5 Arynes Carbenes Nitrenes and Related Species By D. W. KNIGHT Department of Chemistry University of Nottingham Nottingham NG7 2RD 1 General Rearrangement reactions of carbenes and nitrenes have been summarized in a recent monograph.' A book describing various vacuum pyrolytic methods has also appeared2 which will be of interest to those working with thermally generated carbenes and nitrenes. 2 Arynes Several reports this year have emphasized the value of benzyne intermediates in cycloaddition reactions. Thus the anthranilic acid (1)can be converted into (2) in 86% yield via an intramolecular Diels-Alder reaction of the intermediate benzyne generated in the usual way.3 The reaction also works very well (90"h yield!) when the benzyne is generated from a 1,2-dibromobenzene and n-BuLi.The extra- ordinarily high yields obtained in these reactions serve to emphasize the advantages of many intramolecular cycloadditions in comparison to similar intermolecular procedures and suggest that such reactions with arynes as dienophiles will be of considerable value in synthesis. In this case the method has been used to prepare the natural 0-naphthoquinone Mansonone E. An apparently related reaction is Me Me I the conversion of amide (3) into (5) (in 25% yield) on treatment with n-BuLi; the aryne (4) seems to be a plausible intermediate.4 W. M. Jones 'Rearrangements of Carbenes and Nitrenes' in 'Rearrangements in Ground and Excited States' ed. P. DeMayo Academic Press New York 1980 Vol. 1. R. F. C. Brown 'Pyrolytic Methods in Organic Chemistry.Application of Flow and Flash Vacuum Pyrolytic Techniques'. Academic Press New York 1980. W. M. Best and D. Wege Tetrahedron Lett. 1981,22,4877. W. J. Houlihan Y. Uike and V. A. Parrino J. Org. Chem. 1981,46.4515. 97 D. W.Knight The use of 1,2,4,5-tetrabromobenzenesas dibenzyne equivalents has been repor- ted in full.’ The reaction which is almost certainly a stepwise process can be used to prepare compounds such as (6)by treatment of the halobenzene with n-BuLi and excess anthracene at room temperature. Analogous reactions can be performed using 2,3,6,7-tetrabromonaphthalenes. (6) Another full report concerns the preparation of anthraquinones in ‘one-pot’ by the addition of phthalide anions to arynes generated in situ from halobenzenes.6 In related studies the sulphonylphthalide anion (7)has been found to react regioselectively with benzyne (8) to give intermediate (9),which readily collapses to give anthraquinone (lo).’ Unfortunately the addition is not so regioselective with other benzynes.0 0 OMe + (7) QyJ \ \ OMe I Me0 I ‘S02Ph Me0 0 (9) H. Hart S. Shamouilian and Y. Takehira J. Org. Chem. 1981,46,4427; H. Hart and S. Shamouilian ibid. p. 4874. ‘ D. J. Dodsworth M.-P. Colcagno E. V. Ehrmann B. Devadas and P. G. Sammes J. Chem. Soc. Perkin Trans. 1 1981 2120. ’ R. A. Russell and R. N. Warrener J. Chem. Soc. Chem. Commun. 1981 108. Arynes Carbenes Nitrenes and Related Species 99 Convincing arguments have been presented' in full that pyrolysis (at ca.320 "C) of cis-1,2-dialky1-1,2-diethynylolefins involves the intermediacy of a 1,4-dehy- drobenzene as a discreet biradical most likely in the singlet state and with an estimated lifetime of between lo-' and s at 200 "C. Details of the synthesis and isolation of the di- and tetra-dehydrocyclo-octenes (11)and (12) have been published.' Although potentially anti-aromatic the com- pounds seem to be planar and are stable enough for some reactions to be carried out on them. The spectral characteristics and some chemistry of the acenaphthyne (13) have also been recorded." The intermediacy of 2,3-thiophyne and 2,3-benzothiophyne has been evoked to account for some of the products obtained by flow vacuum pyrolysis of the corres- ponding anhydrides in the presence of various dienes." 3 Nitrenes Photoelectronic spectra of methyl azide at 850 K show the main decomposition product to be methanimine and not methylnitrene in accord with the results of theoretical calculations.'2 The recently reported pyrolytic conversion of aryl azidoformates into benzoxazolones does not seem to involve a spiro-intermediate as a para-substituented azide (14) leads only to isomer (15).13If the two ortho-positions in (14) are blocked a cyclohexa-2,4-dienone is initially formed which subsequently dimerizes.Thermolysis or photolysis of phenyl azide in acetic acid gives azepin-2- one (presumably via l-azacyclohepta-1,2,4,6-tetraene)together with various sub- stituted anilines whose formation can most reasonably be rationalized by assuming the intermediacy of (16) an adduct of phenyl nitrene and acetic acid.I4 HN -0Ac D0&0+ 0 fJo>o -_* R ' N3 R' H (14) (15) (16) T.P. Lockhart P. B. Comita and R. G. Bergman J. Am. Chem. SOC., 1981,103,4082; T. P. Lockhart and R. G. Bergman ibid. p. 4091. H. N. C. Wong and F. Sondheimer Tetrahedron 1981 37 Supp. 1,99. lo 0.L. Chapman J. Gano P. R. West M. Regitz and G. Mass I. Am. Chem. Soc. 1981 103 7033. M. G. Reinecke J. G. Newsom and L.-J. Chen J. Am. Chem. SOC.,1981,103 2760. '* H. Bock R. Dammel and L. Horner Chem. Ber. 1981,114,220. l3 0. Meth-Cohn and S. Rhouati J. Chem. SOC.,Chem. Commun. 1981 241; see also R. N. Carde P. C. Hayes G. Jones and C. J. Cliff J. Chem.Soc. Perkin Trans. 1 1981 1132. 14 H. Takeuchi and K. Koyama J. Chem. SOC., Chem. Commun. 1981,202. 100 D. W.Knight An interesting case of what appears to be stereochemical control in the decompo- sition of a heteroaromatic nitrene has been rep~rted.'~ Thus 5-azido-2-fury1 ketones decompose rapidly at 20 "C to give the nitrile esters (18)presumably via the nitrenes (17; R = Pr' or Bu') whereas the corresponding aldehyde is much more stable. This can be explained by assuming that in the case of the ketones the presence of the bulky substituent R results in the ketone carbonyl being out of plane with the rest of the molecule and hence the contribution from the canonical form (20) which leads to (18) is greater than the alternative (19) which can contribute when the carbonyl group is in plane as in the case of the aldehyde.a& -NQCOR 0- (19) (20) A general route to nitrenes (22) involves the thermally induced fragmentation of hydroxylamine derivatives (21).l6 Kinetic data support a concerted mechanism for the reaction. Sulphenylnitrenes (22;R2 = SR) can be obtained by Pb(OAc) oxidation of 2,4-dinitrobenzenesulphenamides;a full description of this method and of the subsequent trapping of these nitrenes by electron-rich olefins has been given.l7 SiR13 / R~-N 'OSiR13 R2 = C02Et,ArC0 (21) ArS02,Me R2 # SiMe3 Pyrolysis of p-phenethylsulphonyl azides gives rise to the sulphonylnitrenes (23) and not P-phenethylnitrenes uia a Wolff -like rearrangement and SO2 elimination as could perhaps happen." Some unexpected products from (23) are the pyridines (24):a reasonable mechanism for their formation involves azepine intermediates.By contrast the sulphonylnitrenes (25; n = 3 or 4) preferentially undergo intramolecular insertion into the aryl ring to give the sultans (26) whereas nitrene (25; n = 5) gives mainly (27) by insertion into an alkyl C-H bond." A general Is P. J. Newcombe and R. K. Norris Terruhedron Lett. 1981 22,699. l6 Y.H. Chang F.-T. Chiu and G. Zon J. Org. Chem. 1981,46 342. R. S. Atkinson and B. D. Judkins J. Chem. Soc. Perkin Trans. 1. 1981 2615. See also W. Bludssus and R. Mews Chem. Ber. 1981,114,1539. '* R. A. Abramovitch W. D. Holcornb and S. Wake J. Am. Chem. SOC.,1981,103 1525. l9 R.A.Abramovitch S.B. Hendi and A. 0.Kress J. Chem. SOC.,Chem. Commun. 1981 1087. Arynes Carbenes Nitrenes and Related Species route has been developed for the preparation of aryloxysulphonyl azides ArOS02N3 which can serve as precursors to aryloxysulphonylnitrenes.20 Benzoylnitrene (28) can be obtained by photolysis of benzoyl azide but not by thermolysis as under such conditions the substrate undergoes a Curtius rearrange- ment.21 The stereoselectivity of the insertion reactions of (28) into C-H bonds suggests that it is formed in the singlet state. Ethoxycarbonylnitrene reacts with stable mesoionic compounds (29) by addition across the 1,3-dipole followed by a rearrangement to give the imidazolines (30).22 Cyanonitrene NC-N :,apparently in its triplet state has been generated from sodium cyanamide and t-butyl hypochlorite in methan01,*~ and the borylnitrene (Pri2N)2BN can be obtained by thermolysis (450485"C) of the corresponding a~ide.~~ There has been considerable interest in the chemistry of amino-nitrenes during the past year.Phthalimidonitrene (3 1)reacts with nitroso-compounds to give the diazene-1-oxides (32) in a reaction that can be reversed by photolysis or acid (31) (32) hydroly~is.~~ Subsequent nucleophilic attack on (32; R = R'*N) results in the forma-tion of amino-nitrenes (33) whose fate depends upon the attacking species. Oxida- tion of the quinazolones (34) by Pb(OAc) in chloroform leads to the expected 'O M. Hedayatullah and J. C. Hugueny Synth. Commun. 1981,11,643. *' M.Inagaki T.Shingaki andT. Nagai Chem. Lett. 1981 1419. 22 T.Sheradsky and D. Zbaida Tetrahedron Lett. 1981 22 1639. 23 M.G.K. Hutchins and D. Swern Tetrahedron Lett. 1981 22,4599. 24 W.Pieper D. Schmitz and P. Paetzold Chem. Ber. 1981,114,3801. *' L. Hoesch and B. Koppel Helv. Chim. Acta 1981,64,864;L. Hoesch ibid.,p. 890;see also C. Leuenberger L. Hoesch and A. S. Dreiding ibid. p. 1219. 102 D. W.Knight nitrene which can be trapped intermolecularly by olefins such as styrene but which can also undergo intramolecular trapping probably via an intermediate dipolar species (35) formed by overlap of thq empty nitrene orbital with the terminus of the olefinic side-chain.26 A record has been set by the recording of the "N n.m.r. spectrum of the diazene (36)at -90 "Cin dimethyl ether the nitrene nitrogen appears 917 p.p.m.downfield from 15NH3,which is the most highly deshielded nitrogen atom yet observed in a neutral organic compo~nd.~' The pyrrolidine analogue (37) has been found to be stable in solution at -78 "C for several days.28 Preliminary photochemical studies of this compound suggest that its So and S1states have differing geometries (theory predicts that the So state is planar while the S1is pyramidal) and that the S1-Tl gap is large again in agreement with theoretical predictions. The reactions of N-aminopyrroles with dimethylacetylene dicarboxylate have been found to have several interesting The initial reaction presumably involves formation of the Diels-Alder adducts (38) which fragment to give the phthalates (39) and amino-nitrenes (40).Another product isolated from the reaction mixture is dimethyl maleate which could arise from reduction of the starting acetylene by di-imides formed by isomerization of nitrenes (40).Finally the pyrazolines (41)are also formed conceivably by an ene reaction in which (40) behaves as a diazene followed by disrotatory ring closure (Scheme 1). 4 Carbenes STO-3G calculations have further confirmed the influence of substituents on the singlet-triplet energy gap in carbenes ?r-donors stabilize the singlet state more 26 R. S. Atkinson J. R. Malpass K. L. Skinner and K. L. Woodthorpe J. Chem. SOC., Chem. Commun. 1981 549; R. S. Atkinson J. R. Malpass and K. L. Woodthorpe ibid. p. 160. 27 P.B. Dervan M. E. Squillacote P. M. Lahti A. P. Sylwester and J. D. Roberts J. Am. Chern. Soc. 1981,103,1120. 28 P. G. Schultz and P. B. Dervan J. Am. Chem. Soc. 1981,103 1563. 29 A. G. Schultz M. Shen and R. Ravichandran Tetrahedron Lett. 1981 22 1767. Arynes Carbenes Nitrenes and Related Species (40) E E (E = C02Me) Scheme 1 than the triplet whereas the reverse is true with T-acceptor sub~tituents.~' Estimates of these singlet-triplet gaps can be made using either calculated T charges for the corresponding substituted benzenes or empirical ug constants. 6,6-Dimethylfulvene has been shown to be a useful standard in the assessment of carbene philicities. (Scheme 2).31Frontier molecule orbital calculations agree with the outcome of these experiments in that they predict that there will be a predominance of LUMO (CCl,)/HOMO (fulvene) interactions (CC12 is an electrophilic carbene) whereas the reverse is true for a nucleophilic carbene such as C(OMe)2.Nucleophilic or @/ {arnbiphilic:CR2 Styrenes t-@QI-< q& @C electro~hilic R R :cR Scheme 2 It is well known that the selectivity of addition of singlet carbenes to olefins increases with decreasing carbene reactivity. However recent frontier orbital calcu- lations suggest that this is only true for electrophilic carbenes whereas carbenes generally thought of as being nucleophilic [e.g. C(OMe)2 C(NMe,),] should show the reverse effect i.e. increasing selectivity with increasing rea~tivity.~~ In general alkyl- and dialkyl-carbenes cannot be used in synthesis owing to their propensity 30 P.H. Mueller N. G. Rondan K. N. Houk,J. F. Harrimn D. Hooper B. H. Willen and J. F. Liebman J. Am. Chem. SOC., 1981,103,5049. 31 R. A. Moss C. M. Young L. A. Perez and K. Krogh-Jespersen J. Am. Chem. SOC., 1981,103,2413; see also K. Steinbeck T. Schenke and J. Runsink Chem. Ber. 1981 114 1836. 32 W. W Schoeller Angew. Chem. Int. Ed. Engl. 1981 20 698; see also B. Giese and W.-B. Lee Chem..Ber. 1981 114 3306. 104 D. W.Knight for very rapid intramolecular rearrangement to give olefins. In an extension of previous work two closely related iron complexes have been which can effect overall transfer of an ethylidene group to olefins leading to methyl- cyclopropanes often in very high yields (Scheme 3).It may well turn out that higher OMe -[ r 1 Cp(CO)(L)Fe -CH / cp(cy$eT 'Me Me (L = COorPPh3) bMe SPh Cp(CO)*Fe -CH \ Reagents i Me,SiOSO,CF,-CH,CI,-)=( -78 OC; ii FS0,Me-CH,CI,-)=( ,25 "C Scheme 3 homologues of these complexes can also be used in this way. A more conventional method has been used to prepare dimethylcarbene (n-BuLi and 2,2-dibromopropane) which can be trapped in siru by olefins at low temperatures (-70 "C):yields of the gem-dimethylcyclopropanes are however usually fairly An exception to this general rule that alkyl carbenes are unstable with respect to rearrangement is adamantylidene. This species can be generated by photolysis of spiro(adamantane-2',3'-diazirine) and undergoes mainly cyclopropanation reac- tions with olefins together with some C-H insertion into the ~lefin.~~ Probably these reactions succeed because intramolecular rearrangement of the carbene would lead to highly strained adamantene.It is known that bicyclo[ l.l.O]butenes react with carbenes to give penta-1,4- dienes and the mechanism is usually thought to involve a diradical intermediate. Thus the 1,2,2-trimethyl derivative (42) would be expected to give diradical (43) by attack of the carbene at the less-hindered site (Scheme 4) (43) then can undergo two possible modes of cleavage (a) or (b). Pathway (a) seem most likely as this will lead to the more highly substituted diene (44); however only diene (45) is formed.37 This has lead to the proposal that the reaction does not involve a diradical intermediate but rather that it is a concerted process as shown in Scheme 4 pathway (c).Evidence has been found that suggests that the mechanism for the photochemical addition of alcohols to acylsilanes (46) leading to acetals (47) involves the TIstate of (46) but that this gives an intermediate most likely the siloxycarbene (48) before reaction with the alcohol R30H.38"Ry contrast the photochemical forma- 33 M. Brookhart J. R. Tucker and G. R. Husk J. Am. Chem. SOC., 1981 103,979. 34 K. A. M. Kremer P. Helquist and R. C. Kerber J. Am. Chem. Soc. 1981 103 1862. 3s P. Fischer and G. Schaefer Angew. Chem. Znt. Ed. Engl. 1981,20,863. 36 R. A. Moss and M. J. Chang Tetrahedron Lett. 1981 22 3749. 37 G.B. Mock and M. Jones jun. Tetrahedron Lett. 1981 22 3819. 38 (a)R. A. Bourque P. D. Davis and J. C. Dalton J. Am. Chem. Soc. 1981 103 697; (b)J. C. Dalton and R. A. Bourque ibid. p. 699. 105 Arynes Carbenes Nitrenes and Related Species (45) Scheme 4 tion of cyclopropanes (49) from (46) and electron-poor olefins does not involve the carbene (48) but rather both the S1 and TI states of the acylsilane which add directly to the ~lefin.~~~ This does not appear unreasonable as (48) should be an electrophilic carbene. n m Me,SiO R' OV0 svs (49) (50) (51) (52) Calculations indicate that initial cleavage of one C-0 bond in carbene (50) is a lower-energy pathway than a concerted process in the decomposition of (50) to ethylene and carbon dio~ide.~' This may be due to the fact that in a concerted mechanism the carbon dioxide will necessarily have a bent structure when initially formed and further suggests that when possible precursors to (50) e.g.carbonate tosylhydrazones decompose to give olefins the carbene (50)is not an intermediate. Semi-empirical CND0/2 calculations indicate that the carbene (51) from 1,3- dithiole is a stable species with respect to ring opening whereas the isomeric species (52)from 1,2-dithiole is not and will undergo cleavage to give thioacyl thi~ketene.~' This agrees well with experimental observations (51)can serve as a useful precursor to 1,1',3,3'-tetrathiafulvenesby dimerization whereas analoguous reactions with (52) fail. The nucleophilic carbenes (53) are formed when 2-tetrazobenzo[d]thiazolines prepared from the corresponding azidinium salts and LiN3 at -50 "C are warmed to ca.O°C.41 They can be trapped by various reagents such as methanol (to give the 2-methoxy-derivative) sulphur (to give the 2-thione) and diazonium salts 39 D. Feller E. R. Davidson and W. T. Borden J. Am. Chem. SOC., 1981 103,2558. 40 C. Th. Pederson J. Oddershede and J. R. Sabin J. Chem. SOC., Perkin Trans. 2 1981 1062; H. Behringer and E. Meinetsberger Liebigs Ann. Chem. 1981 1729. 41 H. Balli H. Griiner R. Maul and H. Schepp Helu. Chim. Acru 1981,64 648. 106 D. W.Knight N=N 0 OMe 0 OMe OMe (54) (55) (56) (57) (leading to 2-arylazo-thiazolinium salts). Reaction with electron-deficient olefins such as tetracyanoethylene yields the expected cyclopropane derivatives.In agreement with some recent calculations it appears that carbonyl yields can decompose to give carbene~.~~ Thus thermolysis at 80 "C of the oxadiazoline (54) gives similar amounts of the two carbenes (56) and (57) probably uia the ylide (55). Theory indicates that an electron-donating substituent (in this case OMe) is necessary for this fragmentation. Certainly more work is needed to confirm these ideas. A useful new method for the generation of carbenes from gem-dihalo- compounds involves passage of the compound through a heated (24-140 "C) evacuated tube packed with glass turnings coated with methyl-lithi~m.~~ The attraction of this procedure is that it couples the advantages of vacuum pyrolysis techniques (e.g.high dilution and hence suppression of intermolecular reactions) with mild conditions allowing the isolation of unstable products in high yields which is obviously of benefit in any study of intramolecular carbene rearrangements. Using this technique a temperature dependence has been observed in the rearrange- ment of carbene (58) to (59) and (60),which suggests that migration of the ethano bridge in (58) leading to (60) is slower than reaction at the olefin leading to (59) (Scheme 5). A similar dependence has also been found in the vinylcyclopropylidene + cyclopentadiene and vinylallene rearrangement with higher temperatures favouring the latter. Scheme 5 The cyclopropylidene thioethers [(61) generated using MeLi in ether from the corresponding gem- dibromide] undergo two types of intramolecular reaction,44 namely the well known cyclopropylidene + allene rearrangement and insertion into the C-H bond shown the extent of insertion following the usual order of 42 M.Bekhazi and J. Warkentin J. Am. Chem. Soc. 1981 103,2473. 43 U. H. Brinker and J. Ritzer J. Am. Chem. Soc. 1981 103 2116. 44 M. S. Baird J. Chem. Res. (S) 1981 352. Arynes Carbenes Nitrenes and Related Species reactivity i.e. 3’ 21 2y > 1’. By contrast the homologous thioethers (62) gives only allenes whereas (63) rearranges to the tetrahydrothiopens (65) presumably by way of ylide (64). f RS 73 (65) a; R = H b; R = CH2CHCH2 The vinylcyclopropylidene + cyclopentenylidene rearrangement is very well known (cf.ref.43) but the homologous vinylclobutylidene + cyclohexenylidene is not known Such a rearrangement has now been observed in the highly constrained system (66) (Scheme 6).45The intermediate (67) when generated in other ways gives (68) in very high yield which strongly suggests that the reaction proceeds as shown. However (68) is a minor product as the predominant pathway for decompo- sition of (66) involves the usual cyclobutylidene + methylenecyclopropane rear- rangement and vinyl migration. When cyclopropylphenylcarbene (69) is generated by photolysis of the corres- ponding diazo-compound in the presence of an olefin the main product is the cyclobutene (70) very little reaction occurs with the ~lefin.~~ However at lower temperatures (-95 + -130 “C)much more of the bicyclopropyl product is obtained at the expense of (70).The reasons for this are not clear but one explanation is that the formation of (70) involves an excited singlet state of carbene (69) which would be less populated at the lower temperatures. Further studies on the rearrangements of carbene (71) have confirmed that benzvalene (72) is not fo~med.~’ Furthermore MIND0/3 calculations predict that the conformation of (71) is not suitable for the expected intramolecular cyclo- propanation which would lead to (72) to occur. The isolated products are instead the isomeric benzvalene (73; 38%) arising via intramolecular [ 1,4]-addition of the carbene together with cyclopentadiene (74; 0.9’/0) from insertion into the methyl group and toluene (18%) from a [1,2]-carbon shift.45 U. H. Brinker and L. Konig J. Am. Chem. SOC.,1981,103,212. 46 R. A. Moss and W. P. Wetter Tetrahedron Lett. 1981 22 997. 47 U. Burger G. Gandillon and J. Mareda Helu. Chim.Acta 1981,64,844. 108 D. W. Knight (73) (74) (75) X = CH2 NH 0,or S The phenylcarbenes (79 generated by pyrolysis (500 "C) of the sodium salts of the appropriate aldehyde tosylhydrazones undergo intramolecular insertion into an adjacent C-H bond when X = CH or NH to give dihydroanthracenes and dihydroacridines re~pectively.~~ However when X = 0 or S cyclopropanation of the adjacent phenyl ring occurs leading to benzo[b]cyclohepta[d]furans and thio- phens. Overall yields for all these reactions are rather low possibly because of the high temperatures used.The absolute rates of decay of the syn- [e.g. (76)] and anti-[e.g. (77)]rotomers of 1-and 2-naphthylcarbene 9-anthrylcarbene and 2-pyridylcarbene have been measured in low-temperature mat rice^.^' Under such conditions decay is via pseudo-first order hydrogen abstraction from the matrix at a much faster rate than syn-anti-conversion suggesting that the barrier between the two forms must be greater than 4.5-6.3 kcal mol-'. In related four isomeric quinolylcar- benes have been generated photolytically in the solid state. E.p.r. spectra indicate H Y' the presence of two similar triplet states (i.e.rotomers) whose zero-field parameters are quite similar to naphthylcarbene showing unexpectedly that the nitrogen atom has little effect on the .rr-spin distribution.Cyclopentadienylidene (78) has been characterized in low-temperature matrices by both U.V. and i.r. spe~troscopy.~~ The carbene is generated by photolysis of the 5-diazocyclopentadiene via an excited A1state of the latter. Desilylation of the tropylium salt (79)with fluoride ions leads to cycloheptatrieny- lidene (go) which not surprisingly reacts as a nucleophilic carbene giving the expected cyclopropanes on reaction with electron-deficient olefins such as dimethyl 48 W. D. Crow and H. McNab Aust. J. Chem. 1981,34,1037. 49 V. P. Senthilnathan and M. S. Platz J. Am. Chem. SOC., 1981,103 5503. 50 R. S. Hutton H. D. Roth M. L. M. Schilling and J. W. Suggs J.Am. Chem. SOC., 1981,103,5147. 51 M. S. Baird I. R. Dunkin N. Hacker M. Poliakoff and J. J. Turner I. Am. Chem. SOC., 1981 103 5190. Arynes Carbenes Nitrenes and Related Species fumarate or 1,2-di~yanoethylene.~’ The nucleophilicity of (80) is also evident from its reactions with alcohols. In general carbenes react with saturated alcohols in three possible ways namely by direct insertion into a C-H bond by electrophilic (79) Pl .. (82) (83) attack on the alcohol oxygen followed by proton transfer or by protonation in the case of a nucleophilic carbene to give a carbonium ion. Good evidence has been obtained which suggests that (80) does indeed react via the latter pathway to give an intermediate tropylium cation and moreover that cyclopentadienylidene (78) reacts as an electrophile via the second pathway thereby giving rise to a cyclopentadienyl anion.53 Reaction between lithium cyclononatetraenide and p-nitrobenzenesulphonyl azide appears to lead to cyclononatetraenylidene (8 1) via the unstable diazo-derivative judging by the isolation of hydrocarbon (82) a rearrangement product which could reasonably arise from dimerization of triplet (8l),followed by electrocyclic ~earrangement.~~ The activation parameters for intersystem crossing of singlet to triplet fluoreny- lidene (83) have been measured directly in a~etonitrile.~~ The results agree with theoretical work and are consistent with the observation that singlet (83) reacts exothermally with olefins.The measurements were made by monitoring both the appearance of triplet (83) at 400nm or the decay of singlet (83) at 470nm and both give the same answer within experimental error.However it has been pointed outs6 that the disappearance of singlet (83) cannot be used.as a measure of the rate constant for intersystem crossing as this species also abstracts protons from the solvent to give a radical (A,, 500nm). This curious triplet-like behaviour of singlet (83) is also shown by its non-stereoselective addition to cis-~lefins~~ (cf.ref. 58). Cycloalkenylidenes with (4n + 2)~ electrons are known to be nucleophilic (cf. ref. 52,53). A study of one such carbene (84) shows that this species exhibits both typical singlet and triplet reactivity suggesting that the two states are in rapid eq~ilibriurn.~~ (Such an equilibration has previously been used to explain various ’’ R.W. Hoffmann M. Lotze M. Reiffen and K. Steinbach Leibigs Ann. Chem. 1981,581; E. E. Waali J. Am. Chem. SOC.,1981,103,3604. s3 W. Kirmse K. Loosen and H.-D. Sluma J. Am. Chem. SOC.,1981 103 5935; see also T. Harada and A. Oku ibid. p. 5965. 54 E. E. Waali and C. W. Wright J. Org. Chem. 1981,46,2201. ” J. J. Zupancic and G. B. Scltuster J. Am. Chem. SOC.,1981,103,944. 56 P. C. Wong D. Griller and J. C. Scaiano J. Am. Chem. SOC.,1981,103,5934. s7 J. J. Zupancic P. B. Grasse and G. B. Schuster J. Am. Chem. SOC.,1981 103 2423. 58 H. Durr and A. Hackenberger J. Chem. Res. (S) 1981 178. 110 D. W.Knight aspects of the chemistry of diphenylmethylene).The cyclopropanation reactions of (84) do indeed suggest that it is a nucleophilic species. calculation^,^^ in agreement with experimental evidence have shown that viny- lidene H2C=C is very unstable with respect to isomerization to acetylene uia rapid tunneling through a barrier of ca. 4 kcal mol-'. Two useful synthetic procedures involving isopropylidenecarbene (86)have been developed. The carbene is generated by treatment of the silyl vinyl triflate (85) with Bun4NF when this is carried out in the presence of an isocyanate vinyl carbamates (87) are formed probably by way of an intermediate ylide produced by electrophilic attack of (86) on the oxygen of the isocyanate.60 This procedure is significant in that little is known about the chemistry of vinyl carbamates.Carbene (86)also reacts with thiones to give divinyl sulphides (88),presumably by insertion into the S -H bond of the tautomeric enethiok61 The thione + enethiol equilibra- tion is sufficiently slow to allow the isolation of pure enethiols and hence if these were used as substrates the yields of (88)should be considerably higher than those (2540%) obtained by this method. Alkadienylidenecarbenes (89) insert into Group 4 hydrides (RL,MH; M = Si Ge or Sn) to give the novel cumulenes (90) as isolable compounds in 26-88% yields.62 Although alkylidene carbenes (9 1)with n = 2 [e.g. (86)] n = 4 [e.g. (89)l are known,62 in the 'odd-numbered' series only (91; n = 3) has been observed. Evidence has now been found that supports the intermediacy of dimethyltrienylidenecarbene (91; n = 5 R = Me) in the elimina- tion of the elements of hydrogen chloride from 5-chloro-5-methylhexa-1,3-diyne by potassium t-buto~ide.~~ A full report has appeared on the generation of difluorocarbene CF2 by reaction between (CF,),Cd.glyme (formed from (CF3)2Hg and Me2Cd in glyme) and an acyl halide.64 The other product of the reaction which can be carried out at temperatures of -78 "Cor lower is the acyl fluoride.In addition it is evident that 59 Y. Osamura H. F. Schaefer 111 S. K. Gray and W. H. Miller J. Am. Chem. SOC.,1981 103 1904. 60 P. J. Stang and G. H. Anderson J. Org. Chem. 1981,46,4585. " P. J. Stang and S. B. Christensen J. Org. Chem. 1981 46 823. 62 P. J. Stang and M. R. White J. Am.Chem. SOC.,1981,103 5429; see also P. J. Stang and M. Ladika ibid. p. 6437. 63 W. J. le Noble S. Basak and S. Srivastava J. Am. Chem. SOC.,1981,103,4638. 64 L. J. Krause and J. A. Morrison J. Am. Chem. Soc. 1981 103,2995. Arynes Carbenes Nitrenes and Related Species 111 R:C=C=C=C R:C=C=C=C MR; / \ R2C=(C)n-2=C H (89) (90) (91) the carbene is obtained in the singlet state by reason of its stereoselective cyclopro- panation reactions with olefins. Under phase-transfer conditions dichlorocarbene (from NaOH-CHCl,) adds to 3-alkenoic acids to give the dichlorocyclopropanes (92) generally in 55430% yield.65 Such reactions with alkenoic acids have not been previously reported. Steric limitations in the addition of CC12 and CBrF to olefins have been found for example these carbenes give no cyclopropanes with very hindered olefins such as 1,1,2-triphenylpropene and bifluorenylidene.66 The classical Reimer-Tiemann reaction of phenol results in a mixture of 2- and 4-hydroxybenzaldehydes in a ratio of ca.59 :41. If however small amounts of a-cyclodextrin are added to the reaction mixture then the ratio becomes 18:82; presumably this is because of complexation between the cyclodextrin and the phenolate anion thereby giving rise to steric shielding of the ortho An extraordinary feature of the reactions between dihalocarbenes and norbor- nadienes is that homo-[ 1,4]-adducts (93) are formed together with the expected em-[ 1,2]- (94) and endo-[1,2]-(95) adducts. [the latter two products are actually isolated as the rearranged dienes (96)].In previous studies Jefford and Huy found that the exo-[1,2] homo-[1,4] ratio [i.e. (94) :(93)] decreased when substituents of increasing electron-attracting ability were present at C-2 (i.e. more homo-[ 1,4] addition with less electron-rich norbornadienes). This was explained by assuming an increased nucleophilic character for CF in the homo-[ 1,4]-additions. However it has now been pointed out that this ratio decrease could also be due to a decrease in the rate of exo-[1,2]-addition and that substituent effects in these systems should be judged by comparing homo-[1,4] endo-[1,2] ratios6* The reaction of 7-CIP CI OzH 6 xx R f XX (95) 65 T. Fujita S. Watanabe K. Suga and K. Sugahara Synthesis 1981 1004.66 L. Anke D. Reinhard and P. Weyerstahl Liebigs Ann. Chem. 1981 591. 67 M. Komiyama and H. Hirai Bull. Chem. SOC.Jpn 1981 54 2053. 68 G. W. Klumpp and P. M. Kwantes Tetrahedron Lett. 1981 22 831. 112 D. W. Knight substituted norbornadienes with phase-transfer-generated CCl has been studied and it was found that the substituent has relatively little effect on the amounts of homo[l,4]- and endo-[1,2] adducts [(93) and (95)] formed but significantly affects exo-[1,2]-adduct (94) formation. The results are consistent with the known elec- trophilicity of CC12 it may be that homo-[ 1,4]-addition occurs in norbornadienes because unfavourable steric constraints tend to prevent the incoming carbene from approaching the olefin along the best reaction pathway for [1,2] addition.Monochlorocarbene reacts with enol ethers stereoselectively and in high yields to give the cyclopropanes [e.g. (97) from cis-ethylpropenylether] even though such olefins are only weakly nucleophilic presumably the reaction is assisted by com- plexation between the carbenoid species and the ether oxygen.69 The carbene is generated from CH2C12 and MeLi.LiBr and this results in the formation of the bromocyclopropanes corresponding to (97) as by-products. The scope of C-H insertion reactions of carbene (98) has been investigated; the species is generated by thermolysis of tetrachlorocyclopropene.70A complete experimental procedure has been presented for the generation of chloro(pheny1)carbene (99) by thermolysis (reflux in C6H6) of the corresponding diazirine together with its subsequent trapping EtO Me (97) (98) (99) by p-methylstyrene followed by base-induced elimination of HC1 to give 1,2- diphenyl-3-methylcyclopropene.71Carbenes such as (99) also insert into reactive C-H bonds such as the 1-C-H adamantane and the 2-C-H of 1,3-dio~olanes.~~ Yields of the adducts are in the range 10-44°/~ when the carbenes are generated from a,a-dihalotoluenes and potassium t-butoxide at 80°C in the presence of 18-crown-6.(Trifluoromethyl)chlorocarbene CF,CCl can be obtained by photolysis of the appropriate diazirine and undergoes standard cyclopropanation reactions with ole fin^.^^. These are stereoselective with cis-and trans-butene indicating that the carbene is formed in the singlet state and in addition it is evident that the trifluoromethyl group strongly destabilizes the species with respect to chloro(methyl)carbene MeCC1.The rhodium cluster complex Rn,(CO),, has been found to be a very efficient catalyst in the formation of cyclopropanecarboxylic acid esters from olefins and ethyl dia~oacetate.~~ When more conventional catalysts such as rhodium(I1)carboxy- lates are utilized such reactions are often inefficient unless a large excess of olefin is used. However better yields can be obtained without the need for such an excess 69 R. Barlet R. LeGoaller and C. Gey Can. J. Chem. 1981 59,621. 70 E. V. Dehmlow and Naser-ud-Din J. Chem. Res. (S) 1981 144. 71 A. Padwa M. J. Pulwer andT. J. Blacklock Org.Synth. 1981 60 53. 72 K. Steinbeck and J. Klein Angew. Chem. Znt. Ed. Engl. 1981 20 773. 73 R. A. Moss W. Guo D. Z. Denney K. N. Houk and N. G. Rondan J. Am. Chem. Soc. 1981,103 6164. 74 M. P. Doyle W. H. Tamblyn W. E. Buhro and R. L. Dorow Tetrahedron Lett. 1981 22 1783; M. P. Doyle W. H. Tamblyn and V. Bagheri J. Org. Chem. 1981,46 5094. Arynes Carbenes Nitrenes and Related Species if the diazoacetate is added slowly to the reaction mixt~re.~' Rh(OAc)2-catalysed decomposition of methyl a-diazoalkanoates leads to the corresponding cis-cw,@-unsaturated esters; the mechanism of this reaction is as yet unclear.76 In general a-oxocarbenes are not prone to react as 1,3-dipoles. The formation (Scheme 7) of 1,3-dioxole-4-carboxylates(100) from aldehydes and methyl a-diazoacetoacetate seems to represent an exception.77 The a-cyanimino carbene OQ-f *L C0,Me 1 =,Bx 0 C0,Me (100) C0,Me Scheme 7 (101) also appears to react in this manner with benzene giving (102) as the sole product and with acetonitrile leading to the [3 + 2]cycloadduct (103) (Scheme 8).78However alternative mechanisms can be drawn which involve initial cyclopro- pane or 1-azirine formation respectively followed by a vinylcyclopropane to cyclopentene type rearrangement.Scheme 8 In contrast to the usual outcome of a Buchner reaction the addition of diazoace-tate to monosubstituted benzenes when catalysed by Rh2(02CCF3) results in the formation of the kinetic non-conjugated cycloheptatrienes (104)in high yields 75 M.P. Doyle D. van Leusen and W. H. Tamblyn Synthesis 1981,787. 76 N.Ikota N. Takamura S.D. Young and B. Ganem Tetrahedron Lett. 1981,22,4163. 77 M.E.Alonso and A. W. Chitty Tetrahedron Lett. 1981 22 4181. D. Danion B. Arnold and M. Regitz Angew Chem. Int. Ed. Engl. 1981 20 113. For a review of a-iminocarbene chemistry see M. Regitz B. Arnold D. Danion H. Schubert and G. Fusser Bull. SOC.Chim. Belg. 1981,90,615. 114 D. W.Knight (270%) as a mixture of all three possible isomer^.'^ An example of a Buchner-type reaction involving thiophen is the formation of (106) in 11% yield from the diazopenicillanate (105) presumably by way of a spiro-cyclopropyl adduct." R' C02R (105) Solution photolysis of a-diazo-amides produces various compounds such as p-lactams and amides the latter by reaction with the solvent or by Wolff rearrange- ment.81 Although the precise nature of the products is obviously solvent and substituent dependent it appears that carbenoid intermediates only arise from the sym-E-conformation (107) whereas the sym-2-conformation (108) leads to p-lactams etc.via an excited singlet state of the diazo-amide. HO "QO NkNR1R2 H N-R2 Ph Ph,P-PPh2 PhP< 0. II II II PPh, 00 O II 0 (109) (110) (111) In an extension of previous work it has been found that photolysis of (109) gives the carbene (1lo) which rearranges to phosphene (111)by a [1,2]-phenyl migration before it can be trapped.82 5 Silylenes Further work has confirmed that silenes having a proton attached to the silicon [e.g.(1 12)] undergo rapid thermal isomerization to silylenes [e.g. (113)]. Further-more it has been found that the process can be reversed by photolysis in an argon matrix at 50 K and that under such conditions (112) only slowly reverts to (113) and instead dirneri~es.~~ In view of those observations it is not surprising that '9 A. J. Anciaux A. Demonceau A. F. Noels A. J. Hubert R. Warin and P. TeyssiC J. Org. Chem. 1981 46 873; A. Demonceau A. F. Noels A. J. Hubert and P. TeyssiC J. Chem. SOC.,Chem. Commun. 1981,688. L. Chan and S. A. Matlin Tetrahedron Lett. 1981 22 4025; see also E. Wenkert M. L. F. Bakuzis B. L. Buckwalter and P. D. Woodgate Synth. Commun. 1981 11 533. H. Tomioka M. Kondo and Y. Izawa J. Org. Chem.1981,46 1090. M. Regitz F. Bennyarto and H. Heydt Liebigs Ann. Chem. 1981 1044; see also M. Regitz A E. M. Tawfik and H. Heydt ibid. p. 1865; M. Regitz and H. Eckes Tetrahedron 1981 37 1039. T. J. Drahnak J. Michl and R. West J. Am. Chem. SOC.,1981,103 1845. Arynes Carbenes Nitrenes and Related Species thermolysis of the silacyclobutane (114) at 625 “Cleads only to products arising from (113) and not (l12).84Previous theoretical results have also indicated that silenes such as (112) are less stable or at best as stable as the corresponding silylenes (1 13). By contrast some new calculations suggest that the dimethylsilene (1 15) is more stable than the isomeric silylene (1 16) or the isomeric carbene (117)? EHMe Me2Si=CH2 MeSiCH2Me Me2HSiCH (115) (116) (1 17) In general silylenes appear to react with ethers in an electrophilic sense to give ylide intermediates.The isolation of methoxysilanes [e.g. (1 IS)] from the reactions of photochemically generated dimethysilylene with ally1 methyl ethers seems to confirm this (Scheme 9).86 Scheme 9 The silylsilylene (119) can be generated photochemically from (Me,Si),SiPh and adds to olefins stereo~electively.~~ For example reaction with trans-1-chloropropene leads to (120) whereas with cis- 1-chloropropene the cis-isomer (121) is formed. One explanation for this is that the reaction involves an intermedi- ate silacyclopropane which then undergoes rearrangement to give the observed products. I ,Si Me,Si 84 R. T. Conlin and D. L.Wood J. Am. Chem. Soc. 1981 103 1843; however see Y.Yoshioka and H. F. Schaefer 111 ibid. p. 7366. 85 M. Hanamura S. Nagase and K.Morokuma Tetrahedron Lett. 1981,22 1813; see also M. S. Gordon and R. D. Koob J. Am. Chem. Soc. 1981,103,2939; M. S.Gordon and J. A. Pople ibid. p. 2945. 86 D.Tzeng and W. P. Weber J. Org. Chem. 1981 46,693; see also W. Ando M. Ikeno and Y. Hamada J. Chem. SOC.,Chem. Commun. 1981,621. M. Ishikawa K.Nakagawa and M. Kumada J. Am. Chem. SOC.,1981,103,4170.