Formal Total Synthesis of (±)-Conduramine E Utilising the Bryce-Smith-Gilbert Photoamination Reaction

Abstract

Utilising a Bryce-Smith-Gilbert photoamination of benzene as a key step, a synthesis of ()-conduramine E was carried out. A highly regioselective dihydroxylation of a cyclic diene was effected utilising Sharpless AD-mix-b.

We have recently reported on the use of formamide 1, prepared via Bryce-Smith-Gilbert photoamination of benzene, as a precursor for the enantioselective synthesis of (-)-fortamine. [¹] [²] The synthetic potential of this crystalline compound has now been further realized, forming the foundation for a synthesis of (±)-conduramine E (Scheme  [¹] ). [³]

Thus, beginning from formamide 1, [¹] bromonium ion induced cyclisation was investigated to install the relative stereochemistry between the adjacent carbon-nitrogen and carbon-oxygen bonds required for conduramine E. However, contrary to expectation, treatment of 1 with two equivalents of N-bromosuccinimide (NBS) delivered a 49% yield of oxazolidinone 2, presumably via hydration of the intermediate 3 and oxidation of 4 (Scheme  [²] ).

In an effort to improve the yield of this conversion we examined a two-step procedure (Scheme  [³] ). Initial treatment of 1 with polymer-supported Br₃ ^- afforded formate 5, presumably again via 4. [4] It is proposed that the acidic nature of this reagent is sufficient to cause N-protonation of 4, driving its ring opening to give 5. It is noteworthy that, in the presence of 2,6-lutidene, amidinium ion 7 was isolated, presumably via 6. The structure of 7 was confirmed by X-ray crystallographic analysis. [5] In the absence of protonation, 4 would be expected to rearrange to the thermodynamically more stable formamide 6 with subsequent cyclization to afford 7. [6] Overall, this transformation achieves the same stereochemical outcome as a Woodward-Prevost dihydroxylation. [7] Hydrolysis of the formate 5 afforded an amino alcohol that was directly protected with triphosgene to give the desired urethane 2 in an overall, purified yield of 86% from 1.

Treatment of 2 with DBU effected elimination of HBr to afford diene 8 in 90% yield (Scheme  [4] ). At this stage, synthesis of conduramine E required a regio- and stereoselective dihydroxylation to give 9. Treatment of 8 under modified Van Rheenen conditions resulted in dihydroxylation exclusively on the exo face with a 4:1 mixture of regioisomers (9/10) in 55% combined yield. [8] Sharpless asymmetric dihydroxylation reagents are usually ineffective at kinetic resolution but can be regioselective in diene dihydroxylation. [9] Indeed, when we treated 8 with AD-mix-β for five hours between 0 ˚C and -5 ˚C, 9 was obtained as a single regio- and stereoisomer in 76% yield. [¹0] The shape of the bicyclic ring system makes the exo stereo­selectivity unsurprising but the high regioselectivity is more difficult to rationalise. Unfortunately, kinetic resolution was ineffective with only 18% ee being achieved at 40% conversion with AD-mix-β.

During a study directed toward the synthesis of (+)-conduritol E, the meso-diene 12 was effectively desymmetrised to give 13 (85% ee) by treatment with AD-mix-β whilst, as expected, AD-mix-α afforded its enantiomer (Scheme  [5] ). [9] [¹¹]

To examine the effect of the cinchona alkaloid ligand on the outcome of the dihydroxylation of 8, its reaction with AD-mix-α was carried out but the same product (9) was obtained (68% yield). Thus, as suggested by the reaction under Van Rheenen conditions, the selectivity is innate to the structure of 8. Calculating the transition state energies in the exo approach of OsO₄-NH₃, as a model, to either double bond of 8 showed that leading to 9 to be 0.9 kcal mol^-¹ lower in energy than that leading to 10 (Figure  [¹] ). Whilst no firm conclusions can be made on the basis of this small difference in energies, it is consistent with the observed ratio of products obtained in the room-temperature Van Rheenen dihydroxylation. [¹²]

Deprotection of 9 was effected by refluxing with TFA to afford 11 in 76% yield, [¹³] [¹4] which has been previously reported by Prinzbach et al. as an intermediate in their synthesis of (-)-conduramine E. [³a] For completeness, utilising known conditions, 11 was hydrolysed with Ba(OH)₂ to give conduramine E then converted into its tetraacetyl derivative and its ^¹H NMR spectrum found to be in accord with data reported by Chida et al. [³b]

In conclusion, we have further demonstrated the synthetic utility of crystalline formamide 1, obtained by photoamination of benzene, as a precursor for the regio- and diastereocontrolled synthesis of natural products possessing polyhydroxylation.

Acknowledgment

We wish to recognize financial support from Reading Endowment Trust Fund (S.G. and D.C.), the University of Reading Chemical Analysis Facility for access to spectroscopic equipment, and the University of Reading and EPSRC for funds for the Oxford Diffraction Gemini X-ray diffractometer.

References and Notes

• 1a Gibson S. PhD Thesis   University of Reading; UK: 2004. 
• 1b

  Synthesis of 1
  A solution of N-tert-butylcyclohexa-2,5-dienylamine contaminated with N-tert-butylcyclohexa-2,4-dienylamine (2.5 g, 16.5 mmol)^²b and Et₃N (4.65 mL, 33.4 mmol, 2 equiv) in dry Et₂O (375 mL) was stirred under argon at 0 ˚C before formic acetic anhydride (1.75 mL, 19.8 mmol, 1.2 equiv) was added slowly, and the resulting yellow solution left stirring for 4 h whilst allowing to warm to r.t. The reaction was quenched with H₂O and the layers separated. The aqueous layer was extracted with Et₂O (3 × 150 mL), the combined organic extracts dried over MgSO₄, and the solvent removed in vacuo. The resultant amber oil was purified via flash column chromatography (Florisil^®, gradient; hexane-EtOAc = 19:1 to 2:1) to give N-tert-butyl-N-cyclohexa-2,5-dienylformamide (1) as a colourless crystalline solid (1.83g, ca. 62%); mp 44-47 ˚C; R _f = 0.11 (SiO₂, hexane-EtOAc = 7:3). IR (thin film): ν_max = 3030, 2972, 2814, 1668 (C=O stretch), 1220. ^¹H NMR (250 MHz, CDCl₃): δ = 1.42 [9 H, s, C(CH₃)₃, rotamer A], 1.48 [9 H, s, C(CH₃)₃, rot. B], 2.64-2.69 [2 H, m, H(4), rot. A and B], 4.56-4.63 [1 H, m, H(1), rot. A], 4.80-4.95 [1 H, m, H(1), rot. B], 5.57-5.91 [4 H, m, 2 × CH=CH, rot. A and B], 8.16 [1 H, s, C(O)H, rot. B], 8.51 [1 H, s, C(O)H, rot. A], [rot. A/rot. B = 2:1]. ^¹³C NMR (63 MHz, CDCl₃): δ = 25.95, 26.39 [CH₂, C(4), rots. A and B], 29.33 [CH₃, C(CH₃)₃, rot. B], 29.71 [CH₃, C(CH₃)₃, rot. A], 48.07, 49.02 [CH, C(1), rot. A and B], 56.96, 57.24 [C, C(CH₃)₃, rot. A and B], 126.28, 126.41, 126.93, 127.86 [CH, C(2,3,5,6), rot. A and B], 162.72 [C, C(O)H, rot. A], 165.89 [C, C(O)H, rot. B]. HRMS (CI): m/z calcd for C₁₁H₁₇NO [M⁺]: 179.1310; found: 179.1310. Anal. Calcd (%) for (CHN): C, 73.70; H, 9.56; N, 7.81. Found: C, 73.46; H, 9.68; N, 7.67.
• 2a Bellas M. Bryce-Smith D. Gilbert A. J. Chem. Soc., Chem. Commun.  1967,  862 
• 2b Bellas M. Bryce-Smith D. Clarke MT. Gilbert A. Klunkin G. Krestonosich S. Manning C. Wilson S. J. Chem. Soc., Perkin Trans. 1  1977,  2571 
• See also:
• 2c Yasuda M. Yamashita T. Matsumoto T. Shima K. Pac C. J. Org. Chem.  1985,  50:  3667 
  Search in Google Scholar
• 2d Yasuda M. Yamashita T. Shima K. Pac C. J. Org. Chem.  1987,  52:  753 
  Search in Google Scholar
• 2e Yasuda M. Matsuzaki Y. Shima K. Pac C. J. Chem. Soc., Perkin Trans. 2  1988,  745 
• For previous syntheses of conduramine E, see:
• 3a Spielvogel D. Kammerer J. Keller M. Prinzbach H. Tetrahedron Lett.  2000,  41:  7863 
  Search in Google Scholar
• 3b Chida N. Sakata N. Murai K. Tobe T. Nagase T. Ogawa S. Bull. Chem. Soc. Jpn.  1998,  71:  259 
  Search in Google Scholar
• 3c Trost BM. Pulley SR. Tetrahedron Lett.  1995,  36:  8737 
  Search in Google Scholar
• 6a Evans DA. Takacs JM. Tetrahedron Lett.  1980,  21:  4233 
  Search in Google Scholar
• 6b Evans DA. McGee LR. J. Am. Chem. Soc.  1981,  103:  2876 
  Search in Google Scholar
• 6c Phillips AP. Baltzly R. J. Am. Chem. Soc.  1947,  69:  200 
  Search in Google Scholar
• 7 Woodward RB. Brutcher FV. J. Am. Chem. Soc.  1958,  80:  209 
  CrossrefSearch in Google Scholar
• 8 Van Rheenen V. Kelly RC. Cha DY. Tetrahedron Lett.  1976,  17:  1973 
  CrossrefSearch in Google Scholar
• 9 Kolb HC. VanNieuwenhze MS. Sharpless KB. Chem. Rev.  1994,  94:  2483 
  Search in Google Scholar
• 11 Takano S. Yoshimitsu T. Ogasawara K. J. Org. Chem.  1994,  59:  54 
  CrossrefSearch in Google Scholar
• 12 Frisch MJ. Trucks GW. Schlegel HB. Scuseria GE. Robb MA. Cheeseman JR. Montgomery JA. Vreven T. Kudin KN. Burant JC. Millam JM. Iyengar SS. Tomasi J. Barone V. Mennucci B. Cossi M. Scalmani G. Rega N. Petersson GA. Nakatsuji H. Hada M. Ehara M. Toyota K. Fukuda R. Hasegawa J. Ishida M. Nakajima T. Honda Y. Kitao O. Nakai H. Klene M. Li X. Knox JE. Hratchian HP. Cross JB. Bakken V. Adamo C. Jaramillo J. Gomperts R. Stratmann RE. Yazyev O. Austin AJ. Cammi R. Pomelli C. Ochterski JW. Ayala PY. Morokuma K. Voth GA. Salvador P. Dannenberg JJ. Zakrzewski VG. Dapprich S. Daniels AD. Strain MC. Farkas O. Malick DK. Rabuck AD. Raghavachari K. Foresman JB. Ortiz JV. Cui Q. Baboul AG. Clifford S. Cioslowski J. Stefanov BB. Liu G. Liashenko A. Piskorz P. Komaromi I. Martin RL. Fox DJ. Keith T. Al-Laham MA. Peng CY. Nanayakkara A. Challacombe M. Gill PMW. Johnson B. Chen W. Wong MW. Gonzalez C. Pople JA. Gaussian 03, Revision C.02   Gaussian, Inc.; Wallingford USA: 2004. 
• 13 Knapp S. Patel DV. J. Org. Chem.  1984,  49:  5072 
  CrossrefSearch in Google Scholar
• 15 CrysAlis   Oxford Diffraction Ltd.; Abingdon UK: 2006. 
• 16 Sheldrick GM. Acta Crystallogr., Sect. A: Fundam. Crystallogr.  2008,  64:  112 ; Shelxs97 and Shelxl97, Programs for Crystallographic Solution and Refinement
  CrossrefPubMedSearch in Google Scholar
• 17 ABSPACK   Oxford Diffraction Ltd.; Oxford UK: 2005. 

Amberlite^® IRA-900 Br₃ ^- form, purchased from Fluka.

Crystal Data
C₁₁H₂₀BrNO₃, M = 294.19, monoclinic, Z = 4, spacegroup P2₁/a, a = 11.297 (14) Å, b = 9.511 (11) Å, c = 13.553 (14) Å, β = 106.50 (1)˚, U = 1396 (3) Å^³. 2765 data were collected with MoKα radiation at 150 K using the Oxford Diffraction X-Calibur CCD System. The crystal was positioned at 50 mm from the CCD. 321 frames were measured with a counting time of 10 s. Data analysis was carried out with the CrysAlis program.^¹5 The structure was solved using direct methods with the Shelxs97 program.^¹6 The nonhydrogen atoms were refined with anisotropic thermal parameters. The hydrogen atoms bonded to carbon were included in geometric positions and given thermal parameters equivalent to 1.2 times those of the atom to which they were attached. An absorption correction was applied using ABSPACK.^¹7 The structure was refined on F^² using Shelxl97 [2] to R1 = 0.0903, wR2 = 0.1744 for 872 reflections with I > 2σ(I). Details of the structure have been deposited at the Cambridge Crystallographic Data Centre as CCDC 752251.

Procedure
AD-mix-β [K₃Fe(CN)₆ (0.35 g, 1.08 mmol, 3 equiv), K₂CO₃ (0.16 g, 1.08 mmol, 3 equiv), (DHQD)₂PHAL (2.5 mol%) and K₂OsO₄˙2H₂O (2.5 mol%)] were dissolved in t-BuOH-H₂O (1:1; 20 mL) and stirred for 5 min before addition of MeSO₂NH₂ (29 mg, 0.36 mmol, 1 equiv). The solution was cooled to 0 ˚C before addition of (±)-(3aS,7aR)-3-(tert-butyl)-3,3a,7a-trihydrobenzoxazol-2-one (8, 70 mg, 0.36 mmol) in t-BuOH (1 mL). The reaction was left stirring for 5 h between 0 ˚C and -5 ˚C. The reaction was diluted with MeOH and evaporated to dryness in vacuo. The residue was dissolved in CHCl₃-MeOH (9:1) and filtered through a plug of Celite^® upon silica gel. The filtrate was concentrated in vacuo to give the crude diol which was purified by flash chromatography (SiO₂, CHCl₃-MeOH = 9:1) to yield (±)-(3aS,6S,7S,7aS)-3-(tert-butyl)-6,7-dihydroxy-3,6,7,3a,7a-pentahydrobenzoxazol-2-one (9) as a colourless oil that solidified on standing (63 mg, 76%); mp 63-65 ˚C; R _f = 0.17 (SiO₂, CHCl₃-MeOH = 19:1). IR (CHCl₃): ν_max 3441 (OH stretch), 1728 (C=O stretch) cm^-¹. ^¹H NMR (250 MHz, CDCl₃): δ = 1.37 [9 H, s, -C(CH ₃)₃], 3.10 (1 H, br s, OH), 3.54 (1 H, d, J = 7 Hz, OH), 4.24-4.27 [2 H, m, H(3a,7)], 4.37-4.40 [1 H, m, H(6)], 4.56-4.61 [1 H, m H(7a)], 5.75 [2 H, br s, H(4,5)]. ^¹³C NMR (63 MHz, CDCl₃): δ = 28.79 [-C(CH₃)₃], 52.52 [CH, C(3a)], 54.37 [C, C(CH₃)₃], 64.25 [CH, C(6)], 67.51 [CH, C(7)], 73.47 [CH, C(7a)], 124.87 [CH, C(4)], 131.60 [CH, C(5)], 156.43 (C, C=O). MS (CI): m/z (%) = 228 (37) [MH⁺]. HRMS: m/z calcd for C₁₁H₁₈NO₄: 228.1236; found: 228.1242.

The structure of 11 was confirmed by X-ray crystallography.

References and Notes

• 1a Gibson S. PhD Thesis   University of Reading; UK: 2004. 
• 1b

  Synthesis of 1
  A solution of N-tert-butylcyclohexa-2,5-dienylamine contaminated with N-tert-butylcyclohexa-2,4-dienylamine (2.5 g, 16.5 mmol)^²b and Et₃N (4.65 mL, 33.4 mmol, 2 equiv) in dry Et₂O (375 mL) was stirred under argon at 0 ˚C before formic acetic anhydride (1.75 mL, 19.8 mmol, 1.2 equiv) was added slowly, and the resulting yellow solution left stirring for 4 h whilst allowing to warm to r.t. The reaction was quenched with H₂O and the layers separated. The aqueous layer was extracted with Et₂O (3 × 150 mL), the combined organic extracts dried over MgSO₄, and the solvent removed in vacuo. The resultant amber oil was purified via flash column chromatography (Florisil^®, gradient; hexane-EtOAc = 19:1 to 2:1) to give N-tert-butyl-N-cyclohexa-2,5-dienylformamide (1) as a colourless crystalline solid (1.83g, ca. 62%); mp 44-47 ˚C; R _f = 0.11 (SiO₂, hexane-EtOAc = 7:3). IR (thin film): ν_max = 3030, 2972, 2814, 1668 (C=O stretch), 1220. ^¹H NMR (250 MHz, CDCl₃): δ = 1.42 [9 H, s, C(CH₃)₃, rotamer A], 1.48 [9 H, s, C(CH₃)₃, rot. B], 2.64-2.69 [2 H, m, H(4), rot. A and B], 4.56-4.63 [1 H, m, H(1), rot. A], 4.80-4.95 [1 H, m, H(1), rot. B], 5.57-5.91 [4 H, m, 2 × CH=CH, rot. A and B], 8.16 [1 H, s, C(O)H, rot. B], 8.51 [1 H, s, C(O)H, rot. A], [rot. A/rot. B = 2:1]. ^¹³C NMR (63 MHz, CDCl₃): δ = 25.95, 26.39 [CH₂, C(4), rots. A and B], 29.33 [CH₃, C(CH₃)₃, rot. B], 29.71 [CH₃, C(CH₃)₃, rot. A], 48.07, 49.02 [CH, C(1), rot. A and B], 56.96, 57.24 [C, C(CH₃)₃, rot. A and B], 126.28, 126.41, 126.93, 127.86 [CH, C(2,3,5,6), rot. A and B], 162.72 [C, C(O)H, rot. A], 165.89 [C, C(O)H, rot. B]. HRMS (CI): m/z calcd for C₁₁H₁₇NO [M⁺]: 179.1310; found: 179.1310. Anal. Calcd (%) for (CHN): C, 73.70; H, 9.56; N, 7.81. Found: C, 73.46; H, 9.68; N, 7.67.
• 2a Bellas M. Bryce-Smith D. Gilbert A. J. Chem. Soc., Chem. Commun.  1967,  862 
• 2b Bellas M. Bryce-Smith D. Clarke MT. Gilbert A. Klunkin G. Krestonosich S. Manning C. Wilson S. J. Chem. Soc., Perkin Trans. 1  1977,  2571 
• See also:
• 2c Yasuda M. Yamashita T. Matsumoto T. Shima K. Pac C. J. Org. Chem.  1985,  50:  3667 
  Search in Google Scholar
• 2d Yasuda M. Yamashita T. Shima K. Pac C. J. Org. Chem.  1987,  52:  753 
  Search in Google Scholar
• 2e Yasuda M. Matsuzaki Y. Shima K. Pac C. J. Chem. Soc., Perkin Trans. 2  1988,  745 
• For previous syntheses of conduramine E, see:
• 3a Spielvogel D. Kammerer J. Keller M. Prinzbach H. Tetrahedron Lett.  2000,  41:  7863 
  Search in Google Scholar
• 3b Chida N. Sakata N. Murai K. Tobe T. Nagase T. Ogawa S. Bull. Chem. Soc. Jpn.  1998,  71:  259 
  Search in Google Scholar
• 3c Trost BM. Pulley SR. Tetrahedron Lett.  1995,  36:  8737 
  Search in Google Scholar
• 6a Evans DA. Takacs JM. Tetrahedron Lett.  1980,  21:  4233 
  Search in Google Scholar
• 6b Evans DA. McGee LR. J. Am. Chem. Soc.  1981,  103:  2876 
  Search in Google Scholar
• 6c Phillips AP. Baltzly R. J. Am. Chem. Soc.  1947,  69:  200 
  Search in Google Scholar
• 7 Woodward RB. Brutcher FV. J. Am. Chem. Soc.  1958,  80:  209 
  CrossrefSearch in Google Scholar
• 8 Van Rheenen V. Kelly RC. Cha DY. Tetrahedron Lett.  1976,  17:  1973 
  CrossrefSearch in Google Scholar
• 9 Kolb HC. VanNieuwenhze MS. Sharpless KB. Chem. Rev.  1994,  94:  2483 
  Search in Google Scholar
• 11 Takano S. Yoshimitsu T. Ogasawara K. J. Org. Chem.  1994,  59:  54 
  CrossrefSearch in Google Scholar
• 12 Frisch MJ. Trucks GW. Schlegel HB. Scuseria GE. Robb MA. Cheeseman JR. Montgomery JA. Vreven T. Kudin KN. Burant JC. Millam JM. Iyengar SS. Tomasi J. Barone V. Mennucci B. Cossi M. Scalmani G. Rega N. Petersson GA. Nakatsuji H. Hada M. Ehara M. Toyota K. Fukuda R. Hasegawa J. Ishida M. Nakajima T. Honda Y. Kitao O. Nakai H. Klene M. Li X. Knox JE. Hratchian HP. Cross JB. Bakken V. Adamo C. Jaramillo J. Gomperts R. Stratmann RE. Yazyev O. Austin AJ. Cammi R. Pomelli C. Ochterski JW. Ayala PY. Morokuma K. Voth GA. Salvador P. Dannenberg JJ. Zakrzewski VG. Dapprich S. Daniels AD. Strain MC. Farkas O. Malick DK. Rabuck AD. Raghavachari K. Foresman JB. Ortiz JV. Cui Q. Baboul AG. Clifford S. Cioslowski J. Stefanov BB. Liu G. Liashenko A. Piskorz P. Komaromi I. Martin RL. Fox DJ. Keith T. Al-Laham MA. Peng CY. Nanayakkara A. Challacombe M. Gill PMW. Johnson B. Chen W. Wong MW. Gonzalez C. Pople JA. Gaussian 03, Revision C.02   Gaussian, Inc.; Wallingford USA: 2004. 
• 13 Knapp S. Patel DV. J. Org. Chem.  1984,  49:  5072 
  CrossrefSearch in Google Scholar
• 15 CrysAlis   Oxford Diffraction Ltd.; Abingdon UK: 2006. 
• 16 Sheldrick GM. Acta Crystallogr., Sect. A: Fundam. Crystallogr.  2008,  64:  112 ; Shelxs97 and Shelxl97, Programs for Crystallographic Solution and Refinement
  CrossrefPubMedSearch in Google Scholar
• 17 ABSPACK   Oxford Diffraction Ltd.; Oxford UK: 2005. 

Amberlite^® IRA-900 Br₃ ^- form, purchased from Fluka.

Crystal Data
C₁₁H₂₀BrNO₃, M = 294.19, monoclinic, Z = 4, spacegroup P2₁/a, a = 11.297 (14) Å, b = 9.511 (11) Å, c = 13.553 (14) Å, β = 106.50 (1)˚, U = 1396 (3) Å^³. 2765 data were collected with MoKα radiation at 150 K using the Oxford Diffraction X-Calibur CCD System. The crystal was positioned at 50 mm from the CCD. 321 frames were measured with a counting time of 10 s. Data analysis was carried out with the CrysAlis program.^¹5 The structure was solved using direct methods with the Shelxs97 program.^¹6 The nonhydrogen atoms were refined with anisotropic thermal parameters. The hydrogen atoms bonded to carbon were included in geometric positions and given thermal parameters equivalent to 1.2 times those of the atom to which they were attached. An absorption correction was applied using ABSPACK.^¹7 The structure was refined on F^² using Shelxl97 [2] to R1 = 0.0903, wR2 = 0.1744 for 872 reflections with I > 2σ(I). Details of the structure have been deposited at the Cambridge Crystallographic Data Centre as CCDC 752251.

Procedure
AD-mix-β [K₃Fe(CN)₆ (0.35 g, 1.08 mmol, 3 equiv), K₂CO₃ (0.16 g, 1.08 mmol, 3 equiv), (DHQD)₂PHAL (2.5 mol%) and K₂OsO₄˙2H₂O (2.5 mol%)] were dissolved in t-BuOH-H₂O (1:1; 20 mL) and stirred for 5 min before addition of MeSO₂NH₂ (29 mg, 0.36 mmol, 1 equiv). The solution was cooled to 0 ˚C before addition of (±)-(3aS,7aR)-3-(tert-butyl)-3,3a,7a-trihydrobenzoxazol-2-one (8, 70 mg, 0.36 mmol) in t-BuOH (1 mL). The reaction was left stirring for 5 h between 0 ˚C and -5 ˚C. The reaction was diluted with MeOH and evaporated to dryness in vacuo. The residue was dissolved in CHCl₃-MeOH (9:1) and filtered through a plug of Celite^® upon silica gel. The filtrate was concentrated in vacuo to give the crude diol which was purified by flash chromatography (SiO₂, CHCl₃-MeOH = 9:1) to yield (±)-(3aS,6S,7S,7aS)-3-(tert-butyl)-6,7-dihydroxy-3,6,7,3a,7a-pentahydrobenzoxazol-2-one (9) as a colourless oil that solidified on standing (63 mg, 76%); mp 63-65 ˚C; R _f = 0.17 (SiO₂, CHCl₃-MeOH = 19:1). IR (CHCl₃): ν_max 3441 (OH stretch), 1728 (C=O stretch) cm^-¹. ^¹H NMR (250 MHz, CDCl₃): δ = 1.37 [9 H, s, -C(CH ₃)₃], 3.10 (1 H, br s, OH), 3.54 (1 H, d, J = 7 Hz, OH), 4.24-4.27 [2 H, m, H(3a,7)], 4.37-4.40 [1 H, m, H(6)], 4.56-4.61 [1 H, m H(7a)], 5.75 [2 H, br s, H(4,5)]. ^¹³C NMR (63 MHz, CDCl₃): δ = 28.79 [-C(CH₃)₃], 52.52 [CH, C(3a)], 54.37 [C, C(CH₃)₃], 64.25 [CH, C(6)], 67.51 [CH, C(7)], 73.47 [CH, C(7a)], 124.87 [CH, C(4)], 131.60 [CH, C(5)], 156.43 (C, C=O). MS (CI): m/z (%) = 228 (37) [MH⁺]. HRMS: m/z calcd for C₁₁H₁₈NO₄: 228.1236; found: 228.1242.

The structure of 11 was confirmed by X-ray crystallography.