The tetrahydropyranyl-protected mandelic acid: A novel versatile chiral derivatising agent

Article abstract of DOI:10.1016/S0957-4166(98)00053-6

A simple synthesis of (2R)-2-phenyl-2-[(2S)-tetrahydro-2- pyranyloxy]ethanoic acid, a versatile chiral derivatising agent (CDA), is proposed. The derivatisation of secondary alcohols aimed at the absolute configurational assignment, determination of enantiomeric purity as well as semipreparative resolution is described.

Full text of DOI:10.1016/S0957-4166(98)00053-6

Pergamon
Tetrahedron: Asymmetry 9 (1998) 885–896
The tetrahydropyranyl-protected mandelic acid: a novel versatile
chiral derivatising agent
Omar Parve,a,∗ Madis Aidnik,b Ülo Lille,a Ivar Martin,a Imre Vallikivi,a Lauri Vares a and
Tõnis Pehk c
aInstitute of Chemistry at Tallinn Technical University, Akadeemia tee 15, EE0026 Tallinn, Estonia
bEstonian Agricultural University, Kreutzwaldi 62, EE2400 Tartu, Estonia
cInstitute of Chemical Physics and Biophysics, Akadeemia tee 23, EE0026 Tallinn, Estonia
Received 19 January 1998; accepted 5 February 1998
Abstract
A simple synthesis of (2R)-2-phenyl-2-[(2S)-tetrahydro-2-pyranyloxy]ethanoic acid, a versatile chiral derivatis-
ing agent (CDA), is proposed. The derivatisation of secondary alcohols aimed at the absolute configurational
assignment, determination of enantiomeric purity as well as semipreparative resolution is described. © 1998
Elsevier Science Ltd. All rights reserved.
1. Introduction
The routine tasks encountered in studying chiral compounds are: (1) the absolute configurational
assignment of stereogenic centres; (2) determination of the enantiomeric purity of the substances
obtained; and (3) semipreparative separation of the enantiomers of bioactive compounds (potential drugs,
insect attractants, etc.), all of them to be tested as samples of high enantiomeric excess.1 The goal of our
recent studies has been the development of techniques using a CDA for an integrated approach to these
tasks2 resulting often in the optimal preparation of samples with required enantiomeric excess (up to
gram-scale) together with their stereochemical characterisation.
For analytical purposes α-methoxy-α-(trifluoromethyl)phenylacetic acid (MTPA; Mosher’s acid)3 is
the most preferable CDA for the derivatisation of secondary alcohols and amines because of its resistance
to racemisation. α-Methoxyphenylacetic acid (MPA) as a CDA has been found to satisfy the needs of
all three procedures (despite the more stringent requirements to derivatisation methodology in order to
avoid racemisation)2 being increasingly popular especially for absolute configurational assignment using
∗
Corresponding author. E-mail: omar@boc.ic.ee
0957-4166/98/$19.00 © 1998 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(98)00053-6

886
O. Parve et al. / Tetrahedron: Asymmetry 9 (1998) 885–896
NMR spectroscopy.2,4 In addition, chiral solvating agents offer a useful method for NMR analysis of
chiral compounds without derivatisation.4a
However, in certain cases, the use of these CDAs can be considered disadvantageous because: (1) they
are relatively expensive to be used in preparative applications; (2) differences between the NMR spectra5
of diastereomers tend to be less in the case of O-methylated derivatives of mandelic acid (MA) than for
those having an unprotected hydroxyl group; and finally, (3) the possibility of further derivatisation6 of
the diastereomer prepared is practically ruled out for the derivatives of O-methylated MA.
The use of unprotected MA as a CDA has been documented by only a few examples.7–9 This is
evidently related to the harsh conditions needed for the direct esterification, as well as the problem of
controlling the composition of the product due to the presence of the hydroxyl group alongside the
carboxylic acid moiety in MA. The direct esterification has given satisfactory results in the resolution
of simple chiral secondary alcohols7,8 being evidently unsuitable for the mandoylation of more labile
complex molecules. The shortcomings related to the esterification of acid-sensitive compounds as well
as to the uncontrolled chemoselectivity of the direct mandoylation were overcome by using O-acetyl MA
as a CDA7a (often giving rise to higher NMR chemical shift non-equivalence of diastereomers than the
O-methyl MA). To produce free mandelates the acetyl group has to be selectively removed by thiolysis
which is the drawback of the method.
Treatment of a chiral bicyclic hydroxyketone9 and nonchiral aldehydes10 with an enantiomerically
pure MA under catalysis by a strong acid with simultaneous azeotropic removal of water has been
reported to afford readily separable isomers of the corresponding dioxolanone in a regio- or (and)
stereoselective, respectively, manner along with a minute amount (<5%) of diastereomeric mandelates
in the former case.
The esterification procedure via bissilylation11 of a homochiral MA followed by the selective chlori-
nation of the dimethyl(tert-butyl)silyl ester moiety with an oxalyl chloride–DMFA mixture affording the
acid chloride to be further coupled with a hydroxy compound has been studied by us.12 Unfortunately,
yields of the esters obtained varied due, most probably, to complications connected with the removal
of residues of the highly reactive oxalyl chloride. Attempts to analyse the silylated diastereomeric
mandelates by HPLC failed. Although the corresponding homochiral mandelates (for instance, 10 and
11, Scheme 3) were readily obtained upon the cleavage of silyl ether by a dilute hydrofluoric acid13
followed by chromatography, this approach had to be rejected as being unreliable, not technically simple
or inexpensive enough.
Our further search for a better approach resulted in the finding that tetrahydropyranylation is a
suitable protecting method for MA to be used for derivatisation under mild conditions. The present
study was aimed at: (1) the preparation of an individual diastereomer of tetrahydropyranyl-protected MA
(THPMA); (2) the description of the stereochemistry of the diastereomers of THPMA by using NMR
spectrometric approach combined with computational chemistry methods; and (3) the characterisation
of some applications of THPMA as a CDA to the preparation of diastereomeric THP-mandelates and
mandelates for the absolute configurational assignment and semipreparative resolution of secondary
alcohols, as well as the estimation of their enantiomeric purity.
2. Results and discussion
2.1. Synthesis and the absolute configurational assignment of THPMA
The THPMA methyl esters have been prepared,14,15 starting from (+)-MA, by the tetrahydropyrany-
lation of the enantiomerically pure ester. Diastereomers obtained were separated by chromatography
and the absolute configuration of the THP group of one of the diastereomers was assigned by X-ray
crystallography.16

O. Parve et al. / Tetrahedron: Asymmetry 9 (1998) 885–896
887
Our initial goal was to prepare a diastereomeric mixture of THPMA to be used in the preparative
synthesis of MA esters. (−)-MA 1 was tetrahydropyranylated17 (Scheme 1) to afford a diastereomeric
mixture of THP-esters 2 which was hydrolysed by alkali to sodium carboxylates 3a+4a. The product was
washed with CHCl3 to remove non-acidic contaminants from the alkaline water solution of 3a+4a, which
was further acidified under simultaneous extraction with CHCl3. Washing the CHCl3 extract of 3+4 with
water to remove hydrophilic impurities (MA, etc.) afforded a crude diastereomeric mixture of THPMA
suitable for preparative derivatisation without further purification. No detectable racemisation was found
to occur under the conditions of the synthesis.18
Scheme 1. Tetrahydropyranylation of mandelic acid
Somewhat unexpectedly, one of the acid diastereomers, the THPMA 3, was readily separated (up to
94–98% purity) from the crude product by single crystallisation over 2–3 h (crystallisation during a
prolonged period has given the product of a significantly lower diastereomeric purity in some cases).
The recrystallisation of 3 afforded the target compound in over 99% diastereomeric purity estimated by
the HPLC of the corresponding methyl esters 5 and 6.19 The enantiomeric excess (e.e.) of THPMA 3
was found to be >98%.20 The pure crystalline diastereomer 3 was stable during a prolonged storage at
4°C while the crystalline mixture of 3 and 4 (3:1) decomposed to a significant extent under the same
conditions.
The preliminary absolute configurational assignment of the THP group in THPMA 3 was based on
the statement that on TLC the methyl ester 5 derived21 from the acid 3 was the more polar diastereomer.
The less polar THPMA methyl ester diastereomer was assigned by X-ray analysis to be of a 2S,20S
configuration16 allowing us to assign the THPMA 3 as being the (2R,20S)-THPMA, respectively.
The independent assignment of the absolute configuration of 3 was performed by using computational
chemistry methods combined with NMR studies22 of diastereomeric THPMA methyl esters relying on
the X-ray crystallographic data for the methyl ester.
The conformational analysis of esters 5 and 6 were performed by the AM1 minimisation of energies23
of the numerous conformers generated by the DGEOM95 program.24 The minimum energy geometries
5A and 6B (Scheme 2) obtained from calculations for 5 and 6, respectively, were compared with the
results of NMR analysis (Table 1). The ring-current effects caused by the phenyl group were in good
agreement with the geometries 5A and 6B. This conclusion can best be explained by the difference in the
chemical shifts corresponding to the 20-hydrogen atom of 5 and 6 being 4.58 and 4.90 ppm, respectively.
Examination of the conformers 5A and 6B shows that in the former case the 20H is shielded by the phenyl
ring, while the shielding of 20H is ruled out for 6B. The same is also true for 30-hydrogens: chemical
shift values are 1.71 ppm for 5 and 1.83 ppm for 6. The energetically most favoured conformer 6B of the
2R,20R diastereomer was also found to be very close (a mirror image) to that determined by the X-ray
analysis for the THPMA methyl ester of the 2S,20S configuration.16 The results of NMR measurements
of the above ring-current effect in the case of the parent THPMAs (3) and (4) (Table 1) are in very good
agreement with the data recorded for the methyl esters 5 and 6.
Thus, by comparison of conformers 5A and 6B, evidently being close to those occurring for the

888
O. Parve et al. / Tetrahedron: Asymmetry 9 (1998) 885–896
Scheme 2. The most favoured conformers 5A and 6B calculated for the diastereomeric THPMA methyl esters 5 and 6,
respectively
Table 1
The 13C and 1H chemical shiftsa of THPMAs 3, 4 and THPMA methyl esters 5, 625
corresponding acid diastereomers, we should conclude that the exceptional crystallisability of the
THPMA 3 may be related to the ease of stacking of the acid molecules having the cyclic fragments
parallel as for conformer 5A.
2.2. Derivatisation of bicyclic alcohols
Two bicyclic secondary alcohols, bromohydrin 726 and hydroxy butyrate 14,27,28 were derivatised with
THPMA (3) in order to study diastereomeric esters for NMR spectroscopic as well as chromatographic
behaviour.
The esterification of 7 and 1420 (Scheme 3) with THPMA 3 proceeded readily under the conditions
of the carbodiimide method21 giving quantitative access to the THP-mandelates25 8, 9 and 15, 16,
respectively, when 1.5 eq. of 3 was used. Neither the kinetic differentiation of the optical antipodes of
bromohydrin 7 by the reagent29 nor the detectable racemisation of the THP-mandoyl moiety was found
to occur throughout esterification.20
The ratio of diastereomeric THP-mandelates 8/9 and 15/16 was found to be easily estimated by both
NMR spectroscopy and HPLC.30 The preparative separation of diastereomers 8 and 9 can be performed
by conventional column chromatography over silica, while for the separation of diastereomers 15 and
16 only the HPLC approach can be suggested. The mandelates9,31 10 and 11 obtainable upon acidic

O. Parve et al. / Tetrahedron: Asymmetry 9 (1998) 885–896
889
Scheme 3. Derivatisation of bicyclic alcohols with THPMA 3
deprotection32 of THP-mandelates 8 and 9 have been ketalised under conventional conditions to afford
diastereomers 12 and 13 of good separability on silica. The separation factor, α, of up to 1.55 has been
obtained when using a C6H6/EtOAc mixture as an eluent.31 The diastereomer 12 has been further treated
with alkali33 to afford the corresponding homochiral endo-epoxide, a known synthon for a number of
biologically active compounds.34
NMR spectroscopic studies of diastereomeric THP-mandelates 8, 9 and 15, 16 have proved the validity
of the same conformational model for the absolute configurational assignment by NMR as stated for
the MA esters (in good approximation; see the 13C NMR spectra25 of 10, 11) as well as MPA esters2
of secondary alcohols. Thus, the carbinyl hydrogen, the carbonyl oxygen of the THP-mandoyl moiety
and the oxygen atom attached to the C2 of the THP-mandoyl group can be considered to be coplanar
(Scheme 3) for the preferred conformer in solution. The most significant appearance of the ring-current
1
diamagnetic shielding effect caused by the phenyl group (Scheme 3: indications with arrows) on H
chemical shifts were observed for the 7-endo hydrogen atom of 8 situated close to the phenyl ring (stated
also by the examination of molecular models) and showed therefore an exceptionally large upfield shift:
2.28 ppm for 8 versus 2.92 ppm for 9 (Table 2). For the diastereomers 15 versus 16 the most prominent
upfield shift of 13C and 1H resonances can be observed for C4 and the endo-hydrogen attached to C4.
In conclusion, the above NMR results prove the suitability of THPMA 3 as a CDA for the derivatisation
of bicyclic secondary alcohols aimed at the absolute configurational assignment. It has been stated
that diastereomeric THP-mandelates can readily be analysed by HPLC; some of them are separable by
low-pressure liquid chromatography over silica gel, while the corresponding diastereomeric MA esters
usually show better separability on silica.

890
O. Parve et al. / Tetrahedron: Asymmetry 9 (1998) 885–896
Table 2
The 1H and 13C NMR chemical shifts of the diastereomeric THP-mandelates of alcohols 7 and 14
2.3. Derivatisation of polyol compounds
The derivatisation of cloprostenol 17 with THPMA was aimed at the semipreparative resolution of this
highly potent luteolytic drug35 as well as at the stereochemical studies of the derivatives.
A racemic cloprostenol36 (Scheme 4) was esterified under pTsOH catalysis in methanol, the cor-
responding methyl ester 18 was purified and THP-mandoylated with the diastereomeric mixture of
THPMA.21 The diastereomeric mixture of CDA was used because based on the previous results we
suggested better chromatographic separability of free trismandelates 21/22 than of THP-protected ones.
Thus, the THP groups of 19 and 20 were cleaved on the acidic hydrolysis.32 The resulting cloprostenol
methyl ester trismandelates 21 and 22 were separated by short-column chromatography over silica (TLC:
Rf 0.18 and 0.27, eluent CHCl3:EtOAc=2:1 were determined for 21 and 22, respectively) and further
purified by semipreparative HPLC37 in order to isolate the minute amounts of impurities detected. HPLC
as well as NMR analysis established the resulting trismandelates 21 and 22 to be pure compounds.
The alkaline hydrolysis of esters 21 and 22 afforded (D)-(+)-cloprostenol and (L)-(−)-cloprostenol,
respectively, in satisfactory yield.38 This final assignment of the absolute configuration of these known
products by polarimetric results was in accordance with our results of the absolute configurational
assignment by the NMR spectra of trismandelates 21 and 22 (Table 3; Scheme 4). Regions of the
molecules affected by the ring-current influence of the phenyl groups of the mandoyloxy moieties were
identified.
2.4. Derivatisation of aliphatic alcohols
The derivatisation of aliphatic secondary alcohols with THPMA 3 as a CDA was first checked
on 2-butanol 23 (Scheme 5). The diastereomeric THP-mandelates25 24 and 25 obtained were sep-
arable by HPLC30 showing differences of expected regularity20 between their NMR spectra.39 The
derivatisation20,29 of 1-octyn-3-ol40 26 with THPMA 3 and related chiral moieties were aimed at a

O. Parve et al. / Tetrahedron: Asymmetry 9 (1998) 885–896
891
Scheme 4. Synthesis of diastereomeric cloprostenol methyl ester trismandelates
Table 3
The 1H and 13C NMR chemical shifts of diastereomeric cloprostenol methyl ester trismandelates 21
and 22
comparison of differential shielding effects from THPMA esters25,41 27 and 28, MA esters25,41 29 and
30, MPA esters25,41 35 and 36, MTPA esters4 as well as from THP-mandoylmandelic acid (THPMMA)
esters25,42 31 and 32 and mandoylmandelic acid (MMA) esters25,42 33 and 34.
1
The compilation of differential H and 13C shielding values from different CDAs on 1-octyn-3-ol is
given in Table 4. It should be stressed that a complete picture gives a reliable basis for the assignment of
the absolute configuration and it can generally be obtained by using 1H–1H and 13C–1H 2D FT diagrams
only.4 Data given in Table 4 show that MA, THPMA, THPMMA and MMA all follow the regularities
observed in MPA esters and are opposite to those of MTPA esters. The shielding effects observed for
1
the H and 13C atoms along the octynol chain of the molecule are the greatest for MA esters due
probably to the existence of the energetically most stabilised (by the intramolecular hydrogen bond)
basic conformation. It should also be noted that the greatest differential shielding values are observed for
both 1H and 13C at the β position to the carbinyl carbon for all the esters studied.

892
O. Parve et al. / Tetrahedron: Asymmetry 9 (1998) 885–896
Scheme 5. Derivatisation of aliphatic secondary alcohols with THPMA 3
Table 4
A differential shielding (in ppm) on the 1H and 13C nuclei in the diastereomeric 1-octyn-3-ol esters
The differential shielding effects observed for the THPMMA and MMA esters can be used for the
absolute configurational assignment of a stereogenic centre despite the influence of the phenyl group of
the terminal mandoyl moiety diminishing primary shielding effects.
Thus, it can be concluded that the best possibilities for the absolute configurational assignment are
afforded by using MA esters. THP-mandelates, in turn, are comparable to MTPA esters giving somewhat
worse results than MPA esters.

O. Parve et al. / Tetrahedron: Asymmetry 9 (1998) 885–896
893
3. Conclusions
– A simple method for the preparation of an enantiomerically pure THP-protected mandelic acid 3 has
been proposed.
– THPMA 3 has been shown to be a versatile chiral derivatising agent for secondary alcohols allowing us
to prepare diastereomeric THP-protected mandelates and mandelates suitable for analysis, the absolute
configurational assignment and the semipreparative optical resolution.
– The absolute configurational assignment of chiral secondary alcohols has been discussed in the case
of some THPMA and MA esters. The conformational model of THPMA esters was found to be close
to that of mandelates and O-methyl mandelates.
Acknowledgements
The authors wish to thank the Estonian Science Foundation for financial support (grants No. 2858 and
No. 2293). The authors also express gratitude to Riina Süld for reading and correcting the manuscript.
References
1. Caron, G.; Tseng, G. W.-M.; Kazlauskas, R. J. Tetrahedron: Asymmetry, 1994, 5, 83–92.
2. Trost, B. M.; Belletire, J. L.; Godleski, S.; McDougal, P. G.; Balkovec, J. M.; Baldwin, J. J.; Christy, M. E.; Ponticello, G.
S.; Varga, S. L.; Springer, J. P. J. Org. Chem., 1986, 51, 2370–2374.
3. Dale, J. A.; Dull, D. L.; Mosher, H. S. J. Org. Chem., 1969, 34, 2543–2549.
4. Pehk, T.; Lippmaa, E.; Lopp, M.; Paju, A.; Borer, B. C.; Taylor, R. J. K. Tetrahedron: Asymmetry, 1993, 4, 1527–1532. (a)
Parker, D. Chem. Rev., 1991, 91, 1441–1457.
5. Dale, J. A.; Mosher, H. S. J. Am. Chem. Soc., 1973, 95, 512–519.
6. MMA esters have been described8 as being readily separable from diastereomeric impurities by crystallisation affording a
simple way to diminish (gradually) the influence of the optical impurity of CDA on the enantiomeric excess of the target
alcohol by introduction of two (or more) CDA molecules. See also Refs. 1 and 38.
7. Whitesell, J. K.; Reynolds, D. J. Org. Chem., 1983, 48, 3548–3551. (a) Parker, D. J. Chem. Soc., Perkin Trans. 2, 1983,
83–88.
8. Whitesell, J. K.; Lawrence, R. M. CHIMIA, 1986, 40, 318–321.
9. Parve, O.; Pals, A.; Kadarpik, V.; Lahe, L.; Välimäe. T.; Lopp, M.; Lille, Ü. Proceedings of the Estonian Acad. Sci.
Chemistry, 1989, 38, 139–140. The treatment of the racemic hydroxyketone K with (R)-(−)-MA under catalysis by H2SO4
with the simultaneous azeotropic removal of water gives rapid access to the dioxolanone D along with three more polar
isomers (ratio 3:2:2:1; yield: ∼70%). The dioxolanone D was separated by short-column chromatography over silica
affording an enantiomerically pure (+)-K upon acidic hydrolysis.
Characterisation of D: 13C NMR (C1–7, C1 ,2 ,s ,ortho,meta,para); 39.68, 58.77, 82.65, 32.28, 51.70, 109.37, 41.40, 170.44, 76.24,
133.70, 126.46, 129.01, 129.55; IR; 3500, 3060, 3030, 2970, 2935, 2920, 1780, 1500, 1330, 1300, 1200, 1130, 960, 770,
0
0 0
20
700, 610, 500 cm−1; mp=92ꢀ1°C;
− 52.5 (c 0.7, C6H6); TLC, Rf=0.381 (C6H6:EtOAc:pyridine 100:3:0.5); Rf-s
[α]
546
0.314, 0.266, 0.238 were measured for three other isomeric products under the same conditions.
10. Seebach, D.; Naef, R.; Calderari, G. Tetrahedron, 1984, 40, 1313–1324.
11. Wissner, A.; Grudzinskas, C. V. J. Org. Chem., 1978, 43, 3972–3974.
12. Our unpublished results.

894
O. Parve et al. / Tetrahedron: Asymmetry 9 (1998) 885–896
13. Newton, R. F.; Reynolds, D. P.; Webb, C. F.; Roberts, S. M. J. Chem. Soc., Perkin Trans. 1, 1981, 2055–2058.
14. Mash, E. A.; Arterburn, J. B.; Fryling, J. A. Tetrahedron Lett., 1989, 30, 7145–7148.
15. Mash, E. A.; Arterburn, J. B.; Fryling, J. A.; Mitchell, S. H. J. Org. Chem., 1991, 56, 1088–1093.
16. Mash, E. A.; Fryling, J. A.; Wexler, P. A. Acta Cryst., 1991, C47, 2708–2709.
17. Synthesis of (2R)-2-phenyl-2-[(2S)-tetrahydro-2-pyranyloxy]ethanoic acid: To benzene (100 ml) in a single-necked round-
bottomed flask (R)-(−)-MA (5 g, 32.86 mmol) was added followed by 3,4-dihydro-2H-pyran (9.4 ml, 103 mmol) and
pTsOH (10 mg). MA dissolved over one hour, stirring was continued for one more hour at RT. Et3N was added and the
volatile material was carefully removed under vacuum on a rotary evaporator; the oily residue was dissolved in acetone
(100 ml) and the NaOH solution (40 ml, 1 N) was added. After stirring at RT for two hours acetone was evaporated,
H2O (50 ml) was added and the alkaline solution was extracted with CHCl3 (3×50 ml). To the remaining water solution
CHCl3 (120 ml) was added. The flask was introduced in a water-bath at RT and the NaHSO4 water solution (45 ml, 1
M) was added under vigorous stirring. The organic layer was separated and the water solution was extracted with CHCl3
(2×50 ml). The CHCl3 extracts were combined and washed with H2O (2×50 ml) and the brine. The extract was dried
over anhydrous Na2SO4, filtered, evaporated and the oily residue was crystallised on stirring at +4°C from CCl4/n-hexane
solution (the crude product was dissolved in 8 ml of CCl4 and 12 ml of n-hexane was added). The THPMA (3) (3.4 g) was
obtained in ∼95% diastereomeric purity. The recrystallisation of this product afforded THPMA (3) (2.9 g; yield 37%) with
homogeneity >99% (mp=104°C; C 66.02%, H 6.79%; theor.; C 66.09%, H 6.83%). The dissolution of the product to be
recrystallised should be assisted by a low-boiling, more powerful solvent than CCl4 (acetone, for instance) that could easily
be evaporated prior to addition of n-hexane.
18. A sample of individual THPMA 3 was stirred in the NaOH water solution (2 N) at RT for 48 hrs. The solution was acidified
with NaHSO4 and THPMA was isolated.17 The specific rotation value of the product was established to be close to that of
the starting material. The product was further methylated21 and analysed by HPLC.19 The diastereomeric purity (∼99/1)
was found to have remained practically unchanged during the alkaline treatment proving the occurrence of no detectable
racemisation throughout the above procedures.
19. The HPLC analysis of THPMA methyl esters was performed instead of that of free acids 3 and 4 because of the
complications encountered on looking for HPLC conditions for the analysis of the latter which still remained unreliable
in our hands.
20. The absolute configurational assignment of diastereomers by NMR was verified by the HPLC identification of the
major/minor components of the samples using the standards synthesised starting from the enantiomerically pure (+)-(7)26
and the enantiomerically enriched (+)-(14)27,28 and (−)-(26)40 of known configuration. The individual diastereomer (8) of
>99% diastereomeric purity (by HPLC) was obtained showing the high optical purity of (3) as well as the occurrence of
undetectable racemisation throughout derivatisation. The assignment of 2-butanol esters (24), (25) is based on the NMR
study and it fits nicely the systematic study of the regularities found in the MPA esters of homological secondary aliphatic
alcohols from butanol to decanol.
21. The procedure used for the acylation of alcohols with THPMA (3): synthesis of the THP-mandelates 8 and 9. N,N0-
Dicyclohexylcarbodiimide (464 mg; 2.25 mmol) was introduced to dry THF (4.5 ml) stirred by a magnetic stirrer. THPMA
3 (354 mg; 1.5 mmol) was added followed by bromohydrin 7 (206 mg; 1.0 mmol) and 4-dimethylaminopyridine (12 mg;
0.1 mmol). The mixture was stirred at RT for 16 h. A quantity (30 ml) of EtOAc was added, the N,N0-dicyclohexylurea was
filtered off and the solution was washed with NaHCO3, H2O and brine, dried (Na2SO4), filtered and evaporated to afford
crude THP-mandelates (8) and (9) which were purified over silica (eluent; benzene:EtOAc 98:2). The total yield of esters
was 96.3% (408 mg).
1
22. H and 13C NMR spectra were measured in CDCl3 solution at room temperature on a Bruker AMX500 instrument. 1H–1H
and 1H–13C 2D FT correlation diagrams were used for the full assignment of 1H and 13C chemical shifts.
23. Energy minimisations were carried out using a semi-empirical method: Austin Method 1 (AM 1), M. J.S. Dewar et al. J.
Am. Chem. Soc., 1985, 107, 3902.
24. Conformational space of 5 and 6 was scanned by using the DGEOM95 package (Chiron Corp.©95).
25. Characterisation of compounds. 1: (R)-(−)-mandelic acid, a commercial product (Merck, Germany); 3: for NMR data see
Table 1; IR (KBr) 3440, 3040, 2950, 2870, 1730, 1640, 1500, 1460, 1420, 1400, 1360, 1300, 1270, 1210, 1180, 1140,
20
1120, 1080, 1060, 1030, 1020, 980, 950, 910, 870, 810, 800, 780, 720, 700, 680, 580 cm−1; mp=104°C;
− 254
[α]
546
(c 1.0, CHCl3); 4: for NMR data see Table 1; 5/6: for NMR data see Table 1, see also Ref. 15; TLC Rf=0.226/0.283
20
[α]
(C6H6:EtOAc=50:1);
− 280/ + 56 (c 0.3, CH3OH); 8/9: for NMR data see Table 2; TLC Rf=0.116/0.134
546
(C6H6:EtOAc=100:2); 10/11: 13C NMR (C1–7; mandoyl) 39.00/39.22, 53.58/53.74, 84.35/84.48, 33.70/33.81, 63.71/63.72,
209.40/209.12, 52.35/52.77, 172.21/171.79, 73.13/73.28, 137.53/137.27, 127.00/126.31, 128.93/128.75, 129.00/128.81; IR
3540/3550, 2930/2940, 3070, 3040, 1790, 1740 cm−1; TLC Rf=0.192/0.163 (C6H6:EtOAc=10:1); 12/13: IR 3450/3460,

O. Parve et al. / Tetrahedron: Asymmetry 9 (1998) 885–896
895
Table 5
The 1H and 13C NMR spectra of diastereomeric THPMA 3 esters of 2-butanol
2960, 2890, 1740, 1460, 1425, 1300, 1180, 1100, 1070, 700 cm−1; TLC Rf=0.333/0.240 (C6H6:acetone=10:1); 14: Ref. 28;
15/16: for NMR data see Table 2; TLC Rf=0.243 (C6H6:EtOAc=100:2); 21/22: for NMR data see Table 3; TLC Rf=0.18/0.27
(CHCl3:EtOAc=2:1); 24/25: for NMR data see Ref. 39; TLC Rf=0.202 (C6H6); 26: Ref. 40; 27/28: for NMR data see Ref.
41; TLC Rf=0.298 (C6H6); 29/30: for NMR data see Ref. 41; TLC Rf=0.157 (C6H6:EtOAc=100:2); 31/32: for NMR data
see Ref. 42; TLC Rf=0.200 (C6H6); 33/34: for NMR data see Ref. 42; TLC Rf=0.226/0.245 (C6H6:n-hexane:EtOAc 5:5:1);
35/36: for NMR data see Ref. 41
26. Newton, R. F.; Paton, J.; Reynolds, D. P.; Young, S.; Roberts, S. M. J. Chem. Soc., Chem. Commun., 1979, 908–909.
27. Parve, O.; Vallikivi, I.; Lahe, L.; Metsala, A.; Lille, Ü.; Tõugu, V.; Vija, H.; Pehk, T. Bioorg. Medicinal Chem. Lett., 1997,
7, 811–816.
28. Parve, O.; Pals, A.; Kadarpik, V.; Lahe, L.; Lille, Ü.; Sikk, P.; Lõokene, A.; Välimäe T. Bioorg. Medicinal Chem. Lett.,
1993, 3, 359–362.
29. A sample of (ꢀ)-7 was acylated with 0.3 eq. of THPMA 3.21 The diastereomers 8, 9 were obtained in the ratio of 53/47 by
HPLC, while the remaining 7 had an specific rotation close to zero. Contrary to this, the above trial for (ꢀ)-26 proved the
occurrence of some kinetic differentiation that was not studied further because the quantitative conversion of the starting
alcohol took place in all derivatisations with 3.
30. HPLC analysis: column: 3×300 mm Separon SGX (5 µm). Detector: UV-spectrophotometer (λ=254 nm). Eluent: n-
hexane:EtOAc (95:5) for 5, 6; 15, 16; 24, 25; 27, 28; n-hexane:EtOAc (90:10) for 8, 9; n-hexane:2-propanol (98:2) for
10, 11; 12, 13; 29, 30; 33, 34; n-hexane:2-propanol:H2O (80:19:1) for 17 and CHCl3:EtOH (99:1) for 21, 22.
31. The chromatographic separation of the diastereomers 10/11 and 12/13 has been studied. See: Lõhmus, M.; Parve, O.;
Kanger, T.; Lopp, M.; Lille, Ü. Resolution of diastereomers as a key step in obtaining optically pure prostanoids In:
Abstracts. The 13th Symp. of Column Liquid Chromatography. Stockholm, 1989, 47.
32. THP groups were cleaved by the treatment of a sample in CH3CN:0.2 N H2SO4 (3:1) for 16 h.
33. The alkaline hydrolysis was performed stirring the sample during 16 h in a MeOH:1 N NaOH (3:1) mixture.
34. Newton, R. F. New Synthetic Routes to Prostaglandins and Thromboxanes; Roberts, S. M. and Sheinmann, F., Eds;
Academic Press: London, 1982, pp. 61–104.
35. MARTINDALE, The Extra Pharmacopoeia. Thirtieth Edition. Reynolds, J. E. F., Ed.; The Pharmaceutical Press: London,
1993, p. 1150.
36. The sample of (ꢀ)-cloprostenol was kindly provided by Kemasol Ltd, Tallinn.
37. Semipreparative HPLC: column: 30×300 mm Separon SGX (7µm). Detector: UV-spectrophotometer (λ=254 nm). Eluent:
CHCl3:EtOH (99:1) for (21), (22) and CHCl3:EtOH:H2O (90:9.6:0.4) for 17.
20
546
38. The alkaline hydrolysis33 of 21 and 22 afforded (D)-(+)-cloprostenol (32 mg;
+ 27 (c 1.0, CHCl3)) and (L)-
[α]
20
[α]
(−)-cloprostenol (35 mg;
− 27 (c 1.0, CHCl3)), respectively, after semipreparative HPLC purification37. The
546
optical resolution procedure was started from 100 mg of (ꢀ)-17: total yield 67%. The present process for the optical
resolution of cloprostenol is supposed to allow a very high control of the enantiomeric purity of the product: the optical
purity of the CDA (e.e.) was ≥98% and all minor diastereomers were separated by semipreparative HPLC (as concluded
by HPLC and NMR analysis). Then the content of the enantiomeric contamination could be calculated as follows:
0.01×0.01×0.01×100%=0.0001%. The value calculated characterises the expected enantiomeric excess of the target
enantiomeric cloprostenols to be as high as 99.9998%. This consideration should be proved by further analysis using proper
methodologies. Compounds of a very high enantiomeric purity can also be obtained by using the repetition of any of the
optical resolution procedure (see also Ref. 6).
39. See Table 5.

896
O. Parve et al. / Tetrahedron: Asymmetry 9 (1998) 885–896
Table 6
The 1H and 13C NMR chemical shifts of diastereomeric mandoyl derivatives of 1-octyn-3-ol
Table 7
The 1H and 13C NMR spectra of 1-octyn-3-ol THPMMA and MMA esters
40. Hashimoto, S.; Kose, S.; Suzuki, A.; Yanagiya, Y.; Ikegami, S. Synth. Commun., 1991, 21, 833–839.
41. See Table 6.
42. See Table 7.