2
Tetrahedron Letters
for their activity at 10 mol% loading in toluene at reflux (Table
such as alkoxides or enolates in the coordination sphere of
zirconium could assist in lactam deprotonation, making it more
susceptible towards nucleophilic attack on the carboxylic acid.
Addition of a strong base such as sodium methoxide was
however not beneficial to increasing the extent of N-acylation
(Entries 9 and 11). Rather than increasing the nucleophilic
character of the lactam moiety in 2, sodium methoxide reacted
with the carboxylic acid 3 or with water to produce sodium
palmitate or sodium hydroxide. The latter may also facilitate
hydrolysis of the methyl ester moiety in 2, which promotes the
formation of N-pyroglutamoylpyroglutamic acid methyl ester and
hence reduces the selectivity towards 4. Finally, the
commercially available solid ZrO2 catalyst was completely
inactive (Entry 15), because the Lewis acidity was much less
pronounced compared to the other homogeneous catalysts and
only zirconium atoms that are exposed at the catalyst’s surface
were able to participate in the reaction.
Scheme 2. Lactamization of glutamic acid dimethyl ester (1) into
pyroglutamic acid methyl ester (2), followed by N-acylation with a
carboxylic acid. The N-acylated lactam can be converted back into
the desired N-acylglutamic acid dimethyl ester by acid-catalysed
methanolysis.
Table 1. Catalyst screening for the N-acylation of pyroglutamic
acid methyl ester (2) with palmitic acid (3)a
Next, the catalytic system was optimized in terms of
temperature by performing the N-acylation of
2
with
Zr(propoxide)4 and ZrOCl2•8H2O in other high-boiling solvents
(Table 2). The selection was limited to non-halogenated aromatic
solvents to enable the azeotropic removal of water from the
reaction medium. In the case of Zr(propoxide)4, the conversion of
2 increased to 97% by using mesitylene at reflux, viz. 165 °C
(Entry 5). Also for ZrOCl2•8H2O the conversion increased to
78% in mesitylene, while maintaining the selectivity at > 99%
(Entry 7).
Entry
Catalyst + additive
X
S
(%)b
< 1
16
6
(%)b
1
-
-
2
Ti(isopropoxide)4
Ti(isopropoxide)4
Sn(2-ethylhexanoate)2
La(OTf)3
77
6
3c
4
< 1
< 1
10
5
-
Table 2. Catalyst and solvent screening in the N-acylation of
pyroglutamic acid methyl ester (2) with palmitic acid (3)a
5
-
6
H3BO3
88
> 99
94d
79d
98
> 99
> 99
2
7
Zr(OAc)4
8
Zr(propoxide)4
Zr(propoxide)4 + NaOCH3
Zr(acac)4
78
73
71
27
63
18
37
< 1
9
10
11
12
13
14
15
Entry
Catalyst
Solvent
Temp.
(°C)
110
X
S
Zr(acac)4 + NaOCH3
ZrOCl2•8H2O
Zr(SO4)2•4H2O
ZrO(NO3)2
(%)b
< 1
78
(%)b
1
2
3
-
Toluene
-
Zr(propoxide)4
Zr(propoxide)4
Toluene
110
94c
91c
92
-
Toluene/
xylene (1:1)
Xylene
130
73
ZrO2 (< 5 µm)
aReagents and conditions: 2 (5.8 mmol), 3 (5.8 mmol), catalyst (0.58 mmol),
93c
94c
additive (5.8 mmol), toluene (20 mL), reflux (110 °C), 18 h, water removal
using a Dean-Stark apparatus. Conversion (X) of 2 and selectivity (S) for N-
palmitoylpyroglutamic acid methyl ester (4) determined by GC-FID and GC-
MS. cMethyl palmitate (5.8 mmol) was used instead of 3. dSelectivity towards
the methyl and propyl ester of N-palmitoylpyroglutamic acid.
4
5
6
7
Zr(propoxide)4
Zr(propoxide)4
ZrOCl2.8H2O
ZrOCl2.8H2O
144
165
110
165
93
97
63
78
b
Mesitylene
Toluene
> 99
> 99
Mesitylene
aReagents and conditions: 2 (5.8 mmol), 3 (5.8 mmol), catalyst (0.58 mmol),
solvent (20 mL), reflux, 18 h, water removal using a Dean-Stark apparatus.
bConversion (X) of 2 and selectivity (S) for N-palmitoylpyroglutamic acid
methyl ester (4) determined by GC-FID and GC-MS. cSelectivity towards the
methyl and propyl ester of N-palmitoylpyroglutamic acid.
Both the catalyst and the Dean-Stark system for water removal
are essential to produce N-palmitoylpyroglutamic acid methyl
ester (4) under these mild conditions, otherwise the reaction did
not proceed (Entry 1). For the titanium-based catalyst, the
conversion was limited to 16% (Entry 2), presumably because Ti
alkoxides are generally highly sensitive to degradation, e.g. by
hydrolysis or by reaction with the carboxylic acid. N-
Pyroglutamoylpyroglutamic acid methyl ester was observed as
the major side product, as a result of the self-condensation of 2.
In an attempt to prevent catalyst degradation, the carboxylic acid
3 was substituted with its methyl ester derivative, which also has
a slightly better leaving group (-OMe vs. -OH), but the yield of 4
could not be improved (Entry 3). Among the other catalysts
tested in the N-acylation of 2, boron-, tin- and lanthanum-based
catalysts appeared to be nearly inactive (Entries 4-6). Higher
activities were observed for several zirconium-based catalysts:
Zr(propoxide)4, Zr(acac)4 and ZrOCl2•8H2O produced 4 in > 60%
yield (Entries 8, 10 and 12), whereas Zr(SO4)2•4H2O and
ZrO(NO3)2 were far less active (Entries 13-14). These
observations can be explained by the higher solubility of
homogeneous zirconium catalysts with organic ligands in the
apolar reaction medium. Moreover, the presence of basic ligands
The long-term stability of active catalysts such as
Zr(propoxide)4 and Zr(acac)4 might be an issue, because
substantial discoloration of the reaction medium was observed.
Moreover, in the presence of Zr(propoxide)4 a complex product
mixture was obtained due to (trans)esterification of the substrates
and the product. On the other hand, ZrOCl2•8H2O represents an
already hydrolyzed zirconium compound and therefore is
expected to be more stable. Furthermore, ZrOCl2•8H2O has low
toxicity (LD50, oral rat = 2950 mg/kg),13 is commercially available,
inexpensive and air stable,14 making it a practical and easily
manageable catalyst.