Pyrolysis of 3-carene: Experiment, Theory and Modeling

Article abstract of DOI:10.1007/s12039-015-0987-7

Thermal decomposition studies of 3-carene, a bio-fuel, have been carried out behind the reflected shock wave in a single pulse shock tube for temperature ranging from 920 K to 1220 K. The observed products in thermal decomposition of 3-carene are acetylene, allene, butadiene, isoprene, cyclopentadiene, hexatriene, benzene, toluene and p-xylene. The overall rate constant for 3-carene decomposition was found to be k / s^-1 = 10^{(9.95 ± 0.54)} exp (- 40.88 ± 2.71 kcal mol^-1/RT). Ab-initio theoretical calculations were carried out to find the minimum energy pathway that could explain the formation of the observed products in the thermal decomposition experiments. These calculations were carried out at B3LYP/6-311 + G(d,p) and G3 level of theories. A kinetic mechanism explaining the observed products in the thermal decomposition experiments has been derived. It is concluded that the linear hydrocarbons are the primary products in the pyrolysis of 3-carene.

Full text of DOI:10.1007/s12039-015-0987-7

c
J. Chem. Sci. Vol. 127, No. 12, December 2015, pp. 2119–2135. ꢀ Indian Academy of Sciences.
DOI 10.1007/s12039-015-0987-7
Pyrolysis of 3-carene: Experiment, Theory and Modeling
N SHARATHa, H K CHAKRAVARTYb,#, K P J REDDYa, P K BARHAIc and E ARUNANb,∗
aDepartment of Aerospace Engineering Indian Institute of Science, Bangalore, India
bDepartment of Inorganic and Physical chemistry, Indian Institute of Science, Bangalore, India
cDepartment of Applied Physics, Birla Institute of Technology, Mesra, Ranchi, India
#Current Address: Max Planck Institute for Chemistry, Mainz, Germany
e-mail: arunan@ipc.iisc.ernet.in
MS received 7 July 2015; revised 28 August 2015; accepted 31 August 2015
Abstract. Thermal decomposition studies of 3-carene, a bio-fuel, have been carried out behind the reflected
shock wave in a single pulse shock tube for temperature ranging from 920 K to 1220 K. The observed products
in thermal decomposition of 3-carene are acetylene, allene, butadiene, isoprene, cyclopentadiene, hexatriene,
benzene, toluene and p-xylene. The overall rate constant for 3-carene decomposition was found to be
0.54)
k/s−1 = 10(9.95
exp(−40.88 2.71 kcal mol−1/RT).
Ab-initio theoretical calculations were carried out to find the minimum energy pathway that could explain the
formation of the observed products in the thermal decomposition experiments. These calculations were carried
out at B3LYP/6-311+G(d,p) and G3 level of theories. A kinetic mechanism explaining the observed products
in the thermal decomposition experiments has been derived. It is concluded that the linear hydrocarbons are the
primary products in the pyrolysis of 3-carene.
Keywords. Shock Tube; Monoterpene; Thermal Decomposition; Isoprene.
1. Introduction
plant sources, namely, Zeravschaniamembranacea and
Pinusroxburghii.
Understanding the chemistry of hydrocarbons in the
combustion and soot formation process is one of the
growing research topics. Molecules like cyclohexane
were previously used as a model molecule to derive and
extend the pyrolysis and combustion mechanism for
higher hydrocarbons.1 It is well known that the ignition
of fuels will be preceded by its pyrolysis. The radicals
formed from the decomposition will either react with
oxygen or with other molecules/radicals to give prod-
ucts. Thus, the decomposition studies of hydrocarbons,
which have the characteristics of a fuel, could be use-
ful in understanding the combustion and soot formation
process.
Kinetics of kerosene based molecules such as JP-10,
which is used as a rocket fuel has been well studied.2–5
3-carene, a bi-cyclic monoterpene, the molecular struc-
ture of which is shown in figure 1, has the same molec-
ular formula as that of JP-10, C10H16. Initial extraction
of 3-carene was carried out by Simonsen from a grass,
Pinuslongifolia, obtained from Himalayas and the hills
ranging from Kashmir to Assam.6–8 Recently, Siavash
et al.9 and Salem et al.10 extracted 3-carene from other
In 2009, Kulkarni et al. suggested a blend of 75%
ethylidenenorbornene and 25% 3-carene as a promis-
ing high energy and density fuel.11 Recently, we car-
ried out ignition delay experiments of 3-carene/O2/Ar
mixture12 and found that under similar conditions its
ignition delay times are lower compared to that of JP-
10, supporting the results from earlier work11 that 3-
carene has all important characteristics of a fuel. As
mentioned above, to understand the initiation of oxi-
dation in a fuel, knowledge of its decomposition pro-
cess at higher temperature is essential. To the best of
our knowledge, the only available study on thermal
decomposition of 3-carene which was carried out in a
heated tube and analyzed using gas liquid chromatog-
raphy was focused on its isomerization process. It also
gave a brief qualitative and quantitative description on
the nature of products and two intermediates formed
at lower temperatures.13 The available results are not
sufficient to derive a complete thermal decomposition
mechanism and also, very low temperature experimen-
tal condition makes it difficult for extrapolating results
to higher temperatures and pressures. Thus, the decom-
position studies of 3-carene are required to understand
its thermal cracking process at elevated temperatures.
∗For correspondence
2119

2120
N Sharath et al.
initial pressure, P1 (450-600 mm Hg), were varied to
obtain the required temperatures. Before each experi-
ment the driven section was evacuated to 10−4 mm Hg
using an ADIXEN - ATP-400 turbo-molecular vacuum
pump. Argon (99.9993% pure) and helium (99.9993%
pure) were used as the driven and the driver gas,
respectively. 3-carene (boiling point 168◦C and density
0.857 g/mL) obtained from Sigma Aldrich was distilled
before use and confirmed to be 99% pure using a GC-
FID. A 2-liter sample tube was used to prepare a highly
diluted 3-carene and Ar mixture. The initial 3-carene
concentration was in the range of 130 ppm to 730 ppm
corresponding to mole fraction of 1.3 × 10−4 to 7.3 ×
10−4 in argon. The required amount of this mixture was
then fed into an evacuated sample section of the shock
tube.
Temperature and pressure were calculated from nor-
mal shock relation using the measured Mach number
obtained from PCB sensors (Model No.113A24) placed
at a distance of 0.505 m apart. A GC-FID was used in
qualitative and quantitative analysis of the post-shock
mixture. A HP-5, 29 m × 0.53 mm × 1.53 μm capil-
lary column was used to separate the product molecules
in the GC-FID. Nitrogen was used as a carrier gas and
its flow rate was maintained at 6 mL/min. The detector,
inlet and oven temperatures were maintained at 250◦C,
150◦C and 65◦C, respectively. A separate GC-MS was
used for qualitative analysis of the post shock mixture.
A HP-5MS, 30 m × 0.250 mm × 0.25 μm capillary
column was used to separate product molecules in GC-
MS. Helium was used as a carrier gas and its flow rate
was maintained at 0.5 mL/min. The oven and inlet were
maintained at same temperatures as that of GC-FID. A
sample cell of volume 200 mL was used to extract post-
shock mixture after completion of experiment from the
shock tube for further analysis. Prior to each experi-
ment, the sample chamber and the sample cell were
evacuated and blank runs were taken to ensure that there
were no contaminations from the previous experiment.
Figure 1. Two-dimensional molecular structure of 3-carene.
Also, 3-carene, being one of the abundant emitters into
atmosphere from the plants, plays an important role
in the atmospheric chemistry.14–16 Somewhat surpris-
ingly its chemistry has not been fully explored. Liter-
ature available on chemistry of 3-carene include deter-
mination of its structure17 and its reaction with triplet
oxygen18 and nitrating agents.19
3-carene is one of the monoterpenes which are
defined as molecules with two isoprene units. Pyrolysis
studies on isoprene, C5H8, were carried out to under-
stand the formation of Polycyclic Aromatic Hydro-
carbons (PAH).20,21 Recent isoprene pyrolysis stud-
ies showed that its decomposition could lead to the
formation of C1 to C6 product molecules.22
This manuscript reports the first detailed study on
thermal decomposition of 3-carene, a complex bicyclic
hydrocarbon. Several possible decomposition pathways
have been identified. The first section of the paper gives
the results from thermal decomposition experiments
carried out to understand the initial bond breaking pro-
cess. It will be followed by discussion on theoretical
results of 3-carene decomposition pathways. Results
from the theoretical calculations carried out to under-
stand the isoprene decomposition process will also be
presented. A kinetic mechanism which explains the
observed experimental results will be discussed as well.
2. Experimental/Computational
2.1 Experimental Setup
2.2 Computational Method
The Chemical shock tube-2 (CST-2), a single pulse Gaussian 0927 package was used to carry out quantum
shock tube was used to carry out pyrolysis experiments chemistry calculations. All geometries were optimized
presented in this paper. The details of CST-2 and the at G3 and B3LYP/6-311+G(d,p) level of theories.
experimental procedure have been described in previ- CHEMKIN28 has been used for numerical simulation of
ous papers.12,23–26 A brief description of the experimen- the reactions to explain the results. The transition state
tal setup and procedure is given here. The shock tube energies at G3 level of theory were manually calculated
with a driver section of length 2 m and driven section using procedure described by Curtiss et al.29 The zero
of 4.25 m and an inner diameter of 40 mm was sepa- point energies (ZPEs) for transition states were calcu-
rated using an aluminium diaphragm. The diaphragms lated using frequencies obtained at MP2/6-31G* level
were made from 1 mm or 0.7 mm aluminium sheets of theory and hence the corresponding scaling factor of
and grooved to different depths. The groove depth and 0.9427 was used.

3-carene pyrolysis
The procedure described in our previous paper26 has
2121
been used to obtain thermodynamic parameters for the
species used in the simulation of thermal decomposition
mechanism.
3. Results and Discussion
3.1 Experimental Observation
The thermal decomposition experiments were carried
out for temperature ranging from 920 to 1220 K. The
observed dwell times in the present study were in the
range of 1.35-1.76 ms. The uncertainty of obtained
Mach numbers were found to be less than 1%. This
corresponds to 20 K in terms of calculated tempera-
ture. The observed products are acetylene, allene, buta-
diene, isoprene, cylclopentadiene, hexatriene, benzene,
toluene and p-xylene. One of the gas chromatograms
of 3-carene post-shock mixture, obtained using the GC-
FID, is shown in figure 2. The sensitivity of FID, the
response of the FID, used in the present work to obtain
the concentration of molecules from the area under of
the peak (in chromatogram) are given in table S1 in
Supplementary Information (SI). The sensitivity factor
for each molecule was obtained by passing a known
concentration of sample through GC-FID. As seen in
figure 2, first three peaks are not resolved. These peaks
were de-convoluted using Eq. 1 to obtain the concentra-
tions of the product molecules corresponding to those
Figure 2. (a) Post shock chromatogram of 3-carene
obtained in GC-FID. T5 = 1185 K ; (b) The same chro-
matogram with time ranging from 1.2 to 2.5 min.
peaks.
2
−2(x−x
)
A
c
2
w
y = y0 + ꢀ e
(1)
π
w
2
Here, Y0: Value of the baseline, xc: x-ordinate of center
of the peak, A: Area under the curve and w: Full width second peak which was left un-assigned was concluded
at half maximum. to be that of a C3 molecule. Propane, propene and
High base line due to the presence of argon prevented propyne did not match the retention time of the second
us from identifying products below m/z of 40 using GC- peak. Hence, the peak was assigned to allene. Theoreti-
MS. The products with m/z >40 were identified after cal calculations and kinetic simulation also support the
matching the retention time in FID and analyzing the formation of allene as the C3 product. This was inferred
mass-spectrum obtained using GC-MS. The first peak after considering the barriers involved in the reactions
in the chromatogram obtained using GC-MS (scanned leading to the formation of propene.
for m/z >48) was found to be of butadiene and iden-
The peaks in the GC were identified and calibrated
tified as the third peak in the chromatogram obtained using standard sample except for three molecules,
using GC-FID. After elution of butadiene, a similar allene, cyclopentadiene and hexatriene. Standard sam-
elusion pattern was observed in GC-FID and GC-MS. ples for these could not be procured. Allene sensitiv-
Hence the first two peaks in the chromatogram obtained ity was approximated to that of propyne. The presence
using GC-FID were concluded to be of m/z <48. In of cyclopentadiene was also confirmed by the thermal
order to identify these peaks, C1, C2 and C3 compounds decomposition studies of JP-10. It is well known that
were passed through GC-FID to match the retention the pyrolysis of JP-10 will result in the formation of
time. Acetylene was confirmed to be the first peak.
cyclopentadiene. The retention time of the peak corre-
Methane, ethane and ethylene did not match the sponding to m/z=66 obtained in the thermal decom-
retention time of the peaks. Hence, the only peak, the position studies of 3-carene is the same as that of the

2122
N Sharath et al.
peak corresponding to m/z=66 obtained in the ther- of soot formation even after all the experiments. The
mal decomposition of JP-10. To obtain the sensitivity assumption can also be justified using the results from
of FID to hexatriene, we studied the relation between previous works where it has been found that below
FID sensitivity and the number of C-H bonds present in 1300 K, nearly 100% of carbon will recovered in gas
C6 molecules. A plot showing the variation of sensitiv- phase.30–32
ity with number of C-H present in the C6 molecules is
[carene]0
=
(10 × [carene]t + 2 × [acetylene]t
+3 × [allene]t + 4 × [butadiene]t
+5 × [isoprene]t + 5 × [cyclopentadiene]t
+6 × [hexatiene]t + 6 × [benzene]t
+7 × [toluene]t
given in figure S1 in SI. Similar analysis was also car-
ried out for cyclopentadiene. Its sensitivity was approx-
imated to that of cyclopentene and appropriate reduc-
tion in sensitivity was done to account for the absence
of two C-H bonds.
+8 × [p − xylene]t)/10
(2)
The initial concentration of 3-carene (pre-shock) was
determined using GC. Since the post shock mixture will
be diluted by helium (driver gas), one needs to find the
amount of initial reaction concentration to which it is
diluted. We have previously used the sum of all prod-
ucts and the final reactant concentration to estimate ini-
tial reaction concentration.25 We have found this to be
more accurate than the concentration measured initially.
However, it is important to ensure that the all reactant
and products have been recovered. Hence, the experi-
mental conditions were chosen to avoid soot formation
and condensation of reactants in the wall. Equation 2
has been used to obtain the initial concentration of the
reactant as the uniformity and the extent of initial reac-
tant concentration dilution were not known accurately.
This was considered with an assumption that the en-
tire carbon was recovered in the gaseous phase. The as-
sumption can be justified on the backdrop of absence
Here, [carene]0 is the initial concentration of 3-
carene and [x]t is the concentration of the respective
molecules after the reaction for a time t given in table 1.
The overall rate constant was calculated using Eq. 3.
Arrhenius plot for overall decomposition of 3-carene is
shown in figure 3.
1
[carene]t
[carene]0
k = − ln
(3)
t
The rate constant expression shown in Eq. 4 was
obtained by linear fitting the data in figure 3.
ꢁ
ꢂ
−40.88 2.71/kcal.mol−1
k/s−1 = 10(9.95 0.54)e
RT
(4)
The product concentrations indicate that the acety-
lene constitutes a major product in the dissociation of
Table 1. Summary of experimental conditions and normalized mole fraction of products obtained in pyrolysis experiments
of 3-carene. T1: Room temperature.
Exp. Dwell Time Temperature Pressure C2H2 C3H4 C4H6 C5H8 C6H6 C7H8 C8H10 C6H8 C5H6 3-Carene
No.
ms
K
bar
1
2
3
4
5
6
7
8
1.76
1.65
1.66
1.53
1.72
1.5
1.65
1.57
1.48
1.64
1.61
1.55
1.49
1.4
1151
974
12.15 0.238 0.067 0.063 0.058 0.027 0.036 0.017 0.032 0.011
0.799
0.992
0.985
0.81
0.973
0.451
0.656
0.673
0.655
0.951
0.791
0.881
0.73
8.71 0.003 0.001 0.001 0.002
9.17 0.01 0.005 0.004 0.01 0.001 0.003 0.001 0.003
11.02 0.2 0.071 0.062 0.071 0.017 0.029 0.016 0.035 0.008
9.5 0.019 0.008 0.006 0.015 0.001 0.005 0.003 0.006 0.002
13.48 0.866 0.209 0.189 0.1 0.088 0.082 0.032 0.054 0.037
14.66 0.4 0.128 0.103 0.106 0.033 0.056 0.026 0.056 0.036
0
0.001
0
0.001 0.008
1038
1163
1057
1220
1172
1172
1198
1066
1136
1101
1163
1027
1009
1066
926
0
15.26 0.373 0.111 0.095 0.101 0.029 0.053 0.027 0.06 0.037
11.67 0.455 0.132 0.108 0.081 0.052 0.064 0.025 0.041 0.022
9.87 0.034 0.015 0.013 0.025 0.002 0.009 0.004 0.012 0.003
10.35 0.19 0.071 0.059 0.082 0.013 0.036 0.018 0.045 0.02
11.27 0.094 0.035 0.031 0.033 0.041 0.019 0.009 0.023 0.004
12.83 0.396 0.087 0.079 0.066 0.04 0.038 0.017 0.04 0.025
9
10
11
12
13
14
15
16
17
18
11.03 0.015 0.005 0.006 0.013 0.004 0.005 0.001 0.006
11.1 0.01 0.003 0.004 0.008 0.004 0.003 0.001 0.004
11.17 0.045 0.018 0.019 0.035 0.005 0.008 0.005 0.019
8.69 0.003 0.001 0.001 0.001 0.002 0.002 0.003
0
0
0
0
0.976
0.984
0.937
0.994
0.692
1.4
1.39
1.35
1.57
0
1150
13.35 0.347 0.099 0.131 0.095 0.032 0.041 0.021 0.054 0.025
Here, C2H2: Acetylene, C3H4: Allene, C4H6: Butadiene, C5H8: Isoprene, C5H6: Cyclopentadiene, C6H6: Benzene, C7H8:
Toluene, C8H10 = p-Xylene, C6H8: Hexatriene.

3-carene pyrolysis
2123
that of flash pyrolysis) suggests a similar behavior by
3-carene as that of isoprene. The aromatic products
observed in isoprene thermal decomposition were also
observed in our experiments. The molecular formula
for 3-carene, C10H16 happens to be twice that of iso-
prene, C5H8. However, structurally they are very differ-
ent and 3-carene is not formed by combining two iso-
prene units. However, isoprene is one of the products
from its decomposition process. Hence, we explored the
possibility of isoprene being the primary product and
formation of other products through isoprene decom-
position. The differences in molecular structure clearly
suggests that direct formation of isoprene from 3-carene
is not possible. Moreover, the observed product profiles
also indicate that this pathway is not likely. For exam-
ple, the minimum energy pathway leading to the for-
mation of acetylene from isoprene involves its decom-
position to propene and vinylidine, which has a bar-
rier of 92.3 kcal mol−1 and will not yield the observed
acetylene concentration. However, in the present work
we have found that at lower temperatures (<1050 K)
the concentrations of isoprene and acetylene are almost
equal and with increase in temperature, the concentra-
tion of acetylene was found to increase more rapidly
compared to that of isoprene. Cyclopentadiene, which
was recently proposed as one of the precursor for PAH33
was observed to form after 1150 K.
Figure 3. Arrhenius plot for overall decomposition of
3-carene.
3-carene. Allene, butadiene and isoprene were the next
major products and these are equally important. Among
the minor products, [Toluene] > [Benzene] > [Hexa-
triene] > [p-Xylene]. The conditions at which exper-
iments were carried out and the normalized product
mole fractions have been given in table 1. The avail-
able literature13 on 3-carene thermal decomposition for
temperature ranging from 723 to 948 K indicated the
formation of aromatic compounds as major products. 3-
carene started to crack at 723 K and at 753 K, one third
3-carene got converted into products. At 833 K, the
entire 3-carene concentration was converted into prod-
ucts. In the above said range (753 to 833 K), toluene,
xylene, and cymene were the major constituents of the
observed products followed by methylstyrene, styrene,
benzene and dehydrocymene. Isoprene and propene
were minor products. In the present work, we observed
around 40% of acetylene. Allene, butadiene and iso-
prene together were found to constitute 30% of the
product concentration. Hexatriene, toluene, xylene and
benzene account for the rest.
All the observed products distribution indicates that
the primary product should be a linear molecule and
not cyclic. Cocker et al.13 in particular discussed the
resistance of cyclopropane ring present in 3-carene to
decompose in vapor phase. This has a minimum BDE
compared to rest of the bonds in 3-carene. The path
leading to the formation of observed products will be
explained in theoretical and mechanism sections.
3.2 Theoretical Observations
Ab-initio theoretical calculations were carried out to
find the minimum energy pathway that could lead to the
formation of experimentally observed products. Tran-
sition states (TS) were optimized to find the energy
barrier involved in the reactions. The theoretical rate
constants were calculated using Eq. 5.
Pyrolysis studies of isoprene which was carried out
for temperature ranging from 673 to 1273 K resulted
in the formation of 49 aromatic products with sev-
eral PAHs.20 Pyrolysis of isoprene at 973 K resulted
in the formation of benzene, toluene and xylene as
major products and accounted for 52% of the observed
products.21 Flash pyrolysis of isoprene resulted in
kBT
h
Q=
QR
E
a
RT
k = l
×
× e−
(5)
the formation of linear molecules and benzene was Here, QR and Q# correspond to partition functions
observed after 1250 K.22 On careful observation, a sim- for reactant and activated complex, respectively. T is
ilar behaviour can also be observed in the pyrolysis temperature, Ea is activation energy, KB is Boltzmann
of 3-carene. The observed aromatic molecules at lower constant, l is reaction degeneracy factor.
temperatures (and longer time) and linear molecules
Partition functions were calculated using the vibra-
at higher temperatures (for test time comparable to tional frequencies and rotational constants obtained at

2124
N Sharath et al.
Figure 4. Molecular structure of 3-carene and intermediates formed in the thermal decomposition
of 3-carene optimized at G3 level of theory. Here, (a) and (f) are in singlet state, (b), (c) and (d) are
in triplet state and, (e), (g), (h) and (i) are in doublet state.

3-carene pyrolysis
2125
Figure 5. Transition state structures for intermediate reaction steps originating from 3-carene
optimized at G3 level of theory (except for TS1, optimized at B3LYP/6-311+G(d,p) level of theory.

2126
N Sharath et al.
B3LYP/6-311+G(d,p) level of theory. The obtained dissociation barriers for C-C and C=C bonds are >80
rate parameters are summarized in table S3.
kcal mol−1. The BDE (Bond dissociation Energy) for
The following section describes the decomposition C7-C13 bond is 83 kcal mol−1. The C-H BDEs and, the
pathways of 3-carene. The pathways involving radicals barriers involved in H abstraction from 3-carene by an
formed from the H-dissociation and H-abstraction from H atom have been shown in figure 6. Hence the struc-
3-carene will also be discussed.
ture shown in figure 4(b) has been considered as the first
intermediate from thermal decomposition of 3-carene.
The C5-C6 bond (in C10H16(T)) shown in figure 4(b)
has a BDE of 14.8 (23.3) kcal mol−1 at B3LYP/6-
311+G(d,p) (G3) level of theory. The ring opening pro-
cess will lead to the formation of a linear intermediate,
C10H16(A), the structure shown in figure 4(c). The H
migration to form structure shown in figure S3, given
in the Supplementary Information is not considered as
it has a higher barrier. At B3LYP/6-311+G(d,p) level
of theory the difference between ring opening and H
migration is 18.4 kcal mol−1. Also, the rate constants
for ring opening process at 1000 K, 3.7 × 109 s−1, can
be compared with 2.8 × 105 s−1, for the H migration
process. The potential energy scans along C-C stretch-
ing in the intermediate structure given in figure 4(b)
showed that the considered reaction path has the least
energy barrier compared to the rest of the C-C bonds.
The C=C BDE is assumed to be >C-C BDE. Hence,
the formation of the linear intermediate shown in
figure 4(c) from the structure shown in figure 4(b) was
found to have a lower BDE and can explain the forma-
tion of observed products. The potential energy scans
along C-C stretching for the structure shown in figure
4(b) has been given in figures S4 and S5 in the Supple-
mentary Information.
3.2.1 3-Carene Decomposition Pathways: The forma-
tion of linear molecules even at low temperatures indi-
cates that 3-carene ring structure should form a linear
molecule/intermediate in the initial stages of its dissoci-
ation process. Theoretical calculations also support the
above statement. The theoretically optimized molecu-
lar/intermediate structures which will be described in
this section are shown in figure 4. The transition state
structures which were optimized to find the minimum
energy path for the reactions are shown in figure 5.
3.2.1a 3-carene ring opening and rearrangement:
Except for the first step, all TS (transition state) energies
were calculated at G3 and B3LYP/6-311+G(d,p) level
of theories. The first step, formation of di-radical, struc-
ture of which has been shown in figure 4(b), C10H16(T),
was analyzed using only B3LYP/6-311+G(d,p) level of
theory. Attempts to optimize transition state structure at
MP2/6-31G* level of theory were not successful. The
energy barrier for the above reaction was found to be
51.0 kcal mol−1. The potential energy scans for all C-C
and C=C bonds are shown in figure S2, given in
the Supplementary Information. It was clear that the
Figure 6. C-H BDEs at G3 and B3LYP/6-311+G(d,p) level of theories and
barriers involved in H abstraction in 3-carene by an H atom.

3-carene pyrolysis
2127
The hydrogen H14 in the structure, C10H16(A), (at 1000 K) to be an order lower than that of H migration
shown in figure 4(c) migrates from C13 carbon to reaction. Similar differences were also found for C13-
C4 carbon forming a structure, C10H16(F), looking like H15 and C9-H12 dissociation reactions. Even though
dimer of isoprene shown in figure 4(d). The activation the C2-H26 BDE is very close to that of H migra-
energy for the above mentioned reaction is 32.9 (36.7) tion reaction, the subsequent reaction from the radi-
kcal mol−1 at B3LYP/6-311+G(d,p) (G3) level of the- cal formed by C2-H26 bond dissociation will lead to
ory. The C-C and C-H BDEs for the structure shown a C10 molecule, which is not observed in the present
in figure 4(c) are given in table S3. The rate con- experiment and hence not considered in the kinetic
stant for C20-H23 bond dissociation reaction was found mechanism presented in the paper.
Figure 7. Schematic diagram depicting energies of molecules, intermediates and transition states involved in 3-carene
decomposition. Solid lines (Black): G3 level of theory, Dotted lines (Red): B3LYP/6-311+G(d,p) level of theory. The values
given in the parenthesis correspond to transition state energy of the corresponding reaction.

2128
N Sharath et al.
3.2.1b Formation of lower hydrocarbons: The disso- been summarized in table S4. The reactions involving
ciation of hydrogen from C19 carbon in structure shown radicals formed from H dissociation from C2 and C4
in figure 4(d) results in the formation of C10H15, struc- [C10H15(FH1) in figure 8] carbon atoms will be dis-
ture of which has been shown in figure 4(e). The bar- cussed later in this paper. Formation of isoprene from
rier for the above reaction at B3LYP/6-311+G(d,p) the structure shown in figure 4(d) preceded by CH3
(G3) level of theory is 48.2(50.2) kcal mol−1. The C- elimination (C7-C9 bond dissociation) is not consid-
C BDEs for the structure shown in figure 4(d) has ered in the mechanism as the detailed analysis showed
Figure 8. Molecular structure of 3-carene radical (C10H15 (A)) and intermediates formed in the
decomposition of 3-carene-H radical (C10H15 (A)) optimized at G3 level of theory.

3-carene pyrolysis
2129
that the barrier involved in the dissociation of C7-C9 is C9H12 (figure 4(f)). The BDE of C7-C9 decompo-
∼60 kcal mol−1 at B3LYP/6-311+G(d,p) level of the- sition at B3LYP/6-311+G(d,p) (G3) level of theory
ory. The potential energy scan along the above said C-C is 52.3(59.7) kcal mol−1. The BDE of C4-C6 bond
stretching is shown in figure S7 in the Supplementary at B3LYP/6-311+G(d,p) (G3) level of theory is 59.4
Information.
(64.5) kcal mol−1. However the scan along C4-C6 bond
The doublet C10H15, structure of which is shown showed that barrier involved in the C4-C6 dissociation
in figure 4(e) then looses CH3 group to form at B3LYP/6-311+G(d,p) level of theory is greater than
Figure 9. Transition state structures for intermediate reaction steps originating from 3-carene
radical (C10H15 (A)) optimized at G3 level of theory.

2130
N Sharath et al.
85 kcal mol−1. The BDE of C2-C4 bond (at B3LYP/6-
311+G(d,p) (G3) level of theory) in the structure shown
in figure 4(f) is 52.3(63.5) kcal mol−1. This bond dis-
sociation will lead to the formation of isoprene radi-
cal (figure 4(g)) and butadiene radical (figure 4(h)). The
BDE of C4-C6 bond (at B3LYP/6-311+G(d,p) level of
theory), the dissociation of which results in the forma-
tion of allene radical is 69.9 kcal mol−1 and hence not
considered in the mechanism presented in this paper.
The schematic diagram showing the relative energies
of the transition state, intermediates and products are
given in figure 7. The transition state energies given
in the figure are calculated at B3LYP/6-311+G(d,p)
level of theory. The transition state energies for same
reactions/steps are given in table S5.
dissociation of 3-carene. The pathways leading from
other radicals formed by H dissociation/abstraction
at other carbon site can be neglected as the barriers
involved in the H dissociation/abstraction and subse-
quent dissociation leading to the products were found to
be higher when compared to C10H15 (A) and C10H15(B)
dissociation pathways.
In this paper the reaction pathway originating from
radicals formed by C2-H26 dissociation/ H26-H abs-
traction will be described. The intermediate structures
involved in the decomposition pathways of C10H15(A)
are shown in figure 8. The transition states involved in
the decomposition pathways of C10H15(A) are shown in
figure 9.
The dissociation pathways of C10H15(A), shown in
figure 8(a), are similar to that of 3-carene. As in 3-
carene, the first step was found to be C3 ring opening
3.2.2 C10H15 Decomposition Pathway: The structures process and will lead to the formation of C10H15(R1),
of the radicals formed after H dissociation and H shown in figure 8(b). The energy barrier involved in
abstraction from 3-carene and, transition state struc- the reaction at B3LYP/6-311+G (d,p) (G3) level of the-
tures for H abstraction have been given in figures S8 ory is 10.2 (15.9) kcal mol−1 and the transition state
and S9. It has been found that the pathways leading structure for the reaction is shown in figure 9(a). The
from carene-H(A)(C10H15(A)), formed from H removal dissociation of C5-C6 bond in C10H15 (R1) leading to
at C2 site, and carene-H(B)(C10H15(B)), formed from the formation of C10H15 (R2) has a barrier of 18.7 (27.0)
H removal at C5 site, play important roles in the kcal mol−1 at B3LYP/6-311+G(d,p) (G3) level of
Figure 10. Schematic diagram depicting energies of molecules, intermediates and transition states involved in 3-carene
radical (C10H15 (A)) decomposition. Solid lines (Black): G3 level of theory, Dotted lines (Red): B3LYP/6-311+G (d,p) level
of theory. The values given in the parenthesis correspond to transition state energy of corresponding reaction.

3-carene pyrolysis
2131
theory and the transition state structure is shown in bonds, for the sake of completion, results of the calcula-
figure 9(b). The above step is similar to ring opening tions which were carried out to understand the decom-
process leading to C10H16 (A) from C10H16(T).
position pathways of rest of the radicals formed by
The third step as for 3-carene is H-migration. The H dissociation/abstraction at other carbon sites will be
hydrogen migration from C9 to C2 carbon atom lead- briefly described below. The potential energy scans of
ing to C10H15 (FH1), shown in figure 8, has a barrier the bonds which will be discussed below are given in
of 36.6 (36.5) kcal mol−1 at B3LYP/6-311+G(d,p)(G3) the Supplementary Information (figures S13-S15).
level of theory. The transition state structure for the
The final products from the dissociation of radicals
above reaction is shown in figure 9(c). Unlike 3-carene, formed by the H dissociation/abstraction of hydrogen at
the next step in the dissociation was found to be C2- C9 (C10H15(C)) and C13 (C10H15 (D)) sites in 3-carene
C4 bond dissociation resulting in the formation of iso- are same as those of C10H15 (A) and C10H15 (B). The
prene, C5H8, and C5H7 (A) radical. The transition state first and second step like 3-carene are C3 ring opening
structure for the above reaction is shown in figure 9(d). and C5-C6 bond dissociation (to form a linear structure
The reaction step has a barrier of 40.5 (57.2) kcal mol−1 from ring structure), respectively. However in the case
at B3LYP/6-311+G(d,p) (G3) level of theory. The C3- of C10H15(C) and C10H15 (D) the H migration reaction
C2 bond dissociation in C5H7 (A) leading to acety- is not required to form the final products. The barrier
lene and C3H5 has a BDE of 35.9 (40.9) kcal mol−1 involved in the second ring opening process leading to a
at B3LYP/6-311+G(d,p)(G3) level of theory. The tran- linear intermediate from ring structure was higher than
sition state structure for the above reaction is given in those of C10H15 (A) and C10H15(B) second ring open-
figure 9(f). The transition state structure for H26 ing process. Dissociation of C2-C4 bond in the resulting
abstraction by H atom is shown in figure 9(e). The structure will lead to the formation of isoprene, C5H8,
schematics showing the relative energies of transition and C5H7 (A) radical.
states and intermediates with respect to C10H15 (A) is
The radical formed by H dissociation/abstraction
shown in figure 10. The reaction pathway originating from C20 site (C10H15 (E)) in 3-carene, which has
from C10H15 (B) was found to be similar to that of a BDE closer to C-H BDE to form C10H15 (A) and
C10H15 (A) and hence not described in detail. How- C10H15(B) from 3-carene, has a different dissociation
ever schematics showing the relative energies of inter- pathway than those explained above. In this case the
mediates, transition states and products calculated at first step of dissociation process is breaking of C1-C2
B3LYP/6-311+G (d,p) level of theory is given in figure bond. A series of dissociation reactions result in the for-
S11 in the Supplementary Information.
mation of C3H3 and C7 molecules. The energy barri-
Even though the BDE of rest of C-H bonds in 3- ers involved in the dissociation process of C10H15 (E)
carene is higher than those of above discussed C-H is considerably higher than those of C10H15 (A) and
Figure 11. Transition state geometries for (a) CH3 decomposition from isoprene;
(b) Hydrogen dissociation from isoprene to form isoprene radical; (c) Propene molec-
ular elimination from isoprene; and (d) C3H5 + vinyldiene elimination from isoprene
radical optimized at G3 level of theory.

2132
N Sharath et al.
C10H15 (B). A detailed pathway for C10H15 (E) disso- lead to a structure with a higher C-C and C-H BDEs.
ciationhasbeengiven in the Supplementary Information. The C1-C2 (C3-C5) bond dissociation in C10H15 (F)
The H dissociation/abstraction from C4 (C10H15 (F)) (C10H15 (G)), even though an endothermic reaction,
and C6 (C10H15 (G)) sites in 3-carene followed by can result in the formation propyne and C5 interme-
breaking of C6-C7 and C4-C7 bonds, respectively, was diate with a lower energy bottle neck. The potential
also analyzed in our study. The C6-C7 and C4-C7 dis- energy scans leading to propyne from C8 intermedi-
sociation reactions are exothermic (∼ 30 kcal mol−1). ate has been given in figure S15 in the Supplementary
However, the above explained exothermic reaction will Information.
Figure 12. Schematic diagram depicting energies of molecules, intermediates and transition states
involved in isoprene decomposition. Solid lines (Black): G3 level of theory, Dotted lines (Red): B3LYP/6-
311+G(d,p) level of theory.

3-carene pyrolysis
2133
3.2.3 Isoprene Decomposition Pathway: Formation The reaction leading to the formation of CH2CHCH2
and vinylidene from isoprene radical has a barrier of
89.5 (96.0) kcal mol−1 at B3LYP/6-311+G (d,p) (G3)
level of theory. The transition state structure for the
above mentioned step is shown in figure 11(d). The C2-
C1 bond dissociation in isoprene radical was found to
result in the formation of allene and vinyl radical. The
above reaction has a barrier of 60.1 (68.3) kcal mol−1 at
B3LYP/6-311+G (d,p) (G3) level of theory. This reac-
tion was found to explain the formation of the observed
C3 molecule concentration. Since the reactant for the
of isoprene radical from 3-carene and, isoprene from
C10H15 (A) and C10H15 (B), respectively, necessitated
the decomposition study of isoprene molecule. Tran-
sition states for two molecular elimination pathways,
one each for isoprene and isoprene radical, have been
found. The reaction which leads to the formation of
propene and vinylidine from isoprene has an energy
barrier of 90.0 (92.3) kcal mol−1 at B3LYP/6-311+G
(d,p) (G3) level of theory. The transition state struc-
ture for the above reaction is shown in figure 11(c).
Figure 13. Normalized concentration profiles of selected molecules with temperature. squares: Experi-
mental values. circles: Values obtained from the mechanism.

2134
N Sharath et al.
above step is formed in the decomposition of 3-carene, isoprene decomposition showed the formation of allene,
transition state structure for the above reaction is also we are proposing the following reaction to explain the
given in the figure 5(g). A schematic diagram show- formation of xylene.
ing relative energies of reactant, intermediates and tran-
sition states with respect to isoprene energy has been
given in figure 12.
[R26] C5H7 + C3H4 = C8H10 + H
Reactions R17 and R18 were included to check
whether xylene could be a secondary product formed
by reaction of toluene and CH3 radical. For Reaction
R17, a range of rate constants in the literature were also
considered. The rate constant for the reaction was var-
ied in the available data range to find out its effect on
xylene formation. The recombination rate constant for
R19 was approximated to the following reaction.
3.3 Kinetic Mechanism
With the help of results described in experimental and
theoretical sections, the mechanism for the 3-carene
thermal decomposition given in table S2, with 44
reactions involving 37 species, has been derived to fit
the product concentrations observed in the present ex-
perimental work. Since a mechanism for pyrolysis of
both 3-carene and isoprene does not exist in the litera-
ture, we have tried to use minimum number of reactions
to explain the product profiles observed in the present
experiments.
After simulating the mechanism, given in table S2, at
reaction conditions given in table 1, it was again simula-
ted at 300 K and 2 atm for 10 ms. This has been done
to take into account of radical reactions which can con-
tinue until all are consumed. The rate parameters used
in the mechanism are obtained from TST calculations
and refined to fit the observed concentrations. For sim-
ple molecular systems, such refinements in TST rate
parameters may not be required. However in the present
case, even with such a detailed analysis for the consid-
ered system we could not explain the observed prod-
uct concentrations without such small variations in
the calculated theoretical rate parameters. To compare
the rate parameters calculated using TST at B3LYP/6-
311+G(d,p) level of theory and the one used in the
mechanism, the former is also given in table S3.
The rate constant for reaction R9 was obtained using
the literature and the present work. Pre-exponential fac-
tor was obtained using the variational transition state
theory and the activation energy was obtained using
reported work.34 The pre-exponential factor for R14
was increased from 6.16 × 1011 to 1.00 × 1012 cm3
mole−1 s−1. The Arrhenius rate parameters, 9.77 ×
1012 cm3 mole−1 s−1 and 0.1 kcal mol−1 given in the
literature35 for reaction R16 can be compared with
present values of 6.0 × 1014 cm3 mole−1 s−1 and 0.1
kcal mol−1. The n factor used the literature for R16 is -
0.08 while in the present work we have used n factor of
0.7 for the same.
[R25] C6H5 + CH3 = C6H5CH3
The obtained results suggested that the reactions R17
and R18 will not be able to explain xylene concentration
obtained in the present experimental work. Thus, xylene
could have formed through reaction R26. Rate constant
for the above reaction was fitted to obtain the observed
xylene concentration. One more possible pathway for
xylene formation could be that of C4H5 dimerisa-
tion process, but, it is not considered in the present
work. The comparison of concentrations obtained in
experiments and simulation is given in figure 13 and the
results are in reasonable agreement.
4. Conclusions
Thermal decomposition experiments of 3-carene have
been carried out for temperatures ranging from 920 to
1220 K. Linear products were found to form in major
concentrations. Ab initio calculations were carried out
to help in deriving and understanding the pyrolysis
mechanism of 3-carene. Opening of C3 ring in 3-carene
as well as in C10H15(A) and C10H15(B) was found to be
the first step in their decomposition process. Theoreti-
cal calculations also showed that 3-carene decomposi-
tion will lead to the formation of linear products. Rate
parameters obtained using transition state theory was
refined and used in the kinetic mechanism presented
in the paper. The mechanism fairly replicated the for-
mation of observed product concentration. This is the
first detailed experimental and theoretical work carried
out on 3-carene decomposition process. We hope that
this work will help in deriving the oxidation mecha-
nism and, stimulate further experimental and theoretical
studies on the pyrolysis of this important molecule.
Pre-exponential factor for R25 was increased from
1.38 × 1013 to 6.0 × 1013 cm3 mole−1 s−1. We are
unable to explain xylene formation through transition
Supplementary Information
state theory. Since isoprene decomposition also resulted Full citation for Ref 27., sensitivity of FID to the
in the formation of xylene, and, previous results of molecules, tables S1-S6, selected potential energy

3-carene pyrolysis
2135
12. Sharath N, Reddy K P J, Barhai P K and Arunan E 2015
Curr. Sci. 108 2083
13. Cocker W, Hanna D P and Shannon P V R 1968 J. Chem.
Soc. (C) 489
14. Feltham E J, Almond M J, Marston G, Ly V P and
Wiltshire K S 2005 Spectrochim. Acta A 56 2605
15. Plemenkov V V, Rul S V, Kuzmina L G, Kataeva O N
and Litvinov I A 1997 J. Mol. Struct. 412 239
16. Luo D, Pierce J A, Malkina I L and Carter W P 1996 Int.
J. Chem. Kinet. 28 1
scans, structures of diradical-2, 3-carene-H radicals and
TS for H abstraction at different sites in 3-carene, Path-
ways for dissociation of 3-carene-H radicals, cartesian
coordinates and normal mode frequencies of molecules,
intermediates and transition states are given in Supple-
mentary Information which is available at www.ias.ac.
in/chemsci.
17. Tkachev A V and Denisov A Y 1991 Mendeleev Com-
mun. 1 98
18. Paulson S E, Orlando J J, Tyndall G S and Calvert J G
1995 Int. J. Chem. Kinet. 27 997
19. Bakhvalov O V, Bagryanskaya I Y, Gatilov Y V,
Korchagina D V and Barkhash V A 2002 Russian J. Org.
Chem. 38 507
20. Oro J, Han J and Zlatkis A 1967 Anal. Chem. 39 27
21. Badger G M, Donnelly J K and Spotswood T M 1966
Aust. J. Chem. 19 1023
Acknowledgements
We acknowledge funding from the following organi-
zations: IISc-ISRO Space Technology Cell, Defence
Research Development Organization, Defence Rese-
arch Development Laboratory and Aeronautics Rese-
arch and Development Board.
References
22. Weber K H and Zhang J 2007 J. Phys. Chem. A 111
11487
1. Granata S, Faravelli T and Ranzi E 2003 Combust.
Flame 132 533
2. Van D B and Anderson S L 2006 Energy Fuels 20
1886
23. Nagaboopathy M, Vijayanand C, Hegde G, Reddy K P J
and Arunan E 2008 Curr. Sci. 95 78
24. Rajakumar B, Reddy K P J and Arunan E 2002 J. Phys.
Chem. A 106 8366
3. Nakra S, Green R J and Anderson S L 2006 Combust.
Flame 144 662
25. Rajakumar B, Reddy K P J and Arunan E 2003 J. Phys.
Chem. A 107 9782
26. Sharath N, Reddy K P J and Arunan E 2014 J. Phys.
Chem. A 118 5927
27. Frisch M J et al. 2009 (Wallingford CT: Gaussian Inc.)
28. CHEMKIN-PRO 15101 2010 (San Diego: Reaction
Design)
29. Curtiss L A, Raghavachari K, Redfern P C, Rassolov V
and Pople J A 1998 J. Chem. Phys. 109 7764
30. Alexiou A and Williams A 1996 Combust. Flame 104
51
4. Chenoweth K, van Duin A C, Dasgupta S and Goddard
III W A 2009 J. Phys. Chem. A 113 1740
5. Davidson D F, Horning D C, Oehlschlaeger M A
and Hanson R K 2001 In Proceedings of the 37th
AIAA/ASME/SAE/ASEE Joint Propulsion Conference
and Exhibit, Salt Lake City, UT, July 8-11, 2001, AIAA
01–3707
6. Simonsen J L 1920 J. Chem. Soc., Trans. 117 570
7. Simonsen J L 1922 J. Chem. Soc., Trans. 121 2292
8. Simonsen J L and GopalaRau M 1923 J. Chem. Soc.,
Trans. 123 549
31. Sivaramakrishnan R, Tranter R S and Brezinsky K 2006
J. Phys. Chem. A 110 9388
9. Saei D S S, Khalighi-Sigaroodi F, Pirali-Kheirabadi K,
Saei-Dehkordi S S, Alimardani-Naghani F and Fallah A
A 2014 J. Food Safety 34 87
10. Salem M Z M, Ali H M and Basalah M O 2014
BioResources 9 7454
11. Kulkarni S G, Bagalkote V S, Patil S S, Kumar U P
and Kumar V A 2009 Propellants Explos. Pyrotech. 34
520
32. Gudiyella S, Malewicki T, Comandini A and Brezinsky
K 2011 Combust. Flame 158 687
33. Wang D, Violi A, Kim D H and Mullholland J A 2006
J. Phys. Chem. A 110 4719
34. Harding L B, Klippenstein S J and Georgievskii Y 2007
J. Phys. Chem. A 111 3789
35. Westmoreland P R, Dean A M, Howard J B
and Longwell J P 1989 J. Phys. Chem. 93 8171