Chiral solvating agents for cyanohydrins and carboxylic acids

Article abstract of DOI:10.1021/jo100445d

We have shown that a structure as simple as an ion pair of (R)- or (S)-mandelate and dimethylamminopyridinium ions possesses structural features that are sufficient for NMR enantiodiscrimination of cyanohydrins. Moreover, ¹H NMR data of cyanohydrins of known configuration obtained in the presence of the mandelate-dimethylaminopyridinium ion pair point to the existence of a correlation between chemical shifts and absolute configuration of cyanohydrins. Mandelate-DMAPH⁺ ion pair and mandelonitrile form a 1:1 complex with an association constant of 338 M^-1 (ΔG ⁰, -3.4 kcal/mol) for the (R)-mandelonitrile/(R)-mandelate-DMAPH ⁺ and 139 M^-1 (ΔG⁰, -2.9 kcal/mol) for the (R)-mandelonitrile/(S)-mandelate-DMAPH⁺ complex. To understand the origin of enantiodiscrimination, the geometry optimization and energy minimization of the models of ternary complexes of (S)-mandelonitrile/(R)- mandelate/DMAPH⁺ and (S)-mandelonitrile/(S)-mandelate/DMAPH ⁺ complexes was performed using DFT methodology (B3LYP) with the 6-31+G(d) basis set in Gaussian 3.0. Further, analysis of optimized molecular model obtained from theoretical studies suggested that (i) DMAP may be replaced with other amines, (ii) the hydroxyl group of mandelic acid is not necessary for stabilization of ternary complex and may be replaced with other groups such as methyl, (iii) the ion pair should form a stable ternary complex with any hydrogen-bond donor, provided its OH bond is sufficiently polarized, and (iv) α-H of racemic mandelic acid should also get resolved with optically pure mandelonitrile. These inferences were experimentally verified, which not only validated the proposed model but also led to development of a new chiral solvating agent for determination of ee of carboxylic acids and absolute configuration of aryl but not alkyl carboxylic acids.

Full text of DOI:10.1021/jo100445d

pubs.acs.org/joc
Chiral Solvating Agents for Cyanohydrins and Carboxylic Acids†
Lomary S. Moon,‡, Mohan Pal,‡ Yoganjaneyulu Kasetti,§ Prasad V. Bharatam,§ and
Ravinder S. Jolly*,‡
‡Department of Chemistry, Institute of Microbial Technology (Council of Scientific and Industrial Research),
Sector 39, Chandigarh 160 036, India, and §Department of Medicinal Chemistry, National Institute of
Pharmaceutical Education and Research, Sector 67, SAS Nagar 160062, Punjab, India. Present address:
Biologics Development Center, Dr. Reddy’s Laboratories, Survey No. 47, Bachupally, Qutubullapur,
R. R. District 500090, Andhara Pradesh, India
jolly@imtech.res.in
Received March 10, 2010
We have shown that a structure as simple as an ion pair of (R)- or (S)-mandelate and dimethylam-
minopyridinium ions possesses structural features that are sufficient for NMR enantiodiscrimination
of cyanohydrins. Moreover, 1H NMR data of cyanohydrins of known configuration obtained in the
presence of the mandelate-dimethylaminopyridinium ion pair point to the existence of a correlation
between chemical shifts and absolute configuration of cyanohydrins. Mandelate-DMAPHþ ion pair
and mandelonitrile form a 1:1 complex with an association constant of 338 M-1 (ΔG0, -3.4 kcal/mol)
for the (R)-mandelonitrile/(R)-mandelate-DMAPHþ and 139 M-1 (ΔG0, -2.9 kcal/mol) for the
(R)-mandelonitrile/(S)-mandelate-DMAPHþ complex. To understand the origin of enantiodiscrimi-
nation, the geometry optimization and energy minimization of the models of ternary complexes of
(S)-mandelonitrile/(R)-mandelate/DMAPHþ and (S)-mandelonitrile/(S)-mandelate/DMAPHþ com-
plexes was performed using DFT methodology (B3LYP) with the 6-31þG(d) basis set in Gaussian 3.0.
Further, analysis of optimized molecular model obtained from theoretical studies suggested that (i)
DMAP may be replaced with other amines, (ii) the hydroxyl group of mandelic acid is not necessary for
stabilization of ternary complex and may be replaced with other groups such as methyl, (iii) the ion pair
should form a stable ternary complex with any hydrogen-bond donor, provided its OH bond is
sufficiently polarized, and (iv) R-H of racemic mandelic acid should also get resolved with optically pure
mandelonitrile. These inferences were experimentally verified, which not only validated the proposed
model but also led to development of a new chiral solvating agent for determination of ee of carboxylic
acids and absolute configuration of aryl but not alkyl carboxylic acids.
Introduction
enantiomers by NMR.1-5 While rigid structures with either
built-in cages6-12 or with strong conformational bias12-16
The ability of a host molecule to form geometrically
different diastereomeric complexes with antipods of a chiral
compound has been exploited for the discrimination of
(2) Seco, J. M.; Quinoa, E.; Riguera, R. Tetrahedron: Asymmetry 2001,
12, 2915–2925.
(3) Seco, J. M.; Quinoa, E.; Riguera, R. Chem. Rev. 2004, 104, 17–117.
(4) Wenzel, T. J. Discrimination of Chiral Compounds Using NMR
Spectroscopy; John Wiley & Sons: New York, 2007.
(5) Wenzel, T. J.; Wilcox, J. D. Chirality 2003, 15, 256–270.
(6) Dignam, C. F.; Richards, C. J.; Zopf, J. J.; Wacker, L. S.; Wenzel, T. J.
Org. Lett. 2005, 7, 1773–1776.
† A part of this work has been published as preliminary report (Moon et al.
Chem. Commun. 2009, 1067-1069); reproduced by permission of the Royal
Society of Chemistry.
(1) Rothchild, R. Enantiomer 2000, 5, 457–471.
DOI: 10.1021/jo100445d
r
Published on Web 07/22/2010
J. Org. Chem. 2010, 75, 5487–5498 5487
2010 American Chemical Society

Moon et al.
JOCArticle
have been among the most used hosts for NMR enantiodis-
crimination, structurally flexible simpler structures17-19
have also been shown to discriminate enantiomers in NMR
with similar resolutions. Recently, we have shown that a
much simpler structure, an ion pair of (R) or (S)-mandelate
and dimethylamminopyridinium ion, possesses structural
features which are sufficient for NMR enantiodiscrimina-
tion of cyanohydrins.20 Moreover, 1H NMR data of cyano-
hydrins of known configuration obtained in the presence of
the mandelate-dimethylaminopyridinium ion pair pointed
to the existence of a correlation between chemical shifts and
absolute configuration of cyanohydrins. Here, we describe
(a) these results in more detail and (b) significant extension of
work leading to development of new chiral solvating agents
for carboxylic acids. The mandelic acid-DMAP combina-
tion reagent has never been reported before, although use of
O-acetyl- and O-methylmandelic acid as derivatizing agent
for determination of ee of secondary alcohols has been
described.21
The asymmetric cyanation of aldehydes and ketones to
produce cyanohydrins is a highly versatile synthetic trans-
formation. Homochiral cyanohydrins are of synthetic inter-
est as they may be transformed into a number of key
functional groups, such as R-hydroxy acids, primary and
secondary β-hydroxyamines, R-aminonitriles, R-hydroxy
ketones, R-hydroxy esters, etc., under conditions that con-
serve optical purity.22 Many of these intermediates can be
used in further stereoselective transformations. Recent ad-
vances in the field of chemical and enzymatic catalyst for
asymmetric cyanohydrin synthesis are set to revolutionize
the use of chiral cyanohydrins in organic synthesis.22-27 This
has resulted in an increased surge of interest in developing
new simple, rapid, and steadfast new methods for ascertain-
ing ee and absolute configuration of cyanohydrins.
Assigning the absolute configuration of cyanohydrins is
not just an extension of the procedure for secondary or
tertiary alcohols because of the complicating effects of cyano
group.28 Even though cyanohydrins appears to have some
resemblance with the structure of secondary or tertiary
alcohols, the presence of the strongly polar -CN substituent
makes the geminal hydroxynitrile moiety a wholly novel
situation from the structural point of view to which the
NMR procedures previously described for secondary alco-
hols cannot be applied without a previous and rigorous
validation.
Although there were few examples of the use of chiral
derivatizing agents and solvating agents for determination of
enantiomeric excess of cyanohydrins, there was no thorough
study available for enantiodiscrimination of cyanohydrins
that can be used for assigning absolute configuration until
2006, when Louzao et al. established the first derivative-
based method for aldocyanohydrins,28 which was later
extended to include ketocyanohydrins as well.29 In these
reports, rigorous validation on MPA esters of cyanohydrins
was done for determining absolute configuration. The method
suffers from drawbacks typical of a derivatization method,
viz. chances of resolution and racemization during deriva-
tization and difficulties in recovering cyanohydrin after the
analysis.
A recent interesting study of binary mixtures of pyridine
and various carboxylic acids using noisy light-based coher-
ent anti-Stokes Raman scattering (I(2)CARS) has shown that
in solution (i) acid-base reaction produces pyridinium ca-
tion and carboxylate anion in highly product favored reac-
tion, (ii) pyridinium and acetate ions exist as ion pairs, and
(iii) the propensity to exist as a pyridinium-carboxylate ion
pair depends on the pKa of acid and remains unaffected by
the steric bulkiness of the carboxylic acid.30 Crystal struc-
tures of several pyridine/carboxylic acid cocrystalline sys-
tems have shown that pyridinium cations interact with
(7) Ema, T.; Ouchi, N.; Doi, T.; Korenaga, T.; Sakai, T. Org. Lett. 2005,
7, 3985–3985.
(8) Ema, T.; Tanida, D.; Sakai, T. Org. Lett. 2006, 8, 3773–3775.
(9) Ema, T.; Tanida, D.; Sakai, T. J. Am. Chem. Soc. 2007, 129, 10591–
10596.
(10) Ma, F.; Ai, L.; Shen, X.; Zhang, C. Org. Lett. 2007, 9, 125–127.
(11) Uccello-Barretta; Balzano, G.; Martinelli, F.; Berni, J.; Villani,
M.-G.; Gasparrini, C.; Francesco Tetrahedron: Asymmetry 2005, 16, 3746–3751.
(12) Yang, D.; Li, X.; Fan, Y. F.; Zhang, D. W. J. Am. Chem. Soc. 2005,
127, 7996–7997.
(13) Luo, Z.; Li, B.; Fang, X.; Hu, K.; Wu, X.; Fu, E. Tetrahedron Lett.
2007, 48, 1753–1756.
(14) Chin, J.; Kim, D. C.; Kim, H. J.; Panosyan, F. B.; Kim, K. M. Org.
Lett. 2004, 6, 2591–2593.
(15) Atwood, J. L.; Szumna, A. J. Am. Chem. Soc. 2002, 124, 10646–
10647.
(16) Rudkevich, D. M.; Hilmersson, G.; Rebek, J. J. Am. Chem. Soc.
1998, 120, 12216–12225.
anions through a moderate to strong NH O bond.31 Since
3 3 3
the second oxygen of carboxylate in the ion pair is available
as a H-bond acceptor, a ternary complex with a H-bond
donor becomes feasible.
The pKa values of mandelic acid and acetic acid are not
very different from each other; therefore, mandelic acid/
DMAP mixture should exist primarily as an ion pair,
mandelate-DMAPHþ, in solution. Due to the powerful
electron-withdrawing effect of the cyano group, the O-H
bond of the hydroxyl group of cyanohydrins is sufficiently
polarized to be able to form a ternary complex with the
ꢀ
(17) Hernandez-Rodrı
7678.
(18) Cuevas, F.; Ballester, P.; Pericas, M. A. Org. Lett. 2005, 7, 5485–
5487.
(19) Garric, J.; Leger, J.-M.; Huc, I. Angew. Chem., Int. Ed. 2005, 44,
1954–1958.
´
guez, M.; Juaristi, E. Tetrahedron 2007, 63, 7673–
ꢀ
(20) Moon, L. S.; Jolly, R. S.; Kasetti, Y.; Bharatam, P. V. Chem.
Commun. 2009, 1067–1069.
(21) (a) Parker, D.; Taylor, R. J. Tetrahedron 1987, 43, 5451–5456.
(b) Benson, S. C.; Cai, P.; Colon, M.; Haiza, M. A.; Tokles, M.; Snyder,
J. K. J. Org. Chem. 1988, 53, 5335–5341. (c) Parker, D. Chem. Rev. 1991, 91,
1441–1457. (d) 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.
mandelate-DMAPHþ ion pair through the OH O bond.
3 3 3
On the basis of the foregoing arguments, we purposed a
(22) Brunel, J. M.; Holmes, I. P. Angew. Chem., Int. Ed. 2004, 43, 2752–
2778.
(27) Klempier, N.; Griengl, H.; Hayn, M. Tetrahedron Lett. 1993, 34,
4769–4772.
(23) (a) Effenberger, F. Angew. Chem., Int. Ed. 1994, 33, 1555–1564. (b)
Brussee, J.; Loos, W. T.; Kruse, C. G.; Van Der Gen, A. Tetrahedron 1990,
46, 979–986.
(24) Gregory, R. J. H. Chem. Rev. 1999, 99, 3649–3682.
(25) Huuhtanen, T. T.; Kanerva, L. T. Tetrahedron: Asymmetry 1992, 3,
1223–1226.
(28) Louzao, I.; Seco, J. M.; Quinoa, E.; Riguera, R. Chem. Commun.
2006, 13, 1422–1424.
(29) Louzao, I.; Garcia, R.; Seco, J. M.; Quinoa, E.; Riguera, R. Org.
Lett. 2009, 11, 53–56.
(30) Berg, E. R.; Green, D. D.; Moliva, A., D. C.; Bjerke, B. T.; Gealy,
M. W.; Ulness, D. J. J. Phys. Chem. A 2008, 112, 833–838.
(31) Jin, Z. M.; Hu, M. L.; Li, Z. G.; Xuan, R. C.; Yu, K. B. J. Chem.
Crystallogr. 2004, 34, 657–660.
(26) Kanerva, L. T.; Kiljunen, E.; Huuhtanen, T. T. Tetrahedron: Asym-
metry 1993, 4, 2355–2361.
5488 J. Org. Chem. Vol. 75, No. 16, 2010

Moon et al.
JOCArticle
NMR enantiodiscrimination of a variety of racemic cya-
nohydrins derived from aliphatic and aromatic aldehydes
and ketones with (S)-mandelate-dimethylaminopyridinium
ion pair was studied (Figure 3). Racemic cyanohydrins were
synthesized by reaction of the corresponding aldehyde or
ketone with potassium cyanide in the presence of sodium
metabisulfite.33 Excellent baseline separation occurred in all
cases with ΔΔδ in the range of 0.055-0.112 ppm for all
aldehyde cyanohydrins studied. The baseline separation
occurred, even when the methine proton appeared as a
doublet, triplet, or quartet (cyanohydrins 9-14, Figure 3).
Excellent baseline separation coupled with the fact that
resolved singlets for aldehyde cyanohydrins appeared in
FIGURE 1. Hypothetical model showing ternary complexes of (R)-
cyanohydrin and (S)-cyanohydrin with (S)-mandelate-DMAPHþ
ion pair. The H in complex A and the R group in complex B of
cyanohydrin face the phenyl group of mandelate.
1
the least complex region of H NMR suggested that the
mandelic acid/DMAP mixture can be used for the accurate
determination of the optical purity of these compounds.
The ΔΔδ for ketone cyanohydrins was of the order of
0.010-0.031 ppm (cyanohydrins 15-19, Figure 3). The
magnitude of resolution for ketone cyanohydrins was less
compared to aldehyde cyanohydrins but sufficient for the
determination of optical purity of these compounds. There
was some improvement in resolution in all examples except
pyridin-4-yl ketone when NMRs were recorded in C6D6
instead of CDCl3. No resolution occurred when DMSO-d6
was used as solvent instead of CDCl3.
Determination of Absolute Configuration of Cyanohydrins.
Having demonstrated enantiodiscrimination of cyanohy-
drins, our next goal was to investigate the suitability of the
method for the determination of their absolute configura-
tion. The NMR spectral behavior of a series of cyanohydrins
of known absolute configuration was studied using DMAP
in combination with (R)- or (S)-mandelic acid to find the
existence of any correlation between the absolute con-
figuration and the NMR chemical shifts. Optically active
cyanohydrins were synthesized using commercially available
Me-HNL, (S)-selective hydroxynitrile lyase from Manihot
esculenta and Pa-HNL, and (R)-selective hydroxynitrile
lyase from Prunus amygdalus.
ΔδRS values, obtained with (R)- and (S)-mandelic acid,
respectively, for several aldo- and ketocyanohydrins of known
configuration are shown in Table 1. Aldehyde cyanohydrins
20-23, 26, 30, and 31, which have the same spatial relation-
ship, showed a positive ΔδRS value, whereas cyanohydrins 24,
25, and 27-29 with opposite spatial arrangement showed
negative ΔδRS values. Similarly, ketone cyanohydrins 32, 34,
35, and 37, which are configurationally related to 1, showed
positive ΔδRS values, whereas cyanohydrins, 33, 36, and 38,
which are configurationally related to 24, showed negative
ΔδRS values. Moreover, the enantiomers of 20, 21, 23, 30, 32,
and 34, which are configurationally related to 24, exhibited
negative ΔδRS values. Similarly, the enantiomers of 25 and
27, which are configurationally related to 20, showed positive
ΔδRS value. Thus, the ΔδRS sign is characteristic for this
enantiomeric series and can be used for the assignment of
absolute configuration. However, the ΔδRS value for pyridine
compound 26 was poor compared to other aldehyde cyanohy-
drin series. Since basic nitrogen on pyridine can alter the nature
of complex formation, assignment of absolute configuration in
the pyridine series may not be reliable.
hypothetical model for ternary complex of mandelonitrile with
mandelate-DMAPHþ ion pair (Figure 1). In the hypothetical
model of the ternary complex, the cylindrically compact,
electronegative cyano group has been shown to prefer a
position in space which is farthest from carboxylate but closer
to the dimethylamino group of DMAP, where the Lewis basic
nitrogen of cyano unit can interact with the partially positive
nitrogen of amino group. Distortion of the aromatic character
and stabilization of imino form has been demonstrated recently
in a dimethylaminopyridinium salt.32
Results and Discussion
To test this hypothesis, we recorded 1H NMR of racemic
mandelonitrile (30 mM) in the presence of 1 molar equiv of
(S)-mandelic acid and DMAP in CDCl3. We were pleased to
note that R-H of two enantiomers of racemic mandelonitrile
appeared as well-resolved singlets (ΔΔδ = 0.084 ppm) in
1
approximate 1:1 ratio based on integral values. H NMR
data was collected on 300 MHz spectrometer. Chemical
shifts (ppm) are internally referenced to TMS signal (0 ppm).
R-H of racemic mandelonitrile appeared as singlet at δ 5.532
(Figure 2a), whereas in the presence of 1 molar equiv of
mandelate-DMAPHþ ion pair, its two enantiomers suf-
fered an unequal upfield shift and appeared as two singlets at
δ 5.435 and 5.351(Figure 2 b). DMAP or mandelic acid alone
caused no shift in resonance of R-H proton. Although no
covalent bond formation was expected, it was confirmed by
recovery of mandelonitrile from samples after recording of
NMR. Thus, after removal of CDCl3, sample was dissolved
in diethyl ether, which was sequentially washed with 2%
sodium bicarbonate, dilute hydrochloric acid, and brine
solution. Removal of solvent under reduced pressure on a
rotary evaporator resulted in >90% recovery of mandelo-
nitrile. Under similar conditions, R-H of (R)- and (S)- man-
delonitrile appeared as singlets at δ 5.436 (Figure 2 c) and δ
5.353 (Figure 2d), respectively. These results clearly show
that mandelate-DMAPHþ ion pair is an effective chiral
solvating agent for mandelonitrile. Since baseline separa-
tion of R-H of two enantiomers of racemate has occurred,
ee of mandelonitrile can be obtained from integral values of
R-H.
(32) Koleva, B. B.; Kolev, T.; Seidel, R. W.; Tsanev, T.; Mayer-Figge, H.;
Spiteller, M.; Sheldrick, W. S. Spectrochim. Acta Part A: Mol. Biomol.
Spectrosc. 2008, 71, 695–702.
(33) Vogel, A. I.Vogel’sTextbook ofPracticalOrganicChemistry; Longman
Scientific & Technical: 1989; Vol. IV.
J. Org. Chem. Vol. 75, No. 16, 2010 5489

Moon et al.
JOCArticle
FIGURE 2. NMR enantiodiscrimination of mandelonitrile with (S)-mandelate-DMAPHþ ion pair. Partial 1H NMR spectra (300 MHz) of
(a) (R,S)-mandelonitrile in CDCl3 and (b-d) (R,S)-, (R)-, and (S)-mandelonitrile, respectively, in the presence of 1 molar equiv of (S)-mandelic
acid and DMAP.
FIGURE 3. NMR enantiodiscrimination of rac-cyanohydrins in the presence of (S)-mandelate-DMAPHþ ion pair. ΔΔδ values for R-H or
R-CH3 are shown in parentheses.
Characterization of the Ternary Complex between Cyanohy-
drin and Mandelate-DMAPHþ Ion Pair. An NMR method as
described previously was used to determine the stoichiometry
of the complex.34,35 The 1H NMR spectra of (S)-man-
delate-DMAPHþ ion pair with (R)-mandelonitrile in various
ratios in CDCl3, keeping the total concentration constant at 40
μM, were recorded. It was found that the R-H of mandelonitrile
underwent a variable upfield shift depending upon the ratio of
mandelonitrile and ion pair. The Jobs plot of ΔδXi (the product
of the chemical shift change and the mole fraction) versus the
mole fraction (Xi) of (R)-mandelonitrile in the mixture was
obtained from these values (Figure 4). A maxima was observed
when the ratio of (S)-mandelate-DMAPHþ ion pair versus
(R)-mandelonitrile was 1:1 (Xi = 0.5), which indicated 1:1
complex formation between (R)-mandelonitrile and (S)-man-
delate-DMAPHþ ion pair.
Stability constants can be determined by NMR spectro-
metry when the species are in rapid exchange on the NMR
time scale and when there is a variation in the chemical shift
of a suitable nucleus on formation of the complex species.36
(34) Job, P. C. R. Hebd. Seances Acad. Sci. 1925, 180, 928–930.
(35) Blanda, M. T.; Horner, J. H.; Newcomb, M. J. Org. Chem. 1989, 54,
4626–4636.
(36) Hynes, M. J. J. Chem. Soc., Dalton Trans. 1993, 311–312.
5490 J. Org. Chem. Vol. 75, No. 16, 2010

Moon et al.
JOCArticle
TABLE 1. Correlation between Chemicals Shifts and Absolute Configuration of Cyanohydrinsa
aΔδRS (ΔδR - ΔδS) values for R-H of 20-31, R-CH3 of 32-36 and 7-endo H of 37 and 38 are shown. Opposite ΔδRS sign for enantiomers of 20-22, 25,
27, 30, 32, and 34 was obtained.
A solution of mandelate-DMAPHþ ion pair in CDCl3 was
placed in 19 5 mm NMR tubes. A predetermined quantity of
a concentrated solution of mandelonitrile in CDCl3 was
added to each of 18 tubes so that finally solutions with
desired relative amounts (equiv) of the mandelonitrile versus
ion pair were obtained. The concentration of ion pair was
always maintained at 20 mM. Volume and concentration
changes were taken into account during analysis. Plots of
concentration versus chemical shift are shown in Figure 5.
An association constant (Ka) of 338 M-1 (ΔG0, -3.4 kcal/mol)
for (R)-mandelonitrile/(R)-mandelate-DMAPHþ complex
and 139 M-1 (ΔG0, -2.9 kcal/mol) for (R)-mandelonitrile/
(S)-mandelate-DMAPHþ complex was obtained by non-
using the WinEQNMR program.36 Values of ΔG0 were calcu-
lated using the equation ΔG0 = -RT ln Ka.
Proposed Model. Modeling studies were performed to un-
derstand the origin of enantiodiscrimination. Justification pro-
vided by our previous report20 has been taken as base for this
study. Although the molecular mechanics optimized models
obtained in our previous study were able to explain enantio-
discrimination, they showed large differences in the calculated
and experimentally observed δ and ΔδRS values (for example,
(S)-lactonitrile showed a ΔδRS deviation of 1.92 ppm), when
subjected to computational estimation of chemical shifts in
implicit chloroform medium (IEFPCM method)37 using the
1
ꢁ
(37) Cances, M. T.; Mennucci, B.; Tomasi, J. J.Chem.Phys. 1997, 107, 3032.
linear least-squares fitting for the H NMR titration curve
J. Org. Chem. Vol. 75, No. 16, 2010 5491

Moon et al.
JOCArticle
FIGURE 4. Jobs plot for the complexation of mandelate-
DMAPHþ ion pair with (R)-mandelonitrile. Δδ is the shift (ppm)
for R-H of mandelonitrile. Maxima at 0.5 mol fraction indicated 1:1
complexation.
GIAO algorithm38 and B3LYP/6-311þG(2d,p) method, there-
by prompting us to reinvestigate the modeling studies. The
modeling studies were performed with Gaussian03 software.39
Initial geometry optimization on the ternary complexes was
carried out using molecular mechanics method in Chem3D
software and a few models of the ternary complex were chosen
for optimization using the density functional theory (DFT)
method with the B3LYP/6-31þG(d) basis set.40 All of the
optimization studies were carried out in the gas phase. The
optimized models were employed to computationally estimate
chemical shifts in implicit chloroform medium (IEFPCM
method)37 using the GIAO algorithm38 and B3LYP/6-311þG-
(2d,p) method. As NMR experimental studies have been
carried out in chloroform solvent, GIAO calculations were
performed in chloroform solvent for reliable comparison.
Calculated chemical shifts for R-H of nitrile are reported as
ppm from the value calculated for TMS after conversion from
shielding values.
FIGURE 5. 1H NMR titration data obtained for the complexation
of (a) (R)-mandelate-DMAPHþ ion pair as host and (R)-mande-
lonitrile as guest; (b) (S)-mandelate-DMAPHþ ion pair as host and
(R)-mandelonitrile as guest. Ka was obtained by nonlinear least-
squares fitting for the 1H NMR titration curve using WinEQNMR
program.36 ΔG0 was calculated using equation ΔG0 = -RT ln Ka.
anisotropic shielding due to the phenyl ring of mandelate in
(S)-nitrile/(R)-mandelate-DMAPHþ complexes, no such
shielding is experienced by the R-H of nitrile in (S)-nitrile/
(S)-mandelate-DMAPHþ complexes. Thus, the ΔδRS value
for (S)-mandelonitrile should be negative, which is consis-
tent with the experimental observation. The calculated ΔδRS
sign in NMR based on models in Figure 6 was in agreement
with the experimentally observed results (Table 2). The cal-
culated δ values for R-H of nitrile for all models in Figure 6
were within (0.28-0.73 ppm compared to the experimen-
tally observed values. Thus, the δ values for R-H of nitrile
calculated based on models in Figure 6 were in good agree-
ment with the experimentally observed values. According
to models shown in Figure 6, the aryl ring of DMAP is
not important for shielding. This observation was supported
by the fact that the substitution of DMAP with aliphatic
triethylamine or N-ethyl-N-isopropyl-2-propanamine did
not result in any appreciable change in δ or ΔΔδ values for
R-H of mandelonitrile (Table 3). rac-Mandelonitrile in the
presence of (S)-mandelic acid showed resonance peaks corre-
sponding to two enantiomers at δ 5.44 and 5.35 with DMAP;
5.38 and 5.26 with triethylamine, and 5.40 and 5.30 with N-
ethyl-N-isopropyl-2-propanamine as base.
The optimized space-filling model representations are
shown in Figure 6. Whereas the R-H of nitrile experiences
(38) (a) London, F. J. Phys. Radium 1937, 8, 397–409. (b) McWeeny, R.
Phys. Rev. 1962, 126, 1028. (c) Ditchfield, R. Mol. Phys. 1974, 27, 789. (d)
Wolinski, K.; Hilton, J. F.; Pulay, P. J. Am. Chem. Soc. 1990, 112, 8251. (e)
Boese, A. D.; Handy, N. C. J. Chem. Phys. 2001, 114, 5497.
(39) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Montgomery, J. A.; Vreven, T., Jr.; Kudin, K. N.;
Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci,
B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada,
M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima,
T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.;
Hratchian, H. P.; Cross, J. B.; Adamo, C.; Jaramillo, J.; Gomperts, R.;
Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.;
Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.;
Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain,
M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;
Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski,
J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.;
Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.;
Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen,
W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian-03; Gaussian, Inc.:
Pittsburgh, PA, 2003.
(40) (a) Parr, R. G.; Yang, W. Density Functional Theory of Atoms and
Molecules; Oxford University Press: New York, 1989. (b) Bartolotti, L. J.;
Fluchick, K. In Reviews in Computational Chemistry; Lipkowitz, K. B.,
Boyd, D. B., Eds.; VCH Publishers: New York, 1996; Vol. 7, pp 187-216.
In fact, triethylamine gave the best ΔΔδ values out of
various amines tested (Table 3). Neither triethylamine nor
5492 J. Org. Chem. Vol. 75, No. 16, 2010

Moon et al.
JOCArticle
FIGURE 6. Space-filling representations for ternary complexes (a) (S)-mandelonitrile/(R)-mandelate-DMAPHþ, (b) (S)-mandelonitrile/
(S)-mandelate-DMAPHþ, (c) (S)-lactonitrile/(R)-mandelate-DMAPHþ, and (d) (S)-lactonitrile/(S)-mandelate-DMAPHþ. R-H of nitrile is
shown in purple. Similar conformations were obtained for (R)-mandelonitrile and (R)-lactonitrile complexes (see the Supporting Information).
TABLE 2. Calculated δ and ΔδRS Values (in ppm) for Various Com-
plexes of Mandelonitrile and Lactonitrile with Mandelate-DMAPHþ
Ion Paira-c
TABLE 4. Enantiomeric Discrimination (ΔΔδ, in ppm) of Mandelonitrile
with Various Carboxylic Acids (1 molar equiv) in the Presence of DMAP
entry
1
carboxylic acid
ΔΔδa (ppm)
δ with
(R)-acid
δ with
(S)-acid
(S)-2-hydroxy-2-phenylacetic acid
(mandelic acid)
0.084
entry
nitrile
ΔδRS
3
4
5
6
(S)-mandelonitrile 5.65 (5.35) 6.19 (5.44) -0.54 (-0.09)
(R)-mandelonitrile 6.26 (5.44) 5.63 (5.35) þ0.63 (þ0.09)
2
(S)-2-(6-methoxynaphthalen-2-yl)-
propanoic acid (naproxen)
(R)-2-(3-chlorophenyl)-2-hydroxyacetic acid
(S)-2-phenylpropanoic acid
(S)-2-methoxy-2-phenylacetic acid
(S)-3,3,3-trifluoro-2-methoxy-2-
phenylpropanoic acid (MTPA)
(S)-2-phenylbutanoic acid
(S)-2-methylbutanoic acid
(S)-2-hydroxy-4-methylpentanoic acid
(S)-2-cyclohexyl-2-hydroxyacetic acid
0.030
(S)-lactonitrile
(R)-lactonitrile
4.75 (4.34) 5.04 (4.45) -0.29 (-0.11)
5.16 (4.45) 4.72 (4.35) þ0.43 (þ0.10)
3
4
5
6
0.048
0.031
0.041
0.115
aSee the Experimental Section for the method of calculating chemical
shifts. bCalculated chemical shifts for R-H of nitrile are reported as ppm
from the value calculated for TMS after conversion from shielding
values. cExperimentally observed values are shown in parentheses.
7
8
9
10
0.030
0.0
0.0
0.0
TABLE 3. Enantiomeric Discrimination (ΔΔδ, in ppm) of Mandeloni-
trile with (S)-Mandelic Acid in the Presence of Various Amines
aR-H of mandelonitrile.
entry
amine
ΔΔδa (ppm)
1
2
3
4
5
6
7
triethylamine
0.124
0.096
0
0.009
0.032
0.032
0.084
summarized in Table 4. The best results were obtained with
(S)-MTPA (entry 6, Table 4), which showed a ΔΔδ value of
0.115, compared to 0.084 ppm obtained with (S)-mandelic
acid. The presence of an aryl group was found to be essential
as aliphatic acids (entry 8-10, Table 4) failed to cause any
enantiodifferentiation. Optically active amines (R)-1-pheny-
lethanamine and (R,R)-1,2-diphenylethane-1,2-diamine in
combination with achiral phenylacetic acid failed to cause
any enantiodifferentiation. A combination of (S)-mandelic
acid and (R,R)-1,2-diphenylethane-1,2-diamine gave ΔΔδ of
0.046 ppm, which was not superior to a combination of (S)-
mandelic acid and triethylamine or DMAP. This indicated
that the chirality of acid and not amine is important for the
enantiodifferentiation of cyanohydrins.
N-ethyl-N-isopropyl-2-propanamine
aniline
pyridine
5-ethyl-2-methylpyridine
2,6-dimethylpyridine
4-(N,N-dimethylamino)pyridine (DMAP)
aR-H of mandelonitrile.
DMAP caused any racemization of mandelonitrile for 6 h
(usually a few minutes are sufficient to complete the
experiment). However, we recommend the use of DMAP
because it is less likely to cause any racemization of cyano-
hydrins.
Further analysis of molecular models shown in Figure 6
suggested the following points. (i) The hydroxyl group of
mandelic acid may not be necessary for stabilization of the
ternary complex; thus, it should be possible to replace the
-OH group with other groups such as methyl. Aromatic ring
appears to be essential for differential shielding of R-H of (R)-
and (S)-madelonitrile. (ii) The ion pair should form a stable
ternary complex with any hydrogen-bond donor, provided its
OH bond is sufficiently polarized to form a hydrogen bond;
e.g., optically active mandelic acid in the presence of DMAP
should be able to resolve another carboxylic acid. (iii) The R-H
of racemic mandelic acid should also get resolved with optically
enriched mandelonitrile, but not with lactonitrile. These infer-
ences were experimentally validated as given below.
We applied the computational model described above for
mandelic acid to MTPA. The optimized space-filling model
representations are shown in Figure 7. The calculated δ values
for R-H of mandelonitrile in (S)-mandelonitrile/(S)-MTPA-
DMAPHþ complex was 5.42, which corresponded well with the
experimentally observed value of 5.39. The calculated δ value
for R-H in (S)-mandelonitrile/(R)-MTPA-DMAPHþ complex
was 6.03, which showed a deviation of only 0.52 ppm from the
experimentally observed value of 5.51. Thus, the δ values for
R-H of mandelonitrile calculated based on models in Figure 7
were in good agreement with the experimentally observed values.
A range of cyanohydrins were tested using readily available
(S)-naproxene, (S)-2-(6-methoxynaphthalen-2-yl)propanoic
acid (39), in presence of DMAP (Table 5). The magnitude of
resolution of cyanohydrins with (S)-naproxene in presence of
DMAP was in the range of 0.006-0.041, which was inferior
compared to a combination of (S)-mandelic acid and DMAP.
Mandelic Acid Can Be Substituted with Other Chiral
Aromatic Acids. Using DMAP as base, various optically
pure acids were tested for their ability to enantiodifferentiate
methine proton of racemic mandelonitrile. The results are
J. Org. Chem. Vol. 75, No. 16, 2010 5493

Moon et al.
JOCArticle
many chiral shift reagents,4 such as amines,48-50 diamines,51-54
amides,12,55 and macrocyclic compounds,8,10,56-59 have been
described for determination of ee of chiral carboxylic acids by
1H NMR.
1
We started our investigation by recording H NMR of
racemic mandelic acid in the presence of 1 molar equiv each
of (R)-mandelonitrile and DMAP at 20 °C in CDCl3. The
singlet due to the methine proton of both enantiomers of
mandelic acid suffered unequal upfield shift and appeared as
two well-resolved singlets at δ 4.96 and 4.98 (ΔΔδ = 0.015
ppm). The resolution was increased to ΔΔδ = 0.017 when
(R)-mandelonitrile was substituted with (R)-4-methoxy-
mandelonitrile, whereas resolution was decreased when
(R)-mandelonitrile was substituted with either (R)-4-methyl-
mandelonitrile (ΔΔδ = 0.013) or (R)-4-choloromandelonitrile
(ΔΔδ = 0.011). Thus, (R)-4-methoxymandelonitrile in the
presence of DMAP was selected as the reagent of choice for
further studies.
The observed resonance in NMR is the average of reso-
nances of complexed and uncomplexed species present in
solution at equilibrium. The magnitude of Δδ is expected to
increase with increase in the concentration of the 1:1 complex
with respect to uncomplexed species. Since the 1:1 complex is
in equilibrium with its components in solution, adding excess
chiral solvating agent to the sample should push the equilib-
rium in favor of the 1:1 complex. Accordingly, the resolution
was improved to ΔΔδ = 0.025 when the ratio of (R)-4-
methoxymandelonitrile was increased to 2.5 molar equiv
(Figure 8). Further, a significant improvement in resolution
occurred when the probe temperature was reduced to -40 °C.
At a probe temperature of -40 °C, ΔΔδwas 0.044 with 1 molar
equiv and 0.054 with 2.5 molar equiv of (R)-4-methoxyman-
delonitrile.
FIGURE 7. Space-filling representations for ternary complexes (a)
(S)-mandelonitrile/(S)-MTPA-DMAPHþ and (b) (S)-mandelonitrile/
(R)-MTPA-DMAPHþ. R-H of mandelonitrile is shown in purple.
TABLE 5. Enantiodiscrimination of Cyanohydrins with (S)-Naproxen
(39)-DMAPHþ Ion Pair
entry
cyanohydrin
resolution ΔΔδa (ppm)
1
2
3
4
5
6
7
mandelonitrile
0.030
0.035
0.029
0.025
0.041
0.029
0.023
0.032
0.029
0.006
0.006
0.007
4-chloromandelonitrile
2-chloromandelonitrile
4-ethoxymandelonitrile
3-bromomandelonitrile
3,4-dimethoxymadelonitrile
4-methylmandelonitrile
10 3-chloro-2-hydroxypropanenitrile
12 2-hydroxybutanenitrile
13 2-hydroxy-2-phenylpropanenitrile
15 2-hydroxy-2-(pyridin-4-yl)propanenitrile
16 2-cyclohexyl-2-hydroxypropanenitrile
aR-H of mandelonitrile.
The mandelate-DMAPHþ ion pair should form a ternary
complex with any hydrogen bond donor, provided its OH
bond is sufficiently polarized to form a hydrogen bond. Such
A range of racemic R-substituted phenylacetic acids were
tested using 1 and 2.5 equiv of (R)-4-methoxymandelonitrile
at 20 or -40 °C. The ratio of DMAP was kept constant at 1
molar equiv in all cases. Baseline separation for methine
proton occurred in all examples of R-substituted phenylace-
tic acids derivatives studied under all conditions. However,
maximum resolution occurred when 2.5 equiv of (R)-4-
methoxymandelonitrile was used at -40 °C. Under these
conditions, ΔΔδ was in the range of 0.030-0.058 ppm for
mandelic acid derivatives (entries 1-4; Table 6). When -OH
of mandelic acid was replaced with Br, ΔΔδ was of the order
1
compounds should also face enantiodiscrimination in H
NMR. Thus, racemic 2-chloromandelic acid was resolved
(ΔΔδ = 0.008 at 25 °C and 0.014 at -40 °C for R-H) with (S)-
mandelic acid-DMAPHþ ion pair, lending further support
to the proposed model.
Mandelic acid should also get resolved with optically active
mandelonitrile. This was found to be true and formed the
basis for development of a new method for determination of
ee of carboxylic acids. Chiral carboxylic acid is an important
functionality present in natural products, biological mole-
cules, metabolic intermediates, and pharmaceuticals. Rapid
developments in the area of search for new chemical and
biological catalysts aided by high-throughput combinatorial
techniques for enantioselective synthesis of chiral carboxylic
acids has given rise to an increasing demand for simple, easy to
use, cheap, and reliable methods for the determination of
enantiomeric purity of these compounds.41-47 In recent years,
(48) Port, A.; Virgili, A.; Jaime, C. Tetrahedron: Asymmetry 1996, 7,
1295–1302.
(49) Port, A.; Virgili, A.; Alvarez-Larena, A.; Piniella, J. F. Tetrahedron:
Asymmetry 2000, 11, 3747–3757.
(50) Enders, D.; Thomas, C. R.; Runsink, J. Tetrahedron: Asymmetry
1999, 10, 323–326.
(51) Yang, X.; Wang, G.; Zhong, C.; Wu, X.; Fu, E. Tetrahedron:
Asymmetry 2006, 17, 916–921.
(52) Fulwood, R.; Parker, D. J. Chem. Soc., Perkin Trans. 2 1994, 57–64.
(53) Staubach, B.; Buddrus, J. Angew. Chem., Int. Ed. 1996, 35, 1344–
1346.
(54) Bailey, D. J.; O’Hagan, D.; Tavasli, M. Tetrahedron: Asymmetry
1997, 8, 149–153.
(55) Bilz, A.; Stork, T.; Helmchen, G. Tetrahedron: Asymmetry 1997, 8,
3999–4002.
(56) Yuan, Q.; Fu, E.-Q.; Wu, X.-J.; Fang, M.-H.; Xue, P.; Wu, C.-T.;
Chen, J.-H. Tetrahedron Lett. 2002, 43, 3935–3937.
(57) Yang, X.; Wu, X.; Fang, M.; Yuan, Q.; Fu, E. Tetrahedron:
Asymmetry 2004, 15, 2491–2497.
(41) Coppola, G. M.; Schuster, H. F. R-Hydroxy Acids in Enantioselective
Syntheses; Wiley-VCH: Weinheim, 1997.
(42) Nicolaou, K. C.; Snyder, S. A. Classics in Total Synthesis II; Wiley-
VCH: Weinheim, 2003.
(43) Nicolaou, K. C.; Sorensen, E. J. Classics in Total Synthesis; VCH:
Weinheim, 1996.
(44) Pu, L. Chem. Rev. 2004, 104, 1687–1716.
(45) Hanessian, S. Total synthesis of Natural Products: The Chiron
Approach; Pergamon: Oxford, 1983.
(46) Lin, J.; Rajaram, A. R.; Pu, L. Tetrahedron 2004, 60, 11277–11281.
(47) Taran, F.; Gauchet, C.; Mohar, B.; Meunier, S.; Valleix, A.; Renard,
P. Y.; Ger’minon, C.; Grassi, J.; Wagner, A.; Mioskowski, C. Angew. Chem.,
Int. Ed. 2002, 41, 124–127.
(58) Gonalez-Alvarez, A.; Alfonso, I.; Gotor, V. Tetrahedron Lett. 2006,
47, 6397–6400.
(59) Gonzalez, S.; Pelaez, R.; Sanz, F.; Jimenez, M. B.; Moran, J. R.;
Caballero, M. C. Org. Lett. 2006, 8, 4679–82.
5494 J. Org. Chem. Vol. 75, No. 16, 2010

Moon et al.
JOCArticle
FIGURE 8. NMR enantiodiscrimination of (R,S)-mandelic acid with (R)-4-methoxymandelonitrile in presence of DMAP. Partial 1H NMR
spectrum (300 MHz) of (R,S)-mandelic acid in CDCl3 (a); (R,S)-mandelic acid in the presence of 1 molar equiv each of DMAP and (R)-4-
methoxymandelonitrile at 20 °C (b) and -40 °C (d); (R,S)-mandelic acid in presence of 1 molar equiv DMAP and 2.5 molar equiv of (R)-4-
methoxymandelonitrile at 20 °C (c) and -40 °C (e).
of 0.071 ppm (entry 5, Table 6). The resolution was increased
to 0.082 when -OH of mandelic acid was replaced with
-OCH3 (entry 6, Table 6). It is significant to note that in
addition to the methine proton located at asymmetric car-
bon, the -OCH3 protons at the para position of phenyl were
also resolved (ΔΔδ = 0.030 ppm; entry 4, Table 6).
1 equiv and 0.018 ppm with 2.5 equiv of (R)-4-methoxyman-
delonitrile at 20 °C, which was improved to 0.030 and 0.040,
respectively, at -40 °C (entry 14, Table 6). The doublet due
to the methyl group was also resolved, but the maximum
ΔΔδ obtained was 0.009 with 2.5 equiv of (R)-4-methoxy-
mandelonitrile at -40 °C. When the methyl group of lactic
acid was substituted with cyclohexyl group, increased reso-
lution was observed with a ΔΔδ value of 0.041 and 0.054 for
methine doublet, with 1 and 2.5 equiv of 2, respectively, at
-40 °C (entry 15, Table 6). However, with 2.5 equiv of (R)-4-
methoxymandelonitrile, there was some overlap of the
methine doublet with the methoxy signal of (R)-4-methoxy-
mandelonitrile. When -OH of lactic acid was replaced with
Br, the resolution was decreased; the maximum ΔΔδ value of
0.023 was obtained with 2.5 equiv of (R)-4-methoxymande-
lonitrile at -40 °C (entry 16, Table 6). The resolution
obtained with 2-bromobutyric acid and 2-methylpentanoic
acid was poor (entries 17 and 18, Table 6). Increasing the
ratio of (R)-4-methoxymandelonitrile and decreasing the
probe temperature to -40 °C resulted only in a marginal
increase in ΔΔδ value in these examples.
Resolution of Carboxylic Acids in the Presence of Triethy-
lamine. To further improve resolution of carboxylic acid,
especially in the case of 2-phenyl- and 2-phenoxypropanoic
acids, DMAP was substituted with triethylamine, which
gave the best resolution in the case of cyanohydrins. The
ratio of triethylamine was kept constant at 1 molar equiv in
all cases. Baseline separation for methine proton occurred
with 1 molar equiv of (R)-4-methoxymandelonitrile at 20 °C
in all examples of mandelic acid derivatives studied, with
Resolution of Racemic 2-Phenyl- and 2-Phenoxypropanoic
Acids in the Presence of DMAP. Determination of ee of
2-phenyl- and 2-phenoxypropanoic acids by 1H NMR meth-
ods is more difficult because in these cases the resonance due
to methine proton appears as quartet due to coupling with
methyl group and as a consequence merger of signals occurs.
When 2.5 equiv of (R)-4-methoxymandelonitrile and a probe
temperature of -40 °C were used, ΔΔδ values in the range of
0.041-0.078 ppm (entries 7-12, Table 6) were obtained, which
are sufficient for the determination of ee of these compounds. In
all of these examples, the doublet due to the methyl group of
propanoates was also resolved with ΔΔδ values in the range of
0.021-0.032 ppm. In the case of 2-phenylbutanaote, the ΔΔδ
value for triplet of methine proton was 0.085 (entry 13, Table 6).
It was interesting to note that the signals of aromatic protons
at the ortho position in 2-(4-cholorophenoxy)- and 2-(2,4-
dicholorophenoxy)propanoic acids also suffered unequal shift
and were resolved with ΔΔδ values of 0.031 and 0.056 ppm,
respectively (entries 10 and 12, Table 6).
Resolution of Racemic Aliphatic Carboxylic Acids in the
Presence of DMAP. Next, resolution of aliphatic carboxylic
acids with 4-methoxylmandelonitrile in the presence of
DMAP was studied. The quartet due to the methine proton
of racemic lactic acid showed a ΔΔδ value of 0.011 ppm with
J. Org. Chem. Vol. 75, No. 16, 2010 5495

Moon et al.
JOCArticle
TABLE 6. Resolution (ΔΔδ, in ppm) of Racemic R-Substituted Phenylacetic Acids, 2-Phenyl- and 2-Phenoxypropanoic Acids, and Aliphatic Caroboxylic
Acids in the Presence of (R)-4-Methoxymandelonitrile and DMAP
(R)-4-methoxymandelonitrile (1.0 equiv) Δδ (ppm)a
20 °C -40 °C
0.017 0.044
(R)-4-methoxymandelonitrile (2.5 equiv) Δδ (ppm)a
20 °C -40 °C
0.025 0.054
entry
acid
1
2
3
4
40
41
42
43
0.019
0.024
0.018
0.053
0.023
0.047
0.029
0.068
0.030
0.058
0.026
0.009; OCH3
0.010
0.021
0.016
0.006; H3
0.013
0.0; H3
0.013
0.003; H3
0.016
0.021; OCH3
0.045
0.056
0.048
0.024; H3
0.048
0.024; H3
0.041
0.020; H3
0.037
0.015; OCH3
0.022
0.037
0.025
0.013; H3
0.030
0.011; H3
0.024
0.007; H3
0.023
0.030; OCH3
0.071
0.082
0.072
0.031; H3
0.078
0.032; H3
0.066
0.024; H3
0.050
5
6
7
44
45
46
8
9
47
48
49
10
0.006; H3
0.009; H20, H50
0.011
0.009; H3
0.028; H20, H50
0.024
0.0; H3
0.033
0.018; H3
0.047; H60
0.049
0.024; H3
0.030
0.013; H3
0.014; H20, H50
0.022
0.009; H3
0.023
0.022; H3
0.024; H60
0.026
0.013 H3
0.018
0.021; H3
0.031; H20, H50
0.046
0.021; H3
0.041
0.023; H3
0.056; H60
0.085
0.038 H3
0.040
11
12
50
51
0.0; H3
0.014
0.010; H3
b
-; H60
13
14
52
53
0.014
0.006; H3
0.011
0.002; C3
0.013
0.009
0.008; C3
0.041
0.014
0.005; C3
0.024
0.017
0.009; C3
0.054
0.023
0.008
0.008
15
16
17
18
54
55
56
57
0.007
0.010
0.006
aδ for H2 unless stated otherwise. bPeaks merged.
improvement in resolution in most of the examples (entries
1-4, Table 7). In the case of 2-phenyl- and 2-phenoxypro-
panoic acids, baseline separation was achieved using 2.5
equiv of 4-methoxymandelonitrile and at a probe tempera-
ture of -40 °C (entries 5-11, Table 7). ΔΔδ values in the
range of 0.084-0.123 ppm were obtained, which are suffi-
cient for the determination of ee of these compounds and
better than resolution obtained with DMAP. Aliphatic
carboxylic acids with (R)-4-methylmandelonitrile in the
presence of triethylamine gave a resolution of 0.004-0.024
ppm (entries 12-14, Table 7).
Determination of Absolute Configuration of Carboxylic
Acids. A procedure that is often used for determination of
absolute configuration is to look for the presence of specific
trends in the shifts that correlate with the absolute config-
uration of the substrate.3,5 The assumption is that, if the
trends are consistent among a series of compounds with
known configurations, then they will be consistent for an
unknown compound with a similar structure. Applying
empirical trends such as these, we have described above the
development of a method for the determination of absolute
configuration of cyanohydrins. On the basis of the analysis
5496 J. Org. Chem. Vol. 75, No. 16, 2010

Moon et al.
JOCArticle
TABLE 7. Resolution(ΔΔδ, inppm) of Racemic 2-Phenyl- and 2-Phenoxy-
propanoic Acids and Aliphatic Carboxylic Acids in Presence of (R)-4-
Methoxymandelonitrile and Triethylamine
as only the magnitude of resolution is expected to be lower with
CSAs as compared to CDAs.
Accordingly, the ΔδRS value of -0.082 (R-H of 4-methoxy-
mandelonitrile) was obtained with (S)-mandelic acid and
þ0.085 with (R)-mandelic acid using 4-methoxymandelonitrile
in the presence of DMAP. Encouraged by these results, we
pursued this method further. It occurred to us that obtaining
two NMRs for each sample, one with optically pure (R)-
mandelonitrile and the other with optically pure (S)-mandelo-
nitrile, to obtain ΔδRS values may not be necessary; instead, a
single NMR using (R)- or (S)-mandelonitrile of lower ee may be
sufficient. Thus, we prepared solution of 4-methoxymandeloni-
trile containing (R)- and (S)-enantiomers in a ratio of approxi-
mately 85:15 (70% ee) by diluting a solution of optically pure
(R)-4-methoxymandelonitrile with an appropriate amount of
racemic 4-methoxymandelonitrile. The actual R to S ratio was
determined by 1H NMR method described above. Using
(R)-4-methoxymandelonitrile of 70% ee, we determined the
ΔδRS values for (R)- and (S)-mandelic acid, which were found
to be þ0.0586 and -0.0618, respectively. Although the ΔδRS
values were compromised to some extent compared to values
obtained when optically pure (R)- and (S)-4-methoxymande-
lonitriles were used, these values are sufficient to look for
existence of any trend in a series of carboxylic acids. ΔδRS
values obtained with a series of optically pure carboxylic acids
with (R)-4-methoxymandelonitriles, ee 70%, in the presence
of DMAP are given in Figure 9. No change in sign of ΔδRS
was observed when these values were obtained using optically
pure (R)- and (S)-4-methoxymandelonitrile; only the magni-
tude of resolution was higher by ∼0.020 ppm in all examples
studied. Under these conditions, all carboxylic acids, which
are spatially related to (S)-mandelic acid, showed a negative
ΔδRS value, whereas those related to (R)-mandelic acid
showed a positive ΔδRS value. Thus, the ΔδRS sign appears
to be characteristic for this enantiomeric series and possibly
can be used for the assignment of absolute configuration.
(R)-4-methoxy
mandelonitrile
(1.0 equiv),
(R)-4-methoxy
mandelonitrile
(2.5 equiv), -40 °C
ΔΔδa (ppm)
entry
acid
20 °C ΔΔδa (ppm)
1
2
3
40
41
43
0.021
0.024
0.019
0.054
0.084
0.057
0.009; OCH3
0.026
0.023
0.032; OCH3
0.103
0.085
4
5
45
46
0.138; H3
0.024;
0.014; H3
0.024
0.010; H3
0.025
0.057; H3
0.088;
0.057; H3
0.098
0.046; H3
0.093
6
7
8
47
48
49
0.013; H3
0.013; H20, H50
0.023
0.011; H3
0.025
0.034; H3
0.047; H20, H50
0.123
0.040; H3
0.084
9
50
51
10
0.013; H3
0.013; H60
0.026
0.060; H3
0.088; H60
0.102
11
52
0.013; H3
0.013
0.005
0.052; H3
0.024
12
13
14
55
56
57
0.004
0.008
aΔΔδ for H2 unless stated otherwise.
1
of H NMR data of cyanohydrins of known configuration
obtained in the presence of the mandelate-DMAPHþ ion
pair, we have shown the existence of a correlation between
chemical shifts and the absolute configuration of cyanohy-
drins. In these cases, the chiral solvating agent, i.e., ion pair,
caused a significant difference in chemical shift for two
enantiomers of the substrate, i.e., cyanohydrin.
We have described above the enantiodiscrimination of
carboxylic acids with chiral mandelonitriles in the presence
of DMAP (Table 6). Although the resolution obtained was
sufficient for the determination of ee of these compounds,
the magnitude was not sufficient to look for correlation
between chemical shift values and absolute configuration.
However, moieties on the substrate that may cause a specific
trend in the shifts in the resonance of chiral derivatizing agent
have also been used for assigning absolute configuration.3 A
chiral derivatizing agent (CDA) could be thought of as CSA
where all the solute is in the complexed form; therefore, the
principles developed for CDAs can be applied to CSAs as well,
Conclusions
Mandelic acid in the presence of DMAP is an effective
chiral solvating agent (CSA) for the determination of ee and
absolute configuration of cyanohydrins. Resolutions of the
order of 0.055-0.112 ppm (16.5-33.6 Hz) for aldocyanohy-
drins and 0.016-0.031 ppm (4.8-9.3 Hz) for ketocyanhy-
drins on 300 MHz NMR were obtained. The magnitude of
resolution for ketone cyanohydrins was less compared to
aldehyde cyanohydrins but sufficient for the determination
of optical purity of these compounds. A model has been
FIGURE 9. Correlation between chemical shifts and absolute configuration of carboxylic acids. ΔδRS (ΔδR - ΔδS) values in ppm for R-H of
nitrile are shown in parentheses.
J. Org. Chem. Vol. 75, No. 16, 2010 5497

Moon et al.
JOCArticle
proposed on the basis of molecular modeling studies. Several
inferences drawn from the proposed model were experimen-
tally verified, including that (i) DMAP may be replaced with
other amines, (ii) the hydroxyl group of mandelic acid is not
necessary for stabilization of ternary complex and may be
replaced with other groups such as methyl, (iii) the ion pair
should form a stable ternary complex with any hydrogen
bond donor, provided its OH bond is sufficiently polarized,
and (iv) R-H of racemic mandelic acid should also get
resolved with optically pure mandelonitrile, which not only
validated the proposed model but also led to development of new
CSA for determination of ee of aryl and alkyl carboxylic acids.
Resolutions of the order of 0.021-0.085 ppm (6.3-25.5 Hz) for
aryl carboxylic acids and 0.008-0.057 ppm (2.4-17.1 Hz)
for aliphatic carboxylic acids on 300 MHz NMR were
obtained, which are sufficient for accurate determination
of ee of these compounds. CSA appears to be suitable for the
determination of absolute configuration of aryl carboxylic
acids but not for aliphatic carboxylic acids.
Modeling Studies. The modeling studies were performed with
Gaussian03 software.39 Initial geometry optimization on the tern-
ary complexes was carried out using the molecular mechanics
method in Chem3D software, and a few models of the ternary
complex were chosen for optimization using the density functional
theory (DFT) method with the B3LYP/6-31þG(d) basis set.40 All
of the optimization studies were carried out in the gas phase. The
optimized models were employed to computationally estimate
chemical shifts in implicit chloroform medium (IEFPCM method)37
using the GIAO algorithm38 and B3LYP/6-311þG(2d,p) method.
As NMR experimental studies have been carried out in chloroform
solvent, GIAO calculations were performed in chloroform solvent
for reliable comparison. Calculated chemical shifts for R-H of nitrile
are reported as ppm from the value calculated for TMS after
conversion from shielding values.
Substitution of DMAP with Other Amines. (S)-Mandelic acid
(18 μmol) and CDCl3 (0.6 mL) were mixed in 5 mm NMR tube
and amine (18 μmol) added to it. Mandelonitrile (18 μmol) was
1
added, and H NMR data was collected on a JEOL ACX 300
MHz spectrometer. Chemical shifts (ppm), internally referenced
to TMS signal (0 ppm), were obtained.
Substitution of (S)-Mandelic Acid with Other Carboxylic Acids.
Carboxylic acid (18 μmol) and CDCl3 (0.6 mL) were mixed in a
5 mm NMR tube, and DMAP (18 μmol) added. Cyanohydrin
(18 μmol) was added, and 1H NMR data was collected on JEOL
ACX 300 MHz spectrometer. Chemical shifts (ppm) internally
referenced to TMS signal (0 ppm) were obtained.
Chiral Solvating Agent for Carboxylic Acids. Carboxylic acid
(18 μmol) and CDCl3 (0.6 mL) were mixed in a 5 mm NMR tube,
and DMAP or triethylamine (18 μmol) was added. (R)-4-
Methoxymandelonitrile (18 or 45 μmol) was added, and 1H
NMR data were collected on a JEOL ACX 300 MHz spectro-
meter. Chemical shifts (ppm) internally referenced to TMS
signal (0 ppm) were obtained.
Experimental Section
Stoichiometry of the complex. NMR method as described pre-
viously was used to determine the stoichiometry of the com-
1
plex.34,35 The H NMR spectra of the (S)-mandelate-DMAPHþ
ion pair with (R)-mandelonitrile in a various ratios in CDCl3 at a
constant total concentration of 40 μM were recorded. It was found
that the R-H of mandelonitrile underwent a variable upfield shift
depending upon the ratio of mandelonitrile and ion pair. Jobs plot of
ΔδXi (the product of the chemical shift change and the mole
fraction) versus the mole fraction (Xi) of (R)-mandelonitrile in the
mixture were obtained.
Determination of Stability Constants. A 20 mM solution of
mandelate--DMAPþ ion pair in CDCl3 was placed in 19 5 mm
NMR tubes. A predetermined quantity of a concentrated solu-
tion of mandelonitrile in CDCl3 was added to each of 18 tubes so
that finally solutions with desired relative amounts (equiv) of the
mandelonitrile versus ion pair were obtained. Volume and
concentrations changes were taken into account during analysis.
The concentration of the ion pair was always maintained at
20 mM. Plots of concentration versus chemical shift were obtained.
An association constant for (R)-mandelonitrile/(R)-mandelate-
DMAPHþ complex and (R)-mandelonitrile/(S)-mandelate-
DMAPHþ complex was obtained by nonlinear least-squares
Acknowledgment. We thank Prof. M. J. Hynes of Na-
tional University of Ireland for providing WinEQNMR.
This work was supported by an NWP006 grant from the
Council of Scientific and Industrial Research (CSIR), New
Delhi, India. L.S.M. and M.P. acknowledge CSIR for the
award of Senior Research Fellowships.
Supporting Information Available: Cartesian coordinates
for models in Figures 6 and 7; tabulated partial 1H NMRs
showing resolution of carboxylic acid and H NMRs for cya-
1
1
fitting for the H NMR titration curve using the WinEQNMR
program.36 Values of ΔG0 were calculated using equation ΔG0 =
-RT ln Ka.
nohydrins 1-19 and carboxylic acids 39-56 in presence of
chiral shift reagent. This material is available free of charge
via the Internet at http://pubs.acs.org.
5498 J. Org. Chem. Vol. 75, No. 16, 2010