Experimental and theoretical study of the reaction of OH radical with sabinene

Article abstract of DOI:10.1039/b604489a

The gas phase sabinene + OH reaction is studied both experimentally and theoretically. Product yields from the reaction of sabinene with OH radicals have been measured in the absence of NO_x in the UCC chamber (Cork, Ireland) and in the presence of NO_x in the LISA chamber. Three primary carbonyl compounds were observed and quantified: acetone in [(24 ± 6)%], formaldehyde in [(25 ± 6)%] and sabinaketone in [(20 ± 6)%]. The simultaneous quantification of these compounds is one of the major results of this work. The mechanism of product formation for this reaction has been studied using the quantum chemical DFT-B3LYP (6-31G(d,p) method. According to these calculations, the H-atom abstraction channel from sabinene by OH in the initial oxidation step may be taken into account to explain the acetone production. Sabinaketone and formaldehyde are mainly products of the addition channels of OH on the -C=CH₂ double bond of sabinene. This is the first theoretical work on the title reaction. the Owner Societies 2006.

Full text of DOI:10.1039/b604489a

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/pccp | Physical Chemistry Chemical Physics
Experimental and theoretical study of the reaction of OH radical with
sabinene
N. Carrasco,a M. T. Rayez,*b J. C. Rayezb and J. F. Doussina
Received 28th March 2006, Accepted 1st June 2006
First published as an Advance Article on the web 13th June 2006
DOI: 10.1039/b604489a
The gas phase sabinene þ OH reaction is studied both experimentally and theoretically. Product
yields from the reaction of sabinene with OH radicals have been measured in the absence of NOx
in the UCC chamber (Cork, Ireland) and in the presence of NOx in the LISA chamber. Three
primary carbonyl compounds were observed and quantified: acetone in [(24 ꢀ 6)%],
formaldehyde in [(25 ꢀ 6)%] and sabinaketone in [(20 ꢀ 6)%]. The simultaneous quantification
of these compounds is one of the major results of this work. The mechanism of product
formation for this reaction has been studied using the quantum chemical DFT-B3LYP
(6-31G(d,p) method. According to these calculations, the H-atom abstraction channel from
sabinene by OH in the initial oxidation step may be taken into account to explain the acetone
production. Sabinaketone and formaldehyde are mainly products of the addition channels of OH
on the –CQCH2 double bond of sabinene. This is the first theoretical work on the title reaction.
0.17. Since sabinaketone is not commercially available, this
Introduction
yield was determined in the absence of an authentic standard,
Large quantities of biogenic volatile organic compounds
thanks to the approximation of the GC-FID response factor
(BVOCs) are emitted by terrestrial and marine sources. The
using the effective carbon number concept.
world-wide emission rate of non-methane biogenic volatile
In this work, the reaction of sabinene with OH radicals has
organic compounds is estimated to be 1150 TgC yearꢁ1
been experimentally studied with and without the presence of
(Guenther et al.1) and of the estimated 84 TgC yearꢁ1 emitted
NOx. In addition, sabinaketone has been synthesized and used
in the US, 20% are monoterpene compounds (C10H16).2
to quantify its yield from the reaction sabinene þ OH. In the
Sabinene is a monoterpene of which the reactivity has been
second part of this work, a theoretical study of the mechan-
rarely studied. There is only one piece of kinetic data available
isms of the oxidation of sabinene by OH is performed,
about its reaction with OH radicals3 (k = 1.17 ꢂ 10ꢁ10
explaining the formation of the primary products observed
moleculeꢁ1 cm3
s
ꢁ1). However, this compound has been
experimentally.
reported to be abundantly emitted by several common tree
species in Europe. Thus, Tollsten and Muller,4 Owen et al.,5
¨
Street et al.6 and Hakola et al.7 have identified sabinene as a
major emission from Fagus sylvatica, Quercus ilex and Betula
pendula. Therefore, its high reactivity with OH (lifetime 1.5 h
for an average OH radical concentration of 1.5 ꢂ 106 molecule
cmꢁ3, ref. 3) and its potentially very abundant emissions
suggest that sabinene is likely to play a significant role in the
chemistry of the rural troposphere.
Experimental
Smog chambers
Experiments were carried out using two different atmospheric
simulation chambers, at LISA (Creteil, France) and at Uni-
´
versity College Cork (UCC, Ireland). Both chambers have
been described in detail elsewhere (Doussin et al.,13 Thuener
et al.14) and only brief details are provided here.
The mechanism of the gas phase degradation of sabinene
with OH radicals has been studied previously by Reissel et al.,8
Hakola et al.9 and Arey et al.10 Reissel et al.8 have observed
the production of acetone with a yield of 0.19 ꢀ 0.03. This
result is of atmospheric importance since it has been shown
that the photolysis of acetone is a significant source of HOx
The LISA chamber is a 977 L cylindrical Pyrex evacuable
reactor, equipped with a multiple-reflection optical system
coupled to a Bomem DA8-ME FTIR spectrometer.13 Between
each experiment, the simulation chamber was conditioned by
lowering the pressure to 10ꢁ3 mbar. Then, the chamber was
filled with synthetic air (80% N2, 20% O2) at atmospheric
pressure and the temperature was kept at 293 ꢀ 2 K during all
the experiments. The reactor is surrounded by 40 visible lamps
(Philips TL03 40W, lmax = 420 nm).
The hydroxyl radical reactions were performed using the
photolysis (at 420 nm) of HONO. The initial reactant con-
centrations (in molecule cmꢁ3) were as follows; [HONO] =
25 ꢂ 1013, [sabinene] = 5 ꢂ 1013. During the reactions, the
contents of the chamber were continuously monitored by
radicals in the upper troposphere (Jaegle
´
et al.,11 Jacob
et al.12). In addition, Hakola et al.9 and Arey et al.10 have
identified another major product, sabinaketone with a yield of
a
`
Laboratoire Interuniversitaire des Systemes Atmospheriques, UMR
7583 CNRS/Universite Paris 7/Universite Paris, 1261 av du General
´
´
´
´ ´
´
de Gaulle, 94010 Creteil Cedex, France
Laboratoire de Physicochimie Moleculaire, UMR 5803 CNRS/
´
b
´
´
Universite Bordeaux 1, 33405 Talence Cedex, France. E-mail:
mt.rayez@lpcm.u-bordeaux1.fr; Fax: (33) 5 40 00 66 45
ꢃc
This journal is the Owner Societies 2006
Phys. Chem. Chem. Phys., 2006, 8, 3211–3217 | 3211

View Article Online
FTIR spectroscopy using a path length of 156 m. Infrared
spectra were obtained at a resolution of 0.5 cmꢁ1 using a MCT
detector and derived from the co-addition of 200 scans
collected over 5 min.
The chamber at UCC is a cylinder made of FEP (fluori-
ne–ethene–propene) foil (4.1 m long, 1.1 m diameter and 0.127
mm thickness) with a volume of 3.91 m3. It is operated at 295 ꢀ
2 K using purified dry air at 0.1–1 mbar above atmospheric
pressure. The chamber is equipped with a multiple reflection
optical arrangement coupled to an FTIR spectrometer (BioRad
Excalibur) for chemical analysis by in situ FTIR spectroscopy.
The chamber is surrounded by 18 Philips TUV (40 W) lamps
with an emission maximum at 254 nm and 18 Philips TL05
(40 W) lamps with an emission maximum at 360 nm. Between
experiments the chamber is cleaned by flushing with the purified
air at a flow rate of 0.15 m3 minꢁ1 for a minimum of 4 h.
The hydroxyl radical reactions were performed in the
absence of NO using the photolysis of hydrogen peroxide
(by the Philips TUV lamps) as the radical source. The initial
reactant concentrations (in molecule cmꢁ3) were as follows;
[H2O2] = 2.5 ꢂ 1013, [sabinene] = 5.0 ꢂ 1013. During the
reactions, the contents of the chamber were continuously
monitored by FTIR spectroscopy using a path length of
(229.6 ꢀ 0.6) m. Infrared spectra were obtained at a resolution
of 1 cmꢁ1 using a narrow-band MCT detector and derived
from the co-addition of 200 scans collected over 4 min.
Fig. 1 Formation of acetone, formaldehyde and sabinaketone during
a reaction between sabinene and OH (19/11/03 CRAC).
Three primary carbonyl compounds have been observed and
quantified, acetone, formaldehyde and sabinaketone. Forma-
tion yields were determined by calculating the initial slopes of
the plots (see Fig. 1). Uncertainties take in account twice the
standard deviation and uncertainties on calibrations. A typical
plot of [product] vs. ꢁD[sabinene] is given in Fig. 1. The
linearity of the plot indicates that the three carbonyl com-
pounds are primary products. As can be seen in Fig. 1, these
yield plots are fairly straight lines indicating that the secondary
chemistry was negligible under our conditions. Consequently,
no correction for secondary reaction was made. Indeed, based
on the literature data for OH þ sabinene and OH þ sabina-
ketone, HCHO and acetone, the maximum corrections of the
measured concentrations for secondary reactions with OH
radicals were less than 2% at the highest extent of reaction
(about 30% based on the initial sabinene concentration—see
Fig. 1), and were neglected. No noticeable difference has been
observed between both NOx conditions. Thus production
yields are averaged and compared to previous literature data
in Table 2.
Hakola et al.9 and Arey et al.10 have observed a sabinake-
tone yield slightly lower than our determination. Nevertheless
the three studies are in good accordance considering the
overall combined uncertainties. The major outcome of this
part of the work was to detect quantitatively sabinaketone by
FTIR after its preliminary synthesis and calibration. Indeed,
this compound is not commercially available and had been
indirectly identified and quantified by Arey et al.10 and Hakola
et al.,9 in GC-MS and GC-FID, respectively, assuming a
sabinaketone response factor equal to that of a similar com-
pound. Only one previous study (Librando et al.17) reported
formaldehyde production yield from the sabinene reaction
with OH radicals. They found a higher yield than the one
presented in this work. From the experimental details given
such a discrepancy is difficult to interpret as it is not clear if
secondary chemistry occurred in the Librando et al.17 study. It
must also be pointed out that they reported chemical instabi-
lity and/or photolysis of sabinene under their experimental
conditions, while the tests we performed did not show any of
these behaviours. The additional formaldehyde could possibly
arise from these pathways.
FTIR spectroscopy
Quantitative analysis was performed by subtraction of cali-
brated reference spectra of known compounds and subsequent
integration of selected absorption bands for sabinene and the
reaction products. The integrated band intensities (IBI) of
spectral regions used during the subtraction procedures are
given in Table 1.
Chemicals. Chemicals were obtained from the following
sources and used without further purification: sabinene (Inter-
chim, >99%), H2O2 (Aldrich, 30%). HONO was synthesised
from the dropwise addition of sulfuric acid to a diluted
NaNO2 solution (Taylor et al.15). Sabinaketone was pre-
viously synthesized by sabinene ozonolysis in the liquid
phase (GCMS, >95%1). The method has been described in
Carrasco et al.16
Results
Product yields from the reaction of sabinene with OH radicals
have been measured in the absence of NOx in the UCC
chamber and in the presence of NOx in the LISA chamber.
Table 1 IBI of the main IR absorption bands, values are given in loge
Main
absorption
Compound
band/cmꢁ1 IBI/cm moleculeꢁ1
Ref.
Sabinene
Acetone
2800–3110 (5.44 ꢀ 0.15) ꢂ 10ꢁ17 This work
1260–1150 (1.02 ꢀ 0.02) ꢂ 10ꢁ17 Picquet-Varrault
et al.32
Formaldehyde 1845–1625 (1.25 ꢀ 0.13) ꢂ 10ꢁ17 Klotz et al.
33
Sabinaketone 1700–1800 (3.7 ꢀ 0.4) ꢂ 10ꢁ17
This work
ꢃc
This journal is the Owner Societies 2006
3212 | Phys. Chem. Chem. Phys., 2006, 8, 3211–3217

View Article Online
Table 2 Production yields observed in this work and compared with previous literature data
Product
Acetone
Formaldehyde
Sabinaketone
Carbon Balance
181103 CRAC (no NOx)
191103 CRAC (no NOx)
020304 LISA (with NOx)
030304 LISA (with NOx)
Averaged yields (this work)
Hakola et al.9
0.22 ꢀ 0.05
0.27 ꢀ 0.05
0.25 ꢀ 0.05
0.21 ꢀ 0.05
0.24 ꢀ 0.06
ND
0.27 ꢀ 0.05
0.23 ꢀ 0.05
0.28 ꢀ 0.06
0.22 ꢀ 0.04
0.25 ꢀ 0.06
ND
ND
ND
0.61
0.23 ꢀ 0.06
0.21 ꢀ 0.06
0.19 ꢀ 0.05
0.19 ꢀ 0.05
0.20 ꢀ 0.06
0.17 ꢀ 0.03
0.17
0.30 ꢀ 0.07
0.30 ꢀ 0.07
0.28 ꢀ 0.07
0.26 ꢀ 0.06
0.28 ꢀ 0.10
0.15 ꢀ 0.03
0.15
Arey et al.10
ND
0.19 ꢀ 0.03
Reissell et al.8
ND
ND
0.06 ꢀ 0.01
Librando et al.17
ND
ND
ND: Not determined
Acetone yield has been previously studied by Reissell et al.,8
leading to a value Racetone = 0.19 ꢀ 0.03, which is slightly
lower than our determination of Racetone = 0.24 ꢀ 0.06 but in
agreement within the error bars. It must be pointed out that
both studies found a primary production of acetone. This
shows that acetone is directly arising from the evolution of the
radical produced from the initial attack of sabinene and not
from the further oxidation of intermediary species. Consider-
ing the chemical structure of sabinene, this direct production is
not trivial.
tational time. For the H-abstraction reactions, calculations
have also been performed at the ab initio MP2/6-31G(d,p)
level of theory. All energies presented in the text and figures
are corrected by zero-point energies. No scaling factors have
been applied to the frequencies. All these quantum chemical
calculations have been performed using the Gaussian 0324
package. The goal of this study is not quantitative but
qualitative in order to explain the experimentally observed
products.
Geometry of sabinene. The geometric structure of sabinene
corresponds to a bicyclic structure, in which a 3-membered
and 5-membered ring have a single C–C bond in common. The
dihedral angle between these two rings is around 901.
Consequently, our study is more exhaustive than previous
studies that were focused on the production of one specific
compound, acetone or sabinaketone. Nevertheless the carbon
balance taking into account the three carbonyl compounds is
only equal to 0.28 ꢀ 0.10 and is still highly incomplete.
Librando et al.17 reported a 300% molar yield for CO2 and
a 250% molar yield for CO. It is not clear from this article if
these yields are primary or secondary but on a carbon balance
basis it would represent around 52%. Under our conditions it
was not possible to precisely measure these species, because of
some strong interferences of the IR absorption of water and
already present CO2. Consequently, it was not possible to
confirm these results but such a huge CO and CO2 production
would significantly increase our carbon balance.
Initial reactions of sabinene. The oxidation of the sabinene
by OH radical can proceed through two main channels: the
addition on the double bond CQC of sabinene or the direct
abstraction of a hydrogen atom from sabinene by OH.25,26
These pathways are complex and involve many intermediary
species.
1. Addition of OH to the CQC double bond of sabinene and
subsequent reactions
The addition of the OH radical to the double bond of sabinene
leads to the adduct ADDa, a primary radical or ADDb, a
tertiary radical, depending on the addition site Ca or Cb (see
the right-hand side of Fig. 2). Our calculations show that this
exothermic addition reaction is barrierless (see the Appendix
for discussion about the addition barrier) and largely exother-
mic by almost 27 kcal molꢁ1 for the formation of the
OH–adduct isomers ADDa and 30 kcal molꢁ1 for the
OH–adduct ADDb.
Theoretical study
In parallel with the experimental study, we have investigated
the potential energy surface (PES) for the OH þ sabinene
reaction in order to propose a mechanism which can explain
the formation of the products experimentally observed. To our
knowledge, no theoretical study has been done so far on the
mechanism of the oxidation of sabinene by OH radical. In this
work, the geometries of the minima and saddle points were
fully optimized for the intermediates of the title reaction using
quantum chemistry methods. The saddle points are character-
ized by the existence of only one negative eigenvalue of the
Hessian matrix corresponding to an imaginary frequency in
the normal mode analysis. We have considered the transition
states to be located at the saddle points along the reaction
paths. The calculations have been performed at the DFT-
B3LYP/6-31G(d,p) level of theory,18 a method which has
already been validated in earlier studies on open-shell sys-
tems.19–23 As a matter of fact, the overall reaction studied is a
doublet open-shell system. For these large systems, this DFT
method is a good compromise between reliability and compu-
Both radicals ADDa and ADDb can promptly react with
O2 to form, the peroxy radicals ADDaO2 and ADDbO2,
respectively. Then, these peroxy radicals react with NO,
yielding ADDa(b)O alkoxy radicals. As shown in recent
publications,26,27 the alkoxy radicals RO* (R = ADDa or
ADDb) formed are chemically activated, and can undergo
thermalisation or unimolecular reactions, dissociation or
isomerisation.
Fate of ADDaO alkoxy radical. As shown in Fig. 3, two
unimolecular reaction pathways can possibly occur to AD-
DaO. b-C–C bond breaking, leading to the hydroxy-sabinenyl
radical and formaldehyde through a barrier (TSD1) which lies
5.5 kcal molꢁ1 above the energy of ADDaO. This low value of
ꢃc
This journal is the Owner Societies 2006
Phys. Chem. Chem. Phys., 2006, 8, 3211–3217 | 3213

View Article Online
Fig. 4 Molecular structure of the transition state TSI1 showing a 1,6-
H atom transfer within the ADDaO carbon structure. The hydrogen
atoms are not shown, except for the one which is migrating.
Fig. 2 Schematic molecular structures and energetics for the OH þ
sabinene initial addition and abstraction reactions from DFT-B3LYP/
6-31G(d,p) calculations. The energies are given in kcal molꢁ1 and
include zero-point energy corrections.
Since, in the presence of NO, the formation of the chemi-
cally activated nascent ADDaO* (from DHIR, see Fig. 3) and
its subsequent unimolecular reactions could be the first and
key steps in the mechanism leading to acetone formation, we
have performed statistical kinetic calculations to estimate the
probability of isomerisation of the alkoxy radical ADDaO*.
These calculations have been performed using the RRKM
master equation analysis described in ref. 28 where tunnelling
has been taken into account. The results show that isomerisa-
tion of ADDaO* through TSI1 is unimportant compared to
the dissociation channel leading to sabinaketone. As a con-
sequence, the yield of acetone formed by this pathway is
predicted to be unimportant.
the transition state energy is in good agreement with available
experimental data for b-hydroxy-alkoxy radicals similar to
this one and calculations performed at the same level of theory
as in this work.19–22 After reaction with O2, the hydroxy-
sabinenyl radical formed will lead to sabinaketone. The other
reaction pathway is isomerisation of ADDaO involving a
1,6-H shift of the isopropylic hydrogen atom to the terminal
oxygen site of the radical alkoxy, forming the di-hydroxy
isopropylic adduct radical via a transition state TSI1. The
geometry of this transition state is presented in Fig. 4. For a
better understanding, the hydrogen atoms are not shown,
except for the specific H-atom which is migrating. It can be
seen that TSI1 exhibits a 7-membered ring with Oꢄ ꢄ ꢄH and
Cꢄ ꢄ ꢄH distances of 1.35 and 1.40 A, respectively. This hydro-
gen shift is associated with a 20 kcal molꢁ1 barrier height
(TSI1) due in part to the constraint of three of the carbon
atoms which are common to this ring and the 5-membered ring
of the sabinene structure. The large difference between the
heights of the two barriers TSI1 and TSD1 (Fig. 3) suggests
that the formation fo the di-hydroxy isopropylic alkyl radical
(DHIR) will have a small probability. If formed, this latter
radical can then react with O2 and NO to give the di-hydroxy
isopropylic alkoxy radical which, in turn, undergoes dissocia-
tion via a barrier of 14 kcal molꢁ1 height leading finally to
acetone.
Fate of ADDbO alkoxy radical. ADDbO can undergo three
dissociation channels as shown in Fig. 5. The first one is a
b-C–C dissociation yielding sabinaketone and formaldehyde
through a 6.7 kcal molꢁ1 barrier height (TSD2). The two other
pathways involve the opening of the five-membered ring by
cleavage of each of the C–C bonds in the a position to the C–O
bond. These latter reaction channels have much higher bar-
riers 12.4 (TSC1) and 18.7 kcal molꢁ1 (TSC2) than the one
corresponding to ADDbO dissociation (TSD2). Moreover,
they do not play any role in acetone, sabinaketone or
formaldehyde formation.
In summary, the two OH-addition channels to sabinene will
lead quite easily to sabinaketone and formaldehyde formation,
Fig. 3 Energetic diagram of the subsequent reactions of ADDaO
from DFT-B3LYP/6-31G(d,p) calculations. The energies are given in
kcal molꢁ1 and include zero-point energy corrections.
Fig. 5 Energetic diagram of the subsequent reactions of ADDbO
from DFT-B3LYP/6-31G(d,p) calculations. The energies are given in
kcal molꢁ1 and include zero-point energy corrections.
ꢃc
This journal is the Owner Societies 2006
3214 | Phys. Chem. Chem. Phys., 2006, 8, 3211–3217

View Article Online
but to a low percentage of acetone, not representative of the
experimental results which show a primary formation of 20%.
2. Direct H-atom abstraction mechanism
A possibility which could be important in the primary
formation of acetone is the direct abstraction by OH of the
isopropylic hydrogen atom of sabinene. In their paper, Kwok
and Atkinson25 consider the H-abstraction pathway to be
minor when compared to OH-addition, but some measure-
ments have shown that in some cases, it can represent up to
30% of the total rate constant.26 We have therefore explored
the PES associated with this possible mechanism. Among the
hydrogen atoms of sabinene, the isopropylic one is the easiest
to abstract, since it is linked to a tertiary carbon atom (Cc)
(Fig. 2, left). The exothermicity of this abstraction pathway is
ꢁ21 kcal molꢁ1 including the zero-point vibrational (ZPE)
correction. An early barrier has been characterized for which
the HOꢄ ꢄ ꢄH distance is equal to 1.81 A and the C–H bond has
lengthened by only 0.05 A. This transition state lies below the
energy of the reactants by 2 kcal molꢁ1 using the B3LYP/6-
31G(d,p) level of theory. Skokov and Wheeler30 have shown
that B3LYP/6-31G(d) values are, on average, within a range
of 1.6 kcal molꢁ1 with respect to the experimental energies at
most for a set of 6 abstraction reactions and that increasing
basis set does not provide better transition state geometries
and barriers heights. Another study by Basch and Hoz31 also
note a good performance for DFT-B3LYP barriers with
respect to experiment and high level ab initio calculations.
We have performed calculations at the ab initio PMP2/6-
31G(d,p) level of theory and found a barrier of 0.8 kcal molꢁ1
including ZPE correction. More sophisticated calculations
are difficult to perform on this system. Moreover, we cannot
a priori calculate barrier heights with a sufficient accuracy to
allow the prediction of the relative percentage of abstraction
with respect to addition. Nevertheless, these low values and
the presence of acetone in the products of the experiment lead
us to admit that this abstraction reaction can take place. The
radical ‘‘sabinenyl’’ called ABS, a product of this abstraction,
can undergo addition with molecular oxygen O2 and the
resulting peroxyl ABSO2 will react with NO to give ABSO*
and NO2. These steps are described in detail in Fig. 6. The
chemically activated nascent ABSO* then undergoes a b-C–C
bond breakage leading to acetone. The transition state for
this decomposition lies 14.9 kcal molꢁ1 above ABSO (includ-
ing the zero-point vibrational correction). This result agrees
well with the structure–activity relationship that we proposed
in a previous paper20 for activation energies of dissociation
reactions of alkoxy radicals. From these results, we can
predict that the abstraction channel has to be taken into
consideration to explain the acetone formation observed
experimentally.
Fig. 6 Energetic diagram of the formation of acetone from the alkyl
radical ADD, product of the isopropylic H-abstraction from the
sabinene by the OH radical. The energies are given in kcal molꢁ1
and include zero-point energy corrections.
and peroxynitrites:
ꢅ
ꢅ
In the absence of NO, self reactions RO2 þ RO2 lead to
(not activated) alkoxy radicals
The limiting process for yielding acetone in the addition
process is the isomerisation transition state (TSI1). RRKM-
ME calculations have revealed that, even if the alkoxy radical
ADDaO is chemically activated, TSI1 is too tight and its
barrier too high to allow the alkoxy radical to isomerize and
therefore yield acetone. Alternatively, both alkoxy radicals
ADDa(b)O will easily decompose to give sabinaketone and
formaldehyde in the same yield. Alternatively, the alkoxy
radical formed from isopropylic H-atom abstraction from
sabinene can decompose easily to give acetone (with or with-
out NO). We can conclude that, in this specific problem, the
way by which the alkoxy radicals are formed does not change
their reactivity. Some hydroxyalkoxynitrates will also be
formed, but in a small quantity.
Conclusion
In this work, parallel experimental and theoretical studies have
been undertaken in order to characterize products and pro-
pose a degradation mechanism for the gas phase oxidation of
sabinene by hydroxyl radicals. Three products have been
experimentally characterized and quantified: sabinaketone
[(20 ꢀ 6)%], formaldehyde [(25 ꢀ 6)%] and acetone [(24 ꢀ
6)%]. The simultaneous quantification of these compounds is
one of the major results of this work. Nevertheless the carbon
balance taking into account the three carbonyl compounds is
only equal to 0.28 ꢀ 0.10 and is still highly incomplete.
Considering the high reactivity of sabinene toward OH3
Discussion of the reaction mechanism with and without
NO27–29. As a result of our calculations, sabinaketone is
mainly formed after OH addition to both sides of the CQC
bond of sabinene, involving formation of peroxyꢅ radical
intermediates RO2ꢅ. In the presence of NO, the RO2 þ NO
reactions lead to vibrationally activated alkoxy radicals ROꢅ*
ꢃc
This journal is the Owner Societies 2006
Phys. Chem. Chem. Phys., 2006, 8, 3211–3217 | 3215

View Article Online
(kOH = 1.1 ꢂ 10ꢁ10 cm3 moleculeꢁ1
s
ꢁ1), it is very unlikely
position with respect to COꢅ and give sabinaketone and
formaldehyde (OH addition on sabinene) and acetone
(H-atom abstraction from sabinene).
that the H-atom abstraction from secondary and primary
carbons would compete with the considered pathways leading
to significant quantities of undetected products. No informa-
tion on the secondary organic aerosol production was avail-
able during this study, but the reactions between terpenes and
OH radicals are generally considered34 to be weak aerosol
generation pathways. Hence it is also unlikely that the missing
carbon would be in the particulate phase. In addition, since
one can not observe any significant differences between the
results in the presence and in the absence of NOx, it is also
unlikely that a huge hydroxynitrate production could account
for the missing pathways.
This work illustrates the usefulness of quantum calculations
in the understanding of the atmospheric chemistry of biogenic
compounds. Indeed these species exhibit unusual chemical
reactivity most of the time because of their complex structures
for which structure activity relationships are often less effi-
cient.
Appendix
Consequence of the long range potential between OH radical
and an ethylenic double bond
Considering that (i) no other major chemical functions can
be detected from our residual experimental FTIR spectra, (ii)
we were not able to precisely quantify CO and CO2 and (iii) a
previous study17 has proposed a yield of 250 and 300%,
respectively, for these compounds, one cannot exclude that
the hydroxyalkoxy radical formed after the OH addition may
undergo some other processes such as carbon-chain fragmen-
tation which would lead to CO and CO2 formation. Such
drastic decomposition may also explain the additional for-
maldehyde production. Indeed, the fact that the formaldehyde
production yield is slightly higher than the sabinaketone
production yield can mean that formaldehyde is not only a
co-product of sabinaketone. This fragmentation occurring
after opening of the sabinene ring could be an alternative
pathway to the acetone production.
The existence of a potential energy barrier for an exchange
reaction like A þ BC - AB þ C can be explained by models
involving the crossing of two correlation lines between states
of reactants and products. Another way to interpret such a
behaviour (in fact not really a different way) is to analyze the
competition between two processes: the BC bond cleavage
which tends to increase the energy and the AB bond forming
which favours the energy decrease. This model is the basis of
the concept of an intrinsic barrier described by Murdoch.35
Coupled to the exo or endothermicity of the A þ BC process,
this concept of an intrinsic barrier leads to the so called early
or late barrier largely used by the chemist invoking the
Hammond postulate and by Polanyi to explain the dynamics
of these exchange three-atom reactions36
The exploration of the different channels of the PES for the
reaction using the DFT-B3LYP/6-31G(d,p) quantum method
allows us to explain the presence of the three primary pro-
ducts. Two main channels have been explored revealing that
sabinaketone and HCHO are formed through the addition of
an OH radical to the CQC bond of sabinene. In the presence
of NO, some peroxynitrates are also formed. Alternatively, the
H-atom abstraction from the tertiary C–H bond in the
–CH(CH3)2 group is predicted to be the most important
abstraction site and leads mainly to acetone (plus sabinenyl
radical). Quantitative theoretical study is quite impossible, due
to the complexity of the reactions, but this qualitative ap-
proach can already give new insight into the sabinene þ OH
reaction.
In our case, we deal with an OH addition on a double bond.
This process can be compared to an exchange reaction since a
single C–O bond is created while a p bond is broken. The
dipole moment of an OH radical is rather large (1.76 D). It
turns out that when OH is approaching the ethylenic double
bond, the attractive long range interaction between this dipole
and the electronic doublet of the ethylenic bond dominates
over the weakening of this double bond which occurs for much
shorter O–C distances. As a matter of fact, neither the CQC
distance nor the p CQC bond order vary significantly as long
as OH approaches CQC over long distances. Therefore, the
domination of this attractive long range potential makes the
approach barrierless.
As stated above, based on the sabinaketone and acetone
yields, the OH addition and the H-abstraction pathways only
account for 44% of the reaction pathways. The missing
channels are thus difficult to predict, considering the complex-
ity of the radicals arising from the ring fragmentation. Addi-
tional experiments focused on CO/CO2 detection and
supplementary calculations would clearly help to clarify this
point.
Acknowledgements
J.C.R. and M.T.R. thank IDRIS (CNRS) for computer facil-
ities and F. Caralp for RRKM-master equation calculations.
N.C. and J.F.D. thank Dr John Wenger (UCC-CRAC) for
simulation chamber facilities. The experimental part of this
work have been achieved thanks to the support of EURO-
CHAMP, an Integrated Infrastructure Initiative of the 6th
framework European program.
No significant difference has been noted between reactions
with and without NOx. The similar carbonyl productions
observed in our study are explained by an efficient peroxy-
to-alkoxy conversion in both the presence and absence of
NOx, knowing that the H-abstraction pathway and the OH
radical addition to the terminalQCH2 group pathway mainly
lead to tertiary peroxy radicals. The calculations show that the
formed alkoxy radicals (vibrationally activated or not) will
preferentially decompose by scission of the C–C bond in the b
References
1 A. Guenther, C. N. Hewitt, D. Erickson, R. Fall, C. Geron, T.
Graedel, P. Harley, L. Klinger, M. Lerdau, W. A. McKay, T.
Pierce, B. Scholes, R. Steibrecher, R. Tallamraju, J. Taylor and P.
Zimmerman, J. Geophys. Res., 1995, 100, 8873.
2 A. Guenther, C. Geron, T. Pierce, B. Lamb, P. Harley and R. Fall,
Atmos. Environ., 2000, 34, 2205.
ꢃc
This journal is the Owner Societies 2006
3216 | Phys. Chem. Chem. Phys., 2006, 8, 3211–3217

View Article Online
3 R. Atkinson, S. M. Aschmann and J. Arey, Atmos. Environ., Part
A, 1990, 24, 2647.
24 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A.
Robb, J. R. Cheeseman, J. A. Montgomery Jr, T. Vreven, K. N.
Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tomasi, V.
Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G. A.
Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R.
Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O.
Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian, J.
B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E.
Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W.
Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J.
J. Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M.
C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Raghava-
chari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S.
Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko, P.
Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A.
Al-Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M.
W. Gill, B. Johnson, W. Chen, M. W. Wong, C. Gonzalez and
J. A. Pople, GAUSSIAN 03 (Revision C.02), Gaussian, Inc.,
Wallingford CT, 2004.
25 E. S. C. Kwok and R. Atkinson, Atmos. Environ., 1995, 29(14),
1685.
26 J. Peeters, S. Vandenberk, E. Piessens and V. Pultau, Chemosphere,
1999, 38(6), 1189.
27 J. J. Orlando, G. S. Tyndall, L. Vereecken and J. Peeters, J. Phys.
Chem. A, 2000, 104, 11578.
28 J. J. Orlando, G. S. Tyndall and T. J. Wallington, Chem. Rev.,
2003, 103, 4657, and references therein.
29 F. Caralp, W. Forst and M. T. Rayez, Phys. Chem. Chem. Phys.,
2003, 5, 476.
30 S. Skokov and R. A. Wheeler, Chem. Phys. Lett., 1997, 271, 251.
31 H. Basch and S. Hoz, J. Phys. Chem. A, 1997, 101, 4416.
32 B. Picquet-Varrault, J. F. Doussin, R. Durand-Jolibois and
P. Carlier, Environ. Sci. Technol., 2002, 36, 4081.
33 B. Klotz, I. Barnes and T. Imamura, Phys. Chem. Chem. Phys.,
2004, 8, 1725.
34 T. Hoffman, J. R. Odum, F. Bowman, D. Collins, D. Klockow, R.
C. Flagan and J. H. Seinfield, J. Atmos. Chem., 1997, 26, 189.
35 J. R. Murdoch, J. Am. Chem. Soc., 1983, 105, 2159.
36 J. C. Polanyi, Acc. Chem. Res., 1972, 5, 161.
4 L. Tollsten and M. Muller, Phytochemistry, 1996, 43, 759.
¨
5 S. Owen, C. Boissard, R. A. Street, S. C. Duckham, O. Csiky and
C. N. Hewitt, Atmos. Environ., 1997, 31, 101.
6 R. A. Street, S. Owen, S. C. Duckham, C. Boissard and C. N.
Hewitt, Atmos. Environ., 1997, 31, 89.
7 H. Hakola, J. Rinne and T. Laurila, Atmos. Environ., 1998, 32, 1825.
8 A. Reissell, C. Harry, S. M. Aschmann, R. Atkinson and J. Arey,
J. Geophys. Res., 1999, 104, 13869.
9 H. Hakola, J. Arey, S. M. Aschmann and R. Atkinson, J. Atmos.
Chem., 1994, 18, 75.
10 J. Arey, R. Atkinson and S. M. Aschmann, J. Geophys. Res., 1990,
95, 18–539.
11 L. Jaegle, D. J. Jacob, W. H. Brune and P. O. Wennberg, Atmos.
´
Environ., 2001, 35, 469.
12 D. Jacob, B. D. Field, E. M. Jin, I. Bey, Q. Li, J. A. Logan, R. M.
Yantosca and H. B. Singh, J. Geophys. Res., 2002, 107(D10), 4100.
13 J. F. Doussin, D. Ritz, R. Durand-Jolibois, A. Monod and P.
Carlier, Analusis, 1997, 25, 236.
14 L. Thuener, P. Bardini, G. J. Rea and J. Wenger, J. Phys. Chem. A,
2004, 108, 11019.
15 W. D. Taylor, T. D. Allston, M. J. Moscato, G. B. Fazekas, R.
Kozlowski and G. A. Takacs, Int. J. Chem. Kinet., 1980, 12, 231.
16 N. Carrasco, J. F. Doussin, B. Picquet-Varrault and P. Carlier,
Atmos. Environ., 2006, 40, 2011.
17 V. Librando and G. Tringali, J. Environ. Management, 2005, 75,
275.
18 A. D. Becke, J. Chem. Phys., 1993, 98, 5648.
19 F. Caralp, P. Devolder, C. Fittschen, N. Gomez, H. Hippler, R.
Me
Chem. Phys., 1999, 1, 2935.
20 R. Mereau, M. T. Rayez, F. Caralp and J. C. Rayez, Phys. Chem.
Chem. Phys., 2000, 2, 3765.
21 R. Mereau, M. T. Rayez, J. C. Rayez and F. Caralp, Phys. Chem.
Chem. Phys., 2003, 5, 4828.
22 R. Mereau, M. T. Rayez, F. Caralp and J. C. Rayez, Phys. Chem.
´
reau, M. T. Rayez, F. Striebel and B. Viskolcz, Phys. Chem.
´
´
´
Chem. Phys., 2000, 2, 1919.
23 L. Vereecken and J. Peeters, J. Phys. Chem. A, 2000, 104, 11140.
ꢃc
This journal is the Owner Societies 2006
Phys. Chem. Chem. Phys., 2006, 8, 3211–3217 | 3217