36 I t readily changes to the salt of tr i-(/3-benzoylethyl)-amine. 3
With cyclohexanone the tertiary amine is obtained directly,6 inanalogy with the reaction of antipyrine Bl 37 (p. 306).
The formation of cyclic products derived from methylamine, byreaction of acetone, formaldehyde, and ammonium chloride, has beenmentioned (p. 305). The reaction with diethyl ketone takes a similar
course, producing a trimethylpiperidone.29 Presumably, m ethylamine isfirst formed from ammonium chloride and formaldehyde.
W ith A cids. From the reaction of benzylmalonic acid, ammonia, andformaldehyde both a primary amine and a secondary amine have beenisolated.19
C02H C02H C02HI I I
C 6H 6CH 2CH -* C 6H 6CH 2C — C H 2N H 2 -> (C 6H 6CH 2C — C H 2)2N H
CO2H CO2H CO2H
In the case of phenylmalonic acid a primary amine is produced and
decarboxylation occurs when ammonia is used.19
CO2H
C 6H BCH -> C 6H 6C H — C H 2N H 2
CO 2H CO2H
When ammonium chlor ide is employed the decarboxylated secondary
amine is obtained.19
C 0 2 H
C 6H 6CH ~> (C 6H 6CHCH 2)2NH
I ICO2H CO2H
RELATED REACTIONS
Aldehydes other than formaldehyde may be used in certain con-densations of the Mannich type . Those which have been studied areacetaldehyde, phenylacetaldehyde, benzaldehyde, and anisaldehyde.These have been employed successfully with acetone, cyclohexanone,and esters of acetonedicarboxylic acid. The reactions appear to belimited to ammonia and primary amines and their salts. W ith acetone,aniline, and benzaldehyde a piperidone is obtained.4"*
In a few cases the products from Mannich reactions decompose
spontaneously. Thus, from monoethyl ethylmalonate, formaldehyde,
and diethylamine there is obtained directly ethyl a-ethylacrylate;
undoubtedly, this is formed by elimination of carbon dioxide and diethyl-
amine from the primary reaction product.43
COOH
C2H6CHCOOC2H6 + H2CO + (C2H6)2NH
COOH
C 2 H B C C O O C 2 H 5
C H 2 N ( C 2 H 6 ) 2 _
+ H2O - • C 2 H B C C O O C 2 H 6 + H2O + CO2 + (C 2 H B )2 N H
I ICH2
Other /3-dimethylaminoketones are sufficiently unstable that they
decompose in the presence of sodium ethylate or dilute alkaline solu-40 Mannich and Honig , Arch. Pharm., 265, 698 (1927).41 Mannich and Heilner , Ber., 55, 356 (1922).42 Mannich and B a uro t h , Ber., 55, 3504 (1922).4 3 M a n n i c h and Riitsert, Ber., 57, 1116 (1924).
that phenyl vinyl ketone and phenylhydrazine react with surprisingease to yield 1,3-diphenylpyrazoline.
CH2—CH2C 6 H 6C O C H = C H 2 + C 6H 6N H N H 2 -> | | + H2O
C6H
6C N C
6H
6\ /
N
When /3-dimethylaminopropiophenone hydrochlor ide and phenylhy-drazine react in the presence of sodium acetate, 1,3-diphenylpyrazolineis formed.1 3 '2 0 ' 4 0 '4 6 '4 7 ' 48
In some cases, the in te rmedia te produc ts mu s t
be treated with ethanolic hydrogen chloride to effect the cyclization.
C 6 H 6 N H N H 2 ->
C 6 H 6 C — C H 2 C H 2 N ( C H 3 ) 2 -> H 2 C — CH2
|| I I + ( C H 3 ) 2 N HN—NHC 6 H 6 C6HBC C H C 6 H B
\ /
N
I t is not imposs ible that the ini t ia l phenylhydrazone decomposes to the
phenylhydrazone of the phenyl vinyl ketone, which then cyclizes to the
1,3-diphenylpyrazoline. Such a mechanism is suppor ted by the work
of Nisbet,49
'B 0
'B 1> 52
whoobserved th a t thephenylhydrazones of /3-dialkyl-aminoketones derived from a, /3-unsaturated ketones isomerize readily
to pyrazolines and in so reac t ing make use of the double bond alreadypresent in the molecule.
RCH CH2
R C H = C H — C — C H 2 C H 2 N R 2 ' - H C l - > | ||| R"N C — C H 2C H 2N R 2 ' -HC1
R " H N — N \ , /where R and R"= aryl N
R' = alkyl
Some of the l ,5-diaryl-3-(/3-dialkylam inoethyl)-pyrazoline salts were
shown by N i s b e t M l 6 1 ' Mto be local anesthetics .
The Use of a Mannich Base as a Source of Unsaturated Ketone for
C ondensations w ith an Active Methylene Compound. A reaction which
offers many possibilities in synthet ic work is the condensat ion of /3-di-
a lkylaminoketones with act ive methylene compounds in the presence
46Jacob and Madinaveitia, J. Chem. Soc, 1929 (1937).
" Harradence and Lions, / . Proc. Roy. Soc. N. S. Wales, 72, 233 (1938).48Harradence and Lions, / . Proc. Roy. Soc. N. S. Wales, 73, 14 (1939).
the alcohol is c o n v e r t e d t o t h e p -a mi n o b e n zo a t e , a n d t h e l a t t e r is used
as t h e hydroch lor ide .
(p) H 2 N C 6 H 4 C O O C H — C H C H 2 N ( C H 3 ) 2 • H C1
CH3 CH3Tutocaine
Products Derived by Transformation of the Aldehyde
Group in {J-Dialkylaminoaldehydes
Certain of the /3-dialkylaminoaldehydes can be transformed intopiperidine deriva tives. Th us , a,a-dimethyl-/3-dimethylam inopropion al-dehy de is conv erted into 1,2,5,5-tetramethy lpiperidine.38
CH3 CH3
4(H)N aO C s H s I V
C H 3C — C H O + C H 3 C O C H 3 -±> C H 3 C — C H = C H C — C H 3 >I 1 2 s tep s
C H 2N(CH 3)2 CH 2N(CH 3)2
CH3C CH2CH2CHCH3 ^ CH3C CH2CH2CHCXI3 >I I I I
C H 2N (C H 3)2 O H C H 2N(CH 3) 2 ClCIl2 C-EI2
/ \ / \
2C C H 2 H e a t (CH3)2C C H 2
I I I IC H 2 CHCH3 CH 2 C H C H 3
\ + / \ /N N
/ \ ICH3 CH3 Cl CH3
The aminoaldehyde a lso may be t ransformed into the corresponding
amino acids 18 by the following series of reactions.
The Mannich bases from one molecule of a primary amine, one offormaldehyde, and one of ketone have been used in a variety of con-
densat ions involving both the ketone group and the secondary aminegroup. T he nitroso deriv ative of /3-methylaminopropiophenone isreadily reduced to the corresponding /3-hydrazinoketone, which cyclizesto l -methyl-3-phenylpyrazoline .64
NO
CH2 CH2
C6H BCOCH2CH2—NCH 3 - • I I + H 2OI C 6H 6C N C H 3
N H 2 \ /N
A similar cyclization occurs in the formation of 2-benzyltetrahydro-
naph th indazo le 3 8 by reduct ion of 2-(benzymitrosaminomethyl)-a- te t-
ralone.C H , C H 2
S C H 2 <^>/ \CH2
C 0 C H 3
ON—N—CHsCeHB N N— CH 2C 6H B
Oth er typ es of cyclic com poun ds ma y result if prop erly con structedmolecules and appro pria te reagents are used. T hu s the compo und
from benzylamine hydrochloride, formaldehyde, and cyclohexanonereacts with potass ium cyan ate to form a urea which u ndergoes dehy dra-tion to an octahydroquinazoline.38
An analogous reaction has been used for the synthesis of l-methyl-2-
keto-4-phenyl- l ,2 ,5,6- te trahydropyrimidine f rom j3-methylaminopropio-phenone .64
CsHsCOCHaCHsNHCHs-HCl-^CeHeCO NH 2 - ^ C 6 H 6 C = = N
II IIC H 2 CO CH 2 COI I - I I
CH 2—NCH3 CH 2 N C H 3
C ondensation Products from O ne M ole of a Primary Am ine, Tw o
Moles of Formaldehyde, and Two Moles of a Ketone
Benzylamine hydrochlor ide , formaldehyde, and acetophenone react
to form a m ixture of pro du cts : 3 S the first from one mole of benzylamine,
one of acetophenone, and one of formaldehyde; and the second from onemole of benzylamine, two of acetophenone, and two of formaldehyde.The second is unstable and cyclizes to a piperidine derivative.
C 6 H 6 C O C H 3 + H C H O + C 6H 6C H 2N H 2-HC1 ->
C6H6COCH 2CH 2NHCH 2C6H 6- HC1
2C 6H 6COCH 3 + 2 H C H 0 + C 6H 5C H 2NH-HC1 ->
C 6H 6 OH\ /C
/ \CeHsCOCHsCHii—N—CH8CeHB-HCl H 2C CH—COC 6H 6
CeHj—CO—CH2— CH 2 H 2C C H 2
\ /N — C H 2C 6H 6-HC1
Benzylamine hydrochloride condenses s imilarly with cyclohexanone,38
and th e prod uct involving tw o m oles of cyclohexanone is conve r ted to areduced isoquinoline der ivat ive dur ing the react ion.
. H C 1L \ / J \ / N C H 2 C 6 H 6 • HC1
C H 2C H 2
A tricyclic r ing system is formed when the diethyl ester of 1-methyl-
3,5-diallyl-4-piperidone-3,5-dicarboxylic acid (obtained from the diethyl
ester of a,a'-diallylacetonedicarboxylic acid, two moles of formaldehyde,and one of methylamine) is hydrolyzed and decarboxylated.32
CH 3CH—O O—CHCH,
CO CH2 C CH2/ \ /C H 2C H = C H 2 \ / \ /C c < CH CHI I X J O O d H , -> I I
V VC H 3 C H 8
EXPERIMENTAL CONDITIONS AND PROCEDURES
Solvents
When aqueous formaldehyde is used the condensation is ordinarilycarried out by shaking or stirring the reactants in the absence of anorganic solvent; in some cases u methanol has been added to suchmixtures. When paraformaldehyde is used an organic solvent is re -quired. If the ketone com ponent is a liquid, such as acetone,38 cyclo-pentanone,47 or cyclohexanone,47 an excess of it may be used as the
solvent. In other cases ethanol (95% or absolute) is added as thesolvent. In condensations involving 2-, 3-, or 9 -acetylphenan threne,paraformaldehyde, and salts of secondary amines, isoamyl alcohol isrecommended as the solvent.66 The condensations proceed much fasterin the higher-boiling solvent, and the formation of certain by-products,obtained by prolonged heating in ethan ol, is avoided. On the otherhand, it is stated that, although in ethanol the condensation between3-acetyl-9-methylcarbazole, formaldehyde, and a secondary amine saltproceeds more slowly than in isoamyl alcohol, it is less subject to sidereactions associated with instability of the aminoketone salts at thehigher temperature.66
Nature of Formaldehyde and Time of Reaction
Formaldehyde is used in the form of a 20-40% aqueous solution or asparaformaldehyde. In certain reactions, such as the condensation ofa-tetralone, formaldehyde, and tetrahydroisoquinoline hydrochloride,
aqueous formaldehyde is said to be superior to paraformaldehyde.16
In a few cases 1 2 '3 8 '4 7 enough concentrated hydrochloric acid is addedat the beginning of the reaction to make the mixture acidic to Congo red ;
66 van de Kamp and Mosettig, J. Am. Chem . Soc, 58, 1568 (1936).66 Ruberg and Small, J. Am. Chem . Soc, 63, 736 (1941).
in other instances "•15> 65 the mixture is acidified at the end of thereaction in order to depolymerize unchanged paraformaldehyde andbring it into solution.
The time required for a Mannich reaction depends upon the natureof the ketone and of the amine salt and upon the boiling point of the
solvent employed. The reaction between furfuralacetone, paraformal-dehyde, and dimethylamine hydrochloride in alcoholic solution is saidto be complete after the mixture has been boiled for a few minutes. 49
When 3-acetyl-9-methylcarbazole, paraformaldehyde, and diethylaminehydrochloride are heated in absolute ethanolic solution for five hoursthe yield of reaction product is 59% but is increased to 83% when themixture is heated for eight hours.66
Relative Am ounts of C omponents
In the preparation of Mannich products, various investigators havemixed the components in the calculated quantities or they have em-ployed an excess of the amine salt and formaldehyde or an excess of theketone. I t is common prac tice to use 1.00 molecular equivalen t of thecarbonyl compound, 1.05-1.10 molecular equivalents of the amine salt,and 1.5-2.0 molecular equivalents of formaldehyde. Excellent yieldsof the basic ketone are obtained by the interaction of cyclohexanone,
aqueous formaldehyde, and dimethylamine hydrochloride,37 or mor-pholine hydrochloride,47 when five tunes the calculated quantity ofketone is allowed to reac t. When excess formaldehyde is used, thematerial is added in several portions during the course of the reaction.Part of the formaldehyde reacts with ethanol, when this is used as asolvent, to form methylene diethy l ether.9
Due consideration should be given to the manner in which unchangedamine salt and formaldehyde can be separated from the desired product
at the termination of the reaction . If difficulties are antic ipated in suchseparations, the advantage to be gained by the employment of any ofthe components in excess may be questioned. If more tha n one reactionproduct is possible, the relative amounts of amine salt and formaldehydemay or may not influence the nature and yield of the product. 18 '19
Isolation of Product
In a number of cases the salt of the desired product precipitates when
the reaction mixture is cooled. Eth er may be added to facilitate sepa-ration of the product. Occasionally the solvent is removed and crystal-lization of the residue brought about by washing it with ether or acetone.Sometimes it is advantageous to liberate the basic product from its salt
( 2 7 % )Tn-(2-cyclohexanonylmethyl)-amine (—)Tn-(4 -an t ipyry lmethyI ) -amine (86%)Tri-(p- tolypyrylmethyl)-amine (72%)Tn-(homoan t ipyry lmethy l ) -amine (70%)[C 6H 6C H ( C O O H ) C H2 ]2NH (63%)
H O C ( C O O H )2 C H 2N H C H 8 ( 3 3 % )[C H 3C ( C O O H )2C H 2]2N CH a ( 3 4 % )C H S C H 2C ( C O O H )2C H 2NHCH3 (—)C eH 6C H 2C ( C O O H )2C H 2N H C H 3 (very good)C 6H 6C O C H 2C ( C O O H )2C H 2N H C H S (good)( 4 )N O 2C 6H 4C H ( C O O H ) C H 2N H C H 3 (20%)
Diethyl l -methyl-3 ,5-die thyl-4-pipendone-3 (5-dicarboxylate(40%)
Die thyl 1 -methyl-3,5-diallyl-4-piperidone-3,5-dicarboxylate(65-70%)
A "pyd in" § (64%)
A " p y d i n " ( > 8 0 % )
A "bispidin" § (74%)
A "bispidin" § (70%)
A tr icycho compound § (45-50%)
Dim ethyl (and die thyl) l -methyl-2 ,6-diphenyl-4-pipendone-3,5-dicarboxylate (65%)
* References 67-74 appea r on p 341t Th e pipendo ne was obtained in smaller amount when am mo nium chloride was used in place of
ammonium bromide, the yie ld was s t i l l lower when ammonia was subst i tu ted for an ammonium sal t1 Malomc acid yielded an unidentified productS S e e p 3 1 4
meth yl -4 - p iperidone -3 , 5 - d icar-boxylate 33
Methylamine hydrochloride, acetal-dehyde, and
Diethyl acetonedicarboxylate 39
Ethylamine, benzaldehyde, and
Diethyl acetonedicarboxylate 4<I
Ethylamine hydrochloride, formalde-hyde, and
2-Methyl-8-ni troquinoline 28
Antipyrine 17
P-Hydroxyethylamine, benzaldehyde,
an d
Dim ethyl acetonedicarboxylate 34
P-Chloroethylamine hydrochloride,
formaldehyde, and
Dimethyl l ,2 ,6- tr imethyl-4-pi-peridone-3,5-dicarboxylate 33
P-Phenylethylamine hydrochloride.formaldehyde, and
Dimethyl l ,2 ,6- tr imethyl-4-pi-peridone-3,5-dicarboxylate 33
(3-Phenylethylamine hydrochloride, tacetaldehyde, and
Dimethyl acetonedicarboxylate 39
AUylamine, formaldehyde, and
Benzylmalonic acid 1 9
Dimethyl l -methyl-2 ,6-diphenyl-4-piperidone-3,5-dicarboxylate 34
AUylamine, benzaldehyde, and
Dim ethyl acetonedicarboxylate 34
Product (Yield)
(C H 3C O C H 2C H 2 )2 N C H 3 (56%)1,4-Dimethyl-3-acetyl-4-hydroxypiperidine (—)C H S C H 2C O C H ( C H 3 )C H 2 N H ( C H 3 ) (—)l,3,5-Trimethyl-4-piperidone (—)C H 3CH [ CH 2NH (CH 3) ]COCH [C H 2N H ( C H 3) ]CH 3 orC H 3C H 2 C O C [ C H 2N H ( C H 3 ) ] 2C H 3 (—)[C H 3C H 2 C O C H ( C H 3 )C H 2 ] 2N C H 3 (—)C 6H 6 C O C H 2C H 2 N H C H 3 ( 7 0 % )(C 6H 5C O C H 2C H 2 )2N C H 3 ( 3 4 % )Methyldi-(2-cyclohexanonylmethyl)-amine (2 .4% + )
Me thyl d i-[2-(a- thenoyl)-e thyl]-amine (61%)Methyldi-(4-antipyrylmethyl)-amine (92%)(C H 3 ) 2 C ( C H O ) C H 2N H C H 3 (70%)
A "b isp id in" (70%)
Die thyl l ,2,6-trimethyl-4-piperido ne-3,5-dicarbox ylate (—)
A "bispidin" (—)Allyldi- (4-antipyry lmeth yl) -am ine (—)
C6H5COCH2NR2; R = 4 -antipyrylm ethyl (98% )
R2NCH 2CO2C 2H 6; R = 4-antipyrylmethyl (—)
CeHfiCHaNHCHCCHzCeHtOCeHflO (1.5% )
C H 3C O C H 2C H 2N H C H 2 C 6H 6 ( > 3 % )C 6H B CH = CH CO CH 2 CH 2 N H CH 2 C6 H 5 ( 2 0 % )l-Benzyl-3-cinnamyl-4-styryl-4-hydroxypiperidine (10%)C 6H 5COCH2CH2NHCH 2C6H5 (53%)
l-Benzyl-3-benzoyl-4-phenyl-4-hydroxypiperidine (—)Benzyl-(2-cyclopentanonylmethyl)-amine (—)Benzyl-(2-cyclohexanonylmethyl)-amine (65%)A decahydroisoquinoline t (10-25% )/3-(Benzylaminomethyl) -a-te tralo ne (55 % )
C 6H 6CH = CH CO CH 2 CH 2 N H CH 2 C6 H 3 ( O 2CH2) (3,4) (52%)C 6H5COCH2CH2NHCH2C6H 3(O2CH2) (3,4) (56%)2- (3 ,4-Methylenedioxybenzylaminomethyl) -cyclopentanone
( 6 7 % )2-(3,4-MethylenedioxybenzyIaminomethyl)-cyclohexanone (—)A decahydroisoq uinoline t (—)0-(3 ,4-Methylenedioxybenzylaminomethyl)-a- te tra lone (70%)
C N C H 2C H 2 N ( C H 3 ) 2 t (—)( 4 )N O 2 C6 H4CH ( CO O H )CH 2N ( CH 3)2 ( 6 7 % )(2,4)(NO 2)2C 6H 3CH [ CH 2N ( CH 3)2]2 (52%)C 6H 5 CO CH 2 CH 2N(CH 3)2 (—)(C H 3 ) 2N C H 2C H C O C O O C H 2 t (56%)
1 1C H 3C O C H 2CH 2 N ( CH 3)2 (42%)C H 3C O C H [ C H 2N ( CH s )2 ] 2 (28%)C H 3 C O C H ( C H 3 )C H 2 N ( C H 3 ) 2 (—)C H 3 C O C H ( C 2 H 6 )C H 2 N ( C H 3 ) 2 t ( 3 0 % )C H 3 C O C H ( C H 2 C H = C H 2 )C H 2 N ( C H 3 ) 2 ( 3 8 % )(CH 3)2NCH 2C H 2 C O C H2C H 2C O O H t (21%)C H ( C O O H ) [ C H 2 N ( C H 3 ) 2 ] 2 (47%)C H 3C ( C O O H ) 2C H 2N ( C H 3 ) 2 (55%)C H 3C H 2C ( C O O H ) 2C H 2N ( C H 3)2 (70%)C H 2 = C H C H 2 C ( C O O H )2 C H 2 N ( C H 3 ) 2 (90%)
C 6H 6C H 2C ( C O O H )2 C H 2 N ( C H 3 ) 2 (90%)C 8H 6 C H ( C O O H ) C H 2 N ( C H 3 ) 2 ( 6 0 % )C6 H 5 CH 2C H 2C H 2C ( C O O H )2C H 2N ( CH S)2 (90%)C 6H 6C O C H 2C ( C O O H )2C H 2N ( C H 3 )2 ( 4 5 % )H O C ( C O O H ) 2C H 2N ( CH 3 )2 (54%)( H O O C )2C ( C H 2C O O H ) C H 2N ( C H 3 ) 2 ( 4 6 % )C 6 H B C = C C H 2 N ( C H 3 ) 2 (—)
( 2 ) N H 2 C 6 H 4 C = C C H 2 N ( C H 3 ) 2 (—)
4-Dimethy laminomethy lan t ipyr ine (60%)2- (Dime thylaminom ethyl) -phenol (—)2,6-Di-(dimethylaminomethyl)-phenol (poor)2 ,4 ,6 -Tn- (d imethy lammomethy l ) -pheno l (86%)2-(Dimethylaminoinethyl)-4-acetyla ininoplienol (—)2-(Dimethylaminomethyl)-6-methylphenol (—)2,4 ,6-Tri-(dunethylaminomethyl)-3-methylphenol (—)2-(Dimethylam.in.omjethyl)-4-iaethylphenol (—)2,6-Di-(dimethylaminomethyl)-4-methylphenol (—)2-Methoxy-6-(dimethylaminomethyl)-phenol (—)4-Methoxy-6-(dimethylaminomethyl)-phenol (—)2-(Dunethylaminomethyl) 3 ,5-dimethylphenol (34%)2-Methyl-4-ethyl-6- (dimethylam inomethyl) -phenol (—)Dunethylaminomethylcatechol (—) X
Dimethylaminomethylresorcinol (—) t
2,5-6is-(Dunethylaminomethyl)-hydroquinone (a lmost quanti-ta t ive) t
Dimethylaminomethylphloroglucmol (—) t
&ts-(DimethylaminomethyI)-phloroglucmol (—) t
3-Dimethylaminomethylmdole (a lmost quanti ta t ive)Dimethylaminomethyl- j3-naphthol (—)
C H 3C O C H 2 C H 2N ( C H 3 ) 2 § (—) (14% )C H 3C O C H [ C H 2N ( C H 3 ) 2 ] 2 § (—) (58% )
C H 3C O C H ( C H 3 ) C H 2 N ( C H 3 )2 (—)C H 3C H 2 C O C H 2C H 2 N ( C H 3 ) 2 (—)C H 3 C O C H ( C 2 H 6 )C H 2N ( C H 3 ) 2 (—)
* References 67-74 appear on p 341t The product could not be obtained m crystalline formt In th is instance the amine sal t was employed§ The amine base was used
l -Ke to -9 -methy l - l ,2 ,3 ,4 - te t ra -hydrocarbazole M
AcetaldehydeJ 8
Propionaldehyde 18
CH 3 CH 2 CO CH ( CH 3)C H 2N ( CH 3)2 (31%)C 6H 5 C O C H 2 C H2N ( CH 3)2 ( 6 0 % )(2)NO 2C 6H 4CO CH j i CH 2N ( CH 8)2 ( 8 0 - 9 0 % )(3)NO 2C«H 4CO CH 2 CH 2 N ( CH8)2 (80-90%)(3) (CH3CO NH )C 6H 4C O C H 2CH2N(CH3)2 (55%)(3) (C 6H 5 CO N H )C6 H4CO CH 2CH 2 N ( CH8)2 (79%)(4)CH 3O C eH 4C O C H 2C H 2N(CH3)2 (—)(3,4) (CH sO)2C 6H 3 CO CH2CH 2N(CH 8)2 (—)C 6 H 6 C H = C H C O C H 2 C H 2 N ( C H 3 ) 2 (25%)( 4 )CH 3 O C 6H 4 C H = C H C O C H2C H 2N ( C H 8)2 (63%)(3,4) (CH 2O2 )C6H 3C H = C H C O C H2C H 2N ( CH 8)2 (—)
(3,4) (CH 3O) (C2H 6O )C6 H 3C H = C H C O C H 2 C H 2N(CH3)2 (—)
(3,4) (C 2H 6O) (CH 3 O )C 6 H 3C H = C H C O C H 2 C H 2 N ( C H 3 ) 2 (—)
(3,4) (6) (CH2O2) (N O 2 )C 6H 2CH = CH CO CH 2 CH 2 N ( CH 3 )2 ( — )
(3,4 ,6)(CH3O)2(NO2)C 6H 2 CH = CH CO CH 2CH2N(CH3)2(20-25%)
/ 3 - C I O H T C O C H 2 C H 2 N ( C H 8 ) 2 (70%)
2-(Dimethylaminomethyl)-cyclopentanone (—)2-(Dimethylaminomethyl)-cyclohexanone (85%)2-(Dimethylaminomethyl)-4-methylcyclohexanone (—)Dimethy laminomethy lmen thone f (54%)(3-(Dimethylaminomethyl)-a- te tra lone (70%)
l -Ke to -2 -d imethy laminomethy l - l ,2 ,3 ,4 - te t rahydrophenan-th rene (65%)
4-Keto -3 -d imethy laminomethy l - lf2,3 ,4- te trahydrophenan-th rene (77%)
2-Furyl /S-dimethylaminoethyl ketone (—)
2-Thienyl /S-dimethylaminoethyl ketone (47%)/3-Dimethylaminoethyl 2-dibenzothienyl ketone (41%) t
ct-(Dimethylaminomethyl)-«thjrl 2-thienyl ketone (60%)j3-Dimethylaminoethyl 4-phenyl-2-thiazolyl ketone (—)4-Dimethy laminomethy lan t ipyr ine (90%)l-Phenyl-2 ,5-dimethyl-4-dimethylaminomethylpyrazolone-3
(74%)(5-Dimethylaminoethyl 2-(9-methylcarbazyl) ketone (18%)/3-Dimethylaminoethyl 3-(9-methylcarbazyl) ketone (61%) X
l -Ke to -2 -d imethy laminomethy l -9 -methy l - l ,2 ,3 ,4 - tet rahydro-carbazole (10-15%)
[(CH 3)2NCH2]2C(CH2OH)CHO (pract ioal ly quanti ta t ive)C H 3CH [ CH 2 N ( CH 3 ) 2]CHO (15%)C H SC [ C H 2N ( C H S ) 2 ]2CHO (—)
* References 67-74 appear on p. 341.t Amixture of isomers seems to be formed.t Yield based on the amount of original ketone not recovered from the reaction mixture.
Diethyl 2 ,6 -d imethy l te t rahydro-pyrone-3,5-dicarboxylate 36
Phenylacetylene 26
2-Ni t rophenylacetylene 26
4-Ni t rophenylacetylene 26
4-Methoxyphenylace ty l ene 26
a-Picoline 27
Quinaldine 7f 27
Diethylamine hydrochloride, formal-dehyde, and
Acetone 8
Acetophenone 1(>2-Ni t roace tophenone 12
3-Ni t roace tophenone 12
Acetovera t rone J 1
Benzalacetone 9
4-Anisalacetone 6 0
2-Butoxybenzalacetone 62
Methylenedioxybenzalacetone9
3,4-Dimethoxybenzalacetone 9p 61
3-Ethoxy-4-methoxybenzalace-tone 6 1
6-Ni t ropiperonalacetone K
6-Ni t roverat raIacetone 56
2-Acetylphenanthrene ^3-Acetylphenanthrene 6B
9-Acetylphenanthrene 66
2-Methylcyclopentanone 8
Cyclobexanone *°2-Methylcyclohexanone 8
l -Keto- l , 2 ,3 ,4- t e t rahydrophe-nanthrene 16
Product (Yield)
C H 3C H 2C H t C H 2N (C H 3) 2 ]C H O (—)C H 3 C H 2 C ( = C H 2 ) C H O (—)(C H 3)2C[CH2N(CH3)2]CHO (70-80%)(C H 3)2C H C H [ C H 2N (C H 3) 2 ]C H O (—)(C H S )2C H (C H 2OH) [CH 2N ( C H a) 2 ]C H O (—)1-Dimethylaminomethylhexahydrobenzaldehyde (—)2-(0-Dimethylaminoethyl) -quinol ine (—)2-(/3-Dimethylaminoethyl)-4-hydroxyquinol ine (—)2-Ethoxy-4-(3-dimethylaminoethyl) -quinol ine (—)
(2,4)(NO 2)2C 6H 3 CH [ CH 2 N ( C 2H 6)2]2 (52%)C H 3C O C H ( C H 2C 6H 5 ) C H 2N (C 2H 6 ) 2 (46%)
C 2H 6O O C C H 2C H 2N (C 2H 6)2 (21%)C 2H 6O O C C H [C H 2N (C 2H 5)2]2 (—)C 2 H 6 O O C C ( = C H 2 ) C H 3 (88%)C 2 H 6 O O C C ( = C H 2) C H 2 C H 3 ( 6 3 % )C 2 H 6 O O C C ( = C H 2 ) C H 2 C H = C H 2 (quant i tat ive)C 2H 6O O C C ( = C H 2 ) C H 2 C6H 6 (73%)
( 2 ) N O 2 C 6 H 4 C s C C H 2 N ( C 2 H 6 ) 2 (—)(4)NO2C 6H 4C sC C H 2N (C 2H 5) 2 (—)(4)CH 3O C 6H 4C s C C H 2 N ( C 2 H 6 ) 2 (—)2-(/S-Diethylaminoethyl)-pyridine (80%)2-(/3-Diethylaminoethyl)-quinoline ( 3 3 % )
C H 3C O C H 2C H 2 N (C 2H 6 ) 2 (66%)C 6H 6C O C H 2C H 2N (C 2H 6)2 ( 4 5 % )(2)NO2C6H4COCH 2CH 2N (C 2H 6)2 (80 -90%)(3)NO2C 6H 4C O C H 2C H 2N(C2H 6)2 (80-90%)(3 ,4)(CH 3O )2C 6H 3C O C H 2 C H 2N (C 2H B )2 (—)C 6H 5 C H = C H C O C H 2 C H 2N (C 2H5)2 (60%)(4)CH aO C 6H 4C H = C H C O C H 2C H 1 !N (C 2H 6)is ( 6 0 % )(2)C 4H 9O C 6H 4C H = C H C O C H 2 C H 2N(C2H5)2 (5-10%)
(3,4) (CH 2O 2)C 6H 3C H = C H C O C H 2 C H 2 N ( C2H 5)2 (60%)(3,4) (C H 3 O ) 2 C 6 H 3 C H = C H C O C H 2C H 2 N ( C 2H 6 ) 2 (60%)
(3,4) (C2H5O) (C H 3O )C 6H 3C H = C H C O C H 2C H 2N ( C 2 H 5 ) 2 (—)(3,4,6) (CH 2O 2) (N O 2)C 6H 2C H = C H C O C H 2 C H 2 N ( C 2 H 5 ) 2
( 5 0 % )(3,4,6)(CH 3O )2(N O 2)C6H 2C H = C H C O C H 2C H 2N(C2H5)2
4 -Ke to - l ,2 ,3 ,4 - te t rahydrophe-nan th rene 15
l -Ke to -9 -methoxy- l ,2 ,3 ,4 - te t ra -hydrophenan th rene 7«
l -Ke to -9 -ace toxy- l ,2 ,3 ,4 - te t ra -h y d r o p h e n a n t h r e n e 7 0
Furfuralacetone 49
Chromanone ' I2-Aoetylthiophene w
2-Acetyldibenzothiophene u
Ant ipyr ine l 7
2-Acetyl-9-methylcarbazole 69
3-Acetyl-9-methylcarbazole 66
2-Acetyl- i-phenylthiazole 13
Iaobu ty ra ldehyde 18
Diethanolamine hydrcchloride, form-
aldehyde, and
2-Acetylfuran 13
Dipropylamine, formaldehyde, and
Ethylaoetoaoetic acid ••
Dipropylamine hydrochloride, form-aldehyde, and
Anisalacetone 60
2-Acetylfuran l s
2-Acetyl-4-phenylthiazole 13
Dibutylami-ne hydrochloride, form-aldehyde, and
2-Aeetylfuran 13
Anisalacetone M
Diisoamylamine hydrochloride.
formaldehyde, andAcetophenone 6 8
Methyldiethylethylenediaminehydrochloride, formaldehyde, and
2-Methyl-4-hydroxyquinoline ^
u-Methylaminopropiophenone
hydrochloride, formaldehyde, andAntipyrine 64
p-Acetyletkylbenzylamine hydro-
chloride, formaldehyde, andAcetone m
Dibenzylamine hydrochloride.
formaldehyde, and
Anisalacetone 60
Product (Yield)
4-Keto-3-die thylaminomethyl- l ,2 ,3 ,4- te trahydrophenan-thren e (51 % )
l-Keto-2-die thylaminomethyl-9-methoxy-l ,2 ,3 ,4- te trahydro-phenan th rene (41%)
l-Keto-2-die thylaminomethyl-9-acetoxy-l ,2 ,3 ,4- te trahydro-phenan th rene (20%)
C 4H 8 O C H = C H C O C H 2 C H 2 N ( C 2 H6)2 (—)3-Diethylaminomethyl-4-chromanone (—)/3-Diethylaminoethyl 2-thienyl ketone (39%)/9-Diethylaminoethyl 2-dibenzothienyl k etone](40% )4-Diethylaminomethylantipyrine (—)/3-Diethylaminoethyl 2-(9-methylcarbazyl) ketone (20-25%)|3-Diethylaminoethyl 3-(9-methylcarbazyl) ketone (83%) t3-Diethylaminoethyl 4-phenyl-2-thiazolyl ketone (—)(C H 3)2C[ CH 2N (C 2H 6)2]CHO (—)
(4)CH 8O C 6H 4 C H = C H C O C H2C H 2N ( C3 H 7 )2 ( 8 S % )3-Dipropylaminoethyl 2-furyl ketone (—)0-Dipropylaminoethyl 4-phenyl-2-thiazolyl ketone (—)
0-Dibutylaminoethyl 2-furyl ketone (—)
(4)CH 3O C6 H 4C H = C H C O C H 2C H 2N(C4H9)2 (16%)
4-Piperidinomethylantipyrine (44%)2-Piperidinomethylcyclohexanone (37%)(4)N02C6H 4CH ( CO O H )CH 2NC5Hio (64%)(2,4) (NC>2)2C6H3CH(CH2NC5Hio)2 (41%)(2) NO2 C 6H 4C(OH) (COOH )CH 2N C BHio (75%)OgxlB'—''JV^il2l-'Xl2J\05X110 \a\) /o)
C 6H i o N CH 2CHCO COO CH2 t (43%)1 1
C H 8C O C H ( C H 3 ) C H2N C 6Hio (60%)CH3COC H(C2H5)CH2NCsHio t (—)CH 3 CO CH ( CH 2 CH = CH 2 )CH 2N C 6H i o ( 3 0 ^ 5 % )CH3COCH(CH2C6H6)CH2NC5Hio (46%)CH2(CH2NC 6Hio)COCH2CH 2COOH (48%) tCeH5CH2C(COOH)2CH2NC6Hio (85%)
C( O H )( CO O H )2 CH2N C6 H 1 0 (14%)
Diethyl 2 ,6-dimethyl-3-(piperidinomethyl)- te trahydropyrone-3,5-dicarboxylate (73%)
3-Methoxy-4-ethoxybenjalace-to ne 613-Ethoxy-4-methoxybenzalace-tone «6-Ni t ropiperonalacetone M
6-Ni t roverat ralaoetone W
2-Acetylphenanthrene 66
3-Acetylphenanthrene 65
9-Acetylphenanthrene 66
M e t h y l 0-naphthyl ke tone 6S
(3-Acetotetralin i1
Cyclopen tanone fiiCyclohexanone 40
4-Methylcyclohexanone 40
o-Tetralone 6S
l -Ke to - l ,2 ,3 ,4 - te t rahydro-phenan th rene i 5
4-Keto - l ,2 ,3 ,4 - te t rahydro-phenan th rene i 5
l -Keto-9-methoxy-l ,2 ,3 ,4- te tra-
hydrophenan th rene70
2-Aoetylfuran 13
Furfuralacetone 49
2-Aoetyl th iophene 10 ' u
2-Acetyldibenzothiophene 14
4-Aoetyldibenzothiopheae 14
2-Aoetyl-4-phenylthiazole "Ant ipyr ine 6 ' 17
Ch r o m a n o n e 7 1
I sobu ty ra ldehyde 18
Isovaleraldehyde ls
Hex ahydrob enzaldehy de 18Tetrahydroisoguinoline hydro-
chloride, formaldehyde, andAoetophenone u2-Ace ty lphenan th rene 6B
Product (Yield)
C«H sCO CH ( C« H 5)C H 8N C 6H io (—)(3,4) (CH sO)2C6H 8COCH2CH2NC6Hio (—)C 6 H 6C H = C H C O C H 2C H 2 N C 5 H i o ( 6 0 % )(2)CH 3O C e H 4 C H = C H C O C H 2 C H 2N C5 H i o ( 1 3 % )(2)C2H 6O C 6H 4 CH = CH CO CH 2 CH 2 N C5 H i o ( 3 0 % )(2)C 3H r O C6 H 4CH = CH CO CH 2 CH 2 N C5 H i o ( 2 6 % )( 2 )C4 H 9 0 C 6H 4 C H = C H C O C H 2C H 2 N C 5 H i o ( 2 6 % )
(4)CH 8O C e H 4 C H = C H G O C H 2 C H 2 N C6H io (60%)(3,4) (CH2 0 2)Co H 3C H = C H C O C H 2 C H 2 N C 5 H 1 0 (60%)(3,4) (CH 30)2C 6H 3C H = C H C O C H 2 C H 2 N C6Hio (60%)
(4)CH 3O C6 H 4 CO CH ( CH 3 )CH2N C 6Hio (—)
(3,4) (C H 80) (C2H 60 )C6 H 8C H = C H C O C H 2 C H 2 N C6H io (—)
(3,4) (C2H5O) (C H sO )C6 H 3CH = CH CO CH 2 CH 2 N C6 H i o ( — )(3,4,6) (CH 2O 2) (NO 2) C6H 2C H = C H C O C H 2C H 2 N C 5H io
(60-65%)(3 ,4 ,6 ) (CH,0)2 (N0 2) C 6 H 2 C H = C H C O C H 2 C H 2N C 5 H i o
^-(jS-Piperidinopropionyl)-tetraUn (—)2-Piperidinomethylcyclopentanone (90%)2-Piperidinomethylcyclohexanone (62 % )2-Piperidinomethyl-4-methylcyclohexanone (93%)^-P ipe r id inomethy l -a - te t ra lone (75%)
l -Ke to -2 -p ipe r id inomethy l - l ,2 ,3 ,4 - te t rahydrophenan-threne (—)
4-Keto -3 -p ipe r id inomethy l - l ,2 ,3 ,4 - te t rahydrophenan-threne (—)
1 -Keto-2-piperidino-9-methoxy-l ,2 ,3 ,4- te trahydrophenan thren e
(63%)|3-Piperidinoethyl 2-furyl ketone (—•)C 4H 3 O C H = C H C O C H 2 C H 2 N C 6Hio (—)/3-Piperidinoethyl 2-thienyl ketone (74 % )/3-Piperidinoethyl 2-dibenzothienyl ketone (55%) t/S-Piperidinoethyl 4-dibenzothienyl ketone (40%) fP-Piperidinoethyl 4-phenyl-2-thiazolyl ketone (—)4-Piperidinomethylantipyrine (70 %)3-Piperidinomethyl-4-chromanone (28%)(C H s)2C( CH 2 N C 5H i o )CH O (—)(C H 8) 2 C H C H ( C H 2N C 6Hio)CHO (—)(C H 3 )2C H C ( C H 2O H ) ( CH 2 N C 6 H i 0 )CH O ( 7 0 % )1-Piperidinomethylhexahydrobenzaldehyde (—-)
2-(3-Benzoylethyl)- l ,2 ,3 ,4- te trahydroisoquinoline (—)2-( jS-l ,2 ,3 ,4-Tetrahydroisoquinolinopropionyl)-phenanthrene
(—)
* References 67-74 appear on p . 341.t Yield based, on the amount of original ketone not recovered from the reaction mixture.
( H O O C)2 CH CH 2 N CH 2 CH 2 N ( CH 2C H 2CO O H )CH 2 CH 2 t(17% ) 1 1
H O O CCH 2CH2NCH2CH2N(CH2CH2COOH)CH 2CH2 (19%)
I 1N,N'-Di-(ant ipyrylmethyl)-piperazine (—)
* A gu m m y pro duc t was obta ined w hen /3-hydrindone was used,t The piperazine base was used.
67 Mannich and Stein, Ber., 58, 2659 (1925).68 Blicke and Maxwell, J. Am. Chem. Soc, 64, 428 (1942).69 Ruberg and Small, / . Am. Chem. Soc, 60, 1591 (1938).70
Burger, / . Am. Chem. Soc, 60, 1533 (1938).71 Harradence, Hughes, and Lions, / . Proc. R oy. Soc. N. S. Wales, 72, 273 (1938).72 Mosettig, Shaver, and Burger, / . Am. Chem. Soc, 60, 2464 (1938).7 3 H a r r a d e n c e a n d L i o n s , J. Proc. Roy. Soc. N. S. Wales, 7 2 , 2 8 4 ( 1 9 3 8 ) .74 Kilhn and Stein, Ber., 70, 567 (1937).
The Low-Temperature Reaction in Nitrobenzene. 2-Methyl-4-hydroxy-acetophenone 354
The Preparation of a p-Hydroxyketone in the Absence of a Solvent. 3-Meth-
yl-4-hydroxybenzophenone 355
The Preparation of an o-Hydroxyketone. 2-Hydroxy-5-methylbenzophenone 355
Formation and Separation of a Mixture of o- and p-Hydroxyketones. o- and
p-Propiophenol 355
TABULAR SURVEY OP THE FR IE S REACTION 356
INTRODUCTION
The Fries reaction consists in the conversion of an ester of a phenol toan o- or p-hydroxyketone, or a mixture of both, by treatment with alum-inum chloride.
A second method avai lable for the syn thes i s of s imi la r compounds is the
Fr iedel-Craf ts react ion , in which a p h en o l , or an e th e r of a phenol , is
condensed with an acid chloride or ac id anhydr ide in the presence of
a luminum ch lo r ide . In sp i te of the f ac t tha t the Fr ies react ion requires
tw o s t ep s —th e p r ep a r a t io n of the es te r and the r e a r r a n g e m e n t to theh y d r o x y k e to n e—as co mp ar ed to the single step in the Fr iedel-Craf ts
syn thes i s , the Fr ies method usua l ly is to be prefer red for the p r ep a r a -
t ion of pheno l ic ke tones . The yields are ord inar i ly be t te r and the
exper imen ta l p rocedure does not h a v e to be modif ied great ly to a d a p t
i t to a v a r i e t y of es te r s .
Three d if ferent mechanisms for theFr ies r ea r rangemen t have r ece ived
ser ious considera t ion . In one of t h e m the es te r is as s u med to r eac t wi th
a luminum ch lo r ide to give an acid chloride and a p h e n o x y a l u m i n u mchlor ide which combine to form a der iva t ive of the h y d r o x y k e to n e .
OCOR OA1C12
+ A1C1, + RC0C1
OA1C12 OAICI2 OAICI2
+ RC0C1 -> HC1 + I J and
COR
In another scheme, it is proposed that one molecule of the phenyl ester
is acylated by another molecule.
OCOR OCOR OH OCOR
AlClj
COR
COR COR
In the third mechanism, the Fries reaction is considered to be a true
intramolecular rearrangement in which the acyl group shifts directly from
the oxygen atom to the carbon atom of the ring.Certain experimental facts can be cited to support each of these
mechanisms,1
but it has not yet been possible to prove or disprove any
The structure of the phenyl ester determines whether or not a Friesreaction will take place. If the reaction does occur with a particularester, the product may be either the o- or p-hydroxyketone or a mixtureof the two. The nature of the product is influenced not only by thestructure of the ester, but also by the temperature, the solvent, and theamount of aluminum chloride used. By variation of these last threefactors it is often possible to direct the course of the reaction so thateither of the isomeric ketones may be the major product from the sameester. Since it is usually possible to separate the twoketones, the syn-thesis is often useful even when it cannot be directed to the exclusiveproduction of one isomer.
Temperature
A tempera ture effect in the Fries reaction hasbeen observed by manyworkers.2
A striking example has been reported by Rosenmund andSchnurr,3
who found that at 25° only the p-hydroxyketone (80%) wasobtained from m-cresyl acetate and aluminum chloride, while at 165°only the o-hydroxyketone (95%) was formed.
OH
Similar observations were made in the rearrangement of m-cresyl ben-zoate; below 100°only the p-hydroxyketone (60%) wasformed, and at175° theortho isomer was the sole product (95%).
I t is not always possible to obtain at will either of the two possibleproducts simply byvarying the reaction temperature. Forexample, thehigher aliphatic esters of m-cresol yield theo-hydroxyketones even at lowtemperatures,4 and the esters of o-cresol yield p-hydroxyketones as theprincipal products even at high temperatures.3 Generally, however, lowreaction temperatures favor the formation of the p-hydroxyketones,andif these are desired it is good practice to keep the reaction temperatureat 60° or less.3
2
E y k m a n n , Chem. Weekblad, 1, 453 (1904).3 Rosenmund and Schnurr , Ann., 460, 56 (1928).4 Baltzly and Bass , J. Am. Chem. Soc, 65, 4292 (1933).
The Fries reaction can be carried out in the absence of a solvent, butthe temperature at which the reaction proceeds at a useful rate is low-
ered by the presence of nitrobenzene.3
'B
No data are available to showwhether the other solvents which have been employed, such as tetra-chloroethane 6 and chlorobenzene,7 also exert this influence. Carbondisulfide has been used in the Fries reaction in a rath er special wa y: thereaction is begun in this solvent, the carbon disulfide then removed bydistillation, and the reaction is completed by heating in the absence of asolvent.8 '9 There is little information at present concerning the effect ofdifferent solvents on the ratio in which the two isomers are produced.
However, it has been shown tha t, in the rearrangement of pheny l capryl-ate a t 70°, the proportion of the p-hydroxyketone formed in nitrobenzeneis higher than in tetrachloroethane (7 1% as compared to 63% ).10
Ratio of Ester to Reagent
The aluminum chloride and the phenyl ester are generally employed inapproximately equimolar quantities. However, in the rearrangem ent ofguaiacol acetate, two moles of aluminum chloride are required.11 The
suggestion has been made that one mole of aluminum chloride is used bycomplex formation with the alkoxyl group.11 It would be desirable tohave information on the effect of using two moles of aluminum chlorideper mole of ester with other, similarly constituted esters, for examplethe aceta te of resorcinol monom ethyl ether. I t has been found th at theproportion of p-hydroxyketone produced by rearrangement of phenylcaprylate in the presence of two moles of aluminum chloride is higher(63% para, 30% ortho) than th at in experiments in which only one mole
of the reagent is used (45% para, 33.5% ortho).™ It should be noted thatthe increase in yield of the p-hydroxyketone is at the expense of unre-acted material, no t a t th e expense of o-hydroxyketone.
Structure of the Acyl Radical
The acyl radical of the phenyl ester may be either aliphatic or aro-matic. Este rs of aliphatic acids as large as stearic acid have been used
6 Barch, J. Am . C hem. Soc, 57, 2330 (1935).6 Blicke and Weinkauff, J. Am. Chem . Soc, 54, 330 (1932).7 Wojahn, Arch. Pharm., 27 1 , 417 (1933).8 Cox, J. Am . C hem. Soc, 52, 352 (1930).9 Fieser and Bradaher, J. Am . C hem. Soc, 58, 1739, 2337 (1936).
10 Rals ton, McC orkle , and Bauer , / . Org. Chem., 5, 645 (1940).11 Coul tha rd , Marsha l l , and Pym an, / . Chem. Soc, 280 (1930).
successfully. Es ters of haloacetic acids and of alkyl, alkoxyl, and halo-gen-substituted aromatic acids have been employed. Este rs of purelyaliphatic unsaturated acids appear not to have been tried, but certainesters of cinnamic acid have been found to rearrange.
Rosenmund and Schnurr3 studied the relative rates at which the Fries
reaction takes place with different esters of the same phenol (thymol).Their observations are summarized in the following list, in which theacyl groups are arranged in the order of decreasing rate of reaction.
As an example of the magnitude of the differences in rate , it may be noted
th at after five hours in nitrobenzene solution a t 20° the rearrangem ent ofthymyl acetate was 60% complete and that of thymyl benzoate only 4%complete. The order of reactiv ity of the acyl groups is the same forrearrangement to either the o- or p-hydroxyketone.
The stability of esters containing the less reactive acyl groups some-times limits the usefulness of the Fries rearrangement. For example,the benzoate of a-naphthol does not undergo a Fries reaction at theordinary temperature.12 The butyrate of a-naphthol furnishes, afterseventeen to eighteen hours at 0°, 35% of 4-butyryl-l-naphthol and22% of 2-butyryl-l-naphthol.12 At 100-120°, the same ester furnishes3% of 4-butyryl-l-naphthol, 55% of 2-butyryl-l-naphthol, and 2% of2,4-dibutyryl-l-naphthol. 13
The rearrangement of p-cresyl cinnamate does not take place at temper-atures below th a t at which the ester undergoes decomposition.
With the aliphatic esters of certain phenols, an increase in the size ofthe acyl group favors the formation of o-hydroxyketone. This is par-ticularly true of the aliphatic acid esters of m-cresol, for only the acetatecan be converted to a p-hydroxyketone. The same tendency, althoughless pronounced, has been observed with aliphatic acid esters of a-naphthol. As indicated by the examples jus t cited, the importance ofthe size of the acyl radical in determining the course of the rearrangem ent
depends on the struc ture of the phenolic residue. Although it is prob-ably correct to state that an increase in the size of the acyl group of a
12 Lederer , J. prakt. Chem., [2] 135, 49 (1932).13 Stoughton , J. Am. Chem . Soc, 57, 202 (1935).
particular ester will increase the tendency toward formation of the o-hydroxyketone, it is still possible to prepare p-hydroxyketones contain-ing very large acyl groups. Phenyl palm itate and phenyl stea ra te, 10 '14
for example, furnish the p-hydroxyketones (19.7% , 21% ) when the reac-tion is carried out by heating the esters to 70° with aluminum chloride intetrachloroethane. The yield of p-hydroxyketone from phenyl palmi-tate is less than half that from phenyl caprylate but the ratio of para- toor^io-hydroxyketones is not greatly different with these two esters (1.32to 1.35).
10
Structure of the Phenoxyl Group
The struc ture of the phenolic portion of the ester is the factor of grea t-
est importance in determining whether a Fries reaction will take placeand whether the product will consist principally of the o- or p-hydroxy-ketone. The importance of this factor is revealed by examination of theproducts from esters of monosubstituted phenols. The presence of ameto-directing group on the aromatic portion of the phenyl ester usuallyinterferes with the Fries reaction. For example, the reaction does notoccur if the phenolic residue carries a nitro or benzoyl group in either theortho or para position; the presence of an acetyl or carboxyl group in theortho position hinders the reaction, and, in the para position, preventsit.3' 8
If the phenyl ester contains a single alkyl group in the phenolic ring,then the position of this substituent has a profound influence on thena ture of the product. For example, esters of o-cresol yield predomi-nantly p-hydroxyketones, esters of m-cresol yield predominantly o-hydroxyketones, and esters of p-cresol yield exclusively o-hydroxy-ketones. The effect of a para substituent has been observed with avariety of alkyl groups and with halogen; the effect of ortho substituents
has been observed with several alkyl groups; the effect of a meta substitu-ent has been determined only with esters of m-cresol.
The rearrangement products of more than fifty disubstituted phenolesters are shown in Part C of the tabular survey of the Fries reaction(p. 360). I t will be noticed th at with three esters, 2,5-dimethylphenylacetate, 2-ethyl-5-methylphenyl acetate, and 2-methyl-6-ethylphenylacetate, products formed by migration of an alkyl group were isolated.It is probable that these migrations were the result of the use of high
temperatures and prolonged reaction times and that they would notoccur if more gentle experimental conditions were used. Th us, the car-vacryl and thymyl esters yielded the expected p-hydroxyketones withoutmigration of alkyl groups when mild experimental procedures were used.
Again, 2-methyl-6-ethylphenyl aceta te yie lded 50% of the normalproduct together with some rearrangement product when the react ion
m ixtu re was hea ted for five hou rs. W hen more gentle conditions wereemployed the yie ld of the no rmal prod uct rose to 7 3 % and no rearrangedprodu ct was repor ted . In m an y of th e es ters the acyl group migrated
to the ortho posi t ion even though the para posi t ion was vac ant . This isdue in part to the presence of alkyl groups in the meta posi t ions . Appar-ently, i t is also due in part to the high temperatures used, s ince thees ters of carvacrol and thymol furnished the p-hydroxyketones under
the mild condit ions employed in their rearrangement.
COCH 3 C H 3
OCOCH3
COCH 3
CHaCOCOCH3 OH so%(73%)
In Part D of the tabular survey of the Fries reaction are given the
products obtained from the acetates of seven tr ialkylphenols , each ofwhich has a t leas t one vacant ortho or pa ra position. In th e second
experiment, the only product isolated was one involving migration of analkyl grou p, and on ly in the firs t and se ven th were such prod uct s entirely
absen t. In th e thi rd experimen t th e transfer of a m eth yl group fromone molecule to an oth er also occurred. A comparison of th e dat a ofPar t D of the tabular survey with those of Par t C indicates that migra-tions of alkyl groups occur the more readily as the number of such
groups is increased. Ho w ever, even w ith th e heav ily alky lated pheny lesters i t is probable that these migrations result from the drastic treat-
ment with a luminum chlor ide and that they are not an integral par t ofthe Fries reaction.
15 Auwers, Bund esman n, and W ieners , Ann , 447, 162 (1926).16 Auwers and Mauss , Ann., 460 , 240 (1928).17 Auwers and Janssen, Ann., 483, 44 (1930).
The migrat ion or removal of an a lkyl group somet imes permits the
Frie s reaction to occur even with esters of 2,4,6-trialk ylphen ols. H ow -
ever, s ince very drast ic condit ions are required, these forced reactions
must be investigated in each individual instance before they can be rel ied
upo n for pre par at ive purposes . P ar t D of th e tab ula r surve y of theFries reaction shows the products obtained from the 2,4,6-trialkyl-
phenols .
The esters of the three dihydroxybenzenes, catechol, resorcinol , and
hydr oquin one, undergo the Fries react ion . Es ters of catechol y ie ld pre-
dominantly 4-acylcatechols and secondari ly 3-acyleateehols.
CH3CO1
The usual technique may be employed with these esters , but i t is prefer-
able to treat an equimolar mixture of a diester and catechol with alum-
' inum chlor ide .18 ' 19 Resorcinol esters can be converted to 4-acylresor-
cinols or to 4,6-diacylresorcinols using a variety of techniques, 2 0 '21> 22 b u t
4-acylresorcinols are so readily obtained directly from resorcinol and the
acids or acid chlorides that the Fries reaction is seldom used for their
p rep a ra t i o n .2 3
'2 4
OCOCH3 OH OH
O CH3CO(T
OH
COCH3
The Fries rearrangement of the acetate of 4-acetylresorcinol furnishes
th e 2,4- (58% ) an d th e 4,6-diacylresorcinol (4 2% ). T he form ation of
the 1,2,3,4-tetrasubsti tuted derivative is explained as a consequence ofchelat ion which stabil izes the Kekule form leading to the 2,4-diacyl
compound. 2 5
OCOCH3 OH OH
LyUOJtl3 L/Xl3OUf| ^ |
OH l ^ ^ O HCOCH3 COCH3 CO CH 3
"Rosenmund and Lohfert, Ber., 6 1, 2601 (1928).19
Miller, H ar tu n g , Rock, and Croasley, / . Am. Chem. Soc, 60, 7 (1938).20 K la r m an n , J. Am. Chem. Soc, 48, 23S8 (1926).21 Rosenmund and Schulz, Arch. Pharm., 265 , 308 (1927).22
Acyl derivatives of a-resorcyclic acid (3,5-dihydroxybenzoic acid) arereported not to give the Fries reaction.26
Th e ace tate of guaiacol furnishes th ree prod uc ts in the F ries reaction.27
Part icular ly to be note d is the presence of a m-hy droxyketon e amon g theproducts , for the formation of m-hydroxyketones in the Fries reaction is
exceedingly rare.
OCOCH3 OH OH OH
hOCHs
COCH 3
26.3% 1% 5.6%
The Friedel-Crafts reaction with guaiacol and acetyl chloride furnishes
the same three products , making i t evident that the formation of them-hydroxyketone is re la ted to the ortho methoxyl group and is not a
pecu liarity of th e Fries reaction . T he resorcinol de riva tive yields an0- and a p-hydroxyke tone (12%, 11%) but no m-hydroxyke tone .26
OCOCH3 OH OH
OH j C o
O C H 3 +^
COCH 3
Esters of pyrogallol,28 phlorogluc inol , 2 8 '2 9 '3 0 1,2,4-tr ihydroxyben-zene,30 and of a number of hydroxydimethoxybenzenes and dihydroxy-methox ybenzenes hav e been s tudied. Th e prod ucts obtained from these
esters are, with few exceptions, those to be expected and the yields areusually q uite small. T he use of mo re th an one mole of alu m inu m chlorideper mole of th e ester m igh t give be tte r resu lts . I t has been repo rted t h at2,6-dimethoxyphenyl aceta te , with zinc chlor ide a t room temperature
in acetyl chloride as the solvent, furnishes the unsymmetrical product,
the acetyl group taking a meta position.31
The same es ter on treatmentwith a luminum chlor ide yie lds the p-hydroxyketone.32
OH OCOCH3 OH
C 0 C H 3
8% 7.5%26 M a u t h n e r , J. prakt. Chem., [2] 136, 205 (1933).27
Reichs te in, Helv. Chim. Acta, 10, 392 (1927).28 Heller, Ber., 45, 2389 (1912).29 Heller, Ber,, 42, 2736 (1909).30 M a u t h n e r , / . prakt. Chem., [2] 139, 293 (1934).31 M a u t h n e r , J. prakt. Chem., [2] 118, 314 (1928).32 M a u t h n e r , J. prakt. Chem., [2] 121 , 255 (1929).
The diaceta te of 2-methoxy-l ,4-dihydroxybenzene undergoes a Fr iesreact ion a nd yie lds the dihyd roxyk etone (38% ) .26
OCOCH3
OCH3
CH 3O
OCOCH3
Esters of a-naphthol furnish 4-acylnaphthols a t low tempera-tures.12 ' 1 3 '3 3 With an increase in the s ize of the acyl group, the rate offormation of the 4-acylnaphthols falls off to such an extent that them etho d is of l i t t le valu e for their prep aratio n. Increasin g th e tem pe ra-ture results in the formation of 2-acylnaphthols and 2,4-diacylnaphthols .
/3-Naphthyl aceta te furnishes l -acetyl-2-naphtho l ( 3 3 ^ 0 % ) togetherwith 6-acetyl-2-naphthol ( 5 % ) . 3 3 ' 3 4 ' 3 5
COCH3
CH 3C
In the phenanthrene series the Fries reaction offers no advantage overth e Friede l-Crafts m eth od for it eithe r leads to difficultly sepa rable orinseparable m ixtures (2-acetoxy- and 3-ace toxy ph enan thren e) or fur-
nishes the same products as the Friedel-Crafts reaction but in nobetter yields (9-acetoxyphenanthrene) .3 6
With one interesting exception, the directive influence of the phenylgroup in esters of the hydroxybiphenyls is s imilar to that of the methylgro up in esters of th e cresols. T hu s, esters of 2-h yd rox ybi ph eny l furnish3-acyl- and 5-acyl-2-hydroxybiphenyls ,37 the yield of the former increas-
ing with the size of the acyl group.38 Esters of 3-hydroxybiphenyl fur-nish 4-acyl-3-hydroxybiphenyls .38 However, with esters of 4-hydroxy-biphenyls the acyl group migrates to the para position of the secondbenzene ring, yielding 4'-acyl-4-hyd roxybiph enyls as well as the expected
3-acyl-4-hydroxybiphenyls .6 ' 9 ' 3 9 ' 40
33Witt and Braun, Ber., 47, 3216 (1914).
34Friea, Ber., 54, 709 (1921).
36 Fries and Schimmelschmidt, Ber., 58, 2835 (1925).36
Moaettig and Burger, J. Am. Chem. Soc, 55, 2981 (1933).37
Auwers and Wittig, J. prakl. Chem., [2] 108, 99 (1924).38
Harris and Christiansen, J. Am. Pharm. Assoc, 23, 530 (1934).39 Hey and Jackson, J. Chem. Soc, 802 (1936).
With the aceta te of 4-hydroxybiphenyl the 4-hydroxy-3-ketone is the
pr incipal product; with the benzoate the 4-hydroxy-4 ' -ketone is thepr incipal product .
The Fries reaction of esters of hydroxycoumarins proceeds normally
to yield the o-hydroxyketones.41 ' 42> 43' 44 The react ion with the acylderivatives of 4-methyl-7-hydroxycoumarin, made from resorcinol andacetoacetic ester, provides a synthesis of 2-acylresorcinols.45
RCO
COR
Although es ters of hydroxycoumarins rearrange readi ly, a t tempts tocarry out the Fries reaction with acetates of the following hydroxy-chromanones have been unsuccessful.46
OH
,COCH3
4 1
Desa i and Hamid , C. A., 32, 1254 (1938).42 Limaye , Ber., 67, 12 (1934).43 Limayo and Munje , C. A., 32, 2096 (1938).44 Sethna, Shah, and Shah, C. A., 32, 549 (1938).46 Russel l and Frye, Org. Syntheses, 21, 22 (1941).46 Kelkar and Limaye, C. A., 31 , 2214 (1937).
nected with an air or water condenser. I t is advisable to use a largeflask as the mixture often foams during the reaction. The flask is placedin an oil bath, heated slowly to 120°, and kept at that temperature forfifteen minutes. The heating should be done cautiously as the heat ofreaction is often large. The upper tem pera ture may be higher than
120°, bu t it is desirable to keep the tempera ture as low as possible. Aftercooling, ice and dilute hydrochloric acid are added.
Boron fluoride has been used to bring about the Fries reaction, but nodetails of its use are available.47
Several procedures are available for working up th e reaction mixtures.Nitrobenzene or tetrachloroethane, when present, can be removed bydistilling with steam. Alternatively, the reaction mixture can be ex-tracted with ether and the product isolated by extraction of the ether
solution with aqueous sodium hydroxide.Mixtures of o- and p-hydroxyketones often can be separa ted by virtue
of the fact th a t the latte r are not volatile with steam. If the o-hydroxy-ketone is of such large molecular weight th a t it is not volatile with steam,a separation may be effected by distillation at ordinary or reduced pres-sure. Thus, o-heptanoylphenol boils a t 135-140° (3 mm .) while the paraisomer boils at 200-207° (4 mm.).48 If the o- and p-hydroxyketones areboth solids, a separation often can be effected by tak ing advantage of the
fact that the ortho isomer will be the more soluble in ligroin. Again, it isfrequently possible to separate a pair of isomeric o- and p-hydroxy-ketones by extracting with dilute sodium hydroxide an ether solutioncontaining bo th isomers. Th e p-hydraxyketone is extracted morereadily.
EXPERIMENTAL PROCEDURES
The Low-Temperature Reaction in Nitrobenzene
Preparation of 2-Methyl-4-hydroxyacetophenone.3
To a solution of10 g. of o-cresyl acetate in 50 g. of nitrobenzene is added in small por-tions 10 g. of aluminum chloride. The reaction mixture is left to standfor twenty-four hours at room temperature and then is poured onto iceand dilute hydrochloric acid. The nitrobenzene is removed by steamdistillation, and the residual crude 2-methyl-4-hydroxyacetophenone ispurified by vacuum distillation. The yield is 8.0-8.5 g. (80 to 85%) ofpure ketone, m .p. 128°.
The Preparation of a ^-H ydroxyketone in the A bsence of a S olvent
Preparation of 3-Methyl-4-hydroxybenzophenone.49 Fifty grams ofo-cresyl benzoate is heated to 130° and stirred while 40 g. of aluminumchloride is added. The tem perature is raised to 160° and kept there forforty-five minutes. After cooling, the reaction mix ture is decomposedwith dilute hydrochloric acid, and the crude product is filtered and dried.On distillation, the material boils at 240-260° (12-15 mm.) and furnishes45.5 g. (90%) of pure ketone, m.p. 173-174°.
The P reparation of an o-H ydroxyketone
Preparation of 2-H ydroxy-5-methylbenzoph enone. In a 1-1., three-necked, round-bottomed flask fitted with a thermometer and an air con-denser are placed 75 g. (0.35 mole) of p-cresyl benzoate and 60 g. (0.44mole) of aluminum chloride. The reac tants are mixed by shaking, andthe flask is then placed in an oil ba th a t 90°. After the reaction mixturehas melted, heat is applied to the bath, rapidly until the temperature ofthe mixture reaches 120°, then slowly until it reaches 140°. The reac-tion m ixture is kept at this temperature for ten m inutes, the thermometer
is removed from the flask, and the flask is removed from the bath. Whenthe reaction mixture is cold, it is added to a stirred mixture of 250 g. ofice and 150 cc. of concentrated hydrochloric acid. After the ice hasmelted, the solid product is filtered and dried. The yield is 70-73 g. of ayellow solid which is pure enough for m ost purposes bu t which contains asmall amount of impurity that lowers the melting point considerably.The ketone may be purified by distillation with superheated steam fol-lowed by crystallization from ethanol. I t then melts at 83-84°, and theyield is 60 g. (80% ).
Formation and Separation of a M ixture of o- and ^-H ydr oxy keton es
Preparation of o- and ^-Propiophenol.6 0 In a 2-1., three-necked,round-bottomed flask fitted with a reflux condenser, a sturdy mechanicalstirrer, and a 100-cc. dropping funnel are placed 374 g. (2.8 moles) ofaluminum chloride and 400 cc. of carbon disulfide. Stirring is begun, and375 g. of phenyl propionate is added a t such a rate th a t the solvent boils
vigorously. When the addition is complete, the reaction mixture isboiled on the steam ba th for about two hours ; then the reflux condenseris turned downward and the solvent is removed by distillation. The
49 Cox, J. Am. Chem . Soc, 49, 1029 (1927).60 Mil le r and Har tung, Org. Syntheses, 13, 90 (1933).
flask is next heated for three hours in an oil bath maintained at 140-150°, stirring being continued as long as possible.
The reaction mixture is allowed to cool and is decomposed by thecautious addition of a mixture of 300 cc. of water and 300 cc. of concen-
tra ted hydrochloric acid, followed by 500 cc. of water. On stand ingovernight, most of the p-propiophenol in the upper oily layer solidifiesand is removed by filtration . I t is crystallized from 400 cc. of methanoland furnishes 129-148 g. (34-39%) of light yellow material melting at145-147°. A second crystallization raises the melting point to 147-148°.
The oily filtrate and the concentrated mother liquors from the aboverecrystallization are dissolved in 500 cc. of 10% aqueous sodiumhydroxide and extracted with two 100-cc. portions of ether to removenon-phenolic products . The alkaline solution is acidified, and the oilylayer is separated, dried over anhydrous magnesium sulfate, and dis-tilled. The distillation furnishes 120-132 g. (32-35% ) of o-propiophenolboiling at 110-115° (6 mm.) and 40 g. of p-propiophenol boiling at 135-150° (11 m m.). The total yield of crude p-propiophenol is 169-188 g.(45-50%).
TABULAR SURVEY OF THE FRIES REACTION
The use of one mole of aluminum chloride per mole of ester is to be
understood unless a different ratio of aluminum chloride to ester or adifferent reagent is specified. The position of the acyl group in theproduct is always given with reference to the hydroxyl group as 1; ifmore than one hydroxyl group is present, the numbering is such as togive the lowest numbers to the carbon atoms carrying the hydroxylgroups. Where a product is listed but no yield is given, the product wasreported in the litera ture with no information abou t the yield.
| hr . a t 160-180°48 hr. at room temp.\ hr. at 160-180°\ hr. at 160-180°15 min. at 140°
Add reagent at 130°; 45min. at 160°
100-120°C6H5NO2; overnight at
room temp., 3-4 hr. at50-60°
170°C 6HfiNO2;24hr. at 20°165°
5 hr. at 150°C 6 H 6 N O 2 ; 1 0 d a y s a t 2 o
120-140°C 6 H 6 N O 2 ; 1 0 d a y s a t 2 °120-140°
C 6 H B N O 2 ;24hr . a t 20°C 6H 6 NO2;24hr . a t 20°120-140°C 6H 6N O 2; 24 hr. at 20°120-140°C 6H 6NO2; 5 hr. at 60°15 min. at 175°CS2; 3 hr. at room temp.,
Hea t on s team bath130-140°130-140°6 hr. at 130-140°3 hr. at 130-140°
130-140°5i hr. at 130-140°130-140°
V
10 min. at 120°
10 min. at 140°
2 hr. at 120°18 hr. at room temper-
ature, heat to 120°
130-140°
C 6H 6N O 2; 24 hr. at 25°
C 6 H 6 N O 2 ; 5h r . at 60°
C 6H 6 N O 2 ; 18hr . at 20°C 6 Hj j NO 2 ; 24hr .a t20°
C 6H 6N O 2; 5 hr. at 60°
C 6H 6NO 2; 48 hr. below20°
Overnight at roomtemp., 6 hr. at 120°
Products
%6-Acyl
60
69No yield
747770
No yield6763
No yieldQuanti-tat ive a
92No yield
2,4-Dimethyl-
6-acetyl,17
2-Ethyl-4-methyl-6-acetyl,
40—
—
——
—
—
—
%4-Acyl
—————
———..—.—
—• —
70
—
90 *
60
828 7 c
70
80
81
Refer-ence *
16
5715171617
151617173
35315
15
3
3
33
3
3
16
* References 51-64 appear on p. 369.
° Comparable results are reported 3 with the propionate and butyrate of 2-chloro-4-methylphenol.Comparable results are reported 3 with the propionate, butyrate, and isovalerate of carvacrol.
c Comparable results are reported3 with the propionate, butyrate, isovalerate, phenylacetate,caprylate, and hydroeinnamate of thymol.
4-Acyl, QuantitativeNo pure product4-Acyl, 2 6 % ; 5-acyl,
5.6%;6-acyl , 1%
4-Acyl, 30%4-Acyl, 5 0 % ; 4-acyl-catechol °
4-Acylcatechol, 5 1 %4-Acylcatechol, 23-
6 2 %4-Acylcatechol, 30-
4 7 %4-Acylcatechol, 8-
17%
2,6-Diacylb
2,6-Diacyl , 40-50% c
4-Acyl, 1 1 % ; 6-acyl, 12%
6-Acyl, Quantita-tive d
Refer-ence *
1818
18
18
19
19
18
19
19
18
18
2 7
1811
19
19
19
19
2220
26
22
* References 51-64 appear on p. 369.a Comparable results are reported u with the butyrate, valerate, and heptanoate.
Comparable results are reported ^ with the dipropionate, dibutyrate, and divalerate.c Comparable results are reported 20 with the dipropionate, dibutyrate, dicaproate, and dilaurate.d Withou t the added mole of 4-ethylresorcinol, the yield is 47% . Comparable results are reported M
with the dipropionate, dibutyrate, diisovalerate, dicaproate, and dibenzoate.
2 A l C l3 ; 3 - 4 h r . a t 6 0 -70°or5hr . a t 110°
2 A l C l 3 ; 3 - 4 h r . a t 6 0 -70°or5hr . a t 110°
Same as 4-ethylresor-cinol diacetate in
2 AlC l3 ;4hr .a t60-70°
24 hr. at room tem-perature
Add one mole of 4-benzylresorcinol,3-4 hr. at 50°
2 hr. at 75°
2 hr. at 75°
Z n C l 2 ;2hr . a t 145°24 hr. at room tem-
perature
24 hr. at room tem-perature
ZnCl2; 3hr. at 120°
ZnCl2; 4 weeks a troom temperature
8 hr. at 100°
C"
Products
2,6-Diacyl, 50%
2,6-Diacyl, 50%
6-Acyl •
2,6-Diacyl, 40%
6-Acyl, 50 %
6-Acyl, 8 5 % !
2-Acyl, 23%
2-Acyl, 24%
5,6-Diacyl, 26%6-Acyl, 6 1 %
4-Acyl, 7. 5%
3-Acetyl-6-methoxy-1,2-dihydroxyben-zene, 8 %
3-Acetyl-6-methoxy-1,2-dihydroxyben-zene, 10%
3-Chloroacetyl-pyrogallol
Refer-ence *
22
22
22
22
26
7
18
18
2826
32
31
31
31
* References 51-64 appear on p. 369.
* Comparable results are reported 22 with the dipropionate, dibutyrate, divalerate, dicaproate, anddibenzoate.
' Comparable results are reported 1 with the dipropionate, dibutyrate, and diisovalerate of 4-benzyl-resorcinol and with the same esters of 4-(/S-phenylethyl)-resoroinol.
" A u w e r s , Ber., 47, 3319 (1914)."Fr ies , Hasse lbach, and Schroder , Ann., 405, 369 (1914).67 Smith and Opie, J. Org. Chem., 6, 427 (1941).68 Auwers, Ber., 48, 90 (1915).69 Auwers and Borsche, Ber., 48, 1708 (1915).60 Imoto , J. Chem. Soc. Japan, 58, 932 (1937) [C. A., 32, 534 (1938)].6 1 Fries and Frells tedt, Ber., 54, 717 (1921).62 Moset t ig and Stuar t , J. Am. Chem. Soc, 61, 1 (1939).8 3 Limaye , Ber., 65, 375 (1932).64 Limaye and Shenolikar, C. A., 32, 2096 (1938).
The migra t ion of an alkyl group or a halogen a tom in a sulfonic acidderived from a polyalkylbenzene, a halogenated polyalkylbenzene, or a
polyhalogenated benzene is known as the Jacobsen react ion. The reac-t ion isne arly alw ays effected by t rea t ing thehydrocarbon or halogenatedhydrocarbon with concentrated sulfuric acid and allowing the resultingsulfonic acid to remain in contact with the sulfuric acid. The firstobservat ion of a rea r rangement of this kind wasm a d e by He rzig * (1881),
who recorded the rea r rangement of a polyhalogenated benzenesulfonicacid. However, the react ions have taken the n a m e of Oscar Jacobsen 2
(1886), who discovered the rea r rangement of polyalkylbenzenesulfonicacids.
The migra t ions of the Jacobsen react ion may be divided into two
general types : (a) intramolecular , in which the migrat ing group movesfrom one position to ano the r in the same m olecule; and (6) intermolec-ular , in which there is a transfer of one or more groups f rom one mole-cule to another . In most cases, migrations of both types occur s imulta-
neously. An important character is t ic of the reaction is the migrat ion of
the alkyl groups to vicinal positions. The rea r rangement of durenesul-fonic acid is a typical example.
It is known with certainty that the rearrangements of the Jacobsenreaction involve the sulfonic acids, not th e hydrocarbons.3 This is shownby the fact that durenesulfonic acid rearranges in contact with phos-phorus pentoxide, a reagent which has no effect on durene itself. Also,the sulfonic acid from pentamethylbenzene rearranges when left in a
desiccator over concentrated sulfuric acid, whereas the hydrocarbon isunchanged under the same conditions. As yet no completely satis-factory explanation has been advanced regarding the function of thesulfonic acid group in prom oting the rearrangem ent.4 Nor is it possibleto account for the side reactions which occur during the course of theJacobsen reaction . The by-products are sulfur dioxide and polymericmaterials ranging from tars to insoluble, infusible solids. I t is knownthat part of the sulfur dioxide is in some way liberated from the sulfonic
acid during rearrangement, while the remainder results from the oxidiz-ing action of sulfuric acid on the organic substances present in the reac-tion mixture.
THE SCOPE OF THE REACTION
The Jacobsen reaction has been limited, with few exceptions, to thepolyalkylbenzenes, halogenated polyalkylbenzenes, and halogen deriva-tives of benzene. The substituents which have been shown capable of
migration are CH3 and C2H5 (the only two alkyl groups studied), I, Br,Cl, and SO3H. No Jacobsen rearrangement of compounds containingamino, nitro, methoxyl, or carboxyl groups is known.
The ease with which rearrangem ent takes place depends on the groupsattached to the benzene ring. If only halogen is present, rearrangementoccurs even when the benzene ring carries but one substituent. If bo thhalogen and alkyl groups are attached to the ring, then rearrangementoccurs the more readily the greater the number of alkyl groups, provided
th at at least one unsubstituted position is present. If only alkyl groupsare present, then rearrangement occurs only with the tetra- and penta-alkyl deriva tives. Thus, the sulfonic acids derived from the trialkyl-benzenes, hemimellitene,6 pseudocumene,3 mesitylene,3 1,2,4-triethyl-benzene,6 and 1,3,5-triethylbenzene 6 are stable to sulfuric acid.
The synthetic value of the Jacobsen reaction lies in the formation ofvicinal derivatives by migration of the alkyl groups of compounds con-taining these groups in non-vicinal positions. Thus, the tetram ethyl-,
tetraethyl- and trimethylethyl-benzenes of non-vicinal orientation rear-3 Smith and Cass , J. Am. Chem . Soc, 54, 1614 (1932).4 Moyle and Smith, J. Org. Chem., 2, 112 (1937).6 Smith and Moyle , J. Am . C hem. Soc, 58, 1 (1936).6 Smith and Guss , J. Am . C hem. Soc, 62, 2631 (1940).
Tetramethylbenzenes. Theequation for therearrangement of durene-
sulfonic acid2> 8 is shown on p. 371. Prehnitenesulfonic acid has beenobtained in 70% yield when the reaction wascarried out by sulfonatingdurene with concentrated sulfuric acid and allowing the sulfonationmixture to stand for twenty-five days at room temperature.3
Theotherproducts were sulfur dioxide, carbon dioxide, andvery small amounts of5-pseudocumenesulfonic acid and hexamethylbenzene. About 30% ofthe reaction product was a brown amorphous m aterial.
Isodurenesulfonic acid rearranges to prehnitenesulfonic acid, but the
yield is somewhat less than that obtained from durene.8
'9
The by-products areessentially the same as those from durene.
CH3
O S O 3 H Main reaction
C H 3 50% ^
C H g C H3
Isodurenesulfonic acid Frehnitenesulfonic acid
Prehnitene is sulfonated by sulfuric acid, and the sulfonic acid doesnot rearrange.3
The 1,2,4,5- and 1,2,3,5-tetraethylbenzenes 6l 1OFU rearrange to giveproducts analogous to those obtained from the tetramethyl derivatives.However, the reactions with the tetraethylbenzenes are much morerapid (15minutes at 100°), and the yield of 1,2,3,4-tetraethylbenzene is90-92%. The rearrangements of the tetraethylbenzenes are the onlyrecorded instances of Jacobsen reactions in which the tarry, polymeric
by-product is entirely absent andpractically nosulfur dioxide is evolved.Ethyltrimethylbenzenes.12 '
13The sulfonic acids of l,2,4-trimethyl-5-
ethylbenzene (5-ethylpseudocumene) and l,3,5-trimethyl-2-ethylben-zene (ethylmesitylene) rearrange to that of l,2,4-trimethyl-3-ethyl-benzene (3-ethylpseudocumene). The yields are relatively low, owingto side reactions which involve elimination of the ethyl group or one ofthe methyl groups. In the chart on p. 375, the sulfonic acid groups arenot included in the formulas because their exact positions are unknown.
Pentamethylbenzene and Pentaethylbenzene. The rearrangement ofpentamethylbenzenesulf onic acid Ui 15 is intermolecular, a methyl groupbeing transferred from one molecule to another.
+ 30% amorphousmaterial
Pentaethylbenzene 6i 10> 12> 16 undergoes a similar reaction, but the*yields (20-30%) are inferior to those obtained in the pentamethyl-benzene rearrangement. The by-products are formed in much largerquan tities. This is in contras t to the tetraethylbenzenes, which rear-
range more readily than th e tetram ethyl derivatives.Hexamethylbenzene 2l 3 is not affected by sulfuric acid.O ctahydroanthracene. Octahydroanthracene-9-sulfonic acid 17 is re-
arranged by sulfuric acid to octahydrophenanthrene-9-sulfonic acid.The reaction is rapid (20 minutes at 90-100°), and yields as high as85% have been obtained . This is one of the rare cases in which theaction of sulfuric acid is improved by the presence of a diluent. Sulfuricacid containing a little acetic acid is used to effect sulfonation and rear-
rangement.
SO3H SO3H
Octahydroanthracene-9- Octahydrophenanthrene-9-flulfo nic acid sulfonio acid
H alogenated Polyalkylbenzenes
4-Iodo-m-xylene.18'19 When 4-iodo-m-xylene is treated with concen-trated sulfuric acid and the reaction mixture is allowed to stand forseveral weeks the products isolated (in unspecified yields) are di- andtetra-iodoxylenes and an iodoxylenesulfonic acid.
14 Jaoobsen, Ber., 20, 896 (1887).16 Smi th and Lux, / . Am. Chem. Soc, 51 , 2994 (1929).16 Smith and Guss , J. Am . Chem . Soc, 62, 2634 (1940).17 Schroeter and Gotzsky, Ber., 60, 2035 (1937).18 Hammer ich , Ber., 23, 1634 (1890).19 Tohl and Bauch, Ber., 23, 3117 (1890); 26, 1105 (1893).
No rearrangements have been reported formonohalogen derivatives of o-and p-xylenes.
5- (and 6 -)H alopseudocumenes. The sulfonic acids of 1,2,4-tri-
methyl-5-chloro- and l,2,4-trimethyl-6-chloro-benzenes rearrange to thesulfonic acid of l,2,4-trimethyl-3-chlorobenzene (3-chloropseudocumene)in yields of 71 and 44%, respectively.5 Apparently both reactions
involve intramolecular migration of the halogen atom.
The corresponding 5-bromo compound is converted to the sulfonic acidof l,2,4-trimethyl-3-bromobenzene (3-bromopseudocumene) in 90%yield.6-20
A small amount of l,2,4-trimethyl-3,5,6-tribromobenzene(tribromopseudocumene) is obtained as a by-product.
l,2,4-Trimethyl-5-iodobenzene 21 (5-iodopseudocumene) gives rise totwo diiodopseudocumenes, an iodopseudocumenesulfonic acid, andpseudocumene-5-sulfonic acid; the yields are not reported.
i
CHCH
!H3
5-IodopBeudo-cumene
3-Iodopseudo-cumene-5-auLfonic
acid
3,6-Diiodo-pseudocumene
5,6-Diiodo-pseudocumene
O 3H
Pseudocumene-5-sulfonic acid
H alomesitylenes.5'22 '23
The sulfonic acid of chloromesitylene appearsto be stable, but that of bromomesitylene rearranges easily to give amixture of mesitylenesulfonic acid, dibromomesitylene, and tribromo-
mesitylene. The sulfonation of iodomesitylene leads to analogous2 0
methyl-4-chlorobenzene, l,2,3,4-tetramethyl-5-chlorobenzene) rearrangeto pentamethylchlorobenzene and l,2,4-trimethyl-3-chloro-5-benzene-sulfonic acid. In these reactions migration of a methyl group mustoccur.
Cl Cl ClC H 8 f r ^ S C H 3
ClChlorodurene Pentamethylchloro-
benzene3-Chloropseudo-
cumene-5-sulfonicacid
CH3
CH3Chloroisodurene
C H 3l l v ^ J C H 3
CH3Chloroprehnitene
The corresponding bromo compounds, l,2,4,5-tetramethyl-3-bromo-benzene (bromodurene 2 4
'2 6
) , l,2,3,5-tetramethyl-4-bromobenzene (bro-moisodurene 26), and l,2,3,4-tetramethyl-5-bromobenzene (bromoprehni-tene) react differently. No migration of a methy l group occurs, bu tthe bromine migrates intermolecularly to give dibromo compounds. Thesulf onic acid from which the bromine has been removed is th at of durene,
24 Smith and Moyle , J. Am. Chem . Soc, 55, 1676 (1933).26 Tohl , Ber., 25, 1527 (1892).26 Jacobsen, Ber., 20, 2837 (1887).
4,6-Dibromo-m-xylene 28 rearranges in the same way, forming 2,4-dibromo-m-xylene (about 2 5 % yield) . Fro m the behavior of other
halogen compounds , i t is l ikely that the halogen a tom is the migrat inggroup, a l though the same products would be produced by migrat ion of a
methyl group.Dihalogen derivatives of o- and p-xylenes have been reported to rear-
range, but the yields of definite products were very low.27
5,6-Dibromopseudocumene.2 9 l ,2 ,4-Trimethyl-5,6-dibromobenzene
(5,6-dibromopseudocumene) was treated with chlorosulfonic acid byJacobsen . Sulfonation was accom panied by th e forma tion of tr ibro m o-pseudo cum ene an d l ,2,4-tr imethyl-6-bromobenzen e-3-sulfonic acid. T heyields were not repor ted, but the main product isola ted was t r ibromo-
3 - (and 6 -)H alo-5-fluoropseudocum enes.30 Only a few fluoro com-
poun ds have been inves t igated in connect ion w ith the Jacobsen react ion.N o instanc e of m igration of a f luorine ato m has been reported . Fo rexample, 5-fluoropseudocumene undergoes no rearrangement when it issulfonated and the sulfonic acid is left in contact with sulfuric acid for
thre e m on ths . W hen 3- (or 6-)bromo-5-fluoropseudocumene is trea tedwith sulfuric acid, rearrangement involving intermolecular migration ofthe bromine a to m occurs . T he me thyl groups are unaffected. T heanalogous chloro-5-fluoropseudocumenes give the corresponding dichloro-
Br Br SO3H
CHJ
3-Bromo-5-fluoro- 3,6-Dibromo-5- 5-Fluoropseudocumene-pseudocumene fluoropseudoc um ene 3-aulfonic acid
f luoropseudocumene and the same fluoropseudocumenesulfonic acid;
yields are not reported.28 Jacobsen, Ber., 21 , 2827 (1888).29 Jacobsen, Ber., 19, 1221 (1886).30 Tohl and Muller , Ber., 26, 1108 (1893).
The reactions of bromobenzene, p-dibromobenzene, and 1,3,5-tri-
bromobenzene x with sulfuric acid have been studied . In all cases sul-fur dioxide and carbon dioxide are evolved and only small yields ofdefinite products result. Bromobenzene is converted to a dibromoben-zenesulfonic acid, probably the 1,3,5-isomer; p-dibromobenzene yields1,2,4,5-tetrabromobenzene and hexabromobenzene; 1,3,5-tribromoben-zene yields hexabromobenzene.
Iodobenzene31'32 is converted by sulfuric acid to p-diiodobenzene andbenzenesulfonic acid, with liberation of some iodine and hydriodic acid.o- and p-Iodotoluenes undergo a similar reaction. p-Diiodobenzene and
fuming sulfuric acid give a mixture of tri - and tetraiodobenzenes;33 experi-mental details are lacking.
I t is quite obvious tha t the Jacobsen reaction as applied to halogenatedbenzenes to form polyhalogenated benzenes is not one of practical syn-thetic value.
There appear in the literature 34 some rearrangements of 1,8-dichloro-naphtha lene which resemble the Jacobsen rearrangem ent. When 1,8-dichloronaphthalene is heated with hydrochloric acid at 290°, a rear-
rangem ent to 1,5-dichloronaphthalene occurs. A similar conversionalso results from the action of sulfuric acid, but considerable decomposi-tion occurs simultaneously. Heating with phosphoric acid or in theabsence of any acid fails to bring about a rearrangement of the dichloro-naphthalene. Only the 1,8-dichloro isomer undergoes rearrangement.The l,8-dichloro-4-naphthalenesulfonic acid is hydrolyzed by acid at230° to give 1,8-dichloronaphthalene; the l,8-dichloro-3-naphthalene-sulfonic acid, however, undeigoes hydrolysis only if a temperature of285° is reached, and then a mixture of the 1,8-, the 1,5- and the 1,7-dichloro derivatives results.
EXPERIMENTAL PROCEDURES
1,2,3,4-Tetramethylbenzene (Prehnitene)
From Pentamethylbenzene.18 To 74 g. of pentamethylbenzene(m.p. 52°) hea ted to 65°, 200 g. of concentra ted sulfuric acid is added
and the mixture is shaken vigorously. This procedure results in a mushof fine crystals of the hydrocarbon in the sulfuric acid; lumps must be31
Neumann, Ann., 241, 33 (1887).32
Cass, P h . D . thesis, University of Minnesota, 1931.33
avoided; if any are formed they should be broken up . The reactionmixture of crystals and red liquid is allowed to stand at room tempera-ture for twenty-four hours, the n cooled in an ice-salt bath. To it is nowadded 165-200 g. of cracked ice in three portions with vigorous stirring.
The cold mixture is filtered and the filter cake pressed as dry as possible;the p recip itate is then stirred with 700 cc. of cold water and again filtered.The product is a mixture of hexamethylbenzene and tar while the redaqueous filtrate contains the prehnitenesulfonic acid.
The filtrate is treated with excess of powdered calcium carbonate, andthe precipitated calcium sulfate is filtered and thoroughly washed withwater. The calcium prehnitenesulfonate in the combined filtrate andwashings is converted to the corresponding sodium salt by addition of asaturated aqueous sodium carbonate solution as long as any precipitateforms. The precipitated calcium carbonate is filtered and washed withwater. The filtrate and washings are evaporated to dryness on thesteam ba th . The residue of sodium prehnitenesulfonate weighs 40 g.
Since the prehnitenesulfonic acid undergoes extensive decompositionwhen heated with sulfuric acid, the sodium salt is advantageously hydro-lyzed to the hydrocarbon by a "flash" method . In a steam-distillationflask, provided with openings for a thermometer and dropping funnel, isplaced about 100 cc. of water . Superheated steam is passed into the
flask, and concentrated sulfuric acid is then added slowly from the drop-ping funnel until the temperature of the diluted acid reaches 150-160°. At this point a saturated aqueous solution of 40 g. of sodiumprehnitenesulfonate or a thin paste of solid and water is run into theflask at such a rate t ha t the temperature of the mixture remains at 140-150°. Careful control of this temperatu re is essential. Hydrolysistakes place rapidly, and a pale yellow oil separates from the distillate.The crude oil weighs 20 g. (88% ). Upon distillation, over 90% boils
at 97-98°/24-25 mm .; m.p. —7.4°. Highly purified prehnitene melts at- 6 . 4 ° .
From a Mixture of the 1,2,4,5- and 1,2,3,5-Tetramethylbenzenes(Durene and Isodurene). A mixture of durene and isodurene, b.p.82-84°/15 mm. can be obtained by fractionation of the hydrocarbonsproduced by the reaction of methyl chloride and aluminum chloridewith the mixed xylenes (see ref. 3 for details). A mixture of 100 g. ofthis fraction, 67 cc. of concentrated sulfuric acid and 33 cc. of 60%
fuming sulfuric acid is shaken (in a 500-cc. glass-stoppered Erlenmeyerflask) for about five minutes. The resulting solution is heated to 80°for a period of nine hours. The black, nearly solid reaction mixture isthen broken up and poured over 500 g. of crushed ice. After filtration ofthe insoluble materia l (18 g.) the solution is cooled to + 1 0 ° and the sul-
fonic acid is precipitated by the addition of 250 cc. of concentrated sul-furic acid. After cooling, th e dark-co lored sulfonic acid, m .p. 98 -10 0°,is collected on a filter an d presse d dry . I t is dissolv ed in 200 cc. of w armwater and is then hydrolyzed by the "f lash" method descr ibed above.
The organic layer from the steam distillation weighs 65 g. and on carefulfractionation yields 41.4 g. (41.4%) of prehnitene boiling at 94-96.4°/25mm. and freezing at —7.2°.
1,2,3,4-Tetraethylbenzene6
From the ethylation of benzene,36 a fraction can be obtained (b.p.110-113°/10-11.5 mm.) which contains 1,2,4,5- and 1,2,3,5-tetraethyl-
benzen es. A m ixt ure of 25 g. of th is fraction an d 100 g. of co nce ntra tedsulfuric acid is stirred a t 100° for fifteen min ut es . T he emulsion wh ichfirst forms dark ens in color an d th e hy dro carb on s dissolve. T he cooled
solution is poured onto 100 g. of ice, whereupon the tan-colored sulfonicacid crystallizes. T he prod uc t is purif ied by crystallization from am ixtu re of benzene an d petro leum ethe r (b.p. 60 -68 °). I t forms w hitecrystals , m.p. 118-120°, which contain one molecule of water of crystal-
lization. Th e yield is 34 -35 g. (90.7-92 .3% ).
A m ixtu re of 84 g. of th e sulfonic acid an d 300 cc. of 5 0 % sulfuric acid
is heated. Steam is passed throu gh the solution; when the te m peratu rereaches 130° (thermometer in l iquid), hydrolysis begins and at 140-150°is rapid. T h e oil in th e distil late is removed a nd fractionated throu gh a
column of the Fen ske typ e pack ed w ith glass helices. T he produ ct dis-
ti ls at 119-120°/ll mm. and is pure 1,2,3,4-tetraethylbenzene.36 T h eyield is 50 g. (90 .7% ).
3-H alopseudocumenes
Chlor inat ion or brominat ion of pseudocumene 3 7 (1,2,4-tr imethyl-benzene) produ ces m ixture s of th e 3 - an d 5-halopseudocum enes. T he5-halopseudocumenes hav e re la t ively high m elt ing points and can belargely remo ved by cooling an d filtering th e reaction prod uct . T hefiltrate con sists largely of th e 3-h alop seu do cum ene (see ref. 5, p . 8, for
deta i ls ) .
3-Chloropseud ocum ene. In a 250-cc. glass-s toppered Erlenm eyer
flask, 30 g. of 5-chloropse udocu men e (or 30 g. of th e m ix tu re of the 3- and
5-chloropseudocumenes) is dissolved in 100 cc. of 20% fuming sulfuricacid by vigorous shak ing of th e m ixture. T he solution is hea ted t o
3 6 S m i t h a n d G u s s , J. Am. Chem. Soc, 6 2 , 2 6 2 5 ( 1 9 4 0 ) .3 6 S m i t h a n d G u s s , J. Am. Chem. Soc, 6 2 , 2 6 3 0 ( 1 9 4 0 ) .37
Smith and Cass, J. Am. Chem. Soc, 54, 1603 (1932); Smith and Lund, ibid., 52,4144 (1930).
65-70° for four hours and is then poured over 150 g. of crushed ice. Theresulting mixture is cooled in a salt-ice bath until crystallization of thesulfonic acid of 3-chloropseudocumene is com plete. The cold mixtureis then filtered and th e cake is pressed dry . The sulfonic acid is dissolved
in 75-125 cc. of water, and the insoluble tar (4.5 g.) is filtered and dis-carded. The cold solution is treate d with an excess of 20% sodiumhydroxide, and the precipitate of sodium sulfonate is collected by filtra-tion. The filtrate is concentrated to one-third volume, chilled, andfiltered to yield a second crop. The tota l yield of sodium 3-chloro-pseudocumenesulfonate, after drying a t 110°, is 35.4 g. (71% ).
The sodium sa lt is dissolved in 250 cc. of 50% sulfuric acid, in a 500-cc.flask arranged for steam d istillation. The flask is heated in an oil ba th
un til the interna l tem pera ture is 135-155°. Steam is passed into theliquid un til the distillate is homogeneous. The organic layer of thedistillate is separated, dried over a little calcium chloride, and distilledunder diminished pressure . Th e pure 3-chloropseudocumene, boiling at127°/61 mm., weighs 17.2 g. (79%, based on the sulfonate).
3-Bromopseudocumene. By vigorous shaking, 19.9 g. of crude 5-bromopseudocumene is dissolved in 120 g. of 20% fuming sulfuric acidwhich is maintained a t 70°. After solution is complete the reactionmixture is trea ted as described in the above procedure. The sodium saltof 3-bromopseudocumenesulfonic acid, which weighs 27.1 g. (90% ), ishydrolyzed by steam distillation from 50% sulfuric acid maintained at175°.13 The 3-bromopseudocumene boils at 85.5-86.5°/5 mm. andweighs 14.5 g. (80%, based on the sulfonate).