Bronsted base-assisted boronic acid catalysis for the dehydrative intramolecular condensation of dicarboxylic acids

Article abstract of DOI:10.1021/ol102926n

Bronsted base-assisted boronic acid catalysis for the dehydrative self-condensation of carboxylic acids is described. Arylboronic acid bearing bulky (N,N-dialkylamino)methyl groups at the 2,6-positions can catalyze the intramolecular dehydrative condensation of di-and tetracarboxylic acids. This is the first successful method for the catalytic dehydrative self-condensation of carboxylic acids.(Figure Presented)

Full text of DOI:10.1021/ol102926n

ORGANIC
LETTERS
2011
Vol. 13, No. 5
892–895
Brønsted Base-Assisted Boronic Acid
Catalysis for the Dehydrative
Intramolecular Condensation
of Dicarboxylic Acids
Akira Sakakura,† Takuro Ohkubo,‡ Risa Yamashita,‡ Matsujiro Akakura,§, and
Kazuaki Ishihara*,‡,
EcoTopia Science Institute, Nagoya University, Japan, Graduate School of
Engineering, Nagoya University, Furo-cho, Chikusa-ku, Nagoya, 464-8603, Japan,
Department of Chemistry, Aichi University of Education, Japan, and JST, CREST, Japan.
ishihara@cc.nagoya-u.ac.jp
Received December 6, 2010
ABSTRACT
Brønsted base-assisted boronic acid catalysis for the dehydrative self-condensation of carboxylic acids is described. Arylboronic acid bearing
bulky (N,N-dialkylamino)methyl groups at the 2,6-positions can catalyze the intramolecular dehydrative condensation of di- and tetracarboxylic
acids. This is the first successful method for the catalytic dehydrative self-condensation of carboxylic acids.
Carboxylic anhydrides are widely used as activated car-
boxylic acids for the synthesis of (poly)esters, (poly)amides,
(poly)imides, and so on.1 For example, thermoplastic aro-
matic polyimides are synthesized from aromatic carboxylic
dianhydrides and diamines2 via the intermediary poly-
(amidocarboxylicacid)s.2b The aromatic carboxylic dianhy-
drides are industrially synthesized from the corresponding
tetracarboxylic acids using excess acetic anhydride as a
dehydration reagent. In this case, however, remaining acetic
anhydride and resultant acetic acid should be completely
removed before the subsequent reaction with a diamine
because the diamine preferentially reacts with acetic anhy-
dride or acetic acid to give the corresponding acetamide or
ammonium acetate. In another method, the aromatic car-
boxylic dianhydrides are synthesized by thermal dehydra-
tion of tetracarboxylic acids, which requires extremely high
reaction temperature (220-270 °C).3 From the perspective
of green chemistry, the catalytic dehydrative self-condensa-
tion of carboxylic acids should be an ideal method for the
synthesis of carboxylic anhydrides since it generates water as
the only byproduct.4 In addition, the reaction mixture can
be used for the subsequent reaction with a diamine without
any purification procedures. However, no practical method
of the catalytic dehydrative self-condensation of carboxylic
acids under mild conditions has been reported to date.
It has been reported that electron-deficient arylboronic
acids catalyze dehydrative amide and ester condensation
† EcoTopia Science Institute, Nagoya University.
‡ Graduate School of Engineering, Nagoya University.
§ Department of Chemistry, Aichi University of Education.
CREST.
(1) (a) Sakakura, A.; Kawajiri, K.; Ohkubo, T.; Kosugi, Y.; Ishihara,
K. J. Am. Chem. Soc. 2007, 129, 14775. (b) Otera, J. Esterification; Wiley-
VCH: Weinheim, 2003.
(2) (a) Ghosh, M. K.; Mittal, K. L. Polyimides, Fundamentals and
Applications; Marcel Dekker: New York, 1996. (b) Tomikawa, M.; Yoshida,
S.; Okamoto, N. Polym. J. 2009, 41, 604.
(3) Davidson, D.; Newman, P. J. Am. Chem. Soc. 1952, 74, 1515.
(4) Ishihara, K. Tetrahedron 2009, 65, 1085.
r
10.1021/ol102926n
Published on Web 02/08/2011
2011 American Chemical Society

reactions through the formation of monoacyl boronate
intermediates.5,6 However, these arylboronic acids are not
effective for the catalytic dehydrative self-condensation of
carboxylic acids because of the much lower nucleophilicity
of carboxylic acids than amines and alcohols. In 2006,
Whiting and co-workers reported arylboronic acid 1a
bearing an (N,N-diisopropylamino)methyl group at the
2-position as an efficient catalyst for amide condensation
reaction.7 In some cases, 1a is more advantageous than
electron-deficient arylboronic acids, although the role of
the (N,N-diisopropylamino)methyl group is unclear for
the amide condensation reaction.8 On the basis of these
previous results, we envisioned that an arylboronic acid of
general type 1 bearing an (N,N-dialkylamino)methyl
group7,9-14 would promote the dehydrative self-condensa-
tion of carboxylic acids under mild conditions through
activation of the other carboxyl group using the (N,N-
dialkylamino)methyl group as a Brønsted base in a mono-
acyl boronate intermediate 2 (Figure 1). We report here
Brønsted base-assisted boronic acid catalysis for the in-
tramolecular dehydrative condensation of di- and tetra-
carboxylic acids.
Figure 1. Working hypothesis: Brønsted base-assisted boronic
acid catalysis.
(10 mol %) that promoted the intramolecular dehydrative
condensation of phthalic acid (Table 1). The reaction was
conducted in nonane (bp 151 °C) under azeotropic reflux
conditions with the removal of water. Arylboronic acids
bearing electron-withdrawing substituents, some of which
Table 1. Catalytic Dehydrative Condensation of Phthalic Acida
According to our working hypothesis, we first investi-
gated the catalytic activities of various arylboronic acids
(5) (a) Ishihara, K. In Boronic Acids; Hall, D. G., Ed.; Wiley-VCH:
Weinheim, 2005; pp 377-409. (b) Ishihara, K.; Ohara, S.; Yamamoto, H.
J. Org. Chem. 1996, 61, 4196. (c) Ishihara, K.; Ohara, S.; Yamamoto, H.
Macromolecules 2000, 33, 3511. (d) Ishihara, K.; Kondo, S.; Yamamoto,
H. Synlett 2000, 1371. (e) Ishihara, K.; Ohara, S.; Yamamoto, H. Org.
Synth. 2002, 79, 176. (f) Maki, T.; Ishihara, K.; Yamamoto, H. Synlett
2004, 1355. (g) Maki, T.; Ishihara, K.; Yamamoto, H. Org. Lett. 2005, 7,
5043. (h) Maki, T.; Ishihara, K.; Yamamoto, H. Org. Lett. 2005, 7, 5047.
(i) Maki, T.; Ishihara, K.; Yamamoto, H. Org. Lett. 2006, 8, 1431. (j)
Maki, T.; Ishihara, K.; Yamamoto, H. Tetrahedron 2007, 63, 8645.
(6) (a) Collum, D. B.; Chen, S.; Ganem, B. J. Org. Chem. 1978, 43,
4393. (b) Houston, T. A.; Wilkinson, B. L.; Blanchfield, J. T. Org. Lett.
2004, 6, 679. (c) Mylavarapu, R. K.; GCM, K.; Kolla, N.; Veeramalla,
R.; Koilkonda, P.; Bhattacharya, A.; Bandichhor, R. Org. Process Res.
Dev. 2007, 11, 1065. (d) Al-Zoubi, R. M.; Marion, O.; Hall, D. G. Angew.
Chem., Int. Ed. 2008, 47, 2876.
(7) (a) Arnold, K.; Davies, B.; Giles, R. L.; Grosjean, C.; Smith,
G. E.; Whiting, A. Adv. Synth. Catal. 2006, 348, 813. (b) Arnold, K.;
Batsanov, A. S.; Davies, B.; Whiting, A. Green Chem. 2008, 10, 124. (c)
ꢀ
Arnold, K.; Davies, B.; Herault, D.; Whiting, A. Angew. Chem., Int. Ed.
2008, 47, 2673. (d) Georgiou, I.; Ilyashenko, G.; Whiting, A. Acc. Chem.
Res. 2009, 42, 756. (e) Charville, H.; Jackson, D.; Hodges, G.; Whiting,
A. Chem. Commun. 2010, 46, 1813.
(8) Marcelli, T. Angew. Chem., Int. Ed. 2010, 49, 6840.
(9) (a) Giles, R. L.; Howard, J. A. K.; Patrick, L. G. F.; Probert,
M. R.; Smith, G. E.; Whiting, A. J. Organomet. Chem. 2003, 680, 257. (b)
Coghlan, S. W.; Giles, R. L.; Howard, J. A. K.; Patrick, L. G. F.;
Probert, M. R.; Smith, G. E.; Whiting, A. J. Organomet. Chem. 2005,
690, 4784. (c) Blatch, A. J.; Chetina, O. V.; Howard, J. A. K.; Patrick,
L. G. F.; Smethurst, C. A.; Whiting, A. Org. Biomol. Chem. 2006, 4,
3297.
(10) (a) Yoshimoto, J.; Sandoval, C. A.; Saito, S. Chem. Lett. 2008,
37, 1294. (b) Oishi, S.; Yoshimoto, J.; Saito, S. J. Am. Chem. Soc. 2009,
131, 8748.
(11) (a) Wiskur, S. L.; Lavigne, J. J.; Ait-Haddou, H.; Lynch, V.;
Chiu, Y. H.; Canart, J. W.; Anslyn, E. V. Org. Lett. 2001, 3, 1311. (b)
Zhu, L.; Shabbir, S. H.; Gray, M.; Lynch, V. M.; Sorey, S.; Anslyn, E. V.
J. Am. Chem. Soc. 2006, 128, 1222. (c) Collins, B. E.; Sorey, S.;
Hargrove, A. E.; Shabbir, S. H.; Lynch, V. M.; Anslyn, E. V. J. Org.
Chem. 2009, 74, 4055.
(12) Sumerin, V.; Schulz, F.; Atsumi, M.; Wang, C.; Nieger, M.;
€
a The reaction of phthalic acid (2.5 mmol) was conducted in the
presence of an arylboronic acid (1 or 10 mol %) in solvent (10 mL).
€
Leskela, M.; Repo, T.; Pyykko, P.; Rieger, B. J. Am. Chem. Soc. 2008,
130, 14117.
1
b Determined by H NMR analysis. c Boiling point. d Data in parenth-
(13) (a) Wulff, G. Pure Appl. Chem. 1982, 54, 2093. (b) Lauer, M.;
Wulff, G. J. Organomet. Chem. 1983, 256, 1.
(14) Bosch, L. I.; Fyles, T. M.; James, T. D. Tetrahedron 2004, 60,
11175.
eses refer to conversion yield in the presence of 1,2,2,6,6-pentamethylpi-
peridine (20 mol %).
Org. Lett., Vol. 13, No. 5, 2011
893

are highly active dehydrative amide condensation
catalysts,5-7 showed low catalytic activities (entries 1-3).
Bulky 2,4,6-triisopropylphenylboronic acid also showed
poor catalytic activity (entry 4). In contrast, as expected,
arylboronic acid 1a7 bearing an (N,N-diisopropylamino)-
methyl group at the 2-position showed high catalytic
activities in nonane, even without electron-withdrawing
substituents (entry 5). Interestingly, arylboronic acids 1b
and 1cbearing an(N,N-diisopropylamino)methylgroupat
the 3- and 4-positions also showed high catalytic activities
(entries 6 and 7). Theoretical calculations suggest that the
(N,N-diisopropylamino)methyl group at the 2- and 3-posi-
tions can form hydrogen bonding with the carboxyl group of
the monoacyl boronate intermediates generated from 1a and
1b.15 In the case of 1c, the 4-(N,N-diisopropylamino)methyl
group may interact with the carboxyl group with the aid of
the other molecules of 1c or phthalic acid through hydrogen
bondings. Detailed mechanistic studies are in progress.
Since the use of heptane (bp 98 °C) as a solvent sig-
nificantly decreased the yield of phthalic anhydride due to
the low reaction temperature, further investigation was
conducted to promote the reaction under mild conditions.
As a result, we found that the introduction of a methyl
group at the 6-position surprisingly increased the catalytic
activity despite the steric hindrance around the boronic
acid moiety (catalyst 1d, entry 8). Furthermore, the use of
arylboronic acid 1f bearing two (N,N-diisopropylamino)-
methyl groups at the 2,6-positions successfully improved
the reactivity (entry 10), while arylboronic acid 1e bearing
two (N,N-diisopropylamino)methyl groups at the 3,5-posi-
tions showed lower activity than 1f (entries 9 versus 10).
Arylboronic acid 1g bearing sterically bulky 1-(2,2,6,6-
tetramethylpiperidinyl)methyl groups at the 2,6-positions
showed the highest catalytic activity (entry 11). Despite the
high catalytic activities of 1f and 1g for the intramolecular
condensation, they showed poor catalytic activities for the
intermolecular dehydrative self-condensation of 4-phenyl-
butyric acid.15
Whiting and co-workers reported that for the dehydra-
tive amide condensation, 1a, which showed no NfB
chelation in both solution and solid states, was more active
than phenylboronic acid bearing a sterically less hindered
(N,N-dimethylamino)methyl group at the 2-position,
which showed NfB chelation.7a,9b,11a,13,14 The steric hin-
drance of the N,N-diisopropylamino group might prevent
the intramolecular interaction between the boronic acid
group and the N,N-diisopropylamino group (NfB
chelation) that causes the inactivation of the boronic acid
group (Figure 2). The crystal structure of 1f shows that the
N,N-diisopropylamino groups interact with protons of the
boronic acid group.15 Saito and co-workers also reported
the intramolecular hydrogen bonding between (N,N-
dialkylamino)methyl groups and boron-coordinated
water in aqua-aminoorganoboron compounds.10a Based
on these analyses and the high catalytic activities of 1f and
1g, we can propose some plausible but significant roles of
the bulky (N,N-dialkylamino)methyl groups in the present
Figure 2. Plausible reaction mechanism.
catalysis (Figure 2). The condensation of phthalic acid
would occur via monoacyl boronate 3 as an active inter-
mediate. Then, the N,N-dialkylamino nitrogen and the N,
N-dialkylammonium proton of 3 would synergistically
promote the intramolecular cyclization to 4: the former
would act as Brønsted base to activate the carboxyl group,
and the latter would act as Brønsted acid to activate the
carbonyl group through two consecutive hydrogen-bond-
ing interactions supported by the B-hydroxyl group
(CdO H-O H-Nþ). The subsequent elimination
3 3 3
3 3 3
of phthalic anhydride from the intermediate 4 would also
be synergistically promoted by the N,N-dialkylamino ni-
trogen and the N,N-dialkylammonium proton of 4. The
intramolecular hydrogen-bonding interaction between the
B-hydroxyl group and the ortho-Brønsted base substituent
of arylboronic acids has also been suggested by the theo-
retical study for the dehydrative amide condensation
catalyzed by 2-iodophenylboronic acid.8 In addition, the
introduction of two bulky substituents at the 2,6-positions
might prevent formation of less active species such as
triarylboroxines 5.11b,13b
Propionitrile (EtCN, bp 97 °C) was suitable as a solvent,
probably due to the high polarity and good solubility. In
EtCN, only 1 mol % of 1f and 1g could catalyze the reaction
to give phthalic anhydride in respective yields of 71 and 99%.
The combination of phenylboronic acid (10 mol %) and
1,2,2,6,6-pentamethylpiperidine (20 mol %) showed no cata-
lytic activity (entry 12). This control experiment indicates that
an intramolecular Lewis basic function on 1g was essential to
the successful promotion of the condensation.7a
With highly active dehydration catalyst 1g in hand, we
next examined the catalytic dehydrative condensation of
aromatic tetracarboxylic acids 6 to dianhydrides 7 (Table 2).
The reactions of 6a-e were successfully catalyzed by 1g
(15) See the Supporting Information for details.
894
Org. Lett., Vol. 13, No. 5, 2011

Table 2. 1g-Catalyzed Dehydrative Condensation of Aromatic
Tetracarboxylic Acids 6a
Table 3. 1g-Catalyzed Dehydrative Condensation of Aliphatic
Dicarboxylic Acids 10a
a Unless otherwise noted, the reaction of 6 (2.5 mmol) was conducted
in the presence of 1g (1 mol %) in n-PrCN (10 mL) for 12 h. b Isolated
yield. Data in parentheses refer to conversion yield in the absence of 1g.
c TsOH (10 mol %) was used as a catalyst. d Conversion yield was
>99%.
a Unless otherwise noted, the reaction of 10 (2.5 mmol) was con-
ducted in the presence of 1g (1 mol %) in EtCN (10 mL) for 12-24 h.
b Determined by 1H NMR analysis. c Data in parentheses refer to
conversion yield in the absence of 1g. d Data in parentheses refer to yield
in the presence of TsOH (10 mol %) instead of 1g. e The reaction was
conducted in n-BuCN for 24 h. f The reaction was conducted in n-PrCN.
(1 mol %) in butyronitrile (n-PrCN, bp 115 °C) under
azeotropic reflux conditions to give the corresponding
dianhydrides 7a-e in excellent isolated yields. Carboxylic
dianhydrides 7 were easily isolated by filtration of the
reaction mixture. The use of n-PrCN was effective mainly
because of the rather good solubility of 6 and 7 under the
reaction conditions. In contrast, dianhydrides 7 were
obtained in poor yields in the absence of 1g. TsOH was
almost inert for the reaction of 6a (entry 2).
The reaction of cyclic cis-1,2-dicarboxylic acid10a gavethe
cis-anhydride 11a in 99% yield (entry 1). Interestingly,
cyclic trans-1,2-dicarboxylic acid 10b selectively gave the
corresponding cis-anhydride 11a via epimerization (entry
2). Not only cyclic 1,2-dicarboxylic acids but also linear
1,2-dicarboxylic acids, such as succinic acid (10c) and N-
Boc-aspartic acid (10d), could be converted to the corre-
sponding anhydrides 11c and 11d17 in good yields (entries 3
and 4). Although glutaric acid (10e), a 1,3-dicarboxylic
acid, was less reactive than 1,2-dicarboxylic acids, the use
of valeronitrile (n-BuCN, bp 140 °C) as a solvent signifi-
cantly improved the yield of 11e (entry 5). 3-Substituted
1,3-dicarboxylic acids 10f-h significantly improved the
reactivity to give the corresponding anhydrides 11f-h in
high yields (entries 6-8).
Scheme 1. Catalytic Synthesis of Diimide 8 and One-Pot
Synthesis of Bis(amidocarboxylic acid)s 9
In conclusion, we have developed bifunctional Brønsted
base-assisted boronic acid catalyst 1g, an arylboronic acid
bearing two bulky (N,N-dialkylamino)methyl groups at
the 2,6-positions, for the green synthesis of carboxylic
anhydrides. This is the first successful method of the
catalytic intramolecular dehydrative condensation of di-
and tetracarboxylic acids.
On the basis of the present catalytic condensation, the
direct synthesis of a diimide from a tetracarboxylic acid
and a primary amine was investigated as a model reaction
for the synthesis of polyimides2 (Scheme 1). When the 1g-
catalyzed condensation of 6d was conducted in the pre-
sence of aniline (2.5 equiv), diimide 8 was obtained in 98%
yield by simple filtration of the reaction mixture. Catalyst
1g could be easily recovered by concentration of the
filtrate, and could be reused repeatedly without any loss
of activity. Furthermore, one-pot synthesis of bis-
(amidocarboxylic acid)s 9 could be achieved by the direct
addition of aniline (3 equiv) to the reaction mixture of 7a at
ambient temperature (Scheme 1).2b After being stirred for
3 h, a 1:1 isomeric mixture of 9 was obtained in 99% yield
by filtration of the reaction mixture.
Acknowledgment. Financial support for this project
was provided by JSPS.KAKENHI (20245022, 20655019),
the Toray Science Foundation, NEDO, Mitsubishi Rayon
Co., Ltd., and the Global COE Program of MEXT. Calcu-
lations were performed at the Research Center or Compu-
tational Science (RCCS), Okazaki Research Facilities,
National Institutes of Natural Science (NINS).
Supporting Information Available. Experimental pro-
cedures and full characterization of new compounds. This
material is available freeof chargevia the Internetat http://
pubs.acs.org.
(16) (a) Atodiresei, I.; Schiffers, I.; Bolm, C. Chem. Rev. 2007, 107,
5683. (b) Hamersak, Z.; Ivsic, T. Tetrahedron: Asymmetry 2009, 20,
1095. (c) Peschiulli, A.; Procuranti, B.; O’ Connor, C. J.; Connon, S. J.
Nature Chem. 2010, 2, 380.
We next investigated the 1g-catalyzed dehydrative con-
densation of aliphatic dicarboxylic acids 10 (Table 3).16
(17) The enantiomeric excess of the obtained 11d was 41%.
Org. Lett., Vol. 13, No. 5, 2011
895