UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Organic metals : studies related to the synthesis of acceptors derived from TCNQ Hammond, Gerald Bruce 1979-03-02

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
831-UBC_1979_A6_7 H35.pdf [ 4.73MB ]
Metadata
JSON: 831-1.0060874.json
JSON-LD: 831-1.0060874-ld.json
RDF/XML (Pretty): 831-1.0060874-rdf.xml
RDF/JSON: 831-1.0060874-rdf.json
Turtle: 831-1.0060874-turtle.txt
N-Triples: 831-1.0060874-rdf-ntriples.txt
Original Record: 831-1.0060874-source.json
Full Text
831-1.0060874-fulltext.txt
Citation
831-1.0060874.ris

Full Text

ORGANIC METALS: STUDIES RELATED TO THE SYNTHESIS OF ACCEPTORS DERIVED FROM TCNQ by Gerald Bruce Hammond B.Sc, Pontificia Universidad Catolica de Peru, 1975 A Thesis Submitted in Partial Fulfillment of the Requirements for the Degree of Master of Science in The Faculty of Graduate Studies (Dept. of Chemistry) We accept this thesis as conforming to the required standard The University of British Columbia April, 1979 © Gerald Bruce Hammond, 1979 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the Head of my Department or by his representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of Chemistry  The University of British Columbia 2075 Wesbrook Place Vancouver, Canada V6T 1W5 HatP May 16, 1979 ABSTRACT Since a report that the complex of TTF.TCNQ showed novel solid state properties including high room temperature conductivity, there has been considerable interest in the study of these so called "organic metals." This thesis describes the synthesis of new acceptors derived from TCNQ. First the synthesis of 2-methoxycarbonyl-7,7,8,8-tetracyanoquino-dimethane (56) is presented. Bromination of methyl 2,5-dimethylbenzoate followed by treatment with sodium cyanide yielded the dicyano compound 60 which was converted to the dihydro-TCNQ 62_ via the corresponding phenylenetetracyanodiacetate intermediate. Oxidation of 6J2 afforded 2-methoxycarbonyl-7,7,8,8-tetracyanoquinodimethane (56) , which was found to be unstable. The synthesis of 2,3-dimethoxycarbonyl-7,7,8,8-tetra-cyanoquinodimethane was attempted in the following manner: the Diels-Alder cyclization of trans,trans-2,4-hexadiene and dimethyl acetylenedi-carboxylate followed by dehydrogenation of the adduct gave dimethyl 3,6-dimethylphthalate. Benzylic bromination of this material followed by displacement of the bromines by cyanide yielded the dicyano compound 73. However, several attempts to convert _73 to the corresponding tetraester 78 were unsuccessful. The synthesis of 2,3-dimethyl-7,7,8,8-tetracyanoquinodimethane is presented. The intermediate 2,3-dimethylcyclohexane-l,4-dione (87) was prepared from 2,3-dimethylphenol by standard means. It was then condensed with malononitrile, and the product was treated with bromine/pyridine to give the dihydro-DMTCNQ 89. When this compound was treated with a mixture of palladium and sulfur at 180°C the desired 2,3-dimethyl-TCNQ 83 was obtained in low yield. -iii— The second part of this thesis describes the preparation of 11,11, 12 ,12-tetracyano-4,5,9,10-tetrahydropyreno-2,7-quinodimethane (TCNTP) from mesitylene. This compound was bisbrominated to give 3,5-bis-(bromomethyl)toluene, which was then coupled with 3,5-bis(mercaptomethyl)-toluene to produce the dithia[3.3jmetacyclophane 98. This was converted to the dimethyl[2.2 jmetacyclophane 9_6 via Wittig rearrangement followed by reduction with lithium in liquid ammonia. This cyclophane was then brominated and the resulting dibromo compound was converted to the corresponding dicyano-94. The latter compound was treated with bromine in the presence of iron powder to produce 2,7-bis(cyanomethyl)-4,5,9,10-tetrahydropyrene. Introduction of the remaining two nitrile groups was accomplished via use of the tetracyanodiacetate intermediate 9_2. Hydrolysis and decarboxylation of 9_2 followed by oxidation with NBS/triethylamine at -78°C furnished the desired TCNTP (90). 7 £ —v-Table of Contents Page Title Page i Abstract ii Table of Contents v List of Illustrations vList of Schemes viAcknowledgements vii1. Introduction 1 2. General Background 2.1 TCNQ Complexes and Salts 4 2.2 Extended TCNQ's 16 2.3 Substituted TCNQ's 23 3. Results and Discussion 3.1 Synthesis of substituted TCNQ's 29 3.2 Synthesis of 11,11,12,12-tetracyano-4,5,9,10- 48 tetrahydropyreno-2,7-quinodimethane. 3.3 Overall results 64 Experimental 66 Bibliography 93 Spectral Appendix 97 -vi-List of Illustrations Page 1. Scale of approximate conductivities. 2 2. Dimensions and relative conductivities of a 9 TTF.TCNQ crystal. 3. Temperature dependence of d.c. conductivity (o) 10 of TTF.TCNQ. 4. a) Crystal structure of TTF.TCNQ 11 b) Molecular overlap in the columnar stacks of TTF and TCNQ. 5. Charge migration using a hopping model. 13 6. UV-Vis spectrum of 2,7-bis(dicyanomethyl)-4,5,9,10- 60 tetrahydropyrene after treatment with aqueous bromine. 7. UV-Vis spectrum of TBA+ TCNDQ~ and TBA+ TCNTP~(114). 63-64 -vii-List of Schemes Page I. Synthesis of TCNQ. 5-6 II. Synthesis of TTF. 9 III. Synthesis of TNAP (14). 17 IV. Improved synthesis of TNAP (14). 18 V. Synthesis of the dihydro-TCNDQ 27. 20 VI. Oxidation states of TCQQ (28). 2 VII. Attempted synthesis of TCQQ (28). 23 VIII. General method for the preparation of 24 substituted TCNQ's. IX. Synthesis of some p-xylylene dihalides. 24 X. Synthesis of tetrafluoro-TCNQ (42) and 25 2,5-dicyano-TCNQ (43). XI. Synthesis of the polyurethane ^8_. 28 XII. Preparation of MTCNQ (49). 29 XIII. Synthesis of 2-methoxycarbonyl-7,7,8,8-tetracyano- 33 quinodimethane (56). XIV. Synthesis of 3,6-bis(cyanomethyl)phthalic 39 anhydride (67). XV. Proposed synthesis of 2,3-dimethoxycarbonyl-7,7,8,8- 42 tetracyanoquinodimethane (115). XVI. Proposed route for the synthesis of 2,3-dimethyl- 45-46 TCNQ (83) . XVII. Proposed synthesis of TCNTP (90) from 5,13-dimethyl- 49-50 [2.2]-metacyclophane (96). XVIII. Alternate pathways to anti-5,13-dimethyl[2.2]meta- 52 cyclophane (96). XIX. Synthesis of 6,15-dimethyl-2,11-dithia[3.3]meta- 52 cyclophane (98) . -viii-Acknowledgements I am very grateful to Dr. Larry Weiler for his guidance, kind encouragement and understanding throughout my period of research. I also wish to express my gratitude to all the members of the Organic Metals group, past and present, for their many helpful dis cussions. Special thanks are due to Drs. Y. Hoyano and M. R. Bryce, and Ms. J. Lee. I also thank Anna Wong for typing the manuscript. If a man does not keep pace with his companions, perhaps it is because he hears a different drummer. Let him step to the music which he hears, however measured or far away... H. D. Thoreau A mis padres, con todo el amor que solo ellos son capaces de inspirar -1-1. INTRODUCTION In 1973 researchers at JohnsHopkins University1detected signs of very high conductivity (around 1500 ohms *"cm 1 at room temperature) from a 1:1 organic complex of 7,7,8,8-tetracyano-p-quinodimethane (TCNQ) (1_) , and tetrathiafulvalene (TTF) (2) . There quickly followed 2 an exciting and controversial report from a group at U. of Pennsylvania who claimed that some crystals of TTF.TCNQ showed "superconducting fluctuations," with increases in conductivity at 58°K of several hundred times the room temperature value. Research in the field of organic conductors has been very active ever since. VO S s-^ 2 This new class of organic solids, typified by TTF.TCNQ, displays unexpected "metallic like" behavior over a wide temperature range. The materials are composed solely of non-metallic atoms, such as carbon, nitrogen, sulphur and oxygen; hence, the term "organic metal" has been used to describe these solids. These organic metals are part of a larger class of compounds known as charge-transfer salts; they are formed when an electron is transferred from an organic donor to an acceptor molecule. The interest in the novel transport properties of these organic metals can be fully appreciated by comparing the conductivities of most organic compounds: typical values are 10 11 to 10 7 ohm 1cm 1 (insulators) -5 -3 -1 -1 and 10 to 10 ohm cm (semiconductors). However, the complex of hexamethylene-tetraselenafulvalene (HMTSF) and TCNQ has a room tem-the highest of any known 3 -1 -1 perature conductivity of 2 x 10 ohm cm organic substance. Organic Metals Nichrome Silicon Most molecular crystals log a |+ 6 Metals + 4 + 2 0 Molten salts -2 -4 Alkali TCNQ salts -10 Carbazole-iodine complex -12 Ferrocene 1-14 -16 Anthracene Figure 1. Scale of approximate conductivities The organic molecules which form these complexes are relatively large and planar. They form stacks in the solid state with their TT-molecular orbitals interacting strongly along the stacks, and only weakly -3-between them, which gives them a quasi-one dimensional nature that is reflected in their physical and electronic properties. TTF.TCNQ was identified as the first organic metal and has become the prototype organic conductor. The crystals are highly anisotropic, and diverse studies of their electronic and physical properties are all consistent with a metallic nature at room temperature. The driving force behind the considerable amount of research done on one-dimensional metals has been the search for high temperature superconductivity. The quest for a totally organic superconductor remains a major but unsuccessful goal. Nevertheless, a by-product of this ambition has been the wealth of exciting physics and chemistry which came as a result of these endeavors, and that is finding practical applications in an ever growing number of areas, such as electrolytic capacitors, solid state batteries and printing systems. -4-2. GENERAL BACKGROUND 2.1 TCNQ Complexes and Salts Research in organic conductors has been progressing for some time and the published data have centered around charge-transfer complexes, free radicals and free radical salts, long chain compounds and polymers. Charge transfer complexes have been studied extensively although in many cases their structures were not well characterized. It was found that -3 a surprising minority of these complexes had conductivities above 10 ohm "'"cm \ These types of complexes containing iodine and aromatic hydrocarbons or aromatic amines provided the first examples of moderately high electrical conductivity in organic materials (graphitized complexes are not included). It was found that tetracyanoethylene (TCNE) (_3) could polymerize with 4 5 some metal chelates ' forming charge transfer complexes that had re latively low resistivities (around 0.04 ohm cm). These results increased the exploration of the chemistry of other unsaturated systems containing the highly electronegative nitrile group, particularly, structures in which the double bond of TCNE is replaced by a conjugated system. On the other hand, quinones are known to be good electron acceptors, also forming stable complexes with some electron donors.^ C—C NC / \ CN 3 -5-NC •CN NC CN TCNQ (1) 1 In 1962, D. S. Acker and W. R. Hertler' synthesized a substituted quinodimethane , namely 7 ,7 ,8,8-tetracyanoquinodimethane (1_), a compound with several exceptional properties; among them the formation of stable anion radical derivatives which have unusually low electrical resisti vities in the solid state. The synthesis of TCNQ is illustrated in Scheme I. It involves the Knoevenagel condensation of 1,4-cyclohexane-dione (4) with malononitrile in the presence of the Prout catalyst, 8-alanine, to give the tetrahydro intermediate 5^, in 94% yield; this compound is then dehydrogenated either with bromine/pyridine at 0°C in acetonitrile, or with manganese dioxide by brief refluxing in g toluene. 4 5 -6-Vridine ^ TCNQ (1) Scheme I. Synthesis of TCNQ (l)7 The unique characteristics of this molecule, namely, its high electron affinity, its planar structure and high symmetry, allowed the formation of electrically conducting donor-acceptor complexes, with TCNQ as the 9 acceptor. According to Melby et al. these complexes could be sub divided into three groups characterized by their composition and con duction behavior: i) TCNQ TT Complexes. These are crystalline complexes formed with aromatic hydrocarbons, amines, chelates, and other neutral compounds, _3 usually in a ratio of 1:1. They have conductivities ranging from 10 -12 -1 -1 to 10 ohm cm . From these early studies it was clear that there was no correlation between the basicity of the donor and the electrical conductivity of the complex. Steric factors and crystal packing seemed to play an important role. ii) TCNQ salts of the composition M+n(TCNQ') . The conductivity of salts of this group (M+n stands for a metal ion or an organic cation) is usually in the range of 10 ^ to 10 ^ ohm "'"cm "*". The esr absorption is very weak. Usually the conductivity in these simple salts is isotropic. iii) TCNQ Salt Complexes of the composition M+n(TCNQr)^ TCNQ°. These complex TCNQ salts which, beside the TCNQ' radical anion, contain -3 a neutral TCNQ molecule, display high conductivities ranging from 10 to +2 -1 -1 10 ohm cm . They show variable esr absorption. Furthermore, the -7-conductivity of single crystals displays strong anisotropy. Table I shows some examples of the different molecular complexes of TCNQ. Table I. Conductivities of some TCNQ Complexes and Salts. Type Tr-complex H2N Me2N Donor NH, NMe, Conductivity (ohm "'"cm L) 3 x 10 -4 10 -6 Simple salts Cs + 3 x 10 -5 Cu 5 x 10 -3 Complex salts 100 (C2H5)3N 1.5 x 10 -1 During the mid 1960's the search for new and better organic semi conductors failed to yield a true organic metal. Some of the different theories of organic conductivity in organic solids that emerged were controversial and lacked experimental support. However, some facts received attention. It was realized that if a low resistivity 1:1 complex is to be formed, there should be considerable, but not complete, charge transfer. Acker and Blomstron^ found that a condition for high conductivity -8-is that the cation be a planar, aromatic, heterocyclic molecule. Further more, Le Blanc11 suggested the importance of having a polarizable cation. 12 In 1965, Melby made a complex between N-methylphenazinium (NMP)(6) and TCNQ. Its single crystal conductivity reached 380 ohm "'"cm 1 at room temperature. This remained the most conductive organic solid until 1973, when Perlstein, Cowan et^ al1 mixed TCNQ and TTF to produce a new compound with exciting and novel solid state properties, including a room tem perature conductivity ranging from 600 to 1800 ohm 1cm 1 (variable from sample to sample). The conductivity of this 1:1 complex increases with decreasing temperature down to 60°K (metallic region) where there is a transition to an insulating state (T<60°K). The peak conductivity at T=59°K is comparable to the conductivity found in some metals such as 4 -1 -1 lead and tin (around 10 ohm cm ). The organic donor, TTF, is a planar molecule with an extensive TT electron system; it is highly symmetric and very similar in size to TCNQ. This molecule is also highly polarizable because of the presence of four sulfur atoms, and it can be readily oxidized to a radical ion, thereby number 14,15,16 13 forming conducting cation radical salts with some halides. A number of synthetic pathways to TTF and its derivatives are available. 16 One route to TTF developed by Melby et_ al_. (Scheme II) involves depro-tonation of a 1,3-dithiolium ion, synthesized from the commercially available dimethyl acetylenedicarboxylate (_7_) and ethylene trithio-carbonate (8) . -9-c Me + CO Me -S" s-2 Me 02C •F6P H-^+ 1) HOAc,A^ (T^Vc 2) pyridine,A^-^/ 11 13 CH3C03H ^2 NaPF* "04SH H-^S" 12 Scheme II. Synthesis of TTF (2) 16 The 1:1 complex of TCNQ and TTF produces lustrous black plates when recrystallized from acetonitrile. Figure 2 shows a diagramatic representation of a single crystal, including approximate dimensions, along with the corresponding room temperature conductivities."'"7 This material exhibits a highly anisotropic conductivity which reaches a 0.5 mm 5mm. -0.05mm. cr = 600-1800 ohm 1cm 1 b io-= 150 Figure 2. Dimensions and relative conductivities of a TTF.TCNQ crystal. -10-maximum along the crystallographic b axis. The temperature dependence of the dc conductivity along the b axis is shown in Figure 3.^ The conductivity peaks at 59-60°K and begins to fall at lower temperatures,going through two transitions at 53°K and 38°K. The magnitude of oT /o300°K max is still a subject of controversy: it is found to be around 20 in most 2 cases, although a case was reported in which the values for certain crystals reached as high as 150. o-(T) cr ( 300 K) T (K) Figure 3. Temperature dependence of d.c. conductivity (a) of TTF.TCNQ15 Insert shows the transitions at 53°K and 38°K. 18 The crystal structure of this material is shown in Figure 4. TTF.TCNQ consists of segregated, uniform columns of TTF and TCNQ molecules -li st acked face-to-face along the highly conducting b axis. The molecular planes are tilted with respect to the b stacking axis, and TTF and TCNQ have opposite tilts relative to each other. The intrastack distances are shorter than the van der Waals radii of TTF or TCNQ, and permit a strong overlap of Tr-molecular orbitals along the stacks. a) \ W__x___y/ X___X_X X / XI7" b) X \ "A \ Figure 4. a) Crystal structure of TTF.TCNQ. b) molecular overlap in the columnar stacks of TTF and TCNQ, showing the lateral displacements of their molecular centres (5). Above the transition temperature the TTF and TCNQ molecules are uniformly spaced within the stack. As mentioned earlier, TTF is a good electron donor, and TCNQ is a 18 good electron acceptor. According to several recent investigations , the charge transfer in the solid state in TTF.TCNQ amounts to 0.6±0.01 of one electron. In TCNQ radical anion the HOMO is a TT-MO. The interaction of these MO's on two adjacent TCNQ within the same stack is bonding, when each -12-uiolecule is displaced laterally with respect to the molecule below. This gives rise to the "ring-double bond" overlap shown in Figure 4. In addition, calculations of the energy surface for the dimer (TCNQ^ show a well defined minimum when <S, distance between the molecular centres °19 ° (Figure 4), is 2.2A (observed value: 2.14A). Similar calculations on TTF+ ions show a strong interaction of the HOMO's, but the energy surface calculations for TTF dimer show a minimum only for the case of two TTF's stacked directly over each other. The fact that this type of TTF packing has been observed in the highly conducting mixed valence halides such as 13c TTF.I 71 , indicates that the interaction of HOMO's on the TCNQ stack and crystal forces are responsible for the packing of the solid in TTF.TCNQ. The picture for TTF.TCNQ that emerges at this point is one of segregated stacks containing planar, partially charged molecules arranged so as to allow relatively strong Tr-orbital overlap within the stack; thus parallel conducting linear chains are formed which are only weakly coupled. There is still controversy over which description best explains the conductivity in these compounds. A useful model that describes the room temperature conductivity of TTF.TCNQ can be obtained by extending the MO approach of an individual TCNQ and TTF+ to an infinite array of linear chains. Thus the corresponding HOMO's of the radical ions generate a conduction band consisting of MO's delocalized along the stack. Calculations' using the extended Huckel Method of Hoffmann, showed that the largest overlap occurs along the highly conducting b axis, and that the TCNQ band is wider than the TTF band. There can be metallic type conductivity in the TCNQ or TTF band as long as it is partially filled, as in normal metals. These 1-D organic solids, however, are subject to lattice dis-21 tortions (Peierls instability) that can create an energy gap in the 20 -13-electronic band. In the case of a half-filled band this will cause the molecules within a stack, to dimerize. This will lead to a semiconducting 22 behavior. It has been proposed that such a distortion may be suppressing the metallic conductivity at 59°K in TTF.TCNQ. The partial, as opposed to full, charge transfer in TTF.TCNQ has two important consequences: it provides crystal stabilization by minimizing intrastack repulsive interactions, and it lowers on site electron-electron repulsive interactions in both the donor and acceptor stacks. This effect is important if we are to consider another approach to the conductivity in these systems: that is, a hopping model, in which charge migrates along the donor and acceptor stacks, with some donors bearing two positive charges and some acceptors bearing two negative charges. This is illustrated in Figure 5. An additional requirement then, would be that these molecules D+ D+ D+. • D+ D+ D D+D+.. D+ D* A A A A~A~ A= A" A: A" A (a) 2D D + D 2A - ± A +• A Figure 5. Charge migration using a hopping model; (a) stacked charge transfer salt; (b) the same salt under the effect of an electric field. -14-should be able to localize two charges. A measurement of the redox potentials of the different species in solution could be useful in providing some information and suggestions for the design of new organic metals. The redox potentials would provide a measure of the electron accepting or donating ability of the components in the charge transfer salt D+ + e" D° E 1D A° -H e" A E 1A D + A D+ 4- A" E - E = AE 1A 1D Finally, an approach that regards the organic metals as resulting from the intermolecular migration of aromatieity was first proposed by 23 Perlstein. Under this concept, the most efficient conductors should contain molecules (a) whose radical ions form a new aromatic sextet upon one-electron oxidation or reduction, and (b) whose aromatieity can migrate by mixed valence interaction. Since the initial report on TTF.TCNQ, an ever-growing number of new compounds have been added to this class of metallic-like organic solids. Much of this effort has focused on preparing derivatives of TTF, either through replacement of the ring heteroatoms, or by addition of substituents Table II contains the electrical conductivity data of TCNQ salts of TTF and some of its derivatives. From this Tnhle it is apparent that sub stitution on the basic skeleton of TTF may produce solids that have similar -15-conductivities Co TTF.TCNQ. Indeed, the substitution of selenium for sulfur in TTF seems to improve the room temperature conductivity. This may result from the greater orbital extent of selenium, which could lead to stronger overlap along the donor stack and, hence, to an increased band width. Nevertheless, the maximum conductivity for these new salts is similar to that of TTF.TCNQ, although the metallic state has been stabilized to higher temperatures. Table II. Electrical conductivity data on the TCNQ salts of TTF and derivatives R Conductivity data for TCNQ salt "2^*2 aRTb a c Td RL .ia max ,a P Name ohm cm ohm~icm-1 (°K) Tetrathiafulvalene X ,=S,R.. . =H 400-500 >104 58 (TTF) 1-4 Diselenadithia- X =S, X =Se, 500 3xl03 64 fulvalene x' _ (DSDTF) R1-4_H Tetraselena- X^^Se, RL_4=H 800 >104 '40 fulvalene (TSF) Hexamethylene- X1_4=S' Ri 2=R3 4 500 2xl03 80 tetrathiafulvalene ,r„ >. ' ' (HMTTF) -(CH2;3 -4 Benzo-tetrathia- X.. . =S, R.. = 10 fulvalene , R3,4 (C4H4) Tetramethyl - TSF X, ,=Se, R, ,= 103 6xl03 71 1-4 1-4 CH3 Hexamethylene - TSF XL_4=Se, ^ 2= 2xl03 7xl03 45-75e R3>4=(CH2)3 a b single-crystal measurement. a = room temperature conductivity. RT , c d a = maximum conductivity at T . Temperature of maximum conductivity, max J p J e Still highly conducting below T^. -16-Some of the chemical and structural features that may be required 2 15 2 A-for metallic behavior are summarized below: ' ' - the existence of unpaired electrons within molecular units. - a uniform crystal structure consisting of parallel and segregated stacked columns of planar molecules with delocalized TT-molecular orbitals. These ir-orbitals must have significant intermolecular overlap. - large Tr-systems in which the open shell cation and anion are of similar size. - a polarizable cation - a certain amount of disorder (including doping) in the one-dimensional system may prevent a uniform periodic distortion of the lattice, and thus may stabilize the metallic state. A different approach to the design of organic metals has been 25 given by Torrance. According to him, the most important properties to consider are stacking, degree of charge transfer, and stoichiometry. Although most of the characteristic features of these organic metals are widely accepted, the relative importance of each of these parameters is still a matter of discussion. 2.2 Extended TCNQ's The unpaired electron in the anion radical salts of TCNQ is largely localized on the terminal dicyano methylene groups. It is the ability of TCNQ to accommodate two electrons at opposite ends of the molecule which distinguishes it from most other acceptors. Another potential quino-dimethane acceptor would be 11, 12, 11 , 12 - tetracyanonaphtho-2,6-quinodimethane (TNAP)(14), an extended, conjugated analog of TCNQ. The -17-2 6 preparation of TNAP, first described by Hertler et al. , is shown in Scheme III. The synthesis started with 2,6-dimethylnaphthalene (15) which was brominated with NBS (N-bromosuccinimide) to yield the dibromide 16. Displacement of bromide with cyanide gave 1_7_ in poor yield. This was then treated with diethyl carbonate and base to give diethyl 2,6-naphthalenedicyanoacetate (18). Treatment of the cyano ester 18 with aqueous ammonia gave the corresponding amide 1_9 as a mixture of diastereo-raers. After purification, 19 was dehydrated with POC£^ to give 2,6-naphthalenedimalononitrile (20). Finally, oxidation of the dimalononitrile (20) with N-iodosuccinimide gave TNAP in 40% yield. Scheme III. Synthesis of TNAP (14) 26 -18-27 This synthetic route was later improved by Sandman and Garito with a procedure that both avoided the stereochemical problem of the earlier synthesis, and improved the overall yield. The reaction of 17_ with sodium ethoxide and diethyl carbonate in toluene, followed by reaction of the resultant dianion with cyanogen chloride, led to a 60% yield of diethyl 2,6-naphthalene a,a,a',a'-tetracyanodiacetate (21). Hydrolysis and decarboxylation of 21_ gave 20_ which was oxidized in situ with aqueous bromine to afford TNAP in essentially quantitative yield (Scheme IV). The anion radical salts of TNAP have compaction resistivities similar -19-to those reported for salts of TCNQ.26 Single crystals of TTF.TNAP. obtained 28 via a slow diffusion process , show uniform, segregated stacks of cation and anion radicals. However, the room temperature conductivity is only 40 ohm "*"cm 1 and exhibits little temperature dependence; a sharp "metal-insulator" transition occurs at 185°K. Next in the series is 13,13,14,14-tetracyanodiphenoquinodimethane (TCNDQ)(22), which has not been isolated as a neutral species. Attempts 27 29a b to prepare it have led only to formation of polymeric material. ' ' However, a number of salts of the anions of TCNDQ have been prepared, and 29a their physical properties studied. The preparation of the dihydro-29a TCNDQ 27_ , the immediate precursor of TCNDQ, was accomplished using biphenyl (2_3) as the starting material. Bisbromomethylation followed by displacement of the bromide with sodium cyanide afforded the dicyano-25 which was converted to the dihydro - TCNDQ _2_7 following the procedure of 27 24a Sandman and Garito (Scheme V). Garito and Heeger suggested that a strong Ti-conjugation between the distant withdrawing groups is required to maintain the molecular stability. If the conjugation is weakened, 22 -20-23 22 Scheme V. 27_ 26 Synthesis of the dihydro - TCNDQ 27.27'29a'b unwanted polymerization occurs. This could explain the difficulty in obtaining a neutral TCNDQ; the inter-ring hydrogen repulsions could 30 account for the weakening of the Tt-conjugation. As stated previously, an analysis of the redox potentials of the donors and acceptors in solution could serve as an experimental approxi-24a mat ion in the evaluation of the Coulombic repulsion energies , and -21-therefore, it could be a practical guide in the search for new donors and acceptors. Table III lists the redox properties of some of these acceptors. a29a,31b,c Acceptor AE TCNE TCNQ 0.15 0.13 0.52 0.21 0.08 •0.57 0.29 0.03 0.17 .08 0.72 0.42 0.49 0.38 0.16 F.-TCNQ TNAP TCNDQ a) reduction potentials, in volts, vs SCE in CH^CN reference electrode The experimental conductivities of the charge transfer salts obtained from some of these acceptors are in agreement with the observed trends in AE (difference between the first two polarographic reduction potentials). All complexes of TCNE are insulators, whereas complexes of TCNQ and TNAP are metallic. A few complexes of F^-TCNQ are semi-con ductors. In the case of TCNDQ, the proximity of the two reduction potentials would indicate a clear reduction of the Coulombic repulsions in the dianion. 30 The search for larger acceptors led Wudl et al_. to attempt the synthesis of 2,7-bisdicyanoquinomethano-2,7,H,H-quinazolino[6,5,4-def]-quinazoline (TCQQ)(28). This is a pyrene analog of TCNDQ in which four carbon atoms have been replaced by nitrogen. This substitution was expected to both enhance the electron affinity of the acceptor,and to increase the number of available oxidation states (Scheme VI). The attempted preparation of TCQQ is outlined in Scheme VII. The tetrahydro intermediate 35 could not be purified and was treated directly with base in acetonitrile -22-Scheme VI. Oxidation states of TCQQ (28) in the presence of either tetrabutylammonium or tetraphenylarsonium chlo ride to give the dianion 3_0 as the only characterizable product. Unfortunately, attempted oxidations of 3J3 to 29_ or 28_ were not successful. 2.3 Substituted TCNQ's A systematic study of the synthesis of substituted TCNQ's, together with the observed conducting properties of their charge transfer complexes 31 with a series of donors, has been performed by Wheland. The reported TCNQ derivatives have substituents in the ring; these include alkyl groups (e.g. Me, Et, i-Pr), halogens (F, C£, Br, I), ethers and thio-ethers (e.g. OMe, O-i-Pr, O-i-Bu, SMe), and nitriles. These syntheses most frequently started with the corresponding p-xylylene dihalide which in most cases could be prepared by direct bis-chioromethylation of an appropriately substituted benzene 3_6 as shown in Scheme VIII. For some dihalides and haloethers derivatives, the p-xylylene dihalide was prepared according to Scheme IX, starting with an -24-C02Me Scheme VIII. General method for the preparation of substituted TCNQ's.31a appropriately substituted p-xylene 37_. The corresponding terephthalic acid 3_8_ was converted to the acid chloride 39^ which was reduced to the glycol 4CL This was then treated with hydrogen halide to give the p-xylylene dihalide 4_1. Once the p-xylylene dihalides had been obtained, Scheme IX. Synthesis of some p-xylylene dihalides. -25-the syntheses of all the substituted TCNQ's were similar, as summarized in Scheme VIII. For the tetrafluoro derivative 4_2 and the dicyano-TCNQ 43, the first step involved initial nucleophilic displacement of aromatic halogen by the anion of tert-butylmalononitrile (44). Thermolysis of the resulting product with the loss of isobutylene, followed by oxidation with aqueous-bromine, gave the TCNQ's 4_2 and 4_3 (Scheme X). Scheme X. Synthesis of tetrafluoro-TCNQ (42) and 2,5-dicyano-TCNQ (43).31a -26-A correlation of the electrical conductivity of different charge 31c transfer complexes, with the redox potentials of each of the components, shows that strong acceptors, for which complete electron transfer is most likely, give poorly conductive complexes; whereas weaker electron acceptors give highly conductive complexes. On the other hand, lowering donor strength to the point where significant electron transfer is unlikely, uniformly gives poorly conductive complexes. Therefore, it can be speculated that the conductivity will increase with increasing amounts of charge transfer up to a maximum occurring at less than full electron transfer; and then conductivity will decrease as one electron transfer is approached. For optimum charge transfer, the redox potentials of both donor and acceptor should be closely matched (|E -E |< 0.25 V). However, regardless of how favourable these factors are, conductivity will be poor if the requirements of crystal structure are not met. In the characteristic segregated stacked structure of TTF.TCNQ, the spacing of the molecules in their stacks is closer than their van der Waals radii. The consequences of this close encounter have been described in previous sections. Steric factors may play a major role in the crystal packing. 31c The effect of substitution on the TCNQ ring with increasingly larger groups appears to have minor effects until five or six side chain carbons or oxygens have been added. Beyond this point there is a sudden decrease in conductivity of the complex, and in some cases the complex does not form at all. Furthermore, the lowest resistivities are found when the acceptor length/donor length ratio is approximately equal to 1, suggesting that mismatching of donor and acceptor sizes upsets the crystal structure. The need for highly polarizable substituents close to the conductive stack has been suggested in several recent theories."'"'2'243 In accord with -27-this, the complex of TCNQ with tetraselenafulvalene (in which the sulfur of TTF has been replaced by heavier selenium) shows a higher conductivity at room temperature than does TTF-TCNQ. The conductivity of the com plexes of substituted TCNQ's increases when a chlorine substituent is replaced by the heavier and more polarizable bromine or iodine atom. This effect is cumulative in that two bromines or iodines are more effective than one. However, there is only a slight increase in con ductivity in going from bromine to iodine, suggesting that the larger size of the iodine might be altering other variables. Some alkyl derivatives of TCNQ have been used to form charge transfer salts with derivatives of TTF and TSF. Results from these studies show that when small amounts of methyl-TCNQ 49_ are added as a dopant to TSF.TCNQ, the characteristic sharp phase transition is broadened, 4 and the conductivity at low temperatures is 10 times greater than for 32 the corresponding undoped system. Interest in synthesizing on organic polymer having metallic con ductivity led to the preparation of a polymer containing covalently 33 bound TCNQ's. Hertler synthesized 2,5-bis(2-hydroxyethoxy)-7,7,8,8-tetracyanoquinodimethane (45) starting from 1,4-bis(2-acetoxyethoxy) benzene (46) . A charge transfer polymer was prepared by reaction of 4_5 with 1,1-di-isocyanatoferrocene (47) to give a black polyurethane 4_8, which contained electron acceptor TCNQ units alternating with electron donating ferrocene units (Scheme XI). However, the conductivity of the compacted powder was -3 -1 -1 only 3x10 ohm cm heme XI. Synthesis of the polyurethane 48. -29-3. RESULTS AND DISCUSSION 3.1 Studies related to the synthesis of substituted TCNQ's In view of the interesting properties observed in TCNQ and its derivatives, it was decided to initiate a synthetic program leading to various analogs of this molecule. It seemed particularly attractive to investigate the possible formation of derivatives having C-substituents with oxidation levels different from that of saturated hydrocarbons. Furthermore, the construction of symmetrical or unsymmetrical dimers containing the TCNQ moiety could be facilitated by the use of functionalized substituents serving as synthetic handles to achieve dimerization. In the 31a past, work done in this area included the preparation of ethers, thio-ethers, halides, nitriles and alkyl groups as substituents on the TCNQ ring. Our initial study involved the functionalization of an alkyl derivative of TCNQ, in particular, 2-methyl-7,7,8,8-tetracyanoquinodi-methane (MTCNQ) (49) . The synthesis of 49^ is outlined in Scheme XII. The preparation of the intermediate 2-methyl-l,4^cyclohexanedione (52) was Scheme XII. Synthesis of MTCNQ (49). -30-accomplished by a modification of the procedure reported by Hertler et al. shown below. The starting material in both cases was the readily available OMe 50 52 methyl-p-hydroquinone (50). In our case it was hydrogenated to the corresponding diol-51 with rhodium on powdered charcoal as a catalyst, in 90% yield. The crude mixture of stereoisomers was oxidized with a 34 modification of Jones reagent to produce the dione-52 in 54% yield.2-Methylcyclohexane-1,4-dione (52) was condensed with malononitrile in the presence of g-alanine to give 2-methyl-l,4-bis(dicyanomethylene)-cyclohexane (53)(85% yield). This could be used without further puri fication in a reaction with bromine and pyridine to produce 49 in 54% yield after purification by sublimation. The low yield in this reaction 2 6 is attributed to the simultaneous formation of a dimer of 4j}. The 26 physical and spectroscopic data of 49 were similar with those reported. Its nmr spectrum showed two singlets, at 62.70 (3H,-CH^) and 67.15 (IH, -r=CH) , and two doublets at 67.43 (2H,J«=4Rz, vinyl). CH3 Initial attempts to functionalize the methyl group of 4_9 with NBS in carbon tetrachloride were unsuccessful. A change of solvent or initiator had no positive effect, and in all cases starting, material was recovered -31-unchanged. It was hoped that the vinyl methyl group of 4_9 could undergo an allylic oxidation on treatment with selenium dioxide. However, when SeC^ was added to a solution of MTCNQ in refluxing ethanol, starting material was still present after 24 hours (by tic analysis). The resistance of _49_ to oxidation might be due to its high oxidation potential, this would make the olefinic bond less accessible for electrophilic attack 35 to form the intermediate allylic selenite ester 54. Treatment of MTCNQ with sodium hydride in THF (tetrahydrofuran) at room temperature, followed by addition of phenyl selenium chloride at -78°C, yielded only starting material. Similar results were obtained when potassium hydride in DME (1,2-dimethoxyethane) was used, followed by addition of bromine. Warming the reaction mixture to room temperature led, in both cases, to complex mixtures of compounds (at least four by tic), including starting material. In light of the difficulty encountered with the above experiments, a different approach to substituted TCNQ's was taken. It involved introduction of an alkyl halide group onto the ring of a dihydro derivative of TCNQ, namely p-phenylenedimalononitrile (55). This compound was prepared by reduction of TCNQ with thioglycolic acid in glacial acetic acid7(85% yield). Haloalkylation was attempted with paraformaldehyde and hydrogen bromide following the procedure -32-developed by Reynolds and Durham. Hydrogen bromide was bubbled through a refluxing suspension of the dihydro-TCNQ 55 and paraformaldehyde in acetic acid. Analysis of the resulting gummy solid showed disappearance of the nitrile band in the ir spectrum. In addition, strong bands at 3600-3200 cm L and 1710 cm * indicated that hydrolysis of the nitriles had taken place. It was decided at this point to introduce the functionalized carbon substituent at the beginning of the synthesis, prior to the construction of the TCNQ frame. An ester functionality seemed an appropriate choice, and the synthesis of 2-methoxycarbonyl-7,7,8,8-tetracyanoquinodimethane (56) was designed in the manner illustrated in Scheme XIII. 56 -33-fir0 NC 56 CN Q„ Me Scheme XIII. Synthesis of 2-methoxycarbonyl-7,7,8,8-tetracyano-quinodimethane (56). The commercially available 2,5-dimethyl benzoic acid (57) was esteri-37 fied under Fischer conditions to produce the corresponding methyl ester 58 in 85% yield. The next step involved bromination on the two methyl groups. When NBS in carbon tetrachloride was used with benzoyl peroxide (radical initiator), the desired dibromo compound _59_ was obtained in 38% yield. By-products included both monobromo isomers and the tribromo -34-derivative. The amount of side products obtained, together with the low yield of 5_9 prompted the search for better reaction conditions. Vogtle 38 et al. reported that the use of methylene chloride as solvent in the benzylic bromination of 1,8-di-p-tolylnaphthalene (63) afforded the dibromo product Jp_4_ in 80% yield; in carbon tetrachloride the yield was around 50%. In our case best results were obtained when a slight excess 63 64 of NBS (2.2 equivalents) was added to a mixture of 58_ and AIBN (azobis-isobutyronitrile) in methylene chloride, followed by refluxing for 2.5 hours. The resulting dibromo ester 59_ was isolated from the reaction mixture by fractional crystallization in methanol. The yield was 70% after recrystallization. The nmr spectrum of this compound showed absorptions at 63.93 (s,3H,-OCH ), 64.43(s,2H,-CH2Br), 64.90(s,2H,-CH2Br), 67.40(s,2H,aryl) and 67.93(s,IH,aryl), which were consistent with structure 59. 39 Reeves et al. have reported the facile conversion of primary alkyl bromides into alkyl cyanides under phase transfer conditions employing primary, secondary or tertiary amines as catalysts. They noted that several requisites are apparent for efficient catalysis: in particular, the use of long chain aliphatic amines, with a chain length of at least -35-six carbons. In addition, cyanide displacements are known to occur more rapidly when acetonitrile is used as solvent. Displacement of the bromines in 59_ by cyanide took place when a solution of the dibromo ester in acetonitrile was stirred at room temperature for ten hours with an excess of aqueous sodium cyanide. Tri-n-octylamine was used as catalyst; the yield in this reaction was 70%. The absorption at 2290 cm 1 in the ir spectrum of 60_ showed the presence of the nitrile group, and the aryl ester carbonyl has an absorption at 1720 cm The nmr spectrum of the dicyano product showed an upfield shift for the methylene protons. Introduction of the remaining two nitrile groups was accomplished via the phenylenetetracyanodiacetate intermediate 61_, according to the 29a procedure of Wheland and Martin. Toluene instead of benzene was used as solvent for the reaction. Treatment of ^0 with sodium methoxide in toluene-dimethyl carbonate establishes an equilibrium with dianion _6_5 which may be driven to completion by distilling off a toluene-methanol azeotrope. Addition of cyanogen chloride to the reaction mixture at 0°C, is followed by gradual warming of the mixture up to 80°C. The tetracyano-diacetatej61, is obtained in relatively good yield (59%). The ir spectrum -36-C^o)2co -MeOH f of the crude product showed a new absorption band at 1770 cm indicating the presence of a saturated ester. Its "*"H nmr spectrum had a new singlet at 64.00 corresponding to the new methoxycarbonyl groups at C-7 and C-8, and lacked the two singlets at 63.77 and 4.12 attributed to the cyano-methylene groups. The mass spectrum of the compound showed the expected parent peak at m/e 380. Attempted purification of the crude oil by chromatography on silica gel, resulted in partial hydrolysis of the triester. The crude product could be used without further purification in the next reaction; this involved hydrolysis and decarboxylation of the 31a ester groups to produce the dihydro-TCNQ b2_. Wheland had used aqueous potassium hydroxide or sodium hydroxide, either hot (70°C) or at room temperature, to effect hydrolysis and decarboxylation of a number of TCNQ -37-derivatives. In our case, milder conditions were desired in order to avoid total or partial hydrolysis of the aryl ester group. Sodium carbonate, being a weaker base, served our purposes. Indeed, when a solution of the triester J31_ in THF was stirred at room temperature with an excess of aqueous sodium carbonate, in the presence of small amounts of sodium borohydride, the dianion 66_ was formed. Neutralization with hydrochloric acid gave the dihydro compound _62_. Its ir spectrum lacked the absorption band at 1770 cm but retained the aryl ester band. Purification of this compound presented great difficulties. The crude material streaked on tic; attempted recrystallization proved unsuccessful; and sublimation in vacuum only de composed the product. However, the acidity of the methine protons in 62 permitted purification via acid-base extraction. After work-up the resulting greenish oil partially crystallized on standing. Gradual darkening of the solid occurred on standing at room temperature after several days. The overall yield after purification was 32% (from the dicyano-60). -38-Oxidation to the desired ester-TCNQ _56 was carried out with aqueous 31a bromine at room temperature. Attempted purification of the resulting light green solid proved difficult. Sublimation at 190°C/0.01 mm Hg yielded a green-yellow solid (3%), m.p. 220°-224° (decomp.), which darkened gradually on standing. The ^"H nmr spectrum of this material was of poor quality and was unreliable; this was due to low solubility, and to the presence of radical anions formed by reaction with solvent impurities. Its ir spectrum showed a weak absorption band at 2220 cm 1 (conjugated C=N) and a strong band at 1725 cm 1 (unsaturated ester C=0). Its mass spectrum further substantiated the proposed structure by having 262 (100) as the molecular ion peak. The product could not be purified for elemental analysis. However, high resolution mass spectrometry established the molecular formula of the product to be C,H,N,0„, as 14 6 4 2 required by structure _56. Its electronic spectrum showed an absorption band at 396 nm (36,800) with a shoulder at 375 nm. If this solution is left standing, its color darkens and the UV spectrum shows a shift to 409 nm. Attempts to form a charge transfer complex with TTF in acetonitrile failed. A large planar molecule containing two identical acceptor sites at both ends could be of interest in the study of new acceptors. If one of the acceptor sites was a TCNQ molecule, the resulting system could have a structure such as: -39-In order to build this symmetrical carbon skeleton, a vicinal disubstituted TCNQ was needed. Initially, it was hoped that the use of a Diels-Alder reaction between TCNQ and an appropriate diene would serve our purposes. However, it has been reported7 that this reaction has not given any characterizable adducts when various dienes were used, and in some cases TCNQ was recovered unchanged. It was decided that the use of a disubstituted TCNQ with appropriate functional groups would be more convenient. Accordingly, the synthesis of 3,6-bis(cyanomethyl)phthalic anhydride (67) was planned as follows (Scheme XIV). It was hoped that this compound could be converted to the corresponding TCNQ derivative. 68 69 • NC CN 67 Br 71 70 Scheme XIV. Synthesis of 3,6-bis(cyanomethyl)phthalic anhydride (67). 3,6-Dimethylphthalic anhydride (70) was prepared by a Diels-Alder reaction of 2,5-dimethylfuran (68) and maleic anhydride. The corresponding -40-adduct was dehydrated directly by treatment with concentrated sulfuric 41 acid at -10°C to give 7_0 in 24% overall yield. The low yield obtained might be partially explained by the occurrence of an acid catalyzed 42 reversal of the Diels-Alder reaction, which competes with the acid-catalyzed dehydration to _7_0. When the phthalic anhydride was bisbro-43 minated under Wohl-Ziegler conditions , a mixture of mono and dibromo derivatives was obtained. The dibromo-71 precipitated out from the mixture and was purified by recrystallization (26% yield). Its nmr spectrum showed resonances at 64.86 (s^Hj-CH^Br) and 67.82 (s,2H,aryl). Its ir spectrum retained the three absorption bands at 1856, 1777 and 1260 cm 1 corresponding to the acid anhydride. Displacement of bromide by cyanide in the presence of sodium iodide 44 under anhydrous conditions led to a mixture of compounds. Based on the ^H nmr spectrum of the crude mixture, the formation of a five-membered ring lactone was suspected. In addition, signals at 64.72 and 64.96 were indicative of cyanomethylene and bromomethylene groups, respectively. The formation of a lactone was explained in terms of an intramolecular displacement of bromide by a carboxylate from the anhydride. It was hoped that substitution of the anhydride by an ester group would permit a cleaner substitution of bromine by nitrile. However, an attempt to convert the anhydride 7_1 into the diester _76, using methanol and p-toluenesulfonic 45 1 acid , produced only the bromolactone ester 7_2 in 51% yield. The H nmr spectrum of 72_ contained a three proton singlet at 64.01 attributed -41-Br .Ci Me 2 O Br •O 71 72 to the methoxy group, while at 64.55 occurred a two proton singlet which was assigned to the methylene protons of the bromomethyl group. The singlet at 65.24 corresponded to the two protons of the y lactone ring, while the aromatic protons gave rise to an AB-pattern (J=8Hz) centered at 67.55. Its ir spectrum showed sharp, strong absorptions at 1775 cm_1 (lactone C=0) and 1735 cm"1 (aryl ester C=0). It was decided at this point to try to synthesize the corresponding diester _73 as outlined in Scheme XV. The preparation of dimethyl 3,6-dimethylphthalate (75) via a Diels-Alder reaction of trans,trans-2,4-46 hexadiene and dimethyl acetylenedicarboxylate has been reported. The Diels-Alder cyclization took place in refluxing benzene to give the adduct 74_ in essentially quantitative yield. Dehydrogenation with DDQ (5,6-dichloro-2,3-dicyano-p-benzoquinone) in benzene overnight, led to the desired dimethylphthalate 7_5 in 80% overall yield. Its physical and 46 spectral data were in agreement with those reported in the literature. -42-115 73 76 Scheme XV. Proposed synthesis of 2,3-dimethoxycarbonyl-7,7,8,8-tetracyanoquinodimethane (115) 46 Gundermann et_ al_. have bisbrominated 7_5 under Wohl-Ziegler 43 conditions , in 37% yield, by using 1.7 equivalents of NBS. When a stoichiometric amount of NBS was used, the tribromoderivative 7J7 was obtained. In our hands, the bromination of 7_5 was effected using the 77 conditions developed for the monoester _58_. In this case, a mixture of di- and tribromo compounds 7_6 and 7_7 was produced. The tribromo derivative could be separated from the crude product by dissolving the solid mixture -43-in minimum amounts of boiling ethanol followed by overnight cooling and selective crystallization by addition of methanol. The crude dibromo-76 (64% yield) was treated directly with sodium cyanide, in a manner similar to that described for the monoester 5^. The yield of the desired dimethyl 3,6-bis(cyanomethyl)phthalate (73) was 44% after purification by preparative column chromatography. Analytical and spectroscopic data are consistent with structure 73_. Notably, the ir spectrum shows a weak band at 2290 cm 1 corresponding to nitrile absorption, and again shows a sharp, strong band at 1730 cm 1 (aryl ester C=0). Attempts to introduce the remaining two nitrile groups via use of a phenylenetetracyanodiacetate intermediate failed to give the desired tetraester 78. An unidentified mixture was obtained and no further charac-CC^Me terizations were made. Use of potassium hydride and methyl chloroformate in DME led to similar results. Treatment of _73 with base followed by direct alkylation with cyanogen chloride proved unsuccessful. In both 4 cases complex mixtures of compounds were obtained. It has been reported that intramolecular electrophilic cyclization of the acid chloride 79 leads to the isoquinolone derivative 80. In light of this fact, it is -44-To our knowledge, the synthesis of a vicinal dialkyl derivative of 49 TCNQ has not been reported in the literature. Studies of the tetrathia-and tetraselenafulvalene salts of 2,5-dimethyl-7,7,8,8-tetracyanoquinodi-methane (82) have shown that the conductivities of these salts are comparable -45-to that of other TCNQ analogs. However the metal-insulator transition temperature is lower in the case of the dlalkyl derivative 8_2. This could arise from disorder effects in the solid, or these results might indicate that solid state properties are mainly dominated by the actual crystal structure rather than by electronic properties of donors and acceptors. 2,3-Dimethyl-7,7,8,8-tetracyanoquinodimethane (2,3-DMTCNQ) (83) is larger than TCNQ and has a different molecular symmetry than TCNQ. A characteristic of the conducting solids formed by TCNQ is the occurrence of separate stacks of acceptors and donors. The dimethyl-TCNQ J33 carries the possibility of different stacking patterns. Stacks of 83 might be formed in an alternating arrangement to facilitate minimum steric repulsion of methyl groups of adjacent molecules. Alternatively, the methyl groups can be randomized causing disorder in the lattice. This latter possibility could cause a broadened phase transition and greater conductivities at low temperatures, as is the case when small amounts of methyl-TCNQ 49_ are 32*15? added as a dopant to TSF-TCNQ. » In view of these interesting features offered by 2,3-DMTCNO (83), a synthesis of this molecule was attempted, as depicted in Scheme XVI. Synthesis of the dione-87 was accomplished 85 -46-Scheme XVI. Proposed route for the synthesis of 2,3-DMTCNQ (S3). following Yamakawa's route. Oxidation of 2,3-dimethyl phenol (84) was readily achieved via the Teuber reaction"'1, with two equivalents of potassium nitrosodisulfonate (Fremy's salt) to produce an intermediate p-benzoquinone. This quinone is reduced in situ during the work-up with aqueous sodium thiosulphate, giving the hydroquinone 85 in 75% yield; 48 its m.p. was 167-170°C (lit. m.p. 163-164°C). The hydrogenation step that follows was carried out using the same procedure as for the hydro quinone _50. The crude oil containing a mixture of isomers was isolated in 95% yield. Oxidation of the diol-86, in the manner described for 51, gave 2,3-dimethylcyclohexane-»l,4-dione (87) in 75% yield. Its ir and 1 48 H nmr spectra were identical to those supplied by Yamakawa. The dione-87 was condensed^ with malononitrile to produce the expected tetracyano-88 in 67% yield. The white solid thus obtained lacked a carbonyl absorption in its ir spectrum, but contained a new absorption band at 2250 cm 1 (conjugated C=N). Treatment of £38 with -47-bromine/pyridine failed to give the desired dimethyl-TCNQ 83. Instead, the partially oxidized dihydro derivative 89^ was obtained in 86% yield. Its 1H nmr spectrum showed a singlet at 67.06 corresponding to the two vinyl protons, and it included signals at 61.16 (d,J=7Hz,6H,CH^-CH) and 63.17 (m,2H,-CH-CH3). Further evidence for structure 89^ came from its ir and mass spectra and microanalysis. Efforts to oxidize the allylic position in 89_ with either NBS or NCS (N-chlorosuccinimide) were un successful. In both cases starting material was recovered. The following conditions led to a similar result: manganese dioxide in refluxing toluene, or DDQ in refluxing dioxane. The search for stronger dehydrogenation conditions prompted the use of a mixture of sulfur and 52 palladium as dehydrogenating agents. Small amounts of this mixture were finely ground with 89 and the resulting powder was heated to 180°C for 3 hours. Analysis of the extract (by tic) showed several colored spots including starting material. The band at 0.5 (eluent methylene chloride) was collected and found to be the desired product 83. Its nmr spectrum showed two singlets at 62.35 (6H,-CH3) and 67.43 (2H, vinyl). Its mass spectrum further substantiated this assignment by having 232 as the highest molecular ion recorded. Unfortunately, when -48-this reaction was scaled up the results were different and 83_ could not be obtained. No further investigations were made. 3.2 Synthesis of -11,11,12,12-tetracyano-4,5,9,10-tetrahydropyreno-2,7-quinodimethane (TCNTP) (90). Among the extended conjugated analogs of TCNQ studied to date, TCNDQ (22) suggested some interesting possibilities for further studies. Un-27 29 fortunately, the neutral species has not been isolated. ' It was hoped that the TCNDQ analog 90_, with its pyrene skeleton, would be more stable 30 because biphenyl inter-ring hydrogen repulsions are eliminated. The NC\/CN NCv^^CN -49-proposed synthesis of TCNTP (90) is outlined in Scheme XVII. This scheme shows how a figurative dissection to simpler systems could help in the search for an appropriate starting material. TCNTP could be obtained from the dihydro precursor 91_ using mild oxidation conditions. In turn 91 could be prepared from the dicyano-93 via hydrolysis and decarboxylation of the phenylenetetracyanodiacetate intermediate 92_. Moving backwards in stepwise fashion, obtention of 93_ requires a transannular reaction of the dicyano[2.2]metacyclophane 94_- This compound can be formed from 95 by displacement of bromide with cyanide. The dibromo metacyclophane 9_5 is available from the corresponding dimethyl compound 96^ by treatment with NBS." -50-Scheme XVII. Proposed synthesis of TCNTP (90) from 5,13-dimethyl-[2.2]metacyclophane (96). The metacyclophane _96_ contained all the necessary elements for the synthesis of 90; and it has been prepared in the past by intermolecular coupling of 3,5-bis(bromomethyl)toluene (97), via use of the Wurtz 54 55 reaction, in the presence of sodium (12% yield), or phenyllithium (no yield reported). The yields in this type of reaction are usually low 5 6 and the amounts of by-products large. On the other hand, the formation of the corresponding dithiacyclophane _98_ generally occurs in high yield, -51-and so it has been used as precursor for the synthesis of 9_6 in three ways: (i) the extrusion of sulphur has been accomplished by conversion into the corresponding bis-sulphone 9J9, followed by elimination of 53 sulphur dioxide on pyrolysis at 800°C (76%)(Scheme XVIII a); (ii) the extrusion of sulphur has been achieved in one step by irradiation of small amounts of 9_8_ in the presence of trimethyl phosphite;"^ (iii) 58 recently, Mitchell _et_ al. found that the Wittig rearrangement of 98, with n-butyllithium, followed by alkylation with methyl iodide, gave 100 as a mixture of isomers in excellent yield. This mixture was converted cleanly to anti-5,13-dimethyl[2.2]metacyclophane (96) by reduction with lithium in liquid ammonia (Scheme XVIII b). 58 It was decided to use Mitchell's route to £6 in view of the fact that both steps could be carried out on a large scale and in good yields. The synthesis of 6,15-dimethyl-2,ll-dithio[3.3]metacyclophane (98) is 59 depicted in Scheme XIX. Bromination of mesitylene (101) with NBS afforded 3,5-bis(bromomethyl)toluene (97) in 50% yield after recrystalli-zation. Conversion of the dibromo-97 to the dithiol-102 was achieved in 95% yield using the thiourea method.5^'^u This method involves the formation of the diisothiouronium bromide 103, via nucleophilic attack on the halide by the sulfur. Hydrolysis of this salt in refluxing potassium Scheme XIX. Synthesis of 6,15-dimethyl-2,11-dithia[3.3]-metacyclophane (98). -53-hydroxide gave 102. Its nmr spectrum showed signals at 61.72 (t,J=8Hz, 2H.-SH), 62.31 (s,3H,-CH3), 63.64 (d,J=8Hz,4H,-CH2-) and 66.97 (s,3H,aryl). Coupling of 97_ and 102 occurred in an ethanolic solution of potassium hydroxide under high dilution conditions^1 and over a period of 7 days. The crude product was purified by filtration through silica gel. Thus, the dithiacyclophane 98 was obtained in 85% yield. A recrystallized sample had a melting point 101.5-103.5°C (lit.53 mp 102-102.5°C). 59 Spectral data were in agreement with those reported previously Treatment of 98 with n-butyllithium in THF at -20°C, followed by methylation of the resulting thiolate afforded the mixture of isomers 100. The crude mixture was not purified but was used directly in the reduction step that follows. Cleavage of the carbon-sulfur bond was achieved by reduction of 100 with lithium in liquid ammonia. The crude reaction product was purified by preparative column chromatography; subsequent recrystallization from ethanol furnished 96 in 30% overall yield, mp 145-146°C (lit.53mp 145-146°C). -54-It was found that NBS bromination of 96_ displayed a remarkable solvent dependence: in methylene chloride a transannular reaction occurred with the consequent formation of the tetrahydropyrene 104; while in carbon tetrachloride, the bisbromide 95 was formed exclusively, in 70% yield. The "*"H nmr spectrum of 9_5 showed the characteristic pattern of 104 f2.2]metacyclophanes.°Z In particular, the intraannular protons exhibited an unusual high field shift at 64.25 (s) due to the ring current effect of the other benzene ring. Since the ten-membered ring system of _95 exists in a rigid chair conformation, the bridging ethyle signal appears as an A2B2~pattern arising from axial and equatorial protons: 6axial=2.14, 6 equatorial=3.13. Furthermore, the benzylic -55-methylene groups bearing bromine appear as a singlet at 64.56. In the case of 104, its ^"H nmr spectrum consists of three singlets at 62.3b (methyl), 62.87 (methylene) and 66.96 (aryl) in a ratio of 3:4:2 respect ively. It has been reported that in the NBS bromination of naphthaleno-38 paracyclophane 105, when methylene chloride was used as solvent, bromination occurred exclusively in the aromatic ring to form the cyclo phane 106 while if carbon tetrachloride was used, the mono bromo-107 was formed alongside the gem-dibromo-108. 108 XzY = Br A similar solvent effect may be operating in the NBS bromination of 96. The tetrahydropyrene 104 now has six benzylic positions, and NBS bromination on this molecule would be expected to yield a mixture of -56-compounds. When 104 is treated with NBS in carbon tetrachloride, one of the major products of the mixture is 2,7-diraethylpyrene (109). Its 1H nmr spectrum showed two singlets at 62.80 (6H,-CH3) and 67.98 (8H,aryl) which are consistent with the literature values.^ Displacement of bromine in dibromo-95 by cyanide was achieved in a manner similar to that described for the mono- and diester ^9 and 7_6 respectively. Due to solubility problems, the reaction was carried out in refluxing THF. The resulting dicyano compound _94 was obtained in 70% yield. It exhibited an absorption band in the ir, at 2300 cm 1 (C=N) , and its ^"H nmr spectrum showed an expected upfield shift for the cyanomethyl groups to 63.80 (s,4H). Compound 9_4 could now be converted into the desired pyrene framework by transannular reaction between the 64 two rings. It has been reported that treatment of 110 with bromine in the presence of catalytic amounts of iron powder, leads to the 53 formation of 111. More recently, Misumi e_t al , have used these 109 104 conditions to achieve a transannular oxidation of 95. The cyclophane 94 -57-was converted into the dicyanopyrene 93 by stirring the mixture of 94, bromine in chloroform, and iron powder at 0°C for 1.5 hours. After work-up the crude reaction mixture is chromatographed, and the dicyano-pyrene is obtained in 70% yield. Evidence for structure 93_ was supplied by its ''"H nmr spectrum which showed the disappearance of the signal at 64.24 (intraannular aromatic protons). In addition, since structure ^3 is planar, the bridging ethylene protons are all equivalent, and the •^2^2~pattern coHaPse^ to a singlet at 62.83 (8H). Based on the 64 mechanistic studies on the transformation of 110 into 111, the following mechanism was suggested for the transannular reaction of 94. Intermediate 112 is formed by the electrophilic attack of the bromonium ion on one ring, accompanied by a simultaneous transannular attack of the generated phenonium ion on the second ring. Regeneration of the aromatic character by loss of a proton followed by elimination of the elements of hydrogen bromide gave rise to the observed product. -58--59-Introduction of the remaining two cyano groups was achieved via formation of the diester 9_2, in a manner similar to that described previously for the synthesis of the triester 6^1. The ir spectrum of 92 showed a strong absorption at 1745 cm 1 (saturated ester C=0). Its nmr spectrum exhibited a new signal at 63.91 (s,6H,-OCH^), and its mass spectrum showed the expected M~^~ at 450. The crude product was not purified any further but was used directly in the hydrolysis step that follows. Hydrolysis and decarboxylation of 92 afforded the dihydro-TCNTP 91. This reaction was carried out using the same procedure as for the triester 61. Disappearance of the signal corresponding to the methoxy carbonyl group in the "'"H nmr spectrum of 91 is accompanied by the appearance of a new singlet at 65.00 (2H,(CN)2CH-). Purification methods such as chromato graphy, recrystallization and sublimation failed to give any positive results, and, as yet, this compound has not been obtained in the pure state. It was decided, however, to attempt the synthesis of TCNTP (90) using crude compound 91_ in view of the fact that, from our past experience, TCNQ derivatives have been more easily purified than the corresponding dihydro compounds. Oxidation of 91_ to the desired TCNTP proved more difficult than g anticipated. The use of manganese dioxide in refluxing toluene led only to unidentified products; treatment with aqueous bromine at room temperature produced a highly insoluble green-brown solid, mp 206-210°C (decomp.). Its ir spectrum showed a band at 2220 cm 1 indicative of a conjugated nitrile, (TCNQ shows a band at 2220 cm ^)^. Furthermore its mass spectrum showed two peaks at 334 and 332 in a 2:1 ratio, together with small peaks at 412 and 414 (1:1 ratio). The latter peaks were believed -60-to be due to the presence of some bromine containing impurities and will be discussed later. It was felt that the M+ + 2 peak observed at 334 ori ginated from compound 91_, formed by aromatization of TCNTP in the probe (T=270°C). A small sample (5 mg) was recrystallized from acetonitrile (ca. 400 ml), mp 210-212°C (decomp). This compound gave an intense purple color in solution. Elemental analysis, however, did not correspond with the desired values. Its UV-Vis spectrum is shown in Figure 6. When this compound was treated with thioglycolic acid in acetic acid under the Figure 6. UV-Vis spectrum of 2,7-bis(dicyanomethyl)-4,5,9,10-tetrahydropyrene (91) after treatment with aqueous bromine. same conditions used for the reduction of TCNQ to the dihydro-TCNQ 55, the corresponding dihydropyrene 91 was the only material isolated from the reaction mixture. -61-Other oxidation conditions were tried, among them, the use of NCS and triethylamine in methylene chloride, at -20°C; in this reaction a drastic color change took place when triethylamine was added, and the color of the reaction mixture turned intense purple. A tic examination of the product mixture revealed the presence of two main products; analysis of the less polar band isolated by preparative tic (eluent chloroform) yielded a blue solid, mp 227-230°C, readily soluble in most organic solvents. The presence of two chlorine atoms in the molecule was evident by the characteristic isotopic distribution exhibited by the molecular ions at 402, 404, 406 (9:6:1) in its mass spectrum. In addition, its "'"H nmr spectrum showed two singlets at 63.50 and 67.56 in a 4:1 ratio, perhaps suggesting chlorine substitution on the benzene rings. Treatment of 9_1_ with NBS in methylene chloride at -78°C followed by addition of triethylamine yielded a yellow-brown solid. Tic analysis showed one spot, Rf^0.5 (eluent chloroform); the solution ir spectrum of the crude reaction product showed a weak absorption at 2228 cm 1 (conjugated CHN). Purification of the crude product was attempted by filtration through silica-gel followed by recry-stallization from benzene; the green microcrystals so obtained melted at 206-208°C (decomposition). Its ir spectrum (KBr pellet) showed two weak absorptions at 2225 and 2265 cm 1; the nmr spectrum showed signals at 62.94 and 67.20, assigned to the methylene and vinyl protons of 90 respectively. However, due to the presence of unidentified products an accurate percentage yield of 90 could not be calculated. The mass spectrum of 9_0 contained peaks at 332 and 334 corresponding to TCNTP (90) and dihydro-TCNTP 91 respectively, together with two lower intensity peaks at 412 and 414 indicating the presence of some bromine containing impurities, as observed previously. The mass spectrum was then recorded at different -62-temperatures ranging from 280°C to 450°C. In this interval, the ratio of the peaks at 334 and 332 decreased gradually with increasing temperature, from a value of 3.1 at 280°C to a value of 1.8 at 450°C. At the same time the peaks at 412 and 414 disappeared gradually. These results seem to indicate: (a) the peaks at 412 and 414 are derived from a compound other than 90; (b) the dihydro compound (m/e=334) is probably formed from TCNTP (m/e=332) in the mass spectrometer. The UV-Vis spectrum of 90 is shown below. 700 -63-On treatment of _9_0 with excess tetrabutylammonium iodide in benzene at room temperature, the color of the solution went' from purple to light red. After evaporation of the solvent, a mixture of TCNTP~ TBA+ (114) (TEA"1": tetrabutylammonium cation) and unreacted tetrabutylammonium iodide was obtained. Its ir spectrum showed a strong sharp nitrile absorption at 2180 cm and a smaller absorption at 2145 cm . Such a low absorption frequency for the nitriles is consistent with the value of 2200-2175 cm 1 9 observed for the tetraethylammonium salt of TCNQ . Furthermore the UV-Vis spectrum of 114 in benzene is similar to that observed for the radical anion of TCNDQ (2_2) , as would be expected for the formation of TCNTP~. Both spectra are shown in Figure 7. I + TCNTP -> -I + TCNTP € xlO"3 -64-Figure 7. UV-Vis spectrum of (i) TBA TCNDQ and (ii) TBA TCNTP (114). 3.3 Overall Results In summary, the work done towards the synthesis of a pyrene analog of TCNDQ (22) has resulted in the preparation of 11,11,]2,12-tetracyano-4,5,9,10-tetrahydropyreno-2,7-quinodimethane (TCNTP)(90). Pre]iminary results obtained from some exploratory experiments show that this molecule -65-could be potentially useful in the study of new donors derived from extended TCNQ analogs. From the work described in the first section of this chapter it can be concluded that the synthesis of new TCNQ deri vatives containing functionalized carbon substituents should start with appropriately substituted benzene or cyclohexane derivatives as opposed to the rather unreactive alkyl substituted TCNQ's. This viewpoint resulted in the synthesis of 2-methoxycarbonyl-7,7,8,8-tetracyano-quinodimethane (56). Finally, the synthesis of 2,3-dimethyl-7,7,8,8-tetracyanoquinodimethane (83) has proved surprisingly difficult because of the stability of compound 89 to oxidation. A new route should be devised, in view of interesting structural features in 83. -66-EXPERIMENTAL GENERAL Spectral Data All infrared (IR) spectra were taken in chloroform solution, unless otherwise stated. They were recorded on a Perkin-Elmer model 700 spectrophotometer and were calibrated with the 1601 cm ^ band of poly styrene. The assignment of each band is noted in parentheses after it. ''"H NMR spectra were recorded on a Varian model T-60 or model XL-100 spectrometer. The solvent used is reported in parentheses at the beginning of each spectrum. Chemical shifts (6) are quoted in p.p.m. downfield from tetramethylsilane, used as an internal standard. The multiplicity, coupling constants (quoted in Hz), integrated peak areas and proton assignments are indicated in parentheses after each signal. Ultraviolet-visible (UV-Vis) spectra were recorded on a Cary 17D spectrometer. The molar extinction coefficients of the main bands are 2 listed in parentheses after the absorptions and are expressed in cm , -1 mole Mass spectra were obtained using an Atlas CH-4B mass spectrometer, and high resolution determinations were obtained using an AEI MS-902 or MS-50 mass spectrometer. Both instruments were operated at an ionising potential of 70 eV. The value in parentheses after each mass is its relative intensity. Physical Data. Melting points (mp) were determined on a Kofler hot stage microscope and are uncorrected. Elemental microanalyses were performed by Mr. Peter -67-Borda, University of British Columbia. Chromatography Preparative thin layer chromatography (TLC) was carried out using 0.9 mm thickness silica gel (E. Merck, grade PF-254 + 366) on either 20 x 20 cm or 5 x 20 cm glass plates. Small analytical plates were prepared by dipping microscopic slides into a stirred solution of silica gel in chloroform. Column chromatography was done using silica gel of grade finer than 200 mesh. Solvents The term "dry" in this thesis refers to solvents dried in the following ways: ethyl ether and tetrahydrofuran (THF) were dried by re fluxing over lithium aluminium hydride. Toluene and benzene were dried over sodium. Carbon tetrachloride, methylene chloride and acetonitrile were distilled from phosphorous pentoxide. Methanol was dried by refluxing over magnesium methoxide. Techniques Anhydrous magnesium sulfate was used to dry organic solutions and was removed by filtration. All stirring, unless otherwise noted, was carried out using teflon-coated magnetic stirring bars. 2-Hethyl-l,4-cyclohexanedione (52)* Methyl-p-hydroquinone (50) (8.8 g, 0.071 mole), rhodium (10%) on powdered charcoal (1.55 g) and a methanol-water mixture (198 ml, 10:1 V/V) were placed in a hydrogenation bottle which was then attached to a Parr-low-pressure apparatus. Air was evacuated and the flask was flushed with hydrogen several times. The reduction was then carried out at room -68-temperature at an initial hydrogen pressure of 50-60 psi. The reaction was complete when a constant hydrogen pressure was reached (approximately 22 hours). The catalyst was removed by filtration, washed with methanol (50 ml), and dried in vacuo. The catalyst could be used again after it had been reactivated - two days in the oven at 110°C. Solvents from the filtrate were removed under reduced pressure; traces of methanol and water were azeotropically distilled with chloroform (30 ml). The resultant oil (8.3 g, 90%) was shown, by ir and "''H nmr spectra, to be 2-methyl-l,4-cyclohexanediol (51) which was not purified but was used directly in the oxidation step that follows. A solution of chromium trioxide (26.72 g, 9.27 mole), water (77 ml), concentrated sulfuric acid (23 ml), and manganese sulfate (0.2 g) was prepared, and added slowly to a stirred solution of the crude diol-51 (1.52 g, 0.012 mole) in chloroform (50 ml) at 5°C. The addition was stopped after 3 equivalents (13 ml) of chromium trioxide had been introduced (30 minutes). Stirring was continued for 4 hours, and then the chloroform layer was separated. The aqueous layer was extracted with chloroform (3 x 50 ml) and the combined extracts were washed with an aqueous solution of concentrated ammonium hydroxide (1 ml) in saturated sodium chloride (10 ml). After drying, the solvent was removed under reduced pressure to yield 2-methyl-l,4-cyclohexanedione (52) (0.79 g, 54%) . IR v 1715 cm"1 (C=0). max 1H NMR(CDC£3)6 1.15(d,J=6,3H,-CH3>, 2.3-2.95(m,3H,-CH2-CH-) and 2.60 ppm(s,4H,-CH2-). Mass spectrum m/e 55(54) 56(100), 57(41), 68(58), 69(75), 71(32), 82(16), 83(19), 111(43) and 126(91). * Experiment first performed by Ms. J. Lee. -69-2-Methyl-7,7,8,8-tetracyanoquinodimethane (49) The compound was prepared by the same procedure as that reported by 2 6 Hertler et a_l. The reagents used were: 2-methyl-l,4-cyclohexanedione (52) (939 mg, 7.45 mmole), malononitrile (993 mg, 15.1 mmole) and a 2% aqueous solution of 8-alanine (1.68 ml), which gave 1.407 g (85%) of 2-methyl-l,4-bis(dicyanomethylene)cyclohexane (53) , identified by its ir and nmr spectra. Compound 52^ (440 mg, 1.98 mmole) was not purified but was treated with a 10% solution of bromine in acetonitrile (2 ml) and pyridine (1 ml) to give 460 mg of 2-methyl-7,7,8,8-tetracyanoquino dimethane (49). Purification of compound 49 was achieved by sublimation at 150°C/0.01 mm (250 mg, 54%), mp 198-200°C (lit26 mp 200-201°C). Its ir, nmr and mass spectra were in accord with the proposed structure. Methyl 2,5-dimethylbenzoate (58) The preparation of methyl 2,5-dimethylbenzoate (58) followed the 37 method described by Falk and Schlogl. The reagents used were: 2,5-dimethylbenzoic acid (57) (16.82 g, 0.112 mole), methanol (46 ml) and concentrated sulfuric acid (5 ml), which gave 15.20 g (82%) of methyl 2,5-dimethylbenzoate (58) after vacuum distillation, bp 70-72°C/0.5 mm (lit37 bp 106-107°C/13 mm). IR v 1715 cm_1 (aryl ester C=0). max J 1H NMR (CDC£3) 62.28 (s,3H,m-CH3), 2.48 (s,3H,o-CH3) 3.77 (s,3H,-OCH3); 6.97 (s,2H,aryl) and 7.55 ppm (s,lH,aryl). Mass spectrum m/e 77(21), 104(32), 105(58), 132(88), 133(100), 149(17) and 164(62). -70-Methyl 2,5-bis(bromomethyl)benzoate (59) An oven-dried 200-ml flask was equipped with a condenser, magnetic stirrer, and heating mantle. The condenser was stoppered with a septum cap, fitted with nitrogen inlet and outlet syringe needles, and flushed with nitrogen. The reaction vessel was then charged with methyl 2,5-dimethyl benzoate (58) (5.53 g, 0.034 mole), NBS (recrystallized from water, 13.5 g, 0.075 mole), azobisisobutyronitrile (0.3 g) and dry methylene chloride (60 ml). The mixture was stirred and heated to reflux for 2.5 hours, during which time it turned yellow and then orange. The reaction mixture was then cooled and filtered. The filtrate was diluted with 100 ml of methylene chloride and washed successively with saturated aqueous solutions of sodium bisulphate (1 x 25 ml) and sodium bicarbonate (1 x 25 ml) to remove excess bromine and hydrobromic acid. After washing with water (2 x 30 ml), the organic layer was dried and the solvent re moved under reduced pressure. The resulting oily solid was covered with methanol and kept at -5°C overnight. The white solid was collected by vacuum filtration and recrystallized from methanol-pentane to yield 7.66 g (70%) of pure compound (59) as white crystals, mp 81.5-82.5°C. IR v 1720 cm"1 (aryl ester C=0). max 1H-NMR(CDC£3) 63.93(s,3H,-0CH3), 4.43(s,2H,m-CH2Br), 4.90(s,2H, o-CH2Br), 7.40(s,2H,aryl) and 7.93 ppm(s,lH,aryl). Mass spectrum m/e 77(26), 119(25), 121(22), 122(17), 147(35), 162(67), 163(48), 209(17), 211(19), 241(100), 243(96), 291(13), 320(10), 321(17), 322(20) and 324(10). Analysis calculated for C^H^Br^: C 37.30, H 3.13, Br 49.63; found: C 37.25, H 2.98, Br 49.40. -71-Methyl 2,5-bis(cyanomethyl)benzoate (60) Sodium cyanide (9.0 g, 0.183 mole) was weighed into a 250-ml flask and then dissolved in 30 ml of water. A solution of 59_ (9.15 g, 0.028 mole) in acetonitrile (73 ml) was added, followed by 0.33 ml of tri-n-octylamine (3% by weight of 59). The mixture was stirred for 10 hours at room temperature and poured into water (200 ml). The mixture was extracted with methylene chloride (3 x 100 ml), and the extracts were washed successively with aqueous sodium bisulphate (50 ml), dilute hydrochloric acid (50 ml), and saturated sodium chloride solution (50 ml). After drying, the solvents were evaporated under reduced pressure to give 6.1 g of a buff-colored solid. Purification of 60_ was achieved by careful recrystallization from methanol, followed by chromatography of the mother liquors using a carbon tetrachloride-ethyl ether mixture (9:1 V/V). Methyl 2,5-bis(cyanomethyl)benzoate (60) was isolated as a white microcrystalline powder (4.19 g, 70%), mp 98-100°C. IR v 2290(C=N) and 1722 cm"1 (aryl ester C=0). max """H NMR(CDC£3) 63.77 (s,2H,m-CH2CN) , 3. 90 (s , 3H ,-0CH3) , 4.17(s,2H, o-CH2CN), 7.50(s,2H,aryl) and 7.95 ppm(s,lH,aryl). Mass spectrum m/e 83(26), 101(18), 127(19), 128(26), 154(26), 155(33), 177(20), 172(19), 182(100), 183(68) and 214(84). Analysis calculated for C12H10N2°2:C 67-28» H 4.71, N 13.08; found C 67.32, H 4.88, N 12.66. Dimethyl a,a, a",a^-tetracyano-2-methoxycarbonyl-l,4-phenylenediacetate (61). An oven-dried, three-neck, 250-ml round-bottom flask was equipped with a 2 in magnetic stirring bar, an oil bath and distillation head. The flask was stoppered with a ground glass bubbler and a septum cap fitted -72-with a nitrogen inlet needle. The whole system was flamed and allowed to cool to room temperature under a nitrogen flow. Fresh sodium methoxide was prepared in situ by addition of sodium metal (0.8 g, 0.034 mole) to dry methanol (10 ml) followed by removal of the excess alcohol under reduced pressure. The methoxide was suspended in dry toluene (80 ml) and a solution of methyl 2,5-bis(cyanomethyl)benzoate (6QK2.632 g, 0.0123 mole) in dry THF-toluene mixture (40 ml, 6:4 V/V) was syringed into the flask. The reaction mixture was stirred for 15 minutes at room temperature; the mixture turned dark brown. Dimethyl carbonate (freshly distilled from molecular sieves, 10.3 ml, 0.123 mole) was syringed in and the reaction mixture was then heated to gentle boiling and refluxed for 4 hours. Then methanol was distilled off azeotropically with toluene. Distillation was continued until the temperature of the vapor reached 110°C (approximately one hour). After cooling the mixture to 0°C in an ice bath, an additional 95 ml of toluene was added. Cyanogen chloride (2.7 ml, 0.052 mole) was condensed inside a graduated cylinder fitted with a septum cap and cooled in an ice bath. Then the cyanogen chloride was distilled into the reaction flask through a stainless steel canulae, when the ice bath was removed. The mixture loosened up and was stirred at 0°C for 1 hour. The flask was then warmed to room temperature, gradually heated and kept at 80°C for 1 hour. After stirring overnight at room temperature, the reaction mixture was evaporated to dryness under reduced pressure. The crude solid mixture of j>l_ and sodium chloride was stirred with cold water. The brown cake obtained by filtration was dissolved in methylene chloride, treated with Norite and magnesium sulfate, and filtered using Celite. -73-Evaporation of the solvent gave 2.68 g (58.6%) of 61 as a green thick oil. IR v 2280 (nonconjugated C=N) , 1770 (saturated ester C=0) and max 1720 cm_1 (aryl ester C=0). 1H NMR(CDC£3) 63.95(s,3H,-0CH3), 4.00(s,6H,-(CN)2C-C02CH3), 8.02 (s,2H,aryl) and 8.40 ppm(s,1H,aryl). Mass spectrum: a) high resolution calculated for C.0rL „N.0,:380.0757 lO 1/ 4 0 amu, found 380.0779 m/e; b) low resolution m/e 59(68), 91(30), 92(21), 139(7), 175(11), 177(13), 193(15), 203(12), 247(14), 262(14), 277(100), 278(21), 303(24), 316(15), 336(25), 349(21) and 380(1). 2-Methoxycarbonyl-l,4-phenylenedimalononitrile (62) A solution of 61_ (450 mg, 1.18 mmole) in THF (20 ml) was added into a 100-ml flask. The flask was fitted with a magnetic stirring bar, septum cap and flushed with nitrogen. Sodium borohydride (2 mg) was added and the mixture was stirred at room temperature while a 5% aqueous solution of sodium carbonate (16 ml, 7.8 mmole) was syringed in dropwise (8 minutes). After stirring for 1 hour, the dark solution was acidified with 1.5 N hydrochloric acid, extracted with methylene chloride (2 x 50 ml), and washed with water (2 x 25 ml). The extracts were dried and the solvents removed under reduced pressure to yield 268 mg (86%) of compound 6^2_. The crude product was purified as follows. Crude product (120 mg) was dissolved in aqueous sodium carbonate (50 ml) and extracted with ethyl acetate (3 x 30 ml). The aqueous layer was acidified with 6 N hydrochloric acid and extracted with ethyl acetate (2 x 30 ml). After -74-drying, the solvent was evaporated under reduced pressure to give 72 mg of 62_ as a light green oil which partially crystallized on standing. After standing several days the color of this product darkened gradually. IR v 2260 (nonconjugated C=N) and 1725 cm"1 (aryl ester C=0). max 1H NMR (acetone-d6) 63.93(s,3H-OCH3), 6.0(bs,2H,(CN)2CH-); 7.87(s, 2H,aryl) and 8.2(s,lH,aryl). Mass spectrum m/e 59(28) 87(77) 85(52) 151(26) 152(26) 177(25), 178(37), 207(27), 208(22), 221(26), 222(26), 232(74), 233(100), 238(33), 249(33), 263(28), 264(93) and 265(22). 2-Methoxycarbonyl-7,7,8,8-tetracyanoquinodimethane (56) Ester 62 (0.38 g) was dissolved in 5% aqueous sodium carbonate (50 ml) and neutralized with 6 N hydrochloric acid. A slight excess of 0.1 M bromine solution in water was added gradually at room temperature (positive test with starch-iodine paper). The resulting green solid was filtered, washed with water (50 ml), and dried in vacuo over phos phorous pentoxide to yield 0.38 g. A small scale purification was achieved by sublimation of _56 (36 mg) at 190°C/0.01 mm overnight; the resulting green-yellow solid (1 mg, 3%) had a mp 220-224°C (decomposition). IR(KBr)v 2220 (conjugated C=N) and 1725 (unsaturated ester C=0). max UV-Vis(CH0CN) A 396 nm (36,800) and 375 nm (sh). 3 max 1H NMR(CDC£3) 6 4.4(s,3H,0CH3) and 7.4-8.6 ppm (m,3H,vinyl). Mass spectrum; high resolution calculated for C^^^^0^: 262.0491 amu found 262.0495 m/e; low resolution m/e 55(72), 57(100), 69(64), 71(71), 111(44), 123(33), 125(39), 149(89), 203(37), 204(42), 231(97), 232(58), 262(100) and 263(26). -75-3,6-Dimethylphthalic anhydride (70) To a well-stirred suspension of maleic anhydride (24.3 g, 0.247 mole) in dry ethyl ether (40 ml) was added slowly 2,5-dimethylfuran (68) (26.4 ml, 0.247 mole). After stirring at room temperature for 20 hours, during which the maleic anhydride dissolved and crystals of the adduct began to separate, the mixture was cooled in an ice-salt bath for 1 hour. The crystals of adduct 69_ (32.8 g, 69%) were collected and dried, 41 mp 57-59°C (lit mp 59-63°C). Compound 69_ was dehydrated directly by slow addition into concentrated sulfuric acid (322 ml) at -10°C The solution was stirred for 1 hour at -10°C before it was allowed to rise to 10°C. After pouring onto 1 kg of crushed ice the light brown precipitate was collected, washed with ice water and air-dried. After successive re-crystallizations from benzene and toluene, the desired' 3,6-dimethylphthalic 41 anhydride (70) was obtained in 24% yield (9.96 g), mp 137-138°C (lit mp 142-143°C). IR v 1845 and 1765 cm 1 (aryl acid anhydride C=0) and 1260 cm 1 max J J (C-0 anhydride). hi NMR(CDC£3) 62.60 (s,6H,-CH3) and 7.39 ppm (S,2H,aryl). Mass spectrum m/e 103(30), 104(68), 120(25), 132(55), 148(65) and 176(100). -76-3,6-Bis(bromomethyl)phthalic anhydride (71) Recrystallized NBS (37 g, 0.208 mole) was weighed into an oven-dried 2-H flask equipped with a magnetic stirring bar. The flask was fitted with a reflux condenser stoppered with a septum cap, and the whole system was flushed with nitrogen. A solution of 3,6-dimethyl-phthalic anhydride (70) (8.33 g, 0.047 mole) in dry carbon tetrachloride (710 ml) was added through the condenser followed by 0.1 g of benzoyl peroxide, and the mixture was heated at reflux. After 8 hours, heating was stopped and the reaction mixture was allowed to cool to room tem perature. Succinimide and unreacted NBS were filtered off and the filtrate was evaporated under reduced pressure. The resulting oil containing a mixture of mono- and dibromo derivatives was left to stand at room temperature. The desired dibromo compound 7_1_ crystallized on standing (4.1 g, 26%) and was further purified by recrystallization from a carbon tetrachloride-hexanes mixture (3:7 V/V). The white crystals obtained had a mp 121-125°C. IR v 1856 and 1777 cm"1 (C=0, aryl anhydride) and 1260 cm"1 max (C-0 anhydride). NMR(CDC£3) 64.86(s,4H,-CH^Br) and 7.82 ppm(s,2H,aryl). Mass Spectrum m/e 63(13), 74(12), 75(15), 76(16), 102(34), 117(13), . 129(9), 146(43), 173(24), 174(100), 175(40), 252(75), 253(87), 254(84), 255(87), 332(7), 334(12) and 336(6). Analysis calculated for C,~H-Br~0_:C 36.85, H 1.85, Br 49.02; found: C 36.52, H 1.87, Br 49.20. 6-(Bromomethyl)-7-methoxycarbonylphthalide (72) A 50-ml flask was equipped with a magnetic stirring bar, a heating -77-mantle, and a reflux condenser protected by a calcium chloride tube. The reaction vessel was charged with 2,5-bis(bromomethyl)phthalic anhydride (71) (0.5 g, 1.5 mmole), methanol (25 ml) and p-toluenesulfonic acid (10 mg). The mixture was refluxed for 2 hours and then cooled to room temperature. Methanol was evaporated under reduced pressure and the resulting solid was dissolved in ethyl ether (50 ml). The ethereal solution was extracted with a saturated solution of sodium bicarbonate (2 x 20 ml) and washed with saturated sodium chloride solution (25 ml); after drying, the solvent was removed by evaporation under reduced pressure to yield 0.216 g (50%) of crude compound _72_. A small sample was purified by preparative tic (chloroform); the band between 0.6 and 0.8 was removed and extracted with methylene chloride. After evaporation of the solvent white crystals of 72 were obtained, mp 125-128°C. IR v 1775 (lactone OO) and 1735 cm 1 (aryl ester C=0) . max J """H NMR(CDCi>3) 64.01(s,3H,-0CH3) , 4 . 55 (s, 2H,-CH2Br) , 5. 24 (s, 2H ,-CH2" 0-CO) and 7.55 ppm (dd,J=8.0,2H,aryl). Mass spectrum: a) high resolution calculated for C^H^BrO^: 283.9684 amu, found 283.9685 m/e; b) low resolution m/e 89(31), 90(18), 91(9), 117(17), 118(25), 129(12), 145(16), 146(24), 147(14), 149(15), 173(60), 174(75), 175(33), 176(18), 205(100), 206(23), 252(87), 253(28), 254(90), 255(28), 284(11) and 286(11). Dimethyl 3,6-dimethylphthalate (75) A 100-ml flask was equipped with a reflux condenser and a calcium chloride tube, a magnetic stirring bar, and a heating mantle. Dimethyl acetylenedicarboxylate (12.34 ml, 0.10 mole), trans ,trans-2,4-hexadiene -78-(13.88 ml, 0.12 mole) and dry benzene (25 ml) were added to the flask and the mixture was stirred and refluxed for 24 hours. The light yellow solution was cooled to room temperature and the solvent was removed under reduced pressure. The crude oil obtained was distilled in vacuo to afford the adduct 7_4; bp 101-103°C (0.4 mm). Its spectroscopic data were in agreement with structure 7_4_. This compound (13.336 g, 0.0595 mole) was treated directly with one equivalent of DDQ (13.97 g, 0.0615 mole) in benzene (100 ml) and the mixture was stirred and refluxed for 20 hours. The reaction mixture was then diluted with two volumes of petroleum ether, bp 30-60°C, and the precipitate of 2,3-dichloro-5,6-dicyano hydroquinone and unchanged DDQ was filtered off. The filtrate was passed through a short (16 cm x 5 cm) alumina column which was washed with additional petroleum ether until no more 7_5 could be eluted. The eluates were evaporated to give dimethyl 3,6-dimethylphthalate (10.5 g, 80% overall 46 1 yield), mp 74-77°C (lit. mp 72-75°C). Its ir, H nmr and mass spectra • , 46 were in agreement with those reported. Dimethyl 3,6-bis(bromomethyl)phthalate (76) This compound was prepared by a similar procedure as that employed in the preparation of methyl 2,5-bis(bromomethyl)phthalate (59). The only difference was that the refluxing time was 3 hours in the case of 75. The reagents used were: dimethyl 3,6-dimethylphthalate (75) (10.5 g, 0.047 mole) in 130 ml of dry methylene chloride, NBS (19.3 g, 0.108 mole), and AIBN (0.5 g). The crude solid obtained after work-up was dissolved in the minimum amount of boiling ethanol and the solution was left to stand overnight. On addition of methanol white crystals of dimethyl 6-bromo-methyl-3-dibromomethylphthalate (77) precipitated and were collected by -79-filtration (3.8 g), mp 118-126°C (lit.47 mp 123°C). Its ir, 1H nmr and mass spectra were consistent with structure 7_7_. The filtrate was evaporated under reduced pressure to give 11.4 g (64%) of crude dimethyl 3,6-bis(bromomethyl)phthalate (76). A small sample (187 mg) was purified by tic, using chloroform as eluent. The band at 0.6 was removed and eluted with methylene chloride. Evaporation of the solvent gave 115 mg (40%) of compound 7_6 as a colorless oil. IR v 1730 cm"1 (aryl ester C=0) max J 1H NMR(CDC£3) 63.77(s,6H,-0CH3),4.60(s,4H,-CH2Br) and 7.38 ppm (s,2H,aryl). Mass spectrum a) high resolution calculated for C ^R^^Sr^0^: 377.9102 amu, found 377.9095 m/e; b) low resolution m/e 77(31), 89(27), 90(21), 102(30), 103(26), 118(21), 129(22), 132(19), 145(67), 162(20), 174(19), 189(50), 205(86), 220(22), 267(40), 269(43), 299(64), 301(67), 347(53), 349(100), 350(44), 351(58), 378(5), 380(8) and 382(5). Dimethyl 3,6-bis(cyanomethyl)phthalate (73) The procedure used in this reaction was the same to that employed in the preparation of compound 60_. The following reagents were used: dimethyl 3,6-bis(bromomethyl)phthalate (76) (7.1 g, 0.0186 mole) in acetonitrile (91 ml), 30% aqueous sodium cyanide (8.8 g, 0.18 mole), and tri-n-octylamine (0.4 ml). After the work-up the crude mixture was purified using a 4 cm x 26 cm column of silica gel eluted with carbon tetrachloride-ethyl ether (95:5 V/V). The band corresponding to a spot at R^ 0.65 on analytical tic was collected and, after evaporation of the solvents, gave 7_3 (2.2 g, 44%), mp 66-68°C. -80-IR Vx 2290 (CEN) and 1730 cm 1 (ary! ester C=0) max 1 "H NMR(CDC2.3) 63 .88 (s, 6H ,-0CH3) , 3.91 (s , 4H,-CH2CN) and 7.58 ppm (s,2H,aryl). Mass Spectrum m/e 127(7), 128(8), 153(7), 154(10), 182(18), 213(4), 214(5), 225(5), 240(100), 241(59) and 272(1.2). Analysis calculated for C^H^N^: C 61.76, H 4.44, Y. 10.29; found: c 61.89, H 4.34, N 10.17. 2,3-Dimethyl-l,4-cyclohexanedione (87) This compound was prepared by a similar procedure to that employed in the preparation of 2-methyl-l,4-cyclohexanedione (52). The following reagents were used: 2,3-dimethylhydroquinone (85) (1.09 g, 0.0079 mole), rhodium catalyst (0.41 g) and methanol-water mixture (22 ml, 10:1 V/V), which gave 7.08 g (95%) of 2,3-dimethyl-l,4-cyclohexanediol (86), identi fied by its ir and nmr spectra. Compound 8_6 (0.94 g, 6.5 mmole) was not purified but was oxidized directly with three equivalents of chromium trioxide solution (7.22 ml) in chloroform (25 ml) at 5°C. After the addition of chromium trioxide (10 minutes), the reaction mixture was stirred for 2 hours, and worked up as described previously for 51, to give 0.684 g (75%) of 2,3-dimethyl-l,4-cyclohexanedione (87). IR v 1710 cm"1 (C=0). max """H NMR (CDC£3) 61.08(d,J=7,6H,-CH3), 2.70(s,4H,-CH2-), and 2.2-3.0 ppm (m,2H,-CH-). Mass spectrum m/e 56(61), 69(21), 73(24), 125(87) and 140(100). -81-2,3-Dimethy1-1,4-bis(dicyanomethylene)cyclohexane (88) To 2,3-dimethyl-l,4-cyclohexanedione (87) (335 mg, 2.39 mmole) in a 50-ml flask, was added a mixture of malononitrile (347 mg, 5.25 mmole) and a catalytic amount of 8-alanine (30 mg) in water (10 ml). The reaction mixture was heated at reflux for 3.5 hours. The product was isolated by filtration and was washed with water (40 ml). The yield of crude 2,3-dimethyl-l,4-bis(dicyanomethylene)cyclohexane (88) was 379 mg (67%). A small sample, recrystallized from acetonitrile, melted at 186-189°C. IR(KBr) v 2250 (C^N) and 1600 cm"1 (C=C). max 1H NMR(acetone-d6) 61.30(d,J=7,6H,-CH3), 3.10(s,4H,-CH2") and 3.33 ppm(q,2H,-CH-). Mass spectrum m/e 91(6), 92(6), 118(12), 119(18), 131(16), 145(100), 146(12), 194(18), 207(6), 208(9), 209(6), 219(8), 221(28), 222(5), 234(8), 235(6) and 236(14). Analysis calculated for C^H^N^ C 71.17, H 5.12, N 23.71; found: C 71.27, H 5.16, N 23.96. 5,6-Dimethylrl,4-bis(dicyanomethylene)-2-cyclohexene (89) An oven-dried, 50-ml flask containing compound 8£5_ (304 mg, 1.29 mmol), acetonitrile (15 ml) and bromine (0.073 ml, 1.42 mmol) was stoppered with a septum cap and fitted with a magnetic stirring bar. It was flushed with nitrogen while cooled to 0-5°C in an ice bath. Pyridine (0.42 ml, 5.16 mmol) was slowly syringed into the reaction flask over a period of 15 minutes. After completion of the addition, the mixture was stirred at 0°C for 30 minutes and then for two more hours at room temperature. Cold -82-water was added and the mixture was stirred for about 15 minutes. The resultant solid product was removed by filtration, washed with cold water (50 ml), and dried over phosphorous pentoxide in vacuo (258 mg, 86%). An analytical sample was recrystallized from acetonitrile, mp 203-205°C. Ir v 2260 (C=N) and 1560 cm"1 (conjugated C=C). max 1H NMR(CD3CN) 61.16(d,J=7,6H,-CH3), 3.17(m,2H,-CH-) and 7.06 ppm (s,2H,vinyl). Mass Spectrum m/e 138(20), 139(20), 140(20), 142(33), 156(16), 165(26), 169(33), 192(22), 206(22), 207(32), 219(100), 233(33), 234(90) and 235(18). Analvsis calculated for C, .H., _N. :C 71.78, H 4.30, N 23.92; found: 14 10 4 C 71.96, H 4.40, N 24.02. 2,3-Dimethyl-7,7,8,8-tetracyanoquinodimethane (83) 5,6-Dimethyl-l,4-bis(dicyanomethylene)-2-cyclohexene (89) (64 mg, 0.27 mmole), 10% palladium on powdered charcoal (22 mg) and sulfur (32 mg) were ground in a mortar. The homogeneous mixture was placed inside a small sublimation apparatus and heated to 180°C for 3 hours. Some of the sulfur deposited on the cold finger, and the black residue was extracted with boiling acetonitrile. After filtration, the acetonitrile was evaporated under reduced pressure. Purification was achieved by tic: 50 mg of the crude material was chromatographed on a 5 cm x 20 cm silica coated plate, using methylene chloride as eluent. The two main components isolated from this chromatography were, in order of elution, starting material and compound 83_ (20 mg, 32%). 2,3-Dimethyl-7,7,8,8-tetracyano-quinodimethane (83) was identified by its nmr and mass spectra. -83-XH NMR(CD CN) 62.35(s,6H,-CH3) and 7.43 ppm (s,2H,vinyl). Mass spectrum m/e 81(100), 100(61), 120(28), 132(43), 170(26), 182(78), 220(14) and 232(16). 3,5-Bis(bromomethyl)toluene (97) A 2-1 three neck flask equipped with two efficient condensers and a strong mechanical stirrer was charged with dry carbon tetrachloride (400 ml), distilled mesitylene (138 ml, 1 mole), NBS (recrystallized from water, 356 g, 2 mole) and benzoyl peroxide (3 g). The mixture was stirred and heated at reflux. After 15-30 minutes of reflux, a vigorous reaction took place and the heating mantle was then replaced by an acetone-dry ice bath. Vigorous reflux continued for about 15 minutes, after which time the reaction was complete. The reaction mixture was then cooled and filtered and the filtrate was evaporated under reduced pressure. The product 96_ (128 g, 46%) was obtained by direct recrystal lization of the crude reaction mixture from ethanol, mp 45-50°C (lit.^ mp 60°C )• Its spectroscopic data were in agreement with the proposed structure. 3,5-Bis(mercaptomethyl)toluene (102) In a 250-ml flask fitted with an efficient reflux condenser and a heating mantle, were placed thiourea (10.9 g, 0.143 mole) and 95% ethanol (60 ml). The mixture was brought to the reflux temperature and 3,5-bis(bromomethyl)toluene (97) (19.8 g, 0.071 mole) was added in batches (ca 10 minutes), and the reflux continued for 2 hours. After cooling, the white precipitate was filtered and dried. The crude salt 103 so obtained was mixed with potassium hydroxide (51 g, 0.92 mole) in water (110 ml) and boiled under reflux for 6 hours. The reaction mixture was -84-then cooled in an ice bath and acidified with 9 M sulfuric acid; it was extracted with ethyl ether (3 x 100 ml), and washed with water (100 ml). After drying, the solvent was removed under reduced pressure to give 12.47 g (95%) of 3,5-bis(mercaptomethyl)toluene (102) as a colorless oil. It was not purified but was used directly in the next reaction. xll NMR(CDC£3) 61.72(t,J=8,2H,-SH), 2.31(s,3H,-CH3), 3.64(d,J=8,4H, -CH2-), and 6.97 ppm (s,3H,aryl). 6,15-Dimethyl-2,ll-dithia[3.3]metacyclophane. (98)* A 5-1 three neck flask was fitted with nitrogen inlet and outlet syringe needles, 3 in magnetic stirring bar and a Hershberg dropping funnel. The reaction vessel was charged with potassium hydroxide (9 g, 0.16 mole) dissolved in ethanol-water (9:1 V/V, 2£) and was stirred under argon at room temperature. A solution of 3,5-bis(bromomethyl)toluene (97) (19.02 g, 0.0684 mole) and 3,5-bis(mercaptomethyl)toluene (102)(12.47 g, 0.0677 mole) in benzene (1.6£) was placed in the funnel; it was then added dropwise over a period of seven days. After the addition was complete, the reaction mixture was filtered; the filtrate was evaporated under reduced pressure and the crude product was purified by column chromatography, using silica-gel (10 cm x 20 cm) and a mixture of benzene and petroleum ether, bp 30-60°C, (2.5:7.5 V/V) to give compound 98_ in 86% yield (17.48 g). An analytical sample, prepared by recrystallization 53 from cyclohexane, melted at 101.5-103.5°C (lit. mp 102-102.5C). IR v 1601 cm"*1 (C=C aromatic), max 1H NMR(CDC£3) 62.13(s,6H,-CH3), 3.62(s,8H, -CH2), 6.58(bs,2H, internal ArH) and 6.66(bs,4H,ArH). Mass spectrum m/e 119(68), 120(80), 149(27), 150(42), 151(52) and -85-300(100). Analysis calculated for C18H2uS2: C 71.95, H 6.71, S 21.34; found: C 72.15, H 6.74, S 21.22. 5,13-Dimethyl[2.2]metacyclophane (96)* An oven-dried, 1-1 flask in charged with 6,15-dimethyl-2,11-dithia [3.3]metacyclophane (98) (17.48 g, 0.058 mole) dissolved in dry THF (290 ml), equipped with a magnetic stirring bar and stoppered with a septum cap. The flask was then fitted with a nitrogen inlet and outlet needle, flushed with nitrogen and cooled to -20°C (carbon tetrachloride -dry ice). n-Butyllithium, as a 1.55 M solution in hexane, (94 ml, 0.145 mole) was added dropwise to the reaction which was stirred for-a further 40 minutes before the addition of methyl iodide (18 ml, 0.29 mole). The mixture turned yellow and a yellow precipitate was formed. The reaction mixture was stirred for an additional 30 minutes at -20°C and then warmed to room temperature. It was quenched with cold water (200 ml), extracted with ethyl ether (2 x 100 ml) and washed with water (100 ml). The aqueous layer was further extracted with methylene chloride (100 ml). The combined organic extracts were washed with saturated sodium chloride solution (100 ml), dried, and the solvents were evaporated under reduced pressure to yield 100 (17.72 g) , which was identified by its ir and nmr spectra. Compound 100 was dissolved in dry THF-ether (200 ml, 6:4 V/V) and added to 250 ml of a blue solution of liquid ammonia containing chips of lithium. The reaction vessel was kept at -78°C during the addition, and lithium chips were added whenever the blue color faded. After the addition was over, the reaction mixture was stirred for 1.5 hours before ammonium chloride was added to discharge the blue color. The flask was -86-then warmed to room temperature and ammonia was allowed to evaporate overnight. Water (100 ml) was added to the resulting mixture which was then acidified with 6 N hydrochloric acid. It was extracted with ethyl ether (2 x 100 ml), washed with water (100 ml), and saturated sodium chloride solution (100 ml). After drying, the solvents were evaporated under reduced pressure. The resultant light yellow solid was purified by filtration through silica gel using petroleum ether, bp 30-60°C, subsequent recrystallization from ethanol afforded 96_ (3.8 g, 30%), as white crystals, mp 145-146°C (lit.53 mp 145-146°C). IR v 1600 cm"1 (aromatic C=C). max """H NMR(CDC£3) 62. 02 (d , J=8 ,4H,axial CH2), 2 . 93 (d ,J=8, 4H .equatorial CH2), 2.23(s,6H,-CH3), 4.03(s,2H,aryl) and 6.67 ppm(s,4H,aryl). Mass spectrum m/e 118(21), 205(18), 206(28), 208(23), 219(20), 221(88), 222(24), 235(18), 236(100) and 237(26). Analysis calculated for C^H^: C 91.47, B 8.53; Found: C 91.30, H 8.52. * Due to the possible toxicity of an intermediate or unknown by-products, this reaction should be totally carried out in a well vented hood. -87-5,13-Bis(bromethyl)[2.2]metacyclophane (95) An oven-dried, 200-ml flask was equipped with a magnetic stirring bar, a reflux condenser fitted with a nitrogen inlet, and a heating-mantle. The reaction vessel was charged with 5,13-dimethyl[2.2] metacyclophane (96) (580 mg, 2.46 mmole), NBS (recrystallized from water, 920 mg, 5.16 mmole), dry carbon tetrachloride (83 ml) and benzoyl peroxide (40 mg). The reaction mixture was then stirred and heated at reflux for 1.5 hours under nitrogen. Heating was discontinued and the mixture was stirred at room temperature overnight. It was then filtered and the solvent from the filtrate was removed by evaporation under reduced pressure. The residue was washed with petroleum ether, bp 30-60°C, and dried in vacuo to give 686 mg (70%) of compound 95_. A small sample was recrystallized from acetonitrile, mp 209-210°C (lit.53 mp 205-206°C) IR v 1600 cm"1 (C=C aromatic) and 1780, 1755 cm"1 (1,3,5 trisub-max stituted benzene). 1H NMR(CDC£3) 2.14(d,J=8Hz,4H,axial-CH2-), 3.13(d,J=8Hz,4H, equatorial-CH2-), 4.25(s,2H,aromatic), 4.56(s,4H,-CH2Br), 7.10(s,4H,aromatic). Mass spectrum m/e 101(14), 108(12), 109(11), 115(11), 116(25), 202(15), 203(18), 204(11), 217(11), 219(29), 220(15), 233(42), 234(69), 235(18), 299(11), 301(11), 313(100), 314(26), 315(95), 316(24), 392(24), 394(43) and 396(23). 5,13-Bis(cyanomethyl)[2.2]metacyclophane (94) A 200-ml flask was equipped with a magnetic stirring bar, a reflux condenser and a heating mantle. The reaction vessel was charged with 5,13-bis(bromomethyl)[2.2]metacyclophane (95)(264 mg, 0.67 mmole), sodium cyanide (197 mg, 4.02 mmole) in water (1 ml), THF (20 ml), and tri-n-octylamine(one drop). The mixture was stirred and heated at reflux -88-overnight. It was then allowed to cool to room temperature, poured into cold water (25 ml) and extracted with methylene chloride (3 x 30 ml). The organic layer was washed with water (30 ml), saturated sodium chloride (30 ml) and dried. After evaporation of the solvents under reduced pressure a light yellow solid was obtained, this was then recrystallized from methanol (with cooling in a carbon tetrachloride-dry ice bath) to give compound 9Jt_ (135 mg, 70%). An analytical sample, mp 190-195°C, was prepared by sublimation at 180°C/0.2 mm. IR v 2300 (C=N), 1600 (C=C aromatic), max 1H NMR(CDC£3) 62.14(d,J=8,4H,axial-CK^-), 3.13(d,J=8,4H,equatorial-CH2"), 3.80(s,4H,-CH2CN), 4.24(s,2H,aromatic) and 7.06 ppm(s,4H,aromatic). Mass spectrum m/e 203(12), 205(19), 206(100), 207(23), 245(13), 246(47), 247(11), 254(12), 286(59) and 287(13). Analysis calculated for C„nH10N0:C 83.88, H 6.34, N 9.78; found: C 83.49, H 6.30, N 9.62. 2,7-Bis(cyanomethyl)-4,5,9,10-tetrahydropyrene (93) 5,13-Bis(cyanomethyl)[2.2]metacyclophane (94)(0.28 g, 1 mmole) was dissolved in dry chloroform (passed through alumina, 15 ml) and added to an oven-dried 50 ml flask. The reaction vessel was then charged with a catalytic amount of iron powder (350 mesh), equipped with a magnetic stirring bar and flushed with nitrogen. The mixture was stirred at 0°C in an ice bath for 5 minutes before a 10% solution of bromine in chloroform (1.5 ml, 2.9 mmol) was added dropwise with a syringe; then, the reaction mixture was stirred vigorously for 1.5 hours at 0°C. The reaction was quenched when saturated solution of sodium thiosulfate (15 ml) was added followed by stirring at room temperature for 15 minutes. The color of the -89-solution went from red-orange to light green. The reaction mixture was then extracted with methylene chloride (2 x 25 ml), washed with water (2 x 20 ml) and dried. The solvents were evaporated under reduced pressure to give the crude compound 93_, which was purified by column chromatography, using chloroform as eluent, to give 199 mg (70%) of compound J93, mp 230-235°C. IR v 2300(CEN) and 1620 cm"1 (C=C aromatic), max 1H NMR(CDC£3): 2.84(s,8H,-CH2~), 3.66(s,4H,-CH2CN) and 6.95 ppm (s ,4H,aromatic). Mass spectrum: a) high resolution calculated for C„.H., ,N0:284.1313 zU lo z amu, found 284.1317 m/e; b) low resolution m/e 206(100), 246(40), 258(28), 284(97) and 285(23). 2.7-Bis(methyl a,a-dicyanoacetate)-4,5,9,10-tetrahydropyrene (92) This compound was prepared by a similar procedure as that employed in the preparation of dimethyl-a,a,a",a^-tetracyano-2-methoxycarbonyl-1,4-phenylenediacetate(61). The only difference was that the refluxing time was 1.5 hours in this case as opposed to 4 hours in the case of the monoester 6J). The reagents used were: 2 ,7-bis (cyanomethyl)-4 ,5 ,9,10-tetrahydropyrene (93) (519 mg, 1.83 mmole), fresh sodium methoxide (5.13 mmole from 118 mg of sodium) dry toluene (30 ml + 25 ml) dry THF (10 ml), distilled dimethyl carbonate (2 ml, 24 mmol) and cyanogen chloride (0.47 ml, 9.09 mmole). Work-up: solvents and volatile components of the reaction mixture were distilled off under reduced pressure and the resulting light brown solid was stirred withl.5 N hydrochloric acid (40 ml) and extracted with ethyl acetate (3 x 50 ml). After drying, the solvent was evaporated under reduced pressure to yield compound _92_ as a dark brown solid, mp 110-115°C (662 mg). -90-IR v 2265 and 2210 cm 1 (C=N) and 1770 (saturated ester C=0) max 1H NMR(CDC£3) 62.94(s,8H,-CH2~), 3.91(s,6H,-0CH3) and 7.33 ppm (s,4H,aryl). Mass spectrum: a) high resolution calculated for C^^-^s^^^1 450.1328 amu, found 450.1335 m/e; b) low resolution m/e 59(100), 69(57), 91(36), 92(29), 202(25), 203(16), 206(20), 217(16), 219(15), 258(46), 282(17), 317(39), 341(31), 355(17), 400(15) and 450(5). -91-2,7-Bis(dicyanomethyl)-4,5,9,10-tetrahydropyrene (91) A solution of 92 (250 mg, 0.55 mmole) in THF (13 ml) was added into a 50-ml flask. The flask was fitted with a magnetic stirring bar, septum cap and flushed with nitrogen. Sodium borohydride (^2 mg) was added and the mixture was stirred at room temperature while a 5% aqueous solution of sodium carbonate (8 ml, 4.0 mmol) was syringed in dropwise (5 minutes). After stirring for 1.5 hours, the dark solution was acidified with 6 N hydrochloric acid, extracted with ethyl acetate (3x50 ml), and washed with water (2x25 ml). The extracts were dried and the solvents removed under reduced pressure to yield 162 mg of 2,7-bis(dicyanomethyl)-4,5,9,10-tetrahydropyrene (91), mp 70-80°C. The crude product was identified by its ir and nmr spectra, and was used directly in the reaction that follows. IR v 2260 and 2215 cm"1 (C=N) max 1H NMR (CDC£3) 62.93 (s,8H,-CH2"), 5.00 (s,2H,-CH(CN)2) and 7.14 ppm (s,4H,aryl). 11,11,12,12-Tetracyano-4,5,9,10-tetrahydropyreno-2,7-quinodimethane (90) Compound _9_1 (116 mg, 0.34 mmol) was dissolved in dry methylene chloride (25 ml) and placed in an oven-dried 50-ml flask. The flask was fitted with a magnetic stirring bar, septum cap and flushed with nitrogen. The solution was cooled to -78°C (dry ice-acetone bath) and NBS (26 mg, 0.15 mmole) was added followed by triethylamine (2 drops). The solution turned dark red instantaneously and it was stirred for 5 minutes at -78°C before it was warmed up to 0°C. Hydrochloric acid was then added to acidify the reaction mixture and this was diluted with methylene chloride (25 ml). The mixture was washed with water (1x10 ml) and the solvent was dried and evaporated under reduced pressure to yield a dark brown solid (99 mg). -92-The crude product was filtered through a short silica gel column (eluent: methylene chloride); evaporation of the solvent gave 24 mg of compound 90. This compound was further purified by recrystallization from benzene to give compound 90^ (19 mg) as green microcrystals, mp 206-208°C (decomposition). IR(KBr) v 2265 and 2225 cm"1 (C=N) max 1H NMR (CDC£3) 62.94 (s,8H,-CH2") and 7.20 ppm (bs,4H,vinyl). Mass spectrum a) high resolution calculated for ^22^12^: 332.1062 amu, found 332.1071 m/e; b) low resolution m/e 99(23), 202(33), 203(25), 204(37), 332(35), 334(100) and 335(25). -93-BIBLIOGRAPHY 1. J. Ferraris, D. 0. Cowan, V. Walatka Jr., J. H. Perlstein, J. Am. Chem. Soc, 95, 948 (1973). 2. L. B. Coleman, M. J. Cohen, D. J. Sandman, F. G. Yamagishi, A. F. Garito, A. J. Heeger, Solid State Commun., 12, 1125 (1973). 3. A. N. Bloch, D. 0. Cowan, K. Bechgaard, T. 0. Poehler, Phys. Rev. Lett., 34, 741 (1975). 4. F. Gutmann, L. E. Lyons, "Organic Semiconductors", Wiley, N.Y., 1967, Chapter 8. 5. S. D. Levina, Dokl. Akad. Nauk SSSR, 145, 602 (1962). 6. a) M. E. Peover, J. Chem. Soc, 4540 (1962). b) M. E. Peover, Nature, 193, 475 (1962). 7. D. S. Acker, W. R. Hertler, J. Am. Chem. Soc, 84, 3370 (1962). 8. A. J. Fatiadi, Synthesis, 133 (1976). 9. L. R. Melby, R. J. Harder, W. R. Hertler, W. Mahler, R. E. Benson, W. E. Mochel, J. Am. Chem. Soc, 84, 3374 (1962). 10. D. Acker, D. Blomstrom, U.S. Pat. 3,162,641 (1964); Chem.Abst.60, 11918(1964) 11. 0. H. Le Blanc Jr., J. Chem. Phys., 37, 916 (1962). 12. L. Melby, Can. J. Chem., 43, 1448 (1965). 13. a) F. Wudl, D. Wubschall, E. J. Hufnagel, J. Am. Chem. Soc. , _94, 670 (1972). b) S. J. la Placa, P. W. R. Coofield, R. Thomas, B. A. Scott, Solid State Commun., 17, 635 (1975). c) C. K. Johnson, C. R. Watson Jr., J. Chem. Phys., 64, 2271 (1976). 14. M. Narita, C. U. Pittman, Synthesis, 489 (1976). -94-15. D. Cowan, P. Shu, C. Hu, W. Krug, T. Carruthers, T. Poehler, A. Bloch in "Chemistry and Physics of One-Dimensional Metals" edited by H. J. Keller, Plenum, N.Y., 1977, p.25. 16. L. R. Melby, H. D. Hartzler, W. A. Sheppard, J. Org. Chem., 39, 2456 (1974). 17. A. J. Heeger, A. F. Garito in "Low-Dimensional Cooperative Phenomena" edited by H. J. Keller, Plenum, N.Y., 1975, p.89. 18. E. M. Engler, Chemtech. b_, 274 (1976) and references cited therein. 19. A.J.Berlinsky, L. Weiler, Solid State Commun., 19, 1165 (1976). 20. A. J. Berlinsky, J. F. Carolan, L. Weiler, Solid State Commun., 15, 795 (1974). 21. R. E. Peierls, "Quantum Theory of Solids", Oxford University Press, London, 1955, p.108. 22. A. F. Garito, A. J. Heeger, Acc. Chem. Res., 7^ 232 (1974). 23. J. H. Perlstein, Angew. Chem. Int. Ed. Engl., 16, 519 (1977). 24. a) A. F. Garito, A. J. Heeger, Acc. Chem. Res., ]_, 232 (1974). b) A. N. Bloch in "Energy and Charge Transfer in Organic Semiconductors" edited by K. Masuda and M. Silver, Plenum Press, N.Y., 1974, p.159. 25. J. B. Torrance, Acc. Chem. Res,, 12, 79 (1979). 26. J. Diekmann, W. R. Hertler, R. E. Benson, J. Org. Chem., 28, 2719 (1963). 27. D. J. Sandman, A. F. Garito, J. Org. Chem., 39, 1165 (1974). 28. P. Berger, D. Dahm, G. Johnson, M. Miles, J. Wilson, Phys. Rev. B, 12, 4085 (1975). 29. a) A. W. Addison, N. S. Dalai, Y. Hoyano, S. Huizinga, L. Weiler, Can. J. Chem., 55, 4191 (1977). b) W. R. Hertler, U.S. Patent No. 3 153 658 (1964); Chem. Abstr. _62, 4145 (1965). •95-30. F. Wudl, M. L. Kaplan, B. K. Teo, J. Marshall, J. Org. Chem., 42, 1666 (1977). 31. a) R. C. Wheland, E. I. Martin, J. Org. Chem., 40, 3101 (1975). b) R. C. Wheland, J. L. Gillson, J. Am. Chem. Soc, 98, 3916 (1976). c) R. C. Wheland, J. Am. Chem. Soc, 98_, 3926 (1976). 32. E. M. Engler, R. A. Craven, Y. Tomkiewicz, B. A. Scott, K. Bechgaard, J. R. Andersen, Chem. Commun., 337 (1976). 33. W. R. Hertler, J. Org. Chem., 41, 1412 (1976). 34. P. Eaton, R. Mueller, G. Carlson, D. Cullison, G. Cooper, T. Chou, E. Krebs, J. Am. Chem. Soc. , 9£, 2751 (1977). 35. a) V. Bhalerao, H. Rapoport, J. Am. Chem. Soc, 93, 4835 (1971). b) J. P. Schaefer, B. Horvath, H. P. Klein, J. Org. Chem., 33, 2647 (1968). 36. D. D. Reynolds, K. R. Durham, US Patent #2789971 (1957); Chem.Abst.51,14814c(195: 37. H. Falk, K. Schlogl, Monatshefte fur Chemie, 49, 578 (1968). 38. F. Vogtle, W. Bieber, Angew. Chem. Int. Ed. Engl., 16, 175 (1977). 39. W. P. Reeves, M. R. White, Syn. Commun., 193 (1976). 40. F. L. Cook, C. W. Bowers, C. L. Liotta, J. Org. Chem., 39, 3416 (1974). 41. M. S. Newman, B. T. Lord, J. Am. Chem. Soc, 66, 734 (1944). 42. M. S. Newman, V. Lee, J. Org. Chem., 4_2, 1478 (1977). 43. a) J. March, "Advanced Organic Chemistry: Reactions, Mechanisms, and Structure." McGraw-Hill, N.Y., 1968, p.535. b) K. Ziegler, A. Spath, E. Schaaf, W. Schumann, E. Winkelmann, Anna1en, 551, 80 (1942). 44. K. Rorig, J. D. Johnston, R. W. Hamilton, T. J. Telinski, Org. Syn., Coll. Vol. IV, 576. 45. A. C. Cope, E. C. Herrick, Org. Syn., Coll. Vol. IV, 304. -96-46. H.'J. Kuhn, K. Gollnick, Chem. Ber., 692 (1973). 47. D. Gundermann, D. Roker, J. Liebigs Ann.Chem., 140 (1976). 48. G. Simchen, Angew. Chem. Int. Ed. Engl., 5_, 663 (1966). 49. J. R. Andersen, C. S. Jacobsen, G. Rindorf, H. Soling, K. Bechgaard, Chem. Commun., 883 (1975). 50. K. Yamakawa, private communication. 51. H. J. Teuber, Chem. Ber., 86, 1036 (1953). 52. M. Crawford, V. R. Supanekar. J. Chem. Soc, 2380 (1964). 53. T. Umemoto, S. Satani, Y. Sakata, S. Misumi, Tet. Lett., 3159 (1975). 54. T. Sato, S. Akabori, S. Muto, K. Hata, Tet., 24, 5557 (1968). 55. H. Irgantinger, R. G.Kirrstetter, C. Krieger, H. Rodewald, H. A. Staab, Tet. Lett., 1425 (1978). 56. F. Vogtle, P. Newmann, Synthesis, 85 (1973). 57. V. Boekelheide, D. Reingold, M. Tuttle, Chem. Commun., 406 (1973). 58. R. H. Mitchell, R. Carruthers, J. Zwinkels, Tet. Lett., 2585 (1976). 59. R. H. Mitchell, R. Carruthers, Can. J. Chem., 52, 3054 (1974). 60. A. J. Speziale. Org. Synt., Coll. Vol. IV, 401. 61. R. H. Mitchell, T. Otsubo, V. Boekelheide, Tet. Lett., 219 (1975). 62. a) F. Vogtle, P. Neumann, Angew. Chem. Int. Ed. Engl., 11, 73 (1972). b) S. Akabori, T. Sato, K. Hata., J. Org. Chem., 33, 3277 (1968). 63. E. Clar, B. A. McAndrew, M. Zander, Tet, 23, 985 (1967). 64. T. Sato, M. Wakabayashi, Y. Okamura, T. Amada, K. Hato, Bull. Chem. Soc. Japan, 40, 2363 (1967). 65. W. Ried, F. J. Konigstein, Chem. Ber., 92, 2532 (1959). 97a SPECTRAL INDEX soo 300 -112--113-HS SH 102 300 I 200 100 • m 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0060874/manifest

Comment

Related Items