Investigation of N-carbamoylamino acid nitrosation by {NO + O2 in the solid-gas phase. Effects of NOx speciation and kinetic evidence for a multiple-stage process

Article abstract of DOI:10.1002/poc.1164

Nitrosation of N-carbamoylamino acids (CAA) by gaseous NO + O₂, an interesting synthetic pathway to amino acid N-carboxyanhydrides (NCA), alternative to the phosgene route, was investigated on N-carbamoyl-valine either in acetonitrile suspension or solventless conditions, and compared to the classical nitrosating system NaNO₂ + CF₃COOH (TFA), the latter being quite less efficient in terms of either rate, stoichiometric demand, or further tractability of the product. The rate and efficiency of the NO + O₂ reaction mainly depends on the O₂/NO ratio. Evaluation of the contribution of various nitrosating species (N ₂O₃, N₂O₄, HNO₂) through stoichiometric balance showed the reaction to be effected mostly by N ₂O₃ for O₂/NO ratios below 0.3, and by N ₂O₄ for O₂/NO ratios above 0.4. The relative contribution of (subsequently formed) HNO₂ always remains minor. Differential scanning calorimetry (DSC) monitoring of the reaction in the solid phase by either HNO₂ (from NaNO₂ + TFA), gaseous N ₂O₄ or gaseous N₂O₃, provides the associated rate constants (ca. 0.1, 2 and 10⁸ s^-1 at 25°C, respectively), showing that N₂O₃ is by far the most reactive of these nitrosating species. From the DSC measurement, the latent heat of fusion of N₂O₃, 2.74 kJ ·mol^-1 at -105 °C is also obtained for the first time. The kinetics was investigated under solventless conditions at 0°C, by either quenching experiments or less tedious, rough calorimetric techniques. Auto-accelerated, parabolic-shaped kinetics was observed in the first half of the reaction course, together with substantial heat release (temperature increase of ca. 20°C within 1-2 min in a 20-mg sample), followed by pseudo-zero-order kinetics after a sudden, important decrease in apparent rate. This kinetic break is possibly due to the transition between the initial solid-gas system and a solid-liquid-gas system resulting from water formation. Overall rate constants increased with parameters such as the specific surface of the solid, the O ₂/NO ratio, or the presence of moisture (or equivalently the hydrophilicity of the involved CAA), however without precise relationship, while the last two parameters may directly correlate to the increasing acidity of the medium. Copyright

Full text of DOI:10.1002/poc.1164

JOURNAL OF PHYSICAL ORGANIC CHEMISTRY
J. Phys. Org. Chem. 2007; 20: 271–284
Published online 29 March 2007 in Wiley InterScience
(www.interscience.wiley.com) DOI: 10.1002/poc.1164
Investigation of N-carbamoylamino acid nitrosation
by {NO R O2} in the solid-gas phase. Effects
of NOx speciation and kinetic evidence
for a multiple-stage process
*
Olivier Lagrille, Jacques Taillades, Laurent Boiteau and Auguste Commeyras
`
´
Dynamique des Systemes Biomoleculaires Complexes, UMR CNRS 5247, Chemistry Department, University of Montpellier 2,
`
Place Eugene Bataillon, F-34095 Montpellier Cedex 5, France
Received 14 September 2006; revised 23 November 2006; accepted 12 December 2006
ABSTRACT: Nitrosation of N-carbamoylamino acids (CAA) by gaseous NO þ O2, an interesting synthetic pathway
to amino acid N-carboxyanhydrides (NCA), alternative to the phosgene route, was investigated on N-carbamoyl-valine
either in acetonitrile suspension or solventless conditions, and compared to the classical nitrosating system
NaNO2 þ CF3COOH (TFA), the latter being quite less efficient in terms of either rate, stoichiometric demand, or
further tractability of the product. The rate and efficiency of the NO þ O2 reaction mainly depends on the O2/NO ratio.
Evaluation of the contribution of various nitrosating species (N2O3, N2O4, HNO2) through stoichiometric balance
showed the reaction to be effected mostly by N2O3 for O2/NO ratios below 0.3, and by N2O4 for O2/NO ratios above
0.4. The relative contribution of (subsequently formed) HNO2 always remains minor. Differential scanning
calorimetry (DSC) monitoring of the reaction in the solid phase by either HNO2 (from NaNO2 þ TFA),
gaseous N2O4 or gaseous N2O3, provides the associated rate constants (ca. 0.1, 2 and 108 sꢀ1 at 258C, respectively),
showing that N2O3 is by far the most reactive of these nitrosating species. From the DSC measurement, the latent heat
of fusion of N2O3, 2.74 kJ ꢁ molꢀ1 at ꢀ105 8C is also obtained for the first time. The kinetics was investigated under
solventless conditions at 08C, by either quenching experiments or less tedious, rough calorimetric techniques.
Auto-accelerated, parabolic-shaped kinetics was observed in the first half of the reaction course, together with
substantial heat release (temperature increase of ca. 208C within 1–2 min in a 20-mg sample), followed by
pseudo-zero-order kinetics after a sudden, important decrease in apparent rate. This kinetic break is possibly due
to the transition between the initial solid-gas system and a solid-liquid-gas system resulting from water formation.
Overall rate constants increased with parameters such as the specific surface of the solid, the O2/NO ratio, or the
presence of moisture (or equivalently the hydrophilicity of the involved CAA), however without precise relationship,
while the last two parameters may directly correlate to the increasing acidity of the medium. Copyright # 2007 John
Wiley & Sons, Ltd.
Supplementary electronic material for this paper is available in Wiley Interscience at http://www.interscience.
wiley.com/jpages/0894-3230/suppmat/.
KEYWORDS: amino acid N-carboxyanhydrides; nitrosation; N-carbamoylamino acids; kinetics; solid-phase reaction
INTRODUCTION
present major drawbacks such as the toxicity of involved
reagents (e.g. phosgene), or the difficulty of separating
side-products, in spite of important efforts made to
improve them by the use of either more efficient leaving
groups, or reagents less toxic than phosgene.4,5
Our previous work dealing with the relevance of
N-carbamoylamino acids (CAA) 2 as possible prebiotic
precursors of peptides,6 led us to examine CAA activation
pathways and to discover that their nitrosation by a
NO þ O2 gaseous mixture resulted into smooth, fast NCA
formation (Scheme 2) with nitrogen and water as the only
side products.7 NMR monitoring (in DMSO-d6) showed
The synthesis of amino acid N-carboxyanhydrides (NCA,
also known as Leuchs anhydrides) 1 is subject to
important research efforts, due to their interest as key
monomers in the synthesis of polypeptide materials.
Classical synthetic routes to NCA usually involving
Leuchs1 or Fuchs–Farthing2,3 methods (Scheme 1),
*Correspondence to: Dr L. Boiteau, DSBC (UMR5247)-CC.017,
University Montpellier-2, Place Eugene Bataillon, F-34095 Montpel-
lier Cedex 5, France.
E-mail: laurent.boiteau@univ-montp2.fr
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 271–284

272
O. LAGRILLE ET AL.
O
R
R
O
R
PBr
3
Br
R
EtO
H2N
N
H
CO2H
EtO
O
N
H
(Leuchs)
O
O
HN
O
O
R
Cl2CO
NCA
CO2H
Cl
N
H
CO2H
(Fuchs-
Farthing)
Scheme 1
Scheme 2
the reaction to proceed through the fast formation of an
intermediate (assumed to be an external N-nitrosourea 3),
followed by its slower decomposition into nitrogen, water
and NCA, most likely through a transient isocyanate
intermediate 4.7 The early intermediacy of an amidium-
type species 5 resulting from the O-nitrosation of urea, is
very probable.8,9 However a further addition-elimination
mechanism on 5 involving the vicinal carboxy group is
likely disfavoured according to Baldwin’s rules10,11 since
it would involve a 5-endo-trig cyclisation, so that its
rearrangement into 3 is more plausible. Such additio-
n-elimination pathway (involving either 3 or 5) may occur
as an alternative to isocyanate formation when the latter is
impossible, for example from N,N-dialkyl ureas.12
Nevertheless such mechanistic issues are not the focus
of the present paper and will be discussed in a future
publication.
One of the major features of CAA nitrosation is that it
occurs in the heterogeneous phase since CAAs are
insoluble in most organic solvents, including acetonitrile
used in previous investigations. This led us to investigate
the reaction without solvent, meanwhile the prebiotic
relevance of such a pathway was also discussed.13 This
new synthetic route to NCA opened new application
perspectives as a potent eco-friendly challenger to
phosgene routes, thus pointing to the need for further
investigations.
Monitoring the NO R O2-mediated nitrosation of
N-carbamoyl valine, C-Val, crystals by optical micro-
scopy showed the reaction not to proceed in a truly
solid phase: the solid surface readily wets, while gas
bubbles evolve. While the reaction actually starts as a
solid-gas process, the water released together with NCA
formation partly dissolves the solid material and NOx
species, thus allowing the reaction to progress also in the
liquid phase. At the end of the reaction however, the crude
product is often recovered as a solid and not as an oil:
nitrogen evolution and exothermicity of the reaction
probably contribute to evaporate the water as and when it
is formed.
Previous investigation of solid-gas nitrosation of C-Val
in a fluid-bed reactor by various NO/O2 stoichiometries
with monitoring the total NOx consumption showed
zero-order kinetics,13 however without allowing to
investigate other reaction parameters. Furthermore,
those experiments were poorly reproducible especially
regarding the final CAA-to-NCA conversion. Further
investigation was then necessary to better understand the
reaction in view of future synthetic applications.
In this paper, we present deeper insights in the
efficiency and kinetics of solventless CAA nitrosation by
NO R O2, investigating valine as the model amino acid,
with comparison to the reaction in acetonitrile suspen-
sion, and to the more classical nitrosating system
{nitrite R trifluoroacetic acid (TFA)}, the latter being
considered as the reference for further evaluation of
{NO R O2} efficiency. NOx speciation (O2/NO ratio) is
chiefly concerned since several species coexist in
Reaction in the solid phase. Although probably
proceeding according to the same pathway, the reaction in
the solid phase appears somewhat more complex.
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 271–284
DOI: 10.1002/poc

N-CARBAMOYLAMINO ACID NITROSATION BY NO þ O2
273
{NO R O2} mixtures, those of major interest here
being N2O3 and N2O4 (both of them are formally NOR
donors14). Since water is generated during NCA
formation, we also consider the nitrosating species
HNO2 separately generated from the system {nitri-
te R TFA}.
the reaction is significantly retarded by any excess of
nitrite over TFA, the catalytic effect of TFA is obvious,
in particular when in excess over nitrite, which suggests
the involvement of the acidic form. The origin of such
catalysis can be either general acid catalysis with
formation of the faster nitrosating species NOþ
(Eqn 1), or the formation of nitrosyl trifluoroacetate
CF3CO2NO as an intermediate (Eqn 2). This species has
been reported as an efficient nitrosyl-transferring
agent,16 as well as other nitrosyl carboxylates, for
example acetate.17–19
RESULTS AND DISCUSSION
N-carbamoyl valine C-Val was prepared according to our
previously published procedures,15 by reaction of the free
amino acid with aqueous potassium cyanate, under pH
regulation to 7–8, then easily purified by recrystallisation
from water. NOx nitrosating mixtures were obtained in all
cases by mixing appropriate amounts of NO and O2 at
atmospheric pressure, usually in stoichiometric excess
compared to CAA so that the pressure and stoichiometric
composition of the equilibrated NOx mixture should not
significantly vary along the reaction progress.
CF3CO2H
CF3CO2H
–
+
NO2
HNO2
H2NO2
NO+ + H2O
(1)
CF3CO2H þ HNO2 CF CO NO þ H O
(2)
!
3
2
2
It deserves to note that, although efficient and easy to
carry out, the nitrosation of CAA by the {TFA þ nitrite}
system offers poor perspectives in a NCA preparative
scope compared to the {NO þ O2} method, since the
presence of salts and water in the reaction mixtures makes
the further purification of NCA very difficult.
Nitrosation of C-Val by nitrite R TFA in
acetonitrile (the reference system)
We investigated HNO2-mediated nitrosation of C-Val
by using the conventional system {sodium nitri-
te þ trifluoroacetic acid} in acetonitrile (Table 1). The
use of moderate excess of nitrite and acid gives good
conversion (evaluated by integration of the Ha peaks
on NMR spectra) within 2 h. To estimate the relative rate
of this system compared to that of the NO þ O2 system,
the conversion was measured after 5 min, evaluating the
remaining C-Val by HPLC analysis after quenching
the reaction by an excess of sodium hydroxide. While
Nitrosation of C-Val by NO R O2
Gaseous mixtures of NO and O2 react at room
temperature, readily forming rather complex mixtures
of O2, NO, NO2, N2O3, N2O4,14 the two latter species
being responsible for the nitrosation reaction. Equi-
librium stoichiometry of such {NO þ O2} mixtures is
mainly governed by Eqns 3–5 (equilibrium constants at
258C were evaluated from standard formation enthalpies
Table 1. Nitrosation of C-Val (100 mg) by NaNO2 þ TFA (various amounts) in acetonitrile (50 ml) at room temperature: effect of
reagent stoichiometry (NCA yield not measured when quenched after 5 min reaction)
Initial conditions
Working conditions
Time/
Conv.
%
NCA
yield %
Run min [NOꢀ2 ]/[CAA] [TFA]/[NOꢀ2 ] [TFA]/[CAA] [NO2ꢀ]/[CAA] [HNO2]/[CAA] [NOþ]/[CAA]
1
2
3
4
5
6
7
8
9
10
11
12
13
30
120
120
5
5
5
5
5
5
5
1.5
1.5
1
12
6
3
3
1
2
4
1
1.5
0.75
1
1.33
2
0.25
0.5
1
2
1
0.5
0.25
2
1.5
2
2
3
3
3
6
1
1
1
2
1.5
0.75
0
0
0
9
3
0
0
0
1
3
0
0
0
1.5
1
0
3
3
3
0
1
1
1
1
1.5
0.75
0
0.5
1
0
0
0
3
0
0
0
0
0
0
95
71
87
15
44
58
100
11
16
5
38.5
31
71
63
79
n.m.
n.m.
n.m.
n.m.
n.m.
n.m.
n.m.
n.m.
n.m.
n.m.
5
5
5
1
1
12.5
Working conditions: equilibrated TFA-nitrite mixture according to Eqns 1–2 considered as completely shifted to the right.
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 271–284
DOI: 10.1002/poc

274
O. LAGRILLE ET AL.
DHf0 and entropies DSf0 of the involved species20–22):
absence of oxygen. With a starting stoichiometry NO/
CAA ¼ 1, the conversion increases upon increasing the O2/
NO ratio up to ca. 60% at O2/NO ¼ 0.2–0.3, roughly
corresponding to the N2O3 overall stoichiometry, then
decreases to ca. 40%. When higher NO/CAA stoichi-
ometries are used, the same trend is observed (in
acetonitrile) with higher conversion, up to 100%; the
minimum O2/NO/CAA stoichiometry required for a
quantitative conversion is 0.6/2/1 and again, the reaction
efficiency decreases with increasing O2/NO ratios. It is
intriguing to observe a so small difference between
acetonitrile and solventless conditions: Bravo et al.23
observed significant solvent effects in various acetonitrile/
water mixtures, with mechanistic change with the water/
acetonitrile ratio and higher rates in pure solvents than in
mixtures. However, our solventless conditions may not be
properly described by a dilute aqueous solution. Retarding
effects have also been observed in acetonitrile at a lower
temperature (ꢀ408C) by Darbeau et al.,9 who suspected
acetonitrile to behave as a NOþ transfer agent in spite of
the supposed inertness of the nitrile group.
NO2 NO þ 1=2O
!
2
(3)
ðK1 ¼ 2:77 ꢂ 10ꢀ12 atm1=2 ; 25ꢃCÞ
N2O3 NO þ NO ðK2 ¼ 2:12 atm; 25ꢃCÞ
(4)
(5)
!
2
N2O4 2NO ðK3 ¼ 0:148 atm; 25ꢃCÞ
!
2
Due to the very low K1 (Eqn 3 completely shifted to the
left), if O2/NO > 0.5 the species NO and N2O3 are almost
absent from equilibrated mixtures, with pressures of
both NO2 and N2O4 independent of the O2/NO ratio.
Conversely, if O2/NO < 0.5, the amount of N2O4
increases almost linearly with the O2/NO ratio,
while N2O3 remains a minor component (at most a few
mol% around O2/NO ¼ 0.25). This occurs over a wide
range of NO initial pressures.
The nitrosation of C-Val by {NO þ O2} was examined
in both acetonitrile suspension and solid-gas phase
(without solvent). The evolution of the (exothermic)
reaction could be easily followed by the rapid discolour-
isation of the gaseous phase. In all cases, the reaction was
complete within ca. 10–30 min. Meanwhile a lagoon-blue
colour (characteristic of condensed N2O3 and N2O4)
transiently appeared in the condensed phase, together
with heat release.
Involved nitrosating species and their relative
efficiency (NOx speciation)
CAA nitrosation involving either N2O3 or N2O4 (Eqns
6–7), as well as N2O3/N2O4 hydrolysis (by water
subsequently released together with NCA formation,
Scheme 2) releases nitrous and/or nitric acid, respectively.
The resulting acidification of the medium (certainly
very strong under ‘dry phase’ conditions where reactive
species are highly concentrated in the condensed
phase) allows HNO2-mediated nitrosation to take place
In the organic suspension, the insoluble, solid CAA
progressively disappeared and was converted into soluble
NCA. In the solid phase, water droplets condensed onto
the walls of the flask. Previous optical microscopy
monitoring showed that a liquid phase forms onto the
solid material, together with gas bubble evolution.
However, in most cases the final product was recovered
as a solid.
24
through the protonated, very reactive species H2NO2þ
The conversion appears mainly dependent on the CAA/
NO/O2 stoichiometry (Figure 1). No reaction occurs in the
(as with the {nitrite þ TFA} system), a pathway much
faster than the slow dehydration of HNO2 into N2O3.25
CAA þ N2O3 ! 3 þ Hþ þ NO2ꢀ
CAA þ N2O4 ! 3 þ Hþ þ NO3ꢀ
CAA þ H2NOþ2 ! 3 þ H3Oþ
(6)
(7)
(8)
Measurement of nitrite and nitrate anions in crude
reaction mixtures (by capillary electrophoresis after
alkaline quenching) therefore leads to the evaluation of
the relative contributions of N2O3, N2O4 and HNO2 to the
overall nitrosation process – N2O4 being quantitatively
related to nitrate (Eqn 7), N2O3 to nitrite (Eqn 6) after
correction of the overall imbalance corresponding to
HNO2/H2NOþ2 consumption (Eqn 8, the latter pathway
releasing no anion). Direct hydrolysis of N2O3/N2O4
turned out to be negligible (not exceeding a few per cent
in blank experiments). We also assumed the dismutation
of HNO2 into NO, H2O and HNO3 (usually occurring at
high temperatures) to be sufficiently slow under our
conditions to be negligible as well.14 What could not be
Figure 1. Nitrosation of C-Val by NO þ O2 at room
temperature and atmospheric pressure: conversion after
30 min reaction as a function of the O2/NO stoichiometric
ratio. Reactions carried out in acetonitrile with 625 mmol
C-Val (open symbols) or in the solid-gas phase with
187.5 mmol (filled symbols) with NO/C-Val stoichiometric
ratios of 1 (diamonds), 2 (squares) or 3 (circles). Experimental
points are connected visually for an easy reading
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 271–284
DOI: 10.1002/poc

N-CARBAMOYLAMINO ACID NITROSATION BY NO þ O2
275
would be required to ensure HNOþ2 -mediated nitrosation
from HNO2, while catalytic amounts were expected to be
sufficient since protons are supposedly regenerated after
the nitrosation step (Eqn 8). Possible reasons can be
kinetic, with for instance the slow decay of subsequent
cationic intermediates.
Thermoanalytical kinetic investigation
Thermochemical and kinetic parameters of the nitrosation
reaction involving either N2O4, N2O3 or HNO2/NOþ
without solvent, were investigated through differential
scanning calorimetry (DSC) monitoring, a technique
allowing to measure both the reaction heat DHr, as well as
its apparent rate constant k and activation energy EA. DH
is directly given by DSC peak integration. Assuming that
the reaction is pseudo first-order ([ꢀdj/dt ¼ k(T) ꢁ j] with
j, the reaction extent) and follows the Arrhenius law
[k(T) ¼ A ꢁ exp(–EA/RT)], parameters EA and A were
determined by fitting of simulated thermograms (easily
computed from given DH, Ea and A at various
temperature scanning rates) to experimental ones.
Figure 2. Contribution of the different nitrosating species
in dry nitrosation of solid C-Val by {NO þ O2} at room
temperature, compared to overall conversion, as a function
of O2/NO ratio: N2O3 (squares), HNO2 (circles) and N2O4
(triangles). Hashed area: unreliable results because of uncer-
tainty in the (measured) low conversion range (<10%)
neglected is the formation of additional nitrite anions
from dismutation of unreacted NO during alkaline
quenching, a phenomenon occurring for O2/NO ratios
below 0.5 and inducing up to 25% bias in nitrite
measurement. Quantification of the phenomenon on
blank experiments allowed correcting it satisfactorily.
Results of the relative contribution of NOx species to
overall conversion versus O2/NO ratios are shown in
Figure 2. Relative contributions of N2O3 and N2O4 do not
follow their equilibrium ratio in the gas phase, meaning
that equilibria Eqns 1–3 are actually shifted during the
nitrosation reaction. N2O4 is the major contributor
with O2/NO ratios larger than 0.4, while N2O3 is the
major contributor when O2/NO ratios are smaller than
0.3. The apparent trend inversion observed at the
lowest O2/NO ratios (<0.15) is questionable because
of the inaccuracy of measured low conversion values and
would require further investigation. In all cases HNO2
(H2NOþ2 ) contributes to a smaller extent to the overall
process, its contribution appearing roughly as the
intersection of the N2O3 and N2O4 curves. This suggests
that a stoichiometric amount of either N2O4 or HNO3
Nitrosation by HNO2/NOR (the reference system)
was investigated through the {NaNO2 R TFA} system,
with TFA/nitrite stoichiometric ratios of 1/1, 2/1 and 10/3
(the nitrite/C-Val stoichiometric ratio was 3/1 in all
cases). DSC curves exhibit first a small endotherm (TFA
melting), then two exotherms corresponding to TFA-
nitrite neutralisation and to C-Val nitrosation, respect-
ively. An increase in TFA/nitrite ratios resulted into
sharper nitrosation peaks, meaning a significant increase
in the reaction rate.
Measured thermochemical and kinetic data are shown
in Table 2. The obtained reaction heat (DH) for a TFA/
nitrite ratio of 1/1 (S136 kJ molS1), far below those
measured for TFA/nitrite ratios of 2/1 and 10/3
(S230 kJ molS1), suggests the reaction to be incomplete
in the first case (ca. 60% extent), as observed in
acetonitrile suspension, thus questionning the reliability
of associated kinetic parameters. The substantial increase
Table 2. DSC data for nitrosation of C-Val (12.5 mmol) by sodium nitrite (3 eq.) þ TFA (3, 6, 10 eq.): reaction enthalpy DH
(average of several measurements), DSC peak temperature Tm; calculated 1st-order kinetic parameters (activation energy EA,
pre-exponential factor A, extrapolated rate constant k at 258C)
[TFA]/[NOꢀ2 ]
1 (3/3)
2 (6/3)
3.33 (10/3)
DSC rate/8C minꢀ1
5
8
5
8
5
8
ꢀDH (kJ ꢁ molꢀ1
)
134.1
138.9
228.6
232.4
228.5
231.8
Average
ꢀ136.5
ꢀ230.5
ꢀ230.2
Tm (8C)
ꢀ1.7
ꢀ2.0
1.9
107.4
3.1
97.7
0.0
193.1
3.0
185.6
EA (kJ ꢁ molꢀ1
)
104.5
96.1
Average
)
Average
100
103
189
A (sꢀ1
1.5 ꢂ 1018
5.1 ꢂ 1017
2.8 ꢂ 1018
1.9 ꢂ 1016
1.5 ꢂ 1035
1.1 ꢂ 1034
18
1.0(ꢄ0.5) ꢂ 10
2.8(ꢄ0.02) ꢂ 1018
3.6(ꢄ0.4) ꢂ 10ꢀ1
0.8(ꢄ0.7) ꢂ 1035
k258C (sꢀ1
)
7.0(ꢄ0.4) ꢂ 10
2.4(ꢄ0.6) ꢂ 101
ꢀ1
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 271–284
DOI: 10.1002/poc

276
O. LAGRILLE ET AL.
in the rate constant with TFA/nitrite ratio strongly
suggests a catalytic effect of TFA, either due to general
acid catalysis (responsible for the faster formation
of NOR intermediates), or due to the involvement of
trifluoroacetyl nitrite as an intermediate.26 From our
calculated kinetic parameters, the rate constant extrapo-
N2O4 stoichiometry. . .). From our calculated kinetic
parameters, the rate constants extrapolated at 25-C lie
around kN O ꢅ 108 sS1 for N2O3, and kN O ꢅ 1.7 sS1
2
3
2 4
for N2O4. However, such values do not take in account the
quite probably different vapour pressures of N2O3
and N2O4, under the respective experimental conditions
(different temperatures).
lated to 25-C is around 0.1 sS1 (kHNO ).
2
Nitrosation by (gaseous) N2O3 or N2O4 was
investigated on C-Val in the solid-gas phase. In this
purpose, a small PTFE (Teflon) cup containing solid
C-Val was placed inside the DSC crucible containing the
frozen NOx, to avoid premature contact between the
reacting species. All reactions were carried out with an
excess of nitrosating reagent. N2O3 was prepared
according to the well-known procedure involving nitric
acid reduction by As2O3 just prior to DSC experiments.27
In blank experiments without C-Val, melting
endotherms occurring at S105-C (N2O3) or S13.2-C
(N2O4), allowed to determine the latent melting heat of
these species, 2.74 and 16.77 kJ ꢁ molS1, respectively.
While our measurement for N2O4 is in good agreement
with previously published data (14.65 kJ ꢁ molS1 at N2O4
triple point28), this is the first report of an experimental
value of N2O3 melting heat. With both N2O3 and N2O4 the
nitrosation exotherm observed far below the boiling point
of the NOx species (ca. R3-C and R21-C for N2O3
and N2O4, respectively), confirms that the nitrosation was
effected by gaseous NOx at saturating vapour pressure.
Thermogramms involving N2O4 were processed after
subtraction of N2O4 phase-transition endotherms (DSC
curves of blanks) otherwise interfering with the nitrosa-
tion reaction. Conversely to the {TFA R nitrite} case, the
use of various stoichiometric excesses of N2O3 and N2O4,
resulted in only small changes in thermochemical and
kinetic results (Table 3). Measurements of reaction
enthalpies for C-Val nitrosation involving both N2O3
and N2O4, (Table 3) appear more accurate than those
involving HNO2, showing consistent DH values (S336
and S356 kJ ꢁ molS1 for N2O3 and N2O4, respectively)
under various conditions (temperature scan rate, N2O3/
Discussion. Measured kinetic parameters and extrapo-
lated rate constants at 25-C are quite consistent with
literature reports on the reactivity of such nitrosating
species in organic media.14 However several questions
arise from these experiments. Firstly, there is an important
difference in the measured DH values which depend on
the nitrosating agent, while similar values would have
been expected. Much lower DH values with N2O3
(S336 kJ ꢁ molS1) and N2O4 (S356 kJ ꢁ molS1) than with
NaNO2 R TFA (S230 kJ ꢁ molS1) suggest the presence of
additional exothermic processes in the former cases, for
instance the hydrolysis of either N2O3 or N2O4 with water
released by Val-NCA formation.
Also, the high values of activation energies (EA ꢅ
100–200 kJ ꢁ molS1, Tables 2–3) for these fast processes,
primarily suggest these reactions to be dominated by
entropic effects. Actually these reactions may be
more complex than first-order, what might also explain
the spectacular EA increase in the nitrosation
by {NaNO2 R TFA} by increasing TFA/nitrite ratio
(Table 2). Such EA increase is probably related to
important qualitative (mechanistic) changes. In all cases
this means that the 1st-order kinetic model used probably
provides quite inaccurate kinetic parameters EA and A.
Parallelly, such high EA values may also be due to
insufficient heat exchange between the reacting material
and its environment, considering the high exothermicity
of the process. In this respect, DSC monitoring at rate of
8-C minS1 which should minimise such effects (com-
pared to those at 5-C minS1), indeed results in smaller EA
values, cf. supporting information.
Further kinetic analysis by more sophisticated models is
also complicated by the coexistence of several isomers of
Table 3. DSC data of reaction peak for nitrosation of C-Val (12.5 mmol) by excess N2O3 or N2O4 (stoichiometries in table):
reaction enthalpy DH, DSC peak temperature Tm; calculated 1st-order kinetic parameters (activation energy EA, pre-exponential
factor A, extrapolated rate constant k at 258C)
N2O3
N2O4
1
Reagent DSC rate (8C minꢀ )
eq./C-Val
5
8
5
8
7.9
10.5
8.9
5.9
6.9
5.5
29.8
5.6
4.6
23.9
6.4
–DH (kJ.molꢀ1
Tm (8C)
)
345.7 343.9 328.2 326.9 346.8 369.2 363.2 352.3 319.2ꢆ 342.2 365.2
ꢀ28.9
ꢀ28.7
ꢀ8.0
ꢀ4.2
EA (ꢀk1J.molꢀ1
)
249.3 ꢄ 0.2
102 ꢄ 4
7.8 (ꢄ0.02) ꢂ 1018
1.7 (ꢄ0.5)
A (s
)
5.9 (ꢄ1.7) ꢂ 1051
k (sꢀ1, 258C)
1.0 (ꢄ0.2) ꢂ 108
ꢆ These values were excluded in the calculation of averaged values.
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 271–284
DOI: 10.1002/poc

N-CARBAMOYLAMINO ACID NITROSATION BY NO þ O2
277
O
O
N
O
O
O
O
N
O
N
N
N
O
O
O
N
N
O
O
O
O
N
6
8
7
9
Scheme 3
29
both species N2O4 and N2O3 (Scheme 3), for which
insufficient thermodynamic data are available, for example
about interconversion equilibria. For instance, while the
symmetrical isomer 6 is known as the most stable form of
N2O4, its unstable, asymmetric isomer 7 (so-called nitrosyl
nitrate) is often believed as the one effecting nitrosation.14,29
Also, the asymmetric structure 8 is described as the most
stable form of N2O3 in either gaseous or condensed
phase, while nitrosation reactions may rather involve the
unstable, symmetric isomer 9 preferentially formed by
addition of NO þ NO2\cdot in condensed media,29 but
until now characterised only in solid argon matrix.30–33
The major influence of the O2/NO ratio on the
efficiency of solid C-Val nitrosation by the gaseous
mixture NO þ O2, may therefore be interpreted in terms
of changes in the availability of the involved nitrosating
species, namely N2O3, N2O4 and HNO2, the former being
the most reactive.
excess NOx, so that the pressure and stoichiometric
composition of the equilibrated NOx mixture should not
significantly vary along the reaction.
We first investigated the evolution of the reaction
extent. Solid C-Val (60–80 mesh powder) was reacted in a
flat-bottom flask at 08C under inert atmosphere (N2).
Gaseous NO was injected, then (at t ¼ 0) O2, so that O2/
NO ¼ 1/4. At time t the reaction was quenched by a fast
nitrogen flush, and the product distribution was deter-
1
mined by H NMR in DMSO-d6. Together with the
expected Val-NCA, the reaction produced significant
amounts of free valine, resulting from the hydrolysis of
the NCA by water released during the reaction. The
kinetics was monitored by repeating the reaction with
quenching at various times. Results reported in Figure 3
show the process to be somewhat complex: we first
observe an 1–2-min latency time (A), followed by a
strong acceleration (B) for a few minutes, then a slower,
linear evolution (C), until ca. 95% conversion is reached.
Complete conversion was achieved within 20 min, where
ca. 15% of NCA was hydrolysed into free valine.
Noteworthy that the observed latency-acceleration
sequence closely resembles a parabolic or exponential
growth (cf. discussion section infra: Latency and
acceleration steps).
Kinetic investigations of the solid-gas phase
nitrosation
To circumvent the rapidity of the reaction at room
temperature (usually less than 5–10 min), kinetic inves-
tigations were carried out at 08C in the presence of
Figure 3. Nitrosation of C-Val in the dry phase at 08C by 5 eq. NO þ 1.25 eq. O2 (O2/NO ¼ 0.25). Time evolution of conversion
(squares) and products: Val-NCA (triangles), Val (diamonds) (measured by 1H NMR); experimental points are connected visually
for an easy reading. Curves: time evolution of temperature: of the gaseous phase (dash-dot line), of a blank experiment without
C-Val (dotted line), of the solid substrate (continuous line) deposited onto the thermometer probe. Time domains A, B, C refer to
latency, acceleration and zero-order kinetic steps, respectively (see text)
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 271–284
DOI: 10.1002/poc

278
O. LAGRILLE ET AL.
100
This important increase in reaction rate, as well as the
exothermicity of the overall process (occurring in a
heterogeneous system), led to assume that the tempera-
ture of the condensed phase may be neither constant nor
homogeneous, and to carry out calorimetric investi-
gations. The use of microcalorimetric techniques was
unsuccessful however: in the available set-up it was
impossible to inject sufficient amounts of NOx reagent
(diluted in an N2 stream) for any reaction to be
observable. Alternately, we carried out rougher calori-
metric investigations in the same glassware setup as
above, by the use of the metallic probe of a conventional
electronic thermometer (results on Figure 3). No
significant temperature changes were recorded in the
gas phase, except a small exotherm during the first 30 s,
corresponding to NO autoxidation with O2 since it was
also present in a blank experiment without C-Val (cf.
discussion infra). To monitor the temperature changes in
the solid phase, a small amount of C-Val was
recrystallised directly onto the thermometer probe then
placed inside the set-up instead of being spread in the
flask. In the typical temperature evolution profile (also
shown on Figure 3), we observe first a small exotherm
during the first 30 s (also observed when monitoring the
temperature of the gas phase or in a blank experiment
without C-Val), then a second, intense exotherm
coincident with the acceleration step (B), followed by
a pseudo-exponential decay coincident with the zer-
o-order step (C).
80
60
40
20
0
t / min
0
5
10
15
20
Figure 4. Effect of average C-Val granulometry hGi (circles:
60 mm, triangles: 211 mm, squares: 358 mm) on nitrosation
of solid C-Val (0.125 mmol) by 5 eq. of NO and 1.25 eq. of O2
(O2/NO ¼ 0.25). Curves (dotted lines: hGi ¼ 358 mm): expo-
nential growth (Eqn 9) and zero-order (with SUC model: Eqn
12) kinetic fits
accelerated process (Figure 4). This effect of solid specific
surface was not investigated by calorimetric experiments
where this parameter was not easily tunable.
Effect of NO/O2 stoichiometric ratio. Symmetri-
cally, the effect of NOx mixture stoichiometry was also
investigated on the 60–80-mesh (hGi ¼ 211 mm) fraction.
Kinetic results are shown in Figure 5. Again, we observed
similar kinetic profiles, exhibiting the latency/acceleration/
zero-order sequence, with a streching/shrinking along the
time axis depending on the O2/NO ratio: the highest O2/
NO, the fastest the kinetics is. Temperature-monitoring
experiments provided similar results (Figure 6): an
Monitoring of either or both reaction extent or
temperature was used to investigate various reaction
parameters. Although our calorimetric technique did not
allow to control the specific surface of the solid material,
it appeared useful for investigating the reaction kinetics,
in a more expedient way than quench/NMR techniques,
thus allowing several complementary parameters to be
investigated.
100
Effect of the specific surface of solid C-Val. The
reaction was carried out on solid C-Val of various gran-
ulometries, obtained from sieving a single batch of
recrystallised, then ground C-Val. Three fractions were
obtained: 40–50 mesh (hGi ¼ 358 mm, Ssp ¼ 136 cm2 ꢁ gS1),
80
60
60–80 mesh (major: hGi ¼ 211 mm, Ssp ¼ 231 cm2 ꢁ gS1
)
and 180–400mesh (hGi ¼ 60mm, Ssp ¼ 810 cm2 ꢁ gS1).
Specific surfaces Ssp are evaluated after solid C-Val density
and average grain size. Using these three C-Val
granulometries with the same O2/NO stoichiometric ratio
of 0.25, we observed very similar kinetic curves, all
exhibiting the latency/acceleration/zero-order sequence.
The curves appear almost superimposable after an overall
streching/shrinking along the time axis. While the
kinetic curves observed with the medium granulometry
(hGi ¼ 211 mm), or with the thickest granulometry
(hGi ¼ 358 mm) are almost identical as that obtained on
the whole batch of ground C-Val, using the finest
granulometry (hGi ¼ 60 mm) results into a significantly
40
20
t / min
0
0
5
10
15
20
Figure 5. Effect of stoichiometric O2/NO ratio (squares:
0.15, triangles: 0.25, circles: 0.40) on the kinetics of nitrosa-
tion of solid C-Val (0.125 mmol with average granulometry
211 mm) by 5 eq. of NO and either 0.75 (squares), 1.25
(triangles) or 2 eq. (circles) of O2. Curves: exponential growth
(Eqn 9) and zero-order (with SUC model: Eqn 12) kinetic fits
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 271–284
DOI: 10.1002/poc

N-CARBAMOYLAMINO ACID NITROSATION BY NO þ O2
279
moisture (e.g. insufficient precautions taken by rainy
weather). Conversely, the addition of a droplet (1–2 ml) of
water into the set-up prior to O2 injection results in an
earlier, sharper exotherm. The same effect was observed
(same changes in thermogramms) when using a
premixed NOx reagent, however with poorer reproduci-
bility except under strictly dry conditions (e.g. NO—O2
transfer under nitrogen counterflow). Such poor reprodu-
cibility is probably due to the high hygroscopicity of
the NOx mixture.
Alternately, a similar moisture effect was observed on
thermogramms (both latency delay and main exotherm
shape) when nitrosating various CAA such as C-Ala,
C-Phe, C-Tyr by 5 eq. NO R 0.75 eq. O2 under dry
conditions: the observed relative reactivity of these CAA
roughly followed the hydrophilicity order of their
side-chains: C-Tyr ꢅ C-Ala > C-Val > C-Phe. While the
final conversion was complete after 1 h for C-Val and
C-Ala, it was only 65% for C-Phe and ca. 51% for C-Tyr.
Upon nitrosation of C-Tyr, a yellow-orange mixture was
formed, possibly due to further nitrosation of the aromatic
ring (prone to electrophilic substitutions because of the
phenol group). Increasing the oxygen/NO ratio up to
0.40 in this case did not result in significant increase in
conversion. Most likely, hydrophilicity (resp. hydropho-
bicity) of the side-chain of the amino acid contributes to
retain (resp. repell) water evolved from NCA formation,
thus having the same influence on the reaction kinetics as
above.
Figure 6. Effect of NOx stoichiometry on nitrosation of solid
C-Val at 08C by 5 eq. NO þ various amounts of O2. Tempera-
ture of the solid (deposited onto the thermometer probe) as
a function of time, under various O2/NO ratios: 0.15 (dotted
line), 0.25 (dash-dot line), 0.40 (solid line)
increase in the oxygen stoichiometry decreased the latency
delay (from 1.5 min with O2/NO ¼ 0.15 to 40s with O2/
NO¼ 0.4), meanwhile the main exotherm appeared
sharper. This is consistent with different nitrosating ability
along NOx speciation as observed above, but it may also be
correlated to faster/slower acidification of the medium (see
Discussion below).
Effect of the NO/O2 injection sequence. We
investigated the possible coupling of NO autoxidation
with O2 (NO) with the nitrosation reaction, by varying
the O2/NO injection protocol. When NO and O2 were
injected sequentially over C-Val, we observed almost
no difference (within experimental error) whether
NO or O2 was injected the first. When NO and O2 were
mixed prior to contact with C-Val, the first, small
exotherm disappeared – thus confirming it is due to NO
autoxidation – and the main exotherm (nitrosation) was
both delayed (3.8 min vs. 2.3–2.5 min) and lowered. In all
cases however, the conversion was complete after 1 h
reaction. The boosting of the nitrosation step by the
NO R O2 autoxidation step may be due to either the
exothermicity of the latter, or to the direct contact of
C-Val with reactive NOx intermediates (e.g. N2O3),
possibly enhanced by the fact that NO R O2 autoxidation
may occur between adsorbed species.
Discussion: kinetic analysis
The behaviour of the system appears more complex here
(Figures 4 and 5) than previously observed kinetics under
fluid-bed conditions.13 In such heterogeneous environ-
ment, the only convenient kinetic variable is the reaction
extent j of the overall nitrosation reaction.
Latency and acceleration steps. In all experiments,
the latency time A is roughly proportional to the
acceleration time B, both steps being then possibly parts
of a parabolic-shaped kinetic curve. Indeed steps (A–B)
together can be accurately fitted with a function of type:
Nevertheless in further experiments we adopted the
NO-then-O2 sequence (always used unless otherwise
specified), for all NO R O2 reactions with CAA, which
provided both the easiest procedure, the most efficient
reactivity, and the best reproducibility (e.g. against
moisture effects, cf. infra).
j ¼ ðk0=k1Þ ꢂ ðexpðk1 ꢁ tÞ ꢀ 1Þ
which is a solution of the differential equation:
(9)
dj=dt ¼ k0 þ k1 ꢁ j
(10)
with j being the reaction extent and k0 the apparent rate
constant at the beginning of the reaction (08C). Mean-
while there is no evidence of any true latency delay before
the parabolic growth (kinetic fits are not improved by
replacing t by t–t0 in Eqn 9). Both rate constants k0 and k1
globally increase with either the O2/NO ratio or the
specific surface Ssp of solid C-Val, however without any
precise linear relationship.
Effects of moisture and amino acid hydrophobi-
city. The presence of moisture from the beginning of the
reaction also accelerates the reaction. When the reaction
is carried out under strictly dry conditions (careful-
ly-dried vessel and reagents), both substantially increased
latency time and weaker exotherm are observed
compared to the reaction carried out without avoiding
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 271–284
DOI: 10.1002/poc

280
O. LAGRILLE ET AL.
Equation 9 is a simple model for the kinetics of either
simplified model assuming the powder to be made of
identical, spherical particles, the kinetic law writes as:
an autocatalytic reaction occurring at constant tempera-
ture (k1 then being the autocatalytic rate constant), or of
an exothermic, zero-order reaction occurring in an
adiabatic system, the latter in the approximations of
both invariant heat capacity along the reaction and linear
dependence of the rate constant k with temperature (e.g. a
linear approximation of the Arrhenius law over a small
temperature range).
Both the reaction exothermicity and the substantial
temperature increase observed (Figs 3 and 6), strongly
argue in favour of the second hypothesis, namely thermal
self-acceleration. This would originate from quite slow
heat diffusion throughout the solid C-Val or towards the
(thermostated) walls of the vessel, compared to the
nitrosation rate, surprisingly even for the slowest
kinetics, for example with O2/NO ¼ 0.15, where Eqn 9
still accurately fits experimental data. Thermal self-
acceleration has already been proposed to account for
unaccountably high activation energies evaluated through
DSC monitoring (cf. previous section).
A combination of both phenomena (thermal self-
acceleration and autocatalysis) is also possible. It is
worthy to note that, when water is present at the beginning
of the reaction, its rate substantially increases over all
steps; since the reaction itself releases water (together
with NCA), this should logically act as an autocatalyst.
Such autocatalysis may be chiefly physical however,
through the formation of a liquid phase and dissolvation
of solid reactants, thus increasing their molecular
mobility.
2=3
dj=dt ¼ kSUC ꢂ ð1 ꢀ jÞ
which integrates as
(11)
3
j ¼ 1 ꢀ ð1 ꢀ kSUC ꢂ ðt ꢀ t0Þ=3Þ
(12)
where kSUC is the (particle-size dependent) rate constant
and t0 is the time at which the reaction is complete (j ¼ 1).
This model fits our data from 40–50% to 100% conversion
(Table 4) with an accuracy similar to the linear fits of
the 50–85% conversion range. Both kSUC or kzero rate
constants increase with either the O2/NO ratio or the
specific surface Ssp of solid C-Val, however without
precise linear relation.
The break in kinetic curves. We hold no definite
explanation yet for the sudden break between self-
accelerated kinetics (parabolic growth) and zero-order
kinetics, systematically occurring at 40–50% conversion
and coincident with the end of the exothermic peak on
thermogramms, and which obviously denotes a signifi-
cant change in the reaction process. As a first hypothesis
(consistent with the rate increase with specific surface), if
the self-accelerated step corresponds to the reaction of
CAA with previously adsorbed NOx species, then the
break corresponds to either exhaustion of these adsorbed
species, or to their desorption because of the temperature
increase. Then the rate-limiting process in the subsequent
zero-order step would be the slow adsorption of NOx
species.
Another, more plausible hypothesis is that this break
would be the transition between a two-phase (solid-gas)
and three-phase (solid-liquid-gas) system during the
reaction, due to the accumulation of both water and nitric
acid generated by the reaction. The formation of such a
liquid phase may have several consequences, both
physical an chemical:
Zero-order kinetics. The linear evolution sequence C
(ca. 40–50% to ca. 90% conversion, assumed to occur at
constant temperature) fits very well a zero-order kinetics,
of rate kzero (Table 4). To better account for the reaction in
a powdered solid, we fitted our data with a shrinking
unreacted core (SUC) model commonly used in
solid-phase chemistry, which hypothesises the (zero-
order) reaction to start at the surface of the material, then
to propagate towards the core of the material, without
diffusion of the reactants into the unreacted solid. With a
ꢇ The formation of such a continuous liquid film between
the (finely divided) solid C-Val substrate and the wall
of the vessel (thermostated to 0-C) should ensure a
Table 4. Kinetic fits for solid C-Val nitrosation at 08C by gaseous NOx under various reaction conditions. Parameters of various
(least square) fitting equations: Eqn 9 for exponential-growth sequence, j ¼ j0 þ kzero ꢂ t for zero-order fit, Eqn 12 for
zero-order with SUC model
Reaction parameters
Exponential growth fit
Zero-order fit
SUC fit
pN O
(10ꢀ3 atm)
ꢆ
Ssp
(cm2 ꢁ gꢀ1
hGi
k1
k0
kzero
kSUC
2
4
O2/NO
(mm)
(minꢀ1
)
(minꢀ1
)
R2
(minꢀ1
)
R2
(minꢀ1
)
R2
0.15
0.25
0.25
0.25
0.4
12.1
23.3
23.3
23.3
41.6
211
358
211
60
231
136
231
810
231
0.425
0.824
0.885
3.09
0.00829
0.0238
0.0144
0.0419
0.0678
0.9997
0.9968
0.9981
1.0000
0.9938
0.034
0.044
0.047
0.17
0.9967
0.9989
0.9997
0.9983
0.9969
0.061
0.099
0.11
0.38
0.23
0.9954
0.9722
0.9789
0.9862
0.9906
211
1.08
0.099
*Calculated at 0 8C and constant volume in an equilibrated mixture of 0.14 atm of NO with appropriate O2 pressure, cf. Supporting Information.
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 271–284
DOI: 10.1002/poc

N-CARBAMOYLAMINO ACID NITROSATION BY NO þ O2
281
more efficient heat transfer, with sudden chilling of the
substrate, significantly lowering the reaction rate and
breaking the self-accelerated kinetics.
measurement of the latent heat of fusion of N2O3
(2.74 kJ ꢁ molꢀ1) for the first time. Comparison of the
{NO þ O2} nitrosating systems to {sodium nitrite þ
TFA} for NCA synthesis shows the former to be more
efficient in terms of either reagent stoichiometric demand,
reaction rate, or further tractability of the crude product.
Kinetic investigation of solid-phase nitrosation of CAA
by gaseous NOx is a challenging problem due to the
number of both physical and chemical elementary steps
involved, with insufficient knowledge on many of them.
Moreover such kinetic investigation of solid-gas reactions
is not frequent in fine organic chemistry. Our investi-
gations on solid N-carbamoyl valine nitrosation by
gaseous NOx revealed two distinct kinetics in the course
of the reaction: first an unexpected auto-accelerating step
with exponential growth, followed by a (pseudo)zero-
order step beyond ca. 40–50% conversion. While the
rationalisation of each step is possible separately if
considering temperature increase due to exothermicity,
possible (auto)catalytic effects related to water formation,
or pseudo-zero-order kinetics in a SUC model, several
hypotheses/explanations remain possible concerning the
neat transition between these two kinetic behaviours,
which may involve both physical (e.g. formation of a
liquid phase) and chemical (depletion of adsorbed
reactive species, or change in reaction intermediates)
factors.
ꢇ Retarded water evaporation, occurring only once a
threshold temperature is attained in the substrate: such
heat dissipation may contribute to lower the tempera-
ture of the substrate, thus decreasing the reaction
apparent rate. This effect is probably weak however:
the vaporisation heat of water (ca. 10 kcal ꢁ molS1) is
indeed significantly lower than the nitrosation exother-
micity (ca. 80 kcal ꢁ molS1, vide supra).
ꢇ A change in the nature of the nitrosating species (vide
supra): while in the solid-gas process the reaction
is probably mediated by reactive covalent species
(N2O3, N2O4), once a liquid phase is formed the
nitrosating species might be ions, such as NOR (from
dissociation of either N2O3 or N2O4) or H2NOR2 (from
protonation of HNO2), due to the acidity of this
liquid phase also resulting from partial hydrolysis
of N2O3 and N2O4 into HNO2 and HNO3, respectively.
We observed (cf. above) the intrinsic reactivity
of H2NOR2 /NOR to be lower than that of either
N2O3 or N2O4.
Acidity and N2O4. Nevertheless this change in
nitrosating intermediates is probably not the main cause
of the kinetic break. Indeed, we observed sharper
exotherms and increased rates of both self-accelerated
and zero-order steps when the reaction was carried out
with either added water at the reaction beginning, or with
increased O2/NO ratios (Figures 5 and 6), while both
factors contribute to the release of more nitric acid,
therefore to the formation of increased NOR amounts.
This acidification of the medium is probably the cause of
the apparent catalytic effect of N2O4: when varying
the O2/NO ratio, increments in rate constants (Table 4)
almost linearly follow the equilibrium partial pressure
of N2O4 (itself increasing with the O2/NO ratio). Further
investigations would be necessary to definitely conclude
on this point.
Beyond the interest of this reaction in the field of
molecular origins of life as a possible prebiotic pathway
from amino acids toward peptides on the primitive
Earth,13,34 further investigation is in progress to improve
the efficiency and selectivity of this NCA synthesis in a
preparative scope, also examining alternative nitrosating
systems, such as nitrosyl halides.
EXPERIMENTAL SECTION
Caution: hazardous compounds
Nitric oxide is skin- and mucous-irritant. Nitrosoureas are
potent carcinogens. All operations involving the handling
of these compounds or their solutions were carried out
wearing gloves in a fume hood. Effluents were destroyed
by treatment with concentrated aqueous sodium hydroxide.
CONCLUSIONS
We have highlighted the major influence of the
O2/NO ratio on the efficiency of solid C-Val nitrosation
by gaseous NO þ O2 mixtures, probably due to the
changing availability of involved nitrosating species,
namely N2O3, N2O4 and HNO2 (the latter formed in
contact with water). Thermoanalytical investigation of
the reaction effected by either of the above species
allowed to evaluate and compare their rate constants,
showing that N2O3 is by far the most reactive under our
conditions. To the best of our knowledge, this is the first
report of such kinetic results concerning the nitrosation of
CAA in the solid phase. Meanwhile, we also report the
Materials and methods
Methanol was obtained from Baeckeroot (France).
Alanine, valine, phenylalanine, tyrosine were obtained
from Aldrich. Potassium cyanate was obtained from
Prolabo. Hydrochloric acid (36%) was obtained from
Carlo Erba (France). Nitric oxide and oxygen were
obtained from L’Air Liquide (France).
Potassium cyanate was washed with methanol then
dried in vacuo prior to use in order to remove ammonia
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 271–284
DOI: 10.1002/poc

282
O. LAGRILLE ET AL.
traces. All other reagents and solvents were used as
received. Gaseous NO and O2 were handled in separate
glass syringes at room temperature and under atmos-
pheric pressure.
400–315 mm sieve pairs. X-ray analysis of C-Val were
performed on a C-Val single crystal prepared by slow
recrystallisation from water.
N-carbamoyltyrosine C-Tyr was prepared using the
procedure described for C-Val, using 5 g (27.6 mmol)
tyrosine and 4.48 g (55.2 mmol) potassium cyanate in
50 ml water at 508C, pH 7.5 for 15 h. The reaction mixture
was concentrated in vacuo at 308C to half-volume then
acidified to pH 2. After recrystallisation from boiling
water, 4.58 g (74%) pure material was obtained. dH 2.80
(2H, ABX syst., Hb, Jab ¼ 5.5 Hz, Ja ꢁ b ¼ 7.5 Hz,
Jbb’ ¼ 14 Hz); 5.65 (2H, s, —NH2), 4.26 (1H, ABX
syst., Ha, Ja ꢁ b ¼ 5.5 Hz, Jab ¼ Ja–NH ¼ 8.0 Hz), 6.08 (1H,
d, —NH—, Ja–NH ¼ 8.0 Hz), 6.75 (2H, d, H1), 6.98 (2H,
d, H2), 9.23 (1H, s, —OH), 12.56 (1H, s, —COOH). dC
37.67 (Cb), 54.78 (Ca), 115.80 (C1), 128.96 (—CAr—
CH2—), 130.97 (C2), 156.76 (—CAr—OH), 158.95
(—NH—CO—NH2), 174.90 (COOH).
Physical measurements
1H-NMR spectra were recorded on a Bruker AC200 or
AC250 spectrometer (200 or 250 MHz, respectively) in
DMSO-d6 solution.
Capillary electrophoretic (CE) separations were per-
formed using an automated CE apparatus (Beckman P/
ACE MDQ, Fullerton, CA, USA) with a negative polarity
on the inlet side of the capillary. Fused-silica capillary –
(Cil Cluzeau) 50 mm i.d. ꢂ 30 cm (20 cm to the detector) –
was initially conditioned with sodium hydroxide (0.1 M
for 15 min under a pressure of 20 psi). Prior to analysis,
the capillary was washed with 0.1 M NaOH (20 psi for
1 min), MV water (20 psi for 0.5 min) and background
electrolyte (20 psi for 10 s). The capillary was thermo-
stated at 258C. Samples were introduced hydrodynami-
cally by the application of a positive pressure of 0.3 psi for
10 s. Nitrate and nitrite were monitored by UVabsorption
at 214 nm. The experiments were carried out applying a
constant voltage of ꢀ10 kV. The background electrolyte
was a 10 mM (pH 8.5) borate buffer containing 150 mM
of sodium chloride.
Nitrosation of C-Val by NaNO2 R TFA in
acetonitrile suspension
In a 100-ml glass flask fitted with a magnetic stirrer were
introduced 100 mg (625 mmol) C-Val, appropriate
amount of sodium nitrite and acetonitrile (50–y ml, y
being related to the amount of TFA to be added). The flask
was closed with a silicon rubber cap, then flushed with
either helium or nitrogen for 30 min, then y ml of a
0.375 M solution of TFA in acetonitrile was injected into
the vessel by means of a glass syringe. Either of the
following procedures was then applied:
DSC analyses were performed using a Mettler-Toledo
DSC–30 apparatus coupled with a TA4000 processor and
monitored using the GraphWare TA72 software, using
sealable, stainless-steel crucibles (120 ml, 20 bar)
obtained from Mettler-Toledo.
a) After stirring for 5 min at room temperature, the
reaction was quenched by injection of an excess 1 N
aqueous sodium hydroxide into the vessel. After
in vacuo evaporation of the solvents, the residue
was dissolved in 50 ml of HPLC eluent (water/aceto-
nitrile: 90/10 v/v), and analysed by HPLC for remain-
ing C-Val titration.
b) After reaction completion (30–120 min), the setup was
flushed by nitrogen for 10 min then the solvent was
removed in vacuo. The residue was then dissolved in
DMSO-d6 for NMR analysis.
X-ray diffraction analysis of C-Val was carried out on a
4-circle CAD4 diffractometer with l ¼ MoKa. Although
the molecular structure was unambiguously established,
standard deviations on bond lengths and angles were too
large (due to poor quality of C-Val single crystals) to be
worthy of separate publication (cf. Supporting Infor-
mation). Powdered solid C-Val density was measured
using a Micromeritics AccuPyc 1330 pycnometer.
Microcalorimetry experiments were carried out using a
Microscal Ltd Flow Microcalorimeter (FMC) system
including a thermal conductivity detector (1308C), a
Bronkhorst High-Tech flowmeter and a Kipp & Zonen
integrator. Temperature monitoring of C-Val reaction
was realised by means of a Checktemp 1 electronic
thermometer (metallic thermal probe).
Nitrosation of C-Val by NO R O2 in acetonitrile
suspension
A 100-ml glass flask fitted with a magnetic stirrer and
containing 100 mg (625 mmol) of C-Val in 50 ml
acetonitrile was closed with a silicon rubber cap and
flushed by nitrogen for 20 min (by means of steel
needles through the silicon cap). 14, 28 or 42 ml (0.625,
1.25 or 1.88 mmol, respectively) of gaseous NO was
then injected into the setup, followed by the appropriate
amount of O2 by means of glass syringes. After stirring
N-Carbamoylamino acids
N-carbamoylvaline C-Val, N-carbamoylphenylalanine
C-Phe and N-carbamoylalanine C-Ala were prepared
according to our previously published procedure.15 Solid
C-Val of selected granulometry was obtained by grinding
pure C-Val then sieving through 80–40, 246–175 and
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 271–284
DOI: 10.1002/poc

N-CARBAMOYLAMINO ACID NITROSATION BY NO þ O2
283
for 30 min at room temperature, the setup was flushed
by nitrogen for 10 min, and the solvent was removed
in vacuo (rotary evaporator). Either of the following
procedures was then applied:
injection of 5 ml of 0.2 N NaOH then worked up as above
for CE analysis.
DSC monitoring of C-Val nitrosation
a) The residue was dissolved in DMSO-d6 for NMR
analysis.
b) The residue was dissolved in 50 ml of HPLC eluent
(water/acetonitrile: 90/10 v/v), and analysed by HPLC
for remaining C-Val titration.
To avoid any premature reaction, all reagents were frozen
with liquid-nitrogen prior to any contact with each other
until DSC experiments.
Nitrosation by HNO2. Finely powdered C-Val (2 mg,
12.5 mmol) and NaNO2 were introduced into the DSC
crucible, which was then frozen (liq. N2) prior to
introduction of frozen TFA. The crucible was sealed then
stored in liquid nitrogen until being placed into the
(pre-frozen) DSC cell. Temperature scanning: S100-C to
Nitrosation of C-Val (solid-gas phase)
A 29-ml, flat-bottom glass reactor containing 30 mg
(187.5 mmol) of C-Val (ground to mesh-size 246–
175 mm) was closed with a silicon rubber septum and
flushed by nitrogen for 10 min (by means of steel needles
through the silicon cap). 4.2 ml (187.5 mmol, 1 eq.) gaseous
NO was then injected into the setup, followed by the
appropriate amount of O2 (0.2 to 6.0 ml) by means of glass
syringes. After 30 min stirring at room temperature, either
of the following procedures was applied:
R30-C at rate 5 or 8-C minS1
.
Nitrosation by N2O4. Solid (liq. N2-frozen) N2O4 was
introduced into the cold DSC crucible, which was then
immediately frozen in liquid nitrogen. A small, open
PTFE cup (hand-made from PTFE 1/800 tubing) contain-
ing 2 mg (12.5 mmol) of finely powdered C-Val, was then
placed into the crucible (on top of solid N2O4). The sealed
crucible was stored in liquid nitrogen until being placed
into the (pre-frozen) DSC cell. Temperature scanning:
a) The setup was flushed by nitrogen for 10 min, then the
residue was dissolved in DMSO-d6 for NMR analysis.
b) 5 ml aqueous NaOH 0.2 N was injected into the reac-
tion mixture, which was then stirred for 1 h at rt. 1 ml
aqueous KBr (c ¼ 8.412 ꢂ 10ꢀ3 mol ꢁ Lꢀ1) was added
to the mixture which was then diluted to 100 ml in
water. An 1 ml aliquot of this solution was submitted
to CE analysis for nitrite and nitrate quantification.
Migration times (min) under the used conditions: 2.45
(Brꢀ: reference), 2.85 (NO2ꢀ), 3.05 (NOꢀ3 ). Extinction
coefficients were previously calibrated using standard
solutions.
S80-C to R50-C at either 5 or 8-C minS1
.
Nitrosation by N2O3. Same procedure as above, using
solid (liq. N2-frozen) N2O3 instead of N2O4. Temperature
scanning: S150-C to R10-C at either 5 or 8-C minS1
.
Preparation of N2O3:27 15 ml of 70% aqueous nitric
acid was added dropwise to 8 g, (40.4 mmol) of dry As2O3
placed under inert atmosphere in a magnetically stirred,
chilled (ice-bath), closed flask. The evolving dark-
brown-reddish gas was directly condensed at S10-C in
a separate flask as a dark-blue liquid, where it was stored
as a liquid under saturating NO atmosphere.
Blank experiments
Estimation of N2O3/N2O4 hydrolysis. The same
setup as above was flushed with nitrogen for 10 min, then
appropriate amounts of NO and O2 were injected by
means of glass syringes. 2 ml of either saturated aqueous
NaCl or 50% aqueous HNO3 were then injected into the
setup. After 10 min reaction, 5 ml of 0.2 N aqueous NaOH
were injected into the setup, while this was quickly
flushed with nitrogen for a few seconds. The resulting
aqueous mixture was stirred for 1 h then worked up as
above for CE analysis.
Solid-Phase nitrosation of C-Val: kinetic
experiments
In a 100–ml glass-reactor was introduced 20 mg
(125 mmol) of C-Val of appropriate granulometry. The
reactor closed with a silicon-rubber septum was flushed
by nitrogen (by means of steel needles through the
septum) for 10 min at room temperature, then for 15 min
at 08C (ice-water bath). After a few minutes standing,
14 ml (625 mmol, 5 eq.) NO then a selected amount of O2:
2.1 ml (93.8 mmol, 0.75 eq.), 3.5 ml (156 mmol, 1.25 eq.)
or 5.6 ml (250 mmol, 2 eq.) were rapidly injected into the
reactor by means of glass syringes. After a given time t
(O2 injection being done as t ¼ 0) the reaction was
quenched by a fast nitrogen flush and the setup was
warmed to rt, then the crude product was immediately
Evaluation of alkaline dismutation of NO. The
above 29-ml reactor containing 30 mg (187.5 mmol) of
C-Val (ground to mesh-size 246–175 mm) and appropriate
amounts of NaNO2 and NaNO3, was closed with a silicon
rubber septum and flushed by nitrogen for 10 min at room
temperature, prior to the injection of the appropriate
amount of NO. The system was immediately quenched by
1
dissolved in DMSO-d6 and analysed by H NMR.
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 271–284
DOI: 10.1002/poc

284
O. LAGRILLE ET AL.
Nitrosation of C-Val with temperature
monitoring
Acknowledgements
This work was financially supported by the CNRS, the
French Ministry of Education and Research, and by
Degussa AG (Germany). We are grateful to Dr Francis
´
Carre from Chemistry Departmen, University of Mon-
tpellier 2 for X-ray diffraction measurements.
The same setup as above was used, with the probe of an
electronic thermometer fitted through the silicon-rubber
cap. 25 mg (156 mmol) C-Val was deposited onto the
metallic thermal probe by repeated recrystallisation from
a saturated methanol solution (2–3 recrystallisation layers
were usually necessary; the amount of C-Val deposited
was checked by weighting the probe before and after
deposition). The probe covered by C-Val was dried in
vacuo at 308C prior to introduction into the reactor.
The setup was then thermostated to 08C (crushed-ice
bath) and flushed with nitrogen prior to injection of NO
and O2 (same amounts and protocol as above). The
temperature was monitored every 5 or 10 s along the
reaction (O2 injection was taken as t ¼ 0). At the end of
the reaction (ca. 20 min), the setup was flushed with
nitrogen and warmed up to rt, then the crude solid was
immediately dissolved in DMSO-d6 and analysed by
1H NMR.
REFERENCES
1. Leuchs H. Ber. Dtsch. Chem. 1906; 39: 857–861.
2. Farthing AC, Reynolds RJW. Nature 1950; 165: 647.
3. Coleman D, Farthing AC. J. Chem. Soc. Abstr. 1951; 3218–3222.
4. Kricheldorf HR. a-Aminoacid-N-Carboxy-Anhydrides and
Related Heterocycles. Synthesis, Properties, Peptide Synthesis,
Polymerisation. Springer-Verlag: Berlin, 1987.
5. Wilder R, Mobashery S. J. Org. Chem. 1992; 57: 2755–2756.
6. Taillades J, Beuzelin I, Garrel L, Tabacik V, Commeyras A. Orig.
Life Evol. Biosphere 1998; 28: 61–77.
7. Collet H, Bied C, Mion L, Taillades J, Commeyras A. Tetrahedron
Lett. 1996; 37: 9043–9046.
8. Meijide F, Vazquez Tato J, Casado J, Castro A, Mosquera M.
J. Chem. Soc., Perkin Trans. 2 1987; 1759–1765.
9. Darbeau RW, Pease RS, Perez EV. J. Org. Chem. 2002; 67:
2942–2947.
Reverse O2-NO injection sequence. The same
protocol as above was applied except O2 was injected
before NO. NO injection was taken as time t ¼ 0.
10. Baldwin JE. J. Chem. Soc. Chem. Commun. 1976; 734–736.
11. Baldwin JE, Cutting J, Dupont W, Kruse L, Silberman L, Thomas
RC. J. Chem. Soc. Chem. Commun. 1976; 736–741.
12. Suzuki M, Yamazaki T, Ohta H, Shima K, Ohi K, Nishiyama S,
Sugai T. Synlett 2000; 189–192.
13. Taillades J, Collet H, Garrel L, Beuzelin I, Boiteau L, Choukroun
H, Commeyras A. J. Mol. Evol. 1999; 48: 638–645.
14. Williams DLH. Nitrosation. Cambridge University Press: Cam-
bridge, 1988.
15. Taillades J, Boiteau L, Beuzelin I, Lagrille O, Biron J-P, Vayaboury
W, Vandenabeele-Trambouze O, Giani O, Commeyras A. J. Chem.
Soc., Perkin Trans. 2 2001; 1247–1254.
16. Rice DE, Crawford GH. J. Org. Chem. 1963; 28: 872–873.
17. Stedman G. J. Chem. Soc. Abstr. 1960; 1702–1709.
18. Casado J, Castro A, Lopez Quintela MA, Mosquera M, Rodriguez
Prieto MF. Monatsh. Chem. 1983; 114: 647–660.
Premixed-NOx injection. The same protocol as above
was applied; the mixing of O2 and NO was realised
immediately prior to injection into the nitrosation setup.
2.1 ml of O2 was first loaded into the glass syringe, which
was then connected to the NO delivery setup. A pressure
slightly exceeding atm pressure was applied to load 14 ml
NO, cheked on syringe graduations. The gas mixture was
allowed to stand for ca. 30 s prior to injection into the
nitrosation setup. When dry conditions were required, the
whole O2/NO handling process was carried out under
nitrogen flush.
19. Casado J, Castro A, Mosquera M, Rodriguez Prieto MF, Vazquez
Tato J. Monatsh. Chem. 1984; 115: 669–682.
20. Verhoek FH, Daniels F. J. Am. Chem. Soc. 1931; 53: 1250–1263.
21. Beattie IR, Bell SW. J. Chem. Soc. 1957; 1681–1686.
22. Lide DR. Handbook of Chemistry and Physics, 81st edition. CRC
Press: Boca Raton, FL, 2000.
Nitrosation of C-Ala, C-Phe, C-Tyr with
temperature monitoring
`
23. Bravo C, Herves P, Leis JR, Pena ME. J. Chem. Soc., Perkin
Trans. 2 1991; 2091–2095.
24. Singer K, Vampley PA. J. Chem. Soc. 1956; 3971–3974.
25. Ridd JH. Quart. Rev. 1961; 15: 418–441.
26. Kim JC, Cho IS, Choi SK. J. Korean Chem. Soc. 1991; 35:
240–248.
27. Schenk PW. In Handbook of Preparative Inorganic Chemistry,
Brauer G (ed.). Academic Press: New York, 1963.
28. In Encyclopedie des gaz - L’air Liquide. Elsevier: Amsterdam,
1976; 1065–1072.
29. Challis BC, Kyrtopoulos SA. J. Chem. Soc., Perkin Trans. II 1978;
1296–1302.
The same procedure as for C-Val was applied using either
21 mg (159 mmol) of C-Ala, 30 mg (144 mmol) of C-Phe,
or 36 mg (161 mmol) of C-Phe (all crystallised onto the
thermometer probe from saturated methanol solution)
instead of C-Val. The NO-then-O2 injection sequence was
applied in all cases.
30. Hisatsune IC, Davlin JP, Wada Y. J. Chem. Phys. 1960; 33:
714–719.
Supplementary Material
31. Varetti EL, Pimentel GC. J. Chem. Phys. 1971; 55: 3813–3821.
32. Nour EM, Chen LH, Laane J. J. Phys. Chem. 1983; 87: 1113–1120.
33. Wang X, Qin Q-Z. Spectrochim. Acta A 1998; 54: 575–580.
34. Commeyras A, Collet H, Boiteau L, Taillades J, Vandenabeele-
Trambouze O, Cottet H, Biron J-P, Plasson R, Mion L, Lagrille O,
Martin H, Selsis F, Dobrijevic M. Polym. Int. 2002; 51: 661–665.
Supporting information is available in Wiley InterScience
at http://www.interscience. wiley.com/jpages/0894-3230/
suppmat/.
Copyright # 2007 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2007; 20: 271–284
DOI: 10.1002/poc