Asymmetric direct amide synthesis by kinetic amine resolution: A chiral bifunctional aminoboronic acid catalyzed reaction between a racemic amine and an achiral carboxylic acid

Article abstract of DOI:10.1002/anie.200705643

(Chemical Equation Presented) Let's resolve this: A planar chiral benzimidazole ferroceneboronic acid catalyst demonstrates low to moderate asymmetric induction in the kinetic resolution of racemic α-substituted benzylamines by direct amide formation with

Full text of DOI:10.1002/anie.200705643

Angewandte
Chemie
DOI: 10.1002/anie.200705643
Amidation
Asymmetric Direct Amide Synthesis by Kinetic Amine Resolution: A
Chiral Bifunctional Aminoboronic Acid Catalyzed Reaction between a
Racemic Amine and an Achiral Carboxylic Acid**
Kenny Arnold, Bryan Davies, Damien HØrault, and Andrew Whiting*
The formation of amide bonds by direct reaction of amines
with carboxylic acids by thermal[1,2] or catalytic methods[2–5] is
generally a high temperature process. There have been no
reports to date of any reactions that involve asymmetric
induction during direct amide formation, with the exception
of an enzyme-catalyzed process.[6] Lower temperatures are
often preferred for any reaction as it can reduce the amount of
reagent, reactant, or thermal product degradation. A more
important consideration is that asymmetric induction pro-
cesses are usually more efficient at lower temperatures
because small differences in energy between diasteroisomeric
transition states are amplified; however, there are an
increasing number of examples where this is not the case
and improved asymmetric induction can be obtained at higher
temperatures.[7] With these considerations in mind, we
endeavored to develop a catalytic direct amide formation
under mild conditions, though still well above room temper-
ature. The current limiting temperature appears to be 858C,
the temperature at which reactions can be carried out with
bifunctional aminoboronic acid catalysts by using azeotropic
water removal in fluorobenzene.[2] Although these reaction
conditions are relatively mild compared to all other direct
amide formation reactions with boron-derived catalyts,[4,5]
they do not completely preclude the thermal direct amide
formation with the more reactive carboxylic acid/amine
combinations.[2] Though the prospect of developing asym-
metric catalysts for amide formation may not look promis-
ing,[7] we report herein preliminary results that demonstrate
asymmetric processes are possible under elevated temper-
atures with the development of a planar chiral ferrocene-
derived bifunctional aminoboronic acid catalyst.
Our recent interests in the development of bifunctional
aminoboronic acids as potential catalysts[8] led to the devel-
opment of N,N-di-iso-propylbenzylaminoboronic acid type
catalysts[2] such as commercially available 1 and planar chiral
ferrocene analogue 2. We recently reported using a (À)-
sparteine-directed metallation to provide 2 in 96% ee.[9]
Attempts to prepare other planar chiral aminoboronic acids
such as 3 were hindered by the lack of direct access by
asymmetric deprotonation methods of ferrocene benzimida-
zole precursors.[8d] However, this was circumvented by the
synthesis of 3 from the known[10] (pS)-bromoferrocene
aldehyde 4 (Scheme 1).
Scheme 1. Synthesis of the (pS)-2-(2-boronoferrocenyl)-N-n-butylbenz-
imidazole (3).
[*] K. Arnold, Dr. D. HØrault, Dr. A. Whiting
Department of Chemistry
Durham University, Science Laboratories
South Road, Durham, DH1 3QZ (UK)
Fax: (+44)191-384-4737
Coupling aldehyde 4 with diamine 5[8c] in the presence of
Oxone gave bromoferrocenylbenzimidazole 6 in 89% yield
and 99% ee (determined by chiral HPLC methods, see the
Supplementary Information), which compares with a litera-
ture value of 98% ee for aldehyde 4.[10] Subsequent lithium–
halogen exchange of 6 provided enantioenriched catalyst 3
(60% yield).
A series of parallel experiments were conducted with
catalysts 2 and 3, in which carboxylic acids 7a–b were reacted
with amines 8a–c in refluxing fluorobenzene by using
activated 3-Š molecular sieves in a Soxhlet (solvent drying
system). The uncatalyzed (background) reaction was com-
E-mail: andy.whiting@durham.ac.uk
B. Davies
Discovery Research Systems Chemistry
GlaxoSmithKline Pharmaceuticals
Stevenage, Hertfordshire, SG1 2NY (UK)
[**] We thank the EPSRC for research grants (GR/R24722/01, GR/
S85368/01, and GR/T22759/01) and GlaxoSmithKline Pharma-
ceuticals for a CASE award (K.A.). We also thank the EPSRC mass-
spectroscopy service, Swansea.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2008, 47, 2673 –2676
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2673

Communications
pared with the reactions having 10 mol% catalyst (either 2 or
3); the reactions were monitored every four hours over a
48 hour period by HPLC methods (Table 1 and Figures 1 and
2).
Table 1: Evaluation of the catalysts with different amines and carboxylic
acids during the amide formation.
Figure 2. Formation and ee value of (1-phenylethyl)-4-phenylbutyramide
Entry Catalyst
7
8
Conversion [%]
ee [%][e] Product
^
9b versus time relative to the thermal reaction =background;
(equiv)
9
~
&
=catalyst; =ee value).
1[a]
2[a]
3[a]
4[a]
5[a]
6[a]
7[a]
background
1
2
3
a
a
a
a
a
b
b
a(1)
a(1)
a(1)
a(1)
a(2)
a(1)
a(1)
0(48 h) n.a.
52(48 h) n.a.
38(48 h)
a
a
a
a
a
b
b
0
in a 52% yield with catalyst 1 (Table 1, entry 3 versus entry 2)
without any asymmetric induction. In contrast, catalyst 3
provides amide 9a in 41% ee (Table 1, entry 4), albeit with a
reduced conversion of 21%, which is approximately two
turnovers in 48 hours for the 10 mol% catalyst loading used.
The concentration of the amine substrate was increased to
two equivalents in an attempt to increase the potential for
kinetic resolution (Table 1, entry 5), however, this did not
increase the asymmetric induction and, more problematically,
it retarded the reaction. Additional examination of catalyst 3
with other racemic a-alkylbenzylamines (Table 1, entries 7, 8,
10, 11, 13 and 14) did not reveal higher levels of asymmetric
induction. However, as expected,[2] the conversions to amides
were considerably increased when the more reactive carbox-
ylic acid 7b was used. Asymmetric induction was still
observed, with the highest being 29% ee after only 12 hours
(Table 1, entry 7).
21(48 h) 41(S)
13(48 h) 18(S)
11(48 h) n.a.
3
background
3
34(12 h), 73(48 h) 29(S),[c]
19(S)[d]
8[a]
3
b
b
b
b
b
b
b
b
b
a(2)
b(1)
b(1)
b(2)
c(1)
c(1)
c(2)
a(1)
a(1)
67(48 h) 15(S)
13(48 h) n.a.
85(48 h) 9(S)
64(48 h) 6(S)
12(48 h) n.a.
65(48 h) 7(S)
63(48 h) 8(S)
<1(48 h) n.a.
21(48 h) 16(S)
b
c
c
9[a]
background
3
3
10[a]
11[a]
c
12[a] background
d
d
d
b
b
13[a]
14[a]
3
3
15[b] background
16[b]
3
[a] Reaction carried out in refluxing fluorobenzene with 1 equiv 7. Np=
napthyl; n.a. =not applicable. [b] Reaction carried out in iPr2O with
1 equiv 7. [c] The ee value after 12 h. [d] The ee value after 48 h.
[e] Absolute configuration indicated in parentheses.
The lower asymmetric induction observed when employ-
ing more reactive carboxylic acid 7b can be explained by
examining Figure 2. As the conversion increases, the back-
ground reaction contributes 11% to the formation of product
amide 9b, and therefore, the ee value drops from an initial
29% to 19%. Also notable is that during all of the reactions
involving catalyst 3 (Table 1), partial degradation of 3 by
proto-deboronation to form 10 was observed[11] when using
Figure 1. Formation of (1-phenylethyl)benzamide 9a versus time.
~
&
^
=background; =catalyst 1; =catalyst 2; ”=catalyst 3).
Catalyst 2 is more active than 3 (Figure 2), as we expected
on the basis of previous investigations involving the corre-
sponding phenyl-derived bifunctional catalysts;[2] however, 2
is less reactive than corresponding phenyl-derived catalyst 1.
There is no background reaction for the formation of amide
9a under these reaction conditions (Table 1, entry 1). Amide
9a is obtained in a 38% yield in the presence of catalyst 2, and
phenylbutyric acid 7b (re-isolation of 35–40% of intact
catalyst 3) and benzoic acid 7a (70–80% of intact catalyst 3
isolated; see the Supporting Information). This competing
side reaction may reduce the asymmetric efficiency of the
reaction because of the slow increase in the concentration of
boric acid,[5] particularly in the reactions with phenylbutyric
2674 www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 2673 –2676

Angewandte
Chemie
acid, which may increase the formation of the racemic amide
product (above any direct background contributions).[2a,5]
Hence, the low enantiomeric excess observed might be
explained by the competitive formation of boric acid, coupled
with direct background contributions for the more reactive
carboxylic acid (7b). Although boric acid is being produced in
these reactions, the boric acid is only a more efficient catalyst
at higher temperatures than employed herein.[2] Therefore,
the major contributions to asymmetric induction are from the
actual catalyst and the background reaction as indicated in
Figure 2. Over time the concentration of catalyst 3 is
decreases and the background reaction increasingly contrib-
utes to product formation. To examine whether it was possible
to overcome the competitive effects of both catalyst 3
decomposition and the background contribution to the
formation of amide 9b, an alternative lower boiling solvent
for azeotroping water was examined. Parallel reactions were
carried out in iPr2O (688C). The background reaction was
essentially negligible under the lower temperature conditions
(Table 1, entry 15), and the reaction with catalyst 3 (Table 1,
entry 16) had a conversion of 21% (73% in fluorobenzene)
and the asymmetric induction was 16%. Although a drop in
boiling point did result in preventing background contribu-
tions to the reaction, it did not result in improved asymmetric
induction.
this unpromising, yet green reaction may be manipulated to
achieve good enantioselectivity and engender new interest in
this old process. The discovery that catalyst 3 is able to react
with one enantiomer of a chiral a-substituted benzylamine
with low to moderate selectivity, and couple it to a moderately
activated acylating agent demonstrates the major potential of
this type of process. The current challenges are to develop
more reactive catalysts that are capable of producing higher
asymmetric induction, and to examine alternative methods of
removing water from these reactions to drive them efficiently
to completion. Developments in these directions will be
reported in due course.
Since N,N-diisopropylbenzylamine-2-boronic acid (1) is a
more active catalyst than ferrocene analogue 2, phenylbenz-
imidazole boronic acid 11 was screened to see if the same
would hold true in the benzimidazole series. However,
phenylbenzimidazole boronic acid 11 did not have any
activity in any the direct formation of amide bonds, including
those reported herein. This could be due in part to B–N
chelation, which is not present in diisopropylamine-derived
catalysts 1 and 2, and ferrocene benzimidazole 3 show. In
contrast, phenylbenzimidazole boronic acid 11 exists as a
mixture of B–N chelates, free boronic acid, and boroxine in
solution as evidenced by 11B NMR spectroscopy. Also bor-
oxine shows partial B–N interactions in the solid state[8c] (two
of the three boron atoms are partially tetrahedral or four-
coordinate). In addition, the benzimidazole moiety is closer to
the boronic acid group in 11 than in 3, suggesting that the
benzimidazole is acting as more than just a steric blocking
group and that the distance between the boron and nitrogen
atoms is crucial for an efficient reaction involving proton
transfer, and therefore catalysis. One explanation for the
observed asymmetric induction obtained from catalyst 3 is
that the benzimidazole group does not deprotonate the
ammonium salt, but preferentially selects (S)-amine 8 by
hydrogen bonding and effectively delivers the incoming
nucleophile to activated diacylboronate[12] 12.[13] In contrast,
the lack of asymmetric induction afforded by corresponding
catalyst 2 may result from the increased basicity of the N,N-
diisopropylamine group, which simply deprotonates the salt
to allow either enantiomer of an uncomplexed free amine
substrate to approach the acylating agent (13).
Received: December 10, 2007
Published online: February 27, 2008
Keywords: boronates · homogeneous catalysis ·
.
kinetic resolution · selectivity
[1] a) A. W. Hofmann, Ber. Dtsch. Chem. Ges. 1882, 15, 977 – 978;
b) J. A. Mitchell, E. E. Reid, J. Am. Chem. Soc. 1931, 53, 1879 –
1883; c) D. Davidson, P. Newman, J. Am. Chem. Soc. 1952, 74,
1515 – 1516; d) B. S. Jursic, Z. Zdravkovski, Synth. Commun.
1993, 23, 2761 – 2770.
[2] a) K. Arnold, B. Davies, R. L. Giles, C. Grosjean, G. E. Smith, A.
Whiting, Adv. Synth. Catal. 2006, 348, 813 – 820; b) K. Arnold,
A. S. Batsanov, B. Davies, A. Whiting, Green Chem. 2008, 10,
124 – 134.
[3] a) Y. I. Leitman, M. S. Pevzner, Zh. Prikl. Khim. 1963, 36, 632 –
639; b) W. Walter, H. Besendorf, O. Schnider, Helv. Chim. Acta
1961, 44, 1546 – 1554.
[4] a) K. Ishihara, H. Kurihara, H. Yamamoto, J. Org. Chem. 1996,
61, 4196 – 4197; b) K. Ishihara, S. Ohara, H. Yamamoto, Macro-
molecules 2000, 33, 3511 – 3513; c) K. Ishihara, S. Kondo, H.
Yamamoto, Synlett 2001, 1371 – 1374; d) K. Ishihara, S. Kondo,
H. Yamamoto, Org. Synth. 2002, 79, 176 – 185; e) T. Maki, K.
Ishihara, H. Yamamoto, Org. Lett. 2005, 7, 5043 – 5046; f) T.
Maki, K. Ishihara, H. Yamamoto, Org. Lett. 2006, 8, 1431 – 1434.
[5] P. Tang, Org. Synth. 2002, 81, 262 – 272.
[6] V. Gotor, Bioorg. Med. Chem. 1999, 7, 2189 – 2197.
[7] a) D. Enders, H. Gielena, J. Runsinka, K. Breuerb, S. Brodeb, K.
Boehnb, Eur. J. Inorg. Chem. 1998, 913 – 919; b) Y.-G. Wang, M.
Ueda, X. Wang, Z. Han, K. Maruoka, Tetrahedron 2007, 63,
6042 – 6050; c) P. Bachu, J. S. Gibson, J. Sperry, M. A. Brimble,
Tetrahedron: Asymmetry 2007, 18, 1618 – 1624; d) A. T. Axtell, J.
Klosin, K. A. Abboud, Organometallics 2006, 25, 5003 – 5009.
[8] a) R. L. Giles, J. A. K. Howard, L. G. F. Patrick, M. R. Probert,
G. E. Smith, A. Whiting, J. Organomet. Chem. 2003, 680, 257 –
262; b) S. W. Coghlan, R. L. Giles, J. A. K. Howard, M. R.
Although direct amide formation has been known and
employed for 150 years, asymmetric direct amide formation is
less well-known, and the reaction itself has attracted remark-
ably little attention until recent years. This demonstrates that
Angew. Chem. Int. Ed. 2008, 47, 2673 –2676
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
2675

Communications
Probert, G. E. Smith, A. Whiting, J. Organomet. Chem. 2005,
[11] a) H. C. Brown, D. Basavaiah, S. U. Kulkarni, J. Org. Chem.
1982, 47, 3808 – 3810; b) H. C. Brown, D. Basavaiah, S. U.
Kulkarni, H. D. Lee, E.-I. Negishi, J.-J. Katz, J. Org. Chem.
1986, 51, 5270 – 5276.
[12] The intervention of a diacylboronate species as the active
acylating agent is not proven, but this type of intermediate has
been detected by electrospray mass-spectrometry. a) Alterna-
tively a monoacylboronate could be involved; see reference [4a].
[13] Although hydrogen bonds become weaker at higher temper-
atures (see C. Czeslik, J. Jonas, Chem. Phys. Lett. 1999, 302, 633 –
638), the interaction shown in 12 would be enhanced by a
positive-charge-assisted hydrogen bond (see P. Gilli, V. Berto-
lasi, V. Ferretti, G. Gilli, J. Am. Chem. Soc. 1994, 116, 116 – 122).
690, 4784 – 4793; c) A. J. Blatch, O. V. Chetina, J. A. K. Howard,
L. G. F. Patrick, C. A. Smethurst, A. Whiting, Org. Biomol.
Chem. 2006, 4, 3297 – 3302; d) D. Hꢀrault, K. Aelvoet, A. J.
Blatch, A. Al-Majid, C. A. Smethurst, A. Whiting, J. Org. Chem.
2007, 72, 71 – 75.
[9] A. S. Batsanov, D. Hꢀrault, J. A. K. Howard, L. G. F. Patrick,
M. R. Probert, A. Whiting, Organometallics 2007, 26, 2414 –
2419.
[10] a) O. Riant, O. Samuel, H. Kagan, J. Am. Chem. Soc. 1993, 115,
5835 – 5836; b) O. Riant, O. Samuel, T. Flessner, S. Taudien,
H. B. Kagan, J. Org. Chem. 1997, 62, 6733 – 6745.
2676 www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 2673 –2676