Stereoselective Synthesis of Optically Active 2-Hydroxymandelic Acids and Esters via Friedel-Crafts Coordinated Reaction: Crystal Structure of Chiral Dichloro[2-(1-oxido-1-menthoxycarbonylmethyl)-4-methoxyphenoxido-O,O,O] titanium?

Article abstract of DOI:10.1021/jo982170o

Highly diastereoselective hydroxyalkylation of phenols with chiral glyoxylates was performed. Hydrolysis or reduction of the esters 3 afforded 2-hydroxymandelic acids 6 and diols 5 in high yields and enantiomeric pure form. Mechanistic evidence of chelati

Full text of DOI:10.1021/jo982170o

5004
J . Org. Chem. 1999, 64, 5004-5009
Articles
Ster eoselective Syn th esis of Op tica lly Active 2-Hyd r oxym a n d elic
Acid s a n d Ester s via F r ied el-Cr a fts Coor d in a ted Rea ction :
Cr ysta l Str u ctu r e of Ch ir a l Dich lor o[2-(1-oxid o-1-m en th oxy-
ca r bon ylm eth yl)-4-m eth oxyp h en oxid o-O,O,O]tita n iu m †
Franca Bigi,*,‡ Gabriele Bocelli,§ Raimondo Maggi,‡ and Giovanni Sartori‡
Dipartimento di Chimica Organica e Industriale dell’Universita`, Parco Area delle Scienze 17A, I-43100
Parma, Italy, and Centro di Studio per la Strutturistica Diffrattometrica del C.N.R., Parco Area delle
Scienze 17A, I-43100 Parma, Italy
Received October 28, 1998
Highly diastereoselective hydroxyalkylation of phenols with chiral glyoxylates was performed.
Hydrolysis or reduction of the esters 3 afforded 2-hydroxymandelic acids 6 and diols 5 in high
yields and enantiomeric pure form. Mechanistic evidence of chelation-controlled reaction comes
from the crystal structure determination of titanium phenoxide complex 8.
In tr od u ction
lacked general utility for preparation of 2-hydroxyman-
delic derivatives.4
Our results1 on selective functionalization of phenolic
systems promoted by the metal template complex of
substrate and reagent with Lewis acid prompted us to
investigate the possibility of stereocontrolled electrophilic
aromatic substitution.
Resu lts a n d Discu ssion
The diastereoselective hydroxyalkylation of phenols 1
with chiral glyoxylates 2 leading to optically active
2-hydroxymandelic esters 3 and 4,5 and the removal of
the chiral auxiliary, is described. New findings that relate
to the mechanism of the ortho-coordinated alkylation of
phenols as well as to the mechanism of diastereoselective
reactions, like ene reactions,4c,6 involving chiral glyoxy-
lates in the presence of Lewis acids, are reported.
We have previously performed ortho hydroxyalkylation
of phenols with good to excellent levels of stereochemical
control using a chiral ketoester1c and chiral1d,e as well as
achiral1b aldehydes.
This paper addresses regio- and stereocontrol in the
synthesis of optically active 2-hydroxymandelic acids.
Mandelic acid and some derivatives are used as drugs2
and are usually prepared as racemic mixtures by mul-
tistep syntheses.3a 2-Hydroxymandelic acid has recently
received attention.3b,c The stereoselective syntheses re-
ported give low values of diastereomeric excess and/or
We first examined the model reaction of 3-tert-bu-
tylphenol (1a ) with (-)-menthyl glyoxylate (2i). The
different Lewis acids employed (Et2AlCl, EtAlCl2, Et-
MgBr, InCl3, BCl3, SnCl4, and TiCl4) affected both the
yield and the diastereoselectivity; 2S diastereoisomer 3i
was favored. Titanium proved to be the most efficient
metal promoter of the reaction for both stereocontrol (de
50%) and yield (53%), operating at -30 °C for 6 h.
Introduction of different chiral alcohols as titanium
ligands ((-) and (+)-menthol, (-)-8-phenylmenthol, as
well as (+)-diisopropyltartrate7) failed to significantly
improve the diastereoselectivity.
† Dedicated to the memory of Professor Giuseppe Casnati on the
6th anniversary of his death.
‡ Dipartimento di Chimica Organica e Industriale.
§ Centro di Studio per la Strutturistica Diffrattometrica.
(1) (a) Casnati, G.; Casiraghi, G.; Pochini, A.; Sartori, G.; Ungaro,
R. Pure Appl. Chem. 1983, 55, 1677. (b) Bigi, F.; Casiraghi, G.; Casnati,
G.; Sartori, G.; Gasparri Fava, G.; Ferrari Belicchi, M. J . Org. Chem.
1985, 50, 5018. (c) Casiraghi, G.; Bigi, F.; Casnati, G.; Sartori, G.;
Soncini, P.; Gasparri Fava, G.; Ferrari Belicchi, M. J . Org. Chem. 1988,
53, 1779. (d) Bigi, F.; Casnati, G.; Sartori, G.; Araldi, G.; Bocelli, G.
Tetrahedron Lett. 1989, 30, 1121. (e) Bigi, F.; Casnati, G.; Sartori, G.;
Araldi, G. Gazz. Chim. Ital. 1990, 120, 413.
Since the stereochemical outcome was unaffected by
reaction time, whereas the yield was increased, the
reactions were carried out for 20 h obtaining isomers 3i
and 4i in good yields and about 50% de (Table 1, runs
1-4). The major isomers 3ia -d were easily isolated by
crystallization from hexane/dichloromethane.
(2) The Merck Index, 11 ed.; Budavari, S., Ed., Merk & Co.: Rahway,
NJ , 1989; p 898.
(3) (a) Turner, A. B. Rodd’s Chemistry of Carbon Compounds, II ed.;
Elsevier Scientific Publ. Co.: Amsterdam, 1974; Vol. III, Part E,
Chapter 18, p 105. (b) Hoefnagel A. J .; Peters J . A.; van Bekkum, H.
Recl. Trav. Chim. Pays-Bas 1988, 107, 242. (c) Hoefnagel A. J .; Peters
J . A.; van Bekkum, H. Recl. Trav. Chim. Pays-Bas 1996, 115, 353.
(4) (a) Vavon, G.; Antonini, A. C. R. Acad. Sci. 1950, 230, 1870. (b)
Horeau, A.; Kagan, H. B.; Vigneron J . P. Bull. Soc. Chim. Fr. 1968,
3795. (c) Whitesell, J . K.; Deyo, D.; Battacharya, A. J . Chem. Soc.,
Chem. Commun. 1983, 802. (d) Tanaka, K.; Mori, A.; Inoue, S. J . Org.
Chem. 1990, 55, 181. (e) Seebach, D.; J aeschke, G.; Gottwald, K.;
Matsuda, K.; Formisano, R.; Chaplin, D. A.; Breuning, M.; Bringmann,
G. Tetrahedron 1997, 53, 7539. (f) Carpentier, J .-F.; Mortreux, A.
Tetrahedron: Asymmetry 1997, 8, 1083. (g) Kirschning, A.; Draeger,
G.; J ung, A. Angew. Chem., Int. Ed. Engl. 1997, 36, 253.
(5) Bigi, F.; Casnati, G.; Sartori, G.; Dalprato, C.; Bortolini, R.
Tetrahedron: Asymmetry 1990, 1, 861.
(6) Whitesell, J . K.; Bhattacharya, A.; Buchanan, C. M.; Chen, H.
H.; Deyo, D.; J amed, D.; Liu, C. L.; Minton, M. A. Tetrahedron 1986,
42, 2993.
(7) (a) Seebach, D.; Beck, A. K.; Imwinkelried, R.; Roggo, S.;
Wonnacott, A. Helv. Chim. Acta 1987, 70, 954. (b) Narasaka, K.;
Iwasawa, N.; Inoue, M.; Yamada, T.; Nakashima, M.; Sugimori, J . J .
Am. Chem. Soc. 1989, 111, 5340.
10.1021/jo982170o CCC: $18.00 © 1999 American Chemical Society
Published on Web 06/17/1999

Synthesis of Optically Active 2-Hydroxymandelic Acids
J . Org. Chem., Vol. 64, No. 14, 1999 5005
Ta ble 1. Syn th esis of 2-Hyd r oxym a n d elic Ester s 3 a n d 4a
run
phenol
R
aldehyde
R*
major compd
T (°C)
time (h)
% total yield
3:4c
1
2
3
4
5
6
7
1a
1b
1c
1d
1a
1b
1e
3-But
4-OMe
H
2i
2i
2i
2i
2j
2j
2j
(-)-menthyl
3ia
3ib
3ic
3id
3ja
3jb
3je
-30
-30
-30
-30
20
20
20
20
20
6
82b
80b
77b
78b
95
75:25
75:25
74:26
76:24
97:3
(-)-menthyl
(-)-menthyl
(-)-menthyl
(-)-8-Ph-menthyl
(-)-8-Ph-menthyl
(-)-8-Ph-menthyl
4-But
3-But
4-OMe
4-Cl
20
20
6
6
97
62
97:3
97:3
a
b
All reactions were carried out in a ratio of phenol/glyoxylate/Lewis acid ) 1:1:1 in CH2Cl2. The use of aldehyde 2i in the hydrate
form or 0.5 equiv of Lewis acid gave slightly lower yields. c The diastereomeric excesses were determined by HPLC (see Experimental
Section).
Sch em e 1. Rem ova l of th e Ch ir a l Au xilia r y
Diastereoselectivity was significantly improved by the
use of (-)-8-phenylmenthyl glyoxylate 2j6,8,9 (Table 1,
runs 5-7).
Titanium tetrachloride is the most efficient promoter
for this reaction, of the Lewis acids examined. Indeed,
the reaction of 3-tert-butylphenol (3a ) and glyoxylate 2j
in the presence of TiCl4 gave the product 3ja in es-
sentially quantitative yield and with excellent diastereo-
selectivity even at room temperature.10 The temperature
decrease (down to -60 °C) affected the yield more
markedly than the stereoselectivity. Thus, the reaction
was extended to activated and weakly deactivated phe-
nols at room temperature, providing 2-hydroxymandelic
esters 3j with high diastereoselection (94%) and good
yields (62-97%).
As far as absolute configuration of the new chiral
carbon is concerned, the assignment of S configuration
was made at C-2 of the major diastereoisomers 3ia -d
purified by crystallization, was chosen as an example,
since (S)-2-hydroxymandelic acid is reported to be dex-
trorotatory.15
1
on the basis of H NMR studies.5,11
This attribution is confirmed by the crystal structure
of compound 4ic, which has the R configuration at C-2.12
The S configuration at C-2 of the isomer 3ja was
attributed on the basis of chemical correlation with
compound 3ia . Convergent reduction of (2S)-3ia and 3ja
to the same diol (-)-1-(2-hydroxy-4-tert-butylphenyl)-1,2-
dihydroxyethane (5a ) confirmed the 2S attribution of
3ja .13
Unlike other hydroxy acids, e.g., 4-hydroxymandelic
acid, the isolation of 2-hydroxymandelic acid is difficult,
as recently stated3c and witnessed by conflicting analyti-
cal data.15,16
The hydrolysis was accomplished using LiOH in THF/
H2O,17 which gave (+)-(S)-2-hydroxymandelic acid 6c in
high yield (92%) with concomitant recovery of (-)-
menthol auxiliary. By this route, 3ia afforded the corre-
sponding acid (+)-(S)-6a in 90% yield.
It is noteworthy that reduction with NaBH4/LiCl14
opens the way to the synthesis of optically pure polyhy-
dric alcohols, which represent useful chiral ligands.
Finally, we have removed of the chiral auxiliary by
hydrolysis of the esters 3 and 4 to give the corresponding
acids 6 (Scheme 1). The diastereomer (2S)-3ic, easily
We have checked the enantiomeric purity of the acid
(+)-(S)-6c as well as of the diol 5a via the de’s of their
Mosher derivatives18 (see Experimental Section) and by
HPLC chiral separation.19 The hydrolysis and reduction
procedures do not cause any racemization of the com-
pounds.
1
(8) (a) Corey, E. J .; Ensley, H. E. J . Am. Chem. Soc. 1975, 97, 6908.
(b) Ensley, H. E.; Parnell, L. A.; Corey, E. J . J . Org. Chem. 1978, 43,
1610.
(9) (a) Whitesell, J . K. Acc. Chem. Res. 1985, 18, 280 and references
therein. (b) Whitesell, J . K.; Buchanan, C. M. J . Org. Chem. 1986, 51,
5443. (c) Solladie-Cavallo, A.; Khiar, N. J . Org. Chem. 1990, 55, 4750.
(d) Whitesell, J . K. Chem. Rev. 1992, 92, 953. (e) Bedekar, A. V.;
Koroleva, E. B.; Andersson, P. G. J . Org. Chem. 1997, 62, 2518.
(10) By using BCl3 or Ti(OPri)4 (low de values) we were sure of
identifying the minor isomer.
(11) (a) Adamczeski, M.; Quinoa, E.; Crews, P. J . Org. Chem. 1990,
55, 240. (b) Pehk, T.; Lippmaa, E.; Lopp, M.; Paju, A.; Borer, B. C.;
Taylor, R. J . K. Tetrahedron: Asymmetry 1993, 4, 1527.
(12) Unpublished results from our laboratories.
(13) The stereochemical outcome is in agreement with the preferred
si-face attack of the reagent 2j.
The H NMR spectra of the esters 3 and 4 exhibit an
interesting feature. The benzylic methine signal in (-)-
8-phenylmenthyl esters 3j and 4j is shielded in compari-
son with values observed for the corresponding (-)-
(15) McKenzie, A.; Stewart, P. A. J . Chem. Soc. 1935, 104.
(16) (a) Ladenburg, K.; Folkers, K.; Major, R. T. J . Am. Chem. Soc.
1936, 58, 1292. (b) Howe, R.; Rao, B. S.; Heyneker, H. J . Chem. Soc.
C 1967, 2510. (c) Sterk, H.; Kappe, T.; Ziegler, E. Monatsh. Chem. 1968,
2223.
(17) We modified the conditions reported in: Evans, D. A.; Ellman,
J . A.; Dorow, R. L. Tetrahedron Lett. 1987, 28, 1123.
(18) (a) Dale, J . A.; Dull, D. L.; Mosher, H. S. J . Am. Chem. Soc.
1969, 34, 2543. (b) Dale, J . A.; Mosher H. S. Ibid. 1973, 95, 512.
(19) Galaverna, G.; Panto`, F.; Dossena, A.; Marchelli, R.; Bigi, F.
Chirality 1995, 7, 331.
(14) (a) Hamada, Y.; Shioiri, T. Tetrahedron Lett. 1982, 23, 1193;
(b) Chem. Pharm. Bull. 1982, 30(5), 1921.

5006 J . Org. Chem., Vol. 64, No. 14, 1999
Bigi et al.
Sch em e 2. Rea ction Mech a n ism
menthyl derivatives 3i and 4i. In particular, significant
upfield shifts are observed for the (2S)-3j isomers, being
∆δ ∼1.1 ppm. These strong shieldings suggest a proxim-
ity to the phenyl group, as has been observed for similar
compounds.20 This is probably due to π-π interaction
between the phenyl moiety and the ester.21
Concerning the mechanism, on the basis of a large
number of synthetic results,1 NMR study22 and more
recently X-ray structure of a complex between phenoxy-
magnesium bromide and p-isopropylbenzaldehyde,23 we
have proposed the formation of an “oriented complex”
between the reagent and the substrate (7), in which the
metal plays a fundamental role, that accounts for the
high ortho regioselectivity (Scheme 2). To rationalize the
stereochemical outcome of the present reaction, we
suggest that the reagent 2 coordinates the metal with
both carbonyl groups, recognizing that nonbonded inter-
actions are important for efficient chirality transfer.
Other authors6,9a presented a chelation mechanism for
reactions involving chiral glyoxylates and Lewis acids,
which can account for the absolute stereochemistry of the
products. However, there is no direct evidence for this
chelation, except a photophysical study.24 Here, we
present the X-ray structure of a complex obtained as dark
red crystals from the reaction mixture of 4-methoxyphe-
nol (1c) and (-)-menthylglyoxylate (2i), carried out in
the presence of TiCl4 at room temperature in Cl2CdCCl2
as solvent.
F igu r e 1. View of complex 8.
oxygen and one chlorine atoms in the basal plane and
with a second chlorine and a carbonyl oxygen in the apex
positions. The chlorine atoms of each titanium adopt a
cis orientation being Cl1 trans to the O2 carbonyl oxygen
and Cl2 trans to the O1 benzylic atom.
On the basis of this crystal structure, we can state with
confidence that the chiral glyoxylate can behave as a
chelating ligand as drawn in Scheme 2, with both the
oxygen atoms in the coordination sphere of titanium in
the complex formed with the reaction product.
Complex 8 has some remarkable features. The alcohol
functionality created in the reaction is present as a
titanium alkoxide, giving a stable chelation complex that
accounts for the absence of diphenylmethane26 byprod-
ucts arising from further reaction of the benzylic alcohol.
The distances between the metal and the benzylic oxygen
atoms are in agreement with the calculated single
covalent bond for Ti-O (2.01 Å), whereas the distances
between the metal and the phenolic oxygen atoms are
shorter, 1.785(2) and 1.834(2) Å, which indicates that
there is some double-bond character between Ti and the
phenolic oxygen atoms, as recently reported.27 Moreover,
the depicted structure of the titanium complex and the
crystal structure of ester 4ic confirm the preference for
the (-)-menthyl esters to accommodate the axial methine
hydrogen atom nearly coplanar and syn to the ester
carbonyl.1c,9a
The reaction of TiCl4 with 4-methoxyphenol proceeded
with spontaneous evolution of HCl to give the metal
phenoxide that reacted with the glyoxylate 2i.
The structure of the titanium complex 8 (Scheme 2, R
) 4-OCH3, R* ) (-)-menthyl) is shown in Figure 1.25 The
complex exists in the solid state as a dimer in which both
the diastereoisomers 3ic and 4ic are present. The
titanium atoms show octahedral coordination with three
(20) Solladie-Cavallo, A.; Khiar, N. Tetrahedron Lett. 1988, 29, 2189.
(21) (a) J ones, G. B.; Chapman B. J . Synlett 1997, 439. (b) J ones,
G. B.; Chapman B. J .; Mathews, J . E. J . Org. Chem. 1998, 63, 2928.
(22) Dradi, E.; Pochini, A.; Ungaro, U. Gazz. Chim. Ital. 1980, 110,
353.
(23) Bocelli, B.; Cantoni, A.; Sartori, G.; Maggi, R.; Bigi, F. Chem.
Eur. J . 1997, 3, 1269.
(24) Whitesell, J . K.; Younathan, J . N.; Hurst, J . R.; Fox, M. A. J .
Org. Chem. 1985, 50, 5499.
(25) For titanium phenoxides structures, see: (a) Floriani, C.;
Corazza, F.; Lesueur, W.; Chiesi-Villa, A.; Guastini, C. Angew. Chem.,
Int. Ed. Engl. 1989, 28, 66. (b) Floriani, C. Polyhedron 1989, 8, 1717.
(c) Ripa, A.; Colombo, P.; Porri, L. XI FECHEM, Parma, Italy, Sep
10-15, 1995; Abstract P35.
(26) Casiraghi, G.; Casnati, G.; Cornia, M.; Pochini, A.; Puglia, G.;
Sartori, G.; Ungaro, R. J . Chem. Soc., Perkin Trans. 1 1978, 318.
(27) Matilainen, L.; Klinga, M.; Leskela¨, M. J . Chem. Soc., Dalton
Trans. 1996, 219.

Synthesis of Optically Active 2-Hydroxymandelic Acids
J . Org. Chem., Vol. 64, No. 14, 1999 5007
using SHELX9331 with the anisotropic thermal parameters
only for titanium, chlorine, and oxygen atoms. Hydrogens were
not located. Most of the atoms were found to have a consider-
able freedom of thermal motion, but any attempts to sharply
elucidate the ordering of these atoms over two or more
positions was limited by the quality of data. The final R factors
for 309 refined parameters were 0.078 for 1036 reflections with
Fo > 4σ(Fo) and 0.105 for all 1695 data.
(2S)- a n d (2R)-2-Hyd r oxy-2-(2-h yd r oxy-4-ter t-bu tylp h e-
n yl)eth a n oic Acid (-)-Men th yl Ester s (3ia a n d 4ia ).
Typ ica l P r oced u r e. To a solution of 3-tert-butylphenol (1.5
g, 10 mmol) in anhydrous CH2Cl2 (34 mL) was added at -30
°C 5.0 mL of a 1 M solution of TiCl4 freshly prepared in
anhydrous CH2Cl2, under a stream of dry nitrogen. After the
mixture was stirred for 30 min, anhydrous (-)-menthyl
glyoxylate (2i) (10 mmol) was added. Anhydrous (-)-menthyl
glyoxylate (2i) was obtained as a viscous oil by heating the
hydrate form (white solid) at 90 °C under high vacuum before
use, and a CH2Cl2 solution, about 1 M, was prepared. The
reaction mixture was stirred for 16 h at -30 °C, quenched with
saturated aqueous ammonium chloride solution, and extracted
with CH2Cl2 (3 × 70 mL). After the solution was dried (Na2-
SO4), the solvent was removed under reduced pressure, and
3ia and 4ia were separated from the residue by chromatog-
raphy on silica gel using CH2Cl2 or hexane/ethyl acetate (80:
20) to give 2.97 g (82% yield). HPLC analysis of the major
isomer 3ia showed the lower tR.
Con clu sion s
We have presented the syntheses of optically active
2-hydroymandelic esters and acids via direct hydroxy-
alkylation of phenols with very high yields and diaste-
reoselectivities (up to 96% de).
The reaction occurs with complete ortho regioselectivity
with mono attack on the phenol ring, and no diphenyl-
methane derivatives are formed.
The hydrolysis and reduction of the products allow
isolation of enantiomerically pure 2-hydroxymandelic
acids 6 and 2-hydroxyphenyldiols 5, respectively.
The crystal structure of the titanium complex repre-
sents strong support for the chelation mechanism pre-
sented for glyoxylate reactivity and supports our proposal
for coordinated ortho alkylation of phenols.
Exp er im en ta l Section
Gen er a l Meth od s. Melting and boiling points are uncor-
rected. 1H NMR spectra were recorded at 200 MHz and at 400
MHz in the indicated solvents. 13C NMR spectra were recorded
at 25.18 MHz with proton decoupling. Mass spectra were
obtained in EI mode. Microanalyses were carried out by
Dipartimento di Chimica Generale ed Inorganica, Chimica
Analitica, Chimica Fisica dell’Universita` degli Studi di Parma,
Italy.
Crystallization from hexane/CH2Cl2 afforded pure (2S)-
3ia : colorless needles; mp 187-188 °C; [R]25 ) -24.9 (c 0.4,
D
CH2Cl2); IR (KBr) 1722 cm-1; UV λmax 278 nm (ꢀ 2003); 1H
NMR (200 MHz, CDCl3) δ 0.77 (d, J ) 6.9 Hz, 3H), 0.86 (d, J
) 6.4 Hz, 3H), 0.89 (d, J ) 6.9 Hz, 3H), 1.29 (s, 9H), 0.8-2.1
(m, 9H), 3.5 (br s, 1H), 4.80 (td, J ) 10.8, 4.4 Hz, 1H), 5.28 (s,
1H), 6.91 (dd, J ) 8.6, 1.9 Hz, 1H), 6.92 (d, J ) 1.9 Hz, 1H),
7.09 (d, J ) 8.6 Hz, 1H), 7.15 (br s, 1H); 13C NMR (CDCl3) δ
16.35, 20.69, 21.91, 23.42, 26.35, 31.20, 31.37, 34.05, 34.58,
40.23, 46.85, 72.26, 77.09, 114.67, 117.46, 119.58, 127.94,
153.54, 154.74, 172.81; MS m/z (rel intensity) 362 (9, M+), 224
(20), 180 (100), 179 (38). Anal. Calcd for C22H34O4: C, 72.89;
H, 9.45. Found: C, 73.21; H, 9.30.
Diastereomeric excesses were determinated by reversed-
phase HPLC using an octadecylsilyl column (ODS 5 µm) with
methanol/water solvent system with UV detector at 280 nm.
TLC analyses were carried out on Merck 60PF254 silica gel
plates using mixtures of hexanes-ethyl acetates.
All reagents were commercial materials from freshly opened
containers. Dichloromethane was stored over 4 Å molecular
sieves. Glyoxylates esters were prepared by the reported
methods.6,9c,28
(-)-Menthol was purchased from Carlo Erba. (-)-8-Phenyl-
menthol was synthesized from (+)-pulegone (from EGA-
Chemie) following the reported procedure.8
Satisfactory microanalyses and spectroscopic data were
obtained for all new compounds described.
(2R)-4ia : mp 110-113 °C; [R]25 ) -118 (c 0.7, EtOH); IR
D
(KBr) 1723 cm-1; H NMR (200 MHz, CDCl3) δ 0.45 (d, J )
1
6.9 Hz, 3H), 0.59 (d, J ) 6.9 Hz, 3H), 0.91 (d, J ) 6.5 Hz, 3H),
1.27 (s, 9H), 0.7-2.2 (m, 9H), 3.1 (br s, 1H), 4.67 (td, J ) 10.8,
4.3 Hz, 1H), 5.23 (s, 1H), 6.88-6.92 (m, 2H), 7.0 (br s, 1H),
7.09 (d, J ) 8.5 Hz, 1H); 13C NMR (CDCl3) δ 15.72, 20.53,
21.98, 23.09, 25.50, 31.25, 31.41, 34.12, 34.58, 40.66, 47.08,
72.67, 77.01, 114.32, 117.24, 119.99, 128.68, 153.54, 154.85,
172.96. MS m/z (rel intensity) 362 (40, M+), 224 (29), 180 (100),
179 (73). Anal. Calcd for C22H34O4: C, 72.89; H, 9.45. Found:
C, 73.26; H, 9.20.
Sin gle-Cr ysta l X-r a y Diffr a ction An a lysis of Com p lex
8. A prismatic crystal of approximately 0.09 × 0.11 × 0.14 mm
was mounted in a glass capillary, sealed under nitrogen, and
diffracted on a Siemens AED diffractometer using Cu KR
radiation with wavelength 1.5418 Å. The space group of this
C
38H52O10Cl4Ti2 crystal was P21 with a ) 10.928(3) Å, b )
27.371(4) Å, c ) 8.374(3) Å, â ) 98.92(4)°. Other crystal
parameters were as follows: cell volume ) 2474.46 Å3, Z ) 4,
dcalc ) 1.22 g cm-3, and µ ) 5.11 cm-1. A total of 1750
reflections were collected in the 3-50° θ range, and 1695 were
unique (Rint ) 0.075). A decay of a check reflection, monitored
periodically every 100 reflections, of about 54% was found
during the data collection time. The effects of this decomposi-
tion were corrected, together with those for Lorentz and
polarization effects, during the data reduction procedure,29 and
an empirical absorption correction was applied. This signifi-
cant decomposition, the limited θ range, and the small dimen-
sions of the available crystal affect the quality and quantity
of experimental data. Use of the SIR9730 package permits the
structure solution and the definition of the molecular skeleton.
The structure was refined by full-matrix least-squares methods
The following diastereomeric pairs of esters listed in Table
1 were prepared in a similar way:
2-H yd r oxy-2-(2-h yd r oxy-5-m et h oxyp h en yl)et h a n oic
a cid (-)-m en th yl ester ((2S)-3ib): white needles (hexane/
CH2Cl2); mp 123 °C; [R]25 ) -10.7 (c 0.4, EtOH); IR (KBr)
D
1725 cm-1; UV λmax 297 nm (ꢀ 4818); 1H NMR (200 MHz,
CDCl3) δ 0.76 (d, J ) 6.9 Hz, 3H), 0.85 (d, J ) 6.2 Hz, 3H),
0.89 (d, J ) 6.9 Hz, 3H), 0.7-2.1 (m, 9H), 3.68 (d, J ) 3.6 Hz,
1H), 3.75 (s, 3H), 4.80 (td, J ) 10.8, 4.3 Hz, 1H), 5.26 (d, J )
3.6 Hz, 1H), 6.7 (s, 1H), 6.7-7.0 (m, 3H); MS m/z (rel intensity)
336 (44, M+), 198 (100), 152 (73). Anal. Calcd for C19H28O5:
C, 67.83; H, 8.39. Found: C, 67.50; H, 8.19.
2-Hyd r oxy-2-(2-h yd r oxyp h en yl)eth a n oic a cid (-)-m en -
th yl ester ((2S)-3ic): colorless needles (hexane/CH2Cl2); mp
119-120 °C; [R]25 ) +4.5 (c 0.8, EtOH); IR (KBr) 1719 cm-1
;
D
UV λmax277 nm (ꢀ 3364); 1H NMR (400 MHz, CDCl3) δ 0.77 (d,
J ) 7.0 Hz, 3H), 0.85 (d, J ) 6.5 Hz, 3H), 0.89 (d, J ) 7.0 Hz,
3H), 0.7-2.0 (m, 9H), 3.55 (d, J ) 4.0 Hz, 1H), 4.80 (td, J )
10.9, 4.4 Hz, 1H), 5.30 (d, J ) 4.0 Hz, 1H), 6.8-7.0 (m, 2H),
7.14 (s,1H), 7.20 (d, J ) 8.5 Hz, 1H), 7.24 (d, J ) 7.7, 1H); MS
(28) (a) J urczak, J .; Zamojski, A. Roczn. Chem. 1970, 44, 865; Chem.
Abstr. 1971, 74, 140858. (b) Kornblum, N. J . Am. Chem. Soc. 1966,
88, 5.
(29) Belletti, D.; Cantoni, A.; Pasquinelli, G. Gestione on line di
diffrattometro a cristallo singolo Siemens AED con sistema IBM PS2/
30. Centro di Studio per la Strutturistica Diffrattometrica del C.N.R.,
Parma, 1988; internal report 1/88.
(30) Altomare, A.; Cascarano G.; Giacovazzo C.; Guagliardi A.;
Moliterni A. G. G.; Burla, M. C.; Polidori, G.; Camalli, M.; Spagna R.
Private communication, 1997.
(31) Sheldrick, G. M. SHELX93. Program for Crystal Structure
Refinement. University of Go¨ttingen, Germany, 1993.

5008 J . Org. Chem., Vol. 64, No. 14, 1999
Bigi et al.
m/z (rel intensity) 306 (6, M+), 168 (15), 123 (89). Anal. Calcd
for C18H26O4: C, 70.56; H, 8.55. Found: C, 70.67; H, 8.90.
(2R)-4ic: colorless prismatic crystals from an enriched
Anal. Calcd for C24H29O4Cl: C, 69.13; H, 7.01; Cl, 8.50.
Found: C, 69.00; H, 7.23; Cl, 8.25.
R ed u ct ion P r oced u r e.14 (-)-1-(2-H yd r oxy-4-ter t-b u -
tylp h en yl)-1,2-d ih yd r oxyeth a n e (5a ). To a solution of (2S)-
3ia (0.36 g, 1 mmol) in anhydrous THF (8 mL) and absolute
EtOH (12 mL) were added NaBH4 (0.37 g, 10 mmol) and LiCl
(0.42 g, 10 mmol), and the mixture was refluxed for 2 h with
stirring, providing complete reduction. After being cooled to
room temperature, the reaction was quenched with water and
extracted with hexane (3 × 15 mL) in order to recover (-)-
menthol. The aqueous phase was acidified with HCl 1 N to
pH ∼3 and extracted with diethyl ether (3 × 30 mL). After
the aqueous phase was dried (Na2SO4), the solvent was
removed under reduced pressure and 5a was purified by
chromatography on silica gel using hexane/ethyl acetate (70:
30) to give a white solid (0.16 g, 76%). Attempts to perform
the reduction using LiAlH4 in diethyl ether resulted in
unselective reaction. The enantiomeric excess of compound
(2S)-5a was determined by using MPTA (R-methoxy-R-trifluo-
romethylphenylacetic acid) derivatives.18 1H NMR (400 MHz,
CDCl3) spectra of the corresponding Mosher monoester, formed
by esterification of the primary alcoholic function, revealed
that 5a is enantiomerically pure. On the other hand, the (-)-
MPTA derivative of enriched 5a (de 50%) showed distinguish-
ible signals for the two diastereoisomers: three dd at δ 4.49
and 4.56 for methylene hydrogens and δ 5.12 for methine
proton of the major isomer and the corresponding three dd at
δ 4.47, 4.58, and 5.13 of the minor isomer.
diastereomeric mixture in hexane/CH2Cl2; mp 122-123 °C;
1
[R]25 ) -144.7 (c 0.6, EtOH); IR (KBr) 1718 cm-1; H NMR
D
(CDCl3) δ 0.48 (d, J ) 6.9 Hz, 3H), 0.61 (d, J ) 6.9 Hz, 3H),
0.91 (d, J ) 6.5 Hz, 3H), 0.7-2.2 (m, 9H), 3.62 (d, J ) 3.0 Hz,
1H), 4.69 (td, J ) 10.8, 4.4 Hz, 1H), 5.26 (d, J ) 3.0 Hz, 1H),
6.8-7.3 (m, 4H), 7.06 (s, 1H); MS m/z (rel intensity) 306 (5,
M+), 168 (24), 123 (100). Anal. Calcd for C18H26O4: C, 70.56;
H, 8.55. Found: C, 70.76; H, 8.42.
(2S)-2-Hyd r oxy-2-(2-h yd r oxy-5-ter t-bu tylp h en yl)eth a -
n oic a cid (-)-m en th yl ester (3id ): white wooly needles from
hexane/CH2Cl2; mp 159-160 °C; [R]25 ) -8.8 (c 0.7, EtOH);
D
IR (KBr) 3520, 3300, 1703, 1230 cm-1
;
1H NMR (400 MHz,
CDCl3) δ 0.77 (d, J ) 7.0 Hz, 3H), 0.85 (d, J ) 6.5 Hz, 3H),
0.89 (d, J ) 7.0 Hz, 3H), 0.8-2.0 (m, 9H), 1.28 (s, 9H), 3.53 (d,
J ) 4.5 Hz, 1H), 4.79 (td, J ) 10.9, 4.4 Hz, 1H), 5.33 (d, J )
4.5 Hz, 1H), 6.82 (d, J ) 8.5 Hz, 1H), 7.04 (s, 1H), 7.17 (d, J
) 2.4 Hz, 1H), 7.23 (dd, J ) 8.5, 2.4 Hz, 1H); MS (CI) m/z (rel
intensity) 362 (14, M+), 224 (100), 207 (47), 179 (62). Anal.
Calcd for C22H34O4: C, 72.89; H, 9.45. Found: C, 73.10; H,
9.20.
(2S)- a n d (2R)-2-Hyd r oxy-2-(2-h yd r oxy-4-ter t-bu tylph e-
n yl)eth a n oic Acid (-)-8-P h en ylm en th yl Ester s (3ja a n d
4ja ). Typ ica l P r oced u r e. To a solution of 3-tert-butylphenol
(1a ) (0.75 g, 5 mmol) in anhydrous CH2Cl2 (17 mL) was added
5.0 mL of a freshly prepared 1 M solution of TiCl4 in anhydrous
CH2Cl2 at room temperature under nitrogen. After the mixture
was stirred for 30 min, (-)-8-phenylmenthyl glyoxylate (2j)
(obtained as a mixture of anhydrous and hydrate form) (1.53
g, 5 mmol) in CH2Cl2 (10 mL) was added. The reaction was
stirred for 6 h at room temperature. Workup as above
furnished a residue from which pure diastereomeric esters 3ja
and 4ja were separated and purified by chromatography on
silica gel using CH2Cl2 as eluent, 2.08 g (95% yield). HPLC
analysis of the major isomer 3ja showed the higher tR.
5a : white solid; mp 88-89 °C; [R]25 ) +36 (c 0.3, EtOH);
D
IR (KBr) 3275, 1078, 1013 cm-1; 1H NMR (400 MHz, CDCl3) δ
1.29 (s, 9H), 2.05 (br s, 1H), 3.15 (s, 1H), 3.8-3.9 (m, 2H), 4.94
(dd, J ) 7.5, 4.9 Hz, 1H), 6.87 (dd, J ) 7.9, 1.9 Hz, 1H), 6.92
(d, J ) 1.9 Hz, 1H), 6.94 (d, J ) 7.9 Hz, 1H), 7.92 (br s, 1H);
MS m/z (rel intensity) 210 (20, M+), 192 (27), 180 (100), 163
(50). Anal. Calcd for C12H18O3: C, 68.55; H, 8.63. Found: C,
68.20; H, 8.30.
Hydr olysis P r ocedu r e:17 (+)-Hydr oxy-2-(2-h ydr oxyph e-
n yl)eth a n oic Acid (6c). To a solution of 3ic (0.61 g, 2.0 mmol)
in THF/H2O (2:1) (33 mL) was added LiOH‚H2O (0.25 g, 6.0
mmol) under nitrogen.
(2S)-3ja : white solid; mp 60-62 °C; [R]25 ) +19.2 (c 0.8,
D
EtOH); IR (KBr) 1730 cm-1; 1H NMR (200 MHz,CDCl3) δ 0.80
(d, J ) 6.5 Hz, 3H, CH3), 1.19 (s, 3H, CH3), 1.25 (s, 9H, (CH3)3
C), 1.32 (s, 3H, CH3), 0.7-2.2 (m, 8H), 3.22 (d, J ) 3.1 Hz,
1H, OH), 4.14 (d, J ) 3.1 Hz, 1H, CHOH), 4.91 (td, J ) 10.6,
4.4 Hz, 1H, CHO), 6.67 (d, J ) 8.8 Hz, 1H, H-6), 6.75-6.90
(m, 2H, H-3 and H-5), 6.94 (br s, 1H, OH), 7.1-7.4 (m, 5H,
Ph); 13C NMR (CDCl3) δ 21.59 (CH3), 22.91 (CH3), 26.20 (CH2),
29.60 (CH3), 31.15 ((CH3)3 and CH), 34.32(CH2), 34.40 (C),
39.37 (C), 40.75 (CH2), 50.19 (CH), 71.58 (CHOH), 76.23(CHO),
114.17 and 116.96 (C3 and C5), 119.34 (C), 125.32 (Ph), 128.10
(C6 and Ph), 151.80 (C), 153.07 (C), 154.50 (C), 172.34 (C);
the assignment of signals between 21 and 50 ppm was
established on the basis of DEPT experiments, and the peaks
between 71 and 128 ppm were assigned on the basis of 1H,
13C 2D correlation in the 1H decoupled version; MS m/z (rel
intensity) 438 (7, M+), 224 (33), 214 (11), 199 (12), 179 (71),
163 (13), 119 (80), 105 (100), 91 (44). Anal. Calcd for
The solution was stirred at room temperature overnight,
giving a complete hydrolysis. After addition of water, the
reaction was extracted with hexane (3 × 30 mL). From the
organic phase, after drying (Na2SO4) and solvent removal, (-)-
menthol (0.30 g, 1.9 mmol, 97% yield) was recovered. The
aqueous layer was acidified with 0.5 N HCl. After extraction
with ethyl acetate (4 × 20 mL), drying (Na2SO4), and removal
of the solvent under reduced pressure, crude 6c was obtained
as a viscous oil that became a sticky solid after warming in
vacuo at 70 °C for 3 h (0.31 g, 92% yield): 1H NMR (400 MHz,
DMSO-d6) 5.22 (s, 1H), 5.52 (br s, 1H), 6.75 (t, J ) 7.5 Hz,
1H), 6.77 (d, J ) 8.1 Hz, 1H), 7.07 (pseudo td, J ) 8.1, 7.5, 1.5
Hz, 1H), 7.22 (dd, J ) 7.5, 1.5 Hz, 1H), 10.9 (br s, 1H); MS
m/z (rel intensity) 168 (12, M+), 150 (21), 123 (46), 121 (100).
It was not possible to exactly measure the [R]D value because
the purification on silica gel (using CH2Cl2/MeOH 80:20)
C
28H38O4: C, 76.67; H, 8.73. Found: C, 76.30; H, 8.41.
The following diastereomeric pairs of esters listed in Table
resulted in product transformation (probably polymerization3c
)
1 were prepared in a similar way:
giving a white solid, insoluble in ether and CDCl3 ,that shows
a methine signal in 1H NMR (DMSO) at lower δ (4.7 ppm);
this material decomposed without melting. However, polari-
(2S)-2-H yd r oxy-2-(2-h yd r oxy-5-m et h oxyp h en yl)et h a -
n oic a cid (-)-8-p h en ylm en th yl ester s (3jb): [R]25D ) +16.5
1
(c 0.3, EtOH); H NMR (200 MHz, CDCl3) δ 0.78 (d, J ) 6.5
metric measurement revealed that 5c is dextrorotatory ([R]25
D
Hz, 3H), 1.19 (s, 3H), 1.32 (s, 3H), 0.6-2.2 (m, 8H), 3.49 (d, J
) 4.0 Hz, 1H), 3.71 (s, 3H), 4.09 (d, J ) 4.0 Hz, 1H), 4.92 (td,
J ) 10.6, 4.4 Hz, 1H), 6.3-6.9 (m, 3H), 6.50 (s, 1H), 7.1-7.5
(m, 5H); MS m/z (rel intensity) 412 (2, M+), 198 (75), 119 (82),
105 (100). Anal. Calcd for C25H32O5: C, 72.79; H, 7.82. Found:
C, 73.07; H, 8.03.
) +118.8 (c 0.5, EtOH)), as expected for the 2S configuration.15
The enantiomeric excess of compound 6c was determined
directly by HPLC chiral separation or, after esterification with
diazomethane, by Mosher ester analysis. Reaction of crude 6c
(84 mg, 0.5 mmol) in anhydrous Et2O (8 mL) with diaz-
omethane at 0 °C for 4 min gave quantitatively the methyl
ester 9c. The enantiomeric excess of compound 9c was
determined using MPTA derivatives.18 Reversed-phase HPLC
analysis of the (-)- and (+)-MPTA diester 10 revealed that
the ee of 9c is > 99%. 1H NMR (300 MHz, CDCl3) spectra
confirmed this result; indeed, the benzylic methine peaks for
the (-)- and (+)-MPTA derivative are easily distinguishible
at δ 6.27 and 6.17 ppm, respectively. 9c: viscous oil that
(2S)-2-Hyd r oxy-2-(2-h yd r oxy-5-ch lor op h en yl)eth a n o-
ic a cid (-)-8-p h en ylm en th yl ester s (3je): mp 63-66 °C;
1
[R]25 ) + 26.5 (c 0.4, EtOH); IR (KBr) 1720 cm-1; H NMR
D
(200 MHz, CDCl3) δ 0.81 (d, J ) 6.4 Hz, 3H), 1.18 (s, 3H),
1.31 (s, 3H), 0.6-2.3 (m, 8H), 3.4 (br s, 1H), 4.01 (s, 1H), 4.90
(td, J ) 10.7, 4.3 Hz, 1H), 6.65 (d, J ) 2.6 Hz, 1H), 6.70 (d, J
) 8.6 Hz, 1H), 7.07 (dd, J ) 8.6, 2.6 Hz, 1H), 7.15-7.5 (m,
6H); MS m/z (rel intensity) 416 (1, M+), 119 (100), 105 (88).
solidified on standing overnight; mp 65-67 °C;3c [R]25
)
D

Synthesis of Optically Active 2-Hydroxymandelic Acids
J . Org. Chem., Vol. 64, No. 14, 1999 5009
1
+127.2 (c 0.2, EtOH); H NMR (CDCl3) δ 3.6 (br s, 1H), 3.79
Ack n ow led gm en t. The authors acknowledge the
support of the Ministero dell’Universita` e della Ricerca
Scientifica e Tecnologica (MURST), Italy, and the
University of Parma (National Project “Stereoselezione
in Sintesi Organica. Metodologie ed Applicazioni”).
(s, 3H), 5.33 (s,1H), 6.88 (d, J ) 8.0 Hz, 1H), 6.91 (td, J ) 7.5,
1.1 Hz, 1H), 7.1 (br s, 1H), 7.15-7.3 (m, 2H); MS m/z (rel
intensity) 182 (82, M+), 123 (100).
(+)-Hyd r oxy-2-(2-h yd r oxy-4-ter t-bu tylp h en yl)eth a n o-
ic Acid (6a ). The acid 6a was prepared by hydrolysis of the
ester 3ia (0.72 g, 2.0 mmol). The acidified aqueous layer was
extracted with diethyl ether (4 × 20 mL), giving pure hydroxy
acid 6a (0.40 g, 90% yield) as a white solid: mp 89-92 °C,
97-98 °C when dehydrated by warming in vacuo at 70 °C;
[R]25D ) +113.1 (c 0.5, EtOH); IR (KBr) 3350, 1729, 1248, 1062
Su p p or tin g In for m a tion Ava ila ble: X-ray crystallo-
graphic data for complex 8 and a crystal structure with the
labeling scheme. 1H NMR spectral data for all compounds
reported (3-6, 9). This material is available free of charge via
the Internet at http://pubs.acs.org.
cm-1; H NMR (400 MHz, DMSO-d6) δ 1.22 (s, 9H), 5.19 (s,
1
1H), 5.42 (br s, 1H), 6.77-6.84 (m, 2H), 7.12 (d, J ) 8.4 Hz,
1H), 9.27 (br s, 1H), 12.22 (br s, 1H); MS m/z (rel intensity)
224 (26, M+), 179 (100), 163 (18). Anal. Calcd for C12H16O4:
C, 64.3; H, 7.2. Found: C, 63.9; H, 7.5.
J O982170O