Spontaneous resolution among chiral glycerol derivatives: crystallization features of ortho-alkoxysubstituted phenyl glycerol ethers

Article abstract of DOI:10.1016/j.tetasy.2007.08.019

Five chiral arylglycerol ethers 2-R-C₆H₄-O-CH₂CH(OH)CH₂OH (R = OMe, OEt, OPrⁿ, OPrⁱ, OBu^t) have been prepared in racemic and enantiopure form. The melting points and enthalpies of fusion of every species were measured by differential scanning calorimetry. Binary phase diagrams were reconstructed for the whole family, the entropies of the mixing of the enantiomers in the liquid state, and Gibbs free energy of formation of the racemic compound, as well as Pettersson i-values were derived from the thermal data. The differences in the phase behavior of the investigated compounds were associated with the conformations of the alkoxy fragments.

Full text of DOI:10.1016/j.tetasy.2007.08.019

Tetrahedron: Asymmetry 18 (2007) 1964–1970
Spontaneous resolution among chiral glycerol derivatives:
crystallization features of ortho-alkoxysubstituted
phenyl glycerol ethers
Alexander A. Bredikhin,* Zemfira A. Bredikhina, Dmitry V. Zakharychev and
Larisa V. Konoshenko
A.E. Arbuzov Institute of Organic and Physical Chemistry, Russian Academy of Sciences, Arbuzov Street, 8,
Kazan 420088, Russian Federation
Received 12 July 2007; accepted 16 August 2007
Abstract—Five chiral arylglycerol ethers 2-R-C6H4-O-CH2CH(OH)CH2OH (R = OMe, OEt, OPrn, OPri, OBut) have been prepared in
racemic and enantiopure form. The melting points and enthalpies of fusion of every species were measured by differential scanning cal-
orimetry. Binary phase diagrams were reconstructed for the whole family, the entropies of the mixing of the enantiomers in the liquid
state, and Gibbs free energy of formation of the racemic compound, as well as Pettersson i-values were derived from the thermal data.
The differences in the phase behavior of the investigated compounds were associated with the conformations of the alkoxy fragments.
ꢀ 2007 Elsevier Ltd. All rights reserved.
1. Introduction
no have finished their review, thus ‘the understanding and
prediction of spontaneous resolution—an apparently mod-
est goal when compared with other contemporary ends—
remains one of the true challenges for science in the 21st
century’.5
The increasing demand for enantiopure chemicals has pro-
moted the development of different optical resolution
methods among which crystallization based ones play an
exceptional role.1 Such separations are most often per-
formed by the formation and crystallization of diastereo-
meric derivatives, also direct resolution methods (e.g.,
preferential crystallization2) could be less expensive and
simpler. The last methods, however, can only be applied
to racemates that form a conglomerate, that is, a mechan-
ical mixture of enantiopure crystals.3 The phenomenon of
conglomerate formation, referred to as the spontaneous
resolution of racemates upon crystallization, or simply
spontaneous resolution, has attracted considerable interest
as evidenced by the substantial number of the reviews pub-
lished on the subject over the last year.2,4–6
The modern level of theory has given no way to solve the
problem ‘from the first principles’. Occasional attempts
to predict a chiral substance crystallization type have pro-
ven to be very abstract7 or rather vulnerable with regards
to their axiomatization.8 We believe that a less sophisti-
cated but also more practical way of thinking about empir-
ical models linking the crystal structure changes with the
chemical structure variations in the series of closely related
compounds would be of great use. The members of such a
series must be selected in such a way that every compound
would have minimal but regular distinctions from each
other. In addition, it would be desirable to have not only
two qualitative categories ‘conglomerate’ and ‘racemic
compound’ for the characterization of crystalline type,
but to introduce a quantitative measure allowing us to rank
the observed properties.
With regards to spontaneous resolution, all reviewers are in
agreement that the understanding of the factors controlling
the two enantiomers’ behavior in solution or melt upon
crystallization is rather limited. Perez-Garcia and Amabili-
This measure is not easy to introduce for conglomerate
forming substances, while for the racemic compound form-
ing substances, one can use at least two numeric values.
*
Corresponding author. Tel.: +7 843 2727392; fax: +7 843 2731872;
e-mail: baa@iopc.knc.ru
0957-4166/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2007.08.019

A. A. Bredikhin et al. / Tetrahedron: Asymmetry 18 (2007) 1964–1970
1965
The relative stability of a racemic compound can be evalu-
R
OH
*
OH
+ ClCH2CH(OH)CH2OH
R
ated based on the melting phase diagram of a chiral sub-
stance using that introduced by Pettersson dimensionless
units i according to the equation
NaOH
-NaCl
O
OH
3a-e
1a-e
2
i ¼ ðT Rm ꢀ T mE Þ=ðTmA ꢀ TmE Þ
ð1Þ
R = OCH3 (a); OCH2CH3 (b);O(CH2)2CH3 (c);
OCH(CH3)2 (d);OC(CH3)3 (e)
where Tm are the melting temperatures of the racemate
(upper index R), pure enantiomer (index A), and the eutec-
tic (index E). According to Pettersson9,3 the values i < 0.5
characterize unstable racemic compounds, 0.5 6 i 6 1.5
are indicative of their moderate stability, and i > 1.5 are
typical of chiral compounds that form stable racemic com-
pounds with a 1:1 composition. For all conglomerates
TRm ¼ TEm, hence the i value is always equal to zero. Another
way to evaluate the relative stability of racemic compound
is to estimate the Gibbs free energy changes accompanied
by the reaction of the racemic compound formation from
the enantiopure components. Based on a thermodynamic
cycle involving the solid and liquid phases of the enantio-
mers and racemic species, formulas for the DG0 calcula-
tions were proposed by Grant et al.10
(S)-Diols were obtained from (S)-chloropropane-1,2-diol
and (R)-chloropropane-1,2-diol gave (R)-diols.
tert-Butoxy substituted phenol 1e was obtained by follow-
ing the scheme of Main et al.13
BzO
OH
HO
OH
PhCO2H
P2O5 / H3PO4
4
HO
OBut
BzO
OBut
NaOH
EtOH
CF3SO3H
ðTfR ꢀ TfAÞDHfA
DG0Tf ¼ ꢀ
ꢀ TfRR ln 2 ðif TRf ꢀ TfA < 0Þ
1e
5
TAf
R
ð2aÞ
2.2. IR spectroscopy
ðTfR ꢀ TfAÞDHfR
DG0Tf ¼ ꢀ
ꢀ TfAR ln 2 ðif TRf ꢀ TfA > 0Þ
To investigate the type of crystallization for our com-
pounds, we compared the IR spectra of the racemic and
highly enantiomerically enriched crystalline samples of
3a–e in KBr pellets, since the IR spectra of the optically ac-
tive and the racemic form should be identical for the con-
glomerate formation and different for racemic compounds.
TRf
A
ð2bÞ
Here TfR, T Af , DHRf , DHfA are the melting point temperatures
and fusion enthalpies of the racemate and pure enantiomer,
respectively. The Gibbs free energy of formation is always
negative for a racemic compound, if it can exist, while for a
racemic conglomerate, this value must be (but not always
is) close to zero.
To substantiate this comparison, the spectra were subjected
to a procedure of normalization and baseline correction,
as described in our previous paper.14 For this purpose,
coefficients that minimize the difference As ꢀ [a0 + a1m +
Ar(a2 + a3m)], where As and Ar are the molar absorption
coefficients of the scalemic and racemic samples, respec-
tively; m is the IR radiation frequency corresponding to
A, and an are the desired regression coefficients, were
selected by the least-squares method. It was reasonable to
introduce the regression terms a1m and Ara3m to correct
the spectral differences caused by the nonspecific (not
related to particular absorption bands) interaction of IR
radiation with matter (probably, by radiation scatter on
heterogeneities of the sample). It should be noted that the
use of polynomials of higher powers (quadratic and cubic)
for the generation of differential spectra does not improve
the statistical parameters characterizing regression. The
ratio between the mean-square deviation of the differential
curves and the mean-square deviation of spectral curves for
the racemate, that is, the ratio of error to variation (%),
was used as a quantitative characteristic for differential
curves.
Recently, we disclosed the conglomerate nature and devel-
oped an effective direct resolution procedure of 3-(2-meth-
oxyphenoxy)-1,2-propanediol, the popular chiral drug
guaifenesin.11 The purposes of this work are to obtain a
series of chiral o-alkoxysubstituted phenyl glycerol ethers
as racemates and as enantiomers; to compare in pairs the
solid state IR spectra of the samples; to investigate the
melting of the aryl glycerol ethers as enantiomers, as race-
mates, and in some cases as mixtures of intermediate com-
positions by means of differential scanning calorimetry; to
evaluate the stabilities of racemic compounds under exper-
iment; and to compare the crystallization peculiarities with
the substitution pattern.
2. Results
2.1. Chemistry
The availability of both enantiomers of 3-chloro-2,3-pro-
panediol 2 by Jacobsen kinetic hydrolytic resolution of a
racemic epichlorohydrin12 makes it possible to obtain all
the series of aryloxypropanediols 3a–e through the use of
a single process:
Figure 1 shows good coincidence between the pairs of spec-
tra for compounds 3a and 3d under visual comparison; the
same is almost true for the normal propoxy derivative 3c,
whereas the spectra of racemic and enantiopure crystalline
samples for tert-butoxy compound 3e, and especially for

1966
A. A. Bredikhin et al. / Tetrahedron: Asymmetry 18 (2007) 1964–1970
phenyl glycerol ethers. At the same time, the IR-test con-
firms the conglomerate nature of guaifenesin 3a. There is
a great probability that a racemic conglomerate is also
formed by isopropoxy derivative 3d. The question of a
crystalline type for compound 3c cannot be solved on this
basis.
2.3. Thermochemical investigations
This part of the work deals with the binary mixtures of (R)-
and (S)-compounds 3a–e using differential scanning calo-
rimetry (d.s.c.) as a research method. The temperature data
were determined according to the method of Ho¨hne et al.,15
and were treated as described previously.16
The results obtained for the temperature and the enthalpy
of fusion of the pure enantiomers and the pure racemates,
as well as the calculated3,10 values of entropy of mixing
for the liquid enantiomeric compounds, DSml , and the free
energy of formation of racemic compounds in the solid
state, DG0, are presented in Table 1. The calculated and
experimentally obtained eutectic melting temperatures
and eutectic compositions, as well as Pettersson’s i-values
calculated according to Eq. 1 are presented in the same
table.
According to the i-values all but one (compound 3e) of the
racemic compounds in the series studied are very unstable,
if they exist at all. It is not feasible to draw a precise line of
distinction between compounds 3a–d with only this crite-
rion in mind as the experimental accuracy of the Tfeu deter-
mination may amount to 0.5 degree. As a result, other
criteria with a more distinct physical sense must be used
to attribute the crystalline type of chiral compound in the
boundary cases.
Thus, the entropy of mixing for enantiomers 3a and 3d in
the liquid state is equal to 5.31 and 5.45 J Kꢀ1 molꢀ1
,
which is slightly less, but close to the ideal value of
5.75 J Kꢀ1 molꢀ1 (Rln2) for conglomerates. The near zero
value for DG0 also points to the same peculiarity of chiral
3a and 3d.10 The relatively high negative value for DG0
for tert-butoxy derivative 3e is a good diagnostic for a sta-
ble racemic compound formation in the crystalline state.10
The intermediate DSml and especially DG0 values for ethoxy
and propoxy diols 3b and 3c preclude conglomerate forma-
tion, and are compatible with the assumption of a rather
unstable racemic compound formation.
Figure 1. IR spectra of the crystalline samples 3a–e. Red curves—
racemates, blue curves—scalemates, black curves are differential curves
(see text).
From the d.s.c. data, the idealized melting temperatures
against the composition diagrams were reconstructed and
are depicted in Figure 2. The binary phase diagrams for
compounds 3a and 3d have an obvious single eutectic V-
shape typical for a racemic conglomerate.3 It follows that
the eutectic ee for guaifenesin 3a (already known) and for
the isopropoxyphenyl ether 3d are equal to zero. The phase
diagram for 3e is very typical for a racemic compound. The
eutectic ee in this case found as mutual point(s) for Schro¨-
der–Van Laar and Prigogine–Defay curves branches is
equal to 57%, that is, a sample with about 79% or more
of one enantiomer which could only be enantioenriched
by crystallization. The phase diagrams for 3b and 3d repre-
the ethoxy substituted one, differ noticeably. A similar
pattern can be observed for the differential curves: for
compounds 3a and 3d, and to a greater or lesser extent
for compound 3c, differences between the spectra of the
racemate and enantioenriched sample are about the same
level as instrumental background. At the same time, they
are rather substantial for compound 3e and dramatic for
compound 3b. This is in agreement with the assumption
that the racemic compounds are formed upon crystalliza-
tion of racemic ethoxy, and tert-butoxy ortho-substituted

A. A. Bredikhin et al. / Tetrahedron: Asymmetry 18 (2007) 1964–1970
1967
Table 1. D.s.c. measured melting point (Tf) and enthalpy of fusion (DHf) of racemic (low index R) and enantiopure (low index A) compounds 3a–e and
calculated thermodynamic characteristics for these substances, calculated and measured eutectic (low index eu) fusion temperature and eutectic
enantiomeric composition, along with Pettersson’s i-values (see text)
Compd
TfA (ꢁC)
TfR (ꢁC)
DHfA
DHfR
Tfeu
calcd (ꢁC)
,
Tfeu; exp (ꢁC)
eeeu (%)
DSml
DG0
i-Value
(kJ molꢀ1
)
(kJ molꢀ1
)
(J Kꢀ1 molꢀ1
)
(J molꢀ1
)
3a
3b
3c
3d
3e
97.2
78.7
88.8
80.8
54.7
79.9
65.6
73.4
62.6
53.1
43.0
36.5
38.6
37.9
29.0
36.9
36.2
37.9
33.4
30.5
79.7
65.1
73.2
62.7
47.4
79.7
65.1
73.0
62.6
47.4
0
22
12
0
5.31
3.99
4.70
5.45
0.46
ꢀ15
ꢀ595
ꢀ353
14
0.0
0.04
0.02
0.0
57
ꢀ1732
0.77
3. Discussion
There are no monotone phase behavior changes (e.g., the
smooth rise of a racemic compound stability) in the series
of alkoxy substituted glycerol ethers with the monoto-
nously changing substituents. In as much as the molecular
structures of compounds 3a–e are substantially the same,
the differences in the crystal structure are bound to the alk-
oxy fragment structure in the solid state.
Until now only the published crystallographic information
on the family has been available for guaifenesin rac-3a.17
As has been established in our earlier work,18 conglomerate
racemic guaifenesin crystallizes in the chiral space group
P212121. As it can be evaluated from the cif-file17 (refcode
PANKUL, available free of charge from the Cambridge
Crystallographic Data Centre), the main supramolecular
pattern in the homochiral crystal lattice is the endless 3D
‘wall’ formed by the intermolecular hydrogen bonds be-
tween glycerol moieties. Each glycerol hydroxyl is simulta-
neously a donor (H) and an acceptor (O) for two hydrogen
bonds with two different molecules. The tightly hydrogen
bonded region positions itself in the central part of every
wall. The physical bonding between the walls occurs
through hydrophobic interactions of the peripheral phenyl
and alkoxy groups. The thus formed crystal packing is sta-
ble enough to exclude racemic compound formation. One
can try to evaluate the factors destabilizing (or stabilizing)
this homochiral mode of packing in the case of other alk-
oxy substituted compounds 3b–e.
The statistics of the crystal structures for methoxysubstitut-
ed benzenes19,20 shows that the mean torsion angle
s1 = CAr–CAr–O–CMe is close to zero, if at least one of
the ortho-positions is free of anything but H substituents.
We have analyzed the Cambridge Crystallographic Data-
base21 (CCD) information for all organic molecules having
another alkoxysubstituted (Alk = Et, n-Pr, i-Pr, tert-Bu)
phenyl ring with at least one near alkoxy group unsubsti-
tuted position. The torsion angles nearest to an aromatic,
that is, s1 = CAr–CAr–O–CAlk
, as well as torsions
Figure 2. Experimental (circles) points and calculated (solid lines) binary
melting phase diagrams for compounds 3a–e.
s2 = CAr–O–CAlk–CAlk for ethoxy and propoxy derivatives,
and s02 = CAr–O–CAlk–H for isopropoxy substituted
aromatics were analyzed. Some results are summarized in
Table 2.
sent rather conglomerate-like curves with very shallow
plateaus in the racemate region. The theoretical eutectic
ee for compound 3b is about 22%, an even smaller value
of 12% characterizes the eutectic for propoxy substituted
ether 3c.
From Table 2 it follows that the methoxy, ethoxy, propoxy,
and isopropoxy substituted aromatics can be characterized
as near planar within the CArCArOCAlk fragment. CCD
contains structure information about only two organic

1968
A. A. Bredikhin et al. / Tetrahedron: Asymmetry 18 (2007) 1964–1970
Table 2. The Cambridge Crystallographic Databasea statistics for some torsion angles in the alkoxy substituted aromatics Ar–OR with at least one near
alkoxy group unsubstituted position
R in Ar–O–R
Number of entries
s1 = CAr–CAr–O–CAlk (ꢁ)
s2 = CAr–O–CAlk–CAlk (ꢁ)
s02 = CAr–O–CAlk–H (ꢁ)
a, Meb
b, Et
ꢁ0
248
62
133
2
0.7 1.0
0.5 2.8
5.7 3.7
ꢁ90
178.6 1.0
179.0 2.3
c, Pr
d, Pri
e, But
43.0 3.2
a Ref. 21.
b For details see Refs. 19 and 20.
compounds having tert-butoxy substituted phenyl rings,
namely two short peptides with the O-tert-butyltyrosyl moi-
eties. Both have the (CH3)3C–O bond orthogonal to the
aromatic plane, s1 ꢁ 90ꢁ. We believe that the bulky tert-but-
oxy substituent located from only one side of phenyl ring
hinders the two 3e molecules approaching each other, there-
by not allowing homochiral hydrogen bonded wall
(HHBW) formation. This is probably the reason for the
substantial differences in the phase diagram for compound
3e when compared with the other members of the series.
fall into two groups. The first, consisting of only one
representative (tert-BuO, compound 3e), demonstrates the
formation in the solid state of a molecular racemic
compound of moderate stability. The representatives of
the second group, compounds 3a–d, form very unstable
molecular racemic compounds 3b and 3c, or undergo
spontaneous resolution, crystallizing as racemic conglom-
erates, that is, form no racemic compound at all (3a and
3d). We believe that the main difference between the two
groups is related to the conformation of an alkoxy substi-
tuent in the solid state: the expected value of the di-
hedral angle CAr–CAr–O–CAlk is about 90ꢁ for the tert-
butoxy derivative and close to zero for all other alkoxysub-
stituted phenyl glycerol ethers. The fine structural differ-
ences between the alkyl fragments could be the reason for
controlling the crystallization type within the second
group.
The reasons for a pairwise similarity in the phase behavior
between compounds 3a,d and 3b,c, and on the contrary for
a pairwise distinction between the above mentioned pairs
could be concealed in the conformation of the alkyl part
of the alkoxy substituents. For the ethoxy and normal pro-
poxy aromatics, the fragment CAr–O–CAlk–CAlk is near
planar trans, so one would expect a near planar organiza-
tion of the alkoxyaromatic fragment for compounds 3a–
c. One additional feature needs to be noted. The ethyl
and propyl fragments are more extended when compared
to the methyl group. This is not important for hydrogen
bonding patterned after 3a, but it is important for periph-
eral hydrophobic bonding. The shortest C/C distance be-
tween the carbon atoms of the OCH3 groups amounts to
5. Experimental
5.1. General
˚
˚
as much as 4.98 A within the HHBW, and only 3.59 A
for the neighboring walls in the 3a crystal lattice, hence
there is no way to put an additional pair of CH2 (even more
so for CH2CH2) moieties toward one another from oppo-
site directions conserving alkoxy group planarity. This will
lead either to a change in the optimum conformation of the
alkoxy substituent within the homochiral packing or, as it
is realized for racemic 3b and 3c, for changing the homo-
chiral packing patterned after 3a to a heterochiral one.
The NMR spectra were recorded on a Bruker Avance-600
spectrometer in CDCl3 with TMS or the signals of the sol-
vent as the internal standard. The IR spectra of the poly-
crystalline samples of rac- and scal-compounds under
investigations in KBr pellets were recorded on a Bruker
IFS-66v Fourier-transform spectrometer. Optical rotations
were measured on a Perkin–Elmer model 341 polarimeter
(concentration c is given as g/100 mL). Melting points for
general purposes were determined using a Boe¨tius appara-
tus and are uncorrected.
As for the isopropoxy derivative, the CCD statistics allows
us to make rough estimates of the two s2 values as
s02 120ꢁ. For positive s02 this means 163 and ꢀ77ꢁ; that
is, no C–CH3 bond lies in the plane of molecule 3d, and
the optimal conformation for 3d is compatible with the
crystal lattice of the guaifenesin type. Moreover, excess
CH3 groups located in the hollows of HHBW could enlarge
the hydrophobicity of the walls periphery, and for this rea-
son to stabilize the homochiral packing.
Melting curves were measured on a Perkin–Elmer Dia-
mond DSC differential scanning calorimeter in aluminum
pans with a rate of heating of 10 ꢁC minꢀ1. The mass of
the samples amounted to approximately 2.5 mg. Tempera-
ture scale and heat flux were calibrated against the data for
indium, phenol, and naphthalene.
HPLC analyses were performed on a Shimadzu LC-20AD
system controller, and UV monitor 275 nm was used as a
detector. The column used, from Daicel, Inc., was Chiralcel
OD (0.46 · 25 cm). All experiments except in the case of
compound 10 were run with column temperature 40 ꢁC;
eluent hexane/isopropanol/diethylamine = 80:20:0.1; and
flow rate 1.0 ml/min.
4. Conclusion
In conclusion, we are able to say that the investigated
series of chiral o-alkoxysubstituted phenyl glycerol ethers

A. A. Bredikhin et al. / Tetrahedron: Asymmetry 18 (2007) 1964–1970
1969
5.2. Synthesis
1H, CH2O), 4.18 (dd, J = 9.7, 3.1 Hz, 1H, CH2I), 6.88–
6.97 (m, 4H, Ar). 13C NMR (150.864 MHz) d 14.77
(CH3), 64.03 (CH2CH3), 64.32 (CH2OH), 69.80 (CH),
73.46 (CH2O), 113.10 ðC3 Þ, 115.99 ðC6 Þ, 121.08 ðC4 Þ,
Racemic epichlorohydrin (99%), guaiacol 1a (98%), 2-eth-
oxyphenol 1b (98%), 2-isopropoxyphenol 1d (97%), and
rac-guaifenesin rac-3a were purchased from Alfa Aesarꢂ;
rac-3-chloropropane-1,2-diol rac-2 (99+%) was pur-
chased from Acros Organicsꢂ. (R)- and (S)-3-chloro-1,2-
propanediol (R)-2 and (S)-2 were prepared through
Jacobsen kinetic hydrolytic resolution of rac-2 without
modifications.12
Ar
Ar
Ar
122.64 ðC5ArÞ, 148.18 ðC1ArÞ, 149.38 ðCA2 rÞ.
5.2.5. (S)-3-(2-Ethoxyphenoxy)-propane-1,2-diol, (S)-3b.
20
Yield 68%, mp 78–79 ꢁC; ½aꢂD ¼ þ8:8 (c 1, hexane/EtOH
20
20
4:1), ½aꢂD ¼ þ11:7 (c 1, EtOH); {lit.:24 ½aꢂD ¼ þ3:9 (c 1,
acetone)}. 99.2% ee [chiral HPLC analysis; tR = 16.5 (min-
or), 19.3 min (major)]. NMR spectra were identical with
that cited above for rac-7.
5.2.1. 2-n-Propoxyphenol, 1c. A solution of NaOH (2.8 g,
0.07 mol) in water (11 ml) was added to a solution of pyro-
catechol (7.7 g, 0.07 mol) in ethanol (50 ml), and the result-
ing mixture was stirred and heated at reflux for 45 min. A
solution of 1-bromopropane (9 g, 0.07 mol) was then added
dropwise within 1 h, and the mixture was further stirred
and heated at reflux for 6 h. After removal of the solvent
in vacuo, the residue was purified by distillation. Yield
7.0 g (65%); bp 81–84 ꢁC (0.5 Torr); n2D0 1:5120; {lit.22 bp
80–83 ꢁC (4 Torr)}. 1H NMR (600 MHz) d 1.05 (t,
J = 7.6 Hz, 3H, CH3), 1.85 (m, 2H, CH2CH3), 4.01 (t,
J = 6.6 Hz, 2H, OCH2), 5.70 (s, 1H, OH), 6.95–6.79 (m,
4H, Ar).
5.2.6. rac-3-(2-Propoxyphenoxy)-propane-1,2-diol, rac-3c.
1
Yield 74%, mp 73–74 ꢁC (hexane). H NMR (600 MHz)
d 1.04 (t, J = 7.3 Hz, 3H, CH3), 1.81–1.87 (m, 2H,
CH2CH3), 2.75 (t, J = 5.8, 6.0 Hz, 1H, OH), 3.40 (d,
J = 5.2 Hz, 1H, OH), 3.77–3.83 (m, 2H, CH2IH), 3.94 (t,
J = 6.6 Hz, 2H, CH2CH2), 4.01–4.03 (m, 1H, CH), 4.05
(dd, J = 9.4, 5.8 Hz, 1H, CH2O), 4.16 (dd, J = 9.4,
3.2 Hz, 1H, CH2O), 6.87–6.96 (m, 4H, Ar). 13C NMR
(150.864 MHz)
d 10.46 (CH3), 22.57 (CH2), 64.03
(OCH2), 69.87 (CH2O), 70.44 (CH), 73.17 (CH2O),
113.22 ðC3ArÞ, 115.82 ðCA6 rÞ, 121.03 ðC4ArÞ, 122.55 ðCA5 rÞ,
148.23 ðC1ArÞ, 149.52 ðCA2 rÞ.
5.2.2. 2-tert-Butoxyphenol, 1e. This was obtained follow-
ing the published scheme.13 Through acylation of pyrocat-
echol by benzoic acid in polyphosphoric acid 2-hydroxy-
phenyl benzoate 4 was obtained. Yield 55%; mp 132–
5.2.7. (S)-3-(2-Propoxyphenoxy)-propane-1,2-diol, (S)-3c.
20
The yield was 74%, mp 88–90 ꢁC (hexane); ½aꢂD ¼ þ3:1
20
(c 1, hexane/EtOH 4:1); ½aꢂD ¼ þ6:8 (c 1, EtOH); 99.8%
1
133 ꢁC; (lit.13 mp 130 ꢁC). H NMR (600 MHz) d 5.5 (br
ee [chiral HPLC analysis; tR = 14.9 min (minor), 16.9 (ma-
jor)]. NMR spectra were identical with that cited above for
rac-3c.
s, 1H, IH), 7.0 (dt, J = 7.8, 1.3 Hz, 1H), 7.1 (dd, J = 8.1,
1.1 Hz, 1H), 7.21 (m, 2H), 7.55 (t, J = 7.8 Hz, 2H), 7.7 (t,
J = 7.6 Hz, 1H), 8.2 (d, J = 7.3 Hz, 2H); (cf. lit.13). Triflu-
oromethanesulfonic acid catalyzed addition of 4 to isobu-
tene has led to 2-tert-butoxyphenyl benzoate, 5. Yield
70%; mp 62 ꢁC. 1H NMR (600 MHz) d 1.33 (s, 9H),
7.12–7.15 (m, 1H), 7.19–7.22 (m, 2H), 7.54 (t, J = 7.4,
7.9 Hz, 2H), 7.65 (t, J = 7.4 Hz, 1H), 8.25 (d, J = 7.4 Hz,
2H). Saponification of 5 by aqueous NaOH followed by
distillation of the crude product allows us to obtain 2-
tert-butoxyphenol, 1e. Yield 60%; bp 60 ꢁC (0.4 Torr);
5.2.8. rac-3-(2-Isopropoxyphenoxy)-propane-1,2-diol, rac-
3d. The yield was 88%, mp 63–64 ꢁC (hexane). 1H
NMR (600 MHz) d 1.35 (d, J = 4.2 Hz, 6H, CH3), 2.79
(br s, 1H, OH), 3.48 (br s, 1H, OH), 3.75–3.81 (m, 2H,
CH2IH), 3.99–4.02 (m, 1H, CH), 4.04 (dd, J = 9.4,
5.9 Hz, 1H, CH2O), 4.14 (dd, J = 9.4, 2.9 Hz, 1H,
CH2O), 4.50–4.54 (m, 1H, CH), 6.88–6.95 (m, 4H, Ar).
13C NMR (150.864 MHz) d 22.08 (CH3), 64.00 (OCH2),
69.86 (CH), 71.57 (CH), 73.39 (CH2O), 115.85 ðC3 Þ,
1
{lit.13 bp 60 ꢁC (0.4 Torr)}. H NMR (600 MHz) d 1.44
Ar
(s, 9H), 5.77 (s, 1H), 6.78–6.81 (m, 1H), 6.96–6.99 (m,
116.41 ðC6ArÞ, 121.47 ðCA4 rÞ, 122.58 ðC5ArÞ, 148.23ðCA1 rÞ ,
2H), 7.04 (d, J = 7.9 Hz, 1H); (cf. lit.13).
149.27 ðC2ArÞ.
Racemic diols 3b–e and enantiomeric diols 3a–e were syn-
thesized by analogy with a published procedure23 from
racemic or enantiomeric 3-chloropropane-1,2-diols and
corresponding phenol; (S)-aryloxypropanediols were ob-
tained from (S)-chloropropane-1,2-diol, and vice versa.
Only the (S)-enantiomers are reported.
5.2.9. (S)-3-(2-Isopropoxyphenoxy)-propane-1,2-diol, (S)-
3d. The yield was 79%, mp 81–82 ꢁC (light petroleum);
½aꢂD ¼ þ7:8 (c 1, EtOH); ½aꢂD ¼ þ6:6 (c 1, hexane/EtOH
4:1); 99.3% ee [chiral HPLC analysis; tR = 26.3 (minor),
27.7 min (major)]. NMR spectra were identical with cited
above for rac-3d.
20
20
5.2.3. (S)-3-(2-Methoxyphenoxy)-propane-1,2-diol, (S)-
3a. The yield was 77%, mp 98–99 ꢁC (lit.:11 mp 98–
5.2.10.
rac-3e. The yield was 65%, mp 50–52 ꢁC (pentane). 13C
NMR (150.864 MHz) d 27.90 (CH3), 62.96 (OCH2), 69.51
rac-3-(2-tert-Buthoxyphenoxy)-propane-1,2-diol,
20
99 ꢁC); ½aꢂD ¼ þ9:5 (c 1.0, MeOH); 99.9% ee [chiral HPLC
analysis; tR = 10.3 (minor), 17.3 min (major)].
(CH), 70.73 (CH2O), 79.73 (CMe3), 114.49 ðC3 Þ, 120.74
Ar
ðC6ArÞ, 123.75 ðCA4 rÞ, 124.71 ðCA5 rÞ, 144.15ðCA1 rÞ , 152.21
5.2.4. rac-3-(2-Ethoxyphenoxy)-propane-1,2-diol, rac-3b.
ðC2ArÞ.
1
Yield 76%, mp 64–65 ꢁC. H NMR (600 MHz) d 1.45 (t,
J = 7.1 Hz, 3H, CH3), 2.59 (t, J = 6.0, 6.6 Hz, 1H, OH),
3.28 (d, J = 5.5 Hz, 1H, OH), 3.79–3.84 (m, 2H, CH2O),
4.00–4.04 (m, 1H, CH), 4.05–4.09 (m, 2H, CH2CH3 and
5.2.11. (S)-3-(2-tert-Buthoxyphenoxy)-propane-1,2-diol, (S)-
3e. The yield was 60%, mp 47–52 ꢁC (pentane);
20
½aꢂD ¼ þ8:8 (c 1, EtOH); 87.8% ee [chiral HPLC analysis;

1970
A. A. Bredikhin et al. / Tetrahedron: Asymmetry 18 (2007) 1964–1970
9. Pettersson, K. Ark. Kemi 1956, 10, 297–323; Pettersson, K.
Chem. Abstr. 1957, 51, 10474h.
10. Li, Z. J.; Zell, M. T.; Munson, E. J.; Grant, D. J. W. J.
Pharm. Sci. 1999, 88, 337–346.
11. Bredikhina, Z. A.; Novikova, V. G.; Zakharychev, D. V.;
Bredikhin, A. A. Tetrahedron: Asymmetry 2006, 17, 3015–
3020.
12. Schaus, S. E.; Brandes, B. D.; Larrow, J. F.; Tokunaga, M.;
Hansen, K. B.; Gould, A. E.; Furrow, M. E.; Jacobsen, E. N.
J. Am. Chem. Soc. 2002, 124, 1307–1315.
column temperature 27 ꢁC; eluent: hexane/isopropanol/
diethylamine = 95:5:0.1; tR = 21.1 min (minor), 22.8 (ma-
jor)]. H NMR (600 MHz) d 1.39 (s, 9H, CH3), 2.47 (t,
J = 6.0, 6.3 Hz, 1H, IH), 3.48 (d, J = 4.7 Hz, 1H, IH),
3.73–3.76 (m, 1H, ICH2), 3.79–3.82 (m, 1H, ICH2), 3.99–
4.02 (m, 1H, CH), 4.04 (dd, J = 9.6, 6.4 Hz, 1H, CH2O),
4.14 (dd, J = 9.6, 3.3 Hz, 1H, CH2O), 6.92–6.97 (m, 2H,
Ar), 7.04–7.07 (m, 2H, Ar).
1
13. Beech, C. L.; Coope, J. F.; Fairley, G.; Gilbert, P. S.; Main,
B. C.; Ple, K. J. Org. Chem. 2001, 66, 2240–2245.
14. Bredikhin, A. A.; Bredikhina, Z. A.; Zakharychev, D. V.;
Pashagin, A. V. Tetrahedron: Asymmetry 2007, 18, 1239–
1244.
15. Ho¨hne, G. W. H.; Cammenga, H. K.; Eysel, W.; Gmelin, E.;
Hemminger, W. Thermochim. Acta 1990, 160, 1–12.
16. Bredikhin, A. A.; Strunskaya, E. I.; Zakharychev, D. V.;
Krivolapov, D. B.; Litvinov, I. A.; Bredikhina, Z. A.
Tetrahedron: Asymmetry 2005, 16, 3361–3366.
17. Wang, Y.; Li, M.; Liu, L.; Zhou, L.; Wang, J. Acta
Crystallogr., Sect. E 2005, 61, o1999–o2000.
Acknowledgments
The authors are indebted to Dr. A. V. Pashagin for chiral
chromatography measurements. The authors thank the
Russian Fund of Basic Research for financial support
(Grant No. 06-03-32508).
References
1. Fogassy, E.; Nogradi, M.; Kozma, D.; Egri, G.; Palovics, E.;
Kiss, V. Org. Biomol. Chem. 2006, 4, 3011–3030.
2. Coquerel, G. Top. Curr. Chem 2007, 269, 1–51.
3. Jacques, J.; Collet, A.; Wilen, S. H. Enantiomers, Racemates
and Resolutions; Krieger Publishing Company: Malabar, FL,
1994, 447 pp.
18. Bredikhin, A. A.; Bredikhina, Z. A.; Lazarev, S. N.; Savel’ev,
D. V. Mendeleev Commun. 2003, 13, 104–105.
19. Nyburg, S. C.; Faerman, C. H. J. Mol. Struct. 1986, 140, 347–
352.
20. Hummel, W.; Burgi, H. B. Helv. Chim. Acta 1988, 71, 1291–
1302.
4. Levkin, P. A.; Torbeev, V. Yu.; Lenev, D. A.; Kostyanovsky,
R. G. Top. Stereochem. 2006, 25, 81–134.
21. Cambridge Structural Database System. Version 5.27. Cam-
bridge Crystallographic Data Centre; 2006.
5. Perez-Garcia, L.; Amabilino, D. V. Chem. Soc. Rev. 2007, 36,
941–967.
22. Klarmann, E.; Gates, L. W.; Shternov, V. A. J. Am. Chem.
Soc. 1932, 54, 1204–1211.
6. Tamura, R.; Takahashi, H.; Fujimoto, D.; Ushio, T. Top.
Curr. Chem. 2007, 269, 53–82.
7. Vlot, M. J.; vanMiltenburg, J. C.; Oonk, H. A. J.; vander-
Eerden, J. P. J. Chem. Phys. 1997, 107, 10102–10111.
8. Gourlay, M. D.; Kendrick, J.; Leusen, F. J. J. Cryst. Grow.
Des. 2007, 7, 56–63.
23. Egri, G.; Kolbert, A.; Balint, J.; Fogassy, E.; Novak, L.;
Poppe, L. Tetrahedron: Asymmetry 1998, 9, 271–283.
24. Kitaori, K.; Furukawa, Y.; Yoshimoto, H.; Otera, J. Tetra-
hedron 1999, 55, 14381–14390.