Nonaromatic amidine derivatives as acylation catalysts

Article abstract of DOI:10.1021/ol0623419

(Chemical Equation Presented) Catalytic activity of nonaromatic bicyclic amidines and bicyclic isothioureas in acylation reactions was found to be remarkably dependent on the sizes of both rings. DBN and especially its thia-analogue (THTP) have been identified as highly active acylation catalysts.

Full text of DOI:10.1021/ol0623419

ORGANIC
LETTERS
2007
Vol. 9, No. 1
37-40
Nonaromatic Amidine Derivatives as
Acylation Catalysts
Vladimir B. Birman,* Ximin Li, and Zhenfu Han
Department of Chemistry, Washington UniVersity, Campus Box 1134,
One Brookings DriVe, St. Louis, Missouri 63130
birman@wustl.edu
Received September 22, 2006
ABSTRACT
Catalytic activity of nonaromatic bicyclic amidines and bicyclic isothioureas in acylation reactions was found to be remarkably dependent on
the sizes of both rings. DBN and especially its thia-analogue (THTP) have been identified as highly active acylation catalysts.
4-(Dimethylamino)pyridine (DMAP) 11 and N-methylimi-
dazole (NMI) 22 are among the best known achiral acylation
catalysts (Figure 1). As such, their structures have been used
extensively as catalaphores3 for designing nonenzymatic
asymmetric acylation catalysts.4 Finding effective chiraphores
for these planar aromatic heterocycles, however, is not a
trivial matter.5,6 Three years ago, we identified dihydroimi-
dazo[1,2-a]pyridine (DHIP) 3 as a moderately active acyl-
ation catalyst.7a Although DHIP was less active than DMAP
1 and even NMI 2 in a preliminary catalytic activity test, it
had one important advantage from the viewpoint of asym-
metric catalyst design: a tetrahedral carbon R to the nucleo-
philic nitrogen, which permitted effective and straightforward
discrimination of the two faces of the molecule.
Two easily obtainable chiral DHIP derivatives, CF3-PIP
45a and Cl-PIQ 55b, proved to be effective catalysts for ki-
netic resolution of benzylic and allylic alcohols. Subse-
quently, we demonstrated that tetramisole 6, lacking the pyri-
dine ring, is also competent in this capacity, albeit not very
active. Its benzannellated analogue BTM 7 displayed im-
proved catalytic activity and far superior enantioselectivity.5c-e
Despite the modest catalytic activity of tetramisole itself,
we wished to explore nonaromatic heterocycles of this type
further. The presence of tetrahedral carbon atoms in both of
the rings would make them potentially attractive catalaphores,
by providing an opportunity for introducing additional
stereocenters, compared to the partly aromatic structures (cf.
4, 5, and 7) and thus leading to greater diversity of possible
structural variations. The key questions were as follows: (a)
What makes tetramisole catalytically active? (b) How can it
be made more active without introducing the benzene ring?
Initially, we set out to determine which part of our catalyst
design constitutes “the minimal catalaphore”. Would the
amidine moiety itself be sufficient for a compound to
function as an acylation catalyst? In order to test this
possibility, we decided to investigate the catalytic activity
(1) For reviews, see: (a) Ho¨fle, G.; Steglich, W.; Vorbru¨ggen, A. Angew.
Chem., Int. Ed. 1978, 17, 569; (b) Spivey, A. C.; Arseniyadis, S. Angew.
Chem., Int. Ed. 2004, 43, 5436.
(2) See, e.g.: Connors, K. A.; Pandit, N. K. Anal. Chem. 1978, 50,
1542.
(3) For discussion of terms “chiraphore” and “catalaphore”, see: Mulzer,
J. Basic Principles of Asymmetric Synthesis. In ComprehensiVe Asymmetric
Catalysis; Jacobsen, E. N., Pfaltz, A., Yamamoto, H., Eds.; Springer-
Verlag: Berlin, Heidelberg, 2004; Vol. 1, Chapter 3.
(4) For review of nonenzymatic kinetic resolution of alcohols and leading
references to various catalyst designs, see: Vedejs, E.; Jure, M. Angew.
Chem., Int. Ed. 2005, 44, 3974.
(5) For the most successful DMAP-based design, see: Fu, G. C. Acc.
Chem. Res. 2004, 37, 542 and references cited therein.
(6) For the most successful imidazole-based designs, see: (a) Miller, S.
J. Acc. Chem. Res. 2004, 37, 601 and references cited therein. (b) Ishihara,
K.; Kosugi, Y.; Akakura, M. J. Am. Chem. Soc. 2004, 126, 12212.
(7) (a) Birman, V. B.; Uffman, E. W.; Jiang, H.; Li, X.; Kilbane, C. J.
J. Am. Chem. Soc. 2004, 126, 12226. (b) Birman, V. B.; Jiang, H. Org.
Lett. 2005, 7, 3445. (c) Birman, V. B.; Li, X. Org. Lett. 2006, 8, 1351. (d)
Birman, V. B.; Jiang, H.; Li, X.; Guo, L.; Uffman, E. W. J. Am. Chem.
Soc. 2006, 128, 6536. (e) Birman, V. B.; Guo, L. Org. Lett. 2006, 8,
4859.
10.1021/ol0623419 CCC: $37.00
© 2007 American Chemical Society
Published on Web 12/09/2006

Figure 1. Known and potential acylation catalysts.
of simple bicyclic amidines, such as 8-12. Naturally, we
first turned our attention to the commercially available DBN
8 and DBU 9.8 Although these reagents have long been
regarded as non-nucleophilic bases, there is now sufficient
evidence in the literature that this is not the case.9 However,
we have not been able to find any reports of their use as
catalysts in acylations with carboxylic anhydrides. We began
by comparing DBN and DBU with the known achiral
acylation catalysts 1-3 using a simple test: acylation of
methanol with acetic anhydride in the presence of stoichio-
metric amounts of Hu¨nig’s base. A 5 mol % catalyst loading
was adopted for most compounds in this study in order to
permit direct comparison of their catalytic activities; however,
the most active catalysts, such as DMAP, had to be employed
in a 50-fold lower concentration to obtain conveniently
measurable reaction times. The reactions were monitored by
1H NMR to determine the time required for 50% conversion
(t1/2). The results are summarized in Table 1.
To our surprise, DBN (Table 1, entries 5 and 6) turned
out to be a highly active catalyst, whereas DBU (entry 7)
displayed virtually no activity. Intrigued by these results, we
synthesized the rest of the amidines shown in Figure 1
(Scheme 1).10 None of these compounds showed any
appreciable activity (entries 8-10). On one hand, these
findings confirmed that simple bicyclic amidines can be
highly catalytically active. On the other hand, the striking
difference between DBN and all the other bicyclic amidines
indicated the critical importance of the sizes of both rings
for the catalytic activity. We reasoned that systematic
variation of the ring sizes could also hold the key to
improving the catalytic activity of the bicyclic isothiourea
catalaphore 13 present in our original nonaromatic “lead
compound”, tetramisole (6).
While this work was in progress, we became aware of a
related recent study by Okamoto and Kobayashi.11 They
compared the catalytic activities of 3 (DHIP) and 17 (the
Table 1. Methanol Test
a
entry
catalyst
none
catalyst loading (mol %)
t1/2
1
2
3
4
5
6
7
8
18 h
DMAP 1
NMI 2
DHIP 3
DBN 8
DBN 8
DBU 9
10
0.1
5
5
5
1
16 min
1.3 h
1.8 h
4.0 min
40 min
12 h
5
5
18 h
9
11
5
16 h
10
11
12
13
14
15
16
17
18
19
12
13
14
THTP 15
16
DHPB 19
CF3-PIP 4
Cl-PIQ 5
Tetramisole 6
BTM 7
5
5
5
0.1
5
0.1
5
5
5
5
18 h
14 h
7.5 h
25 min
3.5 h
10 min
38 min
12 min
16 h
(8) For review of traditional synthetic applications of DBU and DBN,
see: Oediger, H.; Mo¨ller, F.; Eiter, K. Synthesis 1972, 591.
(9) For examples of nucleophilic catalysis with bicyclic amidines, see:
(a) DBU in Morita-Baylis-Hillman reaction: Aggarwal, V. K.; Mereu,
A. Chem. Commun. 1999, 2311. (b) DBU in esterification of carboxylic
acids with dimethyl carbonate: Shieh, W.-C.; Dell, S.; Repicˇ, O. J. Org.
Chem. 2002, 67, 2188. (c) DBU and DBN in cyanoacylation of ketones:
Zhang, W.; Shi, M. Org. Biomol. Chem. 2006, 4, 1671. (d) A chiral
derivative of 8 has been used in the synthesis of â-lactams: Taggi, A. E.;
Hafez, A. M.; Dudding, T.; Lectka, T. Tetrahedron 2002, 58, 8351. (e) For
leading references to other examples of nucleophilic behavior of DBU,
see: Ghosh, N. Synlett 2004, 574.
9.0 h
a Time required to achieve 50% conversion was determined by monitor-
1
ing the reactions by H NMR.
38
Org. Lett., Vol. 9, No. 1, 2007

Scheme 1. Synthesis of Bicyclic Amidines
Table 2. Acylation of Secondary Alcohols
a
entry
catalyst (mol %)
substrate (concn, M)
t1/2
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
none
PhCH(OH)Me (1.0)
PhCH(OH)Me (1.0)
PhCH(OH)Me (1.0)
PhCH(OH)Me (1.0)
PhCH(OH)Me (1.0)
PhCH(OH)Me (1.0)
PhCH(OH)Me (1.0)
PhCH(OH)Me (1.0)
PhCH(OH)Me (1.0)
PhCH(OH)Me (1.0)
PhCH(OH)Me (1.0)
PhCH(OH)Me (1.0)
PhCH(OH)Me (1.0)
PhCH(OH)Me (1.0)
PhCH(OH)Me (1.0)
PhCH(OH)Me (0.1)
PhCH(OH)Me (0.1)
PhCH(OH)Me (0.1)
PhCH(OH)Me (0.1)
PhCH(OH)Me (0.1)
PhCH(OH)Me (0.1)
Cyclohexanol (0.1)
Cyclohexanol (0.1)
Cyclohexanol (0.1)
i-PrCH(OH)Me (0.1)
i-PrCH(OH)Me (0.1)
i-PrCH(OH)Me (0.1)
27 h
DMAP 1 (0.1)
NMI 2 (5)
DHIP 3 (5)
DBN 8 (5)
DBU 9 (5)
10 (5)
11 (5)
12 (5)
13 (5)
14 (5)
3.5 min
9.0 min
5.2 h
15 min
17 h
11 h
12 h
18 h
8.7 h
core structure of BTM) with those of their ring-expanded
analogues 18 and 19 and found the latter (abbreviated herein
DHPB, for 3,4-dihydro-2H-pyrimido[2,1-b]benzothiazole) to
be remarkably active in acetylation of 1-phenylethanol,
surpassing even the “benchmark” catalyst DMAP.
Okamoto’s findings provided further stimulus for our effort
to identify active catalysts among nonaromatic bicyclic
isothioureas. Compounds 13-16 were easily prepared by
cycloalkylation of cyclic thioureas as shown in Scheme 2.12
DHPB 19 was also synthesized for comparison purposes.
1.6 h
7.5 min
2.4 h
THTP 15 (1)
THTP 15 (0.1)
16 (5)
2.7 h
DHPB 19 (0.1)
DMAP 1 (1)
THTP 15 (1)
DHPB 19 (1)
DMAP 1 (5)
THTP 15 (5)
DHPB 19 (5)
DMAP 1 (5)
THTP 15 (5)
DHPB 19 (5)
DMAP 1 (5)
THTP 15 (5)
DHPB 19 (5)
24 min
34 min
30 min
10 min
5 min
2 min
<2 min
18 min
46 min
10 min
24 min
4 h
Scheme 2. Synthesis of Bicyclic Isothioureas
To our delight, 15 (dubbed THTP, for 2,3,6,7-tetrahydro-
5H-thiazolo[3,2-a]pyrimidine) displayed outstanding catalytic
activity (Table 1, entry 13), comparable to DMAP (entry 2)
and DHPB (entry 15). All other bicyclic isothioureas were
several orders of magnitude less active than THTP (entries
11, 12, and 14).
Keeping in mind that the catalytic activity in acylation of
methanol might not necessarily correlate with that displayed
in the acylation of other types of substrates, we subjected
chiral catalysts 4-7 to the methanol test for comparison.
Indeed, only catalysts 4 and 5, but not 6 and 7, proved
reasonably active (Table 1, entries 16-19), despite the fact
that all four of them had proved competent in acylation of
secondary benzylic alcohols.
Clearly, we could not rely on the methanol test alone to
identify new catalysts. Thus, we decided to screen all the
achiral catalyst candidates again using acetylation of (()-
1-phenylethanol as the test reaction, to make sure that we
did not miss any potential catalysts for acylation of benzylic
substrates (Table 2).
20 min
a Time required to achieve 50% conversion was determined by monitor-
ing the reactions by H NMR.
1
In contrast to the methanol test, bicyclic isothioureas 14
and 16 exhibited modest activity in the acetylation of
phenylethanol (Table 2, entries 11 and 14). However,
amidines 9-12 and isothiourea 13 again proved virtually
inactive (entries 6-10). The failure of 13 in this test was
unexpected, given the fact that its phenyl-substituted ana-
logue 6 was a moderately active catalyst for the propionyl-
ation of the same substrate.7c,13 Initially, we were surprised
to find that under the standard conditions of our phenyl-
ethanol test DHPB 19 (entry 15) and especially THTP 15
(entries 12 and 13) fell far behind DMAP (entry 2). These
results were in stark contrast to both Okamoto’s observa-
tions11 with the same substrate and our methanol test (vide
supra). However, when we lowered the concentration of the
reactants to 0.1 M, while keeping the catalyst concentrations
constant at 0.001 M, both DHPB and THTP proved superior
to DMAP (entries 16-18). At 5 mol % catalyst loadings,
this effect was even more pronounced (entries 19-21).
Finally, we compared the relative activities of 1, 15, and 19
in acetylation of two other secondary alcoholsscyclohexanol
(10) Preparation of amidines 10-12 has been described: Rokach, J.;
Hamel, P.; Hunter, N. R.; Reader, G.; Rooney, C. S.; Anderson, P. S.;
Cragoe, E. J., Jr.; Mandel, L. R. J. Med. Chem. 1979, 22, 237.
(11) Kobayashi, M.; Okamoto, S. Tetrahedron Lett. 2006, 47, 4347. We
thank Prof. Okamoto for communicating their results to us prior to
publication.
(12) Preparation of bicyclic isothioureas 13-16 (isolated as hydrobro-
mides or hydrochlorides) has been described, e.g.: (a) Cort, L. A. J. Chem.
Soc. C 1966, 1226. (b) Dehuri, S. N.; Nayak, A. J. Indian Chem. Soc. 1982,
59, 1170.
(13) Preliminary studies suggest that this behavior may be due to rapid
catalyst deactivation, rather than inherent lack of activity.
Org. Lett., Vol. 9, No. 1, 2007
39

(entries 22-24) and 3-methyl-2-butanol (entries 25-27).
With these substrates, DHPB was still more active than
DMAP, albeit by a smaller margin than in the case of
1-phenylethanol, whereas THTP proved much less active.
Plots of percent conversion vs time provided in the Sup-
porting Information indicate that the initial reaction rate with
THTP is considerably faster than that with DMAP in
acetylation of phenylethanol and comparable in the acetyl-
ation of the rest of the substrates examined. However,
THTP undergoes gradual deactivation under the reaction
conditions, which accounts, at least partly, for its lower
overall efficacy.
While the validity of the above interpretation remains to
be verified by future investigations, our findings clearly
demonstrate that nonaromatic amidine derivatives can indeed
be highly catalytically active in acylation reactions. THTP,
an easily obtainable and extremely active achiral acylation
catalyst, holds significant promise as a catalaphore for
asymmetric catalyst design. Investigation of its utility in other
reactions requiring nucleophilic catalysis and preparation and
study of its chiral derivatives are ongoing and will be reported
in due course.
Acknowledgment. We gratefully acknowledge financial
support of this study by NIGMS (NIH R01 GM072682).
Mass spectrometry was provided by the Washington Uni-
versity Mass Spectrometry Resource, an NIH Research
Resource (Grant No. P41RR0954).
It is noteworthy that the structures of the three most
successful catalysts identified in our and Okamoto’s studiess
8, 15, and 19sall contain five-membered rings fused to a
tetrahydropyrimidine ring. The structure-activity trends
observed within this series may be interpreted as follows:
(1) the amidine moiety in itself constitutes a viable catala-
phore (8), but is subject to stringent geometric requirements;
(2) introduction of the sulfur atom results in a sharp increase
in catalytic activity (8 f 15), which may be ascribed to the
nonbonded S-O interactions with the carbonyl oxygen
leading to the stabilization of the N-acylated form;14 (3)
benzannellation leads to further enhancement of the catalytic
activity (15 f 19), which may be due to the aromatic
stabilization of the N-acylated form.
Supporting Information Available: Experimental pro-
1
cedures and H and 13C NMR spectra of compounds. This
material is available free of charge via the Internet at
http://pubs.acs.org.
OL0623419
(14) (a) For examples of intramolecular nonbonded S-O interactions in
structurally similar cases, see, e.g.: Nagao, Y.; Hirata, T.; Goto, S.; Sano,
S.; Kakehi, A.; Iizuka, K.; Shiro, M. J. Am. Chem. Soc. 1998, 120, 3104.
(b) We thank Prof. Glaser (University of MissourisColumbia) for bringing
this phenomenon to our attention.
40
Org. Lett., Vol. 9, No. 1, 2007