Heterocycles from heterocycles. 1,3-Diaryl-4,5-imidazolidinediones from 1,3,5-triarylhexahydro-1,3,5-triazines and oxalyl chloride

oo4o-4020193 s6oo+oo 1993 Perganon Press Ltd

Heterocycles from Heterocycles. 1,3-Diaryi4,5-irmdazolidinediones from 1,3,5-Tnarylhexahydro-1,3,54nazmes and Oxalyl Chlonde

Giancarlo Verardo, Angelo G. Giumanir& * Faust0 Gorass~~, Manlena Tolazv and

Pa010 Strazzohni

Department of Chemical Sc~cncer and Technologies, Unlverslty of

Udine. Via de1 Cotontficio 108, 33100 Udine. Italy

(Recewed m UK 26 Juiy 1993, accepted 10 September 1993)

Abstract: 1.3-Dlaryl-4,5-imidazclidinedicnes (IS) are easily

synthesized from 1,3,5-trierylhexehydro-1,3,5-triez~nes CL) and

CXClYl chloride <S) L” a reection not likely to involve the

zwltterionlc lntermediste (3) of the N-methylenearyIam~ne dlmer,

but VIewing the sequential pack up of two un1tt of the mcncmer

(1) by oxsly chloride (2) The essential role of ethyl alcohol

added to the reaction mixture 1s recognized Reactlo” conditions

have been cptlmized and scme ten ~midazoi~d~nediones (6-1 were

prepared I” good to excellent yields Geometric parameters of S

were obtolned by X-ray dlffrsctlon snslys~r all the nuclei are

found almost in one plane except for a small twist of the phcnyl

rings about the C-N bond

IRTRODUCTIOR

1,3,5-Tnarylhexahydro-1,3,5-triazines Q), the cychc tnmersi of the elusive N-me-

thylenearylaminesz Q) may be induced to react as 2 or the dxmenc zwittenon3 2 by

interaction with substrates with strongly polarized bonds.4 It is not clear in most cases

whether depolymerization occurs first, either thermally or induced by an added catalyst

or whether the substrate attacks 1 dxectly, causing a reaction sequence all the way

down to the observed products.

The employment of & to yield heterocycks has been recorded

the patented preparation of 1,3-drary1-1,3-(5-thiadiarine)-6-thlones

1 with carbon disulphide.5 On a different front, the rea&on of

documented,6 but not that with bdun&onal compounds of the type

10609

m the literature for

from the rea&on

1 with G-X(=O)Hal

G(X[=OlHal),.

of

is

10610 G VERARDO et al

Ar

I N

(‘1

\-6

4 * Ar-N=Ct$ /

Ar -N

ArHNVNNAr 2 \

CH,Hal

1 t!

Most of the ptinent work has appeared m the form of patents and 4 were usually

poorly characterized rntermedrates of processes leatig to other final products.

RESULTS AND DISCUSSION

The reactin between 1 and oxalyl chlmde 3 has now been mvestigated and

developed into a very convenient way of obtimg N,N’-draryl-4,5rmrdazohdinetines (4)

(Table l), a class of compounds so far unreported.

Table 1. L,3-Dlaryl-4.5-lmldazolldlnedlones (5) Prepared.

Trlazlne (Ar)

Product Yield’ Crystaillzatlon solvent (W g,

A!? (3-He-C,H,)

Ic (4-He-C,H,)

(4-?&H,)

If (L-F-CsH,)

(3-F%,H,)

G! (4-F-C,H,)

li (3-Cl-C,H,)

(4&H,)

llz (3-Br-CGH4)

83

80

90

55

60

90

80

93

90

67

269

217

>290

261

182

251b

266

247

267

268

ethyl

ethyl

ethyl

ethyl

ethyl

ethyl

acetate

acetate/dlchloromethane

acetate/dlchloromethane

acetate

acetate/trlchloromethane

acetate/dlchloromethane

ethyle acetate/dlchlorcmethane

ethyl acetate/dlchloromethane

ethyl acetate/dlchloromethane

ethyl acetate/dlchloromethane

‘Yield of recrystallized product bDecomposltlon temperature.

Hctcrocycles from hetemcycles 10611

The parent molecule itself, an isomer of the well known compound hydantoln (irm-

dasohdine-2,3-lone), is not known and only two 1,3-denvataves, namely 1,3-dimethyl-

-4.5~inudazokdinedlone7 and 1,3-drbenzoyl-4,5_irdasolidinedione,8 were described to

date.

Our synthetic procedure calls for the portionwise addition of 5 to an equlmolecular

amount of 1 in dlethyl ether at OV, immediately followed by a necessary termnation of

the reaction with ethanol. Two different courses may be envisaged to rationahze the

outcome. If the monomer (2), either in fast eqmhbnum with the tnmer or ongmating

from the induced decomposition of 1 is to be involved, pathway A should be at work

(Scheme 1) On the other hand, the drmenc zwittenonic species 1 may be active, thus

facilitating pathway B.

Headspace analysis ruled out the presence of CH,Cl,, but both diethoxymethane (1)

and ethoxymethylchlonde (i3) were detected, the former coming from ethanolysrs of the

latter in solution. This observation pointed to an essential role of the alcohol in

carrying out the reaction and ruled out the rnitzal intervention of 3 In fact, it is

conceivable that 2, d formed, would rapidly evolve to l0, in turn bound to undergo fast

reaction with the naked chlonde ion to produce _6: but, besrde the absence of CH,Cl,,

ethanol was found to play an essential role

The monomer 2 is, on the other hand, expected to react promptly with 5 to form

the intermediate Q and furthermore to give lJ, which, apparently unable to undergo

nng closure to lJ, survives until ethanol rs added and the key intermediate j.J 19

generated. This, in turn, either undergoes ring closure to 6a_k or further competitive

ethanolysis to the constantly observed by-product oxalamrdes 14a_k.

Indirect support for this hypothesis was offered by the reaction between 1,3,5-

-tn(4-anlsyl)hexahydro-1,3,5-tnazin e (le) and oxalyl chlonde 5 in its devious behaviour:

enhanced stabilization of 2 sparked route B,

oxaiyl-di(4-anlsylarno)methane (15), detected

rmxture, together with another side product,

-hydroxymethyl-di(4-anlsylamino)methane (l6)

which did not lead to Se, but to N-ethoxy-

by direct inlet MS analysis on the reaction

tentatively identified as N-chloromethyl-N’-

(~-AII-N=CH~)~

le

I I

0 N-MI +

I I p-An/NvN\p-An

It IS possible

attack of 3

p-in p-in I!?

15

that for ahphatic 1 to react, decomposition must be induced by direct

10612 G Wmumoetal

Scheme 1

Ar

Route A I N

f‘l

Ar HNvNIAr

la-k \ +5

Ar

b 1 8

b Ar-N&Ii2 + f

2 Ar ON% 2

/- 3 +5

0 0

73

ArMN \ c1 cH,Cl

Ar -N

+ IztOH

0 / -EtOCH,CI

0 +EtOH *

Ar NNACH cH~cl

-EtOCH,Cl

I 2 Cl

Ar f

Cc1,Cl

lo

Ar xNAN'--CH Cl

+f

2

Cl 4 x

0

0 2

N-Ar

14

-CH,Ci,

-

6a-k

0 0

PC + 9

ci -Cl Route B

s

Heterocycles from hctcxocycles 10613

No systematrc work was carrzed out to optrnuxe the yield of 6 in the reactins

between &g and _5, but a few useful expenmental observatrons were made m adlustrnq

the condltrons for the reactin of la, which were essentilly apphed to the other

substrates. Good solvents for 1, hke droxane and chloroform, were detrrmental; the best

results were obtarned when 1 was added to 5. The use of anhydrous solvents and

absolute ethanol gave better yields

An alternate route to 6 was attempted by reactin of amrnal N,N’-drphenylamr-

nomethane (17a) with 2 under a variety of expenmental conditrons, but m all instances

the oxamrde & was by far the matn product, accompanred by much lower yrelds of &

(Scheme 2).

Schsme 2

(ArW ,CH,

m. Ar = Ph

m: Ar = 4-N02-C6H4

+EtOH

/

-a

0 0 0

7-f ArRNvN-Ar

& (yie. 8-16x)

a (traces)

0 0

Y-3 Ar-N N-Ar

‘H H’

&& (yie. 80-90X;)

141 (yic. quantitative)

10614 G VERARDO et al

In the above scheme the cause for the low yields is to be found m the productin

of acltity which tres up the mtermetite @) ln an unreactive form. In fact, we found

that practically all of & was formed m this reactin before alcohol ad&tin.

All the products 6a_k showed msolubllity u most common solvents and excepbnally

high melting pcunts which are also lndlcatlve of thw extraordinary thermal stablhty

The steadfastness of the crystal structure points to strong intermolecular polar

Interactins. m fact, the isolated molecules present two tides of opposite polanties

Ar

In view of the novelty of the molecules and these propeties of the sohds, we

undertook an X-ray crystal structure deternunatin of &.

Worth notrcrng are the following features of & (Figure 1, mean values are quoted)

Ph

0 0

126 H 117 107

,-c_"+::zyph

20 23

Figure 1. Average values of molecular dimensions and bond angles of &.

The shght twisting m the same dvectron of the phenyl groups, placzng them almost on

the same plane with respect to the heterocychc ring, would not lnhlblt some conlugatin

as indicated by the Ar-N distance (141 A), above that (1 39 A) found for a perfectly

planar arylarmne,g and definitively larger than for less conventinal armdes of this type

(1 36 A) 10 The angle at the saturated carbon EI smaller than that expected for a pure

sp3 hybnduatin, leaving a larger s contnbutlon for the bonds with the hydrogens A

comparison with open ch;lm oxarmdes 11.12 has to take into account that these systems

show a torsional angle of about 90° around the OC-CO anus which 15 extraordmanly

elongated (ca 154 A), the result of the hkely charged oxygen, carbon and nitrogen

repulsions (Figure 2)

Hetemcycles from heterocycles 10615

The only possible response of & is the widening

value of 126’ agatnst 117O for the open charn case which

distance As a consequence the nng NCC bond angles are

107’ from 117’ observed in the open chsun cases.

6’ N x OS9

6’ N 06’ II

of the OCC bond angles to a

would cause a quite short O-O

squeezed to a meager value of

III

Figure 2. Conformations of open oxamides

Planar nng enclosed oxanudes, hke 6+ can only exist in the high energy form II,

where hkely charges are the closest. Different stenc requirements UI open oxanndes may

allow for perpendicular geometnes III more or less approaching the ideal lowest

electrostatic energy configuration I Entropic factors play a role in the equilibrium

positions.

Literature structural data for the heterocycbc system present m & are few and

some were collected for more substituted denvatives (Table 2) Noteworthy is the

contraction by cd. 0 04 A of the C,-C, bond (Figure 3 and Table-a) found 19 @ in

comparison with all other s~~1a.r cases 13 Surnlarly the CC0 angle is strongly widened

by the oxygen repulsion to a value of 126“ in all these systems as well as in &. The

N-CO bond of &a is some 0 02 A larger than in formamide and ca. 0.04 A larger than

that of the 1,2,3-tnsubstituted 1,3-rnndazohdme-4.5dlones mvestigated,lz differences

which are too small to evince any special conclusion. The practrcally mvanant C=O

titance, actually coincident with that of formaldehyde, * 6 though, point to an imperfect

NC0 amide conlugation. Oddly enough, the N-atom seems to be more conlugated with the

ring. But, as we shall see, the 1 H NMR data will introduce a contradiction here

The nature of the obtained products 6 was confirmed by elemental analysis and

their MS and spectroscopic properties. Mass spectra of 6a_k exhibited parent ions of

medium intensity with the exception of the low intensity observed for the

chloro-derivatives &i and 6-J. The common features of the ion decompositions are shown

in scheme 3, which specifically refers to the ion derived from @; the composition of the

ion at m/z 105 was secured by high resolution mass spectrometry whch ruled out the

presence of PhCOe

The peak formally ascribed to the N-methylenearylamine radical cation was

consistently the base peak in all spectra of 6a_k. Of interest is the formal loss of an

azindone (133 mu) from the parent ion of Q to produce what is likely the molecular ion

of an aryhsocyanate molecule (Table 3)

N-C

N

-exo

C

N-C

O C

=O

CC

-CO

N. C

.N

C.

cc

N

N.

.C

C

(A)

(A)

(A)

(A)

(“)

(“)

(“)

co)

(“)

(“)

Ref

0 0

!?!I

9-t

1.4(

i 1.4

1

IPh’

N”N

NPh

’ 3!

3 1-

44

Me

” N

N

NM

e

c Ph

- \

S 14

3 14

5

10

0

W

Me .

-N

N’t

le

x E

)n

Ph

175

145

0 0

Y-t

(M

e) ,N

N

W)

z

1.46

0 0

Y-4

(i-

Pr)

jlN

N(i

-Pr)

, 11

1.38

L

.21

1.49

10

3 0

.

1.38

1.

19

1 52

10

6.2

111

6

1.35

12

2 1.

52

102.

8 11

2 6

105

8

113.

2 10

6 0

118

0

117.

0

127

126

127

5 12

7 3

0 13

ab

124

118

0

0 th

is

wor

k=

0 13

b

13b

‘Ave

rage

va

lues

be

twee

n 6a

’ an

d x

bRep

orte

d va

lues

re

fer

to

S=C

-N(P

h)-C

O-C

O

syst

em

Hctc~les from heterocycles 10617

6a, m/z 252

m/z 224

I - Ph-N=C=O

Ph-N- -CH21e

m/z 105

- Ph-N&l,

*

\

0

P m/z 133

NxPh

Ph-N=C- -01.

m/z 119

/

-NC0

v? -Cd+, * we m/z 77 m/z 51

9 Ph-N=CH

m/z 104

Comparison of this spectral pattern with the fragmentataon of the closely related

open oxarmde N,N’-dimethyl-N,N’-lphenyloxarmde (19) indicated that the trigger for all

of the observed fragments is the cleavage of the sigma bond between the two carbonyls

(Scheme 4)

Table 3 Propertles of 1,3-Dlaryl-4,5-imldazolidinediones (5)

Compound

IRa (cm-l)

6a

1725~8, 159Os, 1490vs,

14556, 1405V8, 12908,

1275s, 745~s. 680s

!&

1715~8, 14858, 1400~8,

1290s. 118Om, 79010,

685111

Sr

1735~8, 16108, 151Os,

14OOs, 13008, 128Os,

815s

2930m, 1725vs, 15lOm,

1430m, 13858, 1300m,

1270m. 830s

1740~8, 1500vs, 14606,

1410~9, 13108, 12708,

1230s. 8108, 755~s

1730~8, 16108, 15908,

149ove, 14608, 14208,

14008, 11908, 770s

173OV8, 171OV8, 15006,

14009, 1240~11, 116Om,

1090m. 830s

1H NMRb (b , ppm; J, Hz)

5.66(s, 2H), 7.27-7.35(1n, 2H).

7.47-7.57(m, 4H), 7.93(d, 4H,

Js8.00)

2.43(s, 6H), 5.46(s, 2H), 7.10-

7.63(88, a)

2.36(s, 6H), 5.40(s, 2H), 7.27

(d, 4H Jx9.00). 7.65(d, 4H,

Jl9.00)

1.33(s, 18H). 5.44(s, ZH), 7.48

(d, 4H, Jn9.00). 7.68(d, 4H,

J=9.00)

5.48(s, 2H). 7.19-7.45(m, 6H),

7.70-7.80(m,

2H)

5.49(s, 2H), 6.99-7.12(m, 2H),

7.40-7.55(m, 4X), 7.66-7,75(m,

2H)

5.47(s, 2H), 7.15-7.30(m, 4H),

7.70-7.85(m, 4H)

MS

C

(m/z

, rel%)

252(M+, 23), 119(6), 106(7),

105(100), 77(38), 64(4), 51

(16)

280(M+, 69). 133(13), 120

(22), 119(100), 118(57), 104

(6), 91(60), 77(8), 65(21),

53(6)

280(M+, 60), 133(21), 120

(22), 119(100), 104(7), 91

(70)) 77(8), 65(23), 51(8)

364(M+, 32). 349(42), 175

(9), 167(20), 161(47), 160

(80), 146(100), 132(18),

118(19), 106(6), 91(7), 77

(7), 44(21)

288(M', 59), 137(11), 124

(17), 123(100), 122(85),

109(7), 95(38), 75(18),

57(5)

288(M+, 31), 137(6), 123

(loo), 122(45), 96(5), 95

(29), 75(10)

288(M+, 26), 137(g), 123

(loo), 122(47), 95(29), 75

(9), 57(3)

(continued)

Table 3. (continuation)

Compound

IRa (cm-l)

IH NMRb (6

t ppm;

J,

Hz)

si

1725~8, 1590s. 14708,

141Os, 1270m, 1105m.

5.49(s, 2H), 7.25-7.48(m, 4H),

870m, 775s. 670m

7.70-7.81(m, 4H)

MS

c (m/z, rel%)

324(M+, l), 322(M+, 8), 320

14), 155(l)

153(5)

%32),

139(1OOj 138(31j

111(21), 77(7), 75(12), Si

(7)

5i

1735vs, 14958, 1395vs,

1270m, 1090m, 825vs,

5.48(s, 2H), 7.47(d, 4H. J=9.00)

810111

7.72-7.80(d, 4H. J=9.00)

324(H+, 3), 322(M+, 17), 320

:!;i42), 139(1OOj

138(45j,

27). 155(5)

153(15)

P

111(27), 77(6), 7&15)

6k

1725vs, 15808, 1470~8,

1415s, 14oovs, 13056,

5.50(s, 2H), 7.32-7.47(m, 4H),

7.75-7.93(m, 4H)(H+, lo), 199(5), 1::;:::' 9), 410(M+, 21), 408

12608, 1090s, 760vs,

665s

185(99), 183(100),157(19),

W;l9),

90(13), 77(26), 51

ft

aSpectra were recorded in KBr. CSpectra recorded in CDC13 eolution using TM

as Internal standard.

Upectra recorded via direct inlet.

Heterocycles from heterocycles 10621

Table 4. Most Significant Gecmetrlcal Characteristics of the Refined Molecules

O(1) - C(1) O(2) - C(2) O(3) - C(16) O(4) - C(17) N(1) - C(1) N(1) - C(4) N(1) - C(3) N(2) - C(2) N(2) - C(3)

C(4) - N(1) - C(3) C(1) - N(1) - C(3) C(1) - N(1) - C(4) C(3) - N(2) - C(10) C(2) - N(2) - C(10) C(2) - N(2) - C(3) C(18) - N(3) - C(19) C(l6) - N(3) - C(19) C(16) - N(3) - C(18) C(18) - N(4) - C(17) C(25) - N(4) - C(17) C(25) - N(4) - C(18) O(3) - C(16) - N(3) N(3) - C(18) - C(17) O(3) - C(16) - C(17) N(4) - C(25) - C(30) N(4) - C(26) - C(26) O(1) - C(l)- - N(i) N(1) - C(1) - C(2) O(l) - C(1) - C(2) N(2) - C(2) - C(1)

Bond Lengths (A)

1.203(7) N(2) - C(10) 1 396(g) 1.216(7) N(3) - C(16) 1.375(8) 1 222(7) N(3) - C(18) 1 458(7) 1 205(8) N(3) - C(19) 1 409(8) 1.377(8) N(4) - C(25) 1 424(8) 1 422(8) N(4) - C(18) 1 452(7) l-458(7) N(4) - C(17) 1 380(8) 1.379(8) C(16) - C(17) 1.502(9) 1.467(8) C(1) - C(2) 1.485(8)

Bond Angles (“)

119.8(S) 111 7(S) 128.6(S) 120.8(S) 127.1(S) 112.1(S) 121.1(S) 127.3(S) 111.6(S) 112.9(S) 126.8(S) 120.3(S) 127.5(S) 107.3(S) 125.2(6) 119.9(6) 120.4(6) 126.5(6) 107 4(S) 126.2(6) 106.2(S)

O(2) - C(2) - C(1) 126.3(6) O(2) - C(2) - N(2) 127 S(5) N(3) - C(l8) - N(4) 102 7(4) N(4) - C(17) - C(16) 105.2(S) O(4) - C(17) - C(16) 126.2(8) O(4) - C(17) - N(4) 128.6(6) N(l) - C(4) - C(9) 121.3(6) N(1) - C(4) - C(5) 117.9(5) N(l) - C(3) - N(2) 102.6(4) N(3) - C(19) - C(24) 119.1(S) N(3) - C(19) - C(20) 121.1(S) N(2) - C(10) - C(15) 120.1(6) N(2) - C(10) - C(11) 122.2(6)

Selected Torsion Angles (")

C(1) - N(1) - C(4) - C(5) 21.0(9) C(3) - N(2) - C(10) - C(15) -24.0(9) C(l8) - N(3) - C(19) - C(24) 19.3(8) C(l8) - N(4) - C(25) - C(30) -22.0(8)

The molecules are stacked m the cryatala like parallel columns; the columns,

contactig each other by a shght overlapping of the me ta-Poe&on of a phenyl nng and

the pars-posrtron of a phenyl nng of the adjacent column, are made up of piles of

alternatig molecules &ia/ and K, where the heterocychc nng of one 19 almoat parallel

to the second phenyl nng of the other. The relative posrtrons of theae molecular

sectxnns are mdrcatrve of some electrostatrc interactin between the negatively charged

oxygens and the electron lmpovenshed phenyl nng.

1 H NMR spectra of 6 III CDCl, showed the expected pattern with a sharp singlet for

the methylene group locahzed at 6 values between 5.40 and 5.66 ppm from the standard

tetramethylsllane (Table 3). Whereas the sryl protona appeared I.II three well separated

regrons, para-substituted 5 showed two types of resonances: all of them were located at

between 7.10 and 8.00 ppm (Table 3), posrtions definitively at lower field than for

aruhne,l7 N,N_drmethylanrhne18 and N-methyl-N-formylanAnel e and the correapondlng

denvatlvea These data are an mdlcatron of a somewhat strong electron depletxon away

G VERARLW et al

or __

Heterocycles from hetemycles

from the aromatic nng into the amrde fun&on.

An lndrrect confrrmatron of the poslbve polandion of thr aromatic rugs came

10623

from

the packing of the molecules &’ and +&’ m the crystals with the oxygens of one lust

on top of the carbons of phenyl ring, two such Interaction bag active for every single

molecule with shghtly different posltiolllngs This appears to be due to the strong polar

intermolecular attractron playmg an important part m holtig the molecules together 111

the crystals.

A full view of the crystal packrng rs offered by figure 4a and 4b, where a side by

side columnar paclung of the stacked molecules, with any two columns barely “touchmg”

with the free phenyls, 15 evidenced. The closest distance between the phenyl nng of

one molecule 6a’ and the related heterocyclic nng of the other (6a”) was found to be -

3 6 A, well beyond any charge transfer Interactin.

Figure 4a. General view of crystal packing of &.

The huge hypsochrormc shift of the carbonyl stretching frequency (1725 cm-l) of

& and the range 1715-1740 cm- 1 for 6b_k (Table 3) compared with the pratrcally

cmncrdental values for N-methylacetamrde (20) (1656 cm-l) and N,N’-lmethyl-N,N’-

-drphenyloxamzde Clg) (1650-1665, doublet) rs due to the combined effect of enhanced

carbonyl character and, therefore lesser delocalizatin of electrons from the nrtrogen

atom, and nng strm 20

Thrs may be simply a descreenlng effect by the nitrogen lone pans or an

overwhelming field electron wlthdrawlng o-effect. Thrs effect overlaps with some much

less effectrve u-electron transfer from the nrtrogen to the nng.

10624 G VMARDO~~ al

b

. ,

. 9

Heterocycles from heterocycles 10625

EXPERIMENTAL SECTION

llaterials. Oxalyl chlonde (3) and pnmary aromatic amines (2) were commercially

available (Aldrich, Milano, Italy). they were conveniently purified before use and used to

prepare the 1,3,5-triarylhexahydro-1,3,5-tnananes (1) accord.ing to the amine

paraformaldehyde method.1’ Diarylarmnomethanes (17) were prepared according to a

described procedure.1 b Dry solvents were obtained following standard procedures.2 *

TLC plates (neutral alumina on alumuuum plates) were obtied from Merk, Italy.

Equ ~pment. lhgh pressure liquid chromatography analyses were performed with a

Waters Milhpore instrument, equipped with an inverse phase C,, Bondapak column

(lengh 30 cm, 1.d 3.9 mm) and a &ted wavelength (240 nm) uv detector. The system

uses two independent pumps and a processing unit enabling eluent composition control.

Water-acetonitrile mixtures were found su&able for our analysea operating at a flow rate

of cd. 1 ml/min.

Infrared spectra were recorded with a Jasco Mod. DS-702G spectrophotometer by

the KBr pellet technique

Electron impact (70 eV) mass spectra were obtsnned from a Finnegan MAT 1020 with

automatic continuous data recording. During direct inlet vaponxataon of the whole

sample rnto the ion source, the full recordmg was carefully inspected ~II order to detect

any side product and check sample punty. Headspace analyses were performed by

injecting the vapours over the solutions kept in inert atmosphere into a

gaschromatograph pnor to the electron impact with continuous ms mo&onng of the

eluate for the detection of gaseous products. The most intense peaks with therr relative

intensrty (%) are reported for each product.

ill NMR data were secured from a Bruker Mod. AC-F 200 spectrometer using

tetramethylsllane as internal standard. The high insolubility of 6a_k presented a

practical difficulty in recording of the 1 3 C NMR spectra.

Elemental analyses were obtained with a Carlo Frba Mod. 1106 elemental analyzer

for all isolated compounds and were satifactory.

X-ray diffraction analyses were obtatned fmm a crystal of @ cd. 0.2 x 0.2 x 0.5

mm that was mounted on a CAD4 single crystal drffractometer with graphite

monochmmatrzed MO Ka radiation, 25 reflections with 9 in range 10 s s s 16’ used for

measuring lattace constants (Table 5).

For data collection 3 5 s 5 26O (-8 J; h s 8, 0 5 k s 32, 0 s 1 5 14). o - 29

scans, o-scan width (0.80 + 0.35 tana); intensities of three reflections monitored every

2h of exposure time showed no slgruficant variation 4537 unique reflections were

collected; 1095 with I 2 3s(I). The structure was solved with MULTAN 8022 m default

setting and refined with SEELX 76.2 3 At convergence R = 0.059 for 1095 observed data.

Hydrogen atoms were located at calculated positions. Atomrc scattering factors were

10626 G. VERARDO et al

taken from Cromer & Mann.24

Table 5. Crystal Data and Experimental Details

Formula

M W.

Space Group

a/A

b/A

c/A

B/”

v/A3

z

D ..,,fg cmm3

~(Mo Ka)A

p/cm-’

F(OO0)

C15H1202N2

252.3

p2,/c

7.259(3)

27.466(2)

12.525(2)

99.23(3)

2464.6(6)

8

1.36

0.71069

0.86

1024

Cryst. size /mn

0 range /”

h range

k range

1 range

scan mode

Measd. Reflections

Solution of Structure

Ref lnement

Final R factor

Final R, factor

Room temperature

0.2 x 0 2 x 0.5

3.26

-8.8

0.32

0 14

0 - 28

2402

MuLTAN80

SHELX76

0.058

0.059

General Procedure for the Preparation of ga_k. A suspension of the

appropriate 1,3,5-trxarylhexahydro-1,3,5-triazine (l, 10 mmol) III anhydrous ether (ca. 50

ml) was slowly added to neat oxalyl chlonde (5, 30 mmol) kept at O°C under efficient

stxnng in an atmosphere of Argon. About 10 nunutes after the end of the addltin

anhydrous ethanol (30 ml) EI added at O°C slowly, wUe hydrogen chltide 19 evolved

and a sohd separates. The preupltatin is completed at rcom temperature by addition of

ether. The preupltate, separated by filtratin, IZI recrystallized from a sortable solvent.

Headspace analyw was performed m the case of the syntheu of &, by sampling the

atmosphere over the reactron mxture after the additin of ethanol and analysrng It by

GC-MS. The procedure with inverted order of addition of the ~~taal reagents or not

using ethanol to end the process yielded products 6 m substantially lower yields

The above procedure IS IJI part the result of the study of the variatxon of a

number of reactin parameters, when & was used as a substrate (Table 6)

Reaction between N,N’-Diarylaminomethane (l7) and 2. The react&on was

carned out according to the optamal procedure described for the preparation of 6. When

G was the substrate only ca 10% yield of & was obhned; N,N’-dlphenyloxalanude

(&) bexng the other observed product.

Hetcrocyclesfromhetcxocycles 10627

A slrmlar result was obtained when N,N'-&(4-r&.rophenyl)armnomethane (E) was

used, but the cyclic product &I was not present at all.

Table 6

mol z/m01 &

6

3

12

Reaction

time

10 min

10 min

10 min

Quenching

reagent’

Et,Ob

EtDH

EtQH

Reaction

solvent=

-_

__

-_

Yield

(W

5oc

40o

2Sd

1 10 min EtCtl Et,0 45

3 10 min EtDH Et,0 75=

6 10 min EtDH Et,0 58

3 10 min EtDH Dioxane mixe

3 10 mln EtDH CHCl s mixP

6 16 hours EtCH Hexane 65

‘Anhydrous materials were used. bComnercial diethyl ether was used as received ‘DI-MS analysis of mother liquors obtained after filtration of & showed the presence of N,N’-diphenyloxamlde [l&q MS (m/z): 240 (M@, 43), 121(30), 120(31). 105(12), 93(100), 92(22). 77(58)] dDI-MS analysis of mother liquours obtained after filtration of & showed the presence of 14a and bis(N-ethoxyoxalyl-N-phcnyl)diaminomethane [MS (m/e)* 398(<1), 325(18). 252(4). 206(28), 176(100). 134(6). 120(10), 106(83). 93(34). 77(38)]. ‘A very complex mixture was obtained which was not worked up.

Acknowledgements This work was supported in part by grants to AGG (CNR

86.01649.03, CNR 8903765.03, MPI 1987-1989 40% and 60%) and to GV (MPI 1987 60% and

1990 40%). The autors are grateful to Mr. P. Padovam for expert instrumental

mntenance

REFERENCES

(a) Glumamm, A G , Verardo, G, Randacuo, L., Bresciaru-Pahor, N.; Traldr, P. J

Prakt Chem 1985, 327, 739-748. (b) Glumanini, A G., Verardo, G.; Zangrando, E.;

Lasslam, L. J. Prakt Ches 1987, 329, 1087-1103.

(a) Drstefano, G, GUXM~.UU, A. G., Modelh, A.; Poggr, G. J Chea sot. Perkin

Trans 2 1985, 1623-1627. (b) GUXMIUIU, A. G., Verardo, G ; Poiana, M. J Prakt

Chem 1988, 330, 161-174.

Bertoluzza, A.: Bonora, S., Fim, G., J Prakt Chem 1990, 332, 595604.

(a) Verardo, G., Giumanml, A. G.; Cauci, S Actd Chim Hung. 1986, 122, 3-5. (b)

10628 G VERARDO et al

Giumanm A. G.; Verardo, G.; Caucz, S. J. Prakt. Chest 1387, 329, 417-432.

5 &=I=, R., Hllghetag, G.: Mti, A.; NeJedly, 0.; Schlegel, J. Arch. Pharm. 1960,

293, 957-967

6. (a) Ulnch, H.: Richter. R.; Wlutman, P J.; Saylgh, A. A. R. J Org Chem. 1974, 39,

2897-2899. (b) Gronowltz, S.; Lldert, 2. Syn thesl s 1979, 810-814. (c) Rathgeb, P.;

Tournayre, J. C., Vogel, C. GP 2306274/1972 (Chem Abs tr. 1973, 79, 115359x).

7 Hanafin, J; Ben-Ishsu, D. J. Heterocycl Chem. 1976, 13, 889-890.

8. Isaksson, R.: ~l~efors, T., Sandstrom, J J Chem Res. Synop 1981, 43.

9. Giumaxunl. A G., Randacclo, L.; Zangrando, E.; Ga, M. Ii. J. Prak t Chem. 1987,

329, 187-194.

10. Allen, F H.; Kennard, 0 ; Watson, D. G., Brammer, L.; Orpen, A. G., Taylor, R. J

Chest. Sot. Per&in Trans. 2 1987, 1-19.

11. Toda, F.; Tagarm, Y., Mak, T. C. W. Bull. Chem Sot. Jpn. 1977, 59, 1189-1194.

12. A&w.tdlala, G.; Voss, J. Chem. Ber 1977, 110, 1159-1166.

13. (a) Ried, W.; Broft, G. W.; Bats, J. W. Chem Ber. 1983, 116, 1547-1563. (b)

Gotthardt, V H.; Huss, 0. M. Acts Cryst. sect. 8 1982, 38, 875-879.

14. Lide, D. R. Tetrahedron 1962, 17, 229-233.

15. Fryer, R. I.; Earley, J. V.; Blount, J. F. J Org. Chem. 1977, 42, 2212-2219.

16 Kwa, G. II.; Curl, R. F. J. Chem, Phys. 1960, 32, 1592-1594.

17. Pouchert, J. C.; Campbell, J. R. The Aldrich Library of NtlR Spectra, Aldnch

Chesucal Company Inc., Milwaukee, Wisconsm, 1974, V, 37A.

18. Pouchert. J. C.; Campbell, J. R. The Aldrich Library of NMR Spectra, Al&h

Chesucal Company Inc., Milwaukee, Wisconlun, 1974, Y, 40C.

19. Pouchert, J. C.; Campbell, J. R. The Aldrfch Library of NUR Spectra, Aldrich

Chemical Company Inc., Milwaukee, Wisconsin, 1974, VI I, 65C.

20. Zahn, H.; Kunde, J. Chen. Ber. 1961, 94, 2470-2475.

21. Perrin, D. D.; Armarego, W. L. F.; Pernn, D. R. Purification of Laboratory

Chemicals, A. Wheaton 61 Co. Ltd., Exeter, Great Britaxn, 1980, 2nd Ed..

22. Main, P.; Fiske, S. J.; Hull, S. E.; Lesssnger, L.; Germ, G.; Declercq J. P.;

Woolfson, M. M. MIJLTANIO, A System of Computer Programs for the

Automatic Solution of Crystal Structures from X-ray Diffraction Data,

Unrversrty of York, England, 1980.

23. Sheldrick, G. M. SHELX-7s. Program for Crystal Structure Determination,

Unlvcrslty of Cambndge, England, 1976.

24. Cromer, D. T.; Mann, J. B. Acta Cryst. se c t B 1968, A24, 321-322.