toward nucleophiles.15 This could lead to lower and also more
variable loadings. Therefore, the effective loading of the resin
must be verified for each batch with an “end-use” test. With
this improved synthesis, the counteranion is the non-
nucleophilic triflate, circumventing this problem. The use
of polymer-supported resin 1 was tested for the reaction
between carboxylic acids and amines (Table 1). The reagent
proved to be extremely effective at promoting these cou-
plings, giving complete conversions (HPLC analysis) after
short reaction times in most cases.
Scheme 1. Synthesis of Polymer-Supported Reagent 1
The reaction is operatively simple and does not require
preactivation of the carboxylic acid. All reagents were just
dissolved in DCM prior to addition of the resin. Very
importantly, the required workup was very simple and easy
to apply to parallel chemistry: a filtration on an amino-
functionalized solid-phase extraction plug,16 followed by
washing with DCM and solvent evaporation afforded the
products in high purity (HPLC, HPLC-MS, and NMR). A
very slight excess of carboxylic acid (1.05 equiv) was
employed to ensure complete reaction of the amine reagent,
the excess of acid being trapped either on resin 1 or on the
SPE column.17
of a series of polymer-supported N-alkyl-2-halo pyridinium
salts and their use as coupling reagents for the synthesis of
esters and amides.
The first strategy investigated for the preparation of the
polymer-supported pyridinium salts involved the reaction of
2-chloropyridine with chloromethylpolystyrene (Merrifield
resin). Various conditions were tested (from room temper-
ature to refluxing THF, with or without the addition of
equimolar NaI). Contrary to the results reported by Convers,
all resins obtained had a very poor loading (ca. 10% from
elemental analysis).
It was therefore decided to employ a resin bearing a better
leaving group than a simple halide. Wang resin12 was
activated in situ with trifluoroacetic anhydride, forming a
triflate ester, which was immediately substituted by the
2-chloropyridine present (Scheme 1).13 The excess of 2-chlo-
ropyridine also acts as the base necessary for the formation
of the triflate ester, so that the addition of another base is
not necessary. The same reaction has been applied to
2-bromo and 2-fluoropyridine with excellent results.
The reaction gave very good results, especially with
sterically hindered or other unactivated amines. For example,
couplings of proline or valine methyl ester with Boc-
tryptophan proceeded with good yields (Table 1, entries 2,
3, and 6). The coupling between Boc-protected tryptophan
and valine methyl ester was shown to proceed with very little
racemization, giving a de of 99% (HPLC). The reaction is
relatively insensitive to steric bulk on the carboxylic acid
partner: (4-nitro)-phenylalanine methyl ester could be ef-
ficiently coupled to pivalic acid in just 2 h (entry 9).
However, the coupling of a C-terminal dipeptide (Cbz-Gly-
PheOH) with valine methyl ester gave poor isolated yields
(entry 7).
N-Methylaniline, poorly nucleophilic because of both steric
and electronic effects, coupled very efficiently with acetic
acid (entry 11). The reaction with the more sterically hindered
pivalic acid (entry 13) required harsher reaction conditions:
using 3 equiv of acid and heating the reaction mixture under
microwave irradiation for 10 min at 100 °C gave a product
containing only ca. 10% of starting amine, which could be
easily removed by filtration on a sulfonic acid derivatized
SPE column.18
Compared to the procedure reported by Convers, which
gives yields of 33-72% depending on scale, mode of stirring,
and loading of the starting Merrifield resin, this operationally
simple one-pot protocol consistently gave complete conver-
sion of the Wang resin to the halo-pyridinium salts, as judged
by elemental analysis, on several 5-g batches.14 It was also
found that mass increase of the resin upon reaction (from
5.0 to 7.1 g starting from 1.7 mmol/g Wang resin) correlates
well with the loading measured by elemental analysis and
can be used as a quick and cheaper way to determine an
approximate loading of the resin.
Another major advantage of the reported synthesis is that
iodide-based additives are not needed. It is well-documented
how the iodide anion can, in some circumstances, substitute
the chlorine in the pyridinium salts and that the resulting
2-iodopyridinium salts do not activate carboxylic acids
Formamides were obtained in good yield and excellent
purity without requiring the use of an excess of formic acid
(entry 14).
(15) (a) Folmer, J. J.; Acero, C.; Thai, D. L.; Rapoport, H. J. Org. Chem.
1998, 63, 8170-8182. (b) Bradlow, H. L.; Vanderwerf, C. A. J. Org. Chem.
1951, 16, 1143-1152. (c) Sutherland, J. K.; Widdowson, D. A. J. Chem.
Soc. 1964, 4650-4651.
(12) Hydroxymethylphenoxymethyl polystyrene (cross-linked with 1%
divinylbenzene) 150-300 µm, loading 1.7 mmol/g, from Polymer Labo-
ratories.
(13) Experimental Procedure. Wang resin (5.00 g, 8.5 mmol) was
suspended in dry DCM (50 mL). 2-Chloropyridine (4.0 mL, 42.5 mmol, 5
equiv) was added, and the mixture was cooled with an ice bath. Trifluo-
romethanesulfonic anhydride (2.0 mL, 11.9 mmol, 1.4 equiv) was added
dropwise, and after 5 min the ice bath was removed. The mixture was stirred
overnight at room temperature. The resin was collected by filtration and
washed with DCM, DMF, and DCM and dried under vacuum.
(14) Elemental analysis of the resin gave 1.83% N, 4.22% S, and 5.85%
Cl, which corresponds to a loading of 1.3 mmol/g (nitrogen and sulfur
analyses) or 1.6 mmol/g (chlorine analysis, probably due to ca. 1% w/w
DCM still trapped in the resin).
(16) Isolute NH2 SPE column, 2 g, part no. 470-0200-D.
(17) Typical Experimental Procedure. The carboxylic acid (0.16 mmol,
1.05 equiv), the alcohol or amine (0.15 mmol, 1 equiv), and TEA (0.45
mmol, 3 equiv) were dissolved in anhydrous DCM (3 mL). When needed,
PS-DMAP was added (20 mg, 0.03 mmol, 0.2 equiv). Resin 1 (240 mg,
0.3 mmol, 2 equiv) was added, and stirring was continued until completion
of the reaction (HPLC). The resin was removed by filtration, and the filtrates
were passed through an amino-functionalized SPE column (2 g). The resin
and SPE column were washed with 10 mL of DCM, and then the combined
filtrates were evaporated under vacuum to afford the desired ester or amide.
(18) Isolute SCX SPE column, 2 g, part no. 530-0200-D.
4580
Org. Lett., Vol. 6, No. 24, 2004