Catalytic parallel kinetic resolution under homogeneous conditions

Article abstract of DOI:10.1021/jo100695z

(Figure Presented) Two complementary chiral catalysts, the phosphine 8d and the DMAP-derived ent-23b, are used simultaneously to selectively activate a mixture of two different achiral anhydrides as acyl donors under homogeneous conditions. The resulting activated intermediates 25 and 26 react with the racemic benzylic alcohol 5 to form enantioenriched esters (R)-24 and (S)-17 by fully catalytic parallel kinetic resolution (PKR). The aroyl ester (R)-24 is obtained with near-ideal enantioselectivity for the PKR process, but (S)-17 is contaminated by ca. 8% of the minor enantiomer (R)-17 resulting from a second pathway via formation of mixed anhydride 27 and its activation by 8d.

Full text of DOI:10.1021/jo100695z

pubs.acs.org/joc
Catalytic Parallel Kinetic Resolution under Homogeneous Conditions
Trisha A. Duffey, James A. MacKay,† and Edwin Vedejs*
Department of Chemistry, University of Michigan, Ann Arbor, Michigan 48109. †Present Address:
Department of Chemistry and Biochemistry, Elizabethtown College, 1 Alpha Drive, Elizabethtown, PA 17022
edved@umich.edu
Received April 10, 2010
Two complementary chiral catalysts, the phosphine 8d and the DMAP-derived ent-23b, are used
simultaneously to selectively activate a mixture of two different achiral anhydrides as acyl donors
under homogeneous conditions. The resulting activated intermediates 25 and 26 react with the
racemic benzylic alcohol 5 to form enantioenriched esters (R)-24 and (S)-17 by fully catalytic parallel
kinetic resolution (PKR). The aroyl ester (R)-24 is obtained with near-ideal enantioselectivity for the
PKR process, but (S)-17 is contaminated by ca. 8% of the minor enantiomer (R)-17 resulting from a
second pathway via formation of mixed anhydride 27 and its activation by 8d.
Introduction
Two strategies have been developed that avoid the mass
action problem by maintaining the enantiomer ratio near 1:1
throughout the reaction, (1) dynamic kinetic resolution
(DKR)3 and (2) parallel kinetic resolution (PKR).4 In prin-
ciple, DKR is the simpler approach, but it is limited to
systems where interconversion of enantiomers is possible
during KR. The more recently recognized alternative of
PKR also has limitations, but they are related more to
reagent selectivity and experimental design.
In PKR, two kinetic resolutions are performed simulta-
neously (“in parallel”) so that each enantiomer is converted
to distinct products via two enantiodivergent pathways.5 If
the two kinetic resolutions convert each enantiomer at a
similar rate, then the starting material remains racemic
throughout, and the mass action problem is avoided. For
the ideal PKR experiment, the two kinetic resolutions should
(1) occur with similar rates, (2) use two reagents with
The kinetic resolution (KR) of racemic mixtures is a power-
ful method for obtaining enantioenriched substances.1 Even
withrelatively modestenantioselectivity, s (kfast/kslow) = 20or
above, a substantial fraction of the less reactive enantiomer
can be recovered with >90% ee if the reaction is run to g55%
conversion. However, this approach usually results in <50%
recovery of highly enriched material due to the well-known
consequence of competition kinetics and the problem of
mass action. As the faster-reacting enantiomer is consumed,
the ratio of less reactive to more reactive enantiomers
increases, and eventually, only the less reactive enantiomer
remains. This is the key advantage of KR, but it ensures that
conversion of the slower reacting enantiomer competes
increasingly over time. One consequence is that the product
of simple KR cannot be obtained with acceptable yield and
ee unless s is of the order of 200 or better.1,2 Over the past
decade, nonenzymatic catalysts have begun to reach this
level of enantioselectivity, but applications that exploit en-
antioenriched products of KR remain quite rare.
(3) (a) Pellissier, H. Tetrahedron 2008, 64, 1563. (b) Pellissier, H. Tetra-
hedron 2003, 59, 8291. (c) Huerta, F. F.; Minidis, A. B. E.; Backvall, J. E.
Chem. Soc. Rev. 2001, 30, 321. (d) Ratovelomanana-Vidal, V.; Genet, J. P.
Can. J. Chem. 2000, 78, 846. (e) Noyori, R.; Tokunaga, M.; Kitamura, M.
Bull. Chem. Soc. Jpn. 1995, 68, 36. (f) Ward, R. S. Tetrahedron: Asymmetry
1995, 6, 1475.
(4) Vedejs, E.; Chen, X. J. Am. Chem. Soc. 1997, 119, 2584.
(5) (a) Dehli, J. R.; Gotor, V. Chem. Soc. Rev. 2002, 31, 365. (b) Eames, J.
Angew. Chem., Int. Ed. Engl. 2000, 39, 885.
(1) Reviews: (a) Kagan, H. B.; Fiaud, J. C. Top. Stereochem. 1988, 18,
249. (b) Keith, J. M.; Larrow, J. F.; Jacobsen, E. N. Adv. Synth. Catal. 2001,
343, 5. (c) Vedejs, E.; Jure, M. Angew. Chem., Int. Ed. 2005, 44, 3974.
(2) Spivey, A. C.; Maddaford, A.; Redgrave, A. J. Org. Prep. Proced. Int.
2000, 32, 331.
4674 J. Org. Chem. 2010, 75, 4674–4685
Published on Web 06/17/2010
DOI: 10.1021/jo100695z
r
2010 American Chemical Society

Duffey et al.
JOCFeatured Article
complementary enantioselectivies, (3) take place without
mutual interference by the reagents, and (4) afford readily
separated products. These requirements involve consider-
able planning and optimization, but numerous examples of
PKR using stoichiometric reagents are now known,5 includ-
ing nitrone cycloaddition,6 Horner-Wadsworth-Emmons
(HWE) olefination,7 Michael addition,8 reactions of lithia-
ted oxazolidinones,9 and transesterification.10
SCHEME 1. Stoichiometric PKR
Alternative approaches are available for converting the
enantiomers of a racemic mixture into distinct products
using a single chiral reagent or catalyst. Such reactions also
involve two enantiodivergent pathways and were cited in the
original paper discussing PKR,4 but they are conceptually
distinct and have been termed divergent reactions of a
racemic mixture.11 In contrast to the PKR experiments,
single reagent methods do not exploit the relative rates of
enantiomer conversion, do not encounter or deal with the
mass action problem, and do not benefit from maintaining a
1:1 ratio of enantiomers throughout.12
In a representative PKR experiment,4 racemic alcohol
5 was treated with equimolar amounts of preformed
(6) (a) Cardona, F.; Valenza, S.; Goti, A.; Brandi, A. Eur. J. Org. Chem.
1999, 1319. (b) Pietrusiewicz, K. M.; Holody, W.; Koprowski, M.; Cicchi, S.;
Goti, A.; Brandi, A. Phosphorus, Sulfur, Silicon Relat. Elem. 1999, 146, 389.
(c) Cardona, F.; Lalli, D.; Faggi, C.; Goti, A.; Brandi, A. J. Org. Chem. 2008,
73, 1999.
(7) (a) Pedersen, T. M.; Jensen, J. F.; Humble, R. E.; Rein, T.; Tanner, D.;
Bodmann, K.; Reiser, O. Org. Lett. 2000, 2, 535. (b) Pedersen, T. M.; Hansen,
E. L.; Kane, J.; Rein, T.; Helquist, P.; Norrby, P. O.; Tanner, D. J. Am.
Chem. Soc. 2001, 123, 9738. (c) Strand, D.; Rein, T. Org. Lett. 2005, 7, 199.
(d) Strand, D.; Norrby, P.-O.; Rein, T. J. Org. Chem. 2006, 71, 1879.
(8) (a) Davies, S. G.; Diez, D.; El Hammouni, M. M.; Garner, A. C.;
Garrido, N. M.; Long, M. J. C.; Morrison, R. M.; Smith, A. D.; Sweet, M. J.;
Witheya, J. C. M. Chem. Commun. 2003, 2410. (b) Davies, S. G.; Garner,
A. C.; Long, M. J. C.; Smith, A. D.; Sweet, M. J.; Withey, J. M. Org. Biomol.
Chem. 2004, 2, 3355. (c) Davies, S. G.; Garner, A. C.; Long, M. J. C.;
Morrison, R. M.; Roberts, P. M.; Savory, E. D.; Smith, A. D.; Sweet, M. J.;
Withey, J. M. Org. Biomol. Chem. 2005, 3, 2762. (d) Cailleau, T.; Cooke,
J. W. B.; Davies, S. G.; Ling, K. B.; Naylor, A.; Nicholson, R. L.; Price, P. D.;
Roberts, P. M.; Russell, A. J.; Smith, A. D.; Thomson, J. E. Org. Biomol.
Chem. 2007, 5, 3922. (e) Aye, Y.; Davies, S. G.; Garner, A. C.; Roberts, P. M.;
Smith, A. D.; Thomson, J. E. Org. Biomol. Chem. 2008, 6, 2195. (f) Abraham,
E.; Davies, S. G.; Docherty, A. J.; Ling, K. B.; Roberts, P. M.; Russell, A. J.;
Thomson, J. E.; Toms, S. M. Tetrahedron: Asymmetry 2008, 19, 1356.
(g) Davies, S. G.; Durbin, M. J.; Hartman, S. J. S.; Matsuno, A.; Roberts,
P. M.; Russell, A. J.; Smith, A. D.; Thomson, J. E.; Toms, S. M. Tetrahedron:
Asymmetry 2008, 19, 2870.
quasi-enantiomeric pyridinium salts 3 and 4 to afford two
mixed carbonates 6 and 7 in high yield and enantiomeric
excess (Scheme 1). Recovery of carbonate 6 in 49% yield and
95% de corresponds to s = 125 in a standard kinetic resolu-
tion. Such high selectivity is remarkable considering that
classical KR of alcohol 5 using a single chiral pyridinium salt
(either 3 or 4) occurs with s = 41-42. However, like nearly
all reported PKR experiments,5-10 the example of Scheme 1
has the disadvantage that it uses stoichiometric chiral
reagents.
In principle, PKR is also possible using a combination of
two chiral catalysts with two achiral reagents in place of the
two stoichiometric chiral reagents. This is a more difficult
experiment because requirement (3), above (noninterfering
reagents), can only be satisfied if each of the chiral catalysts
activates only one of the two achiral reagents with high rea-
gent selectivity. So far, only one study has demonstrated
PKR under catalytic conditions, and the necessary reagent
selectivity could only be achieved by relying on phase iso-
lation methods.13 Thus, a soluble, chiral phosphine catalyst
ent-8a14a was used in concert with the lipase catalyst Chiro-
CLEC 9 to derivatize the racemic alcohol 5 (Scheme 2).
Because the insoluble lipase has little reagent selectivity and
activates a broad range of achiral acyl donors, it was essential
to differentiate reagents according to solubility and to design
an insoluble acyl donor that would be activated only by the
(soluble) phosphine catalyst ent-8a. The success of this PKR
experiment therefore depends on (1) selective activation
of soluble vinyl pivalate (11) by the insoluble ChiroCLEC
(9), (2) no reactivity between 11 and phosphine ent-8a, (3)
activation of the insoluble, polymer-supported anhydride 10
by the soluble ent-8a, and (4) no acyl transfer between the
(9) Liao, L.; Zhang, F.; Dmintrenko, O.; Bach, R. D.; Fox, J. M. J. Am.
Chem. Soc. 2004, 126, 4490.
(10) (a) Coumbarides, G. S.; Eames, J.; Flinn, A.; Northen, J.; Yohannes,
Y. Tetrahedron Lett. 2005, 46, 849. (b) Coumbarides, G. S.; Dingjan, M.;
Eames, J.; Flinn, A.; Northen, J.; Yohannes, Y. Tetrahedron Lett. 2005, 46,
2897. (c) Coumbarides, G. S.; Dingjan, M.; Eames, J.; Flinn, A.; Motevalli,
M.; Northen, J.; Yohannes, Y. Synlett 2006, 101. (d) Chavda, S.; Coulbeck,
E.; Coumbarides, G. S.; Dingjan, M.; Eames, J.; Ghilagaber, S.; Yohannes,
Y. Tetrahedron: Asymmetry 2006, 17, 3386. (e) Boyd, E.; Chavda, S.;
Coulbeck, E.; Coumbarides, G. S.; Dingjan, M.; Eames, J.; Flinn, A.;
Krishnamurthy, A. K.; Namutebi, M.; Northen, J.; Yohannes, Y. Tetrahe-
dron: Asymmetry 2006, 17, 3406. (f) Chavda, S.; Coumbarides, G. S.;
Dingjan, M.; Eames, J.; Flinn, A.; Northen, J. Chirality 2007, 19, 313. (g)
Coumbarides, G. S.; Dingjan, M.; Eames, J.; Flinn, A.; Northen, J. Chirality
2007, 19, 321. (h) Boyd, E.; Chavda, S.; Eames, J.; Yohannes, Y. Tetrahe-
dron: Asymmetry 2007, 18, 476. (i) Coulbeck, E.; Eames, M. Tetrahedron:
Asymmetry 2007, 18, 2313. (j) Boyd, E.; Coulbeck, E.; Coumbarides, G. S.;
Chavda, S.; Dingjan, M.; Eames, J.; Flinn, A.; Motevalli, M.; Northen, J.;
Yohannes, Y. Tetrahedron: Asymmetry 2007, 18, 2515. (k) Coulbeck, E.;
Eames, J. Synlett 2008, 333. (l) Shaye, N. A.; Boa, A. N.; Coulbeck, E.;
Eames, J. Tetrahedron Lett. 2008, 49, 4661. (m) Chavda, S.; Coulbeck, E.;
Eames, J. Tetrahedron Lett. 2008, 49, 7398. (n) Chavda, S.; Coulbeck, E.;
Dingjan, M.; Eames, J.; Flinn, A.; Northen, J. Tetrahedron: Asymmetry
2008, 19, 1536. (o) Coulbeck, E.; Eames, J. Tetrahedron: Asymmetry 2008, 19,
2223.
(11) Kagan, H. B. Tetrahedron 2001, 57, 2449.
(12) Vedejs, E. In Separation of Enantiomers; Todd, M. H., Ed.; Wiley:
New York, 2007.
(13) Vedejs, E.; Rozners, E. J. Am. Chem. Soc. 2001, 123, 2428.
J. Org. Chem. Vol. 75, No. 14, 2010 4675

Duffey et al.
JOCFeatured Article
SCHEME 2. Catalytic PKR (Acyl Transfer)
TABLE 1. Kinetic Resolution of Alcohol 5 with TADMAP Analogues
and (iPrCO)2O a
temp time
%
entry cat. (%)
solvent
(°C) (h) conv s
1
2
3
4
5
6
7
8
18a (1%)
18b (1%)
18b (1%)
18b (1%)
18c (1%)
18d (1%)
18b (2%)
18b (4%)
toluene
toluene
tert-amyl alcohol
tert-amyl alcohol
tert-amyl alcohol
tert-amyl alcohol
rt
rt 92
rt 48
0
0
0
7
40 1.4
35 3.4
41 5.1
40 5.3
52 3.5
50 2.3
27 6.1
21 3.2
48
24
16
3:1 tert-amyl alcohol: DCM -25 37
DCM -25 28
aAll reactions used [5] = 0.1 M in the given solvent with 1-4 mol % of
catalyst, 1 equiv of isobutyric anhydride, and 1.1 equiv of Et3N.
isobutyroylation.15 Following this lead, we have studied the
competition between benzoic anhydride and isobutyric an-
hydride in the esterification of alcohol 5 using the second-
generation catalyst 8c14b (generated in situ from the HBF4
salt using Et3N) to activate a mixture of the two anhydrides
(1.5 molar equiv each relative to 5). The rate advantage for
benzoylation was reflected in the formation of a 3:1 mixture
of the benzoate 16 and the isobutyrate 17 (eq 1). To exploit
this result in PKR, the modest 3:1 reagent selectivity of 8c
would have to be improved, and a second catalyst would
have to be found that selectively activates the aliphatic
anhydride. Initial studies quickly established that isobutyr-
oylation is much faster than benzoylation using p-dimethyl-
aminopyridine (DMAP) or its derivatives as catalysts.
Indeed, a similar anhydride competition experiment with
DMAP þ 5 gave the isobutyrate ester 17 as the sole product
according to NMR assay (eq 1). Accordingly, we turned to
the evaluation of chiral DMAP derivatives16,17 to find a
catalyst having similar enantioselectivity in simple KR com-
pared to 8c, as well as high anhydride selectivity in competi-
tion experiments. These characteristics alone are not
sufficient to predict successful PKR, but they provide a
simple basis for the evaluation of potential PKR catalysts.
insoluble, activated intermediates 12 and 13 (noninterference).
The conditions are satisfied almost perfectly using the three-
phase procedure, and the PKR experiment yields the readily
separated, enantioenriched esters 14 and 15 (93% ee and 97%
ee, respectively). However, these experiments required serious
optimization and the results did not come easily.
Fully catalytic PKR using acyl transfer under homoge-
neous conditions would be more convenient and more scal-
able, but such an experiment is more challenging. By
comparison with the three-phase procedure, homogeneous
catalytic PKR confronts a greater risk of interference be-
tween quasi-enantiomeric, activated intermediates analo-
gous to 12 and 13 because there is no protection against
reversible acyl exchange if all of the reagents are soluble.
However, the most difficult problem is to find two catalysts
that will selectively activate only one of two soluble, achiral
reagents. This becomes especially challenging if the experi-
ment is designed to effect an enantiodivergent derivatization
of the racemic substrate using two chiral catalysts that
function according to similar mechanistic principles. Below,
we describe the demonstration of such an example involving
catalytic PKR acylation experiments under homogeneous
conditions.
Results and Discussion
During an investigation of chiral nucleophilic catalysts in
the phosphabicyclo[3.3.0]octane (PBO) series (8a-d),14 we
noticed that phosphines catalyze the benzoylation of alco-
hols significantly faster than they catalyze the corresponding
The first-generation chiral DMAP catalyst developed in
our laboratory (18a, TADMAP) had been optimized for
acyl migration applications and does not perform well in
the kinetic resolution of 5 (s = 1.4, Table 1, entry 1).18 In an
(16) Shaw, S. A.; Aleman, P.; Christy, J.; Kampf, J. W.; Va, P.; Vedejs, E.
J. Am. Chem. Soc. 2006, 128, 925.
(17) Wurz, R. P. Chem. Rev. 2007, 107, 5570.
(18) Shaw, S. A. Ph.D. Thesis, University of Michigan, Ann Arbor, MI,
2005.
(14) (a) Vedejs, E.; Daugulis, O. J. Am. Chem. Soc. 2003, 125, 4166.
(b) MacKay, J. A.; Vedejs, E. J. Org. Chem. 2004, 69, 6934.
(15) Harper, L. A. Ph.D. Thesis, University of Wisconsin, Madison, WI,
2001.
4676 J. Org. Chem. Vol. 75, No. 14, 2010

Duffey et al.
JOCFeatured Article
SCHEME 3. Synthesis of TADMAP Analogues
TABLE 2. Simple Kinetic Resolution of 5 and 20a,b Using Catalysts 8c,d and 18ba
entry
alcohol
R1
R2
cat. (mol %)
anhydride
time (h)
% conv
s
1
2
3
4
5
6
7
5
5
5
20a
20a
20b
20b
1-Npth
1-Npth
1-Npth
2-MeC6H4
2-MeC6H4
Ph
Me
Me
Me
Me
Me
t-Bu
t-Bu
8c (4%)
8d (4%)
18b (2%)
8d (2%)
18b (2%)
8d (2%)
18b (2%)
nicotinic
nicotinic
3
3
79
68
27
37
50
34
43
7.5
11
6.1
17
14
57
isobutyric
nicotinicb
isobutyricc
nicotinicb
isobutyricc
37
20
64
20
64
Ph
9.9
aAll reactions were conducted at -25 °C using 1 equiv of alcohol, 1 equiv of anhydride, 1.5 equiv of Et3N, and 2-4 mol % of catalyst and were diluted
to 0.1 M substrate in 3:1 tert-amyl alcohol/DCM unless noted. b0.6 equiv of anhydride. c2.5 equiv of anhydride.
effort to improve enantioselectivity, 18a was modified by
replacement of the acetate by other carboxylates. Enantio-
enriched 18a was easily cleaved to the alcohol 19 by reduc-
tion, but the hindered 19 was resistant to esterification with
anhydrides. Strongly basic conditions were problematic due
to facile trityl anion cleavage at the alkoxide stage,16 but the
magnesium alkoxide generated by treating 19 with EtMgBr
reacted normally with anhydrides to afford the desired esters
18b-d (Scheme 3). The new esters were screened as catalysts
for the kinetic resolution of racemic alcohol 5, and the ben-
zoate ester 18b proved to be somewhat more enantioselective
(s = 3.4, Table 1, entry 2) compared to 18a. Enantioselec-
tivity was higher in tert-amyl alcohol compared to toluene
(entry 3), and a further increase was observed at 0 °C
(entry 4). The isobutyrate or formate analogues (18c,d) were
less enantioselective (entries 5 and 6), so benzoate 18b was
selected for detailed optimization. Access to lower tempera-
tures was limited by the freezing point of tert-amyl alcohol,
but addition of DCM as cosolvent allowed KR at -25 °C,
resulting in s = 6.1 (entry 7). The higher s value is due to
the temperature, and not the cosolvent (s = 3.2 in DCM;
entry 8). This improved level of enantioselectivity for catalyst
18b in entry 7 was deemed sufficient for proof of principle
PKR experiments if the other experimental criteria for PKR
could be satisfied.
Although the chiral phosphine 8c catalyzed the acylation
of 5 with complementary anhydride selectivity in favor of
Bz2O over (i-PrCO)2O compared to 18b, the ratio of pro-
ducts (16/17 = 3:1) was too low for PKR. Several electron-
rich aromatic anhydrides were evaluated but did not improve
the results with alcohol 5 using catalyst 8c in toluene solu-
tion. Thus, the competition between isobutyric anhydride vs
(ArCO)2O gave modest product ester ratios (arylcarboxylate
vs isobutyrate 17) as follows: Ar = p-methoxyphenyl, 1:1.9;
Ar = m-methoxyphenyl, 2.7:1; Ar = 3,5-dimethoxyphenyl,
1.5:1). A more promising anhydride selectivity was observed
using phosphine catalyst 8c with the electron-deficient
m-chlorobenzoic anhydride vs isobutyric anhydride (chloro-
benzoate/isobutyrate 17 = 18:1). However, m-chlorobenzoic
anhydride gave disappointing selectivity in the complemen-
tary competition experiment with the DMAP-derived cata-
lyst 18b (chlorobenzoate/isobutyrate 17 = 1:6). Finally, the
competition experiment was repeated using nicotinic an-
hydride as an electron-deficient analogue of the m-chloro-
benzoic anhydride. Under the conditions already optimized
for simple KR (Table 1, entry 7; t-amyl alcohol/DCM at
-25 °C), the competition between nicotinic vs isobutyric
anhydrides with phosphine 8d as catalyst afforded only the
nicotinate ester. On the other hand, the same anhydride
competition using the complementary DMAP-derived cata-
lyst 18b again gave modest, but opposite anhydride selec-
tivity (nicotinate/17 = 1:6). The match of complementary
anhydride selectivities in the nicotinate experiments was not
as close as desired for PKR, but a potential solution to the
problem of finding two sets of complementary catalysts and
anhydrides appeared to be in hand.
Attention was now turned to optimizing the simple kinetic
resolutions of representative racemic alcohol substrates
using individual components of potential complementary
catalyst/anhydride pairings. Under the best conditions,
moderate enantioselectivity factors were found with both
the phosphine- and the DMAP-derived catalysts for KR of
benzylic alcohols 5, 20a, and 20b (Table 2). The simplest
P-phenyl catalyst 8d was more enantioselective than the
hindered analogue 8c in tert-amyl alcohol/DCM (entry 2
vs 1), so the other phosphine-catalyzed experiments used 8d.
Good complementarity was found in the acylation experi-
ments with alcohols 5 and 20a, as judged by similar enantio-
selectivities with each individual catalyst of the comple-
mentary pair. However, alcohol 20b was a much better
substrate for nicotinic anhydride activated by phosphine 8d
compared to the complementary reagent combination of
isobutyric anhydride activated by the DMAP-derived 18b
(entries 6 vs 7).
According to the preliminary experiments, phosphine 8d
preferentially converts (R)-5 into the corresponding nico-
tinic ester 21, while DMAP derivative 18b converts the
(S)-alcohol into isobutyrate 17 (Table 3). These results define
J. Org. Chem. Vol. 75, No. 14, 2010 4677

Duffey et al.
JOCFeatured Article
TABLE 3. Parallel Kinetic Resolutions of Alcohols Using Catalysts 8d and 18ba
catalyst (mol %)
ester ratio
nicotinate 21
butyrates 17/ 22
entry
alcohol
R1
R2
8d
18b
Nic/ iBu
yield (%)
ee (%)
yield (%)
ee (%)
1
2
3
4
5
5
5
5
20a
20b
1-Npth
1-Npth
1-Npth
2-MeC6H4
Ph
Me
Me
Me
Me
t-Bu
(4.5)
(1.3)
(1.3)
(6.0)
(6.0)
(6.7)
(6.7)
(6.7)
(3.0)
(5.0)
1.6:1.0
1.1:1.0
1.0:1.1b
1.2:1.0
1.0:1.9
62
51
44b
52
33
62
65
68
68
79
38
46
46b
44
60
-92
-73
-63
-82
-41
aReactions used 1 equiv of alcohol, 1 equiv of each anhydride, 1.5 equiv of Et3N, and 0.1 M substratein 3:1 tert-amyl alcohol/DCM unlessnoted. b0.75
equiv of nicotinic anhydride.
two viable complementary kinetic resolutions, so the key
PKR experiments were conducted using both anhydrides
and both chiral catalysts simultaneously. Catalyst 18b is
considerably less reactive than 8d, so it was necessary to
use a ca. 5:1 ratio of 18b/8d to obtain the ca. 1:1 ratio of
products 21 and 17 (entry 2 vs entry 1) as desired for optimal
PKR. A further improvement in the product ratio was pos-
sible by decreasing the amount of nicotinic anhydride to 0.75
equiv (entry 3). However, these refinements did not result in
ideal behavior in the PKR experiments, even though the
product ratio was nearly 1:1. Given the enantioselectivities in
simple KR (Table 2, entries 2 and 3; s = 11 for 21 and s = 6.1
for 17) isobutyrate ester 17 should have been obtained with
ca. -72% ee in the experiment of Table 3, entry 3, while
nicotinate 21 should be formed with an ee value above ca.
80%. The lower values found for Table 3, entry 3 (-63% ee
for 17; 68% ee for 21) indicate that one or more of the
conditions for PKR was not fully satisfied.
At first glance, it may seem surprising that 17 was obtained
with -92% ee in Table 3, entry 1. However, this outcome is
not due to a “better” PKR experiment in entry 1. Instead, it is
the consequence of the nonideal product ratio (62:38 21/17)
resulting from the greater reactivity of the phosphine catalyst
8d compared to 18b. According to Horeau’s generalization
relating enantiomeric purity with the mol % of products
resulting from two divergent reactions starting from a race-
mic mixture,19 two products obtained in similar proportions
(near 1:1) should have nearly identical ee values. In contrast,
if one product is obtained in lower proportion, then its ee
value increases compared to the major product. This latter
trend is evident in Table 3, where lower ee is observed for the
product obtained in higher yield with each of the three
substrates (compare entries 3-5). According to the Horeau
criterion, entry 3 comes closest to satisfying the requirements
for PKR because product yields and ee values are similar. As
already mentioned in connection with entry 1, the relatively
high ee value for 21 in entry 5 results from the nonideal
product ratio.
At this stage of our investigation, another generation of
chiral DMAP derivatives became available.20 The catalysts
23a and 23b reported earlier were found to activate a mixture
of isobutyric and m-chlorobenzoic anhydrides with the usual
rate advantage for isobutyrate ester formation. However, the
faster reacting benzylic alcohol substrate enantiomers proved
to have the (R)-configuration, the same as for the phosphine
catalyst 8d. To obtain the complementary chiral DMAP
catalysts, the published routes were repeated starting from
the enantiomeric valinol and tert-leucinol, affording ent-23a
and ent-23b, respectively. Standard evaluations were carried
out, and both ent-23a and ent-23b were found to have promis-
ing enantioselectivity and the expected preference for the (S)
enantiomer of 5, but most of the preliminary studies used
catalysts 23a and 23b prepared from the relatively inexpensive
L-tert-leucinol.
At room temperature, KR of alcohol 5 using the valine-
derived 23a with isobutyric anhydride was more enantiose-
lective in toluene (s = 6.8)18 than in tert-amyl alcohol (s =
2.7; Table 4, entries 1 and 2). The more hindered tert-leucine-
derived 23b gave somewhat better results in toluene (s = 8.9,
(19) Guette, J. P.; Horeau, A. Bull. Soc. Chim. Fr. 1967, 1747.
4678 J. Org. Chem. Vol. 75, No. 14, 2010
(20) Duffey, T. A.; Shaw, S. A.; Vedejs., E. J. Am. Chem. Soc. 2009, 131, 14.

Duffey et al.
JOCFeatured Article
entry 3),18 and selectivity increased to synthetically useful
levels in the temperature range from -40 to -70 °C (s =
20-29, entries 4 and 5). At the lower end of this range, the rate
of acylation decreased dramatically, so further experiments
were conducted at -40 °C. Alcohol 20a was also isobutyr-
oylated with moderate enantioselectivity (entry 6), but the
more hindered tert-butyl-substituted alcohol 20b reacted with
minimal enantiomer discrimination (entry 7; s = 1.8).
Given the promising enantioselectivity in simple KR using
the DMAP-derived catalysts 23a and 23b, the anhydride
selectivity was tested in the usual competition experiments
with isobutyric vs m-chlorobenzoic or nicotinic anhydrides.
Formation of the isobutyrate over the m-chlorobenzoate or
nicotinate esters was favored by ratios of 16:1 and 11:1,
respectively, using 23a (toluene, rt). These selectivities are
significantly improved compared to the corresponding va-
lues observed using 18b, but the best KR results with catalyst
23b were obtained in toluene, while prior experiments with
18b and the phosphine catalyst 8d had been conducted in
tert-amyl alcohol. Fortunately, simple KR of alcohol 5 with
m-chlorobenzoic anhydride activated by 8d proved to be
even more enantioselective in toluene (s = 19.8, -40 °C).
These observations quickly displaced nicotinic anhydride as
a reagent of interest and established m-chlorobenzoic anhy-
dride as a viable alternative for PKR experiments.
phosphine catalyst 8d, the new catalyst combination was
evaluated for PKR of alcohol 5. In contrast to 18b, ent-23b
was more reactive than 8d, and it was necessary to increase
the proportion of 8d to equalize the rates for conversion of
each enantiomer. Using a 2.2:1 ratio of 8d/ent-23b, a 1:1.2
ratio of m-chlorobenzoate 24 to isobutyrate 17 was obtained
(ca. 92% conversion after 3 h; Table 5, entry 1). After
isolation and saponification of the esters, the alcohols were
assayed by HPLC on chiral support. The alcohol derived
from m-chlorobenzoate 24 was obtained in 88% ee and the
alcohol from isobutyrate 17 in -75% ee.
The unreacted alcohol recovered from the above experi-
ment was found to have 27% ee, indicating that one of the
enantiomers had been consumed faster than the other. To
compensate, the ratio of 8d/ent-23b was increased to 2.6:1,
and afforded the ideal 1:1 ratio of products (NMR assay).
However, this did not significantly alter the product ee, even
though the unreacted alcohol (-8% ee) was closer to the
desired 0% ee for an ideal PKR experiment (entry 2).
Doubling the catalyst loading resulted in a small increase
in isobutyrate ee (17; -78% ee) at the expense of m-chloro-
benzoate ee (24; 86%; entry 3), but unreacted alcohol 5
recovered from this experiment was obtained with a nonideal
value of 19% ee (0% ee is ideal). All of the PKR experiments
in Table 5 generated 8d in situ from the HBF4 salt as
precatalyst via deprotonation with Et3N. This procedure
provides better control of catalyst loading and minimizes the
risk of phosphine oxidation to the catalytically inactive
phosphine oxide.14b,21 Despite these precautions, the unex-
pected variations in recovered alcohol ee could not be
suppressed, and the contrast with the relatively consistent
product ester ee values could not be explained.
With similar and complementary enantioselectivities as
well as anhydride selectivities demonstrated for ent-23b vs
TABLE 4. KR of Alcohols with Isobutyric Anhydride Using 23a,ba
entry alcohol
R1
R2 catalyst
solvent
temp (°C)
s
1
2
3
5
5
5
5
5
1-Npth
1-Npth
1-Npth
1-Npth
1-Npth
Me
Me
Me
Me
Me
23a tert-amyl alcohol
23a toluene
23b toluene
23b toluene
23b toluene
rt
rt
rt
2.7
6.8
8.9
Under ideal PKR conditions, the product corresponding
to simple KR with s = 20 should be obtained with 90% ee
(calculated ee at <1% conversion in simple KR), so the
recovery of m-chlorobenzoate ester 24 with 86-88% ee is
near-optimal. If the corresponding simple KR experiment
were taken to 50% conversion, then 24 would be formed with
79% ee, indicating that PKR has significantly enhanced
4
-40 20.5
-70 29
-40
-40
5b
6
20a 2-MeC6H4 Me
20b Ph
23b toluene
t-Bu 23b toluene
9.5
1.8
7
aAll reactions used 1 equiv of alcohol, 1 equiv of isobutyric anhydride,
1.5 equiv of Et3N, 1 mol % of catalyst, 0.06 M substrate, 4 h; catalyst
23a,b consistently gave the R ester. b8 h reaction time.
TABLE 5. Parallel Kinetic Resolutions of Alcohols with Catalysts 8d and ent-23ba
entry
8d (mol %)
ent-23b (mol %)
time (h)
ratio 24/17/5b
24 (% ee)
17 (% ee)
5 (% ee)
1
2
3
2.2
2.6
5.2
1
1
2
3
3
2
1.0:1.2:0.18
1.0:1.0:0.5
1.0:1.2:0.53
88
-75
-76d
-78
27
-8
19
87c
86
aAll reactions used 1 equiv of 5, 1 equiv of each anhydride, 1.5 equiv of Et3N, and 1-5.2 mol % of each catalyst and were diluted to 0.06 M substrate in
toluene. bBy NMR assay. c44% isolated. d33% isolated.
J. Org. Chem. Vol. 75, No. 14, 2010 4679

Duffey et al.
JOCFeatured Article
SCHEME 4. Intermediates in Catalyzed Acyl Transfer
TABLE 6. Mixed Anhydride Activationa
entry
cat. (mol %)
Ar
R
product ratio
enantiomeric purity. However, the ideal PKR conditions
were not fully satisfied because 17 was obtained with -75 to
-78% ee, lower than the expected -90% ee and essentially
the same as the value predicted from simple KR at 50%
conversion (s = 20.5). Furthermore, recovery of alcohol 5
with significant ee indicates that at least one of the PKR
prerequisites had not been fully controlled.
1
2
3
4
DMAP (20)
8c (1)b
Ph
Ph
biphenyl
biphenyl
i-Pr
i-Pr
cyclohexyl
cyclohexyl
1:50 16/17
1:3 16/17
1:4 30/31
1:50 30/31
8d (5)b
18b (5)
aAll reactions used 1 equiv of 5, 1-2 equiv of each anhydride (1:1
ratio), and 1.5 equiv of Et3N and were diluted to 0.2 M substrate in
toluene at room temperature. bThe phosphine was generated in situ from
the HBF4 salt; an additional 0.1 equiv of Et3N was added.
In the PKR experiments of Table 5, the m-chlorobenzoate
24 was obtained in nearly ideal enantiomeric excess, suggest-
ing that the formation of 24 occurs cleanly by the intended
phosphine catalysis resulting from activation of m-chloro-
benzoic anhydride as desired. If this deduction is correct,
then problems in the isobutyric anhydride activation path-
way could be responsible for the observed nonideal behavior,
but it was difficult to imagine how one of two mechanistically
similar pathways might become compromised without af-
fecting the other pathway. On the other hand, there was
no mystery regarding the most delicate stage of the process
where reagent interference might occur. The anhydride acti-
vation stage leading to ion-pair intermediates 25 and 26
was the likely culprit (Scheme 4) because two different
carboxylate anions must be present simultaneously. In prin-
ciple, this would allow carboxylate exchange between the ion
pairs 25 and 26 and would open the door to reversible acyl
transfer, equilibration, and formation of the mixed anhy-
dride 27. Indeed, a simple NMR control experiment in
benzene established that both DMAP and phosphine 8d
promote the scrambling of a 1:1 mixture of benzoic and
isobutyric anhydrides to form a statistical mixture of the
mixed and symmetrical anhydrides within ca. 1 h at room
temperature. Initially, we saw no harm in anhydride scram-
bling and even considered the possibility that mixed anhy-
drides such as 27 or 28 may be sufficient reagents for PKR if
selective activation of the two different acyl groups would
occur with the same selectivity as observed in competition
experiments with the symmetrical anhydrides. However, our
attempts to probe this possibility encountered a very differ-
ent scenario.
internal competition experiment with phosphine 8c as cata-
lyst gave the same isobutyrate 17 as the major product, in
striking contrast to the result of eq 1. At first, we suspected
that this outcome might be an artifact resulting from inter-
vention by contaminants in the mixed anhydride 28 (ca. 5%
of isobutyric anhydride was present). Accordingly, a search
was launched to find a crystalline mixed aryl-aliphatic
anhydride that might be obtained in better purity, and this
search lead to the mixed anhydride 29. To minimize the
possibility of contamination, 29 was used immediately upon
crystallization, and was tested with the “real” catalysts 8d
and 18b (Table 6, entries 3 and 4). Once again, the aliphatic
(cyclohexanecarboxylate) ester was the major product with
both the phosphine- and the DMAP-based catalysts.
The evidence discussed above indicates that formation of
mixed anhydride 27 is likely during the PKR experiments.
This would be detrimental to enantiomeric excess of the
isobutyrate product 17 because an additional fraction of 17
would be formed as the minor enantiomer due to the activa-
tion of 27 by the phosphine catalyst. If this undesired path-
way is responsible for the formation of ca. 8% of the
isobutyrate ester 17 in Table 5, then all of the observed
deviations from ideal PKR behavior with catalyst 8d can be
attributed to mixed anhydride formation. Analogous devia-
tions are not observed in the m-chlorobenzoate product 16
because mixed anhydride activation by the DMAP-derived
ent-23b follows the same pattern of acyl discrimination as
with the symmetrical anhydrides.
The origins of complementary anhydride selectivity for
the phosphine vs the DMAP-derived catalysts remain to be
considered. Nucleophiles activate anhydrides by forming ion
pairs (32 or 33, eq 2), and the rate of acylation depends
on Keq = k1/k-1 (extent of ion pair formation) and also on
k2 (ion pair reactivity).22 For the phosphine catalysts,
Early experiments with mixed anhydrides tested 28, gen-
erated from sodium benzoate and isobutyroyl chloride. As
expected, the presumed internal competition for acyl activa-
tion with DMAP as catalyst gave the same preference for the
aliphatic (isobutyrate) ester 17 (Table 6, entry 1) as observed
under conditions of intermolecular competition using the
symmetrical anhydrides (eq 1). However, the corresponding
(22) (a) Heinrich, M. R.; Klisa, H. S.; Mayr, H.; Steglich, W.; Zipse, H.
Angew. Chem., Int. Ed. Engl. 2003, 42, 4826. (b) Xu, S.; Held, I.; Kempf, B.;
Mayr, H.; Steglich, W.; Zipse, H. Chem.;Eur. J. 2005, 11, 4751. (c) Held, I.;
Villinger, A.; Zipse, H. Synthesis 2005, 9, 1425. (d) Fischer, C. B.; Xu, S.;
Zipse, H. Chem.;Eur. J. 2006, 12, 5779.
(21) Netherton, M. R.; Fu, G. C. Org. Lett. 2001, 3, 4295.
4680 J. Org. Chem. Vol. 75, No. 14, 2010

Duffey et al.
JOCFeatured Article
anhydride selectivity increases as the leaving group ability
of the carboxylate anion increases. This trend is evident
in the improved intermolecular competition using benzoic,
m-chlorobenzoic, and nicotinic anhydrides vs isobutyric
anhydride as discussed previously, and it is also apparent
in the mixed anhydride experiments. Two ion pairs 32 and 33
are possible (Nu: = phosphine) in experiments using mixed
anhydrides, and preferential formation of the aliphatic ester
17 indicates that product formation via ion pair 32 is favored
over ion pair 33. A comparison of the favored ion pair 32
from the mixed anhydrides with the favored ion pair 25 from
the two symmetrical anhydrides in phosphine-catalyzed
acylation reveals that the common element is the arylcarbox-
ylate leaving group. Since this is the more stable carboxylate
anion, preferential formation of 17 can be attributed to an
increase in the value of k1R/k-1R for the equilibrium with 32
compared to the analogous term k1Ar/k-1Ar in the competing
pathway via 33.
analogous transition state, or the related phosphoranes can
be considered. All of these species could play a role, but some
version of carboxylate-assisted nucleophilic attack by the
alcohol substrate at the carbonyl group of ion pair 25 is
necessary. This event would be resisted by C sp3 vs P sp3
repulsions between the quaternary phosphorus and the
developing quaternary character at the isobutyroyl CdO
carbon regardless of mechanistic details. In contrast, the
analogous event in DMAP-derived intermediates such as 26
would involve a less demanding C sp3 vs N sp2 repulsion.
The simplest version of the steric argument is not consistent
with the mixed anhydride experiments because the aliphatic
(isobutyrate) ester is favored with phosphine as well as
DMAP catalysts. On the other hand, no simple argument
can address the role of carboxylate anions on k2R vs k2Ar at
the current stage of mechanistic understanding. Certainly,
the anions are important for enantioselectivity,23 and it
would be premature to assume that their contributions to
the relative rate terms would follow a simple pattern.
Summary
We have presented several proof of principle examples of
fully catalytic parallel kinetic resolution under homogeneous
conditions. These experiments show that it is possible to use
complementary chiral catalysts for the selective activation of
two different achiral acyl donors in solution, resulting in the
formation of distinct enantioenriched esters. The best experi-
ments involve the simultaneous, selective generation of
activated ion pair intermediates 25 and 26 from a mixture
of isobutyric and m-chlorobenzoic anhydrides. Although
complementary anhydride selectivity by the catalysts 8d
and ent-23b is compromised by minor formation of the
undesired mixed anhydride 27, it is clear that the quasi-
enantiomeric24 intermediates 25 and 26 discriminate quite
well between alcohol substrate enantiomers. Near-ideal en-
antiomeric excess was observed for (R)-24, the product of
enantioselective aroylation in the PKR experiment. How-
ever, the complementary process leading to the isobutyrate
(S)-17 encountered interference by a second (minor) path-
way, the activation of mixed anhydride 27 by phosphine 8d.
This undesired pathway generates ion pair 32 (Nu: = 8d), an
intermediate that functions as the quasi-enantiomer of 26
and is responsible for increased product contamination by
the enantiomer (R)-17.
A similar analysis of the DMAP-catalyzed acyl transfer
reactions leads to a different conclusion. Since DMAP
derivatives selectively catalyze the isobutyroylation of alco-
hols regardless of the isobutyroyl donor anhydride, ion pairs
26 and 32 are the dominant isobutyroylating agents. The
common element between 26 and 32 is an acylated pyridi-
nium ion, while the carboxylate anions are different. In this
scenario, selectivity for the DMAP catalysts is determined
more by the relative reactivity of the intermediate ion pairs
(as in 32 vs 33) and the corresponding values of k2R
(isobutyroylation) compared to k2Ar (aroylation). We do
not have sufficient evidence to comment in detail on the
factors that would favor k2R vs k2Ar for the DMAP-derived
catalysts, but a simple rationale can be invoked that is
consistent with the data. Somewhat increased delocalization
is expected for the N-aroylpyridinium subunit in the ion
pair 33 vs 32 (Nu: = DMAP), a factor that would decrease
k2Ar and would result in a slower aroylation compared to
isobutyroylation. The lower anhydride selectivity (aromatic
vs isobutyric) for DMAP-derived catalysts with the relatively
electron poor aromatic anhydrides (nicotinic, m-chloro-
benzoic) compared to electron-rich aromatic anhydrides
discussed earlier supports this rationale. As the π-electron
donor ability of Ar decreases, the N-aroylpyridinium inter-
mediate 33 (Nu = DMAP) has a smaller advantage com-
pared to 32 (Nu = DMAP).
Further studies will be needed to control the behavior of
mixed acyl transfer agents analogous to 27 or to develop
methods that use more robust and more compatible acti-
vated intermediates compared to 25 and 26. Many options
can be considered for this purpose. One possibility would be
to develop acyl donors having more stabilized anionic leav-
ing groups in place of the carboxylate anions to minimize
(23) In one of our attempts to avoid the problem of mixed anhydride
formation, we considered using two activated esters having the same anionic
leaving group attached to different acyl groups. The N-isobutyroyl and N-
benzoyl derivatives of 4-phenyltetrazole were evaluated for this purpose.
Each tetrazolide was activated for the usual esterifications of 5 by the
DMAP-derived catalyst 18b and the phosphine 8d, respectively, in 3:1 tert-
amyl alcohol/DCM at -25 °C. However, 8d gave s = 2.5 for the benzoyla-
tion, while 18b afforded racemic products (s = 1) in the isobutyroylation.
These experiments hint at the variety of options that may be considered for
improving PKR, but they caution against simple assumptions regarding the
role of anions in the product-determining step for acyl-transfer experiments.
(24) For a recent discussion of relevant concepts and terminology, see:
Zhang, Q.; Curran, D. P. Chem.;Eur. J. 2005, 11, 4866.
In principle, the reversed anhydride selectivity for the
phosphine catalysts with the symmetrical anhydrides may
also reflect a contribution from steric effects22d that might
decrease k2R for the isobutyroylations relative to k2Ar for
the competing aroylations. Although mechanistic details
for the acyl transfer stage remain uncertain, pathways from
25 involving a conventional tetrahedral intermediate, the
J. Org. Chem. Vol. 75, No. 14, 2010 4681

Duffey et al.
JOCFeatured Article
nucleophile-induced mixed anhydride formation.23 Another
alternative would be to design PKR experiments where one
of the parallel KR pathways is mechanistically different. Our
first proof of principle examples were chosen to tackle the
most difficult scenario involving parallel, mechanistically
similar pathways where reagent compatibility can be com-
promised by acyl exchange between ion pairs 25 and 26.
Future versions of fully catalytic PKR might improve re-
agent compatibility by replacing one of the two parallel
acylations with an enantiodivergent process that incorpo-
rates distinct functionality.
dissolved in DCM (1.5 mL). The solution was cooled to 0 °C,
and EtMgBr (0.85 M in Et2O, 0.54 mL, 0.46 mmol) was added
via syringe. The cooling bath was removed, and the mixture was
stirred for 1 h at rt. The mixture was recooled to 0 °C, and
benzoic anhydride (208 mg, 0.92 mmol) was added as a solution
in DCM (0.4 mL) via syringe. The solution immediately turned
yellow, was stirred for 5 min, and was warmed to room tempera-
ture. The mixture was then stirred for an additional 2.5 h,
and the solution turned orange. The reaction was quenched
with H2O (1 mL) followed by the addition of NaHCO3 (satd,
10 mL) and extracted with DCM (4 ꢀ 10 mL). The combined
organic layers were washed with NaHCO3 (satd, 10 mL), dried
(MgSO4), and concentrated (aspirator). The crude product was
purified by flash chromatography (3 ꢀ 14 cm) using 4:1 EtOAc/
hexanes with a 1% Et3N buffer as eluent and collecting 9 mL
fractions. Impurities eluted in fractions 6-9. Fractions 10-17
were combined to yield 97 mg (63%) of 18b, a pale yellow foam.
HPLC assay was carried out on a Chiralpak AD analytical
column, 5% isopropanol/hexanes, 1 mL/min flow rate. Reten-
tion times of enantiomers: 11.1 min (S), 20.0 min (R), 95% ee;
HRMS calcd for C34H30N2O2 [M þ H]; ESMS m/z = 499.2386,
found 499.2363; IR (neat, cm-1) 1715, CdO; 1585, CdN; 1265,
CO; 1H NMR (500 MHz, CDCl3, ppm) δ 8.17 (1H, d, J = 5.9
Hz) 8.01 (1H, s) 7.88 (2H, d, J = 7.3 Hz) 7.51 (1H, t, J = 7.6 Hz)
7.40-7.34 (9H, m) 7.26-7.16 (9H, m) 6.73 (1H, d, J = 5.4 Hz)
2.79 (6H, s); 13C NMR (125.7 MHz, CDCl3, ppm) δ 166.2,
159.5, 152.7, 149.9, 143.5, 133.4, 131.1, 130.3, 129.9, 128.7,
126.8, 126.5, 114.3, 73.1, 64.8, 43.8, 27.5.
Experimental Section
Et2O, THF and CH2Cl2 (DCM) were dried by passing through
a column of activated alumina. Ethyl acetate and Et3N were
distilled over CaH2, Et2NH was distilled over KOH, and tert-amyl
alcohol was carefully distilled over molten sodium. Solvents used
for reactions involving phosphines were further deoxygenated by
bubbling a stream of N2 through the solvent (∼30 min) prior
to use. Organolithium reagents were titrated with diphenylacetic
acid in THF. All other reagents were used as received from the
manufacturer. Analytical thin layer chromatography (TLC) was
˚
done using 0.25 mm K6F silica gel 60 A plates. Flash chromato-
graphy followed the Still procedure25 using silica gel/Puracil 60 A
˚
(230-400 mesh). All reactions were performed under an atmo-
sphere of nitrogen in oven-dried glassware. Catalysts 23a and
23b were prepared as previously described,20 and the same route
was used to prepare the previously unreported enantiomeric cata-
lysts ent-23a and ent-23b from (R)-N-benzoylvalinol and (R)-N-
benzoyl-tert-leucinol, respectively. Catalysts ent-23a,b were puri-
fied by recrystallization (>98% dr, NMR assay) as previously
described for 23a,b. As expected, the NMR data for ent-23a,b
are identical to those reported previously for 23a,b.20 Nicotinic
anhydride and m-chlorobenzoic anhydride were prepared accord-
ing to prior literature.28,29
Isobutyric Acid 1-(4-Dimethylaminopyridin-3-yl)-2,2,2-triphe-
nylethyl Ester (18c). Following the representative procedure for
the synthesis of TADMAP derivatives, alcohol 19 (26 mg, 0.066
mmol) was protected as the isobutyrate ester 18c (8 mg, 29%) at
55% conversion using isobutyric anhydride (31 μL, 0.19 mmol):
1H NMR (400 MHz, CDCl3, ppm) δ 8.16 (1H, d, J = 5.4 Hz)
7.70 (1H, s) 7.34 (1H, s) 7.30-7.27 (6H, m) 7.23-7.17 (9H, m)
6.71 (1H, d, J = 5.4 Hz) 2.74 (6 H, s) 2.45 (1H, qq, J = 7.3, 6.8
Hz) 0.99 (1H, d, J = 6.8 Hz) 0.96 (1H, d, J = 7.3 Hz).
Formic Acid 1-(4-Dimethylaminopyridin-3-yl)-2,2,2-triphenyl-
ethyl Ester (18d). Following the representative procedure for the
synthesis of TADMAP derivatives, alcohol 19 (24 mg, 0.062
mmol) was protected as the formate ester 18d (13 mg, 45%) at
56% conversion using formyl pivaloyl mixed anhydride (17 μL,
0.19 mmol): 1H NMR (400 MHz, CDCl3, ppm) δ 8.21 (1H, d,
J = 6.1 Hz) 8.01 (1H, s) 7.84 (1H, s) 7.47 (1H, s) 7.31-7.27
(6H, m) 7.24-7.19 (9H, m) 6.74 (1H, d, J = 6.1 Hz) 2.68 (6 H, s).
General Procedure for Kinetic Resolutions Using DMAP
Derivatives (A): Kinetic Resolution of 5 Using 18b. A solution
of alcohol 5 (17 mg, 0.1 mmol), 18b (0.5 mg, 0.001 mmol, 91%
ee), and Et3N (15 μL, 0.1 mmol) in toluene (0.8 mL) was treated
with isobutyric anhydride (17 μL, 0.1 mmol). The reaction was
monitored by TLC to an estimated 50% conversion and then was
quenched by addition of i-PrNH2 (0.1 mL). The mixture was
concentrated (aspirator) and purified by flash chromatography.
Ester fractions were concentrated, and 5% NaOH/MeOH
(1 mL) was added to saponify the ester prior to assay. The
solution was warmed gently for 5 min and then left at room
temperature for 2 h. Methanol was evaporated, and the residue
was filtered through an 8 cm ꢀ1.2 cm pad of silica gel in DCM.
After solvent removal (aspirator), HPLC assay was carried out
on a chiral support. Conversion and selectivity of the kinetic
resolution were calculated by a best fit method using both the ee
of the recovered alcohol and ee of the hydrolyzed ester, correct-
ing for the ee of the catalyst.26
1-(4-Dimethylaminopyridin-3-yl)-2,2,2-triphenylethanol (19).
To a solution of (R)-TADMAP (18a)16 (175 mg, 0.41 mmol,
96.4% ee) in toluene (4.1 mL) cooled to 0 °C was added DIBAL
(0.89 mL, 1 M solution in toluene) via syringe. The mixture was
stirred at 0 °C for 2 h and then quenched with aqueous sodium
potassium tartrate (satd, 4 mL) and stirred vigorously for 2 h
to break up the emulsion. An additional 2 mL of saturated
aqueous sodium potassium tartrate was added along with
EtOAc (10 mL) and 10% HCl (0.5 mL). The layers were
separated, and the aqueous layer was extracted with EtOAc
(3 ꢀ 10 mL). The combined organic layers were washed with
saturated NaHCO3 (20 mL), and the aqueous later was ex-
tracted with EtOAc (10 mL). All aqueous layers were combined
and extracted with DCM (2 ꢀ 10 mL). The organic layers were
combined, dried (MgSO4), and concentrated (aspirator) to give
123 mg (78%) of a white solid (19) that was used in the sub-
1
sequent reaction without further purification: H NMR (500
MHz, CDCl3, ppm) δ 8.26 (1H, d, J= 5.3 Hz) 7.56 (1H, s)
7.38-7.33 (6H, m) 7.27-7.16 (9H, m) 6.88 (1H, d, J = 5.4 Hz)
6.75 (1H, s) 5.17 (1H, br s) 2.55 (6H, s); 13C NMR (101 MHz,
CDCl3, ppm) δ 160.4, 151.1, 149.3, 144.5, 130.6, 129.9, 127.7,
126.4, 115.0, 73.1, 64.8, 44.5.
Representative Procedure for the Synthesis of TADMAP
Derivatives: Benzoic Acid 1-(4-Dimethylaminopyridin-3-yl)-2,2,
2-triphenylethyl Ester (18b). Alcohol 19 (121 mg, 0.31 mmol) was
General Procedure for Kinetic Resolutions Using Phosphine
Derivatives (B): Kinetic Resolution of 5 Using 8c. A solution of
alcohol 5 (17 mg, 0.1 mmol) and nicotinic anhydride (23 mg,
0.01 mmol) in tert-amyl alcohol (0.5 mL) and DCM (0.25 mL)
was cooled to -25 °C in a Cryocool. To the mixture was added
(25) Still, W. C.; Kahn, M.; Mitra, A. J. Org. Chem. 1978, 43, 2923.
(26) Ismagilov, R. F. J. Org. Chem. 1998, 63, 3772.
(27) Ruble, J. C.; Tweddell, J.; Fu, G. C. J. Org. Chem. 1998, 63, 2794.
(28) Badgett, C. O. J. Am. Chem. Soc. 1947, 69, 2231.
(29) Rule, H. G.; Paterson, T. R. J. Chem. Soc., Trans. 1924, 125, 2155.
4682 J. Org. Chem. Vol. 75, No. 14, 2010

Duffey et al.
JOCFeatured Article
a solution of 8c HBF4 (2 mg, 0.004 mmol) and Et3N (22 μL,
(R1=1-naphthyl, R2= Me): HRMS calcd for C18H15NNaO2
3
0.15 mmol) in tert-amyl alcohol (0.25 mL). The reaction was
stirred for 3 h and then was quenched by addition of i-PrNH2
(0.15 mL). The mixture was concentrated (aspirator) and puri-
fied by flash chromatography. Ester fractions were concen-
trated, and 5% NaOH/MeOH (1 mL) was added to saponify
the ester prior to assay. The solution was warmed gently for
5 min and then left at room temperature for 2 h. Methanol was
evaporated, and the residue was filtered through an 8 cm ꢀ
1.2 cm pad of silica gel in DCM. After solvent removal (aspi-
rator), HPLC assay was carried out on a chiral support. Con-
version and selectivity of the kinetic resolution were calculated
by a best fit method using both the ee of the recovered alcohol
and ee of the hydrolyzed ester, correcting for the ee of the
catalyst.26
Table 1, Entry 2: Isobutyroylation of 5 with 18b. Following the
general procedure A, racemic 5 (17 mg, 0.1 mmol), isobutyric
anhydride (17 μL, 0.1 mmol), and catalyst 18b (0.5 mg, 0.001
mmol, 91% ee) in toluene were stirred at rt for 92 h to afford
unreacted alcohol R-5 22.3% ee and isobutyrate S-17 40.7%
ee (after hydrolysis); s = 3.426 and c = 35%. HPLC analysis:
CHIRALCEL OD analytical column, 10% 2-propanol/hexanes,
1 mL/min flow rate. Retention times 9.8 min (S), according
to (-) sign of optical rotation,27 15.0 min (R).
[M þ Na]; ESMS m/z = 300.0995, found 300.0997; IR (neat,
cm-1) 1717, CdO; 1590, CdN; H NMR (500 MHz, CDCl3,
1
ppm) δ 9.32 (1H, d, J = 2.2 Hz) 8.78 (1H, dd, J = 4.7, 1.5 Hz)
8.34 (1H, dt, J = 8.1, 2.2 Hz) 8.19 (1H, d, J = 8.8 Hz) 7.92-7.82
(2H, m) 7.70 (1H, d, J=6.5 Hz) 7.6-7.47 (3H, m) 7.41-7.36
(1H, m) 6.92 (1H, q, J=6.6 Hz) 1.9 (3H, d, J=6.6 Hz); 13C
NMR (100.6 MHz, CDCl3, ppm) δ 164.5, 153.4, 151.0, 137.1,
137.0, 133.9, 130.2, 129.0, 128.7, 126.5, 126.3, 125.7, 125.3,
123.5, 123.3, 123.0, 70.9, 21.8.
Table 2, Entry 2: Nicotinoylation of 5 with 8d. Following the
general procedure B, racemic 5 (17 mg, 0.1 mmol), nicotinic
anhydride (23 mg, 0.1 mmol), and catalyst 8d (0.004 mg, 0.004
mmol added as a solution of the free phosphine) in 3:1 tert-amyl
alcohol/DCM were stirred at rt for 3 h to afford unreacted
alcohol S-5 97.5% ee and nicotinate R-21 46.4% ee (after
hydrolysis); s = 11 and c = 68%.
Table 2, Entry 3: Isobutyroylation of 5 with 18b. Following the
general procedure B, racemic 5 (17.2 mg, 0.1 mmol), isobutyric
anhydride (17 μL, 0.1 mmol), and catalyst 18b (0.9 mg, 0.002
mmol, 95% ee) in 3:1 tert-amyl alcohol/DCM were stirred
at -25 °C for 37 h to afford unreacted alcohol R-5 24.0% ee
and isobutyrate S-17 66.0% ee (after hydrolysis); s = 6.126 and
c = 27%.
Table 1, Entry 3: Isobutyroylation of 5 with 18b. Following the
general procedure A, racemic 5 (17 mg, 0.1 mmol), isobutyric
anhydride (12 μL, 0.06 mmol), and catalyst 18b (0.5 mg, 0.001
mmol, 91% ee) in tert-amyl alcohol were stirred at rt for 48 h to
afford unreacted alcohol R-5 34.0% ee and isobutyrate S-17
49.8% ee (after hydrolysis); s = 5.126 and c = 41%.
Table 1, Entry 4: Isobutyroylation of 5 with 18b. Following the
general procedure A, racemic 5 (17 mg, 0.1 mmol), isobutyric
anhydride (17 μL, 0.1 mmol), and catalyst 18b (0.5 mg, 0.001
mmol, 91% ee) in tert-amyl alcohol were stirred at 0 °C for 48 h
to afford unreacted alcohol R-5 33.4% ee and isobutyrate S-17
51.1% ee (after hydrolysis); s = 5.326 and c = 40%.
Table 1, Entry 5: Isobutyroylation of 5 with 18c. Following the
general procedure A, racemic 5 (17 mg, 0.1 mmol), isobutyric
anhydride (12 μL, 0.06 mmol), and catalyst 18c (0.5 mg, 0.001
mmol, 93% ee) in tert-amyl alcohol were stirred at 0 °C for 24 h
to afford unreacted alcohol R-5 40.0% ee and isobutyrate S-17
36.8% ee (after hydrolysis); s = 3.526 and c = 52%.
Table 2, Entry 4: Nicotinoylation of 20a with 8d. Following the
general procedure B, racemic 20a (14 μL, 0.1 mmol), nicotinic
anhydride (13 mg, 0.06 mmol), and catalyst 8d (0.05 mg, 0.002
mmol) in 3:1 tert-amyl alcohol/DCM were stirred at -25 °C
for 20 h to afford unreacted alcohol S-20a 48.7% ee, and
nicotinate R-21 82.8% ee (after hydrolysis); s = 17 and c =
37%. GLC analysis after hydrolysis: SUPELCO BETA-DEX
120, 110 °C, 1.9 mL/min carrier gas flow. Retention times
18.2 min (R), 22.2 (S). For 21 (R1 = 2-MeC6H4, R2 = Me):
HRMS calcd for C15H15NNaO2 [M þ Na]; ESMS m/z =
264.0995, found 264.0992; IR (neat, cm-1) 1717, CdO; 1590,
CdN; 1H NMR (500 MHz, CDCl3, ppm) δ 9.29 (1H, d, J = 1.5
Hz) 8.78 (1H, dd, J = 4.9, 1.5 Hz) 8.32 (1H, dt, J = 7.8, 2.0 Hz)
7.50 (1H, dd, J = 7.3, 1.5 Hz) 7.38 (1H, dd, J = 7.8, 4.9 Hz)
7.38-7.14 (3H, m) 6.35 (1H, q, J = 6.8 Hz) 2.45 (3H, s) 1.67 (3H,
d, J = 6.8 Hz); 13C NMR (125.7 MHz, CDCl3, ppm) δ 164.4,
153.3, 150.9, 139.6, 136.9, 134.7, 130.5, 127.8, 126.9, 126.3,
125.2, 123.2, 70.6, 21.5, 19.2.
Table 1, Entry 6: Isobutyroylation of 5 with 18d. Following the
general procedure A, racemic 5 (18 mg, 0.1 mmol), isobutyric
anhydride (17 μL, 0.1 mmol), and catalyst 18d (0.1 mL, 0.01 M
in tert-amyl alcohol, 0.001 mmol, 95.5% ee) in tert-amyl alco-
hol were stirred at 0 °C for 16 h to afford unreacted alcohol
R-5 27.0% ee, and isobutyrate S-17 26.7% ee (after hydrolysis);
s = 2.326and c = 50%.
Table 2, Entry 5: Isobutyroylation of 20a with 18b. Following
the general procedure A, racemic 20a (13.6 mg, 0.1 mmol), iso-
butyric anhydride (42 μL, 0.25 mmol), and catalyst 18b (0.9 mg,
0.002 mmol, 95% ee) in 3:1 tert-amyl alcohol/DCM were stirred
at -25 °C for 64 h to afford unreacted alcohol R-20a 69.9%
ee and isobutyrate S-22a 69.0% ee (after hydrolysis); s = 1426
and c = 50.2%.
Table 1, Entry 7: Isobutyroylation of 5 with 18b. Following the
general procedure A, racemic 5 (17.2 mg, 0.1 mmol), isobutyric
anhydride (17 μL, 0.1 mmol), and catalyst 18b (0.9 mg, 0.002
mmol, 95% ee) in 3:1 tert-amyl alcohol/DCM were stirred
at -25 °C for 37 h to afford unreacted alcohol R-5 24.0% ee
and isobutyrate S-17 66.0% ee (after hydrolysis); s = 6.126 and
c = 27%.
Table 1, Entry 8: Isobutyroylation of 5 with 18b. Following the
general procedure A, racemic 5 (17.2 mg, 0.1 mmol), isobutyric
anhydride (17 μL, 0.1 mmol), and catalyst 18b (2 mg, 0.004
mmol, 95% ee) in DCM were stirred at -25 °C for 28 h to afford
unreacted alcohol R-5 11.2% ee, and isobutyrate S-17 48.0% ee
(after hydrolysis); s = 3.226 and c = 21%.
Table 2, Entry 6: Nicotinoylation of 20b with 8d. Following the
general procedure B, racemic 20b (17 mg, 0.1 mmol), nicotinic
anhydride (14 mg, 0.06 mmol), and catalyst 8d (0.05 mg, 0.002
mmol added as a solution of the free phosphine) in 3:1 tert-amyl
alcohol/DCM were stirred at rt for 20 h to afford unreacted
alcohol S-20b 48.8% ee and nicotinate R-21 94.4% ee (after
hydrolysis); s = 57 and c = 34%. HPLC analysis after hydrolysis:
CHIRALCEL OD analytical column, 3% 2-propanol/hexanes,
1 mL/min flow rate. Retention times 11.8 min (S) (minor) accord-
ing to (-) sign of optical rotation27 18.4 min (R) (major). For 21
(R1 = C6H5, R2 = t-Bu); HRMS calcd for C17H20NO2 [M þ H];
1723, CdO; 1590, CdN; 1H NMR (300 MHz, CDCl3, ppm)
δ 9.33 (1H, d, J = 2.4 Hz) 8.79 (1H, dd, J = 4.8, 1.8 Hz) 8.33 (1H,
dt, J = 7.8, 1.8 Hz) 7.44-7.23 (6H, m) 5.75 (1H, s) 1.04 (9H, s);
13C NMR (75.5 MHz, CDCl3, ppm) δ 164.3, 153.4, 150.8, 137.9,
132.1, 127.8, 127.6, 126.4, 123.3, 89.1, 68.8, 35.4, 26.1.
ESMS m/z = 270.1489, found m/z = 270.1479; IR (neat, cm-1
)
Table 2, Entry 1: Nicotinoylation of 5 with 8c. Following the
general procedure B, racemic 5 (17 mg, 0.1 mmol), nicotinic
anhydride (23 mg, 0.1 mmol), and catalyst 8c HBF4 (1 mg,
3
0.004 mmol) in 3:1 tert-amyl alcohol/DCM were stirred at rt
for 3 h to afford unreacted alcohol S-5 99.3% ee, and nicotinate
R-21 40.7% ee (after hydrolysis); s = 7.5 and c = 79%. For 21
Table 2, Entry 7: Isobutyroylation of 20b with 18b. Follow-
ing the general procedure A, racemic 20b (16.4 mg, 0.1 mmol),
J. Org. Chem. Vol. 75, No. 14, 2010 4683

Duffey et al.
JOCFeatured Article
isobutyric anhydride (42 μL, 0.25 mmol), and catalyst 18b
(0.8 mg, 0.002 mmol, 95% ee) in 3:1 tert-amyl alcohol/DCM
were stirred at -25 °C for 64 h to afford unreacted alcohol R-20b
50.9% ee and isobutyrate S-22b 66.8% ee (after hydrolysis);
s = 9.926 and c = 43%.
tion times 13.4 min (S), according to (-) sign of optical rotation,
21.4 min (R).14a
Table 4, Entry 7: Isobutyroylation of 20b with 23b. Following
the general procedure A, racemic 20b (17 mg, 0.1 mmol), iso-
butyric anhydride (21 μL, 0.1 mmol), and catalyst 23b (0.05 mL,
0.02 M in toluene, 0.001 mmol) in toluene were stirred at -40 °C
for 4 h to afford unreacted alcohol S-20b 10.1% ee and iso-
butyrate R-22b 25.2% ee (after hydrolysis); s = 1.8 and c =
29%. HPLC analysis: CHIRALCEL OD analytical column,
3% 2-propanol/hexanes, 1 mL/min flow rate. Retention times
11.8 min (S) according to (-) sign of optical rotation, 18.4 min
(R).14a
General Procedure for Parallel Kinetic Resolution Using Cat-
alysts 8d and 18b. Under Ar, a solution of 5 (0.1 mmol), nicotinic
anhydride (23 mg, 0.1 mmol), isobutyric anhydride (17 μL,
0.1 mmol, degassed), and Et3N (22 μL, 0.15 mmol) in 3:1 tert-
amyl alcohol/CH2Cl2 (0.8 mL) was cooled to -25 °C in a
Cryocool. Catalysts 8d and 18b (amounts based on the rate
predicted from control experiments A and B assuming a linear
relationship between the amount of catalyst and the rate) were
taken up in 3:1 tert-amyl alcohol/CH2Cl2 (0.2 mL) and added to
the reaction mixture. The reactions were stirred until the alcohol
reactant was consumed (several days). The mixture was con-
centrated (aspirator), and the ratio of products was determined
by crude NMR. The products were purified by flash chroma-
tography. Ester fractions were concentrated to afford 21 and 17.
Next, 5% NaOH/MeOH (1 mL) was added to saponify each
ester prior to assay. The solution was warmed gently for 5 min
and then left at room temperature for 2 h. Methanol was
evaporated, and the residue was filtered through an 8 ꢀ 1.2
cm pad of silica gel in CH2Cl2. After solvent removal (aspirator),
HPLC assay for each product was carried out on a chiral
support.
Table 3, Entry 2: PKR of 5. Following the general procedure,
racemic 5 (17 mg, 0.1 mmol), 8d (ca. 0.001 mmol), and 18b
(3.3 mg, 0.007 mmol, 95% ee) were stirred for 69 h to afford 14
mg (51%) of R-21 65.3% ee (after hydrolysis) and 11 mg (46%)
of S-17 72.7% ee (after hydrolysis).
Table 3, Entry 4: PKR of 20a. Following the general proce-
dure, racemic 20a (14 μL, 0.1), 8d (ca. 0.006 mmol), and 18b
(1.5 mg, 0.003 mmol, 95% ee) were stirred for 6 days to afford
12.6 mg (52%) of R-21 67.7% ee (after hydrolysis) and 9.1 mg
(44%) of S-22a 82.3% ee (after hydrolysis).
Table 3, Entry 5: PKR of 20b. Following the general proce-
dure, racemic 20b (17.5 mg, 0.1 mmol), 8d (ca. 0.006 mmol), and
18b (2.2 mg, 0.005 mmol, 95% ee) were stirred for 6 days to
afford 9.0 mg (33%) of R-21 79.3% ee (after hydrolysis) and
14.1 mg (60%) of S-22b 40.5% ee (after hydrolysis).
Table 4, Entry 1: Isobutyroylation of 5 with 23a. Following the
general procedure A, racemic 5 (17 mg, 0.1 mmol), isobutyric
anhydride (8 μL, 0.05 mmol), and catalyst 23a (0.05 mL, 0.02 M
in tert-amyl alcohol, 0.001 mmol) in tert-amyl alcohol were
stirred at rt for 4 h to afford unreacted alcohol S-5 10.9% ee
and isobutyrate R-17 41.8% ee (after hydrolysis); s = 2.7 and
c = 21%.
Table 4, Entry 4: Isobutyroylation of 5 with 23b. Following the
general procedure A, racemic 5 (17 mg, 0.1 mmol), isobutyric
anhydride (17 μL, 0.1 mmol), and catalyst 23b (0.4 mg, 0.001
mmol) in toluene were stirred at -40 °C for 4 h to afford
unreacted alcohol S-5 47.5% ee and isobutyrate R-17 85.5%
ee (after hydrolysis); s = 20.5 and c = 36%.
Table 4, Entry 5: Isobutyroylation of 5 with 23b. Following the
general procedure A, racemic 5 (17 mg, 0.1 mmol), isobutyric
anhydride (8 μL, 0.05 mmol), and catalyst 23b (0.05 mL, 0.02 M
in toluene, 0.001 mmol) in toluene were stirred at -70 °C for 8 h
to afford unreacted alcohol S-5 7.0% ee and isobutyrate R-17
92.9% ee (after hydrolysis); s = 29 and c = 7%.
m-Chlorobenzoylation of 5 with Phosphine 8d. Following the
general procedure B, racemic 5 (22 mg, 0.12 mmol), m-chloro-
benzoic anhydride (29 mg, 0.1 mmol), and phosphonium salt
8d HBF4 (0.05 mL, 0.04 M in toluene, 0.002 mmol) in toluene
3
were stirred at -40 °C for 6 h to afford unreacted alcohol S-5
97.5% ee and isobutyrate R-24 65.3% ee (after hydrolysis); s =
19.8 and c = 60%.
Representative Anhydride Competition Experiment. A solu-
tion of 1-naphthylmethylcarbinol 5 (17 mg, 0.1 mmol), m-chlo-
robenzoic anhydride (30 mg, 0.1 mmol), isobutyric anhydride
(17 μL, 0.1 mmol, degassed), and the phosphonium salt
8c HBF4 (1.5 mg, 0.004 mmol) in toluene (0.7 mL) was treated
3
with Et3N (21 μL, 0.15 mmol). After the solution was stirred for
5 h, addition of i-PrNH2 (0.1 mL) quenched the reaction, and the
solvent was evaporated (N2 stream). NMR assay of the reaction
mixture revealed 18:1 of the m-chlorobenzoate to isobutyrate
esters based on comparison of the spectrum with the spectra of
the individual esters.
General Procedure for Parallel Kinetic Resolution Using Cat-
alysts 8d and ent-23b. A solution of 5 (0.1 mmol), m-chloroben-
zoic anhydride (23 mg, 0.1 mmol), isobutyric anhydride (17 μL,
0.1 mmol, degassed), and Et3N (22 μL, 0.15 mmol) in toluene
(0.8 mL) was cooled to -40 °C in a Cryocool. The catalysts were
added as solutions: phosphonium salt 8d HBF4 (25 μL, 0.0022
3
mmol, 0.088 M in DCM) and ent-23b (25 μL, 0.001 mmol, 0.04
M in DCM). The reactions were stirred for 2-3 h and then
quenched with i-PrNH2 (0.1 mL). The mixture was concentrated
1
(aspirator), and the ratio of products was determined by H
NMR spectroscopy. The products were purified by flash chro-
matography (2:1 Hex/Et2O). Analytical TLC (2:1 Hex/Et2O):
m-chlorobenzoate, Rf = 0.44; isobutyrate, Rf = 0.32; alcohol,
Rf = 0.12. Ester fractions were concentrated to afford 17 and 24.
Next, 5% NaOH/MeOH (1 mL) was added to saponify each
ester prior to assay. The solution was warmed gently for 5 min
and then left at room temperature for 2 h. Methanol was
evaporated, and the residue was filtered through an 8 ꢀ 1.2
cm pad of silica gel in DCM. After solvent removal (aspirator),
HPLC assay for each product was carried out on a chiral
support as previously described.
Parallel Kinetic Resolutions. Table 5, Entry 1. Following the
general procedure, racemic 5 (17 mg, 0.1 mmol), phosphonium
salt 8d HBF4 (0.0022 mmol), and ent-23b (0.001 mmol) were
3
stirred for 3 h to afford R-24, 88% ee (after hydrolysis), S-17,
75% ee (after hydrolysis), and recovered alcohol 5, 27% ee.
Table 5, Entry 2. Following the general procedure, racemic 5
(17 mg, 0.1 mmol), phosphonium salt 8d HBF4 (0.0026 mmol),
3
and ent-23b (0.001 mmol) were stirred for 3 h to afford R-24 (14
mg, 44%) 87% ee (after hydrolysis), S-17 (8 mg, 33%) 76% ee
(after hydrolysis), and recovered alcohol 5, 8% ee.
Table 4, Entry 6: Isobutyroylation of 20a with 23b. Following
the general procedure A, racemic 20a (17 mg, 0.1 mmol),
isobutyric anhydride (21 μL, 0.1 mmol), and catalyst 23b
(0.05 mL, 0.02 M in toluene, 0.001 mmol) in toluene were stirred
at -40 °C for 4 h to afford unreacted alcohol S-20a 36.2%
ee and isobutyrate R-22a 74.0% ee (after hydrolysis); s = 9.5
and c = 33%. HPLC analysis: CHIRALCEL OB analytical
column, 3% 2-propanol/hexanes, 1 mL/min flow rate. Reten-
Table 5, Entry 3. Following the general procedure, racemic 5
(17 mg, 0.1 mmol), phosphonium salt 8d HBF4 (0.0052 mmol),
3
and ent-23b (0.002 mmol) were stirred for 2 h to afford R-24,
86% ee (after hydrolysis), S-17, 75% ee (after hydrolysis), and
recovered alcohol 5, 19% ee.
Preparation of Mixed Biphenyl Cyclohexyl Anhydride 29. A
solution of 4-biphenyl carboxylic acid (999 mg, 5.0 mmol) in
4684 J. Org. Chem. Vol. 75, No. 14, 2010

Duffey et al.
JOCFeatured Article
THF (15 mL) was cooled to 0 °C, and NaH (204 mg, 5.1 mmol)
was added. After 30 min, the solution was allowed to warm to rt
and stirred for an additional 1 h. The solution was recooled to
0 °C, and cyclohexyl carbonyl chloride (0.61 mL, 4.5 mmol) was
added dropwise. The mixture was stirred for 30 min at 0 °C. The
ice bath was removed, and the mixture was stirred for an
additional 40 min at rt. Activated carbon (∼0.5 g) and hexanes
(10 mL) were added, and the mixture was filtered through Celite
and concentrated to yield a white solid. The solid was taken up
in hot hexanes, filtered (to remove the symmetrical biphenyl
anhydride), and allowed to cool to cool. Slow evaporation of the
hexanes afforded 29 as a white crystalline solid and was used as
soon as possible without further manipulation (478 mg, 35%):
1H NMR (400 MHz, CDCl3, ppm) δ 8.10 (2H, d, J = 8.2 Hz)
7.69 (2H, d, J = 8.2 Hz) 7.62 (2H, d, J = 6.9 Hz) 7.50-7.39 (3H,
m) 2.61 (1H, tt, J = 10.9, 3.5 Hz) 2.14-2.06 (2H, m) 1.88-1.80
(2H, m) 1.72-1.56 (3H, m) 1.42-1.23 (3H, m); 13C NMR (101
MHz, CDCl3, ppm) δ 171.5, 147.0, 139.6, 130.9, 129.0, 128.5,
127.6, 127.4, 127.3, 44.2, 28.5, 25.6, 25.2.
Acknowledgment. This work was supported by the
National Science Foundation and the National Institutes
of Health (CA17918).
Supporting Information Available: 1H and 13C spectra of
new compounds. This material is available free of charge via the
Internet at http://pubs.acs.org.
J. Org. Chem. Vol. 75, No. 14, 2010 4685