Kinetic resolution of N-acyl-thiolactams via catalytic enantioselective deacylation

Article abstract of DOI:10.1021/ol401122g

Methanolysis of N-acyl-thiazolidin-2-thiones and -oxazolidin-2-thiones in the presence of acyl transfer catalyst benzotetramisole (BTM) proceeds in a highly enantioselective fashion thus enabling kinetic resolution of these substrates.

Full text of DOI:10.1021/ol401122g

ORGANIC
LETTERS
XXXX
Vol. XX, No. XX
000–000
Kinetic Resolution of N‑Acyl-Thiolactams
via Catalytic Enantioselective Deacylation
Valentina D. Bumbu, Xing Yang, and Vladimir B. Birman*
Department of Chemistry, Washington University, One Brookings Drive,
St. Louis, Missouri 63130, United States
birman@wustl.edu
Received April 22, 2013
ABSTRACT
Methanolysis of N-acyl-thiazolidin-2-thiones and -oxazolidin-2-thiones in the presence of acyl transfer catalyst benzotetramisole (BTM) proceeds
in a highly enantioselective fashion thus enabling kinetic resolution of these substrates.
Enzymatic resolution of racemic alcohols, thiols, and
amines is routinely accomplished via two basic approaches:
(a) enantioselective acylation and (b) enantioselective hy-
drolysis of the corresponding acylated derivatives. Often,
these complementry transformations can be catalyzed by a
single enzyme, which, as expected, exhibits the same sense of
enantioselection in both directions of the process and thus
can be used to achieve opposite overall enantioselectivities
(see, e.g., Figure 1).
Over the past two decades, considerable progress has
been made in the development of nonenzymatic asym-
metric acylation catalysts, which can now be used to
resolve the aforementioned classes of substrates with good
enantioselectivity.1 However, toourknowledge, there have
been no reports of these catalysts being able to operate in
Figure 1. Enzymatic esterification and ester hydrolysis.
reverse and achieve enantioselective deacylation.2ꢀ6 In this
letter, we disclose the first instances of this transformation.
In the course of our recent studies on enantioselective
(1) For recent reviews, see: (a) Krasnov, V. P.; Gruzdev, D. A.; Levit,
€
Angew. Chem., Int. Ed. 2011, 50, 6012. (c) Pellissier, H. Adv. Synth.
Catal. 2011, 353, 1613. (d) Spivey, A. C.; Arseniyadis, S. Top. Curr.
Chem. 2010, 291, 233.
N-acylation of lactams and thiolactams,7,8 we subjected
(()-4-phenylthiazolidine-2-thione 1 to our standard kinetic
resolution protocol, which calls for quenching the reaction
mixture with methanol to destroy excess anhydride (Figure 2,
eq 1). We noted that the enantiomeric excess of the product
G. L. Eur. J. Org. Chem. 2012, 1471. (b) Muller, C. E.; Schreiner, P. R.
(2) (a) The only previously known nonenzymatic enantioselective
deacylation was achieved via CBS reduction of N-acyl-oxazolidinones
and -imidazolidinones: Hashimoto, N.; Ishizuka, T.; Kunieda, T. Tetra-
hedron Lett. 1998, 39, 6317. (b) In addition, an unsuccessful attempt at
enantioselective ester deacylation (s e 1.5) has been recorded: Yoshizu-
mi, T.; Takaya, H. Tetrahedron: Asymmetry 1995, 6, 1253.
(3) We use the term “enantioselective deacylation” to describe a
reaction in which a racemic acyl donor loses an achiral acyl group to
produce a chiral leaving group in enantioenriched form (cf. the right-
hand side of Figure 1). By contrast, there are many examples of
enantioselective alcoholysis (or esterification) of racemic acyl donors
wherein the chirality resides in the acyl group, while the leaving group
(if any) is achiral. See, e.g., refs 4ꢀ6.
(4) For select examples of conventional kinetic resolution (KR) of
acyl donors via nonenzymatic alcoholysis, see: (a) Narasaka, K.; Kanai,
F.; Okudo, M.; Miyoshi, N. Chem. Lett. 1989, 1187. (b) Berkessel, A.;
Cleemann, F.; Mukherjee, S. Angew. Chem., Int. Ed. 2005, 44, 7466. (c)
Notte, G. T.; Sammakia, T. J. Am. Chem. Soc. 2006, 128, 4230. (d)
Ishihara, K.; Kosugi, Y.; Umemura, S.; Sakakura, A. Org. Lett. 2008,
10, 3191. (e) Shiina, I.; Nakata, K.; Ono, K.; Onda, Y.; Itagaki, M.
J. Am. Chem. Soc. 2010, 132, 11629. (f) Yang, X.; Birman, V. B. Chem.;
Eur. J. 2011, 17, 11296. (g) Bumbu, V. D.; Birman, V. B. J. Am. Chem.
Soc. 2011, 133, 13902.
r
10.1021/ol401122g
XXXX American Chemical Society

and the unreacted starting material diminished considerably
if the quenched mixture was allowed to stand for too long.
As a result, the conversion and the apparent selectivity
factor calculated from these parameters9 plummeted as well.
original “forward” process (eq 2). As expected, in the absence
of added catalyst, the methanolysis proceeded at a negligible
rate (Table 1, entry 1).
Figure 3. Catalysts used in this study.
Table 1. Optimization Studya
Figure 2. Enantioselective N-acylation of (()-1 and the reverse
reaction.
We were aware of prior literature reports describing
DMAP-catalyzed alcoholysis of N-acyl-oxazolidine-2-thione
and N-acyl-thiazolidine-2-thiones.10,11 Thus, we quickly real-
ized that we were witnessing an enantioselective version of
this process catalyzed by our asymmetric acyl transfer cata-
lyst benzotetramisole (BTM) 2 (Figure 3). To confirm our
hypothesis, we subjected racemic 1a, the N-isobutyryl deri-
vative of 1, to methanolysis under similar conditions. Pleas-
ingly, the reaction yielded the same products with the
opposite selectivity and a better selectivity factor than the
entry
catalyst
R
solvent
time
conv %
s
1
none
i-Pr
i-Pr
i-Pr
i-Pr
i-Pr
Et
CDCl3
CDCl3
CDCl3
CDCl3
CDCl3
CDCl3
CDCl3
CDCl3
CDCl3
CD2Cl2
PhMe
Et2O
7 d
<10
55
48
49
50
57
<10
37
47
49
47
42
NDd
26
2
2
4 d
3
2þBzOH
2þBzOH
2þBzOH
2þBzOH
2þBzOH
3þBzOH
4þBzOH
2þBzOH
2þBzOH
2þBzOH
24 h
24 h
24 h
12 h
7 d
78
4b
5c
6
81
84
15
7
Ph
NDd
6.7
7.0
41
8
i-Pr
i-Pr
i-Pr
i-Pr
i-Pr
7 d
(5) For select examples of dynamic kinetic resolution (DKR) of acyl
donors via nonenzymatic alcoholysis, see: (a) Seebach, D.; Jaeschke, G.;
Gottwald, K.; Mastsuda, K.; Formisano, R.; Chaplin, D. A. Tetrahe-
dron 1997, 53, 7539. (b) Liang, J.; Ruble, J. C.; Fu, G. C. J. Org. Chem.
1998, 63, 3154. (c) Tang, L.; Deng, L. J. Am. Chem. Soc. 2002, 124, 2870.
9
3 d
10
11
12
24 h
24 h
7 d
86
3.9
€
(d) Berkessel, A.; Cleeman, F.; Mukherjee, S.; Muller, T. N.; Lex, J.
Angew. Chem. 2005, 44, 807. (e) Lu, G.; Birman, V. B. Org. Lett. 2011,
13, 356. (f) Yang, X.; Birman, V. B. Angew. Chem., Int. Ed. 2011, 50,
5553.
(6) For select examples of desymmetrization of prochiral meso-
anhydrides, see: Tian, S.-K.; Chen, Y.; Hang, J.; Tang, L.; McDaid,
P.; Deng, L. Acc. Chem. Res. 2004, 37, 621.
a General conditions (0.1 mmol of (()-1aꢀc, 0.075 mmol of MeOH,
0.01 mmol of (R)-2, 0.01 mmol of benzoic acid, 0.75 mL of solvent, rt)
were used, unless specified otherwise. b 0.75 equiv of PhCH2OH was
used instead of methanol. c 0.75 equiv of (1-naphthyl)2CHOH was used
instead of methanol. d Not determined.
(7) Yang, X.; Bumbu, V. D.; Liu, P.; Li, X.; Jiang, H.; Uffman, E. W.;
Guo, L.; Zhang, W.; Jiang, X.; Houk, K. N.; Birman, V. B. J. Am. Chem.
Soc. 2012, 134, 17605.
(8) For examples of nonenzymatic enantioselective N-acylation of
other classes of substrates, see: (a) Arai, S.; Bellemin-Laponnaz, S.; Fu,
G. C. Angew. Chem., Int. Ed. 2001, 40, 234. (b) De, C. K.; Klauber, E. G.;
Seidel, D. J. Am. Chem. Soc. 2009, 131, 17060. (c) Fowler, B. S.;
Mikochik, P. J.; Miller, S. J. J. Am. Chem. Soc. 2010, 132, 2870. (d)
Binanzer, M.; Hsieh, S. Y.; Bode, J. W. J. Am. Chem. Soc. 2011, 133,
19698.
(9) (a) Kagan, H. B.; Fiaud, J. C. Top. Stereochem. 1988, 18, 249. (b)
Enantioselectivity in KR is expressed in terms of a selectivity factor (s)
defined as the ratio of reaction rates of the fast- and the slow-reacting
enantiomers of the starting material: s = kfast/kslow. In the KR of
racemic mixtures, it is usually calculated from the ee values of the
product and the unreacted starting material according to Kagan’s
equations (ref 5): (1) conversion C = eeSM/(eeSM þ eePR). (2) Selectivity
factor s = ln[(1 ꢀ C)(1 ꢀ eeSM)]/ln[(1 ꢀ C)(1 þ eeSM)].
(10) (a) Su, D.-W.; Wang, Y.-C.; Yan, T.-H. Tetrahedron Lett. 1999,
40, 4197. (b) Wu, Y.; Sun, Y.-P.; Yang, Y.-Q.; Hu, Q.; Zhang, Q. J. Org.
Chem. 2004, 69, 6141.
It is also noteworthy that the reaction worked only
poorly when a thoroughly purified sample of 1a was used
(Table 1, entry 2). Evidently, the traces of isobutyric acid in
crude 1a played the role of a cocatalyst. With the addition
of benzoic acid (Table 1, entry 3), the reaction protocol
became reproducible and a brief optimization study was
undertaken. Benzyl alcohol and di-(1-naphthyl)methanol
(Table 1, entries 4 and 5) produced virtually the same results
as did methanol. N-Propionyl derivative 1b underwent meth-
anolysis considerably faster than did 1a, but with greatly
diminished enantioselectivity (Table 1, entry 6). Its N-benzoyl
counterpart 1c was essentially unreactive (Table 1, entry 7).
These results were in line with the observations made in the
(11) For reviews of synthetic utility of thiazolidine-2-thiones and
ꢀ
oxazolidine-2-thiones, see: (a) Velazquez, F.; Olivo, H. F. Curr. Org.
Chem 2002, 6, 1. (b) Delaunay, D.; Toupet, L.; Corre, M. L. J. Org.
Chem. 1995, 60, 6604.
(12) (a) Birman, V. B.; Jiang, H.; Li, X.; Guo, L.; Uffman, E. W.
J. Am. Chem. Soc. 2006, 128, 6536. (b) Yang, X.; Bumbu, V. D.; Birman,
V. B. Org. Lett. 2011, 13, 4755.
B
Org. Lett., Vol. XX, No. XX, XXXX

course of our earlier N-acylation studies.7,12 More surpris-
ingly, much lower reaction rates and selectivity factors were
obtained with two other amidine-based catalysts, 3 and 4,
that previously showed good enantioselectivities (Table 1,
entries 8 and 9). Among the solvents examined (see Table 1,
entries 3 and 10ꢀ12), toluene proved to be equal, or even
slightly superior, to chloroform. However, for the sake of
convenient NMR monitoring, we continued using CDCl3 for
the rest of this study. Variation of the aryl substituent
suggested that electron-donating groups increase the enan-
tioselectivity, while electron-withdrawing ones diminish it
(Table 2, entries 1 and 2 vs 3 and 4). Somewhat surprisingly,
replacing the phenyl group with 1- or 2-naphthyl or 2-thienyl
was also detrimental (Table 2, entries 5ꢀ7), even though the
same groups were typically associated with higher levels of
enantioselectivity in our previous studies on N-acylation.
indane-fused substrate 16a, in which the benzene ring is
geometrically constrained (Table 2, entry 12). Similarly
poor results were also obtained in the reverse direction:
acylation of (()-16 (Figure 4). These results were in contrast
with the successful enantioselective N-acylation of the
Figure 4. Enantioselective N-acylation of (()-16.
Table 2. KR of N-Isobutyryl-thiazolidine-2-thionesa
structurally similar indane-fused oxazolidin-2-one and
β-lactam in the presence of Cl-PIQ 3 (selectivity factors 36
and 14 were obtained, respectively).
Alcoholysis of the analogous N-isobutyryl-oxazolidin-
2-thiones was examined next (Table 3). The reaction
proved to be rather slow, but still afforded synthetically
useful levels of enantioselectivity in most cases.13 Signifi-
cantly, we observed that the apparent enantioselectivity of
deacylation of the ester-substituted substrate 22a dimin-
ished with time (Table 3, entries 6 and 7). We surmised
that the deacylated product (R)-22 may undergo slow
acyl group exchange with the unreacted (S)-22a under
the reaction conditions. By the same token, gradual race-
mization should also be observed during enantioselective
acylation of (()-22 in the presence of BTM 2. A control
experiment confirmed that this is indeed the case (Table 4).
The reaction reached essentially complete conversion with-
in 30 min. HPLC analysis of the crude mixture indicated
significant enantiomeric enrichment of both the product
and the starting material corresponding to a selectivity
entry
R1
time, d
conv %
s
1
p-MeOC6H4 (5a)
o-MeOC6H4 (6a)
p-ClC6H4 (7a)
1
49
47
54
50
48
57
45
47
49
0
90
2
1
108
23
3
0.6
1
4
o-ClC6H4 (8a)
36
5
1-naphthyl (9a)
2-naphthyl (10a)
2-thienyl (11a)
trans-Ph-CH=CH (12a)
CO2Me (13a)
5
30
6
2
37
7
1.5
1
32
8
37
9
0.6
7
15
10
11
12
i-Pr (14a)
ND
1.4
2.3
PhCH2 (15a)
7
27
33
; (cf. 16a, Figure 4)
7
a General conditions were used (see Table 1).
Nonetheless, the selectivity factors obtained in these less
favorable cases are still amply sufficient for practical
purposes (s > 20). The styryl group containing an extended
π-system also produced a good result (Table 2, entry 8).
Diminished enantioselectivity was obtained with the ester-
substituted substrate (Table 2, entry 9). No reaction was
observed with the isopropyl group lacking a π-system
(Table 2, entry 10). Similarly, extremely a low reaction
rate and enantioselectivity were observed in the case of the
4-benzyl-substituted substrate (Table 2, entry 11), wherein
the π-system is one atom farther away from the nitrogen.
These negative results were fully consistent with our earlier
N-acylation studies and once again underscored the im-
portance of π-interactions in the transition state.7 On the
other hand, we were disappointed to observe the slow
conversion and the low selectivity factor in the case of the
Table 3. KR of N-Isobutyryl-oxazolidine-2-thionesa
entry
R1
time, d
conv %
s
1
2
3
4
5
6
7
phenyl (17a)
2
4
6
3
3
4
1
36
47
30
57
43
53
29
57
59
38
13
32
2.6
7.9
p-MeOC6H4 (18a)
o-MeOC6H4 (19a)
p-ClC6H4 (20a)
o-ClC6H4 (21a)
CO2Me (22a)
CO2Me (22a)
(13) The analogous N-acyloxazolidin-2-ones do not react under these
conditions. Resolution of racemic oxazolidinones is thus best achieved
via enantioselective acylation, as previously described.7,12a
a General conditions were used (see Table 1).
Org. Lett., Vol. XX, No. XX, XXXX
C

Table 4. Enantioselective N-Acylation of (()-22a
Table 5. Racemization of 22 and 22aa
ee % after 24 h
(R)-22a (S)-22
entry
catalyst
1
2
3
4
5
none
81
18
20
81
82
77
14
11
74
75
2 (10 mol %)
2þ BzOH (10 mol % each)
i-Pr2NEt (50 mol %)
i-PrCO2H (10 mol %)
time,
h
(R)-22a,
(S)-22,
entry
ee %
ee %
conv %
s
a Conditions: 0.1 M (R)-22a (ee = 82%) and 0.1 M (S)-22 (ee = 78%)
in CDCl3, 10 mol % catalyst, rt, 24 h.
1
2
3
4
5
0.5
1.0
2.3
8.0
48
75
57
39
33
3.0
82
62
47
36
3.8
52
48
54
52
56
17
7.5
3.5
2.8
1.1
Indeed, when we mixed roughly equimolar amounts of
enantioenriched (R)-22a and (S)-22 in the presence of
10 mol % (R)-BTM 2, both ee values deteriorated signifi-
cantly over the course of 24 h, whether or not benzoic acid
was also added (Table 5, entries 2 and 3). In contrast, the
ee’s remained virtually unchanged in the absence of
a Single data points are presented for the sake of consistency. For
duplicate results, see Supporting Information.
€
additives or in the presence of Hunig’s base or isobutyric
acid (Table 5, entries 1, 4, and 5). Taken together, these
results clearly demonstrate the reversibility of the deacyla-
tion of 22a, which explains the low apparent selectivity
factors in Table 3, entries 6 and 7. The same racemization
pathway probably also operates to some extent in the case of
other substrates reported in this study, particularly when the
reaction times are long. Thus, the experimentally observed
s-values shown in Tables 2 and 3 are likely to be lower than
the “true” enantioselectivities.
In conclusion, we have demonstrated for the first time
the possibility of catalytic enantioselective deacylation pro-
moted by an acyl transfer catalyst. At present, this metho-
dology provides an alternative means of achieving kinetic
resolution of thiazolidine- and oxazolidine-2-thiones. Other
applications will be reported in due course.
Acknowledgment. We are grateful to the National
Science Foundation for financial support of this study
(CHE-1012979).
Figure 5. Racemization of 22 and 22a via acyl exchange.
factor of 17. However, the ee’s dropped significantly over
the next 2 h leading to diminished apparent selectivity
factors. We hypthesized that the acyl group exchange is
likely to be promoted by BTM, via the acyl transfer mode
of catalysis (Figure 5).
Supporting Information Available. Experimental pro-
cedures and NMR spectra. These materials are available
free of charge via the Internet at http://pubs.acs.org.
The authors declare no competing financial interest.
D
Org. Lett., Vol. XX, No. XX, XXXX