Franz H. Kohnke and J. Fraser Stoddart- The evolution of molecular belts and collars

8/3/2019 Franz H. Kohnke and J. Fraser Stoddart- The evolution of molecular belts and collars

1/6

Pure & A p p / . Chem., Vol. 61, No. 9, pp. 1581-1586, 1989.Printed in Great Britain.0 1989 IUPAC

The evolution of molecular belts and collarsFranz H.Kohnkea and J. Fraser Stoddartba Dipartimento di Chimica Organica e Biologica dell'Universita di Messina,Contrada P apardo, Salita Sperone, 98100 Messina, Italy

Department of Chemistry, The University, Sheffield S3 7HF, United Kingdom

Abstract - The Diels-Alder reaction has been employed to synthesise ahexaepoxyoctacosahydro[l2]cyclacenederivative 6 in a highly efficient mannerfrom the syn-isomer 1 of 1,4:5,8-diepoxy-1,4,5,8-tetrahyd roanthracene and2,3,5,6-tetramethylene-7-oxabicyclo[2.2.l]heptane. Each time a newcyclohexene ring is formed in the macropolycyclisation process, which utilisestwo molar equivalents of both the bisdienophile 1 and the bisdiene 2, thecycloadditions are trebly diastereoselective. A sequence of reactions oncompound 6 involving (i) deoxygenation, (ii) dehydration, and (iii) partialhydrogenation have led to the isolation and characterisation of the D6hsymmetrical dodecahydro[l2]cyclacene erivative 10 containing six benzenerings. The synthetic strategy is new and the macropolycyclic compounds,which can b e prepared according to this kind of molecular 'LEGO, are novel.

I N T R O D U C T I O NMany macrocycles with molecular receptor properties have been m ade in recent years. With a fewnotable exceptions1 most of the synthetic molecular receptors of well-defined shapes with rigidcavities em erge after highly elaborate syntheses. In a search for new and efficient syntheticprocedures for the preparation of rigid collar-like molecules, which would offer a lot of flexibility interms of (i) structure, (ii) size, (iii) shape, (iv) electronic characteristics, (v) functionality, and (vi)properties, we have identified the Diels-Alder reaction2 as a route to a novel range ofmacropolycycles composed of laterally-fused six-membered rings. This well-known reaction, whichhas been much used, particularly in the synthesis of natural products, has a number of attractivefeatures. It (i) can involve cheap and readily-available starting materials, (ii) affords six-memberedrings by a mechanism in which two bonds are formed m ore or less simultaneously, (iii) usuallyproceeds in very high yields, and (iv) requires no reagents, a fact which assists greatly in productisolation. If necessary, it can be (i) performed in a range of solvents including water, (ii)carried outover a wide temperature range, (iii) promoted by very high pressures, and (iv) catalysed by Lewisacids. Moreover, the reaction mechanism is well-understood. I t exhibits (i)high regioselectivity, (ii)complete stereospecificity (cis-addition), and (iii) high stereoselectivity, e.g. endo and ex0configurational control with respect to bicyclic systems. It is quite rem arkable that, with all theseattributes, the Diels-Alder reaction has not been employed previously, at least to our knowledge, inthe synthesis of macropolycyclic compounds. The synthetic strategy is based on the idea(Scheme 1) that suitable bisdienes and bisdienophiles should be able to undergo repetitivecycloadd itions until the 'head bites the tail' of a grow ing polymer strip in a final intramolecular Diels-Alder reaction to afford a m acropolycycle. The challenge resides in identifying both a bisdienophileand a b isdiene with rigid structures, the 'correct' conformations, and the appropriatestereoelectronic Characteristics to demonstrate that the strategy is sound and does work.

1581


2/6

1582 F. H. KOHNKE AND J. F. STODDARTScheme 7

M O L E C U L A R 'LEGO'Recently, we described3 the m aking (Scheme 2) of the m olecular belt compound 6, a possibleprecursor4 of [12 ]beltenes starting from the known617 bisd ienoph ile1and bisdiene 2. There ismuch evidence in the literatures or the highly diastereoselec tive attacks (i) by dienes at the exo-faces of dienophilic units such as those p resent in1and (ii) by dienophiles at the endefaces ofdiene units such as those present in 2. The success of the D iels-Alder approach to makingmacropolycycles s a direct consequence of the treble diastereoselectivity expressed (Scheme 3)every time a new cyclohexene ring is formed in repetitive cycloadditions. For examp le, there areanother 9 diastereo isomeric forms of 4, yet the 2 :l adduct has been isolated in 78% yield from areaction of 1with an excess of 2: the other major product from this first reaction in refluxingtoluene is the 1 I adduct 3 of which there are another 3 diastereo isomers. There is a goodreason why this particular 2:l adduct, i.e. 4, accumulates during the reaction: it is because thebisdiene 2 forms monoadducts ca. 100 times faster than it forms bisadducts.7 The 1 1 adduct 3can be converted (Scheme 2) into the macrocycle 6 - a 3.5% yield has been obtained from thecyclod imerisation of 3 in refluxing xylene. However, by far the most efficient way to obtain 6 is tosubject molar equiva lents of the 2:l adduct 4 and the b isdienophile1 o very h igh pressure (10kbars) in dichloromethane solution at 60%: in this manner, the yield of the macropolycycle6 hasbeen raised to 36% in this final step where 5 is presumably the intermediate that precedes thefinal intramolecular cycloaddition step (Scheme 2). Obviously, strong stereoelectron ic effects areoperating at each cycloadd ition step. In accordance with literature precedent,E the bond-form ingapproaches to the dienophilic components of 1 and the diene components of 2 appear to beexclus ively (Scheme 4) ex0 and endo, respectively. On account of the 4 'c lose' (e x e l , endo-l,exo-2, and endo-2) and 2 'remote' (syn and anto configura tional poss ibilities, there are 8 differentways (i.e. here is a two-fold degeneracy) for 1 o react with 2 to give 4 diastereoisomeric 1 1adducts including3 with the syn/endo-H con figuration, which is the only one isola ted. The other 3diastereoisomers3a, 3b, and 3c which have, respectively, he synlexo-H, anti/endeH, andanti/exeH configura tions have not been detected so far. Inspection of molecular models (in thehands or on the computer) demonstrate that the s tereoelectronic requirem ents or the 'close'stereochemistry dictate that the 'remote' stereochemistry must be syn. For steric reasons, antistereochemistry s 'impossible'. And so, steric and stereoelectronic reasons combine9 to impose


3/6

Evo lution of molecular belts and collars 1583Scheme2

- -C H 2 C I 2 / ~ 10 Kbars

CHpCI2A

10Kbarsc--

XyleneAc--

The making of the m olecular belt compound6 by a stereoelectronically-programmedset of Die ls-Alder reactions starting from the bisd ienophile1 and the bisdiene 2. The way thebicyclic ring systems have been drawn, front and back approaches to the reacting n-systemscorrespond to ex0 and endo attacks, respectively.

Scheme3Reaction via the 1:l 2:l 2:2 adducts

[51

There are 4 10 64 possible isomersThe evolutionary synthetic pathway traced by the reactionof the bisdienophile1withthe bisdiene 2 on their way to the macropolycycle6.


4/6

1584 F. H. KOHNKE AND J. F. STODDART

Scheme 4Favoured by a steric effectI I m The bisdienoDhile 1 and bisdiene 2showing the ex0 and endo faces of

the reactive components and arationalisation for the favouredsyn/exol /end&2 transition stateleading to the 1 :1 adduct 3 withsyn/endo-H stereochemistry.0

Syn/Endo-H Isomer3 Endo-lIreble diastereoselectivity via a syn/exo-l/endo-2 transition state. This act is repeated in a further3 Diels-Alder reactions to produce the macropolycycle 6 with its 16 chiral centres. It would seemthat the stereoelectronic features of 1 and 2 are ultimately expressed in 6 in almost evolutionaryfashion. No reagent control is required. We believe we have discovered a stereoelectronically-controlled molecular ' L E G 0 set.9

H H

H HSyn/ndeH Isomer 3 Syn/xeH Isomer 3a

H H

A n t E n d e H Isomer 3b A ntExo-H Isomer 3c

W O R K I N G R O U N D T H E C LO CKThe reaction steps outlined in Scheme 5 summarise how the macropolycycle 6 can be convertedinto the hydrocarbon 10with 6 benzene rings in an alternating cyclic array. We have christened4this compound [12]collarene. A clock numbering system, as employed in Scheme 5, is useful inreferring to the reaction pathway, 6 + 7 --f 8 --f 9 + 10. Whilst deoxygenation at 3 and 9 o'clockproceed smoothly to afford 7, dehydration of 7 is accompanied by some reshuffling of thearomatic rings so that, instead of isolating 8 , with 2 anthracenes and 2 benzenes, thehydrocarbon 9 with 2 naphthalenes, 1 anthracene, and 1 benzene is obtained. Reduction of 8under m ild Birch conditions affords [12]collarene ( lo ) , which is, of course, a precursor of[lelbeltene.

M O L E C U L A R R E C E P T O RSThe X-ray crystal structures of both 6 and 7 have been determined3s4 Fig. 1 ) . In the case of 7 ,a'free' water molecule is trapped inside its Celtic cross-like hydrophobic cavity with its hydrogenatoms > 2.7 8, away from any potentially interactive sites. Although 6 does not form an inclusioncompound with its so lvent (chloroform) of crystallisation, it has been successfully employed'o as adetector coating for the selective piezoelectric quartz crystal detection of nitrobenzene.


5/6

Scheme 5Evolution of molecular belts and collars

[-H201 [-H2011 2TiCI4LiAIH4

9 3 -THF-2[0"1

6[-H201 [-WlHCIA I q O

0 0

0

1585

[12]BelteneThe conversion of the macropolycycle 6 into [I 2]collarene (10). Note that a clocknumbering system can be used to identify the six-membered rings in [12]cyclacene derivatives.

Fig. 1. Space-filling representation of the solid state structures of (a) the rnacropolycycle 6 and(b) its dideoxy derivative 7 with an included disordered water molecule which is represented as asphere with radius equivalent to the enve lope of a water molecule.


6/6

1586 F. H. KOHNKE AN D J. F. STODDART

F UT U RE PERSPECT IVESThe discovery that compounds such as the macropolycycle 6 can be produced efficiently in gramquantities opens up a great deal of exc iting new po~ s i b i l i t i es g~ ~~n synthetic, structural, andreceptor chemistry. As we have already stated9 elsewhere . . . ." . . . he opportunity exists to ring the structural changes on compound 6 ad infinitum. Molecularbelts, big and small, molecular cages, little and large, and molecular strips, linear and coiled, notto mention molecular nets and stacks, all start to swarm into the mind of the chemist attracted bythe uncharted territory of unnatural product synthesis.'* The fact that the synthetic chem ist cancreate molecular materials that are rigid, ordered, and large, without having to resort tocomplicated multistep syntheses using expensive reagents and often employing difficult toremove templates as structural scaffoldings is an intriguing and exciting prospect. The generalityand usefulness of this concept will depend on just how many different sets of molecules can beprogrammed to express themselves through their mutual reactions in unique structural fashionswithout extensive reagent control."AcknowledgementsWe a re deeply indebted to Dr Neil S. lsaacs (Un iversity of Reading) for h is assistance inperforming high pressure reactions, to Mr Peter R. Ashton (University of Sheffield) for helping toidentify many new compounds in the first instance by fast atom bombardment m ass spectrometry,and to Dr David J. Williams and Miss Alexandra M.Z. Slaw in (Imperial College London) for theirrapid solutions to many awkward X-ray crystal structures. Much of the experimental data, onwhich this article is based, was obtained because of their enthusiastic cooperation. We thank theScience and Engineering Research Council, the Ministry of Defence, and the Johnson MattheyTechnology Centre in the United Kingdom, and the University of Messina in Italy for theirgenerous financial support. One (J.F.S.) of us acknowledges the award of a Research Fellowshipfrom the Leverhulme Trust.

1.2.3.

4.5.6.

7.

8.

9.10.1 1 .12.

REFERENCESD.J. Cram, Science279, 177-1 83 1983); ngew. Chem. lnt. Ed. Engl. 25, 039-1 57(1986); cience 240, 60-767 1988); ngew, Chem. lnt. Ed. Engl. 27, 009-1 20 (1 988).0. iels and K. Alder, Ann. Chem. 460, 8-1 2 (1928).F.H. Kohnke and J.F. Stoddart, Abstracts of 794th American Chemical Society NationalMeeting, New Orleans, 30 Aug - 4 Sept 1987, ARB 32; .H. Kohnke, A.M.Z. Slawin,J.F. Stoddart and D.J. Williams, Angew. Chem. lnt. Ed. Engl.26, 92-894 1987).P.R. Ashton, N.S. Isaacs, F.H. Kohnke, A.M.Z. Slawin, C.M. Spencer, J.F. Stoddart andD.J. Williams, Angew. Chem. Int. Ed. Engl. 27,966-969 1988).R.W. Alder and R.B. Sessions, J. Chem. Soc., Perkin Trans.2,1849-1854 1985).H. Hart, N. Raja, M.A. Meador and D.L. Ward, J. Org. Chem.48,4357-4360 1983);F.H. Kohnke, J.F. Stodda rt, A.M.Z. Slawin and D.J. Williams, Acta Cryst. C44, 38-740and 742-745 1988).P. Voge l and A. Florey, Helv. Chim. Acta 57,200-204 1 974); .-A. Carrupt,J.-P. Hagenbuch, A. Florey and P. Vogel, Helv. Chim. Acta 63, 149-1 57 1980).W.H. Watson (Ed), Stereochemistry and Reactivity of Systems containing a-flectrons,Verlag Chem ie, Deerfield Beach, Florida (1983):n particular, the articles by K .N. Houk(p a l) , .A. Paquette (p.41), R. Gleiter and M.C. B6hm (p.105) and P. Vogel (p. 147).P. Ellwood, J.P. M athias, J.F. Stoddart and F.H. Kohnke, Bull. SOC.Chim. Belg., In press.M.A.F. Elmosa lamy, F.H. Kohnke, G.J. Moody, J.F. Stoddart and J.D.R. Thom as,Analytical Proceedings, Submitted.J,F. Stoddart, J. lncl. Phenom., In press; Chem. Brit., In press.J.F. Stoddart, Nature 334, 0-1 (1988).

Franz H. Kohnke and J. Fraser Stoddart- The evolution of molecular belts and collars

Documents