Because simple carboxylates do not react under the esteri-
fication conditions described here, it is unlikely that the same
mechanism of activation is occurring in these reactions. Both
the neutral ester 2 and anionic species 3 may be present in
alcohol solution, but once esterification occurs to produce
4, neither of these complexes can form. It is unclear at present
whether a free alcohol molecule attacks the carbonyl carbon
of either 2 or 3 to yield product or whether the alcohol is
transferred in an intramolecular manner through 3 to produce
4. In either case, an electrophilic boron species will then be
regenerated to catalyze another esterification (Scheme 1).7
It is interesting to note that this reaction did not occur to
any appreciable extent when equimolar amounts of R-
hydroxyacid, alcohol, and boric acid were mixed in aqueous
solution at room temperature. The amount of borate ester
will be greatly reduced under these conditions, and it is
apparently necessary that the alcohol be present in excess
to react with this species. Even in 2:1 water/methanol solvent
systems, no appreciable reaction occurred after 24 h. This
result was important to us because it ensured that undesired
esterification did not interfere with fluorescent sensing
measurements of R-hydroxyacid binding by boronates.2
Although rigorously dried alcohol solvents were not neces-
sary for the esterification reactions, reagent grade material
was used in all cases. The lower yield observed for lactic
acid is likely due to the fact that this acid contained ca. 12%
water in the commercial form used.
Table 1. Isolated Yields from Boric Acid Catalyzed (10 mol
%) Methyl Ester Synthesis (18 h, Room Temperature)
(>5%) formed when the reaction was run at 0.5 M
concentration. Diluting the reaction mixtures below 0.25 M
reduced the amounts of these biproducts and increased the
yield of the monoester product. It is important to note that
although diacids such as tartrate and malate can form esters
upon prolonged reflux in alcoholic solvents, no appreciable
amount (<5%) of ester formation occurred after several days
at room temperature in the absence of boric acid. Even in
the presence of the catalyst, benzoic acid and succinic acid
were quite unreactive. While N-protected R-amino acids have
the potential to serve as chelating ligands for boric and
boronic acids, neither N-tosylglycine nor N-BOC-alanine
underwent significant esterification under standard condi-
tions.
Two entries in Table 1 demonstrate that the reaction
conditions allow for selective methyl esterification of R-
hydroxycarboxylic acids in the presence of â-hydroxycar-
boxylates. Malic acid contains a hydroxy group that is in
the R-position relative to one carboxylic acid and in the
â-position relative to another. Citric acid contains two such
â-carboxylates in addition to the R-hydroxyacid. The car-
boxylate adjacent to the hydroxy group in both malic and
citric acid reacts more rapidly, providing their respective
monomethyl esters as the major products with very little
diester (<10%) formed in either case. The malate monoester
can also be synthesized by hydrolysis of malic acid anhy-
drides,8 but such selective ring-opening reactions are only
amenable to diacids that are capable of forming cyclic
anhydrides. The protocol described here is not limited in this
regard. The citrate sym-monomethyl ester has previously
been synthesized in two steps using formaldehyde to create
an acetal that reacts with methanol to create the methyl ester.9
This monoester can alternatively be synthesized through
selective saponification of trimethyl citrate.10 The procedure
described here requires only one step.
Yamamoto has used trifluorophenylboronic acid as a
catalyst for amide formation under anhydrous conditions.3b
In the absence of protic solvents, an OH from the boronate
can serve to activate the carbonyl through formation of a
mixed anhydride such as the boxed species in Scheme 1.
Scheme 1. Possible Mechanisms of Catalysis
Other alcohols can be used as solvents to create esters
using this method (Table 2). In these cases, reactions
proceeded more slowly at room temperature (the conversion
to diethyl L-tartrate was only 50% complete after 24 h at
(7) A number of exchanging boron species are likely to be present at
any one time in alcohol solution and these are simply represented here as
B(OR′)3 where R′ ) H and/or Me(Et/iPr).
(8) Miller, M. J.; Bajwa, J. S.; Mattingly, P. G.; Peterson, K. J. Org.
Chem. 1982, 47, 4928.
(9) Lee, B. H.; Miller, M. J. J. Org. Chem. 1983, 48, 24.
(10) Pearce, K. N.; Creamer, L. K. Aust. J. Chem. 1975, 28, 2409.
680
Org. Lett., Vol. 6, No. 5, 2004