On the racemisation of aspartic anhydride during its preparation

Article abstract of DOI:10.1016/j.tetasy.2007.07.003

Due to the possible differentiation of both carboxyl groups, N-protected aspartic anhydride is a useful starting material in synthesis. However, most methods published in the literature for its formation lead to partial racemisation, which is generally not mentioned. Herein we report a comparison between the main methods published, with accurate measurements of the enantiomeric purity of this compound.

Full text of DOI:10.1016/j.tetasy.2007.07.003

Tetrahedron: Asymmetry 18 (2007) 1625–1627
On the racemisation of aspartic anhydride during its preparation
*
´ ´
Frederic Buron, Geoffrey Deguest, Laurent Bischoff, Corinne Fruit and Francis Marsais
Laboratoire de Chimie Organique Fine et He´te´rocyclique, IRCOF-INSA Rouen, CNRS UMR 6014, BP 08,
76131 Mont-Saint-Aignan Cedex, France
Received 15 June 2007; accepted 5 July 2007
Abstract—Due to the possible differentiation of both carboxyl groups, N-protected aspartic anhydride is a useful starting material in
synthesis. However, most methods published in the literature for its formation lead to partial racemisation, which is generally not men-
tioned. Herein we report a comparison between the main methods published, with accurate measurements of the enantiomeric purity of
this compound.
Ó 2007 Published by Elsevier Ltd.
1. Introduction
2. Results and discussion
N-protected aspartic anhydride 1 was first described more
than 70 years ago.1 It has generally been used as a synthon
with potential differentiation of both acid moieties, thus
avoiding selective protection and activation sequences.
Regioselective nucleophilic opening of the latter was
achieved with alcohols,2 amines,3 hydrides4 and carbon
nucleophiles.5 Frequently described syntheses using this
reagent lead to either aspartam 26 or lactone 34 (Scheme 1).
2.1. Determination of optical purities
Unfortunately, the literature lacks reliable values of enan-
tiomeric purities for these compounds. The anhydride itself
is commercially available from AldrichÒ, its [a]D value is
given in methanol. However, since N-protected aspartic
anhydride reacts with methanol, this measurement must
be taken as soon as the anhydride is solubilised. Indeed,
the specific rotation of a methanol solution of N-Cbz aspar-
20
tic anhydride, that is, ½aꢀD ¼ ꢁ25 (c 1, MeOH); t = 1 min;
O
O
slowly decreased overnight. Methanol evaporation gave a
compound whose NMR spectrum was consistent with the
a-monoester Z-Asp(OH)-OMe. In addition, two different
crystalline forms were described7 for the N-Cbz anhydride
and their common [a]D value was given in glacial acetic acid.
RHN
+H3N
RHN
COOMe
Ph
NH
COO-
O
O
O
O
1a : R= Cbz
1b : R= Boc
1c : R= Fmoc
Aspartam 2
3
Searching for a reliable method to measure the racemisa-
tion of the title compound led us to hydrolyse it in water
under mild conditions. Thus, whatever the regioselectivity
of the nucleophilic opening with water, N-protected aspar-
tic acid will be the only product resulting from the hydro-
lysis. This allowed us a comparison with an authentic
sample of enantiomerically pure N-protected aspartic acid.
No racemisation occurred during the opening step, since
enantiomeric excesses of up to 100% could be measured
after the hydrolysis of the anhydride (Scheme 2).
Scheme 1. N-protected aspartic anhydride and products arising from this
reactant.
Despite the usefulness of this molecule, little has been pub-
lished concerning the racemisation of the a-carbon during
the dehydration of the diacid. As we have encountered this
problem during the preparation of N-Cbz aspartic anhy-
dride 1a,5b we have further examined the stereochemical
outcome of the dehydration of N-Cbz aspartic acid, using
standard literature procedures.
These racemisations were measured using a chiral Chiracel
OD column, eluting with [heptane–i-PrOH–TFA =
80:20:0.5]. Retention times were 5.98 min for the (S)-enan-
*
Corresponding author. Fax: +33 235 5229 62; e-mail: laurent.bischoff@
insa-rouen.fr
0957-4166/$ - see front matter Ó 2007 Published by Elsevier Ltd.
doi:10.1016/j.tetasy.2007.07.003

1626
F. Buron et al. / Tetrahedron: Asymmetry 18 (2007) 1625–1627
Table 2. Dehydration of N-Fmoc aspartic acid
O
RHN
RHN
RHN
COOH
COOH
COOH
H2O
dehydration
Entry
1
R
Experimental conditions
ee
86
O
COOH
Fmoc
TFAA 1.4 M equiv,
O
AcOEt (0.26 M), 35 °C, 1 h
DCC 1 M equiv, rt, 12 h
e.e. = 100%
e.e. determined by HPLC
2
Fmoc
100
Scheme 2. Determination of enantiomeric purity after re-opening to the
diacid.
avoided such racemisation and gave the anhydride with a
good 92% ee (entry 7).
tiomer, and 7.51 min for the (R)-enantiomer, first deter-
mined from racemic and L-enantiomer of N-Cbz aspartic
acid.
More reactive dehydrating agents were also proposed for
this transformation. Dehydration with dicyclohexyl carbo-
diimide (DCC) afforded a nearly enantiomerically pure
material (entry 8), whereas the stereogenic centre seemed
slightly more sensitive to treatment with di-tert-butyl dicar-
bonate Boc2O in the presence of pyridine (entry 9).
2.2. Dehydration procedures
Many published reports account for the use of N-protected
aspartic anhydride, but seldom give unambiguous enantio-
meric ratios of the latter. In view of this study, we have
examined most dehydration procedures reported in the
literature, as listed in Table 1.
Treatment with trifluoroacetic anhydride in ethyl acetate
under moderate heating was published on a kg-scale.15
Using this method on a 1 g scale, afforded enantiomerically
pure material. Eventually, we verified a sample of commer-
cially available anhydride, the latter exhibiting a complete
enantiomeric purity.
Dehydration protocols using acetic anhydride are the most
commonly employed methods for obtaining 1a (entries 1–
7). The original reference1 describes the use of Ac2O, but
does not actually provide an accurate protocol in terms
of time and temperature. Our results suggest that racemisa-
tion depends on time, temperature and on the molar excess
of acetic anhydride. The least racemising conditions were
obtained with a low amount of Ac2O, either neat (entries
1–3), or in ethyl acetate solution (entry 4). Increasing the
temperature, reaction times or excess amounts of acetic
anhydride resulted in higher racemisations, up to 40% ee
(entries 2, 5 and 6).
On the basis of these results, we checked whether the least
racemising procedures would be applicable to similar N-
Fmoc anhydride, in order to use it in solid-phase peptide
synthesis (Table 2). N-Fmoc aspartic acid readily dehy-
drated in the presence of TFAA, but unexpected, partial
racemisation was observed. Finally, dehydration of the
latter with DCC gave an excellent enantiomeric excess.
Apparently, N-Fmoc anhydride is more sensitive to racemi-
sation in acidic medium than N-Cbz anhydride.
Nevertheless, the same dehydration carried out during a
very short period of time under microwave irradiation
3. Conclusion
Table 1. Formation of N-Cbz aspartic anhydride 1a using different
dehydration protocols described in the literature
Despite what is generally assumed, the dehydration of N-
protected aspartic acid often leads to high rates of racemi-
sation. This work allows a choice of the experimental con-
ditions for the preparation of this reagent, taking into
account the enantiomeric purity obtained through different
methods.
Entry
1
Experimental conditions
eea
92
Ref.
7
Ac2O (0.7 ml for 1 g Cbz-Asp),
45 °C 15 min then rt 2 h
Ac2O 3 M equiv,
100 °C 5 min
Ac2O 2.1 M equiv,
50–60 °C 2.5 h
Ac2O 1.5 M equiv,
AcOEt (1 M), 54 °C, 2 h
Ac2O 8.5 M equiv,
25 °C, 24 h
2
3
4
5
6
7
40
80
84
45
76
92
8
9
4. Experimental
10
11
12
13
4.1. Procedures for non-racemising dehydration
4.1.1. TFAA protocol.15 To a solution of N-Cbz aspartic
acid (1 g, 3.74 mmol) in ethyl acetate (14 mL) was added
trifluoroacetic anhydride (0.73 mL, 5.24 mmol). After heat-
ing for 1 h at 35 °C, concentration in vacuo afforded pure
N-Cbz aspartic anhydride (0.93 g, 100%).
Ac2O 30 M equiv,
rt, 48 h then 50–55 °C, 2 h
Ac2O 3 M equiv,
microwaves, 210 W, 1 min
DCC 1 M equiv, rt, 12 h
Boc2O 1 M equiv,
8
9
96
86
2
14
0.6 equiv pyridine,
4.1.2. DCC protocol.2 To a solution of N-Cbz aspartic
acid (1 g, 3.74 mmol) in THF (7 mL) was added di-
cyclohexylcarbodiimide (772 mg, 3.74 mmol) at ꢁ5 °C.
After stirring overnight at room temperature, DCU was
precipitated at 0 °C and removed by filtration. Concentra-
tion in vacuo afforded pure N-Cbz aspartic anhydride
(0.93 g, 100%).
CH2Cl2 (0.5 M), rt 8 h
TFAA 1.4 M equiv,
AcOEt (0.26 M), 35 °C, 1 h
Commercially available
material (Aldrich)
10
11
100
100
15
a Enantiomeric excesses of N-Cbz aspartic acid were measured after sub-
sequent hydrolysis, using a HPLC Chiracel OD chiral column.

F. Buron et al. / Tetrahedron: Asymmetry 18 (2007) 1625–1627
1627
4.2. Determination of the rates of racemisation
References
1. Bergmann, M.; Zervas, L. Berichte 1932, 1192.
2. Jouin, P.; Castro, B.; Zeggaf, C.; Pantaloni, A.; Senet, J. P.;
Lecolier, S.; Sennyey, G. Tetrahedron Lett. 1987, 28,
1665.
3. Huang, X.; Luo, X.; Roupioz, Y.; Keillor, J. J. Org. Chem.
1997, 62, 8821.
4. Mc Garvey, G. J.; Hiner, R. N.; Matsubara, Y.; Oh, T.
Tetrahedron Lett. 1983, 24, 2733.
Re-opening by mild aqueous hydrolysis was carried out by
stirring a water suspension (20 mL for 200 mg) of the anhy-
drides in water at room temperature for three hours, and
freeze-drying of the solution. An aliquot was dissolved in
a mixture of heptane and isopropanol (80:20). Enantio-
meric excesses were measured on a chiral Chiracel OD col-
umn, eluting with [heptane–i-PrOH–TFA = 80:20:0.5] at a
flow rate of 1 ml/min, measuring at 254 nm. Retention
times were 5.98 min for the (S)-enantiomer and 7.51 min
for the (R)-enantiomer, first determined from commercially
available racemic and (L)-aspartic acid.
5. (a) Brinkmann, T.; Gilg, A.; Hamm, A.; Luscht, H.;
¨
Morbach, G.; Uzar, H. C. Tetrahedron: Asymmetry 2000,
11, 3827; (b) Deguest, G.; Bischoff, L.; Fruit, C.; Marsais, F.
Tetrahedron: Asymmetry 2006, 17, 2120.
6. Mazur, R. H.; Schlatter, J. M.; Goldkamp, A. H. J. Am.
Chem. Soc. 1969, 91, 2684.
7. John, W. D.; Young, G. T. J. Chem. Soc. 1954, 2870.
8. Aebi, J. D.; Seebach, D. Helv. Chim. Acta 1985, 68, 1507.
9. Ebeling, S. C. Food Chem. 1998, 61, 107.
With N-Fmoc aspartic acid, retention times were 10.79 min
for the (S)-enantiomer and 15.32 min for the (R)-
enantiomer.
10. Yang, C. P.; Su, C. S. J. Org. Chem. 1986, 51, 5186.
11. Kemp, D. S.; Fotouhi, N.; Boyd, J. G.; Carey, R. I.; Ashton,
C.; Hoare, J. Int. J. Pept. Protein Res. 1988, 31, 359.
12. Carson, J. F. Synthesis 1983, 669.
13. Luppi, G.; Tomasini, C. Synlett 2003, 797; Luppi, G.; Villa,
M.; Tomasini, C. Org. Biomol. Chem. 2003, 1, 247.
14. Pozdnev, V. F. Int. J. Pept. Protein Res. 1992, 40, 407.
15. Melillo, D. G.; Larsen, R. D.; Mathre, D. J.; Shukis, W. F.;
Wood, A. W.; Colleluori, J. R. J. Org. Chem. 1987, 52,
5143.
Acknowledgments
Laetitia Bailly and Catherine Fiol (CNRS UMR 6014) are
gratefully acknowledged for the determination of enantio-
meric purities by chiral HPLC. The MENESR (Ministere
de l’Education Nationale, l’Enseignement Superieur et de
la Recherche) is gratefully acknowledged for a grant to
G.D. and a ATER position for F.B.
`
´