Enantioresolution by the chiral phthalic acid method: Absolute configurations of (2-methylphenyl)phenylmethanol and related compounds

Article abstract of DOI:10.1016/S0957-4166(99)00231-1

Racemic (2-hydroxymethylphenyl)phenylmethanol 5 was enantioresolved by the chiral phthalic acid method using acid-amide 2, and the absolute configurations of (-)-(2-methylphenyl)phenylmethanol 3 and related o-substituted diphenylmethanol derivatives were unambiguously determined by X-ray crystallography and chemical correlation.

Full text of DOI:10.1016/S0957-4166(99)00231-1

Pergamon
Tetrahedron: Asymmetry 10 (1999) 2075–2078
Enantioresolution by the chiral phthalic acid method:
absolute configurations of (2-methylphenyl)phenylmethanol and
related compounds
Masataka Watanabe,a Shunsuke Kuwahara,a Nobuyuki Harada,a,∗ Masatoshi Koizumi b and
Takeshi Ohkuma b
aInstitute for Chemical Reaction Science, Tohoku University, 2-1-1 Katahira, Aoba, Sendai 980-8577, Japan
bDepartment of Chemistry, Graduate School of Science, Nagoya University, Chikusa, Nagoya 464-8602, Japan
Received 13 May 1999; accepted 2 June 1999
Abstract
Racemic (2-hydroxymethylphenyl)phenylmethanol 5 was enantioresolved by the chiral phthalic acid method
using acid-amide 2, and the absolute configurations of (−)-(2-methylphenyl)phenylmethanol 3 and related o-
substituted diphenylmethanol derivatives were unambiguously determined by X-ray crystallography and chemical
correlation. © 1999 Elsevier Science Ltd. All rights reserved.
Recently, much attention has been focused on the asymmetric synthesis of various chiral alcohols
using chiral catalysts.1,2 These catalytic asymmetric syntheses are very powerful for preparing chiral
compounds with high enantiomeric excess. However, there are some weak points: (i) although the
absolute configuration of products may be estimated by the reaction mechanism and/or by comparison
with similar cases, in which the absolute configuration of products is known, such assignments of
absolute configurations have remained ambiguous; (ii) the products obtained by asymmetric syntheses
are not always enantiopure. As a method for solving these problems, classical enantioresolution using
new chiral auxiliaries is of much practical use. For example, we have recently developed new chiral
auxiliaries, chiral phthalic acid amides (1S,2R,4R)-(−)-1 and (1S,2R,4R)-(−)-2, which were very useful
for the enantioresolution of various alcohols by HPLC on silica gel and also for the X-ray crystallographic
determination of their absolute configurations (Fig. 1).3,4 One group of the authors has intensively studied
the catalytic asymmetric reduction of various ketones to chiral alcohols using ruthenium(II) complexes,1
and has obtained various chiral substituted diphenylmethanols with high enantiopurity.5 In some cases,
however, the absolute configuration of products has remained undetermined or only estimated in an
empirical manner. Furthermore, the [α]D values of some products are very small. Such facts bring much
ambiguity to the stereochemical studies of asymmetric synthesis. (2-Methylphenyl)phenylmethanol 3
∗
Corresponding author. E-mail: n-harada@icrs.tohoku.ac.jp
0957-4166/99/$ - see front matter © 1999 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(99)00231-1

2076
M. Watanabe et al. / Tetrahedron: Asymmetry 10 (1999) 2075–2078
Figure 1. Chiral phthalic acid amides and o-substituted diphenylmethanols
is one such compound (Fig. 1). The absolute configuration of alcohol 3 had been in much confusion;
Cervinka et al.6 reported it to be (R)-(−) based upon asymmetric reductions, but Seebach et al.2 assigned
it as (R)-(+) by catalytic asymmetric syntheses. In this paper, we report the preparation of enantiopure
alcohol 3 and unambiguous determination of its absolute configuration by the chiral phthalic acid method.
To determine the absolute configuration, we first applied the chiral phthalic acid method to 3 itself;
however, the diastereomeric mixture of esters formed could not be separated by HPLC on silica gel.
Next the method was applied to (4-bromophenyl)-20-methylphenylmethanol 4, from which alcohol
3 is obtainable by reduction. In fact, we have previously succeeded in the enantioresolution of (4-
bromophenyl)-40-methylphenylmethanol by this method, and the enantiopure bromo-alcohol obtained
was converted into (4-methylphenyl)phenylmethanol.4f However, unfortunately the diastereomeric esters
formed from (−)-2 and (ꢀ)-4 were also inseparable by HPLC on silica gel. We therefore took the
alternative route to resolve (2-hydroxymethylphenyl)phenylmethanol 5, which is convertible into alcohol
3.
Commercially available methyl 2-benzoylbenzoate was reduced with LiAlH4 in diethyl ether to
give diol 5, which was then condensed with chiral dichlorophthalic acid amides (1S,2R,4R)-(−)-2;
a mixture of diol (ꢀ)-5, (−)-2 (1.0 equiv.), 1,3-dicyclohexylcarbodiimide (DCC, 1.2 equiv.), and 4-
dimethylaminopyridine (DMAP, 0.2 equiv.) in CH2Cl2 was stirred at room temperature for 2 days
(Scheme 1). The diastereomeric mixture of esters formed was separated by HPLC on silica gel (hex-
ane:EtOAc, 4:1): separation factor α=1.14; resolution factor Rs=1.04. The first-eluted ester (−)-6a (30%,
23
20
[α]D −8.62 (c 2.52, CHCl3)) and the second one (−)-6b (32%, [α]D −98.4 (c 2.36, CHCl3)) were
obtained. The less polar ester (−)-6a was recrystallized from hexane/EtOAc giving large prisms; however,
these were not single crystals when checked by X-ray analysis. On the other hand, single crystals of (−)-
6b were obtained by recrystallization from EtOH, and subjected to X-ray crystallography: orthorhombic;
space group P212121. The structure was solved as usual: R=0.0324 and Rw=0.0415, while R=0.0470 and
Rw=0.0627 for the mirror image structure. The absolute configuration of ester (−)-6b was thus determined
to be S by the heavy atom effect as well as by the internal reference method using the known absolute
configuration of the camphor part of auxiliary (−)-2 (Fig. 2). The R absolute configuration was therefore
assigned to ester (−)-6a.
The secondary hydroxyl group of ester (R)-(−)-6a was blocked as tert-butyldiphenylsilyl (TBDPS)
21
ether as shown in Scheme 1. TBDPS ether (R)-(−)-7a (97%, [α]D −60.6 (c 1.36, CHCl3)). Ester (R)-
21
(−)-7a was hydrolyzed with K2CO3 in MeOH to yield alcohol (R)-(−)-8 (81%, [α]D −1.31 (c 1.87,
CHCl3)), which was then converted to mesylate (R)-9 (97%). Reduction of the mesylate group with
23
LiAlH4 afforded compound (R)-(−)-10 (74%, [α]D −27.3 (c 2.51, CHCl3)), from which the desired
enantiopure alcohol (R)-(−)-3 was finally obtained by treatment with tetrabutylammonium fluoride
23
24
(TBAF) in THF: (R)-(−)-3 (90%, [α]D −7.64 (c 1.51, CHCl3); [α]D −12.9 (c 0.570, EtOH)). The
absolute configuration of (−)-3 was thus unambiguously determined as R.7 Reduction of ester (R)-(−)-6a
24
with LiAlH4 in diethyl ether yielded diol (R)-(+)-5 (87%, [α]D +56.9 (c 1.39, CHCl3)).
The CD spectrum of alcohol (R)-(−)-3 exhibits characteristic positive Cotton effects with vibrational

M. Watanabe et al. / Tetrahedron: Asymmetry 10 (1999) 2075–2078
2077
Scheme 1. (a) DCC, DMAP/CH2Cl2, rt; (b) TBDPSCl, imidazole/DMF, rt; (c) K2CO3/MeOH, rt; (d) MsCl, Et3N/CH2Cl2, 0°C;
(e) LiAlH4/Et2O, 0°C; (f) TBAF/THF, rt
Figure 2. ORTEP drawing of ester (S)-(−)-6b. The atoms are drawn as 50% probability ellipsoids
structure in the 1Lb transition region at 240–280 nm (Fig. 3). In addition, it also shows a positive Cotton
effect at 225.4 nm. Therefore, the enantiomer 3 is fully specified as (R)-[CD(+)225.4]-(−)-3.8
Acknowledgements
This work was supported in part by grants from the Ministry of Education, Science, Sports, and
Culture, Japan (Scientific Research (A) No. 07408032, (B) No. 10554035, Priority Areas (A) No.
10146205, (B) No. 10208202, and International Joint No. 10045022 to N.H.; Exploratory Research No.
09874129 to M.W.) and the Ciba-Geigy Foundation (Japan) for the Promotion of Science (to N.H.).

2078
M. Watanabe et al. / Tetrahedron: Asymmetry 10 (1999) 2075–2078
Figure 3. CD and UV spectra of (R)-[CD(+)225.4]-(−)-(2-methylphenyl)phenylmethanol 3 in EtOH
References
1. Ohkuma, T.; Ooka, H.; Ikariya, T.; Noyori, R. J. Am. Chem. Soc. 1995, 117, 10417. Ohkuma, T.; Ooka, H.; Yamakawa,
M.; Ikariya, T.; Noyori, R. J. Org. Chem. 1996, 61, 4872. Fujii, A.; Hashiguchi, S.; Uematsu, N.; Ikariya, T.; Noyori, R. J.
Am. Chem. Soc. 1996, 118, 2521.
2. Seebach, D.; Beck, A. K.; Roggo, S.; Wonnacott, A. Chem. Ber. 1985, 118, 3673.
3. (a) Harada, N.; Soutome, T.; Nehira, T.; Uda, H.; Oi, S.; Okamura, A.; Miyano, S. J. Am. Chem. Soc. 1993, 115, 7547. (b)
Harada, N.; Nehira, T.; Soutome, T.; Hiyoshi, H.; Kido, F. Enantiomer 1996, 1, 35. (c) See also: Toyota, S.; Yasutomi, A.;
Oki, M. Tetrahedron Lett. 1995, 36, 6297.
4. (a) Harada, N.; Koumura, N.; Robillard, M. Enantiomer 1997, 2, 303. (b) Harada, N.; Hiyoshi, N.; Vassilev, V. P.; Hayashi,
T. Chirality 1997, 9, 623. (c) Harada, N.; Vassilev, V. P.; Hiyoshi, N. Enantiomer 1997, 2, 123. (d) Nemoto, H.; Miyata,
J.; Fukumoto, K.; Ehara, H.; Harada, N.; Uekusa, H.; Ohashi, Y. Enantiomer 1997, 2, 127. (e) Harada, N.; Koumura, N.;
Feringa, B. L. J. Am. Chem. Soc. 1997, 119, 7256. (f) Harada, N.; Fujita, K.; Watanabe, M. Enantiomer 1997, 2, 359. (g)
Harada, N.; Fujita, K.; Watanabe, M. Enantiomer 1998, 3, 64. (h) Kuwahara, S.; Fujita, K.; Watanabe, M.; Harada, N.;
Ishida, T. Enantiomer, 1999, 4, in press.
5. Ohkuma, T.; Koizumi, M.; Ikehira, H.; Yokozawa, T.; Ikariya, T.; Noyori, R. in preparation.
6. Cervinka, O.; Suchan, V.; Masak, B. Collect. Czech. Chem. Commun. 1965, 30, 1693. Cervinka, O.; Belovsky, O.;
Fabryova, A.; Dudek, V.; Grohman, K. Collect. Czech. Chem. Commun. 1967, 32, 2618.
20
7. The [α]D values of (R)-3 were previously measured in ethanol, [α]D −2.50 (c 4.00) (no % ee data),6 and in benzene,
25
[α]D +0.67 (c 1.27) (enantiopure),2 respectively. Since the reported value in benzene is very small, we recommend the
use of chloroform or ethanol instead of benzene.
8. When the [α]D value is small, its measurement necessitates a large amount of sample, and it is hard to obtain a reliable
value. In such a case, the designation of enantiomers by CD data is pertinent, because CD measurement generally needs
a much smaller amount of sample than [α]D measurement: Harada, N.; Iwabuchi, J.; Yokota, Y.; Uda, H.; Okamoto, Y.;
Yuki, H.; Kawada, Y. J. Chem. Soc., Perkin Trans. 1 1985, 1845. Harada, N. Enantiomer 1996, 1, 81.