Organic Letters
Letter
complex (Figure 1E). Gratifyingly, this change, which resulted in
urea 4a, led to a significant improvement in the observed rate
(Figure 2). We found that greater than 90% of thioester A was
converted to amide C within 2 h in the presence of this catalyst.
We prepared several analogues of 4a to improve on this result
and to analyze the key components of this trifunctional catalyst.
Modification of the thiol group to alcohol (4b) resulted in a large
decrease in the reaction rate, supporting the hypothesis that
thioester exchange is a critical component of the reaction
(Figures 1 and 2). Substitution of the dimethylamino group with
a pyridyl (4c) or pyrrolidinyl (4d) group caused an appreciable
decrease in product formation. Thiophenylureas are better
hydrogen bond donors than phenylureas and have been shown
to be generally superior as organocatalysts.5b We prepared
thiourea 4e to test the potential of this enhanced hydrogen-bond
donor on reactivity; however, we found that the thiourea thiol
can participate in thioester exchange with the substrate,
potentially reducing the effectiveness of this catalyst.
Urea 4a is built on the classical diphenyl urea catalyst with one
of the aromatic rings modified with bis-trifluoromethyl groups.
We tested if the addition of an electron-withdrawing group on
the aromatic ring would enhance the observed activity of the
catalyst. Urea 4f, which features a nitro group on the aromatic
ring para to the urea functionality, was found to provide a rate
increase to that of 4a. Lastly, we prepared a squaramide analogue
of 4a to evaluate a different hydrogen-bonding anion recognition
scaffold that has shown promise for the construction of receptors
and catalysts.10
For our model amide bond-forming reaction, squaramide 5
proved to be active but not as effective as 4a. As a control, we also
tested the potential of 4-mercaptophenylacetic acid (MPAA), a
well-known additive in NCL reactions,11 to accelerate amide
bond formation from thioester A. As expected, this simple thiol
did not have an effect on the observed rate of conversion (Figure
2).
The systematic design and evaluation of different scaffolds
provided urea 4a as a lead catalyst for amide bond formation
between model thioester and amine substrates. As part of these
optimization studies, we also gauged the performance of the
catalyst in different solvents (Table S2). As expected on the basis
of the mode of interactions, the catalyst proved to be more
effective in nonpolar solvents. We also prepared and evaluated
phenyl and o-nitrophenyl ester analogues of A and analyzed the
potential of 4a to catalyze their amide bond formation. We found
that oxoesters are not good substrates for the urea catalyst likely
due to their inability to participate in rapid thioester exchange
Table 1. Potential of Urea 4a To Catalyze Dipeptide
Formation from Amino Acid Thioesters
a
b
entry
dipeptide
catalyst (mol %)
time
1
FmocAlaAlaOMe
FmocAlaAlaOMe
FmocAlaValOMe
FmocAlaValOMe
FmocAlaValOMe
FmocAlaSarOMe
FmocAlaSarOMe
FmocPheAlaOMe
FmocProAlaOMe
FmocProAlaOMe
FmocValAlaOMe
FmocValAlaOMe
10
10 min
>20 days
>50 days
c
c
2
no catalyst
3
no catalyst
d
4
10
7 h
d
5
20
3 h
d
6
10
4 h
d
7
20
2 h
8
10
10 min
40 min
>50 days
4.5 h
9
10
no catalyst
10
c
10
11
12
20
2 h
a
Reaction conditions: Fmoc-Xaa-SPh (10 μmol), amino acid methyl
ester HCl salts (20 μmol), Et3N (20 μmol), and catalyst 4a in 1 mL of
b
toluene. Time for >98% conversion of Fmoc-Xaa-SPh based on
c
analysis of HPLC trace of the crude reaction mixtures. Estimated time
d
based on reaction progress after 24 h. Roughly 10−15% hydrolysis of
Fmoc-Ala-SPh was observed for these entries.
valine methyl ester requires roughly 7 h for completion (Table 1,
entry 4); we observe a significant amount (12%) of hydrolysis of
the thioester during the reaction. (We attribute any water present
in the reaction to the hygroscopic nature of the hydrochloric acid
salt of the amino acid methyl esters used directly from
commercial sources.) Addition of 20 mol % catalyst doubles
the reaction rate such that alanine-valine dipeptide may be
formed over 3 h.
We also analyzed the potential of the catalyst for secondary
amine coupling (Table 1, entries 6 and 7). While sarcosine (N-
methylglycine) condenses with alanine thioester within 4 h in the
presence of 10 mol % catalyst, we observed no dipeptide
formation with proline over 24 h (Table S1). As part of these
explorations, we also varied the thioester to determine the
effectiveness of the catalysts to potentially participate in thioester
exchange with various amino acids (Table S1). We observed less
variation in the observed rates with the thioester component than
with the amine, suggesting that reaction of the amine with the
thioester−4a intermediate is the slower step (vide infra).
Phenylalanine, lysine, proline, and arginine thioesters condensed
with alanine methylester in 10, 30, 40, and 60 min, respectively.
However, β-branching on the thioester partner also diminishes
the rate of the reaction on a similar level as observed with the β-
branched amine partner (Table 1, entries 11 and 12). No
epimerization was observed under the reaction conditions after
comprehensive evaluations with catalyst 4a (Figure S17).
We analyzed the kinetics of the amidation reaction between
Fmoc-valine thiophenylester (10 mM) and alanine methylester
(100 mM) in toluene. These substrates were chosen because
their dipeptide formation occurs over a sufficiently longer time
period to allow precise measurement of product formation. Urea
4a provides a 10000-fold rate acceleration for dipeptide
formation over the uncatalyzed reaction (Table 2; details are
requirement for different components of the trifunctional
catalyst with designed controls. Compounds 6−8, in which the
thiol, urea, or the tertiary amine groups are removed from 4a
(Table 2), were synthesized. Removal of either of these
functional groups leads to a significant loss in the observed
Next, we evaluated the suitability of the catalyst for amino
acids protected with the standard Fmoc group. We began by
analyzing the rate of alanine dipeptide formation. Condensation
of 10 mM Fmoc-alanine phenylthioester with 20 mM alanine
methyl ester in toluene leads to 5% formation of the Fmoc-Ala-
Ala-OMe dipeptide after 24 h at 22 °C. In the presence of 10 mol
% of 4a under the same conditions, the reaction is completed in
roughly 10 min (Table 1, entry 1). Encouraged by this finding, we
screened the catalyst for the formation of dipeptides to
determine the scope of the hydrogen-bonding catalyst to accept
The rates of product formation were monitored by HPLC. β-
Branched amino acid residues are often difficult to condense with
activated carboxylic acids. In keeping with their known lower
reactivity, we find that the reaction of alanine thioester with
C
Org. Lett. XXXX, XXX, XXX−XXX