|
1. |
The X-ray crystal structure, conformation and preparation ofanti-3,3,6,6-tetramethylthiepane-4,5-diol: stereochemistry of reduction of a heterocyclic α-hydroxy ketone |
|
Journal of the Chemical Society, Perkin Transactions 1,
Volume 1,
Issue 22,
1997,
Page 3479-3484
Neil Feeder,
Preview
|
|
摘要:
J. Chem. Soc. Perkin Trans. 1 1997 3479 The X-ray crystal structure conformation and preparation of anti- 3,3,6,6-tetramethylthiepane-4,5-diol stereochemistry of reduction of a heterocyclic ·-hydroxy ketone Neil Feeder,a Michael J. Ginnelly,a Ray V. H. Jones,b Susan O’Sullivan,a Stuart Warren *,a and Paul Wyatt a a University Chemical Laboratory Lensfield Road Cambridge UK CB2 1 EW b Process Technology Department ZENECA Grangemouth Works Earls Road Grangemouth UK FK3 8XG The X-ray crystal structure and conformation of the anti title diol is described together with stereoselective syntheses of syn- and anti-diols from a readily available acyloin. Some control of the stereoselective reduction of ·-hydroxy ketones by chelating and non-chelating reducing agents is possible. The hydroxythiepanone 4 can be prepared 1,2 from the chloro acid 1 in three steps the key reaction being an efficient intramolecular acyloin condensation (enolisation is prevented by the tertiary alkyl groups) to give the seven-membered ring.The corresponding dione 5 has been used to prepare crowded molecules such as 4,5-di-tert-butylimidazole 1 6 and the remarkable cycloheptyne 3 7. de Groot and Wynberg also investigated the reduction 4 of the hydroxy ketone 4 to give a ‘diol’ and the rearrangements of this diol to give for example the bicyclic compound 8 in acid solution.5 No mention was made in these papers of the stereochemistry of ‘the diol’ prepared in this way. We now report the stereoselective preparation of both the synand the anti-diols 9 by the reduction of the hydroxy thiepanone 4 and the assignment of their configurations by X-ray crystal structure determination of the anti-diol.6 The stereochemistry of the two diols is interesting because each has a symmetry element the syn-diol is a meso compound with a plane of sym- Cl CO2H CO2Et S EtO2C S O OH CO2H S HO2C N HN S S O O 2 1 5 2.Na2S 1. Na2CO3 Na toluene 4 EtOH H+ 3 6 7 S O 8 CH3CO2H conc. H2SO4 4 reduction "a diol" metry while the anti compound is C2 symmetric and though chiral can be expected to show the same number of signals in the NMR spectrum as the syn compound. The diagrams syn-9a and anti-9a show these features. de Groot and Wynberg obtained 4,5 ‘the diol mp 179– 180 8C’ by reduction of 4 with LiAlH4 without discussing stereochemistry but the characterisation suggests it was a single compound. They report ‘four methylene protons’ in the 60 MHz NMR spectrum at dH 3.01 2.78 2.30 and 2.05.If these are in fact two AB systems the separation of 0.24 ppm would be a coupling constant of 14 Hz. This is a geminal (2JHH) coupling and is neither diagnostic nor characteristic (see below). Others have prepared one or both diols 9 without necessarily identifying the configuration. Johnson and co-workers 7 reduced the hydroxy ketone 4 with LiAlH4 to give an 85 15 mixture of two diols and with NaBH4 to get one diol in >90% yield. They identified the major diol as the syn isomer ‘The stereochemistry of this diol (a meso compound) was determined by using chiral shift reagents.2c A complete report of this approach to diol stereochemistry will be forthcoming.’ Their reference 2c is ‘P. Y. Johnson I.Jacobs and D. J. Kerkman J. Org. Chem. in the press.’ We do not believe that this paper appeared though they did publish a paper 8 on the stereochemistry of the related azepinediols. Their method relied on the NMR spectrum of a prochiral CH2 group in a substituent on the nitrogen atom in the azepines and would not apply to the thiepanes. It seems that they used Applequist’s ingenious and reliable method 9 to assign the configuration of the thiepanediols 9 and we believe their assignments are correct. We reinvestigated this problem to clear up the assignment of stereochemistry and because we needed the anti-diol 9 for S OH HO S OH HO S OH MeA MeB MeA MeB HO HC HC HB HB HA HA S OH MeA MeB MeA MeB HO HC HC HB HB HA HA syn-9a anti-9a sh C2 reduction 4 + syn-9 anti-9 3480 J.Chem. Soc. Perkin Trans. 1 1997 Table 1 Diastereomeric ratios of diols syn- and anti-9 from reduction of 4 Reagent NaBH4 NaBH4 NaBH4/CeCl3 LiAlH4 LiAlH4 Zn(BH4)2 DIBAL DIBAL/ZnCl2 Conditions EtOH 25 8C EtOH room temp. EtOH 278 8C THF reflux Et2O reflux Et2O 0 8C CH2Cl2 278 8C THF/Et2O room temp. Ratio syn- anti-9 >90% syn 100:0 100:0 85:15 (100 0?) 64:36 95:5 39:61 Isolated Yield syn or anti (%) syn 79 syn 92 syn 74 a Both isolated no yields given 64 syn 61 anti 34 — syn 36 anti 56 Ref. Johnson7 This work This work Johnson7 de Groot 4 This work This work This work a After recrystallisation. another project. It is also important to extend the well established control in the reduction of open-chain a-hydroxy ketones to their cyclic counterparts. The reduction of acyclic a-hydroxy ketones can usually be controlled to give either diol since the normal Felkin conformation with the CH]OH bond orthogonal to the plane of the C]] O group can be changed by chelation into one in which the CH]OH bond is in the plane of the C]] O group.The same freedom of rotation is not available to cyclic a-hydroxy ketones and ‘the influence of polar groups on the stereoselectivity of reduction of cyclic ketones has not been widely studied’.10 The only close analogy we can find is Applequist’s six-membered ring example9 10 (which lacks only a sulfur atom in comparison with 4!). Reduction with borohydride is also syn-selective here though not to such a pronounced degree. The assignment of structure referred to above depends on the NMR spectra of the cyclic sulfites derived from 11.They did not attempt to make the anti-diol 11 in high yield. Preparation and characterisation of the syn and anti diols 9 We first improved the preparation of the acyloin 4 by studying the only bad step the formation of the symmetrical sulfide 2 from the hindered acid in alkaline soloution. de Groot and Wynberg reported 1 a 50% yield but we were able to get only 16% by their procedure. However a simple adjustment of conditions chiefly the proportions of the reagents improved this to 89%. de Groot and Wynberg used 1 1 chloro acid 1 to Na2S we simply changed to the correct stoichiometric ratio of 2 1. Minor changes (see Experimental section) included recrystallisation of the Na2S?9H2O from water before use dropwise addition of this reagent to the acid 1 rather than the reverse and recrystallisation of the product 4 from water rather than acetic acid.The diester 3 can be made in 92% yield by continuous azeotropic distillation of water. We then studied the reduction of 4 by various chelating and non-chelating reducing agents measuring the diol ratio from the NMR spectrum of the reaction mixture and separating the diols (easily) by column chromatography. The results are given in Table 1 together with the previously published reductions. All non-chelating reducing agents gave a high proportion of one diol which we identified as the syn isomer (see below). Sodium borohydride in ethanol gave this isomer exclusively and it can be isolated in 92% yield by this method. Reduction under Luche 11,12 conditions (NaBH4–CeCl3) gave surprisingly the same complete selectivity as NaBH4.The cerium chloride evidently had very little effect in this case. Chelation with zinc proved to be the only way to get substantial amounts of the anti-diol 9. Zinc borohydride gave some anti-9 but with DIBAL–ZnCl2 we at last got mostly the anti compound and could isolate it in 56% yield. The diols are easily separated by chromatography on silica eluting with hexane– O OH HO OH HO OH NaBH4 10 syn-11 anti-11 + 82:18 ethyl acetate (6 1). Both diols 9 are crystalline and the syn-diol crystallises particularly easily. The most obvious distinctions between the two diols are the melting points the chromatographic behaviour and the chemical shift of the hydroxy-substituted carbon in the 13C NMR spectrum. These are summarised in Table 2. In the proton spectrum there are only small differences chiefly the wider separation between the methyl signals in the anti isomer the wider separation of the AB signal in the syn isomer (at 250 MHz the anti isomer gives a sharper and more distorted AB system) and the large difference in chemical shift for the hydroxy proton.The 1H NMR spectra are summarised in Table 3. Determination of stereochemistry by X-ray crystal structure analysis Determination of the crystal structures of syn- and anti-diols 9 was attempted by single-crystal X-ray diffraction. Data were collected using an Enraf-Nonius CAD4 four-circle diffractometer with graphite monochromated Mo-Ka radiation (sealed-tube source) and for the syn-diol also on a Rigaku AFC7R four-circle diffractometer. The structures were solved using SHELXS-8613 and refined using SHELX-93.14 Crystals of both diols were grown by slow evaporation from ethanol solutions.Crystals of anti-9 were colourless narrow needles (typical dimensions 0.40 × 0.10 × 0.10 mm) while crystals of syn-diol 9 were colourless plates (typical dimensions 0.30 × 0.20 × 0.10 mm). Since these structures are poorly refined we do not present the three dimensional co-ordinates here. Both structures were refined to give high initial R-factors (syn-diol 9 R = 0.102; anti-diol 9 R = 0.130). In the case of anti-9 this was a result of the weakly diffracting crystal giving a low reflection/parameter ratio (F2 > 3sF2 = 527 number of parameters = 126). The structure of syn-9 was found to exhibit severe disorder around the diol portion of the molecule. This disorder took the form of a number of slightly different ring conformations although each had the same syn-diol stereo- Table 2 Characteristic differences between the syn- and anti-diols 9 Measurement Mp Rf (3:1 hexane–EtOAc) dC CHOH syn-9 180–182 8C 0.17 82.9 ppm anti-9 89–91 8C 0.33 73.6 ppm Lit.syn-9 179–180 8C4 syn-9 183–185 8C7 Table 3 Proton NMR spectra of syn- and anti-diols 9 (refer to diagrams 9a – chemical shifts in ppm and coupling constant in Hz) CHCOH CHAHB CMeAMeB Diol syn-9 anti-9 d(HC) 3.73 3.56 d(OH) 1.78 2.76 d(HA) 2.83 2.52 d(HB) 2.28 2.27 2JAB 14.4 14.7 d(MeA) 1.08 1.09 d(MeB) 1.06 0.95 J. Chem. Soc. Perkin Trans. 1 1997 3481 chemistry. This disorder was difficult to model and we do not present this structure. The geometry of the anti-diol obtained from this structure determination is of sufficient quality to distinguish the syn and anti stereochemistry of the two diols 9.† Stereoselectivity of reduction and conformation of the two diols The syn-diol 9.The syn-diol is the one previously prepared. The stereoselectivity in favour of this diol with non-chelating reducing agents is most simply explained with a Felkin-like conformation 12 of the hydroxy ketone 4 with the OH group at right angles to the plane of the carbonyl group and attack occurring from the face opposite the OH group. This is a similar explanation to the high syn-selectivity in the attack of MeLi on 2-phenylsulfanylcycloheptanone 13 to give 15 syn-14. The anti-diol 9. The anti-diol and the stereoselectivity of its formation are more interesting. The crystal structure shows a single conformation—a rather chair-like puckered ring with pseudo-equatorial hydroxy groups 16 (Fig.1). Using this as a model for the transition state of the reduction we suggest a zinc chelate 15 with pseudo-axial attack opposite the nearer pseudoaxial methyl group. The closest analogy to our work with DIBAL and zinc salts is the work of Solladié and his group16–18 who have used chelation by zinc to reverse the stereoselectivity of the reduction of cyclic b-keto sulfoxides by DIBAL. The analogy is not very close because although the ketones are cyclic the sulfoxide group is always outside the ring and therefore free to rotate. Crossing from the syn- to the anti-series We used the Sharpless 19 orthoester approach to convert the syndiol 9 into the trans-acetoxy chloride 17 as a way of crossing from the syn- to the anti-series.The reaction was reasonably efficient (64% over two steps) and may provide an alternative source of thiepanes with anti-relationship between the substituents. Experimental All solvents were distilled before use. Tetrahydrofuran (THF) and diethyl ether were dried by stirring over lithium aluminium hydride. Dichloromethane hexane and toluene were dried by stirring over calcium hydride. Ether refers to diethyl ether. Thin Fig. 1 X-Ray crystal structure of the anti-diol 9 S OH HO S OAc Cl 1. (MeO)3CMe TsOH CH2Cl2 syn-9 anti-17 2. Me3SiCl CH2Cl2 3 days room temperature † Detailed crystallographic results for this work have been deposited with the Cambridge Crystallographic Data Centre and are available on request. Such a request should be accompanied by a full bibliographic reference for this work together with the reference number 207/147.Details of the deposition scheme are given in Instructions for Authors J. Chem. Soc. Perkin Trans. 1 1997 Issue 1. layer chromatography was carried out on commercially available pre-coated plates (Merck silica Kieselgel 60F254). Flash column chromatography was carried out on Merck Kieselgel 60 (230–400 mesh). 1H and 13C NMR spectra were recorded on a Bruker WM 200 Bruker WM 250 or a Bruker WM 400 Fourier transform spectrometer. The attached proton test (APT) for 13C NMR spectra recorded on the 250 and 400 MHz machines is reported with (1) designating signals in the same direction as the solvent (quaternary carbon and CH2) and (1) the opposite (i.e. CH and CH3). For 13C spectra recorded on the 200 MHz instrument (2) and (1) have the same meaning as above with (q) representing quaternary carbons which do not show up by DEPT.Melting points were recorded on a Reichart hot-stage microscope and are uncorrected. IR spectra were recorded on either a Perkin-Elmer 297 or a Perkin-Elmer 1600 FTIR spectrophotometer. Mass spectra (either electron impact or positive fast atom bombardment) were recorded on an AEI Kratos MS30 or MS890 machine using a DS503 data system for high-resolution analysis. Microanalyses were carried out using Carlo Erba 1106 or Perkin-Elmer 240 automatic analysers. Improved preparation of 2,2,6,6-tetramethyl-4-thiaheptanedioic acid 2 Sodium carbonate (2.01 g 19 mmol) was added to a stirred solution of 3-chloro-2,2-dimethylpropanoic acid 1 (5.27 g 38.4 mmol) in water (3.5 cm3).Sodium sulfide (4.6 g 19 mmol recrystallised from distilled water) in water (3.5 cm3) was added dropwise to the reaction mixture which was then stirred overnight at 45 8C. The mixture was cooled and carefully acidified with 50% aq. H2SO4 and the precipitate collected. Extraction of the precipitate with ethanol and recrystallisation from distilled water gave the diacid 2 (4.49 g 89%) (this acid has been prepared before1 in low yield but not characterised spectroscopically); dH(200 MHz; CD3OD) 2.80 (4 H s 2 × CH2S) and 1.23 (12 H s 4 × Me); dC(63 MHz; CD3OD) 180.4 (2) 45.9 (2) 45.0 (2) and 25.1 (1). Improved preparation of diethyl 2,2,6,6-tetramethyl-4-thiaheptanedioate 3 The diacid 2 (800 mg 3.4 mmol) was added to a stirred solution of dry ethanol (20 cm3) dry benzene (20 cm3) and a few drops of concentrated sulfuric acid.The reaction mixture was refluxed for 24 h using a Dean–Stark apparatus to remove the azeotrope produced. After being cooled to room temperature the reaction mixture was washed with water (40 cm3) and dilute aqueous sodium hydrogen carbonate (40 cm3) dried (MgSO4) and evaporated under reduced pressure. The residue was chromatographed (SiO2 hexane–ethyl acetate 7 1) to give the diester 3 (910 mg 92%) as a colourless oil (this diester has been prepared before 1 but not characterised spectroscopically) Rf [hexane–ethyl acetate (6 1)] 0.3; dH(200 MHz; CDCl3) 4.10 (4 H quartet J 7.1 2 × CO2CH2Me) 2.73 (4 H s 2 × SCH2) 1.22 (6 H t J 7.1 2 × CO2CH2Me) and 1.19 (12 H s 4 × Me); dC(100 MHz; CDCl3) 176.5 (2) 60.6 (2) 45.1 (2) 44.2 (2) 24.6 (1) and 14.2 (1).5-Hydroxy-3,3,6,6-tetramethylthiacycloheptan-4-one 4 The method of de Groot and Wynberg1 gave the hydroxy ketone 4 as needles mp 81–83 8C (from ethanol) (lit.,1 80– 82 8C); Rf [hexane–ether (5 2)] 0.2; nmax(CH2Cl2)/cm21 3499 (OH) and 1697 (C]] O); dH(250 MHz; CDCl3) 4.17 [1 H d J 7.8 CHOH (collapses to a singlet upon D2O shake)] 3.41 [1 H d J 7.8 CHOH (disappears upon D2O shake)] 2.78 (1 H d J 14.7 SCHACHB) 2.69 (1 H d J 15.3 SCHCCHD) 2.60 (1 H d J 15.7 SCHCCHD) 2.46 (1 H d J 14.7 SCHACHB) 1.29 (3 H s MeA) 1.15 (3 H s MeB) 1.11 (3 H s MeC) and 0.78 (3 H s MeD); dC(100 MHz; CDCl3) 216.6 (2) 78.8 (1) 50.3 (2) 47.4 (2) 42.6 (2) 42.2 (2) 27.6 (1) 27.3 (1) 23.5 (1) and 19.2 3482 J. Chem. Soc. Perkin Trans. 1 1997 (1) (Found M1 202.1033. C10H18SO2 requires M1 202.1027); m/z 202 (M1,13%) 147 (87) 118 [M1 2 C(]] O)C(Me)2CH2 30] and 56 [C(Me)2CH2 100].Stereoselective reductions of the ·-hydroxy ketone 4 syn- 3,3,6,6-tetramethylthiacycloheptane-4,5-diol syn-9 Sodium borohydride (75 mg 19.5 mmol) was added to a stirred solution of the hydroxy ketone 4 (80 mg 3.9 mmol) in dry ethanol (1 cm3). The reaction mixture was stirred at room temperature for 2 h after which it was treated with dilute hydrochloric acid and concentrated by evaporation of most of the ethanol under reduced pressure. The residue was dissolved in dichloromethane (5 cm3) and the solution was washed with water (5 cm3) aqueous sodium hydrogen carbonate (5 cm3) and water and then dried (MgSO4) and evaporated under reduced pressure to give a colourless solid. This was purified by chromatography (SiO2 hexane–ethyl acetate 3 1) to remove any remaining starting material to give the syn-diol 9 as prisms (75 mg 92%) mp 180–182 8C (from hexane) (lit.,4 179–180 8C lit.,7 183–185 8C); Rf [hexane–ethyl acetate (3 1)] 0.19; nmax(Nujol)/ cm21 3406 (OH) and 2922 (CH); dH(250 MHz; CDCl3) 3.73 (2 H d J 6.2 2 × CHOH) 2.83 (2 H d J 14.5 2 × SCHAHB) 2.28 (2 H d J 14.5 2 × SCHAHB) 1.78 (2 H d J 6.3 2 × CHOH) 1.08 (6 H s 2 × CMeAMeB) and 1.07 (6 H s 2 × CMeAMeB); dC(100 MHz; CD3COCD3) 82.9 (1) 46.3 (2) 40.5 (2) 27.6 (1) and 19.6 (1) (Found M1 204.1185.C10H20O2S requires M1 204.1184); m/z 204 (M1 48) 120 (78) 86 [CH2CMe2C(OH)H 69] and 56 [CH2CMe2 52]. Luche Reduction Sodium borohydride (10 mg 0.26 mmol) the hydroxy ketone 8 (16 mg 0.08 mmol) and CeCl3?7H2O (33 mg 0.09 mmol) in dry ethanol (0.5 cm3) at 278 8C gave after purification by washing through a plug of silica eluting with hexane–ethyl acetate (1 1) and recrystallisation the syn-diol 9 (12 mg 74%).Preparation of Zn(BH4)2 solution 20 Zinc chloride (1 mol dm23 solution in diethyl ether; 27.5 cm3) was added dropwise to a stirred solution of sodium borohydride (2 g 52.8 mmol) in dry ether (150 cm3). The reaction mixture was stirred at room temperature for 2 days. The solid material was then allowed to settle and the supernatant solution transferred by cannula to a bottle. The solution was stored at 5 8C under argon. Reduction with Zn(BH4)2 Zinc borohydride solution in Et2O (prepared as above) and the hydroxy ketone 4 (88 mg 0.43 mmol) in dry ether (1.5 cm3) at 0 8C for 6 h (with further reducing agent added until all the starting material was consumed) gave after acidification extraction with ethyl acetate (3 × 5 cm3) washing with saturated aqueous sodium hydrogen carbonate drying (MgSO4) and chromatography (SiO2 hexane–ethyl acetate 6 1) the syn-diol 9 (50 mg 62%) and the anti-diol 9 (30 mg 38%) (see below).Reduction with LiAlH4 The hydroxy ketone 4 (20 mg 0.1 mmol) in dry ether (0.5 cm3) and a solution of lithium aluminium hydride (5 mg 0.13 mmol) in dry ether (1.5 cm3) was refluxed for 4 h and quenched by the careful addition of water and then dilute aq. HCl. The organic layer was separated and the aqueous layer extracted with ether (3 × 5 cm3) to give after work-up the syn- and anti-diols 9 in a 16 3 ratio (by NMR). Reduction with DIBAL A mixture of diisobutylaluminium hydride (1 mol dm23 in CH2Cl2; 0.22 cm3) and the hydroxy ketone 4 (9 mg 0.044 mmol) in dry CH2Cl2 was stirred at 278 8C for 3 h and then warmed to room temperature and treated with methanol.After the mixture had been stirred for a further 30 min the precipitate was filtered off and washed several times with dichloromethane. The filtrate and combined washings were evaporated under reduced pressure and the crude material was filtered through a plug of silica to give a 95 5 mixture (by NMR) of the syn- and anti-diols 9. anti-3,3,6,6-Tetramethylthiacycloheptane-4,5-diol anti-9 Reduction with DIBAL and ZnCl2. Zinc chloride (0.5 mol dm23 solution in THF; 3 cm3) was added to a stirred solution of the hydroxy ketone 4 (300 mg 1.49 mmol) in dry THF. The solution was cooled to 278 8C and treated with a solution of DIBAL (1 mol dm23 in THF; 1.5 cm3 1.5 mmol) added dropwise.The reaction mixture was allowed to warm to room temperature and then stirred overnight. After this methanol was added to the mixture which was then stirred for a further 30 min. The resulting inorganic precipitate was filtered off and washed several times with dichloromethane. The filtrate and organic washings were combined and evaporated under reduced pressure. The crude material (a 39 61 ratio of syn- and anti-diols 9 by NMR) was chromatographed on silica eluting with hexane–ethyl acetate (5 2) to give the syn-diol (105 mg 35%) and the anti-diol 9 (160 mg 54%) Rf [hexane–ethyl acetate (3 1)] 0.33 (Found C 58.75; H 9.89. C10H20SO2 requires C 58.78; H 9.87%); nmax(CH2Cl2)/cm21 3614–3495 (OH) and 2958–2871 (CH); dH(250 MHz; CDCl3) 3.56 (2 H br s 2 × CHOH) 2.76 (2 H br s 2 × CHOH) 2.52 (2 H d J 14.7 2 × SCHACHB) 2.27 (2 H d J 14.7 2 × SCHACHB) 1.09 (6 H s 2 × CMeAMeB) and 0.95 (6 H s 2 × CMeAMeB); dC(63 MHz; CDCl3) 73.6 (1) 47.1 (2) 38.8 (2) 28.3 (1) and 19.7 (1) (Found M1 204.1181.C10H20SO2 requires M1 204.1184); m/z 204 (M1 11%) 186 (M1 2 H2O 20) 139 (76) and 130 [M1 2 H2O 2 CH2C(Me)2 100] 86 [CH2C(Me)2C(OH)H 40] and 56 [CH2C(Me)2 52]. 4-Acetoxy-5-chloro-3,3,6,6-tetramethylthiacycloheptane 17 Trimethyl orthoacetate (0.138 cm3 1.1 mmol) was added to a solution of the syn-diol 9 (170 mg 0.83 mmol) and toluene-psulfonic acid (2 mg) in dry CH2Cl2 (2 cm3). After the mixture had been stirred at room temperature for 1 h volatile material was removed under reduced pressure and most of the residual methanol was removed by subjecting the sample to high vacuum for 1 min.The residue was then dissolved in dichloromethane and trimethylsilyl chloride (0.160 cm3 1.25 mmol) was added to the solution. The reaction mixture was stirred at room temperature for 4 days after which it was evaporated under reduced pressure. The residue was chromatographed (SiO2 hexane–ether 10 1) to give the a-chloro acetate 17 (141 mg 64%) as a colourless oil; Rf [hexane–ethyl acetate (3 1)] 0.52; nmax(CDCl3)/cm21 2962–2890 (CH) and 1732 (C]] O); dH(250 MHz; CDCl3) 5.18 (1 H d J 3.5 CHOAc) 3.78 (1 H d J 3.5 CHCl) 3.40 (2 H AB quartet J 11.1 SCHACHB) 2.89 (1 H d J 10.1 SCHCCHD) 2.52 (1 H d J 10.1 SCHCCHD) 2.11 (3 H s CO2Me) 1.12 (3 H s MeC) 1.08 (6 H s MeAMeB) and 0.97 (3 H s MeD); dC(100 MHz; CDCl3) 170.2 (2) 82.6 (1) 56.7 (1) 55.4 (2) 47.7 (2) 41.2 (2) 38.1 (2) 25.4 (1) 24.4 (1) 24.2 (1) 22.8 (1) and 21.3 (1) (Found M1 264.0954.C12H21O2SCl requires M1 264.0951); m/z 266 (M1 15%) 264 (M1 45) 221 [M1 2 C(]] O)Me 2] 204 (M1 2 HCO2Me 4) and 173 [M1 2 Cl 2 C(Me)2CH2 100]. Acknowledgements We thank Zeneca Fine Chemicals and EPSRC for CASE awards (to M. J. G. S. O’S and P. W.). References 1 A. de Groot and H. Wynberg J. Org. Chem. 1966 31 3954. 2 H. Wynberg and A. de Groot J. Chem. Soc. Chem. Commun. 1965 171. 3 A. Krebs and H. Kimling Tetrahedron Lett. 1970 761. 4 A. de Groot J. A. Boerma and H. Wynberg Rec. Trav. Chim. Pays- Bas. 1969 88 994. J. Chem. Soc. Perkin Trans. 1 1997 3483 5 A. de Groot J. A. Boerma and H. Wynberg Tetrahedron Lett. 1968 2365. 6 N. Feeder M. J.Ginnelly R. V. H. Jones S. O’Sullivan and S. Warren Tetrahedron Lett. 1994 35 9095. 7 P. Y. Johnson and M. Berman J. Org. Chem. 1975 40 3046. 8 P. Y. Johnson and D. J. Kirkman J. Org. Chem. 1976 41 1768. 9 D. E. Applequist P. A. Gebauer D. E. Gwynn and L. H. O’Connor J. Am. Chem. Soc. 1972 94 4272. 10 N. Greeves in Comprehensive Organic Syntheses ed. B. M. Trost and I. Fleming Oxford 1991 vol. 8 pp.1–24. 11 A. L. Gemal and J.-L. Luche J. Am. Chem. Soc. 1981 103 5454. 12 J.-L. Luche J. Am. Chem. Soc. 1978 100 2226. 13 G. M. Sheldrick in ‘SHELX86. Program for the Solution of Crystal Structures’ University of Gottingen Germany 1990. 14 G. M. Sheldrick in ‘SHELXL93. Program for the Refinement of Crystal Structures’ University of Gottingen Germany 1993. 15 M. Hannaby and S. Warren J. Chem.Soc. Perkin Trans. 1 1989 303. 16 M. C. Carreño J. L. G. Ruano A. M. Martín C. Pedregal J. H. Rodriguez A. Rubio J. Sanchez and G. Solladié J. Org. Chem. 1990 55 2120. 17 M. C. Carreño J. L. G. Ruano M. Garrido M. P. Ruiz and G. Solladié Tetrahedron Lett. 1990 31 6653. 18 G. Solladié C. Fréchou G. Demailly and C. Greck J. Org. Chem. 1986 51 1912. 19 H. C. Kolb and K. B. Sharpless Tetrahedron 1992 48 10 515. 20 T. Nakata Y. Tani M. Hatozaki and T. Oishi Chem. Pharm. Bull. 1984 32 1411. Paper 6/05572I Received 9th August 1996 Accepted 6th August 1997 © Copyright 1997 by the Royal Society of Chemistry J. Chem. Soc. Perkin Trans. 1 1997 3479 The X-ray crystal structure conformation and preparation of anti- 3,3,6,6-tetramethylthiepane-4,5-diol stereochemistry of reduction of a heterocyclic ·-hydroxy ketone Neil Feeder,a Michael J.Ginnelly,a Ray V. H. Jones,b Susan O’Sullivan,a Stuart Warren *,a and Paul Wyatt a a University Chemical Laboratory Lensfield Road Cambridge UK CB2 1 EW b Process Technology Department ZENECA Grangemouth Works Earls Road Grangemouth UK FK3 8XG The X-ray crystal structure and conformation of the anti title diol is described together with stereoselective syntheses of syn- and anti-diols from a readily available acyloin. Some control of the stereoselective reduction of ·-hydroxy ketones by chelating and non-chelating reducing agents is possible. The hydroxythiepanone 4 can be prepared 1,2 from the chloro acid 1 in three steps the key reaction being an efficient intramolecular acyloin condensation (enolisation is prevented by the tertiary alkyl groups) to give the seven-membered ring.The corresponding dione 5 has been used to prepare crowded molecules such as 4,5-di-tert-butylimidazole 1 6 and the remarkable cycloheptyne 3 7. de Groot and Wynberg also investigated the reduction 4 of the hydroxy ketone 4 to give a ‘diol’ and the rearrangements of this diol to give for example the bicyclic compound 8 in acid solution.5 No mention was made in these papers of the stereochemistry of ‘the diol’ prepared in this way. We now report the stereoselective preparation of both the synand the anti-diols 9 by the reduction of the hydroxy thiepanone 4 and the assignment of their configurations by X-ray crystal structure determination of the anti-diol.6 The stereochemistry of the two diols is interesting because each has a symmetry element the syn-diol is a meso compound with a plane of sym- Cl CO2H CO2Et S EtO2C S O OH CO2H S HO2C N HN S S O O 2 1 5 2.Na2S 1. Na2CO3 Na toluene 4 EtOH H+ 3 6 7 S O 8 CH3CO2H conc. H2SO4 4 reduction "a diol" metry while the anti compound is C2 symmetric and though chiral can be expected to show the same number of signals in the NMR spectrum as the syn compound. The diagrams syn-9a and anti-9a show these features. de Groot and Wynberg obtained 4,5 ‘the diol mp 179– 180 8C’ by reduction of 4 with LiAlH4 without discussing stereochemistry but the characterisation suggests it was a single compound. They report ‘four methylene protons’ in the 60 MHz NMR spectrum at dH 3.01 2.78 2.30 and 2.05. If these are in fact two AB systems the separation of 0.24 ppm would be a coupling constant of 14 Hz.This is a geminal (2JHH) coupling and is neither diagnostic nor characteristic (see below). Others have prepared one or both diols 9 without necessarily identifying the configuration. Johnson and co-workers 7 reduced the hydroxy ketone 4 with LiAlH4 to give an 85 15 mixture of two diols and with NaBH4 to get one diol in >90% yield. They identified the major diol as the syn isomer ‘The stereochemistry of this diol (a meso compound) was determined by using chiral shift reagents.2c A complete report of this approach to diol stereochemistry will be forthcoming.’ Their reference 2c is ‘P. Y. Johnson I. Jacobs and D. J. Kerkman J. Org. Chem. in the press.’ We do not believe that this paper appeared though they did publish a paper 8 on the stereochemistry of the related azepinediols.Their method relied on the NMR spectrum of a prochiral CH2 group in a substituent on the nitrogen atom in the azepines and would not apply to the thiepanes. It seems that they used Applequist’s ingenious and reliable method 9 to assign the configuration of the thiepanediols 9 and we believe their assignments are correct. We reinvestigated this problem to clear up the assignment of stereochemistry and because we needed the anti-diol 9 for S OH HO S OH HO S OH MeA MeB MeA MeB HO HC HC HB HB HA HA S OH MeA MeB MeA MeB HO HC HC HB HB HA HA syn-9a anti-9a sh C2 reduction 4 + syn-9 anti-9 3480 J. Chem. Soc. Perkin Trans. 1 1997 Table 1 Diastereomeric ratios of diols syn- and anti-9 from reduction of 4 Reagent NaBH4 NaBH4 NaBH4/CeCl3 LiAlH4 LiAlH4 Zn(BH4)2 DIBAL DIBAL/ZnCl2 Conditions EtOH 25 8C EtOH room temp.EtOH 278 8C THF reflux Et2O reflux Et2O 0 8C CH2Cl2 278 8C THF/Et2O room temp. Ratio syn- anti-9 >90% syn 100:0 100:0 85:15 (100 0?) 64:36 95:5 39:61 Isolated Yield syn or anti (%) syn 79 syn 92 syn 74 a Both isolated no yields given 64 syn 61 anti 34 — syn 36 anti 56 Ref. Johnson7 This work This work Johnson7 de Groot 4 This work This work This work a After recrystallisation. another project. It is also important to extend the well established control in the reduction of open-chain a-hydroxy ketones to their cyclic counterparts. The reduction of acyclic a-hydroxy ketones can usually be controlled to give either diol since the normal Felkin conformation with the CH]OH bond orthogonal to the plane of the C]] O group can be changed by chelation into one in which the CH]OH bond is in the plane of the C]] O group.The same freedom of rotation is not available to cyclic a-hydroxy ketones and ‘the influence of polar groups on the stereoselectivity of reduction of cyclic ketones has not been widely studied’.10 The only close analogy we can find is Applequist’s six-membered ring example9 10 (which lacks only a sulfur atom in comparison with 4!). Reduction with borohydride is also syn-selective here though not to such a pronounced degree. The assignment of structure referred to above depends on the NMR spectra of the cyclic sulfites derived from 11. They did not attempt to make the anti-diol 11 in high yield. Preparation and characterisation of the syn and anti diols 9 We first improved the preparation of the acyloin 4 by studying the only bad step the formation of the symmetrical sulfide 2 from the hindered acid in alkaline soloution.de Groot and Wynberg reported 1 a 50% yield but we were able to get only 16% by their procedure. However a simple adjustment of conditions chiefly the proportions of the reagents improved this to 89%. de Groot and Wynberg used 1 1 chloro acid 1 to Na2S we simply changed to the correct stoichiometric ratio of 2 1. Minor changes (see Experimental section) included recrystallisation of the Na2S?9H2O from water before use dropwise addition of this reagent to the acid 1 rather than the reverse and recrystallisation of the product 4 from water rather than acetic acid. The diester 3 can be made in 92% yield by continuous azeotropic distillation of water. We then studied the reduction of 4 by various chelating and non-chelating reducing agents measuring the diol ratio from the NMR spectrum of the reaction mixture and separating the diols (easily) by column chromatography.The results are given in Table 1 together with the previously published reductions. All non-chelating reducing agents gave a high proportion of one diol which we identified as the syn isomer (see below). Sodium borohydride in ethanol gave this isomer exclusively and it can be isolated in 92% yield by this method. Reduction under Luche 11,12 conditions (NaBH4–CeCl3) gave surprisingly the same complete selectivity as NaBH4. The cerium chloride evidently had very little effect in this case. Chelation with zinc proved to be the only way to get substantial amounts of the anti-diol 9.Zinc borohydride gave some anti-9 but with DIBAL–ZnCl2 we at last got mostly the anti compound and could isolate it in 56% yield. The diols are easily separated by chromatography on silica eluting with hexane– O OH HO OH HO OH NaBH4 10 syn-11 anti-11 + 82:18 ethyl acetate (6 1). Both diols 9 are crystalline and the syn-diol crystallises particularly easily. The most obvious distinctions between the two diols are the melting points the chromatographic behaviour and the chemical shift of the hydroxy-substituted carbon in the 13C NMR spectrum. These are summarised in Table 2. In the proton spectrum there are only small differences chiefly the wider separation between the methyl signals in the anti isomer the wider separation of the AB signal in the syn isomer (at 250 MHz the anti isomer gives a sharper and more distorted AB system) and the large difference in chemical shift for the hydroxy proton.The 1H NMR spectra are summarised in Table 3. Determination of stereochemistry by X-ray crystal structure analysis Determination of the crystal structures of syn- and anti-diols 9 was attempted by single-crystal X-ray diffraction. Data were collected using an Enraf-Nonius CAD4 four-circle diffractometer with graphite monochromated Mo-Ka radiation (sealed-tube source) and for the syn-diol also on a Rigaku AFC7R four-circle diffractometer. The structures were solved using SHELXS-8613 and refined using SHELX-93.14 Crystals of both diols were grown by slow evaporation from ethanol solutions. Crystals of anti-9 were colourless narrow needles (typical dimensions 0.40 × 0.10 × 0.10 mm) while crystals of syn-diol 9 were colourless plates (typical dimensions 0.30 × 0.20 × 0.10 mm).Since these structures are poorly refined we do not present the three dimensional co-ordinates here. Both structures were refined to give high initial R-factors (syn-diol 9 R = 0.102; anti-diol 9 R = 0.130). In the case of anti-9 this was a result of the weakly diffracting crystal giving a low reflection/parameter ratio (F2 > 3sF2 = 527 number of parameters = 126). The structure of syn-9 was found to exhibit severe disorder around the diol portion of the molecule. This disorder took the form of a number of slightly different ring conformations although each had the same syn-diol stereo- Table 2 Characteristic differences between the syn- and anti-diols 9 Measurement Mp Rf (3:1 hexane–EtOAc) dC CHOH syn-9 180–182 8C 0.17 82.9 ppm anti-9 89–91 8C 0.33 73.6 ppm Lit.syn-9 179–180 8C4 syn-9 183–185 8C7 Table 3 Proton NMR spectra of syn- and anti-diols 9 (refer to diagrams 9a – chemical shifts in ppm and coupling constant in Hz) CHCOH CHAHB CMeAMeB Diol syn-9 anti-9 d(HC) 3.73 3.56 d(OH) 1.78 2.76 d(HA) 2.83 2.52 d(HB) 2.28 2.27 2JAB 14.4 14.7 d(MeA) 1.08 1.09 d(MeB) 1.06 0.95 J. Chem. Soc. Perkin Trans. 1 1997 3481 chemistry. This disorder was difficult to model and we do not present this structure. The geometry of the anti-diol obtained from this structure determination is of sufficient quality to distinguish the syn and anti stereochemistry of the two diols 9.† Stereoselectivity of reduction and conformation of the two diols The syn-diol 9.The syn-diol is the one previously prepared. The stereoselectivity in favour of this diol with non-chelating reducing agents is most simply explained with a Felkin-like conformation 12 of the hydroxy ketone 4 with the OH group at right angles to the plane of the carbonyl group and attack occurring from the face opposite the OH group. This is a similar explanation to the high syn-selectivity in the attack of MeLi on 2-phenylsulfanylcycloheptanone 13 to give 15 syn-14. The anti-diol 9. The anti-diol and the stereoselectivity of its formation are more interesting. The crystal structure shows a single conformation—a rather chair-like puckered ring with pseudo-equatorial hydroxy groups 16 (Fig. 1). Using this as a model for the transition state of the reduction we suggest a zinc chelate 15 with pseudo-axial attack opposite the nearer pseudoaxial methyl group.The closest analogy to our work with DIBAL and zinc salts is the work of Solladié and his group16–18 who have used chelation by zinc to reverse the stereoselectivity of the reduction of cyclic b-keto sulfoxides by DIBAL. The analogy is not very close because although the ketones are cyclic the sulfoxide group is always outside the ring and therefore free to rotate. Crossing from the syn- to the anti-series We used the Sharpless 19 orthoester approach to convert the syndiol 9 into the trans-acetoxy chloride 17 as a way of crossing from the syn- to the anti-series. The reaction was reasonably efficient (64% over two steps) and may provide an alternative source of thiepanes with anti-relationship between the substituents.Experimental All solvents were distilled before use. Tetrahydrofuran (THF) and diethyl ether were dried by stirring over lithium aluminium hydride. Dichloromethane hexane and toluene were dried by stirring over calcium hydride. Ether refers to diethyl ether. Thin Fig. 1 X-Ray crystal structure of the anti-diol 9 S OH HO S OAc Cl 1. (MeO)3CMe TsOH CH2Cl2 syn-9 anti-17 2. Me3SiCl CH2Cl2 3 days room temperature † Detailed crystallographic results for this work have been deposited with the Cambridge Crystallographic Data Centre and are available on request. Such a request should be accompanied by a full bibliographic reference for this work together with the reference number 207/147. Details of the deposition scheme are given in Instructions for Authors J. Chem. Soc. Perkin Trans.1 1997 Issue 1. layer chromatography was carried out on commercially available pre-coated plates (Merck silica Kieselgel 60F254). Flash column chromatography was carried out on Merck Kieselgel 60 (230–400 mesh). 1H and 13C NMR spectra were recorded on a Bruker WM 200 Bruker WM 250 or a Bruker WM 400 Fourier transform spectrometer. The attached proton test (APT) for 13C NMR spectra recorded on the 250 and 400 MHz machines is reported with (1) designating signals in the same direction as the solvent (quaternary carbon and CH2) and (1) the opposite (i.e. CH and CH3). For 13C spectra recorded on the 200 MHz instrument (2) and (1) have the same meaning as above with (q) representing quaternary carbons which do not show up by DEPT. Melting points were recorded on a Reichart hot-stage microscope and are uncorrected.IR spectra were recorded on either a Perkin-Elmer 297 or a Perkin-Elmer 1600 FTIR spectrophotometer. Mass spectra (either electron impact or positive fast atom bombardment) were recorded on an AEI Kratos MS30 or MS890 machine using a DS503 data system for high-resolution analysis. Microanalyses were carried out using Carlo Erba 1106 or Perkin-Elmer 240 automatic analysers. Improved preparation of 2,2,6,6-tetramethyl-4-thiaheptanedioic acid 2 Sodium carbonate (2.01 g 19 mmol) was added to a stirred solution of 3-chloro-2,2-dimethylpropanoic acid 1 (5.27 g 38.4 mmol) in water (3.5 cm3). Sodium sulfide (4.6 g 19 mmol recrystallised from distilled water) in water (3.5 cm3) was added dropwise to the reaction mixture which was then stirred overnight at 45 8C.The mixture was cooled and carefully acidified with 50% aq. H2SO4 and the precipitate collected. Extraction of the precipitate with ethanol and recrystallisation from distilled water gave the diacid 2 (4.49 g 89%) (this acid has been prepared before1 in low yield but not characterised spectroscopically); dH(200 MHz; CD3OD) 2.80 (4 H s 2 × CH2S) and 1.23 (12 H s 4 × Me); dC(63 MHz; CD3OD) 180.4 (2) 45.9 (2) 45.0 (2) and 25.1 (1). Improved preparation of diethyl 2,2,6,6-tetramethyl-4-thiaheptanedioate 3 The diacid 2 (800 mg 3.4 mmol) was added to a stirred solution of dry ethanol (20 cm3) dry benzene (20 cm3) and a few drops of concentrated sulfuric acid. The reaction mixture was refluxed for 24 h using a Dean–Stark apparatus to remove the azeotrope produced. After being cooled to room temperature the reaction mixture was washed with water (40 cm3) and dilute aqueous sodium hydrogen carbonate (40 cm3) dried (MgSO4) and evaporated under reduced pressure.The residue was chromatographed (SiO2 hexane–ethyl acetate 7 1) to give the diester 3 (910 mg 92%) as a colourless oil (this diester has been prepared before 1 but not characterised spectroscopically) Rf [hexane–ethyl acetate (6 1)] 0.3; dH(200 MHz; CDCl3) 4.10 (4 H quartet J 7.1 2 × CO2CH2Me) 2.73 (4 H s 2 × SCH2) 1.22 (6 H t J 7.1 2 × CO2CH2Me) and 1.19 (12 H s 4 × Me); dC(100 MHz; CDCl3) 176.5 (2) 60.6 (2) 45.1 (2) 44.2 (2) 24.6 (1) and 14.2 (1). 5-Hydroxy-3,3,6,6-tetramethylthiacycloheptan-4-one 4 The method of de Groot and Wynberg1 gave the hydroxy ketone 4 as needles mp 81–83 8C (from ethanol) (lit.,1 80– 82 8C); Rf [hexane–ether (5 2)] 0.2; nmax(CH2Cl2)/cm21 3499 (OH) and 1697 (C]] O); dH(250 MHz; CDCl3) 4.17 [1 H d J 7.8 CHOH (collapses to a singlet upon D2O shake)] 3.41 [1 H d J 7.8 CHOH (disappears upon D2O shake)] 2.78 (1 H d J 14.7 SCHACHB) 2.69 (1 H d J 15.3 SCHCCHD) 2.60 (1 H d J 15.7 SCHCCHD) 2.46 (1 H d J 14.7 SCHACHB) 1.29 (3 H s MeA) 1.15 (3 H s MeB) 1.11 (3 H s MeC) and 0.78 (3 H s MeD); dC(100 MHz; CDCl3) 216.6 (2) 78.8 (1) 50.3 (2) 47.4 (2) 42.6 (2) 42.2 (2) 27.6 (1) 27.3 (1) 23.5 (1) and 19.2 3482 J.Chem. Soc. Perkin Trans. 1 1997 (1) (Found M1 202.1033. C10H18SO2 requires M1 202.1027); m/z 202 (M1,13%) 147 (87) 118 [M1 2 C(]] O)C(Me)2CH2 30] and 56 [C(Me)2CH2 100]. Stereoselective reductions of the ·-hydroxy ketone 4 syn- 3,3,6,6-tetramethylthiacycloheptane-4,5-diol syn-9 Sodium borohydride (75 mg 19.5 mmol) was added to a stirred solution of the hydroxy ketone 4 (80 mg 3.9 mmol) in dry ethanol (1 cm3).The reaction mixture was stirred at room temperature for 2 h after which it was treated with dilute hydrochloric acid and concentrated by evaporation of most of the ethanol under reduced pressure. The residue was dissolved in dichloromethane (5 cm3) and the solution was washed with water (5 cm3) aqueous sodium hydrogen carbonate (5 cm3) and water and then dried (MgSO4) and evaporated under reduced pressure to give a colourless solid. This was purified by chromatography (SiO2 hexane–ethyl acetate 3 1) to remove any remaining starting material to give the syn-diol 9 as prisms (75 mg 92%) mp 180–182 8C (from hexane) (lit.,4 179–180 8C lit.,7 183–185 8C); Rf [hexane–ethyl acetate (3 1)] 0.19; nmax(Nujol)/ cm21 3406 (OH) and 2922 (CH); dH(250 MHz; CDCl3) 3.73 (2 H d J 6.2 2 × CHOH) 2.83 (2 H d J 14.5 2 × SCHAHB) 2.28 (2 H d J 14.5 2 × SCHAHB) 1.78 (2 H d J 6.3 2 × CHOH) 1.08 (6 H s 2 × CMeAMeB) and 1.07 (6 H s 2 × CMeAMeB); dC(100 MHz; CD3COCD3) 82.9 (1) 46.3 (2) 40.5 (2) 27.6 (1) and 19.6 (1) (Found M1 204.1185.C10H20O2S requires M1 204.1184); m/z 204 (M1 48) 120 (78) 86 [CH2CMe2C(OH)H 69] and 56 [CH2CMe2 52]. Luche Reduction Sodium borohydride (10 mg 0.26 mmol) the hydroxy ketone 8 (16 mg 0.08 mmol) and CeCl3?7H2O (33 mg 0.09 mmol) in dry ethanol (0.5 cm3) at 278 8C gave after purification by washing through a plug of silica eluting with hexane–ethyl acetate (1 1) and recrystallisation the syn-diol 9 (12 mg 74%).Preparation of Zn(BH4)2 solution 20 Zinc chloride (1 mol dm23 solution in diethyl ether; 27.5 cm3) was added dropwise to a stirred solution of sodium borohydride (2 g 52.8 mmol) in dry ether (150 cm3). The reaction mixture was stirred at room temperature for 2 days. The solid material was then allowed to settle and the supernatant solution transferred by cannula to a bottle. The solution was stored at 5 8C under argon. Reduction with Zn(BH4)2 Zinc borohydride solution in Et2O (prepared as above) and the hydroxy ketone 4 (88 mg 0.43 mmol) in dry ether (1.5 cm3) at 0 8C for 6 h (with further reducing agent added until all the starting material was consumed) gave after acidification extraction with ethyl acetate (3 × 5 cm3) washing with saturated aqueous sodium hydrogen carbonate drying (MgSO4) and chromatography (SiO2 hexane–ethyl acetate 6 1) the syn-diol 9 (50 mg 62%) and the anti-diol 9 (30 mg 38%) (see below).Reduction with LiAlH4 The hydroxy ketone 4 (20 mg 0.1 mmol) in dry ether (0.5 cm3) and a solution of lithium aluminium hydride (5 mg 0.13 mmol) in dry ether (1.5 cm3) was refluxed for 4 h and quenched by the careful addition of water and then dilute aq. HCl. The organic layer was separated and the aqueous layer extracted with ether (3 × 5 cm3) to give after work-up the syn- and anti-diols 9 in a 16 3 ratio (by NMR). Reduction with DIBAL A mixture of diisobutylaluminium hydride (1 mol dm23 in CH2Cl2; 0.22 cm3) and the hydroxy ketone 4 (9 mg 0.044 mmol) in dry CH2Cl2 was stirred at 278 8C for 3 h and then warmed to room temperature and treated with methanol.After the mixture had been stirred for a further 30 min the precipitate was filtered off and washed several times with dichloromethane. The filtrate and combined washings were evaporated under reduced pressure and the crude material was filtered through a plug of silica to give a 95 5 mixture (by NMR) of the syn- and anti-diols 9. anti-3,3,6,6-Tetramethylthiacycloheptane-4,5-diol anti-9 Reduction with DIBAL and ZnCl2. Zinc chloride (0.5 mol dm23 solution in THF; 3 cm3) was added to a stirred solution of the hydroxy ketone 4 (300 mg 1.49 mmol) in dry THF. The solution was cooled to 278 8C and treated with a solution of DIBAL (1 mol dm23 in THF; 1.5 cm3 1.5 mmol) added dropwise. The reaction mixture was allowed to warm to room temperature and then stirred overnight. After this methanol was added to the mixture which was then stirred for a further 30 min.The resulting inorganic precipitate was filtered off and washed several times with dichloromethane. The filtrate and organic washings were combined and evaporated under reduced pressure. The crude material (a 39 61 ratio of syn- and anti-diols 9 by NMR) was chromatographed on silica eluting with hexane–ethyl acetate (5 2) to give the syn-diol (105 mg 35%) and the anti-diol 9 (160 mg 54%) Rf [hexane–ethyl acetate (3 1)] 0.33 (Found C 58.75; H 9.89. C10H20SO2 requires C 58.78; H 9.87%); nmax(CH2Cl2)/cm21 3614–3495 (OH) and 2958–2871 (CH); dH(250 MHz; CDCl3) 3.56 (2 H br s 2 × CHOH) 2.76 (2 H br s 2 × CHOH) 2.52 (2 H d J 14.7 2 × SCHACHB) 2.27 (2 H d J 14.7 2 × SCHACHB) 1.09 (6 H s 2 × CMeAMeB) and 0.95 (6 H s 2 × CMeAMeB); dC(63 MHz; CDCl3) 73.6 (1) 47.1 (2) 38.8 (2) 28.3 (1) and 19.7 (1) (Found M1 204.1181.C10H20SO2 requires M1 204.1184); m/z 204 (M1 11%) 186 (M1 2 H2O 20) 139 (76) and 130 [M1 2 H2O 2 CH2C(Me)2 100] 86 [CH2C(Me)2C(OH)H 40] and 56 [CH2C(Me)2 52]. 4-Acetoxy-5-chloro-3,3,6,6-tetramethylthiacycloheptane 17 Trimethyl orthoacetate (0.138 cm3 1.1 mmol) was added to a solution of the syn-diol 9 (170 mg 0.83 mmol) and toluene-psulfonic acid (2 mg) in dry CH2Cl2 (2 cm3). After the mixture had been stirred at room temperature for 1 h volatile material was removed under reduced pressure and most of the residual methanol was removed by subjecting the sample to high vacuum for 1 min. The residue was then dissolved in dichloromethane and trimethylsilyl chloride (0.160 cm3 1.25 mmol) was added to the solution.The reaction mixture was stirred at room temperature for 4 days after which it was evaporated under reduced pressure. The residue was chromatographed (SiO2 hexane–ether 10 1) to give the a-chloro acetate 17 (141 mg 64%) as a colourless oil; Rf [hexane–ethyl acetate (3 1)] 0.52; nmax(CDCl3)/cm21 2962–2890 (CH) and 1732 (C]] O); dH(250 MHz; CDCl3) 5.18 (1 H d J 3.5 CHOAc) 3.78 (1 H d J 3.5 CHCl) 3.40 (2 H AB quartet J 11.1 SCHACHB) 2.89 (1 H d J 10.1 SCHCCHD) 2.52 (1 H d J 10.1 SCHCCHD) 2.11 (3 H s CO2Me) 1.12 (3 H s MeC) 1.08 (6 H s MeAMeB) and 0.97 (3 H s MeD); dC(100 MHz; CDCl3) 170.2 (2) 82.6 (1) 56.7 (1) 55.4 (2) 47.7 (2) 41.2 (2) 38.1 (2) 25.4 (1) 24.4 (1) 24.2 (1) 22.8 (1) and 21.3 (1) (Found M1 264.0954. C12H21O2SCl requires M1 264.0951); m/z 266 (M1 15%) 264 (M1 45) 221 [M1 2 C(]] O)Me 2] 204 (M1 2 HCO2Me 4) and 173 [M1 2 Cl 2 C(Me)2CH2 100].Acknowledgements We thank Zeneca Fine Chemicals and EPSRC for CASE awards (to M. J. G. S. O’S and P. W.). References 1 A. de Groot and H. Wynberg J. Org. Chem. 1966 31 3954. 2 H. Wynberg and A. de Groot J. Chem. Soc. Chem. Commun. 1965 171. 3 A. Krebs and H. Kimling Tetrahedron Lett. 1970 761. 4 A. de Groot J. A. Boerma and H. Wynberg Rec. Trav. Chim. Pays- Bas. 1969 88 994. J. Chem. Soc. Perkin Trans. 1 1997 3483 5 A. de Groot J. A. Boerma and H. Wynberg Tetrahedron Lett. 1968 2365. 6 N. Feeder M. J. Ginnelly R. V. H. Jones S. O’Sullivan and S. Warren Tetrahedron Lett. 1994 35 9095. 7 P. Y. Johnson and M. Berman J. Org. Chem. 1975 40 3046. 8 P. Y. Johnson and D.J. Kirkman J. Org. Chem. 1976 41 1768. 9 D. E. Applequist P. A. Gebauer D. E. Gwynn and L. H. O’Connor J. Am. Chem. Soc. 1972 94 4272. 10 N. Greeves in Comprehensive Organic Syntheses ed. B. M. Trost and I. Fleming Oxford 1991 vol. 8 pp.1–24. 11 A. L. Gemal and J.-L. Luche J. Am. Chem. Soc. 1981 103 5454. 12 J.-L. Luche J. Am. Chem. Soc. 1978 100 2226. 13 G. M. Sheldrick in ‘SHELX86. Program for the Solution of Crystal Structures’ University of Gottingen Germany 1990. 14 G. M. Sheldrick in ‘SHELXL93. Program for the Refinement of Crystal Structures’ University of Gottingen Germany 1993. 15 M. Hannaby and S. Warren J. Chem. Soc. Perkin Trans. 1 1989 303. 16 M. C. Carreño J. L. G. Ruano A. M. Martín C. Pedregal J. H. Rodriguez A. Rubio J. Sanchez and G. Solladié J. Org. Chem. 1990 55 2120. 17 M. C. Carreño J. L. G. Ruano M. Garrido M. P. Ruiz and G. Solladié Tetrahedron Lett. 1990 31 6653. 18 G. Solladié C. Fréchou G. Demailly and C. Greck J. Org. Chem. 1986 51 1912. 19 H. C. Kolb and K. B. Sharpless Tetrahedron 1992 48 10 515. 20 T. Nakata Y. Tani M. Hatozaki and T. Oishi Chem. Pharm. Bull. 1984 32 1411. Paper 6/05572I Received 9th August 1996 Accepted 6th August 1997 © Copyright 1997 by the Royal Society of Chemistry
ISSN:1472-7781
DOI:10.1039/a605572i
出版商:RSC
年代:1997
数据来源: RSC
|
|