Kinetics of the NH3and CO2 solid-state reaction at low temperature

Article abstract of DOI:10.1039/c4cp02414a

Ammonia and carbon dioxide play an important role in both atmospheric and interstellar ice chemistries. This work presents a theoretical and experimental study of the kinetics of the low-temperature NH₃ and CO₂ solid-state reaction in ice films, the product of which is ammonium carbamate (NH₄⁺NH₂COO^-). It is a first-order reaction with respect to CO₂, with a temperature-dependent rate constant fitted to the Arrhenius law in the temperature range 70 K to 90 K, with an activation energy of 5.1 ± 1.6 kJ mol^-1 and a pre-exponential factor of 0.09 ^+1.1_-0.08 s^-1. This work helps to determine the rate of removal of CO₂ and NH₃, via their conversion into ammonium carbamate, from atmospheric and interstellar ices. We also measure first-order desorption energies of 69.0 ± 0.2 kJ mol^-1 and 76.1 ± 0.1 kJ mol^-1, assuming a pre-exponential factor of 10¹³ s^-1, for ammonium carbamate and carbamic acid, respectively.

Full text of DOI:10.1039/c4cp02414a

PCCP
View Article Online
View Journal | View Issue
PAPER
Kinetics of the NH3 and CO2 solid-state reaction
at low temperature
Cite this: Phys.Chem.Chem.Phys.,
2014, 16, 23604
J. A. Noble,a P. Theule,*a F. Duvernay,a G. Danger,a T. Chiavassa,a
P. Ghesquiere,b T. Minevac and D. Talbib
Ammonia and carbon dioxide play an important role in both atmospheric and interstellar ice chemistries.
This work presents a theoretical and experimental study of the kinetics of the low-temperature NH3 and
CO2 solid-state reaction in ice films, the product of which is ammonium carbamate (NH4+NH2COOꢀ). It
is a first-order reaction with respect to CO2, with a temperature-dependent rate constant fitted to the
Arrhenius law in the temperature range 70 K to 90 K, with an activation energy of 5.1 ꢁ 1.6 kJ molꢀ1 and
Received 2nd June 2014,
Accepted 22nd August 2014
s
ꢀ1. This work helps to determine the rate of removal of CO2
+1.1
a pre-exponential factor of 0.09
ꢀ0.08
and NH3, via their conversion into ammonium carbamate, from atmospheric and interstellar ices. We
also measure first-order desorption energies of 69.0 ꢁ 0.2 kJ molꢀ1 and 76.1 ꢁ 0.1 kJ molꢀ1, assuming
DOI: 10.1039/c4cp02414a
a pre-exponential factor of 1013
s
ꢀ1, for ammonium carbamate and carbamic acid, respectively.
www.rsc.org/pccp
Its chemical capture through conversion into non-volatile
species is a vast field of research.3 Plants naturally fix the
1 Introduction
Ammonia (NH3) and carbon dioxide (CO2) are ubiquitous CO2 they need for their growth in the form of carbamate using
constituents of the Earth’s atmosphere and of the molecular the ribulose 1,5-bisphosphate carboxylase/oxygenase enzyme in
interstellar medium.
the first step of the Calvin cycle.3 It is therefore important to
On Earth, ammonia is considered to be the most important understand how the reactivity between NH3 and CO2 can lead to
of all nitrogen-bearing species which are deposited onto other non-methane volatile organic compounds (NMVOC) or
vegetation and other receptors.1 Ammonia originates from refractory species on aerosols or in terrestrial ice. The reaction
animal waste and the volatilisation of synthetic fertilizers, of carbon dioxide with Bronsted bases is of great importance
biomass burning, losses from soils bearing native vegetation in biology, geology, and for industrial applications.4 As an
or agricultural crops, and fossil fuel combustion.2 Ammonia is example, amines (RNH2) are used to remove carbon dioxide
a very important alkaline constituent in the atmosphere as it from gas streams.5
reacts readily with acidic substances, e.g. sulfuric acid (H2SO4)
In space, water, carbon dioxide, and ammonia are among
or nitric acid (HNO3), to form ammonium (NH4 ) salt aerosols.1 the most abundant species present in interstellar ices. They are
The removal of gaseous NH3 from the atmosphere by the formed during the transition from the diffuse atomic medium
+
+
formation of fine particulate NH4 salts is a very efficient to the dense molecular medium via simple atomic or diatomic
mechanism. The effects of NH3 when combined with other air reactions on the surface of bare interstellar grains. Ammonia is
pollutants, such as the all-pervasive ozone (O3) or the increas- formed by the hydrogenation of the N atom.6 Carbon dioxide
ingly abundant carbon dioxide (CO2), are poorly understood. is thought to be formed from the HO–CO complex7–9 or by the
Atmospheric CO2 is of prime importance in plant and algal CO + O addition reaction.10 Water can be formed by the
photosynthesis, as well as for the greenhouse effect. It originates hydrogenation of the O atom, O2 or O3 molecules, or the OH
from the respiration processes of living aerobic organisms, radical.11–15 In molecular clouds and in solar system bodies,
organic matter, and volcanic outgassing. Its increasing concen- carbon dioxide and ammonia have abundances on the order
tration contributes to global warming and ocean acidification. of 20% and 5%, respectively.16 Under interstellar conditions,
both molecules are transformed into other species either
thermally17–20 or non-thermally, due to the effect of UV photon
a Aix-Marseille Univ., CNRS, PIIM UMR 7345, Marseille, 13397, France.
irradiation,21 proton or heavy ion bombardment,22,23 and/or
electron bombardment.24 Products will either enrich the gas-
phase of the interstellar medium in complex organic molecules
when the ice mantle is desorbed or be a constituent of the
refractory organic residue.
E-mail: patrice.theule@univ-amu.fr; Tel: +33 4 91 28 85 82
b
´
Laboratoire Univers et Particules de Montpellier UMR 5299, CNRS et Universite
`
Montpellier 2, Place Eugene Bataillon, 34095 Montpellier cedex 05, France
c Institute Charles Gerhardt, UMR 5253 CNRS/ENSCM/UM2/UM1, 8 rue de l’Ecole
Normale, 34296 Montpellier cedex 05, France
23604 | Phys. Chem. Chem. Phys., 2014, 16, 23604--23615
This journal is © the Owner Societies 2014

View Article Online
Paper
In this work we investigate the reaction mechanisms and the
PCCP
2 Methods
2.1 Experimental methods
kinetics of the NH3 : CO2 solid-phase system, with the reactants
located on neighboring sites. Low-temperature NH3 : CO2
thermal reactivity has been shown to lead to the formation of
The experiments were performed using our RING experimental
set-up, as described elsewhere.26 In a high-vacuum chamber
(few 10ꢀ9 mbar) a gold-plated copper surface is held at low
temperature using a closed-cycle helium cryostat (ARS Cryo,
model DE-204 SB, 4 K cryogenerator). The sample temperature
is measured using a DTGS 670 silicon diode with an uncertainty of
0.5 K. The temperature is controlled using a Lakeshore Model
336 temperature controller and a heating resistance. Infrared
absorption spectra are recorded in reflection mode by means of
Fourier-transform infrared reflection absorption spectroscopy
(or FTIR) using a Vertex 70 spectrometer with either a DTGS
detector or a liquid N2-cooled MCT detector. A typical spectrum
has a resolution of 0.5 cmꢀ1 and is averaged over a few tens of
interferograms. Mass spectra of the gas-phase species are recorded
using a Hiden HAL VII RGA quadrupole mass spectrometer (QMS).
The ionization source is a 70 eV impact electronic source. Mass
spectra are recorded between 1 and 60 amu.
NH3 : CO2(:H2O) ice films are formed by vapour deposition.
Gas-phase CO2 and NH3 are commercially available in the form
of 99.9995% pure gas and obtained from Linde and Air Liquide,
respectively. The H2O vapour is obtained from deionised water
which has been purified by several freeze–pump–thaw cycles,
carried out under primary vacuum. The different gases are
introduced and mixed together in a defined concentration ratio
in a primary pumped vacuum line using standard manometric
techniques at room temperature. The homogeneously mixed
gas-phase mixture is then sprayed onto the cold gold-plated
copper surface.
+
the ammonium carbamate salt ([NH4 ][NH2COOꢀ]), carbamic acid
(NH2COOH)17–20 and, in the presence of water, the ammonium
bicarbonate salt ([NH4+][HCO3ꢀ]).20 While these previous
studies spectroscopically characterise the reaction products,
the mechanisms and the kinetics of the reaction NH3 + CO2
have not yet been addressed experimentally. High level calcula-
tions on the solvation of CO2 in water–ammonia clusters of four
molecules are reported,25 with barriers of solvation determined
between 60 and 110 kJ molꢀ1. Here, we present the results of
dedicated isothermal laboratory experiments on NH3 : CO2 ice
films of different concentration ratios, measured using Fourier-
transform infrared spectroscopy (FTIR). We measured the
reaction barrier for the formation of ammonium carbonate to be
Etah = 5.1 kJ molꢀ1 ꢁ 1.6 kJ molꢀ1. The substitution of a NH3
molecule by a H2O molecule is also investigated. The question of
the diffusion of the reactants within the ice has been purposefully
separated from the reactivity of the reactants, and will be
addressed in a future publication (Ghesquiere et al., in prepara-
tion). We also performed calculations based on density functional
theory to mimic the reaction of NH3 and CO2 in a water–ammonia
cluster to aid the interpretation of experimental results and
provide insight into the mechanism(s) of the formation of
ammonium carbamate (NH4+NH2COOꢀ) and carbamic acid
(NH2COOH) in a model NH3 : CO2:H2O ice mixture.
Reaction mechanisms
The formation of the products NH2COOH (A) and NH4 NH2COOꢀ
+
The morphology of the NH3 : CO2(:H2O) ice film depends on
the temperature of the gold surface on which the gas-phase
mixture is deposited.27 If the ice film is deposited below 60 K,
the ice film is amorphous and porous. If deposited between
60 K and 150 K, the ice is amorphous and compact, with little or
no porosity. If deposited above 150 K, the ice is crystalline. We
deposit the homogeneous NH3 : CO2 gas-phase mixture at 60 K
in order to measure the NH3 + CO2 reactivity in compact
amorphous ice. Such ice films are better defined and more
reproducible than porous amorphous ice films. Pore collapse
in the latter can change the overall kinetics by introducing
additional reorganisation kinetics.28
(C) from the reactants NH3 and CO2 can occur via three
mechanisms.
In the first mechanism, NH2COOH is a reaction intermediate,
+
and NH4 NH2COOꢀ is the final product of a sequence of two
successive reactions:
NH3 + CO2 " NH2COOH
(1)
(2)
NH3 + NH2COOH " NH4 NH2COOꢀ
+
In the second mechanism, the reaction pathway proceeds via a
+
zwitterion (NH3 COOꢀ, Z) intermediate, rather than carbamic acid:
+
NH3 + CO2 " NH3 COOꢀ
(3)
(4)
The column density (N, molecules cmꢀ2) of each molecular
species is derived immediately after deposition from the IR spectra,
an example of which is presented in Fig. 1(a), using the expression:
+
+
NH3 + NH3 COOꢀ " NH4 NH2COOꢀ.
Ð
If we are not able to observe the zwitterion intermediate, it
cannot be included in the kinetic analysis and thus we cannot
distinguish between reactions (1) and (3).
tndn~
A
N ¼
;
(7)
In the third mechanism, NH4+NH2COOꢀ is produced as a where the optical depth (tn) is equal to ln(10) times the inte-
reaction intermediate (either directly or via a zwitterion, as in grated absorbance and A is the band strength, in cm moleculeꢀ1
.
the second mechanism); NH2COOH is the final product of a Ammonia is identified via its umbrella mode at 1110 cmꢀ1, the
sequence of two successive reactions:
band strength of which is 1.3 ꢂ 10ꢀ17 cm molecule
.
ꢀ1 29 Carbon
dioxide is identified via its asymmetric stretching mode at
2339 cmꢀ1 and its bending mode band at 667 cmꢀ1. The band
strength of the CO2 asymmetric stretching band was measured
+
2NH3 + CO2 " NH4 NH2COOꢀ
(5)
(6)
NH4 NH2COOꢀ " NH3 + NH2COOH
+
This journal is ©the Owner Societies 2014
Phys. Chem. Chem. Phys., 2014, 16, 23604--23615 | 23605

View Article Online
PCCP
Paper
the ice film. The factor one half comes from the reflection mode,
which probes double the column density.
We use two approaches to derive the relevant physical and
chemical parameters of the NH3 : CO2 system:
ꢅ During isothermal kinetic (IK) experiments, infrared spectra
of the solid-phase molecules are recorded at regular time intervals
at a fixed temperature.
ꢅ During temperature ramp experiments, IR spectra of the
solid-state molecules are recorded at regular time (temperature)
intervals, and mass spectra of the gas-phase molecules are
recorded as they desorb during a temperature ramp.
Fig. 1 Infrared absorption spectra of a homogeneously mixed NH3 : CO2
ice film with an excess of NH3 held at 80 K. Spectra shown are: (a) at t = 0,
(b) after 11 hours at 80 K, and (c) the difference spectrum of (b) minus (a)
magnified by a factor of 10. The CO2 bands have decreased while the
carbamate bands have appeared.
2.1.1 Isothermal kinetic experiments. In isothermal kinetic
(IK) experiments, immediately after deposition the ice film is
heated as fast as possible to a fixed temperature (T), in typically
a few tens of seconds. Once the temperature, T, is reached, we
set the initial time, t = 0 s, of our isothermal experiment. NH3
and CO2 react to form reaction products, as can be seen from
the IR spectrum and the difference spectrum of the ice mixture
at the end of the experiment, as displayed in Fig. 1(b) and (c),
respectively. The kinetics of the NH3 + CO2 ice mixture at
the fixed temperature, T, are monitored by measuring the
disappearance of the reactants and appearance of the products
from their characteristic IR absorption bands as a function of
time, as illustrated for a typical isothermal experiment in Fig. 2.
Isothermal kinetic experiments are performed at different
temperatures to study the temperature dependence of the reactivity.
These kinetics will determine the removal of NH3 and CO2 in
atmospheric or interstellar ices. Since both in atmospheric and
interstellar chemistry this reaction is competing with other reac-
tions, such as NH3 + HNCO,28 NH3 + H2CO,33 or CH3NH2 + CO2,34
to be 7.6 ꢂ 10ꢀ17 cm moleculeꢀ1 for the pure solid,30,31 while in
water ice, a value of 1.4 ꢂ 10ꢀ17 cm moleculeꢀ1 was found.32 Water
ice has three characteristic bands at 3280, 1660, and 760 cmꢀ1
corresponding to the OH stretching, HOH bending, and libration
modes, respectively. The corresponding band strengths are
2.1 ꢂ 10ꢀ16, 3.1 ꢂ 10ꢀ17, and 3.1 ꢂ 10ꢀ17 cm moleculeꢀ1
,
respectively.31 The frequencies and band strengths of the carba-
mate are taken from Bossa et al.19 There is an approximately
30% uncertainty on the band strengths and therefore on the
calculated column densities. The characteristic frequencies of
the different products are listed in Table 1.
The ice film thickness is determined from the measured
column density (N, molecule cmꢀ2), assuming r = 0.94 g cmꢀ3
as the amorphous ice density, and using:
N ꢂ 17
l
¼
ꢂ cosð18ꢄÞ=2;
(8)
½cmꢃ
r ꢂ NA
where NA is the Avogadro constant, 17 g molꢀ1 is the molar
mass for NH3, and the cos(181) term comes from the 181
incidence angle between the FTIR beam and the normal to
Table 1 Infrared absorption bands and assignment of the products
formed from the reaction of NH3 + CO2
Assignment
Wavenumbers (cmꢀ1
)
n(NH2) NH2COOH dimer
n(NH2)[NH4+][NH2COOꢀ]
n(NH2)[NH4+][NH2COOꢀ]
n(OH) NH2COOH
3462
3428
3325
3140a
1691
1623
1553
1495
1451
1393
1320
1117
1037
829
n(CQO) NH2COOH
d(NH2)[NH4+][NH2COOꢀ]
n
as(COOꢀ)[NH4+][NH2COOꢀ]
d(NH4+)[NH4+][NH2COOꢀ]
NH2COOH dimer
n(CN) [NH4+][NH2COOꢀ]
NH2COOH dimer
ns(COOꢀ) [NH4+][NH2COOꢀ]
Fig. 2 Isothermal kinetic experiments at 75
K carried out on a
r(NH2) [NH4+][NH2COOꢀ]
homogeneously mixed NH3 : CO2 ice film, deposited at 60 K with an
excess of NH3 (solid lines). The CO2 column density exponentially
d
oop(OCN) [NH4+][NH2COOꢀ]
d(COOꢀ) [NH4+][NH2COOꢀ]
674
decreases with time as the NH4+COOꢀ (from its band at 1393 cmꢀ1
increases (dashed lines).
)
a
Broad band.
23606 | Phys. Chem. Chem. Phys., 2014, 16, 23604--23615
This journal is © the Owner Societies 2014

View Article Online
Paper
PCCP
it is important to know its kinetic parameters in order to
compare it with the other reactions.
The temperature range within which isothermal experi-
ments can be performed is limited. The lowest temperature is
constrained by the maximum time we can wait. This time is
defined as the amount of background deposition onto the ice film
from the vacuum chamber; at 10ꢀ8 mbar, ca. 360 monolayers are
deposited in 10 hours. The highest temperature is constrained
by desorption of our ice film. With pre-exponential factors of
5 ꢂ 1013
s
ꢀ1 and 6 ꢂ 1014
s
ꢀ1 and desorption energies of 25 and
23 kJ molꢀ1 for NH3 and CO2, respectively,35,36 the desorption
rates of NH3 and CO2 can be evaluated. For example, at 80 K, it
takes approximately seven minutes to desorb one monolayer.
2.1.2 Temperature ramp experiments. In temperature pro-
grammed reactivity (TPR) experiments, the temperature is linearly
increased at a ramp rate, b, from the deposition temperature, T0:
Fig. 3 NH4+NH2COOꢀ produced during temperature programmed
reactivity experiments on NH3 :CO2 ice mixtures with an excess of NH3. The
temperature is ramped between 60 K and 160 K at rates of 0.5 (diamonds),
1 (crosses), 2 (triangles) and 5 (circles) K minꢀ1. NH4+NH2COOꢀ abundances
are monitored using the 1393 cmꢀ1 band. IR bands corresponding to
NH2COOH are not observed when NH3 is in excess. Although the initial
quantity of matter (the curve area) is different in the four experiments, the
slope of the curves is the same.
T = T0 + b ꢂ t.
(9)
The abundances of the solid-phase species during the
temperature ramp are measured from FTIR spectra.
In temperature programmed desorption (TPD) experiments, the
temperature is also linearly increased, as in eqn (9). During heating,
the mass spectrum of the sublimating species is recorded;
desorption parameters are derived from TPD experiments.
any temperature below the desorption temperature of CO2. This
means that the reaction barriers are too high and, as a result,
desorption of CO2 is faster than any reaction. This case, where
ice is dominated by a non-polar molecule such as CO2, is less
usual in real ices, which are usually water-dominated. Thus we will
focus on NH3 :CO2 where polar and hydrogen bonded molecules,
such as NH3 and H2O, are dominant.
2.2 Computational methods
3.1.2 Reactivity of the NH3 : CO2 system with NH3 in excess.
We perform TPR experiments on NH3 : CO2 ice films deposited
at 60 K and heated at different temperature ramp rates, as
shown in Fig. 3. The characteristic IR bands of NH4+NH2COOꢀ
increase as it is produced and decrease as it sublimates. NH2COOH
is not identified in our spectra, either as a transitory species or as
a final product, when NH3 is in excess. Indeed, carbamic acid is
observed in TPR experiments, at high temperatures, only when
CO2 is present in an equal concentration with NH3, as previously
shown.19,20 The absence of NH2COOH tends to validate either
the second mechanism, with a non-observable zwitterion inter-
mediate, where the activation energy of reaction (4) is lower
than that of reaction (3), or the third mechanism, a serial
formation pathway of NH2COOH from NH4+NH2COOꢀ, the
second step (reaction (6)) having too large an energy to occur
before carbamate desorption.
Energy profile calculations have been performed within the
framework of density functional theory using the hybrid general-
ised gradient approximation (GGA) B3LYP functional37,38 in
conjunction with a triple zeta atomic basis set extended by
polarisation functions i.e., 6-311G(d,p). Structures at minima
and maxima (transition states) on the reaction paths have
been located by total optimisation using analytical gradients.
Harmonic vibrational analyses have been performed at the
same levels of theory to confirm each stationary point as either
an equilibrium structure (i.e., all real frequencies) or a transi-
tion structure (TS) (i.e., with one imaginary frequency). Intrinsic
reaction coordinate calculations have been conducted in order
to reliably link the transition states with the corresponding
minima. All energies have been corrected for unscaled zero-point
energies calculated at the same theoretical level. All calculations
have been performed using the GAUSSIAN-09 package.39
Thus, the experimental observations can be summarised as:
+
2NH3 + CO2 " NH4 NH2COOꢀ
(10)
3 Results
3.1 Experimental results
We assume that the reaction can be described using kinetic
equations of the form:
We perform several IK and TPR experiments at different tempera-
tures on several CO2 :NH3 ice mixtures with varying concentration
ratios to investigate the reaction mechanism and measure
the kinetics of the reaction. First we study the reactivity in a
non-protic CO2 dominated ice, and second in a protic NH3
dominated ice. The possible influence of H2O is addressed.
dðCO2Þ
a
b
¼ ꢀkðTÞ ꢂ ðNH3Þ ꢂðCO2Þ
dt
(11)
dðNH4þNH2COOꢀÞ
dt
a
b
¼ kðTÞ ꢂ ðNH3Þ ꢂðCO2Þ
3.1.1 Reactivity of the NH3 : CO2 system with CO2 in excess. where k(T) is the temperature-dependent reaction rate constant,
In an NH3 : CO2 ice mixture with a 1 : 20 concentration ratio, a and b are the partial orders of reaction, and (X) the molar
where NH3 is diluted in a CO2 ice, no reactivity is observed at fraction of the X species (with no dimension). In the case of
This journal is ©the Owner Societies 2014
Phys. Chem. Chem. Phys., 2014, 16, 23604--23615 | 23607

View Article Online
PCCP
Paper
excess NH3, (NH3) C 1 and we measure the effective reaction
rate constant k0 = k(T) ꢂ (NH3)a.
In NH3 : CO2 ice mixtures where CO2 is diluted in an excess
of NH3, each CO2 molecule is surrounded by an NH3 molecular
environment and we can assume pseudo first-order kinetics for
every reaction. The NH3 : CO2 ice films are deposited at 60 K.
The IR bands of reactants decrease and product bands increase
as the temperature is increased, which implies that NH3 and
CO2 react.
3.1.2.1 Determination of the partial orders of reaction. Iso-
thermal kinetic experiments on NH3 : CO2 ice films with an
excess of NH3 exhibit a time dependence, with CO2 and
NH4 NH2COOꢀ concentrations evolving as depicted in Fig. 2.
+
Fig. 4 Dependence of the pseudo first-order reaction constant k0 = k ꢂ
(NH3)a on the (NH3) molar fraction at T = 80 K. We limit (NH3) to
approximately 0.66, since we need at least two NH3 molecules to react
with one CO2 molecule.
The exponential decay of CO2 and the corresponding growth of
+
NH4 NH2COOꢀ exhibit pseudo first-order reaction kinetics, with
a rate constant k0, i.e. b = 1. The system of kinetic eqn (11) is
solved analytically as:
0
(CO2)(t) = (CO2)0 ꢂ eꢀk t
(12)
(13)
0
(NH4+NH2COOꢀ)(t) = (CO2)0(1 ꢀ eꢀk t).
Fitting IK experiments with eqn (12) and (13) gives the pseudo
first-order reaction constant, k0, at the fixed temperature T. To
determine the partial reaction order, a, we measure k0 at different
(NH3) molar fractions at the same temperature, as shown in
Table 2 and in Fig. 4. The pseudo first-order reaction constant
k0 shows no dependence on the NH3 molar fraction, which
indicates that a = 0 and therefore that k0 = k. Due to the relatively
large uncertainties on our measurements, the determination of
a is tentative. In this particular case, the unit of k is sꢀ1
.
3.1.2.2 Determination of the temperature dependence of the
reaction rate. In order to determine the temperature depen-
Fig. 5 Logarithm of the pseudo first-order reaction rate constant, k, as a
dence of the reaction rate constant k(T), different IK experi- function of the inverse of the temperature. The uncertainty on each point
is estimated from both the fitting procedures for the CO2 band and the
ments are performed at different fixed temperatures. They are
summarised in Table 2.
four carbamate bands time decay, and from the dispersion between two
measurements taken at 70 K and 75 K.
The temperature dependence of the pseudo first-order reac-
tion rate constant, k, is determined from the values of mixtures
with an excess of NH3 in Table 2, and is displayed in Fig. 5. The
uncertainties are evaluated by (i) taking into account the fitting
Table 2 Reaction rate constants, k, from isothermal experiments with
uncertainty on each fitting procedure, (ii) the dispersion on the
rate constants obtained from the CO2 band and from the four
carbamate bands, and (iii) from the dispersion on two experi-
ments performed at the same temperature. The uncertainty is
dominated by the dispersion, which results from the pore
collapse of the out-of-equilibrium ice.28 The uncertainty on
the temperature measurement is 0.5 K.
different NH3 : CO2 concentration ratios. Uncertainties on the last digit of
the rates are in parenthesis
T(K)
Concentration ratio
Rate constant k0 (sꢀ1
)
CO2 : NH3 ice films
70–95
70
70
75
75
80
85
90
80
Excess of CO2
No reaction
Excess of NH3
Excess of NH3
Excess of NH3
Excess of NH3
Excess of NH3
Excess of NH3
Excess of NH3
1 : 1
9.6(2) ꢂ 10ꢀ6
1.0(1) ꢂ 10ꢀ5
4.8(3) ꢂ 10ꢀ5
2.0(5) ꢂ 10ꢀ5
4.9(3) ꢂ 10ꢀ5
6.4(4) ꢂ 10ꢀ5
6.4(4) ꢂ 10ꢀ5
6.8(5) ꢂ 10ꢀ5
4.9(4) ꢂ 10ꢀ5
7.8(5) ꢂ 10ꢀ5
Fitting the experimental curve with the Arrhenius law gives an
experimental activation energy of 5.1 ꢁ 1.6 kJ molꢀ1 (642 ꢁ 62 K)
and a pre-exponential factor of 0.09
+1.1 sꢀ1 for a temperature
ꢀ0.08
interval between 70 K and 90 K. The small value of the pre-
exponential factor can have two origins: either the measured
pseudo first-order rate represents a sequence of elementary
processes, or there is a contribution from quantum tunnelling
between 70 K and 90 K.40 Such a quantum tunnelling effect has
been observed at lower temperature (8–15 K) in the HNCO + NH3
80
80
1 : 4
1 : 13
CO2 : NH3 : H2O ice films
80
80
1 : 60 : 60
1 : 60 : 60
5.8(4) ꢂ 10ꢀ5
9.0(4) ꢂ 10ꢀ5
23608 | Phys. Chem. Chem. Phys., 2014, 16, 23604--23615
This journal is © the Owner Societies 2014

View Article Online
Paper
PCCP
system28 (although not in the HNCO + H2O system26) and it carbamate has co-desorbed with them. Above 150 K the ice film
contributes to the reaction rate in the15–40 K range, along with is made of carbamate only and carbamate is converted into
the thermal reaction rate. The latter takes over as the reaction carbamic acid, as seen in Fig. 6 between 150 K and 225 K.
temperature increases.
Above 225 K, both the carbamate and carbamic acid desorb.
3.1.3 Determination of the influence of H2O. In order to These experiments clearly show that we have an additional
investigate the change in reactivity induced by replacing a NH3 high-temperature mechanism of conversion of the carbamate
molecule by a H2O molecule, including its possible effect on the into carbamic acid:
height of the reaction barrier, we perform IK experiments on
+
CO2 : NH3 : H2O ice films with different NH3 : H2O ratios. We
take care not to go below a 1 : 1 ratio, since we need at least two
NH3 molecules in the four closest neighbouring molecules of
NH4 NH2COOꢀ " NH3 + NH2COOH.
(15)
This is consistent with reaction (6) of the third reaction
mechanism, although it could also be the reverse reaction of
reaction (2).
the first shell around CO2 to form NH4 NH2COOꢀ. If no
+
NH3 molecule is present in the first shell, it needs to diffuse
toward CO2 through the H2O molecules, which alters the
kinetic measurements. We see in Table 2 that the rates at
80 K are, within the dispersion uncertainty, similar or slightly
higher if the NH3 molecule is replaced by a H2O molecule.
This means that, within the dispersion of our experimental
methods, NH3 and H2O have more or less the same effect on
the reaction barrier.
3.1.5 Determination of the desorption energy of ammonium
carbamate and carbamic acid. To determine the desorption para-
meters of ammonium carbamate, we perform a TPD experiment on
a CO2 :NH3 ice mixture with an excess of NH3. The mixture was
held at 75 K for 12 hours in an IK experiment, then the temperature
was ramped at a rate of 5 K minꢀ1. The TPD profile of m/z 44
displayed in Fig. 7 shows three desorption features, which can
be assigned from the correlation with IR bands. The first
In the presence of ꢀwater, the formation of ammonium
bicarbonate NH4 HCO3 can also take place.25
+
36,41
feature at 120 K corresponds to the desorption of CO2
which did not react with NH3. The second feature at 240 K
corresponds to the desorption of NH4+NH2COOꢀ. The third
feature at 265 K corresponds to the desorption of NH2COOH.
+
ꢀ
NH3 + CO2 + H2O " NH4 HCO3
.
(14)
We do not observe the formation of ammonium bicarbonate in
our experiments, perhaps due to the limited range of temperature.
3.1.4 Reactivity of the CO2 : NH3 system with equal
amounts of CO2 and NH3. When CO2 and NH3 are present in
comparable amounts in ice films, both ammonium carbamate
and carbamic acid are formed.19,20
+
We see in Fig. 6 that ammonium carbamate (NH4 NH2COOꢀ)
forms first, and that carbamic acid (NH2COOH) forms later from
the carbamate. Above 150 K, the initial reactants CO2 and NH3
have desorbed and are not present anymore. Part of the formed
Fig. 6 A temperature programmed reactivity experiment performed at
5 K minꢀ1 on a CO2 : NH3 ice mixture in a 1 : 1.5 concentration ratio after an
isothermal experiment at 80 K lasting 26.75 hours. The evolution of
NH4+NH2COOꢀ (circles, dashed line) and NH2COOH (triangles, dashed
line) is monitored via their bands at 1393 cmꢀ1 and 3462 cmꢀ1, respec-
tively, when the temperature is ramped between 60 K and 300 K. The
(blue) dotted line (circles) is a TPR experiment performed immediately after
the CO2 : NH3 ice mixture deposition (data from Fig. 3, 5 K minꢀ1, circles).
The desorption of reactants prevents the formation of large quantities of
carbamate, and consequently of carbamic acid as well.
Fig. 7 Temperature programmed desorption experiment (m/z 44) at a 5 K
minꢀ1 temperature ramp of (a) a CO2 : NH3 ice mixture with an excess of
NH3, and (b) an approximately equal amount of CO2 and NH3. The
shoulder on the desorption feature is due to the crystallisation. The first
desorption feature at 120 K corresponds to CO2 desorption, the second
one at 240 K to NH4+ : NH2COOꢀ desorption, and the third one at 265 K to
NH2COOH desorption, according to their correlation with IR bands. Fitting
the experimental curve with
a first-order Polanyi–Wigner equation
(dashed line) gives the desorption parameters of NH4+ : NH2COOꢀ and
of NH2COOH.
This journal is ©the Owner Societies 2014
Phys. Chem. Chem. Phys., 2014, 16, 23604--23615 | 23609

View Article Online
PCCP
Paper
Replacing time by temperature, using eqn (9), in the Polanyi– observed, but rather m/z 61, corresponding to NH2COOH, is
Wigner equation, we get:
weakly observed, and the masses corresponding to NH3 and
CO2 dominate. The carbamate decomposes in the gas phase,
either spontaneously or under the electron impact, into NH3
and NH2COOH and then into NH3 and CO2.
N
n
¼ ꢀ exp ꢀEdesRT ꢂ Nn;
(16)
dT
b
where n and Edes are the pre-exponential factor and desorption
energy of the desorption rate constant, respectively, and n is the
order of the desorption. Fitting the carbamate desorption
feature at 240 K with a first-order Polanyi–Wigner equation
while fixing the pre-exponential factor at a typical value of n =
3.2 Theoretical results
Three different clusters of eight molecules were constructed in
order to study the reactivity of NH3 and CO2 in water:ammonia
environments. These clusters were designed to provide a
fully solvated first solvation layer (i.e. six molecules around
the NH3–CO2 pair). Such cluster calculations are inherently gas
phase calculations and thus resemble the energetics of the
solid state reaction only approximately, independent of the
quantum chemical method/DFT functional used.
The first cluster was designed to mimic the reactivity of
CO2 and NH3 in a pure ammonia ice, and therefore it was
constructed from one CO2 and seven NH3 molecules. This
cluster is labeled ‘‘7 : 1 : 0’’, derived from the NH3 : CO2 : H2O
concentration ratio. The two other clusters were designed to
mimic the reaction of NH3 and CO2 in a diluted, water-dominated
environment. One was composed of one CO2, one NH3, and six
H2O molecules, hereafter denoted ‘‘1 : 1 : 6’’. The other was
composed of one CO2, two NH3, and five H2O molecules,
hereafter denoted ‘‘2 : 1 : 5’’.
1013
s
ꢀ1, we get Edes = 69.0 ꢁ 0.2 kJ molꢀ1. Fitting the carbamic
acid desorption feature at 265 K with a zeroth-order Polanyi–
Wigner equation while fixing the pre-exponential factor at a
typical value of n = 1013
The higher desorption energy of carbamic acid is probably due
to its ability to arrange into dimers.19
s
ꢀ1, we get Edes = 76.1 ꢁ 0.1 kJ molꢀ1
.
Fig. 8 shows the mass spectrum of ammonium carbamate
and carbamic acid at their maximum desorption temperatures,
i.e. 240 K and 265 K, respectively. For carbamic acid, the
molecular ion at m/z 61 is weak but detectable and masses
m/z 17, 18, and 44, corresponding to NH3 and CO2, dominate.
Carbamic acid decomposes into NH3 and CO2 either sponta-
neously or under electron impact, but is stable for at least
some time since the molecular ion at m/z 61 is observed. For
ammonium carbamate, the molecular ion at m/z 78 is not
We would like to point out here that our goal with these
calculations is not to provide accurate energy barriers, but rather
to compare different mechanisms in order to assess the catalytic
role of ammonia and water molecules in the formation of
ammonium carbamate and carbamic acid in model ices,
and thus to rationalise the present experimental results. The
structures located in the following energy profiles are all
obtained through an optimisation procedure allowing for the
total relaxation of all eight molecules.
3.2.1 Reaction profile of the 7 : 1 : 0 cluster. The reaction of
NH3 and CO2 is investigated in the presence of six surrounding
ammonia molecules. The energy profile is shown in Fig. 9:
The energy profile shown in this figure reveals four stable
structures. The first one, R, can be viewed as the solvation of
CO2 in an ammonia solvent. It is characterised by a CN bond of
2.8 Å, associated with a frequency of 119 cmꢀ1. This system haꢀs
to overcome a barrier of 6.3 kJ molꢀ1 to produce the NH3+CO2
zwitterion, Z, surrounded by six NH3 molecules. The transfer of
a proton from the zwitterion to a neighbouring ammonia
molecule leads to the formation of the carbamate anion C
(NH2CO2ꢀ) stabilised by the ammonium cation NH4+. This
transformation involves a barrier that is 11 kJ molꢀ1 above R.
Such a small barrier is consistent with the 5.1 kJ molꢀ1 barrier
found experimentally. Structure C is very stable with respect to
R, i.e. 42 kJ molꢀ1 below. One further hydrogen transfer can
transform this carbamate anion into carbamic acid, A. This
transformation involves barriers of 13 kJ molꢀ1 below R and
29 kJ molꢀ1 above C.
3.2.2 Reaction profile of the 1 : 1 : 6 cluster. The reaction of
NH3 and CO2 is investigated in the presence of six surrounding
water molecules. The energy profile is shown in Fig. 10:
Fig. 8 Mass spectrum of ammonium carbamate (NH4+ : NH2CH2COOꢀ)
at T = 240 K (top) and of carbamic acid (NH2COOH) at T = 265 K (bottom)
obtained using a QMS spectrometer with a 70 eV electron impact source.
23610 | Phys. Chem. Chem. Phys., 2014, 16, 23604--23615
This journal is © the Owner Societies 2014

View Article Online
Paper
PCCP
Fig. 9 B3LYP/6-311G(d,p) lowest energy profile for the formation of ammonium carbamate and carbamic acid in the 7 : 1 : 0 cluster. Energies are given in
kJ molꢀ1 and are corrected for unscaled zero-point energies. Green = carbon, red = oxygen, and white = hydrogen. The arrows indicate the atomic
motion during the reaction. R, Z, C, and A stand for reactants, zwitterion, carbamate anion, and carbamic acid structures, respectively. All relative energies
are given with respect to R.
The energy profile shown in Fig. 10 reveals three stable surrounded by six H2O molecules. A water-mediated proton
structures. As in the previous 7 : 1 : 0 structure, R can be viewed transfer involving a high barrier of 55 kJ molꢀ1 above R leads to
as CO2 and NH3 solvated in water. The CN bond of 2.6 Å, the formation of carbamic acid. Such a high barrier for the
associated with a frequency of 160 cmꢀ1, is also consistent with formation of NH2COOH is certainly large, but this formation
the one obtained for R in the 7 : 1 : 0 structure. ꢀThe second pathway will dominate if H2O is more abundant than NH3, as
stable structure Z corresponds to the NH3+CO2 zwitterion long as the temperature is high enough.
Fig. 10 B3LYP/6-311G(d,p) lowest energy profile for the formation of carbamic acid in the 1 : 1 : 6 cluster. Energies are given in kJ molꢀ1 and are
corrected for unscaled zero-point energies. Green = carbon, red = oxygen, and white = hydrogen. The arrows indicate the atomic motion during the
reaction. R, Z, and A stand for reactants, zwitterion, and carbamic acid structures, respectively. All relative energies are given with respect to R.
This journal is ©the Owner Societies 2014
Phys. Chem. Chem. Phys., 2014, 16, 23604--23615 | 23611

View Article Online
PCCP
Paper
Fig. 11 B3LYP/6-311G(d,p) lowest energy profile for the formation of the carbamate anion and carbamic acid in the 2 : 1 : 5 cluster. Energies are given in
kJ molꢀ1 and are corrected for unscaled zero-point energies. Green = carbon, red = oxygen, and white = hydrogen. The arrows indicate the atomic
motion during the reaction. R, Z1, Z2, C and A stand for reactants, zwitterions 1 and 2, carbamate anion, and carbamic acid structures, respectively. All
relative energies are given with respect to R.
3.2.3 Reaction profile of the 2 : 1 : 5 cluster. The reaction of without going through the formation of a carbamate anion.
NH3 and CO2 is investigated in the presence of five water This transformation does, however, involve a high barrier of
molecules and one NH3 molecule. The energy profile is shown 39 kJ molꢀ1 with respect to R. The barrier involved in the
in Fig. 11.
formation of A2 is much higher than the one involved in the
As for the previous energy profiles, structure R can be viewed formation of A1, both structures corresponding to carbamic
as the solvation of CO2 and NH3 in a five water- and one acid but with a different arrangement of the solvent molecules.
ammonia-cluster. It is characterised by a CN distance of 2.6 Å, with The reaction pathway via the carbamate anion structure should
a frequency of 154 cmꢀ1, consistent with the two previous optimised therefore be strongly favoured.
+
ꢀ
R structures. The formation of Z2 from R, a NH3 CO2 zwitterion
surrounded by five water molecules and one ammonia molecule,
involves a very small barrier of 4.6 kJ molꢀ1. A water-mediated
proton transforms this zwitterion into a carbamate anion stabilised
by an NH4+ cation. The barrier for the transformation of Z2 into C is
5 kJ molꢀ1 above R. Structure C is very stable with respect to R, i.e.
31 kJ molꢀ1 below. This 5 kJ molꢀ1 barrier for a mixed NH3 :H2O
solvation is lower than the 11 kJ molꢀ1 barrier for a pure NH3
solvation. The formation of the carbamate anion seems to be better
catalysed by water than ammonia, if water and ammonia are both
present. This is due to higher efficiency of proton transfers. This
phenomenon was not observed in the experiments in Section 3.1.3,
but is clearly evident from the theoretical calculations.
4 Discussion
On the basis of laboratory experiments and ab initio calcula-
+
tions we discuss the NH2COOH and NH4 NH2COOꢀ formation
mechanisms from a low temperature solid-phase NH3 and
CO2 ice mixture. The mechanisms involved in the formation
of the carbamate anion and carbamic acid are concentration
dependent.
In a CO2-dominated ice, the proton transfer cannot take
place and no reactivity occurs.
In a protic environment, i.e. a NH3- or H2O-dominated ice,
the less energetic pathway is the formation of the carbamate
A water-mediated proton transfer can allow for the transfor-
mation of this carbamate anion into the carbamic acid, A1. The
transition state involved in the C - A1 transformation is lower
in energy with respect to R by 23 kJ molꢀ1. This transition state
is 8 kJ molꢀ1 above C.
+
NH4 NH2COOꢀ, via the formation of a zwitterion (scenario 2).
The carbamate then transforms into carbamic acid, according
to the following reactions:
+
NH3 + CO2 " NH3 COOꢀ
An alternative pathway has been investigated for another
zwitterionic structure, Z1, although Z1 being close enough
to Z2 a simple relaxation can transform one into another.
We found this zwitterion to transform into carbamic acid, A2,
NH3 + NH3 COOꢀ " NH4 NH2COOꢀ
+
+
+
NH4 NH2COOꢀ " NH3 + NH2COOH.
23612 | Phys. Chem. Chem. Phys., 2014, 16, 23604--23615
This journal is © the Owner Societies 2014

View Article Online
Paper
PCCP
The formation of carbamic acid involves lower energy barriers water ice, which has a desorption energy of 46.6 kJ molꢀ1
when formed from the carbamate anion. The direct formation (ref. 47)). This implies that some NH4+NH2COOꢀ can stay
of carbamic acid is possible but energetically unfavourable. on the substrate (silicate, carbon, aerosol) surface at high
However, in a H2O dominated ice, where there is a shortage of temperature when water ice has sublimated. Such refractory
NH3 reactants, this energetic pathway is the only one available. species enter the composition of a residue, where a water-
Calculations show that carbamate seems to be better catalysed free chemistry can take place at higher temperature than the
by water than by ammonia, although such a difference was not water ice sublimation temperature. For example, the reaction
observed experimentally in kinetic experiments.
between amines and CO2 produces alkyl-ammonium alkyl-
In a NH3-dominated ice, the kinetics of the reaction are carbamates which, under (solar or interstellar) ultraviolet
partial first-order with respect to CO2 and partial zero-order with irradiation, can lead to the formation of amino acid salts, such
respect to NH3 (for a NH3 molar fraction between 0.5 and 1). The as ethyl-ammonium glycinate.34
kinetic equation for the reaction can therefore be written as:
More generally, carbamates are CO2 reservoirs. They can store
CO2, which can be released at higher temperatures when the
carbamate sublimates, as seen in Fig. 7. The rate of carbamate
formation is important when calculating the overall rate of CO2
removal. Carbamates can also be reaction intermediates in the
synthesis of urea from NH3 and CO2 carried out in industrial
plants.48 Industrial research is presently carried out on methods
of CO2 removal in aqueous ammonia solutions.49 At middle
latitudes, the tropopause temperature ranges from an average
of 15 1C at the sea level to about ꢀ55 1C at the top of the
dðNH4þNH2COOꢀÞ
dt
dðCO2Þ
dt
¼ ꢀ
¼ kðTÞ ꢂ ðCO2Þ
(17)
The temperature dependence of the first-order rate constant
is experimentally determined to follow the Arrhenius law, k(T) =
+1.1
0.09
sꢀ1 ꢂ exp(5.1 ꢁ 1.6 kJ molꢀ1/RT), in the 70 K to
ꢀ0.08
90 K range, when the reactants are located on two neighbouring
sites, i.e. for a NH3 : H2O ratio greater than 1 : 1.
The determination of the reaction barrier of the two reactants
located on two neighboring sites is very important for later studies
of diffusion-limited reactions. Indeed, solid-phase reactions in
water-dominated ice, in either interstellar or atmospheric environ-
ments, are diffusion-limited.42 The reaction will involve reactants
either initially located on neighbouring sites (for a NH3 :H2O ratio
greater than 1) or brought to neighbouring sites by diffusion
(NH3 : H2O ratio lower than 1). The chemical reaction kinetics
must be coupled with the diffusion kinetics of the reactants in
order to fully describe such diffusion-limited reaction kinetics.
As shown in Fig. 12, the necessary diffusion of the reactants
in H2O-dominated ice ((NH3 : H2O ratio lower than 1) slows
down the kinetics of the NH3 and CO2 solid-state reaction.
Very few studies exist on surface and volume diffusion in either
crystalline43,44 or amorphous ice.45,46 Such studies on the diffu-
sion of molecules in ice are a fundamental step toward the
understanding of the kinetics of diffusion-limited reactions in
ice. The coupling between solid-phase reactivity and diffusion in
ice is therefore far from being understood.
+
tropopause. The NH3 to NH4 conversion rates have been
estimated between 10ꢀ3 sꢀ1 (ref. 50) and 5 ꢂ 10ꢀ5 sꢀ1 51
.
Upon
extrapolating the rate we have measured to these temperatures,
we find reaction rates between 0.01 sꢀ1 and 5 ꢂ 10ꢀ3 sꢀ1, with an
average of 7 ꢂ 10ꢀ3 sꢀ1. This extrapolation is tentative since the
temperature ranges are different. However, as long as diffusion
is not involved, the amorphous or crystalline state of the ice
should not be an issue. Our results are slightly higher than
the previous estimates of the rate of removal of atmospheric gas-
+
phase NH3 and its transformation into (NH4 ) salt aerosols. This
can be explained by the effect of the diffusion on the overall
reaction rate, as shown in Fig. 12.
5 Conclusion
We studied, both theoretically and experimentally, the low
temperature solid-phase reactivity of the CO2 : NH3 system.
While NH3 and CO2 do not react in the gas-phase, we have
shown that once in the solid-phase, the molecular environment
can lower the reaction barrier, enabling the reaction between
frozen CO2 and NH3 to occur. In a CO2 dominated environment,
the reaction is not possible. In a hydrogen bonded environment,
such as in the presence of NH3 or H2O, the reaction barrier
is lowered and the reaction is possible. The reaction produces
both carbamic acid (NH2COOH) and ammonium carbamate
We have also shown that NH4+NH2COOꢀ has a zeroth-order
desorption rate constant, kdes(T) = 1013 sꢀ1 ꢂ exp(69.0 ꢁ
0.2 kJ molꢀ1/(R ꢂ T)), and is therefore more refractory than
+
+
(NH4 NH2COOꢀ). However, the formation of NH4 NH2COOꢀ
precedes the formation of NH2COOH. The reaction can occur
at temperatures below the desorption of the reactants, and their
kinetic parameters can be measured using isothermal kinetic
experiments and temperature programmed reactivity experi-
ments on CO2 : NH3 ice films. We concluded that the reaction
Fig. 12 Isothermal decay at 80 K of CO2 in a CO2 : NH3 ice film (dia-
monds) and in a CO2 : NH3 : H2O ice film where H2O is in excess with
respect to NH3 (triangles). The need for reactants to diffuse before reacting
dramatically slows down the reaction.
rate constant follows the Arrhenius law k(T) = 0.09
+1.1 sꢀ1
ꢂ
ꢀ0.08
exp(5.1 ꢁ 1.6 kJ molꢀ1/(RT)) between 70 K and 90 K. The product
of this reaction, ammonium carbamate (NH4 NH2COOꢀ), is
+
This journal is ©the Owner Societies 2014
Phys. Chem. Chem. Phys., 2014, 16, 23604--23615 | 23613

View Article Online
PCCP
Paper
more refractory than water ice and has 69.0 ꢁ 0.2 kJ molꢀ1 first-
order desorption energy for a 1013 sꢀ1 pre-exponential factor.
Carbamic acid (NH2COOH) has 76.1 ꢁ 0.1 kJ molꢀ1 first-order
desorption energy for a 1013 sꢀ1 pre-exponential factor.
Real ices are water dominated, and thus solid-phase thermal
reactions are diffusion-limited. The next steps are therefore to
investigate the kinetics of the diffusion of each reactant and the
kinetics of the reactivity of the CO2 : NH3 : H2O mixture where
H2O is dominant. Understanding how the reactivity is limited
by the diffusion of the reactants is the key to understand
ice chemistry.
16 E. Dartois, Space Sci. Rev., 2005, 119, 293–310.
17 D. L. Frasco, J. Chem. Phys., 1964, 41, 2134–2140.
18 I. Hisatsune, Can. J. Chem., 1984, 62, 945–948.
´
19 J. B. Bossa, P. Theule, F. Duvernay, F. Borget and
T. Chiavassa, Astron. Astrophys., 2008, 492, 719–724.
20 Y. Rodriguez-Lazcano, B. Mate, V. J. Herrero, R. Escribano
and O. Galvez, Phys. Chem. Chem. Phys., 2014, 16, 3371–3380.
21 Y.-J. Chen, M. Nuevo, J.-M. Hsieh, T.-S. Yih, W.-H. Sun,
W.-H. Ip, H.-S. Fung, S.-Y. Chiang, Y.-Y. Lee, J.-M. Chen and
C.-Y. R. Wu, Astron. Astrophys., 2007, 464, 253–257.
22 R. K. Khanna and M. H. Moore, Spectrochim. Acta, Part A,
1999, 55, 961–967.
23 V. Bordalo, E. F. da Silveira, X. Y. Lv, A. Domaracka,
H. Rothard, E. Seperuelo Duarte and P. Boduch, Astrophys.
J., 2013, 774, 105.
24 M. Bertin, I. Martin, F. Duvernay, P. Theule, J. B. Bossa,
F. Borget, E. Illenberger, A. Lafosse, T. Chiavassa and
R. Azria, Phys. Chem. Chem. Phys., 2009, 11, 1838.
25 P. Jackson, A. Beste and M. I. Attalla, Phys. Chem. Chem.
Phys., 2012, 14, 16301–16311.
Acknowledgements
We thank the PCMI (Physique et Chimie du Milieu Interstellaire)
program and the CNES (Centre National d’Etudes Spatiales)
for financial support, and the HPC resources from GENCI-
[CCRT/CINES/IDRIS] (grant 2014 [x2014085116]) for computing
time. J.A.N. is a Royal Commission for the Exhibition of 1851
Research Fellow.
26 P. Theule, F. Duvernay, A. Ilmane, T. Hasegawa, O. Morata,
S. Coussan, G. Danger and T. Chiavassa, Astron. Astrophys.,
2011, 530, A96.
References
27 T. Bartels-Rausch, V. Bergeron, J. H. E. Cartwright,
´
R. Escribano, J. L. Finney, H. Grothe, P. J. Gutierrez,
1 S. Krupa, Environ. Pollut., 2003, 124, 179–221.
2 J. G. J. Olivier, A. F. Bouwman, K. W. Van der Hoek and
J. J. M. Berdowski, Environ. Pollut., 1998, 102, 135–148.
3 J. Bassham, A. Benson and M. Calvin, J. Biol. Chem., 1950,
185, 781–787.
J. Haapala, W. F. Kuhs, J. B. C. Pettersson, S. D. Price,
C. I. Sainz-Diaz, D. J. Stokes, G. Strazzulla, E. S. Thomson,
H. Trinks and N. Uras-Aytemiz, Rev. Mod. Phys., 2012, 84,
885–944.
28 F. Mispelaer, P. Theule, F. Duvernay, P. Roubin and
T. Chiavassa, Astron. Astrophys., 2012, 540, A40.
29 O. Kerkhof, W. A. Schutte and P. Ehrenfreund, Astron.
Astrophys., 1999, 346, 990–994.
4 D. B. Dell Amico, F. Calderazzo, L. Labella, F. Marchetti and
G. Pampaloni, Chem. Rev., 2003, 103, 3857–3897.
5 G. Sartori and D. Savage, Ind. Eng. Chem. Fundam., 1983, 22,
239–249.
30 H. Yamada and W. B. Person, J. Chem. Phys., 1964, 41,
2478–2487.
6 K. Hiraoka, A. Yamashita, Y. Yachi, K. Aruga, T. Sato and
H. Muto, Astrophys. J., 1995, 443, 363–370.
31 P. A. Gerakines, W. A. Schutte, J. M. Greenberg and
E. F. van Dishoeck, Astron. Astrophys., 1995, 296, 810.
32 S. A. Sandford and L. J. Allamandola, Astrophys. J., 1990, 355,
357–372.
33 J. B. Bossa, P. Theule, F. Duvernay and T. Chiavassa,
Astrophys. J., 2009, 707, 1524–1532.
7 Y. Oba, N. Watanabe, A. Kouchi, T. Hama and V. Pirronello,
Astrophys. J., Lett., 2010, 712, L174–L178.
8 S. Ioppolo, Y. van Boheemen, H. M. Cuppen, E. F. van
Dishoeck and H. Linnartz, Mon. Not. R. Astron. Soc., 2011,
413, 2281–2287.
9 J. A. Noble, F. Dulieu, E. Congiu and H. J. Fraser, Astrophys.
J., 2011, 735, 121.
´
34 J.-B. Bossa, F. Duvernay, P. Theule, F. Borget, L. D’Hendecourt
and T. Chiavassa, Astron. Astrophys., 2009, 506, 601–608.
35 S. A. Sandford and L. J. Allamandola, Astrophys. J., 1993, 417,
815–825.
`
10 M. Minissale, E. Congiu, G. Manico, V. Pirronello and
F. Dulieu, Astron. Astrophys., 2013, 559, A49.
11 K. Hiraoka, T. Miyagoshi, T. Takayama, K. Yamamoto and
Y. Kihara, Astrophys. J., 1998, 498, 710.
´
´
36 O. Galvez, I. K. Ortega, B. Mate, M. A. Moreno, B. Martn-
´
Llorente, V. J. Herrero, R. Escribano and P. J. Gutierrez,
Astron. Astrophys., 2007, 472, 691–698.
12 S. Ioppolo, H. M. Cuppen, C. Romanzin, E. F. van Dishoeck
and H. Linnartz, Astrophys. J., 2008, 686, 1474–1479.
13 Y. Oba, N. Watanabe, T. Hama, K. Kuwahata, H. Hidaka and
A. Kouchi, Astrophys. J., 2008, 686, 1474–1479.
14 F. Dulieu, L. Amiaud, E. Congiu, J.-H. Fillion, E. Matar,
A. Momeni, V. Pirronello and J. L. Lemaire, Astron. Astrophys.,
2010, 512, A30.
15 H. Mokrane, H. Chaabouni, M. Accolla, E. Congiu,
F. Dulieu, M. Chehrouri and J. L. Lemaire, Astrophys. J.,
Lett., 2009, 705, L195–L198.
37 A. D. Becke, J. Chem. Phys., 1993, 98, 5648–5652.
38 P. J. Stephens, F. J. Devlin, C. F. Chabalowski and
M. J. Frisch, J. Phys. Chem., 1994, 98, 11623–11627.
39 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone,
B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato,
X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng,
J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda,
23614 | Phys. Chem. Chem. Phys., 2014, 16, 23604--23615
This journal is © the Owner Societies 2014

View Article Online
Paper
PCCP
J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, 44 C. A. Varotsos and R. Zellner, Atmos. Chem. Phys., 2010, 10,
H. Nakai, T. Vreven, J. A. Montgomery Jr., J. E. Peralta, 3099–3105.
F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, 45 F. Mispelaer, P. Theule, H. Aouididi, J. Noble, F. Duvernay,
´
V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari,
A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi,
G. Danger, P. Roubin, O. Morata, T. Hasegawa and
T. Chiavassa, Astron. Astrophys., 2013, 555, A13.
N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, 46 L. J. Karssemeijer, S. Ioppolo, M. C. van Hemert,
V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts,
R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi,
A. van der Avoird, M. A. Allodi, G. A. Blake and
H. M. Cuppen, Astrophys. J., 2014, 781, 16.
C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, 47 H. J. Fraser, M. P. Collings, M. R. S. McCoustra and
V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg,
D. A. Williams, Mon. Not. R. Astron. Soc., 2001, 327,
¨
S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman,
1165–1172.
J. V. Ortiz, J. Cioslowski and D. J. Fox, Gaussian 09 Revision 48 F. Barzagli, F. Mani and M. Peruzzini, Green Chem., 2011,
D.01, Gaussian Inc., Wallingford, CT, 2009.
40 R. L. Brown, J. Res. Natl. Bur. Stand., 1981, 86, 357–359.
41 S. A. Sandford and L. J. Allamandola, Icarus, 1990, 87,
188–192.
13, 1267–1274.
49 Z. Niu, Y. Guo, Q. Zeng and W. Lin, Ind. Eng. Chem. Res.,
2012, 51, 5309–5319.
50 A. W. M. Vermetten, W. A. H. Asman, E. Buijsman,
W. Mulder, J. Slanina and A. Waijers-Ijpelaan, VDI. Ber.,
1985, 560, 241–251.
51 J. W. Erisman, A. W. M. Vermetten, W. A. H. Asman,
A. Waijers-Ijpelaan and J. Slanina, Atmos. Environ., 1988,
22, 1153–1160.
´
42 P. Theule, F. Duvernay, G. Danger, F. Borget, J. B. Bossa,
V. Vinogradoff, F. Mispelaer and T. Chiavassa, Adv. Space
Res., 2013, 52, 1567–1579.
43 F. Livingston, J. Smith and S. George, J. Phys. Chem. A, 2002,
106, 6309–6318.
This journal is ©the Owner Societies 2014
Phys. Chem. Chem. Phys., 2014, 16, 23604--23615 | 23615