Using mixed anhydrides from amino acids and isobutyl chloroformate in N-acylations: A case study on the elucidation of mechanism of urethane formation and starting amino acid liberation using carbon dioxide as the probe

Article abstract of DOI:10.1016/S0040-4039(03)01329-7

A case study on the elucidation of mechanism of urethane by-product formation and starting amino acid liberation during the conventional two-step isobutyl chloroformate mediated N-acylation is described using carbon dioxide offgas as the probe. The main reason for the urethane formation and starting amino acid liberation was found to be the formation of the symmetrical anhydride of the amino acid during the preparation of the mixed carboxylic-carbonic anhydride intermediate, as determined by quantifying the evolved carbon dioxide. New conditions were developed to minimize this side reaction.

Full text of DOI:10.1016/S0040-4039(03)01329-7

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 44 (2003) 5543–5546
Using mixed anhydrides from amino acids and isobutyl
chloroformate in N-acylations: a case study on the elucidation of
mechanism of urethane formation and starting amino acid
liberation using carbon dioxide as the probe
Apurva Chaudhary, Michael Girgis,* Mahavir Prashad,* Bin Hu, Denis Har, Oljan Repicˇ and
Thomas J. Blacklock
Process Research and Development, Chemical and Analytical Development, Novartis Institute for Biomedical Research,
One Health Plaza, East Hanover, NJ 07936, USA
Received 29 April 2003; revised 13 May 2003; accepted 13 May 2003
Abstract—A case study on the elucidation of mechanism of urethane by-product formation and starting amino acid liberation
during the conventional two-step isobutyl chloroformate mediated N-acylation is described using carbon dioxide offgas as the
probe. The main reason for the urethane formation and starting amino acid liberation was found to be the formation of the
symmetrical anhydride of the amino acid during the preparation of the mixed carboxylic–carbonic anhydride intermediate, as
determined by quantifying the evolved carbon dioxide. New conditions were developed to minimize this side reaction. © 2003
Elsevier Science Ltd. All rights reserved.
Isobutyl chloroformate (IBCF)1 is a well-known cou-
pling agent in peptide synthesis2–4 and has been found
to be a reagent of choice for large scale syntheses.5–8
Conventional conditions involve two steps: activation
of the carboxyl group with IBCF in the presence of a
tertiary amine base to form the mixed carboxylic–car-
bonic anhydride, and the reaction of this intermediate
with the amine component.9 One of the drawbacks with
this methodology is a side reaction,4,5,10–15 in some cases
leading to the formation of urethane by-product and
liberation of the starting amino acid. The starting
amino acid detected at the end of the reaction is due to
the side reaction and not to incomplete reaction, an
after thought. Although the aminolysis at the undesired
carbonyl in the mixed carboxylic–carbonic anhydride
has been implicated as the cause of the urethane
formation5,10–15 and liberation of the starting amino
acid, to the best of our knowledge there is no method
known in the literature to ascertain the actual cause of
this side reaction in the two-step procedure. Another
possible pathway that could result in this side reaction
is the symmetrical anhydride formation12 in the first
activation step of the amino acid with IBCF. In this
paper we report the elucidation of the mechanism of
urethane formation and the starting amino acid libera-
tion by determining, and quantifying, the evolved car-
bon dioxide during the first step involving the
preparation of the mixed carboxylic–carbonic anhy-
dride intermediate, which shows that symmetrical anhy-
dride formation is the main cause of this side reaction.
In a development program we needed to develop a
large-scale synthesis of BOC-(S)-3-(2-naphthyl)alanyl-
N-benzyl-N-methylamide (2) from BOC-L-3-(2-naph-
thyl)alanine (1). Synthesis of 2, using conventional
two-step conditions involving the addition of IBCF to a
solution of BOC- -3-(2-naphthyl)alanine (1) and a base
L
(4-methylmorpholine or N,N-dimethylbenzylamine) in
ethyl acetate or toluene at −12°C and then the addition
of N-benzylmethylamine to the resulting mixed carb-
oxylic–carbonic anhydride (I, Scheme 1) led to sub-
stantial amounts of urethane 3 and liberated starting
acid 1. This side reaction was also scale-dependent.5
Urethane 3 did not present any purification problems
since it was easily removed in the mother liquor during
the isolation of crystalline (S)-3-(2-naphthyl)alanyl-N-
benzyl-N-methylamide hydrochloride after the depro-
tection of 2 with HCl. However, multiple additions of
IBCF and N-benzylmethylamine were necessary to
recycle in situ the liberated starting material 1 and drive
the reaction to amide 2 to completion.5 The formation
* Corresponding authors.
0040-4039/03/$ - see front matter © 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0040-4039(03)01329-7

5544
A. Chaudhary et al. / Tetrahedron Letters 44 (2003) 5543–5546
Scheme 1.
of the urethane by-product (3) could either be due to
the aminolysis with N-benzylmethylamine at the unde-
sired carbonyl5,10–15 in the mixed carboxylic–carbonic
anhydride intermediate (I) or because of the symmetri-
cal anhydride (II) formation, leaving unconsumed
IBCF in the reaction mixture that then reacts with
N-benzylmethylamine. In both cases the reaction would
also result in the liberation of the starting material 1.
To avoid the multiple additions of reagents and make
the process more convenient for manufacturing, it
became necessary for us to understand the mechanism
of this side reaction. We have developed a novel
method using the formation of CO2 offgas as the probe
for elucidating the mechanism of the formation of
urethane 3 and liberation of the starting material 1.
This method was based on the rationale that any
symmetrical anhydride (II) formation would be accom-
panied by the formation of an equivalent amount of
CO2 offgas during the preparation of mixed carboxylic–
carbonic anhydride (I) from 1 by treatment with IBCF.
This was because during the addition of IBCF to 1 and
the base, initially formed I could react further with 1,
present in excess under conventional conditions, to
afford the symmetrical anhydride (II), isobutanol, and
CO2, leaving some of isobutyl chloroformate uncon-
sumed. Thus, if the CO2 formation were observed
during the preparation of mixed carboxylic–carbonic
anhydride (I) (before amine addition) then the urethane
3 would form as a result of the reaction of unconsumed
IBCF with N-benzylmethylamine. Otherwise, its origin
must be aminolysis at the undesired carbonyl in I.
Thus, by quantifying the amount of CO2 formed during
the preparation of I, one can also determine whether
there was any aminolysis at the undesired carbonyl in I.
This would establish whether only one or both effects
are operating and to what extent. Although in situ
Fourier-transform infrared spectroscopy16 could be
used to address this problem, application of this tech-
nique, which was tried initially, was complicated by
several factors, including the necessity of synthesizing
authentic samples of various species and the complexity
of spectra due to various carbonyl group containing
species and the possibility of overlapping peaks.
The addition of isobutyl chloroformate to a solution of
1 and N,N-dimethylbenzylamine in toluene led to a
significant amount of CO2 formation. Characterization
and quantification of evolved CO2 indicated the forma-
tion of 17.5% (Table 1) of symmetrical anhydride (II)
along with I.17 Further reaction of this reaction mixture
with N-benzylmethylamine led to only minor (<2.0%)18
aminolysis at the undesired carbonyl in I. These results
were consistent with the amount of liberated amino
acid 1 at the end of the reaction as determined by
HPLC. Formation of substantial amounts of CO2 in
the first step clearly demonstrated that the formation of
the urethane 3 and liberation of the starting material 1
was mainly due to the symmetrical anhydride (II) for-
mation during the preparation of I. No corresponding
isobutyl ester of 1 was detected suggesting that liber-
ated isobutanol did not react with I to produce CO2.
t-Butyloxycarbonyl protecting group on nitrogen is
known19 to prevent the formation of oxazolinone that
would have also resulted in CO2 production from I. We
reasoned that the symmetrical anhydride formation,
and in turn the multiple additions of reagents to recycle
in situ the liberated starting material 1, could be
avoided by reverting the order of addition of IBCF and
the amino acid. One-stage conditions20 involving the
addition of isobutyl chloroformate to a solution of 1,
N-methylbenzylamine and base (4-methylmorpholine)
in THF also led to the urethane formation and libera-
tion of ꢀ30% of 1, again requiring multiple additions
of isobutyl chloroformate to drive the reaction to com-
pletion. We have now developed a new procedure21 that
involves a reverse addition of 1 and N,N-dimethylbenz-
ylamine in toluene to a solution of isobutyl chlorofor-
mate in toluene for the formation of I, followed by the
addition of N-benzylmethylamine. These conditions led
to completion of the reaction and avoided the multiple

A. Chaudhary et al. / Tetrahedron Letters 44 (2003) 5543–5546
5545
Table 1. Comparison of conventional and new method
4. Johnson, E. P.; Cantrell, W. R.; Jenson, T. M.; Miller, S.
A.; Parker, D. J.; Reel, N.; Sylvester, L. G.; Szendroi, R.
J.; Vargas, K. J.; Xu, J.; Carlson, J. A. Org. Proc. Res.
Dev. 1998, 2, 238–244.
5. Prashad, M.; Prasad, K.; Repicˇ, O.; Blacklock, T. J.;
Prikoszovich, W. Org. Proc. Res. Dev. 1999, 3, 409–415.
6. Franzen, H. M.; Bessidskaia, G.; Abedi, V.; Nilsson, A.;
Nilsson, M.; Olsson, L. Org. Proc. Res. Dev. 2002, 6,
788–797.
Conditions
Symmetrical anhydride Aminolysis18 at
(II) based on CO2
undesired carbonyl in
I
Conventional
New
17.5%
1.9%
<2.0%
<2.0%
7. Beaulieu, P. L.; Lavallee, P.; Abraham, A.; Anderson, P.
C.; Boucher, C.; Bousquet, Y.; Duceppe, J. S.; Gillard, J.;
Gorys, V.; Grand-Maitre, C.; Grenier, L.; Guindon, Y.;
Guse, I.; Plamondon, L.; Soucy, F.; Valois, S.; Wernic,
D.; Yoakim, C. J. Org. Chem. 1997, 62, 3440–3448.
8. Brady, S. F.; Freidinger, R. F.; Paleveda, W. J.; Colton,
C. D.; Homnick, C. F.; Whitter, W. L.; Curley, P.; Nutt,
R. F.; Veber, D. F. J. Org. Chem. 1987, 52, 764–769.
9. Bailey, P. D. An Introduction to Peptide Chemistry; John
Wiley & Sons: Chichester, New York, Brisbane, Toronto
and Singapore, 1992; pp. 131–132.
additions of isobutyl chloroformate. Only 1.9% (Table
1) of symmetrical anhydride (II) formed under these
conditions during the reaction of 1 with isobutyl chlo-
roformate as determined by CO2 analysis and confi-
rmed by measuring liberated 1. Such conditions
involving the reverse addition of the amino acid to
IBCF are not reported previously for the N-acylation
(peptide coupling) and they represent a new methodol-
ogy that was racemization-free. This newly developed
reaction was successfully scaled-up in our pilot plant on
a 63.0 kg scale of 1.
10. Sole, N.; Torres, J. L.; Anton, J. M. G.; Valencia, G.;
Reig, F. Tetrahedron 1986, 42, 193–198.
11. Chen, F. M. F.; Lee, Y.; Steinauer, R.; Benoiton, N. L.
Can. J. Chem. 1987, 65, 613–618.
12. Chen, F. M. F.; Benoiton, N. L. Can. J. Chem. 1987, 65,
619–625.
13. Thaler, A.; Seebach, D.; Cardinaux, F. Helv. Chim. Acta
1991, 74, 617–627.
14. Chen, F. M. F.; Steinauer, R.; Bentoin, N. L. J. Org.
Chem. 1983, 48, 2939–2941.
15. Bodanszky, M.; Tolle, J. Int. J. Peptide Protein Res. 1977,
10, 380–384.
16. Landau, R. N.; McKenzie, P. F.; Forman, A. L.; Dauer,
R. R.; Futran, M.; Epstein, A. D. Process Control and
Quality 1995, 7, 133–142.
17. Reactions were carried out in a Mettler–Toledo RC1
reaction calorimeter equipped with a 1 L glass MP-10
vessel. The nitrogen gas flow was metered with a Brooks
Model 5850i mass flow controller. The composition of
the offgas was determined by gas chromatography using
a Hewlett–Packard 5890 GC equipped with a 5 mL sam-
pling loop, an Agilent GasPro capillary column (60 m
length, 0.25 mm ID) and a thermal conductivity detector.
Offgas volume was determined using a Ritter model
TG05 wet test meter. The wet test meter was placed
downstream of the GC sampling loop, and the cumula-
tive volume was recorded using a datalogging system.
During the experiment, which was always performed with
nitrogen purge gas flowing, the cumulative total gas
volume was measured with the wet test meter. The flow
rate of total offgas was determined by numerical differen-
tiation. The flow rate of the CO2 offgas was determined
by subtracting the N2 purge gas flow rate determined
earlier from the total offgas rate during the experiment.
The cumulative CO2 offgas volume was subsequently
determined by numerical integration of the CO2 flow
rate. In the experiments performed here, the gas evolu-
tion occurred during periods in which liquid was added
to the reactor. To prevent incorrectly counting the gas
displaced by the liquid as evolved chemical offgas, it was
necessary to subtract the volume of liquid added from the
cumulative chemical offgas evolved.
In summary, a case study on the elucidation of mecha-
nism of urethane by-product formation and starting
amino acid liberation during the conventional two-step
isobutyl chloroformate mediated N-acylation using car-
bon dioxide offgas as the probe is described. The
formation of the symmetrical anhydride from the
amino acid in the first step involving the preparation of
the mixed carboxylic–carbonic anhydride intermediate,
as determined by quantifying the evolved carbon diox-
ide, was found to be the main reason for the urethane
formation and starting amino acid liberation in our
case under conventional conditions. Only minor
amounts of these two by-products resulted from the
aminolysis at the undesired carbonyl group in the
mixed carboxylic–carbonic anhydride intermediate in
the second step. Based on this mechanistic understand-
ing, a new protocol for the coupling of 1 with N-benz-
ylmethylamine, involving a reverse addition of 1 and
the base to isobutyl chloroformate, was developed that
decreased the formation of symmetrical anhydride to
<2%. This methodology will be useful to address this
side reaction also in peptide synthesis.
Acknowledgements
We thank Drs. Wen Shieh, Prasad Kapa, and Hong-
Yong Kim for a helpful discussion. We are also grateful
to Professor R. K. Boeckman (University of Rochester)
for helpful suggestions.
References
1. Vaughn, J. R. J. Am. Chem. Soc. 1951, 73, 3547.
2. Anderson, G. W.; Zimmerman, J. E.; Callahan, F. M. J.
Am. Chem. Soc. 1967, 89, 5012–5017.
3. Dudash, J.; Jiang, J.; Mayer, S. C.; Joullie, M. M. Synth.
Commun. 1993, 23, 349–356.

5546
A. Chaudhary et al. / Tetrahedron Letters 44 (2003) 5543–5546
The validity of the offgas measurement was confirmed by
setup was charged with toluene (390.0 mL) and isobutyl
chloroformate (28.86 g, 211.3 mmol). The solution was
cooled to an internal temperature at −12 2°C, and the
first forming a known amount of CO2 by the neutraliza-
tion of sodium bicarbonate with sulfuric acid. (A
manuscript describing the details of this methodology is
in preparation).
above-prepared solution of BOC-L-3-(2-naphthyl)alanine
(1) and N,N-dimethylbenzylamine in toluene was added
over a period of 40 min while maintaining an internal
temperature at −12 3°C (jacket temperature −15 3°C).
The addition funnel was washed with toluene (2×5.2 mL)
and added to the reaction mixture. The slurry was stirred
at −12 2°C for an additional 30 min, and a solution of
N-benzylmethylamine (24.6 g, 203.0 mmol) in toluene
(15.6 mL) was added over a period of 30 min while
maintaining an internal temperature at −12 2°C (jacket
temperature −15 3°C). The addition funnel was washed
with toluene (2×5.2 mL) and added to the reaction mix-
ture. The reaction mixture was stirred at −12 2°C for 30
min and then warmed to an internal temperature at
21 2°C over a period of 35 min (the completion of the
reaction was monitored by HPLC). The reaction mixture
was quenched by addition of 1N H2SO4 (125.0 mL) and
stirred for 10 min. The organic layer was separated and
washed sequentially with water (125.0 mL), 5% aqueous
sodium bicarbonate (125.0 mL), and water (125.0 mL).
The organic layer was concentrated at a jacket tempera-
ture of 65 5°C in vacuo (150 mbar) to collect 340 mL
(ꢀ309.0 g) of solvent and obtain a of solution of BOC-
(S)-3-(2-naphthyl)alanyl-N-benzyl-N-methylamide in tol-
uene (2, 310.0 mL, 272.0 g; containing 68.0 g of 2,
assumed 100% yield) that was used crude in the next step.
18. Calculated based on the mass balance on 1 remaining at
the end of the reaction and CO2 formed in the first step
(a manuscript with detailed methodology is under prepa-
ration).
19. Peptide Synthesis; Bodanszky, M.; Klausner, Y. S.;
Ondetti, M. A., Eds.; John Wiley & Sons: New York,
London, Sydney, Toronto, 1976; p. 140.
20. Shieh, W. C.; Carlson, J. A.; Shore, M. E. Tetrahedron
Lett. 1999, 40, 7167–7170.
21. The Mettler–Toledo Labmax reactor, equipped with a 1
L glass vessel, pitched-blade impeller, RTD sensor, addi-
tion funnel, nitrogen inlet/outlet and distillation setup
was charged with BOC-L-3-(2-naphthyl)alanine (1, 51.2 g,
162.3 mmol) and toluene (130.0 mL). The suspension was
cooled to an internal temperature at 20 2°C and N,N-
dimethylbenzylamine (29.64 g, 219.3 mmol) was added
over a period of 10 min while maintaining the internal
temperature at 20 2°C (jacket temperature 17 3°C). The
addition funnel was washed with toluene (5.2 mL) and
the wash was added to the reaction mixture. This solution
was held for further use.
The Mettler–Toledo Labmax reactor, equipped with a
clean 1 L glass vessel, pitched-blade impeller, RTD sen-
sor, addition funnel, nitrogen inlet/outlet and distillation