Stereochemical course of the hydrogen migration in the boron trifluoride etherate-catalyzed rearrangement of 1,1-disubstituted epoxides

Article abstract of DOI:10.1016/S0957-4166(00)00139-7

Mechanistic studies on the BF₃.Et₂O-catalyzed rearrangement of optically active, regioselectively deuterated 1,1-disubstituted epoxides to aldehydic products revealed that the two hydrogens migrate at the migration terminus with oppo

Full text of DOI:10.1016/S0957-4166(00)00139-7

Tetrahedron: Asymmetry 11 (2000) 1859±1868
Stereochemical course of the hydrogen migration in the boron
tri¯uoride etherate-catalyzed rearrangement of
1,1-disubstituted epoxides
Noriyuki Hara, Akiyoshi Mochizuki, Akinori Tatara and Yoshinori Fujimoto*
Department of Chemistry and Materials Science, Tokyo Institute of Technology, Meguro, Tokyo 152-8551, Japan
Received 24 September 1999; accepted 3 April 2000
Abstract
.
Mechanistic studies on the BF3 Et2O-catalyzed rearrangement of optically active, regioselectively deuterated
1,1-disubstituted epoxides to aldehydic products revealed that the two hydrogens migrate at the migration
terminus with opposite stereochemical preferences, i.e. the hydrogen anti to the bulky substituent prefers to
migrate with inversion of con®guration, whereas the hydrogen syn to the bulky substituent prefers to
migrate with retention of con®guration. # 2000 Elsevier Science Ltd. All rights reserved.
1. Introduction
Lewis acid-catalyzed rearrangement of epoxides to carbonyl compounds is a well-known reaction
in organic chemistry.1 It is generally explained that this reaction proceeds by coordination of a
Lewis acid molecule on the oxygen atom, ®ssion of a C±O bond to form an electron-de®cient
carbon center at the more substituted carbon atom, and migration of a substituent to the adjacent
carbon center with concomitant formation of a carbonyl compound. A number of examples have
been reported on epoxides embedded in cyclic systems such as steroidal epoxides.2 In these cases,
the structure of the carbonyl product, including the stereochemistry at the migration terminus,
seems to be governed by the rigid conformation of the ring system and would be predictable. In
contrast, few studies have been done on the stereochemistry at the migration terminus in the
rearrangement of acyclic epoxides.3 5 We previously reported the unique results that the hydrogen
migrates preferentially with retention of con®guration at the migration terminus, whereas an alkyl
.
group migrates exclusively with inversion of con®guration in the BF3 Et2O-catalyzed rearrangement
of acyclic trisubstituted epoxides.6 This ®nding prompted us to study in more detail the mechanism
of Lewis acid-catalyzed rearrangement for acyclic epoxides. We describe here our results on the
* Corresponding author. Tel/fax: +81-3-5734-2241; e-mail: fujimoto@cms.titech.ac.jp
0957-4166/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(00)00139-7

1860
N. Hara et al. / Tetrahedron: Asymmetry 11 (2000) 1859±1868
.
mechanism of and, in particular, the steric course of the hydrogen migration in the BF3 Et2O-
catalyzed rearrangement of acyclic 1,1-disubstituted epoxides to aldehydic products.
2. Results and discussion
In our preliminary studies, optically active epoxides 1 and 2 were individually treated with
.
BF3 Et2O (0.2±2.0 equiv.) in CH2Cl2 or benzene at room temperature, and the resulting aldehydes
were analysed to be essentially racemic.7 We, therefore, selected the optically active epoxide 3
which bears the tertiary and methyl substituents. To distinguish the two hydrogens and facilitate
the analysis of the steric course of the rearrangement, the regioselectively deuterated substrates,
(1S,2S)-epoxide 3a and (1R,2S)-epoxide 3b, were synthesized.
The synthesis of the epoxides 3a and 3b is outlined in Scheme 1. The aldehyde 4 was converted
into the acetylene 5b and its deuterated form 5a via dibromoole®n. Carbozirconation of 5b by
Negishi's protocol8 followed by quenching with D2O gave the E-ole®n 6a (E:Z=16:1). The cor-
responding Z-ole®n 6b (Z:E=21:1) was obtained from 5a by the same procedure, but by
quenching with H2O. Sharpless asymmetric dihydroxylation9 of 6a and 6b gave the (1S,2S)-diols
7a (55% ee) and (1R,2S)-7b (73% ee), respectively. The enantiomeric excesses of 7a and 7b were
improved by repeated recrystallization of their phthalic half-esters, followed by hydrolysis to give
7a (86% ee) and 7b (92% ee). Extension of the methylene chain of 7a and 7b was achieved in
seven steps to give the diols 8a and 8b, respectively.10 The diols were ®nally converted into the
epoxides 3a and 3b via the corresponding mesylates.
Scheme 1. Synthesis of epoxides 3a and 3b. Reagents: (i) CBr4, Ph3P; (ii) nBuLi, H2O or D2O; (iii) Me3Al, Cp2ZrCl2,
D2O or H2O; (iv) (DHQ)2PHAL, K2OsO2(OH)4, K3Fe(CN)6, K2CO3, CH3SO2NH2; (v) (a) phthalic anhydride, py, (b)
recrystallization, (c) NaOH, MeOH; (vi) Me2C(OMe)2, pTsOH; (vii) (a) RuCl3, NaIO4, (b) CH2N2, (c) LiAlH4, (d)
TsCl, py; (viii) (a) Ph(CH2)3MgBr, Li2CuCl4, (b) H+, MeOH; (ix) (a) MsCl, py, (b) K2CO3, MeOH

N. Hara et al. / Tetrahedron: Asymmetry 11 (2000) 1859±1868
1861
.
Brief treatment of (1S,2S)-3a with BF3 Et2O (2 equiv.) in CH2Cl2 at room temperature a€orded
a crude product containing aldehyde 9a±d (21% yield).11 A part of the product was reduced
with LiAlD4 to give the corresponding alcohol which was then converted into the (R)-MTPA
ester 10a±d. The other part of the product was reduced with LiAlH4 and the resulting alcohol was
converted into the same ester derivative 11a±d (Scheme 2).
.
Scheme 2. BF3 Et2O-catalyzed rearrangement of the epoxide 3a and analysis of the steric course of the migration of
the hydrogen or deuterium atom in the forms of 10 and 11
The 1H NMR spectrum of 10a±d exhibited four broad singlets in the oxymethine/oxymethylene
region (ꢀ 4.49, 4.41, 4.03 and 3.95) (Figure 1). Among compounds 10a±10d, only compounds 10a
and 10b which arose from the migration of the deuterium atom exhibited the oxymethine signals
in this spectrum. The two signals (ꢀ 4.41 and 4.03) were assigned to the oxymethine protons of the
diastereomers 10a, and the remaining two to those of the diastereomers 10b.12 The ratio of the
compounds, 10a, 10b and 10c+10d (formed by hydrogen migration), was determined to be
31:14:55 on the basis of the signal intensities of the above oxymethine hydrogens by comparison
1
with those of the benzylic methylene hydrogens (ꢀ 2.60). The H NMR spectrum of 11a±d
exhibited eight sets of doublets in the oxymethine region (Figure 1). The two doublets at ꢀ 4.43,
4.05 (J=10.8 Hz) were readily assigned to those of 11a and the two doublets at ꢀ 4.51, 3.97
(J=10.8 Hz) to 11b in comparison with the spectrum of 10a±d. The two doublets at ꢀ 4.04 (J=8.8
Hz) and 4.42 (J=3.7 Hz) were assigned to the oxymethine protons of the diastereomers 11c and
the last two doublets at ꢀ 3.95 (J=8.8 Hz) and 4.50 (J=3.7 Hz) to those of the diastereomers
11d.9 Although the ratio of the signal intensities of these eight doublets could not be accurately
obtained due to some overlap of the signals, the ratio (58:42) of (2Â11a+11c) and (2Â11b+11d)
was readily obtained from the integration values of their signals. With the aforementioned two
sets of ratios in hand, the ratio for 9a:9b:9c:9d was calculated to be 31:14:22:33.13,14
.
The deuterium labeled (1S,2S)-epoxide 3b was similarly treated with BF3 Et2O and the result-
ing aldehyde 9 was analyzed in the forms of the (R)-MTPA ester derivatives 10 and 11. The ratio

1862
N. Hara et al. / Tetrahedron: Asymmetry 11 (2000) 1859±1868
Figure 1. 1H NMR spectra (270 MHz, CDCl3) of the oxymethylene/oxymethine protons of 10a±d (left) and 11a±d
(right) derived from the aldehyde 9a±d obtained from the epoxide 3a. The signals at ꢀ 4.15±4.36 in the right spectrum
are due to impurities
9a:9b:9c:9d=11:17:50:22 was obtained.14 One may suspect racemization of the once-formed
aldehyde 9 during the reaction conditions. This racemization was found to be negligible by sub-
jecting a synthetic optically active aldehyde15 to the reaction conditions.
Based on the above two sets of the data for 9a±d, it is deduced that the deuterium migration is
retarded by 1.77-fold16 compared to the hydrogen migration. Thus, the ratio of the four rear-
ranged products corresponding to 9a, 9b, 9c and 9d could be estimated as 41:18:16:25 for non-
deuterated epoxide (Scheme 3). This ratio would re¯ect the relative contribution of four transition
conformers A, B, C and D in this order, wherein either hydrogen Ha or Hb is placed in the proper
orientation for the 1,2-hydride shift (Scheme 4). The unstable carbocation (conformation E)17
generated by the C±O bond ®ssion of the epoxide would rotate along the central C±C+ axis to
give rise to conformers A±D. Out of the two possible directions of the rotation, anti-clockwise
rotation seems more preferable than the clockwise, due to the larger non-bonded repulsive inter-
action between the RCMe2 and O-BF3 groups. Among these, conformation A, wherein Hb
migrates with inversion of stereochemistry, seems most readily accessible, and conformation D,
which corresponds to the migration of Ha with retention of stereochemistry, might follow as the
next accessible one. Blackett et al. reported that Hb (the hydrogen anti to the bulky substituent) is
more susceptible to the migration (by ca. 1.9 fold) than Ha by considering only two conforma-
.
tions A and D in the BF3 Et2O-catalyzed rearrangement of a similar 1,1-disubstituted epoxide
without information on the steric course of the hydrogen migration.4 The reported aptitude was
con®rmed in the present experiments in which Hb migrated more readily [1.44 (=59/41) fold]
than Ha.
Scheme 3. The ratio of the rearranged products expected for the non-labeled epoxide

N. Hara et al. / Tetrahedron: Asymmetry 11 (2000) 1859±1868
1863
Scheme 4. Four transition conformations A±D for the hydrogen migration, arising from the initially formed con-
formation E
Concerning the migration of Hb, the retention:inversion ratio (22:50 or 14:31) indicates that
conformation B which corresponds to the hydrogen migration with retention of stereochemistry
has to be considered as well, although its contribution is less than half of that of conformation A.
Similarly, it was found that conformation C, which is responsible for the migration of Ha with
inversion of stereochemistry, cannot be ignored on the basis of the retention:inversion ratio of
Ha (33:22 or 17:11). It can be summarized that the relative signi®cance of the four transition
conformations is A, D, B, C in the decreasing order.
In conclusion, the present study has investigated the steric course of each hydrogen in the
.
BF3 Et2O-catalyzed rearrangement of 1,1-disubstituted epoxide for the ®rst time, and established
that the hydrogen anti to the bulky substituent prefers to migrate with inversion of con®guration
at the migrating terminus, whereas the hydrogen syn to the bulky substituent prefers to migrate
with retention of con®guration. The relative contribution of the four transition conformers in the
rearrangement is also estimated.
3. Experimental
3.1. General
All melting points were measured on a Yazawa hot-stage microscope BY-1 and are reported
uncorrected. The IR spectra were measured on a JASCO FT-IR200 spectrometer in CHCl3
1
solutions and are reported in cm^1. H and 13C NMR spectra were recorded on a JEOL EX-400
or EX-270 spectrometer in CDCl3 solutions and chemical shifts were reported in ꢀ value based on
internal tetramethylsilane (ꢀH 0.0) or the solvent signal (ꢀC=77.0). 2H NMR spectra were recorded
on a JEOL GSX-500 (76.8 MHz) spectrometer in CHCl3 solutions and chemical shifts are referenced
to natural abundance CDCl3 (ꢀD 7.26). EI-MS (70 eV) and high-resolution FABMS spectra were
obtained on a JEOL JMS-AX505HA spectrometer.

1864
N. Hara et al. / Tetrahedron: Asymmetry 11 (2000) 1859±1868
3.2. [1-2H]-3,3-Dimethyl-4-phenyl-1-butyne 5a
A mixture of PPh3 (110 g, 419 mmol) and CBr4 (69.4 g, 209 mmol) in dry CH2Cl2 (136 ml) was
stirred at 0ꢀC for 10 min. A solution of 2,2-dimethyl-3-phenylpropanal 4 (11.3 g, 69.8 mmol) in
CH2Cl2 (26 ml) was added and stirring was continued at 0ꢀC for 10 min and then at room
temperature overnight. The mixture was ®ltered by suction and the ®ltrate was diluted with ether
and aq. NaHCO3. Extractive (ether) workup and chromatography on silica gel (hexane:benzene,
20:1) a€orded the dibromoole®n (19.4 g, 87%) as a colorless oil. IR ꢁmax: 2970, 2930, 1600, 1500,
1470, 1450, 1370. 1H NMR ꢀ: 1.19 (6H, s, 3-CH3Â2), 2.80 (2H, s, 4-H2), 6.49 (1H, s, 2-H), 7.13±
7.31 (5H, m, Ph). 13C NMR ꢀ: 27.15, 39.98, 46.63, 85.99, 126.27, 127.89, 130.51, 137.99, 145.86.
Anal. calcd for C12H14Br2: C, 45.32; H, 4.44. Found: C, 45.16; H, 4.37.
nBuLi (1.70M hexane solution, 39.1 ml, 66.5 mmol) was added to a solution of the di-
bromoole®n (10.8 g, 33.9 mmol) in dry THF (45 ml) at ^78ꢀC under N2 and the mixture was
stirred for 15 min at the same temperature. D2O (5.0 ml) was added to this mixture and the whole
was stirred for 30 min after removal of the cooling bath. Extractive workup (ether) and chromato-
graphy on silica gel (hexane:ether 30:1) gave 5a (5.07 g, 94%) as an oil. IR ꢁmax: 3300, 3010, 2970,
1
2930, 2590, 1490, 1450, 1360. H NMR ꢀ: 1.21 (6H, s, 3-CH3Â2), 2.71 (2H, s, 4-H2), 7.19±7.31
(5H, m, Ph). 13C NMR ꢀ: 28.92, 32.00, 48.68, 68.84 (t, J=38 Hz), 90.87 (t, J=7.4 Hz) 126.38,
127.57, 127.67, 130.40, 130.51, 137.92. HR-EIMS (70 eV) calcd for C12H13D: 159.1157 (M+).
Found: 159.1132.
3.3. 3,3-Dimethyl-4-phenyl-1-butyne 5b
The dibromoole®n (7.92 g, 24.9 mmol) was similarly treated with n-BuLi and the reaction
mixture was quenched by the addition of water (4.0 ml) in place of D2O. Extractive workup and
chromatography on silica gel gave 5b (3.62 g, 92%) as a colorless oil. IR ꢁmax: 3305, 3005, 2970,
2930, 2110, 1490, 1460. 1H NMR ꢀ: 1.20 (6H, s, 3-CH3Â2), 2.10 (1H, s, 1-acetylene), 2.70 (2H, s,
4-H2), 7.18±7.30 (5H, m, Ph). 13C NMR ꢀ: 28.88, 31.97, 48.61, 69.09, 91.23, 126.38, 127.64,
130.37, 130.48, 137.84. HR-EIMS (70 eV) calcd for C12H14: 158.1096 (M+). Found: 158.1081.
3.4. [1-2H]-(1R,2S)-2,3,3-Trimethyl-4-phenyl-1,2-butanediol 7a
Trimethylaluminum (2.0 M toluene solution, 18.2 ml, 36.3 mmol) was added dropwise to a
solution of Cp2ZrCl2 (5.33 g, 18.2 mmol) in dry 1,2-dichloroethane (82 ml) at room temperature
and the mixture was stirred for 10 min. A solution of acetylene 5a (2.91 g, 18.3 mmol) in dry 1,2-
dichloroethane (15 ml) was added and stirring was continued for 2 days. Extractive (ether)
workup and chromatography on silica gel (hexane) a€orded a mixture (2.47 g, 77%) of Z-ole®n
6a (E:Z=1:21) and trans-1,2-disubstituted ole®n as a colorless oil.
A mixture of (DHQ)2PHAL (590 mg, 0.76 mmol), K3Fe(CN)6 (13.4 g, 40.6 mmol), K2CO3
(5.61 g, 40.6 mmol), CH3SO2NH2 (1.29 g, 13.5 mmol) in tBuOH (63 ml) and water (68 ml) was
vigorously stirred at room temperature for 5 min, and then cooled to 0ꢀC. K2OsO2(OH)4 (30 mg,
0.081 mmol) was added and the mixture was further stirred for 40 min. The above ole®n (2.37 g,
13.5 mmol) in tBuOH (5.0 ml) was added dropwise and stirring was continued at 0ꢀC for 3 days and
.
at room temperature for 2 days. Na2S2O3 5H2O (40 g) was added and stirring was continued for
0.5 h. Extractive (AcOEt) workup and chromatography on silica gel (hexane:AcOEt, 4:1) a€orded
7a (1.70 g, 60%) and 4,4-dimethyl-5-phenylpentane-2,3-diol (1.10 g, 39%) as a white solid. 7a: mp

N. Hara et al. / Tetrahedron: Asymmetry 11 (2000) 1859±1868
1865
45±48ꢀC. IR ꢁmax: 3620, 3560, 3450, 3010, 2970, 1370, 1080, 1040. 1H NMR ꢀ: 0.84, 0.85 (3H each,
s, 3-CH3Â2), 1.30 (3H, s, 4-CH3), 2.68 (2H, s, 4-H2), 3.51 (1H, brs, 1-H), 7.13±7.29 (5H, m, Ph).
13C NMR ꢀ: 19.88, 21.35, 21.41, 40.21, 42.19, 65.54 (J=21.8 Hz), 76.85, 125.83, 127.61, 131.09,
138.85. Anal. calcd for C13H19DO2: C, 74.60; H+D, 10.11. Found: C, 74.31; H+D, 10.08. The ee
1
1
of 7a was determined to be 56% by H NMR analysis of the mono-(R)-MTPA ester. H NMR
ꢀ: 0.86 (6H, s, 3-CH3Â2), 1.25 (3H, s, 2-CH3), 1.61±1.71 (1H, br, OH), 2.69 (2H, s, 4-H2), 3.58
(3H, s, OMe), 4.25 (1H, s, 1-H1, for (2S)-isomer), 4.32 (1H, s, 1-H1, for (2R)-isomer), 7.11±7.59
(10H, Ph).
A mixture of the diol 7a (1.64 g, 7.85 mmol) and phthalic anhydride (1.26 g, 8.51 mmol) in
pyridine was stirred at room temperature overnight. Addition of ice chips and extractive (ether)
workup followed by chromatography on silica gel (hexane:AcOEt, 4:1) a€orded phthalate half-ester
as a white solid, mp 128±130ꢀC. The ester was repeatedly recrystallized from hexane and ether to a€ord
an optically enriched ester (1.10 g). Hydrolysis of the ester by treating with 5N aq. NaOH (1.5 ml)
in MeOH (10 ml) followed by chromatography on silica gel furnished 7a (580 mg, 35%, 86% ee).
3.5. [1-2H]-(1S,2S)-2,3,3-Trimethyl-4-phenyl-1,2-butanediol 7b
The acetylene 5b (2.88 g, 18.2 mmol) was subjected to the carbozirconation as described for 5a
and the reaction was quenched by addition of D2O in place of H2O. Extractive workup and
puri®cation gave a mixture (2.30 g, 72%) of 6b (E:Z=16:1) and trans-1,2-disubstituted ole®n as
an oil.
Asymmetric dihydroxylation of the mixture, as described for 6a, a€orded 7b (1.71 g, 62%, 74%
ee) and 4,4-dimethyl-5-phenylpentane-2,3-diol (603 mg, 24%) as white solids. Compound 7b: mp
44±47ꢀC. IR ꢁmax: 3630, 3560, 3450, 3010, 2980, 1450, 1370, 1100, 1030. H NMR ꢀ: 0.84, 0.85
1
(3H each, s, 3-CH3Â2), 1.30 (3H, s, 4-CH3), 2.68 (2H, s, 4-H2), 3.81 (1H, brs, 1-H), 7.12, 7.30
(5H, m, Ph). 13C NMR ꢀ: 19.93, 21.37, 21.44, 40.24, 42.21, 65.57 (J=22.0 Hz), 76.82, 125.88,
127.64, 131.09, 138.85. Anal. calcd for C13H19DO2: C, 74.60; H+D, 10.11. Found: C, 74.31; H+D,
10.00. The ee of the diol 7b was increased in the manner as described for 7a, furnishing 7b (727
mg, 43%, ee 92%).
3.6. [1-2H]-(1R,2S)-2,3,3-Trimethyl-8-phenyl-1,2-octanediol 8a
A solution of the diol 7a (515 mg, 2.46 mmol) and acetone dimethylacetal (0.35 ml, 2.86 mmol) in
dry CH2Cl2 (5 ml) containing p-TsOH (3 mg) was stirred at room temperature for 15 min. Addition
of NaHCO3 (powder, 5 mg), removal of the solvent under reduced pressure and chromatography
of the residue on silica gel (hexane:AcOEt, 10:1) a€orded the acetonide (580 mg, 95%). NaIO4
.
(9.6 g, 45 mmol) and RuCl3 nH2O (44±45% Ru, 26 mg, 0.12 mmol) was added to a solution of
the acetonide (561 mg, 2.25 mmol) in CCl4:CH3CN (1:1 v/v, 11.2 ml) and phosphate bu€er
(NaH2PO4:Na2HPO4=1:1, 11 ml) was added and the mixture was stirred at room temperature
for 36 h. Extractive (CH2Cl2) workup and chromatography on silica gel (hexane:AcOEt, 2:1) gave
a crude acid. This was treated with excess of ethereal CH2N2 and the resulting methyl ester was
reduced with LiAlH4 in dry ether in the usual manner. Extractive (AcOEt) workup and chroma-
tography on silica gel (hexane:AcOEt, 2:1) a€orded the alcohol (270 mg, 59%) as a colorless oil.
The alcohol (255 mg, 1.25 mmol) in pyridine (0.90 ml) was treated with TsCl (310 mg, 1.63
mmol) at 0ꢀC for 1.5 h. Extractive (ether) workup and chromatography on silica gel (hexane:AcOEt,
7:1) a€orded the tosylate (380 mg, 85%) as a colorless oil.

1866
N. Hara et al. / Tetrahedron: Asymmetry 11 (2000) 1859±1868
To a solution of 3-phenylpropylmagnesium bromide, prepared from 3-phenylpropylbromide
(1.06 ml, 6.97 mmol), magnesium (170 mg, 6.99 mmol) and THF (5.0 ml), was added Li2CuCl4
(0.1 M THF solution; 1.0 ml, 0.10 mmol) and the tosylate (359 mg, 1.00 mmol) in THF (2.0 ml)
at ^78ꢀC. After removal of the cooling bath, the mixture was stirred for 4 h. Extractive (ether)
workup and chromatography (benzene) on silica gel gave a coupling product (216 mg, 71%) as a
colorless oil.
A solution of the coupling product (201 mg, 0.66 mmol) in MeOH (50 ml) and 2N HCl (1.0 ml,
2.0 mmol) was heated at 55ꢀC for 1.5 h. Removal of the solvent under reduced pressure and
chromatography of the residue on silica gel a€orded 8a (108 mg, 62%) as a white solid, mp 63±
65ꢀC. IR ꢁmax: 3630, 3470, 3010, 2970, 2940, 2870, 1460, 1370, 1030. 1H NMR ꢀ: 0.87, 0.88 (each
3H, s, 3-CH3Â2), 1.17 (3H, s, 2-CH3), 1.28 (6H, m, 4,5,6-H2), 1.63 (2H, m, 7-H2), 2.08, 2.39 (each
1H, br, OHÂ2), 2.60 (2H, t, J=7.6 Hz, 8-H2), 3.38 (1H, brs, 1-H), 7.16±7.30 (5H, m, Ph). 13C NMR
ꢀ: 19.57, 21.37, 24.17, 30.35, 31.50, 35.96, 36.75, 38.94, 65.46 (J=22.0 Hz), 77.00, 125.55, 128.18,
128.36, 142.73. Anal. calcd for C17H27DO2: C, 76.93; H+D, 11.01. Found: C, 76.73; H+D, 11.16.
3.7. [1-2H]-(1S,2S)-2,3,3-Trimethyl-8-phenyl-1,2-octanediol 8b
Diol 7b (707 mg, 3.38 mmol) was converted, in the same manner as described for 7a, into 8b
(125 mg, 14%), mp 71±74ꢀC. IR ꢁmax: 3630, 3460, 3010, 2970, 2940, 2870, 1460, 1380, 1030, 940.
1H NMR ꢀ: 0.87, 0.88 (each 3H, s, 3-CH3Â2), 1.17 (3H, s, 2-CH3), 1.28 (6H, m, 4, 5, 6-H2), 1.63
(2H, m, 7-H2), 1.90, 2.10 (each 1H, br, OHÂ2), 2.60 (2H, t, J=7.9 Hz, 8-H2), 3.67 (1H, brs, 1-H),
7.15±7.30 (5H, m, Ph). 13C NMR (67.5 MHz, CDCl3) ꢀ: 19.59, 21.37, 24.19, 30.37, 31.52, 35.98,
36.75, 38.94, 65.48 (J=21.4 Hz), 77.05, 125.55, 128.19, 128.37, 142.75. Anal. calcd for
C17H27DO2: C, 76.93; H+D, 11.01. Found: C, 76.89; H+D, 11.01.
3.8. [1-2H]-(1S,2S)-1,2-Epoxy-2,3,3-trimethyl-8-phenyl-1-octane 3a
MsCl (35 ml, 0.45 mmol) was added to a solution of the diol 8a (100 mg, 0.38 mmol) in pyridine
(0.8 ml) at 0ꢀC and the mixture was stirred for 1 h at the same temperature. Extractive (ether)
workup gave a crude mesylate which was treated with sat. K2CO3 in methanol (1 ml) at room
temperature for 5 min. Extractive (ether) workup and chromatography on silica gel (hex-
ane:AcOEt, 10:1) a€orded the epoxide 3a (78 mg, 84%) as a colorless oil. IR ꢁmax: 3020, 2970,
1
2940, 2860, 1460, 1380. H NMR ꢀ: 0.83, 0.90 (each 3H, s, 3-CH3Â2), 1.26 (3H, s, 2-CH3), 1.32
(6H, m, 4,5,6-H2), 1.64 (2H, m, 7-H2), 2.60 (2H, t, J=7.9 Hz, 8-H2), 2.72 (1H, s, 1-H), 7.14±7.30
2
(5H, m, Ph). H NMR ꢀ: 2.46 (pro-R-2H). 13C NMR ꢀ: 18.46, 23.17, 23.42, 23.99, 30.14, 31.39,
35.96, 39.73, 50.80 (J=25.7 Hz), 61.08, 125.55, 128.18, 128.36, 142.73. HR-FABMS calcd for
C17H26DO: 248.2146 (MH+). Found: 248.2108.
3.9. (1R,2S)-[1-2H]-1,2-Epoxy-2,3,3-trimethyl-8-phenyl-1-octane 3b
The diol 8b (120 mg, 0.45 mmol) was converted, in the same manner as described for 8a, into
the epoxide 3b (103 mg, 92%) as a colorless oil. IR ꢁmax: 3020, 2970, 2940, 2860, 1460, 1380. 1H NMR
ꢀ: 0.83, 0.90 (each 3H, s, 3-CH3Â2), 1.26 (3H, s, 2-CH3), 1.31 (6H, m, 4,5,6-H2), 1.63 (2H, m, 7-H2),
2
2.37 (1H, s, 1-H), 2.60 (2H, t, 8-H2, J=8.0 Hz), 7.13±7.30 (5H, m, Ph). H NMR ꢀ: 2.80 (pro-
S-2H). 13C NMR ꢀ: 18.53, 23.20, 23.45, 24.03, 30.17, 31.43, 35.99, 39.77, 50.87 (J=26.9 Hz), 61.12,
125.59, 128.21, 128.39, 142.77. HR-FABMS calcd for C17H26DO: 248.2146 (MH+). Found: 248.2105.

N. Hara et al. / Tetrahedron: Asymmetry 11 (2000) 1859±1868
1867
.
3.10. Treatment of the epoxides 3a and 3b with BF3 Et2O
.
BF3 Et2O (69 ml, 0.55 mmol) was added to a solution of the epoxide 3a (68 mg, 0.28 mmol) in
dry CH2Cl2 (6.8 ml) at room temperature and the mixture was stirred for 3 min. Ether and sat.
NaHCO3 were added at once and extractive (ether) workup and chromatography on silica gel
(hexane:benzene, 10:1) gave a fraction (30 mg) containing aldehyde 9. Half of this was reduced
with LiAlD4 in THF in the usual manner. Chromatography of the crude product on silica gel
(hexane:AcOEt, 10:1) a€orded an alcohol (7 mg). The other half was similarly reduced with
LiAlD4 to give another alcohol (7 mg). Treatment of the two alcohols with (S)-MTPA chloride in
pyridine followed by puri®cation by p-tlc gave the corresponding (R)-MTPA esters 10a±d (7 mg)
1
and 11a±d (7 mg). Compound 10a±d: H NMR ꢀ: 0.78±0.92 (9H, m, CH3), 1.05±1.42, 1.53±1.75
(10.55H, m, CH2 and CH), 2.60 (2H, t, J=8.1 Hz, PhCH2), 3.56 (3H, s, OCH3), 7.23±7.62 (10H,
m, Ph). The signals of the oxymethine and oxymethylene region are illustrated in Fig. 1. HR-FABMS
1
calcd for C27H34D2O3F3: 467.2742 (MH+). Found: 466.2739. Compound 11a±d: H NMR ꢀ:
0.79±0.92 (9H, m, CH3), 1.05±1.40, 1.50±1.76 (10.55H, m, CH2 and CH), 2.60 (2H, t, J =8.1 Hz,
PhCH2), 3.56 (3H, s, OCH3), 7.13±7.58 (10H, m, Ph). The signals of the oxymethine and oxy-
methylene region are illustrated in Fig. 1. HR-FABMS calcd for C27H35DO3F3: 466.2679 (MH+).
Found: 466.2673.
.
The epoxide 3b (93 mg, 0.38 mmol) was treated with BF3 Et2O in the same manner as described
for 3a and the aldehyde was converted into 10a±d (9 mg) and 11a±d (11 mg). Compound 10a±d:
1H NMR ꢀ: 0.79±0.89 (9H, m, CH3), 1.04±1.38, 1.53±1.75 (10.72H, m, CH2 and CH), 2.60 (2H, t,
J=8.1 Hz, PhCH2), 3.55 (3H, s, OCH3), 3.92±4.05 (0.06H, m, CH2 and CDH), 4.38±4.56 (0.01H,
m, CH2 and CDH), 7.12±7.56 (10H, m, Ph). HR-FABMS calcd for C27H34D2O3F3: 467.2742
1
(MH+). Found: 466.2744. Compound 11a±d: H NMR ꢀ: 0.80±0.88 (9H, m, CH3), 1.05±1.38,
1.55±1.75 (10.72H, m, CH2 and CH), 2.60 (2H, t, J=8.1 Hz, PhCH2), 3.56 (3H, s, OCH3), 3.92±
4.10 (0.10H, m, CH2 and CDH), 4.38±4.56 (0.21H, m, CH2 and CDH), 7.12±7.56 (10H, m, Ph).
HR-FABMS calcd for C27H35DO3F3: 466.2679 (MH+). Found: 466.2651.
References
1. For Lewis acid-mediated epoxide rearrangements, see: Maruoka, K.; Hasegawa, M.; Yamamoto, H.; Suzuki, K.;
Shimazaki, M.; Tsuchihashi, G. J. Am. Chem. Soc. 1986, 108, 3827±3829; Suzuki, K.; Miyazawa, M.; Tsuchihashi,
G. Tetrahedron Lett. 1987, 28, 3515±3518; Maruoka, K.; Ooi, T.; Yamamoto, H. J. Am. Chem. Soc. 1989, 111,
6431±6432; Maruoka, K.; Nagahara, S.; Ooi, T.; Yamamoto, H. Tetrahedron Lett. 1989, 30, 5607±5610; Maruoka,
K.; Ooi, T.; Nagahara, S.; Yamamoto, H. Tetrahedron 1991, 47, 6983±6998.
2. Kirk, D. N.; Hartshorn, M. P. Steroid Reaction Mechanisms; Elsevier Publishing Co.: Amsterdam, 1968; pp. 353±
372.
3. Coxon, J. M.; Lim, C. Aust. J. Chem. 1977, 30, 1137±1143; Bach, R. D.; Lucast, D. H. J. Chem. Soc., Chem.
Commun. 1979, 593±594.
4. Blackett, B. N.; Coxon, J. M.; Hartshorn, M. P.; Richards, K. E. J. Am. Chem. Soc. 1970, 92, 2574±2575.
5. Maruoka, K.; Nagahara, S.; Ooi, T.; Yamamoto, H. Tetrahedron 1992, 48, 3303±3312.
6. Fujimoto, Y.; Kanzawa, Y.; Ikuina, Y.; Kakinuma, K.; Ikekawa, N. J. Chem. Soc., Chem. Commun. 1989, 1107±
1109; Fujimoto, Y.; Ikuina, Y.; Kanzawa, Y.; Nagakari, M.; Kakinuma, K.; Ikekawa, N. Heterocycles 1990, 30,
275±278.
.
7. For example, treatment of (S)-1 with 0.2 equiv. of BF3 Et2O in CH2Cl2 a€orded 2-methyl-7-phenylpentanal with
(2R):(2S)=52:48.
8. Negishi, E.; Van Horn, D. E.; Yoshida, T. J. Am. Chem. Soc. 1985, 107, 6639±6647.

1868
N. Hara et al. / Tetrahedron: Asymmetry 11 (2000) 1859±1868
9. Sharpless, K. B.; Amberg, W.; Bennani, Y. L.; Crispino, G. A.; Hartung, J.; Jeong, K.-S.; Kwong, H.-L.;
Moriyama, K.; Wang, Z.-M.; Xu, D.; Zhang, X.-L. J. Org. Chem. 1992, 57, 2768±2771.
.
10. BF3 Et2O treatment of the epoxide which was prepared from 7a in two steps (mesylation and K2CO3 treatment)
a€orded a mixture of products, probably formed by benzyl group participation. The ee of compound 8a could not
be increased by recrystallization via the phthalic half-ester derivative. The detour route was selected due to these
reasons.
11. The epoxide was consumed under this reaction condition. Attempts to increase the yield of the aldehyde were not
successful. For example, when the reaction was carried out at a lower temperature (0ꢀC), an increased amount of
.
¯uorohydrin was produced. Use of a reduced amount of BF3 Et2O (0.2 equiv.) led to a similar result.
12. The chemical shifts were assigned by comparing the 1H NMR data with those of stereochemically de®ned
synthetic compounds; (2R)-5-phenyl-2,3,3-trimethylpentanol (R)-MTPA ester showed oxymethylene proton
signals at ꢀ 4.62 (dd, J=11.0, 3.7 Hz), 4.07 (dd, J=11.0, 8.7 Hz), while the (2S)-counterpart showed the
1
corresponding proton signals at ꢀ 4.54 (dd, J=11.0, 3.7 Hz), 4.14 (dd, J=11.0, 8.7 Hz). The H NMR data for
the (R)-MTPA ester of non-deuterated racemic 10: (2R)-isomer ꢀ: 4.52 (dd, J=10.7, 3.7 Hz), 3.97 (dd, J=10.7, 9.2
Hz); (2S)-isomer ꢀ: 4.54 (dd, J=10.7, 3.7 Hz), 4.05 (dd, J=10.7, 9.2 Hz).
13. The percentages of 10c (x=22) and 10d (y=33) were obtained by solving the following two simultaneous
equations: x+y=55 and 42Â(31Â2+x)=58Â(14Â2+y).
14. The presence of the antipode in the epoxide substrate was not considered.
15. (2R)-5-Phenyl-2,3,3-trimethylpentanal was subjected to the reaction conditions.
16. The deuterium isotope parameter z was determined to satisfy 31z:14z:22:33=50:22:11z:17z.
17. White, J. C.; Cave, R. J.; Davidson, E. R. J. Am. Chem. Soc. 1988, 110, 6308±6314; Nakamura, K.; Osamura, Y.
J. Phys. Org. Chem. 1990, 3, 737±745.