Azoniaazulenes. Part 7. Functionalisation of 10H-azepino1,2-aindoles

摘要

J. CHEM.SOC. PERKIN TRANS. 1 1995 Azoniaazulenes. Part 7.' Functionalisation of 1OH-AzepinoEl ,Z-aindoles Gurnos Jones,*Sa Michael W. Kempa,a Michael B. Hursthouse* and K. Abdul Malikb a Department of Chemistry, Keele University, Keele, Staffordshire ST5 5BG, UK SERC Crystallography Unit, School of Chemistry and Applied Chemistry, University of Wales, Cardiff, PO Box 912, Cardiff CFl 3TB, UK Direct functionalisation of azepinoindole 7 is described. Electrophilic attack by bromine, and by the Vilsmeier, Mannich, and Friedel-Crafts reagents gives 11 -substituted azepinoindoles 8 and 11-1 4. From the bromo compound 8 an 11 -1ithioazepinoindole was obtained. With tert-butyllithium the azepinoindole 7 gave an allylic anion 36, which with trimethylsilyl (TMS) chloride gave the 8H-8- TMS derivative 16 and the IOH-8.10-bis-TMS derivative 17, for which an X-ray structure is provided.Alkylation of anion 36 gave mixtures of 8H-8-alkyl- and 1OH-I 0-alkyl derivatives 18-25, while further lithiation of compound 16 and alkylation gave mixtures of 8.8- and 8.1 0-disubstituted compounds 26-33. Decomposition of azide 43 gave eight identified products 44, 45, 47-51 and 54. We have had, as a long standing objective, the synthesis of the aromatic pyrrolol,2-aazepinium ion 1 or its benzologue the cpX cDJr+' 0 R 1 R=H 2R=H 3R=OH 4 R =OH, X= HS04- 5 R = CHPCN 6 A IOEt, X = BFd- azepinoc 1,2-uindolium ion 2. Hydroxy- and ethoxy-substi- tuted ions such as compounds 3,4and 6have been obtained by protonation or ethoxylation 2,3 of the corresponding azepinones but the only alkyl derivative remains the cyanomethyl compound 5.' We report here further approaches to azepinoindolium salts, which, though unsuccessful have produced methods to functionalize azepinoindoles directly, a process not previously achieved.A number of 10H-azepinoindoles have been made from a standard nitrene insertion reaction, starting from 2-azidodi- phenylmethane~,'-~-~and we have improved the synthesis of the parent 7 to 70 after three stages from the readily avail- able 2-aminobenzophenone (2-aminodiphenylmethanone). Attempts to remove hydride directly from compound 7 have previously failed, possibly because of electrophilic attack by the carbocation on the reactive indolic 11 position.We tried hydride abstraction from 11-methylazepinoindole using the triphenylmethyl carbocation and the potentially more hindered 1,2,4,6-tetramethylcycloheptatrienyliumtetrafluoroborate, but without success. Feeling that a better leaving group was required, we next studied the bromination of azepinoindole 7, using N-bromosuccinimide under a variety of conditions, the best yields of a monobrominated product being obtained in carbon tetrachloride solution at room temperature in an ultrasonic bath. The mass spectrum showed the compound to be a monobromo derivative, but it proved too unstable for consistent microanalysis. The only significant change in the 'H NMR spectrum from that of compound 7 was the absence of the singlet at 6 6.20 diagnostic for the indole proton 11-H, and hence bromination had not taken place at the allylic position as required for ionisation.We confirmed that the substitution was electrophilic since the bromination was unaffected by the presence of glacial acetic acid as a radical trap. A short investigation of the potential in synthesis of this 11-bromo derivative 8 was made (Scheme 1). Treatment with butyllithium, followed by acetone, gave two products, one unstable to heat, when it was converted into the other. The more stable product had the formula C,,H,,N from its mass spectrum and microanalysis, and the 'H NMR spectrum showed, in addition to the normal azepinoindole signals, a methyl singlet at 6 2.20 and two alkene multiplets (each 1 H) at 6 5.00 and 5.30, indicating that the product was an 1l-isopropenylazepino-indole, 9.The second product was shown by microanalysis and 'H NMR spectroscopy to be the alcohol 10. Other carbonyl compounds gave uncharacterised products. The generality of electrophilic attack at position 1 1 was tested with other reagents. Vilsmeier-Haack formylation gave a single product, the 11 -formy1 derivative, 11, in 76 yield. The absence of a signal at 6 6.20 confirmed 11 substitution, and there was an aldehyde signal at 6 10.15. The modified Vilsmeier reaction with dimethylacetamide gave 1 1 -acetylazepinoindole 12. Reaction with N-benzoylmorpholine and phosphoryl chloride gave a very small yield of the 11-benzoylazepinoindole 13, identical with a sample prepared by treating compound 7 with benzoyl chloride and tin(1v) chloride (41 yield).The acetyl derivative 12 gave with methylmagnesium iodide the same mixture of alkene 9 and alcohol 10 as was obtained from the 1 1-lithio derivative and acetone. A Mannich reaction gave 11-dimethylaminomethylazepinoindole 14 (79 yield), which gave the methiodide 15 in quantitative yield. Thermolysis of the quaternary hydroxide gave only the amine 14. The instability of the bromo compound 8 limited our access to 1 1-substituted azepinoindoles, so we investigated the possibility of direct lithiation. No anion was formed when azepinoindole 7 was treated with butyllithium, lithium diisopropylamide (LDA)or butyllithium-N,N,N', N'-tetramethylethylenediamine (TMEDA), but with tert-butyllithium at -78 OC a deep green solution formed, and subsequent treatment with trimethylsilyl chloride gave a single crystalline product in 82 yield.Microanalysis and its mass spectrum showed it to be a monotrimethylsilyl derivative, but the 'H NMR spectrum revealed considerable differences from that of the 1OH-azepinoindole 7. In addition to the singlets at 6 0.0 (9 H, Me3Si) and 6.30 (1 H, 11-H) there was only a single proton (pseudotriplet) at 6 2.55, well upfield of the normal 10-H position. The presence of two doublets, each 1 H, at 6 6.15 and 204 J. CHEM. soc. PERKIN TRANS. I 1995 COR 23' 6' 7 16 7 11, Ill e 15 t 10 Scheme 1 Reagents and conditions: i, N-Bromosuccinimide, CCI,; ii, BuLi; iii, MeCOMe; iv, R',CONR, POCI,,or v, RCOCI, SnCI,; vi, Me,NH, HCHO, AcOH; vii, MeI; viii, Ag,O, heat 6.65 (assigned to 6-H and 10-H), allows only one structure, the 8-substituted 8H-azepinoindole 16 (Scheme 2).We have previously seen only one euro;iH-compound, among those formed when 2-azido-2',4',6'-trimethyldiphenylmethanewas decom-posed,3 and it is notable that the 8-methylazepinoindole prepared by nitrene insertion has the 10H-structure. Compound 16 was relithiated with tert-butyllithium and, with excess trimethylsilyl (TMS) chloride gave a single product with two TMS substituents. From the 'H NMR spectrum this was shown to be the 8,10-disubstituted 10H-azepinoindole 17 (all NMR spectra for 1OH-azepinoindoles are collected in Table 1, all 8H in Table 2).An X-ray structure on compound 17 confirmed the assignment (Fig. 1). Compound 17 could also be prepared directly from azepinoindole 7 by using a large excess of tert-butyllithium, the TMS chloride being added before or after lithiation. No further lithiation occurred and no trisubstituted products were ever observed. The lithiation reaction provides the first direct functionalisation of the seven-membered ring in azepinoindoles, and we next examined the reaction of the lithiated azepino- indole 7 with alkyl halides. Quenching the anion with methyl iodide gave a mixture of equal amounts of two products, both monomethylazepinoindoles. Their behaviour on chromato-graphy was very similar; pure samples could only be obtained by crystallisation of fractions from the front and rear of a single broad band from a Chromatotron.The less polar product had an NMR spectrum identical with that of 10-methyl-IOH-azepinoindole 18, while the more polar material had an NMR 17 Scheme 2 Reagents andconditions: i, Bu'Li, THF, -78 'C; ii, Me,SiC1 Fig. 1 X-Ray crystal structure of compound 17 spectrum very similar to that of compound 16 and was the 8-methyl-8H isomer 19 (Scheme 3). The spectrum of this isomer showed a number of small long-range couplings, from 6-H and 10-H to 8-H, and from 7-H to 9-H, similar to those shown by azepinoindol-8-one. 7 18 R-Me 19 R=M" 20 R=EI fl R=Et 14 IH, h, 22 R=PS 24 R-Bu' 23 REPS 25 R=Bd Scheme 3 Reagents and conditions: i, Bu'Li, THF, -78 "C; ii, R-X;iii, BuLi; iv, Bu'I Further experiments using ethyl iodide, isopropyl bromide or iodide, and tert-butyl bromide or iodide, gave in all cases equal mixtures of the 10Hisomers 20,22 and 24 with the 8Hisomers 21,23 and 25.Only compound 25 was not obtained analytically pure, but all isomers were separated by GC-MS, and mass spectra determined for all. The observation that the anion from the 8-trimethylsilyl derivative 16 gave only one product with TMS chloride suggested that similar clean reactions might be achieved with alkyl halides, but this was not so. When the anion from compound 16 was treated with methyl iodide, a mixture of two compounds of very similar behaviour on chromatography was obtained.From the mixture only one compound could be 80 Table 1 'H NMR shifts (ppm)and coupling constants of new 10H-azepinoindoles CLComp. I-H 2-H 3-H 4-H 6-H 7-H 8-H 9-H 10-H 11-H Other J/Hz W bsol;D VI -8 7.15-7.60 (m) 5.75 (m) 6.10 (m) 5.90 (m) 3.50 (2 H, d) 9 7.1C7.70 (m) 5.80 (m) 6.05 (m) 5.90 (m) 3.55 (2 H, d) 2.20 (3 H, s), 5.00 (1 H, m), 5.30 (I H, m) 10 7.00-7.90 (m) 6.05 (m) 5.80 (m) 5.95 (m) 2.70 (2 H, d) 1.75 (6 H, s), 1.85 (1 H, br s, OH) 11 8.30 (m) 7.20-7.40 (m) 6.10 (m) 5.85 (m) 3.65 (2 H, d) 10.15 (s, CHO) 12 8.00 (m) 7.25-7.50 (m) 6.15 (m) 6.00 (m) 3.90 (2 H, d) 2.70 (3 H, s, CH,CO) 13 7.20-7.80 (m) 6.15 (m) 5.80 (m) 3.40 (2 H, d) 7.2c7.80 (m, C6H,) 14 7.0C7.60 (m) 5.65 (m) 5.95 (m) 5.80 (m) 3.40 (2 H, d) 2.20 6 H, s (CH,),N, 3.50 (2 H, s, CH,N) 17 7.0C7.45 (m) 6.96 (d) 5.55 (d) 6.00 (d) 3.35 (1 H, d) 5.95 (s) 0.05 (9 H, s), 0.15 (9 H, s) 18 7.10-7.55 (m) 5.80 (m) 5.95 (m) 5.60 (m) 3.35 (1 H, d) 6.15(s) 1.60(3H,d) 20 7.1 1-7.25 (m) 6.90 (d) 5.50 (m) 5.75 (m) 5.40 (m) 3.05 (1 H, d) 6.00 (s) 0.95 (3 H, overlapping t), 1.78 (2 H, m) 22 7.1g7.60 (m) 6.15 (s) 0.80, 1.10 (3 H, overlapping d), 2.30 (1 H, m) 24 7.40 (m) 7.1C7.30 (m) 7.55 (m) 5.85 (m) 6.00 (m) 5.60 (m) 3.50 (1 H, m) 6.25 (s) 1.05 (9 H, s) 31 7.00-7.50 5.90 (d) ~ 6.LO (d) 3.20 (1 H, m) 6.20 (s) 0.00 (9 H, s) 35 6.9C7.10 (m) 7.40 (m) 6.80 (d) 5.25--5.25-3.00 (I H, m) 5.90 (s) 0.80(6 H,d), 2.05 (1 H, hept) 5.35 (m) 5.35 (m) 39 7.40 (m) 7.10-7.30 (m) 7.55 (m) 7.1amp;7.30 (m) 5.85 (d) 6.25 (t) 3.45 (2 H, d) 6.20 (s) 44 7.60(d) -7.60(d) -8.85 (d) -~-3.55 (2 H, s) 6.50 (s) 3.50 (3 H, s) ~45 7.50(d) 7.60(d) -8.55 (s) ~ 5.15 (t) 3.30 (2 H, d) 5.90 (s) 3.65 (3 H, s) ~48 7.75 (d) 7.55 (s) -8.25 (d) 5.80 (d) 4.95 (t) 3.30 (2 H, d) 6.10 (s) 3.50 (3 H, s) (m 'H Z) 58'2 (PP)SZ'L (s 'H 6) 00'0 (P) 02'9 (m 'H Z) 00'2 '(1 'H f) 00' I '(s 'H 6) 00'0 (P) 02'9 (S 'H E) Of' I '(S 'H 6) 00'0 (P)s 1'9 (m 'H I)58'1 '(P 'H 9) 56.0 (P) OS'9 (m 'H Z) 09'1 '(1 'H f) 00.1 (P) OS.9 (P 'H f) SZ'I (P)SV9 (S 'H 6) SO0 (m) 09'5 (m) SS'Z (W) 6-P (PI 59'9 (m)OS'L (m) SZ'L3O.L zH/f WIO H-II H-01 H-6 H-8 H-L H-9 H-b H-f H-Z ~-1 punodmo3 207J.CHEM.SOC. PERKIN TRANS. 1 1995 7 36 26 R=Me 27 R=Me 28 R=Et 29 R=Et 30 R=PS 31 R= Pt/l6 32 R-Bd 33 R=B~ 1 34 R=Me R 35 R = PI' Scheme 4 Reagents and condirions: i, Bu'Li, THF.-78 "C; ii, R-X; iii, Bu,N * F ~ isolated crystalline, and the NMR spectrum showed a singlet methyl signal at 6 1.30 and the TMS signal, but no other sp3 protons. The similarity of the rest of the spectrum to that of the 8H isomers shows that this compound is the 8,8-disubstituted derivative 26 (Scheme 4). Inspection of the NMR spectrum of the mixture shows a pair of overlapping triplets at S 3.45, which can be confidently assigned to the 8,10-disubstituted 10H isomer 27. Ethyl iodide gave also a mixture of approximately equal amounts of two products, of which the 8,8-disubstituted isomer 28 could be isolated, and the 8,lO isomer 29 identified from the NMR spectrum of the mixture.Isopropyl iodide and tert-butyl iodide gave mixtures that could not be separated, but linked GC-MS showed the presence of two isomers in each case. compounds 30-33 (with mass spectra for each), and the NMR spectra left no doubt that they had the structures proposed. Treatment of compound 26 with tetrabutylammonium fluoride (TBAF) at room temperature gave very rapidly the known 8-methylazepinoindole 34; under similar conditions compound 30 gave 8-isopropylazepinoindole 35. We have reported that flash vacuum pyrolysis of azepinoindole gave a mixture containing the 6H isomer; we have heated the three 8H tautomers 16, 19 and 21, and the 10H tautomers 17,22 and 24, for extended periods in deuteriated toluene at 105 "C, but no changes were observed. Attempts to obtain an 8H or 10H bromo-or chloro-azepinoindole by reaction between the lithiated compound 7 and bromine, or 1,2-dibromotetra-chloroethane. or hexachloroethane were unsuccessful.Dialkyl- amino groups have been used to direct lithiation.' The Mannich compound 14 reacted more easily with lithiating agents than compound 7. butyllithium causing rapid formation of a lithium derivative. but treatment of the lithium derivative with tert-butyl iodide again gave a mixture of 8-and 10-substituted compounds, with no degree of regioselectivity. We believe that deprotonation of compound 7 gives the allylic anion 36, and that the tautomer which forms when this anion is attacked by an electrophile is stable under the conditions of reaction and work-up. We see no indication of further extension of delocalisation to position 6; no 6-substituted products have been observed and such a fully delocalised azepine would be anti-aromatic if planar.Our anion 36 is stable up to 0 OC. The only unexplained discrepancy is the regiospecificity shown by TMS chloride, but this may indicate kinetic control. Our final efforts to produce the azepinoindolium carbocdtion were aimed at the synthesis of bromoazepinoindoles with the bromine in the seven-membered ring, starting from suitable brominated aminobenzophenones. The 4'-bromo derivative 37 was prepared from acetanthranil and the mono Grignard reagent from 1,4-dibromoben~ene.~ Reduction of benzophen- one 37 by our usual procedure (sodium in ethanol) removed the 4037 R=Br Ii Br 4238 R=Br 41 R=OMe ii, iii R = BrI lii Bryamp;rJBrOMe Br 39 43 Scheme 5 Reagenfsandcondiliuns: i, LiAIH,, AICI,: ii, HNO,, NaN,; iii, TCB, 165 "C; iv, Br,, 3 mol bromine to give 2-aminodiphenylmethane, but reduction with lithium aluminium hydride and aluminium chloride gave the amine 38 (Scheme 5), converted into the azide by a standard procedure.Pyrolysis of the azide in boiling trichlorobenzene gave a single crystalline product, which from its analysis, mass spectrum, and NMR spectrum was the expected 8-bromo-1OH- azepinoindole 39. We were unable to remove bromide, using antimony(v), so the amount of tautomeric 8H isomer must be vanishingly small.A second possibility seemed to be to use tetracyclic compounds of type 40similar to products which we have found from some of our nitrene insertions; attempts to add dibromocarbene to 10-methylpyridoindole under a number of different conditions failed. We have made a single attempt to obtain a suitable cyclo- propapyridoindole via a nitrene insertion. Bromination of 2-amino-4'-methoxydiphenylmethane41 in concentrated hydro- bromic acid gave three products; increasing the amount of bromine maximized the yield of one product, a tribromo derivative. The 'HNMR spectrum showed that there were no hydrogen atoms ortho-or para-to the amino group, and that the third bromine atom was in position 3', giving formula 42.Since substitution on the aniline ring does not affect the final nitrene insertion we converted the amine 42 into the azide 43 and decomposed the azide at 165 "C in trichlorobenzene, giving a mixture of at least seven compounds, and some high molecular weight material. By careful chromatography it was possible to obtain three main bands (A, B and C), and from these to isolate three compounds analytically pure, and another four in sufficient purity to allow identification by spectroscopy. The first pure compound from band A was a tribromomethoxy- azepinoindole. The 'H NMR spectrum showed an intact methylene group (10-H) at S 3.75 as a singlet, indicating substitution, presumably by bromine, at C-9, confirmed by a pair of doublets at 6 8.40 and 5.90 (J 9.5 Hz) assigned to 6- J.CHEM. SOC. PERKIN TRANS. 1 1995 52 48 55 L H+, H20/ 1 OMe -BrTpamp;Br Br ' OMe 44 R'-BrmR* 0;Br 49 R' = Br, R2= OMe 45 50 R' = Meo. R2= Br 46 R=H 47 R=Br Scheme 6 and 7-H. Hence this compound is the 2,4,9-tribromo derivative Models are available in the 6,7-dihydro 52 and in the 9,lO-44.The second crystalline compound from band A was an dihydro derivative 53.3Comparison of the chemical shifts of the isomer of compound 44,with a methylene doublet at S 3.30 two methylene signals in our new compound with those and a singlet (1 H) at 6 8.85 assigned to 6-H and hence reported showed it to be compound 54. These compounds are establishing structure 45.The third compound in the fast summarized, together with our interpretation of their mode of moving band A was identified as the cyclopropapyridoindole formation in Scheme 6. 47, by its very characteristic set of three NMR signals (each 1 H) Some brief comment on Scheme 6 is necessary. No bromo at 6 0.90 (HA),1.95 (HE) and 5.00 (Hc), very similar to those in derivative 55 suitable for conversion into an aromatic system compound 46,6but with a downfield shift in HC.This, and the was obtained. Insertion of the intermediate nitrene into the proves the presence of the adjacent ring could give a spirodiene intermediate," fromabsence of long range coupling in HC, bromine at C-7. From band B two compounds were isolated. which either acridine 49 or 50 could be obtained.Ring The first, obtained pure, was a dibromoazepinoindole, 48, with expansion gives azepinoindoles 44 and 45 and hydrolysis of an NMR spectrum very similar to that of S-methoxy-compound 44 during work-up produces a bromo ketone well the two remaining bromine atoms being on the set up for elimination of hydrogen bromide to give azepino- a~epinoindole,~ benzene ring. The second compound from band B was bright indolone 51. yellow, and highly fluorescent in ultraviolet light. In the NMR We have postulated radical intermediates during formation spectrum all signals were in the aromatic region apart from the of azepinoindoles from cyclopropapyridoindoles.6The isomer methoxy singlet. This product is formulated as an acridine, 55 which we sought, is well set up for radical abstraction of although we cannot distinguish between structures 49 and 50.bromine, to give an azepinoindolyl radical (centred on C-8 and From band C two compounds were obtained; the first was C-lo), which by hydrogen abstraction produces compound 48 bright orange, and had a pattern of signals in the NMR and hence, by hydrolysis the dihydro derivative 54. spectrum very similar to that of azepinoindol-amp;one modified only by the presence of two bromine atoms on the benzene ring; X-Ray Crystal Structure of Compound 17.-The first report hence this compound is the azepinoindol-8-one 51. The last of the formation of an azepinoindole by nitrene insertion from compound isolated was a dibromodihydroazepinoindol-8-one. an azidodiphenylmethane wrongly ascribed an 1 1 H structure, J.CHEM. SOC. PERKIN TRANS. 1 1995 corrected to the accepted 1OH structure 7 on the basis of 'H NMR data. Until now there has been no confirmation of the structure by X-ray diffraction, but with the bis-TMS compound 17 (Fig. 1) it can clearly be seen that the previous suggestion was correct. The axial position of the trimethylsilyl group in the 10 position appears unusual, but there are in fact no bad non-bonded interactions. Experimental M.p.s were determined on a Kofler heated stage, and are uncorrected. NMR spectra were determined on a JEOL 270 MHz spectrometer, for solutions in CDCl,, from tetramethyl- silane (TMS) as standard. S Values are in ppm and J values in Hz.Linked GC-MS determinations were performed on a Hewlett-Packard HP5890 chromatograph coupled to an HP5970 mass sensitive detector, controlled by an HP series 300 computer and HP5970C Chemstation software. UV and visible spectra were recorded on a Varian DMSlOO spectrometer for solutions in 95 ethanol. Chromatography was performed on Aldrich alumina, deactivated to Brockmann grade IV, and Chromatotron separations used plates with 2 mm of PF254 silica gel. Solvent mixtures were of petroleum (b.p. 60-80 "C) and ethyl acetate, proportions given thus (60 :40). Ether refers to diethyl ether. 1OH-Azepino1,2-aindole 7.-The overall procedure was as described previously '* but it was observed that crude material from the reduction of 2-aminobenzophenone (2-aminodiphen- ylmethanone) was sufficiently pure to use in the preparation of azide, which in turn could be used crude for pyrolysis boiling trichlorobenzene (TCB) under argon.Thus, from 20 g of aminobenzophenone 13 g of azepinoindole 7 (72 yield) was obtained after chromatography but before recrystallisation. The product was pure by 'H NMR, and could be used as such in most reactions. 1 1-Bromo-lOH-azepino 1,2-aindole 8.-A solution of aze- pinoindole 7 (0.4 g) in carbon tetrachloride (25 cm3) with N-bromosuccinimide (0.4 g) was stirred in an ultrasonic bath under nitrogen. Progress of the reaction was monitored by NMR spectroscopy (disappearance of the signal at SH 6.15). When the reaction was complete the solution was filtered through Celite and the solvent removed under reduced pressure, keeping the temperature below 20 "C to give 0.44 g of product; any attempt at purification or increase in temperature led to decomposition; m/z 261 (M' + 2,84) and 259 (M', 84), 180 (M' -Br, IOO), 165 (39), 152 (34) and 143 (49). 1l-lsopropenyl-lOH-azepinol,2-aindole9 and 11-(2-Hy-droxypropan-2-yr)- IOH-azepino 1,2-aindole 10.-To a stirred solution of the bromide 8 (0.44 g) in anhydrous ether (25 cm3) at -50 "C: under argon, was added, dropwise, butyllithium (1.2 mol dm-3 in hexane; 2 cm3).After a further 20 min, anhydrous acetone was added, and the mixture allowed to warm to room temperature and then treated with saturated ammoniacal ammonium chloride (10 cm3).The organic layer was separated, the aqueous layer further extracted with ether (3 x 25 cm3), the organic extracts dried (MgSO,), filtered, and then evaporated. The residual oil (0.33 g), was chromato- graphed; elution with petroleum (b.p. 60-80deg;C) gave the isopropenyl compound9, b.p. 160 "C/0.02 mmHg (bulb to bulb) (Found: C, 86.6; H, 6.8; N, 6.55. C,,H,,N requires C, 86.85; H, 6.85; N, 6.35); m/z 221 (M', 6873, 180 (M' -C3H5, loo), 165 (44)and 143 (42). Further elution (85:15) gave the alcohol 10. Attempts to distill the alcohol gave mixtures of alkene and alcohol (Found: C, 80.05; H, 7.2; N, 4.55. C1,HI7NO requires C, 80.4; H, 7.1; N, 5.85). 1l-Formyl-lOH-uzepino 1,2-aindole 11 .-Phosphoryl chlor-ide (3.1 g, 22 mmol) was added dropwise over 10 rnin to stirred anhydrous dimethylformamide (DMF) (10 cm3), at -5 "C, under nitrogen.The mixture was stirred (15 min) and then a solution of azepinoindole 7 (3.3 g, 18.5 mmol) in DMF (5 cm') was added dropwise over 20 min, keeping the temperature below 5 "C. The red solution was heated at 40 "C (3 h) and then evaporated under reduced pressure to reduce the volume by 70.A large excess (-150 cm3) of saturated aqueous sodium carbonate was added to the mixture to give a precipitate. The mixture was extracted with ether, dried (MgSO,), decolourised with charcoal and then filtered to give, after evaporation, a pale yellow solid. Recrystallisation from ethanol gave the aldehyde 11, m.p. 130deg;C (Found: C, 80.2; H, 5.05; N, 6.6.CI4H,,NO requires C, 80.35; H, 5.3; N, 6.9); vmax(CC14)/cm-1 1664, 1462, 1180 and 1052; A,,,/nm 321 (log,, E 4.19), 244 (4.42) and 219 (4.65); m/z 209 (M', 91), 208 (53), 180 (M' -CHO, 100) and 152 (20). 11-Acetyl-lOH-azepino 1,2-aindole 12.-Phosphoryl chlor-ide (3.1 g, 22 mmol) was added dropwise over 10 rnin to stirred, anhydrous N,N-dimethylacetamide (DMA) (15 cm3) at -10 "C under nitrogen. After 20 rnin azepinoindole 7 (2.5 g, 14 mmol) in anhydrous DMA (10 cm3) was added dropwise, keeping the temperature below 0 "C, and the mixture then stirred at 40 "C (48 h). The cooled mixture was quenched with approx. 100 cm3 of saturated aqueous sodium carbonate, giving a dense white precipitate. The mixture was extracted with chloroform (3 x 75 cm3), dried (MgSO,) and then evaporated to give a yellow solid.Chromatography, eluting with petroleum, b.p. 60-80 "C gave starting material (1.6 g) and the acetyl compound 12, m.p. 116deg;C (0.5 g after sublimation, 16 based on unrecovered starting material) (Found: C, 80.6; H, 5.9; N, 6.1. C15Hi3N0 requires C, 80.75; H, 5.8; N, 6.25); v,,,(CCl,)/cm-' 1650, 1460, 1410 and 716; 1,,Jnm 321 (log,, E 4.05), 241 (4.29) and 224 (4.28); m/z 223 (M', 88), 208 (M' -CH,, 44),180 (M+ -CH3C0, 100) and 166 (25). 1l-Benzoyl-IOH-azepino1,2-aindole A solution of 13.4~) N-benzoylmorpholine (2.5 g, 13 mmol) in phosphoryl chloride (3 cm3) was stirred at 0 "C (24 h). A solution of indoloazepine 7 (2.0 g, 11 mmol) in 1,Zdichloroethane (40 cm3) was added to it, stirring continued (3 h), and then saturated aqueous sodium carbonate (75 cm3) added.The resultant green-black suspension was shaken (15 min) and then boiled (15 rnin). Aqueous sodium hydroxide (2 mol dm-3; 50 cm3) was added to the hot solution, and the mixture stirred until it reached room temperature. Ether was added to the mixture and then it was filtered. The aqueous phase was extracted with further ether and the combined organic extracts were worked-up as described for compound 12. Recrystallisation from ethanol gave the benzoylderiuatiue13 (0.13 g, 3), m.p. 223 "C (Found: C, 84.2; H, 5.3; N, 4.75. C2,,Hl5NO requires C, 83.9; H, 5.6; N, 4.9); i,,,/nm 338 (log,, E 3.99), 311 (3.85), 250 (4.36) and 220 (4.48).(b) A solution of benzoyl chloride (1.35 g, 9.5 mmol) and compound 7 (1.5 g, 8.5 mmol) in carbon disulfide (30 cm3) was added dropwise to a cooled (-20 "C) stirred solution of stannic chloride (4 g, 16 mmol) in CS2 (10 cm3) under nitrogen. After 45 rnin a further portion of benzoyl chloride (0.7 g, in 5 cm3 CS,) was added. After a further 15 min the reaction mixture was added carefully to methanol (125 cm3) and then water added until a precipitate formed. The aqueous mixture was extracted with chloroform (3 x 150 an3),the organic extracts dried and the solvent removed to give a residue (3.9 g) which was chromatographed on a column (100 g, 9 :1) to give the benzoyl derivative (1.0 g, 41). 1 1-Dimethylaminomethyl-1OH-azepino1,2-aindole 14.-A mixture of purified dioxane (15 an3),glacial acetic acid (15 cm3), aqueous formaldehyde (1.1 cm3 of a 36 solution) and aqueous dimethylamine (3.2 cm3 of a 25 solution) was placed in a flask fitted with a rubber septum and under a positive pressure of nitrogen.The mixture was stirred in an ice bath (10 min) and then azepinoindole 7 (2.5 g, 13.8 mmol) in dioxane (1 5 cm3) was added via a syringe (25 min), and the mixture was left overnight at room temperature. The mixture was poured into water (150 cm3), filtered through Celite, basified (as. NaOH) and then extracted with ether (3 x 75 cm3). The combined organic extracts were dried (MgSO,), filtered and evaporated, to give a pale yellow oil (2.8 g), Distillation (Kugelrohr) gave the amine 14, b.p.150 "CjO.5 mmHg (2.6 g, 79) (Found: C, 80.55; H, 7.55; N, 1135. C,,H,,N2 requires C, 80.7; H, 7.55; N, 11.75); v,,,(film)/cm~' 2946, 2762, 1634, 1412, 1172 and 1040; i,,,/nm 310 (log,, E 3.89), 273 (4.18) and 227 (4.40). The methiodide, 15, was prepared in acetone and crystallised from methanol, m.p. 219 "C (decomp.) (Found: C, 53.9; H, 5.4; N, 7.15. C,,H2,1N, requires C, 53.7; H, 5.55; N, 7.35); v,,,(Nujol mull)/cm-' 1580, 1560, 858 and 726; i,,,hm 271 (log,, E 4.20) and 221 (4.46). General Procedure for the Direct Lithiation of Azepinoindole 7.-To a cooled (-78 "C) stirred solution of compound 7(0.7 g, 3.8 mmol) in anhydrous THF under argon, was added tert- butyllithium (1.5 mol dm-' in pentane; 2.6 cm').A dark green colour developed over 20 min, and then the co-reagent was added dropwise via a syringe and septum. The reaction mixture was allowed to come to room temperature, quenched with saturated ammonium chloride in methanol (20 cm3) and extracted with ether (3 x 40 cm3). The dried (MgSO,) ether extracts were filtered, evaporated, and the residue purified as described for each product. 8-Trimethylsilyl-8H-azepino1,2-aindole 16. The residue was crystallised from ethanol to give the trimethylsilyl derivative 16, m.p. 91.5-92.5"C (82) (Found: C, 76.2; H, 7.7; N, 5.4. CI6Hl9NSi requires C, 75.9; H, 7.5; N, 5.55); vmaX(CCl4)/cm-' 1665, 1517, 1204, 740 and 634; i,,,/nm 310 (log,, E 4.22), 263 (4.43) and 224 (4.37); mlz 253 (M', 78), 180 M' -(CH3)$i, 1001 and 73 Si(CH,),, 1001.1O-Metlzyl-lOH-azepino 1,2-aindole 18 and 8-methyMH-azepino 1,2-aindole 19. The co-reagent was methyl iodide and separation was by Chromatotron (petroleum, b.p. 60-80 "C). The early fractions gave the 10-methyl derivative 18, which was recrystallised from ethanol, m.p. 59 "C (lit.,, m.p. 59 "C). Recombination of the later bands from the Chromatotron, followed by a further passage across the Chromatotron, gave later fractions sufficiently pure for purification by distillation (Kugelrohr) to give the 8-methyl derivative 19, b.p. 160 'C/0.04 mmHg (Found: C, 86.55; H, 6.65; N, 6.8. C,,H,,N requires C, 86.15; H, 6.65; N, 7.15); imax/nm 323 (log,, E 4.09), 261 (4.16) and 222 (4.34); mjz 195 (M', 66) and 180 (M' -CH,, 100).IO-Ethyl-IOH-azepino 1,2-aindole 20 and 8-ethyl-8H-aze-pinol,2-aindole 21. The co-reagent was ethyl iodide. The separation procedure was as described for compounds 18 and 19 to give the 10-ethyl derivative, m.p. 53 "C (from ethanol) (Found: C, 86.15; H, 7.3; N, 6.55. C15H15N requiresc, 86.1; H, 7.15; N, 6.7); I,,,/nm 275 (log,, E 4.24) and 230 (4.33); rnlz 209 (M', 4473, 180 (M' -C,H,, 100). From the later Chromatotron fractions the 8-ethyl derivative was isolated, b.p. 140 "CjO.05 mmHg (Kugelrohr) (Found: C, 86.1; H, 7.5; N, 6.5. C,5H,5N requires C, 86.1; H, 7.15; N, 6.7), imaX/nm 323 J. CHEM. SOC. PERKIN TRANS. I 1995 (log,, E 4.12), 265 (4.20) and 219 (4.40); m/z 209 (M', 58) and 180 (M* -C,H,, 100).1O-Isopropyl-lOH-azepino1,2-aindole 22 and 8-isopropyl- 8H-azepinoCI ,2-aindok 23. The co-reagent was isopropyl bromide or iodide. The separation procedure was as described for compounds 18 and 19 to give the 10-isopropyl derivative 22, m.p. 56deg;C (from ethanol) (Found: C, 85.9; H, 7.7; N, 6.4. C16H17N requires C, 86.1; H, 7.6; N, 6.3); amp;,,Jnm 311sh. 277 (log,, E 4.21) and 228 (4.31); m/z 223 (M+, 25) and 180 (M' -C3H7, 100). From later fractions the 8-isopropyl derivative 23 was isolated, b.p. 145 "CjO.04 mmHg (Found: C, 85.75; H, 7.7; N, 6.35. C16Hl,NrequiresC, 86.1; H, 7.6; N, 6.3); m/z 223 (M+, 20) and 180 (M+ -C,H,, 100). 10-tert-Butyl-lOH-mepino 1,2-aindole 24 and 8-tert-butyl- 8H-azepino 1,2-aindole 25.The co-reagent was tert-butyl bromide or tert-butyl iodide. The lithiation mixture had to be raised to room temperature to complete reaction. The separation procedure was as described for compounds 18 and 19 to give the 10-tert-butyl derivative 24, m.p. 141-142deg;C (Found: C, 86.3; H, 8.4; N, 5.9. CI,H,,N requires C, 86.05; H, 8.05; N, 5.8); i,,,jnm 319 (log,, E 3.90), 274 (4.23) and 228 (4.37); m/z 237 (M', 9), 180 (M' -C,H,, 100) and 152 (7). The 8-tert-butyl derivative 25 was characterised only from its NMR spectrum and mass spectrum, m/z 237 (M', 10) and 180 (M' -C,H,, 100). General Procedure for the Lithiation of 8-Trimethylsil~vl-8H- azepino 1,2-aindole 16.-To a stirred, cooled (-78 "C) solution of the trimethylsilylazepinoindole 16 (0.5 g, 2 mmol) in anhydrous THF (7 cm') was slowly added tert-butyllithium (1.5 mol dm in pentane; 6 mmol).After 15 min alkyl halide (10 mmol) was added dropwise, and stirring continued while the flask reached room temperature. Following work-up with methanolic ammonium chloride, the crude product was passed across a Chromatotron, giving a single broad band. In the case of the methyl and ethyl derivatives, collection of early portions of this broad band gave one component sufficiently pure to allow crystallisation from ethanol. In the case of the isopropyl and tert-butyl derivatives no single pure fraction was obtained. The mixture of isomers was distilled (Kugelrohr) and individual components identified by 'H NMR and by GC-MS.8,1O-Bis(trimethylsi~yl)-1OH-azepino1,2-aindole17. Com-pound 17 was obtained in 88 yield by the general procedure, m.p. 125-127 "C (from ethanol) (Found: C, 70.05; H, 8.6; N, 4.15. C,,H,,NSi, requires C, 70.15; H, 8.3; N, 4.3); R,,,/nm 261 (log,, E 4.19) and 230 (4.32); mjz 325 (M', 4) and 252 (M+ -C,H,Si, 100). 8-Methyl-8-trimeth,~l.~iI~~l-8H-azepino1,2-aindole 26. Re-crystallised from ethanol, m.p. 140-144 "C (Found: C, 75.9; H, 8.1; N, 5.0. C,,H,,NSi requires C, 76.3; H, 7.85; N, 5.25); /Imax/nm 308 (log,, E 4.26), 266 (4.36) and 222 (4.39); miz 267 (M', 5), 252 (M' -CH,, 4) and 195 (M' -C3H,Si, 100). 8-Eth~~l-8-trimeth~lsily/-8H-azepino1,2-aindole 28. Recrys-tallised from ethanol, m.p. 140-144 "C (Found: C, 76.4; H, 8.25; N, 5.15.C18H23NSi requires C, 76.85; H, 8.2; N, 5.0); m/z 281 (M', 4) and 209 (M' -C,H,Si, 100). 8-Isopropyl-8-trimethylsilyl-8H-azepinol,2-aindole30 and 10-isopropyl-8-trirnethylsilyl-1OH-axpino 1,2-aindole 31. Formed as an inseparable mixture, which on distillation (Kugelrohr) gave microanalytical figures C, 76.8; H, 8.15; N, 4.9. C,,H,,NSi requires C, 77.3; H, 8.45; N, 4.75. Two peaks on GC yielded m/z values of 295 (M', 4) and 223 (M' -C3H,Si, 100) and 295 (M', 9) and 223 (M' -C,H,Si, IOO), respectively. 8-tert-Butyl-8-trimethylsilyl-8H-azepino1,2-aindole 32 and 1O-tert-butyl-8-trimethyl~~lyl-1 OH-azepino 1,2-aindole 33. Ob-tained as a mixture, inseparable by chromatography. A sample J CHEM.SOC. PERKIN TRANS. 1 1995 distilled (Kugelrohr) had microanalytical figures C, 77.35; H, 8.75; N, 4.6. C2,H2,NSi requires C, 77.3; H, 8.45; N, 4.75. Two peaks on GC yielded m/z values of 309 (M', 2), 252 (M' -C,H9, 75) and 236 (M' -C,H,Si, 100) and 309 (Mi, 4), 252 (M+ -C,H,, 70) and 236 (M' -C,H,Si, loo), respectively. amp;Methyl- 10H-azepinoC1 ,2-aindole 34.-A solution of tetra- butylammonium fluoride (1 mol dm-3 in THF; 0.05 mmol) was added to a solution of compound 26 (0.05 mmol) in anhydrous THF (2 cm3) at -78 "C. After 1 min the reaction mixture was taken up in ether (15 cm3), washed with water (3 cm3), dried (MgSO,), filtered, and then evaporated. The residue was almost pure methyl derivative 34,identified by comparison of its spectra with those of an authentic specimen.' Similarly, from compound 28, was obtained the 8-isopropyl derivative 35 (NMR spectrum in Table I).2-Arnino-4'-brornodiphenylmethanone37.-The Grignard re- agent from 1,4-dibromobenzene (78 g, 0.3 mol) and magnesium (8 g, 0.3 mol) prepared as described by Schiemenz9 was slowly added to a vigorously stirred suspension of acetanthranil(53 g, 0.3 rnol) in a mixture of toluene (400 cm3) and ether (300 cm3) at 0 "C. The mixture was warmed at 30 "C with stirring (2 h). Dilute hydrochloric acid (2 mol dm-3; 500 cm3) was added slowly and the mixture stirred (10 min) and then the organic layer was separated. Extraction (ether, 2 x 200 cm3) was followed by evaporation of the combined organic layers to give a yellow oil, which was dissolved in a mixture of concentrated hydrochloric acid (150 cm3) and ethanol (300 cm3).The solution was boiled (2.5 h), cooled, and then basified with 10 aqueous sodium hydroxide. Extraction with ether (4 x 250 cm3) and removal of the solvent from the dried, filtered solution, gave a red oil, which solidified and was recrystallised from ethanol. to give the title compound 37 (44 g, 48). (Bergmann and Barshai reported a 25 yield of the N-acetyl derivative).', 2-Amino-4'-brornodipheriylmethane 38.-Lithium aluminium hydride (8 g) was added slowly to an ethereal solution of freshly sublimed aluminium chloride (40 g in 150 cm3), stirred under a nitrogen atmosphere. A solution of benzophenone 37 (20 g) in the minimum amount of anhydrous ether was added dropwise to the hydride solution, and the mixture vigorously stirred (1 h).Careful addition of damp ether, then water (200 cm3) was followed by separation of the ether layer, and further extraction of the aqueous layer with ether (3 x 100 cm3). The combined extracts were dried (MgSO,) and after evaporation of the filtered solution the residue was chromatographed. Elution with petroleum (b.p. 60-80 "C) gave as the slower running of two bands the bromo amine 38 (9.7 g, 47), sufficiently pure for conversion into the bromoazepinoindole 39. 8-Brorno-1 OH-uzepino 1,2-aindole 39.-(a) Sulfuric acid (4 mol dm 3: 50 cm3) was added slowly to a solution of bromo amine 38 (4.1 g, I5 mmol) in dioxane (50 cm3), and the resultant solution was cooled to -10 "C.A solution of sodium nitrite (1.2 g, 17 mmol) in water (1 0 cm3) was added to the reaction mixture at -10 "C, and stirring was continued (20 min) before slow addition of sodium azide (1.3 g, 20 mmol in 10 cm3 of water). The reaction mixture was stirred at 30 "C (10 min), diluted with water (300 cm3), and then extracted with ether (3 x 100 cm3). The combined ether extracts were washed with sodium hydroxide (50 cm3 of 5 aqueous solution), dried (MgSO,). and evaporated, keeping the temperature below 35 "C. The azide was decomposed directly in boiling TCB (400 cm3). Removal of solvent and column chromatography gave 8-bromoazepinoindole39, m.p. 137.5-138.5 "C(2.3 g, 56xfrom the amine38)(Found: C, 60. I; H, 3.75; N, 5.15.C,,H,,BrNrequires 211 C, 60.0; H, 3.85; N, 5.4); amp;,,x/nm 322 (log,, E 4.00), 269 (4.34), 224 (4.50) and 205 (4.44); m/z 261 (M' + 2, 2073, 259 (M', 20) and 180 (M+ -Br, 100). 2-Amino-3,3',5-tribromo-4'-methoxydiphenylmethane42.-To a solution of the amine 41 (6.3 g, 30 mmol) in concentrated hydrobromic acid (50 cm3) was added bromine (15.1 g, 90 mmol). The reaction mixture was stirred in the dark (48 h) at which point a sample basified showed a single major product. The whole reaction mixture was basified (4 mol dm-3 NaOH) and extracted with ether (3 x 75 cm3), the combined extracts were dried (MgSO,), and the filtrate decolourised using charcoal. Removal of solvent gave a yellow oil, which slowly solidified and was crystallised from ethanol to give the tribromo amine 42, m.p.178-183 "C (Found: C, 37.75; H, 2.65; N, 3.05. C1,H,,Br3N0 requires C, 37.35; H, 2.65; N, 3.10); m/z 453 (M' + 6, 33x1, 451 (M+ + 4, loo), 449 (M' + 2, IOO), 447 (M', 32), 372 (28), 370 (66), 368 (M+ -Br, 39, 210 (58), 195 (35), 167 (67), 139 (48), 105 (47) and 77 (64). 2-Azido-3,3',5-tribromo-4'-metho,~~diphenyimethane 43.-Prepared from amine 42 as described in the preparation of 8-bromoazepinoindole (above) to give the azide 43, m.p. 74-75 "C (fromethanol) (Found: C, 35.2; H, 2.0; N, 8.75. Cl,H,,Br3N30 requires C, 35.30; H, 2.10; N, 8.85); v,,,(KCI disc)/cm-' 21 12, 1490, 1436 and 1254; A,,,hm 209 (log,, E 4.09). Decomposition of 2-Azido-3,3',5-tribromo-4'-metho.uydi-phenylmethane 43.-A solution of the azide 43 (1.0 g) in TCB (5 cm3) was added dropwise to stirred TCB (150cm3) at 165 "C under argon.Heating at 165 OC was continued (4 h), and then the cooled solution was evaporated under reduced pressure. The residue was adsorbed on to alumina, applied to a column of alumina, and eluted with petroleum (b.p. 60-80 "C) and then with mixed solvent containing ethyl acetate (up to 40). All the fractions were combined and evaporated to give total non- polymeric products (0.48 g). Passage across a Chromatotron gave three broad bands A, B and C (eluted with petroleum, b.p. 60-80 "C and mixtures with up to 15 of ethyl acetate). Each band was collected in several fractions. Some compounds were crystallised from enriched fractions; compounds are reported in order of elution.Band A.-2,4,9-Trzbromo-8-methoxy-10H-uzepino1,2-a-indole 44.Recrystallised from petroleum (b.p. 60-80 "C), m.p. 179-182 "C (Found: C, 37.5; H, 2.15; N, 3.05. C,,H,,Br,NO requires C, 37.55; H, 2.25; N, 3.15); v,,,(KCI disc),km 1624, 1384, 1230 and 780; j.,Jnrn 314 (log,, E 3.85) 266 (4.16), 237 (4.28) and 205 (4.32). 2,4,7-Tribromo-8-methoxy-lOH-azepinol,2-aindole45. Re-crystallised from petroleum (b.p. 6amp;80 "C), m.p. 162-165 "C (Found: C, 37.9; H, 2.2; N,3.05); v,,,(KCI disc)/cm-' 1638, 1442, 1394, 1134 and 764; 2.,,Jnrn 313 (log,, E 3.78), 275 (4.06), 237 (4.15) and 205 (4.27). The cyclopropapyridoindole 47 had S 0.90 (1 H, m, J 5.5 and 7.0, HA), 1.95 (1 H, m, J 5.5 and 8.2, HB), 3.40 (3 H, s, OCH3),5.00(1H,m,J7.0and8.2,9-H),6.35(1H,s,lO-H),6.90 (1 H, s, 6-H), 7.50 (1 H, d, J 1.5, I-H) and 7.65 (I H, d, J 1.5).Band B.-2,4-Dibromo-8 -methoxy-1OH-azepinuI, 2-a indole 48. Recrystallised from ethanol, m.p. 144147 "C (Found: C, 45.45; H, 2.95; N, 3.6. C,,H,,Br,NO requires C, 45.55; H, 3.0; N, 3.60); vm,,(KCI disc)/cm-' 1648, 1444, 1398, 1228, 1154 and 776; 2.,,,/nm 313 (log,, E 3.74), 264 (4.08), 229 (4.10) and 210 (4.16). Acridine 49 or 50 had 6 4.15 (3 H, s, OCH,), 7.65 (1 H, s), 8.10 (1 H, d, J 1.8), 8.20 (1 H, d, J l.8), 8.25 (1 H, s) and 8.50 (1 H, S, 9-H). Band C.-2,4-Dibromo-8H-azepino1,2-aindol-8-one 51, S 5.95 (1 H, dd, J 1.5 and 10.5, 7-H), 6.40 (1 H, dd, J 1.5 and 11.0, 9-H), 6.95 (1 H, s, 11-H), 7.25 (1 H, dd, J 1.5 and 11.0, 10-H), 7.75 (1 H, d, J2.0,1-H), 7.85 (1 H, d,J2.0,3-H)and 9.20 (1 H, dd, J 1.5 and 10.5, 6-H).2,4-Dibromo-9,1O-dihydro-8H-azepino1,2-aindol-8-one 54, 6 2.85 (2 H, m, 10-H), 3.20 (2 H, m, 9-H), 5.75 (1 H, d, J 10,5,7-H), 6.40 (1 H, S, 11-H), 7.55 (1 H, d,J 1.7,1-H), 7.65 (1 H, d, J 1.7, 3-H) and 8.70 (1 H, d, J 10.5,6-Hj. X-RQYCrystallography.-Crystals of compound 17 suitable for X-ray work were grown from ethanol. Unit cell and intensity data were recorded using a Delft Instruments FAST TV area detector diffractometer positioned at the window of a rotating anode generator with Mo-Ka radiation (I = 0.710 69 A) following previously described procedures.l4 The structures were solved by direct methods (SHELX-S),rdquo; and refined by full-matrix least-squares on F using SHELX-76.I6 Hydrogen atoms were included in idealised positions with C-H = 0.96 A. Non-hydrogens were refined anisotropically. Details are as follows. Crysfal data. CI9H,,NSi,, Mr = 325.60, monoclinic, Q = 6.293(1), b = 23.000(3), c = 13.851 A, #? = 97.00(2)rsquo;, U = 1988 A3,monoclinic, space group P2,/c, 2 = 4, D, = 1.088 g cm-rsquo;, F(OO0) = 704, m -1.70 cn-rsquo;. Total data measured 11 843 giving 4842 unique (Rin, 0.06) and 2400 observed F, =. 3 s (F,j. R = 0.050, R, = 0.052 with unit weights and 221 refined parameters. Atomic fractional coordinates, thermal parameters and bond lengths and angles have been deposited at the Cambridge Crystallographic Data Centre.* * For details of the deposition scheme, see lsquo;Instructions for Authorsrsquo;, J.Chem. SOC.,Perkin Trans. I, 1995, Issue 1. J. CHEM. SOC. PERKIN TRANS. 1 1995 Acknowledgements We thank the SERC for a maintenance grant (to M. W. K.). References 1 Part 6. P. C. Hayes and G. Jones, J. Chem. Soc.,Perkin Trans. 1,1982, 1871. 2 W. Flitsch, B. Muter and W. Wolf, Chem. Ber., 1973,106, 1993. 3 G. R. Cliff and G. Jones, J. Chem. SOC.C, 1971,3418. 4 R. N. Carde, G. Jones, W. H. McKinley and C. Price, J. Chem. Soc., Perkin Trans. I, 1978, 121 1. 5 R. N. Carde, P. C. Hayes, G. Jones and C. J. Cliff, J. Chem. SOC., Perkin Trans. I, 1981, 1132. 6 G. Jones, B. D. Long and M. P. Thorne, J. Chem. SOC.,Perkin Trans. 2, 1992,903. 7 M. G. Hicks and G. Jones, J. Chem. SOC.,Chem. Commun., 1983, 1277. 8 H. W. Gschwend and H. R. Rodriguez, Org. React. (N.Y.),1979, 26, 1. 9 G. P. Schiemenz, Org. Synth. CON. Vol. V, 1973,496. 10 R. Robinson and J. E. Saxton, J. Chem. Soc., 1952,976. 1 1 See, for example, J. I. G. Cadogan, S. Kulik and M. J. Todd, J. Chem. SOC.C, 1970,2437. 12 G. R. Cliff, E. W. Collington and G. Jones, J. Chem. SOC.C, 1970, 1490. 13 D. Bergmann and R. Barshai, J. Am. Chem. SOC.,1959,81,5641. 14 S. R. Drake, M. B. Hursthouse, K. M. A. Malik and S.A. S. Miller, Inorg. Chem., 1993,32,4653. 15 G. M. Sheldrick, Acta Crystallogr., Sect. A, 1990,46,467. 16 G. M. Sheldrick, University of Cambridge, 1976. Paper 41057301 Received 20th September 1994 Received 13th October 1994