Homogeneous and unimolecular gas-phase thermal decomposition kinetics of methyl benzoylformate: Experimental and theoretical study

Article abstract of DOI:10.1002/poc.3396

The kinetics of the gas-phase thermal decomposition of the α-ketoester methyl benzoylformate was carried out in a static system with reaction vessel deactivated with allyl bromide, and in the presence of the free radical inhibitor propene. The rate coefficients were determined over the temperature range of 440-481 °C and pressures from 32 to 80 Torr. The reaction was found to be homogenous, unimolecular and obey a first-order rate law. The products are methyl benzoate and CO. The temperature dependence of the rate coefficient gives the following Arrhenius parameters: log₁₀ k (s^-1) = 13.56 ± 0.31 and E_a (kJ mol^-1) = 232.6 ± 4.4. Theoretical calculations of the kinetic and thermodynamic parameters are in good agreement with the experimental values using PBE1PBE/6-311++g(d,p). A theoretical Arrhenius plot was constructed at this level of theory, and the good agreement with the experimental Arrhenius plot suggests that this model of transition state may describe reasonably the elimination process. These results suggest a concerted non-synchronous semi-polar three-membered cyclic transition state type of mechanism. The most advanced coordinate is the bond breaking C^δ+-^δ-OCH₃ with an evolution of 66.7%, implying this as the limiting factor of the elimination process.

Full text of DOI:10.1002/poc.3396

Research Article
Received: 15 August 2014,
Revised: 14 October 2014,
Accepted: 20 October 2014,
Published online in Wiley Online Library: 7 December 2014
(wileyonlinelibrary.com) DOI: 10.1002/poc.3396
Homogeneous and unimolecular gas-phase
thermal decomposition kinetics of methyl
benzoylformate: experimental and
theoretical study
Yeljair Monascala, Alexis Maldonadoa, José R. Moraa*, Tania Córdovab
and Gabriel Chuchania*
The kinetics of the gas-phase thermal decomposition of the α-ketoester methyl benzoylformate was carried out in a static sys-
tem with reaction vessel deactivated with allyl bromide, and in the presence of the free radical inhibitor propene. The rate
coefficients were determined over the temperature range of 440–481 °C and pressures from 32 to 80 Torr. The reaction was
found to be homogenous, unimolecular and obey a first-order rate law. The products are methyl benzoate and CO. The tem-
perature dependence of the rate coefficient gives the following Arrhenius parameters: log10 k (sꢀ1) = 13.56 0.31 and Ea
(kJ molꢀ1) = 232.6 4.4. Theoretical calculations of the kinetic and thermodynamic parameters are in good agreement with
the experimental values using PBE1PBE/6-311++g(d,p). A theoretical Arrhenius plot was constructed at this level of theory,
and the good agreement with the experimental Arrhenius plot suggests that this model of transition state may describe rea-
sonably the elimination process. These results suggest a concerted non-synchronous semi-polar three-membered cyclic tran-
sition state type of mechanism. The most advanced coordinate is the bond breaking Cδ+---δ-OCH3 with an evolution of 66.7%,
implying this as the limiting factor of the elimination process. Copyright © 2014 John Wiley & Sons, Ltd.
Keywords: DFT calculations; gas-phase thermal decomposition; kinetic; mechanism; methyl benzoylformate
An experimental and theoretical study of the elimination ki-
netics of methyl oxalyl chloride, ClCOCOOCH3, found this com-
INTRODUCTION
The gas-phase elimination of esters of carboxylic acids is gener-
ally known to proceed through a six-membered cyclic transition
state type of mechanism to give the corresponding acid and ole-
fin [(reaction (1)].[1,2] A Cβ―H bond at the alkyl side of the ester is
required for molecular elimination.
pound to decarbonylate by chlorine migration through a
three-membered cyclic transition state as shown in reaction
(3).[4] Semi-empirical PM3 and MP2/6-31G* level of theories
gave good agreement with the experimental results.
(3)
(1)
A year later, a few ethyl esters of 2-oxocarboxylic acids, ethyl
oxamate and ethyl oxanilinate,[5] were proposed to decarbonylate
as the first step of the reaction. The corresponding ethyl ester in-
When a carbonyl group is interposed between substituent Z
and the acid side of the alkyl ester, that is, alkyl esters of 2-oxo-
acids, interesting complex reactions have been found. In this
respect, a work on the gas-phase thermal decomposition of an
α-ketoester, methyl benzoylformate, was reported to yield
benzaldehyde, formaldehyde and CO gas[3] for which the
* Correspondence to: José R. Mora and Gabriel Chuchani, Centro de Química,
Instituto Venezolano de Investigaciones Científicas (I.V.I.C.), Apartado 21827,
Caracas 1020A, Venezuela.
Arrhenius expression was log k = 13.2–222.9 kJmolꢀ1(2.303RT)ꢀ1
.
The mechanism of this reaction was believed to involve a five-
membered cyclic transition state as depicted in reaction (2).
E-mail: morajr2@yahoo.com; chuchani@ivic.gob.ve
a Y. Monascal, A. Maldonado, J. R. Mora, G. Chuchani
Centro de Química, Instituto Venezolano de Investigaciones Científicas (I.V.I.
C.), Apartado 21827, Caracas 1020A, Venezuela
b T. Córdova
(2)
Department of Medicinal Chemistry, College of Pharmacy, University of
Florida, Gainsville, FL, 32610, USA
J. Phys. Org. Chem. 2015, 28 40–46
Copyright © 2014 John Wiley & Sons, Ltd.

GAS-PHASE THERMAL DECOMPOSITION
termediate then went through a parallel decomposition as shown
in reactions (4).
ing to the mechanism explained previously,[8] to give the cor-
responding aldehyde product [reaction (5)].
(4)
(5)
Since a parallel reaction of the gas-phase elimination ki-
netics of ethyl oxamate and ethyl oxanilinate[5] was postu-
Additional kinetic determination of the above-mentioned
work[7] was the inclusion of the compound of reference ethyl
glyoxalate. The presence of H as a substituent gave a parallel
decomposition where the predominant pathway originated
formic acid, ethylene and CO2, through Step 1 and Step 2 of re-
action (5). The small amount of decarbonylation as a parallel pro-
cess occurred through a mechanism similar to reaction (4). The
product intermediate ethyl formate was thought to undergo a
slow decomposition, as depicted in reaction (1) to yield formic
acid and ethylene.
lated,
a theoretical study was undertaken to verify or
support the above suggested mechanisms.[6] In this respect,
the quantum chemical calculations by using ab initio level
of theory were limited to study two competitive mecha-
nisms [(reaction (4) Steps 2 and 3] leading to ethylene
and ethanol formation. The transition state for ethylene for-
mation (Step 2) implied a non-planar six-membered cyclic
structure, where the most important factor was the polari-
zation of O4―C6 bond (breaking of ester bond). The pro-
cess was proposed to be concerted and semi-polar, with
the hydrogen being transferred located halfway between
carbonyl oxygen O5 and C9 (Scheme 1). In the case of eth-
anol formation from ethyl oxamate and ethyl oxanilate
seemed to occur through a planar four-membered cyclic
transition state structure (Step 3). Calculations suggested
this reaction to proceed in a concerted non-synchronous
type of mechanism, where the dominating process was
the breaking of N1―H2 bond (Step 3). NBO charge analysis
suggested the bond polarization of alkyl oxygen–carbon, in
the sense O4 δꢀ―C6 δ+, as a determining factor Step 2,
while N1―H2 bond breaking was considered the limiting pro-
cess in Step 3. Activation parameters were found to be in
reasonable agreement with experimental values at MP2/6-
31G level of theory for these substrates.
A combined experimental and theoretical study in the elimi-
nation kinetic of methyl trifluoropyruvate[9] was recently re-
ported. The degree of polarization between the substituent at
the acid side and the carbon of the C¼O was considered impor-
tant. The products of decomposition of methyl trifluorobutyrate
are methyl trifluoroacetate and CO gas. Using DFT calculations
with different basis sets led to a result in reasonable good agree-
ment with the experimental values for a 1, 2-migration of the
OCH3 group rather than the F3C substituent. The three-
membered cyclic transition state shown in Scheme 2 suggested
that the F3C group, a strong electron-withdrawing substituent,
appears to be weakly polarizable in the 2-oxo-ester. Conse-
quently, it has a lesser migratory aptitude than the OCH3 group.
Reyes et al.[10] investigated the gas-phase elimination kinetics
of several selected 2-oxo-esters. This investigation aimed at
studying the effect of electron-withdrawing substituents Cl,
piperidyl and phenyl in ethyl oxalyl chloride, ethyl piperidine
glyoxylate and ethyl benzoylformate. Ethyl oxalyl chloride
decomposed to ethyl chloroformate and CO gas, and the mech-
anism was thought to be along the same line as reaction (3).
With regard to both ethyl piperidine glyoxylate and ethyl ben-
zoyl formate they underwent to a parallel elimination of decar-
boxylation similar to the mechanism of reaction (5) (Steps 1
and 2), and decarbonylation similar to the mechanism of reac-
tion (4) (Steps 1 and 2).
Further investigation related to the gas-phase elimination
of esters of 2-oxo-acids [reactions (3) and (4)][4,5] containing
electron-withdrawing substituents (Cl, NH2 and NHPh) was
to change it to electron-releasing groups, i.e. methyl (ethyl-
2-oxo-propionate) and isopropyl (3-methyl-2-oxo-butyrate).[7]
The thermal decomposition of these substrates underwent
as in ethyl carboxylic acids[1,2] through a concerted six-
membered cyclic transition state mechanism, producing the
corresponding 2-oxo-acid and ethylene. Under the reaction
conditions the oxo-acid intermediate decarboxylates, accord-
The above-described investigations on the gas-phase elimina-
tion of ethyl ester of 2-oxo-acids indicate that the presence of H
at the Cβ―H bond at the alkyl side gives the corresponding 2-
oxo-acids and ethylene as delineated in organic esters [reaction
Scheme 1. Transition State for ethylene and ethanol formation
J. Phys. Org. Chem. 2015, 28 40–46
Scheme 2. Transition state of methoxy migration
Copyright © 2014 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/poc

Y. MONASCAL ET AL.
(1)]. However, the absence of the β-hydrogen, that is, the re-
placement of ethyl group by a methyl group in carboxylic esters
has rarely shown abstraction of α-H of the CH3 group. In associ-
ation with this fact, the work on the gas-phase elimination of
methyl benzoylformate described in reaction (2)[3] appeared to
be give an interesting and different result when compared with
the reported papers[5–9] showing no abstraction of the α-H in the
CH3 group.
RESULTS AND DISCUSSIONS
The gas-phase elimination of methyl benzoylformate produces
methyl benzoate and CO gas. The reaction took place in a static
system, with allyl bromide deactivated vessel and using propene
as a free radical inhibitor [reaction (6)].
Our experimental conditions have been considered by
Allinger et al.[11] as true gas-phase reaction because, at sub-
atmospheric pressures, deactivated reaction vessels and under
the presence of free radical suppressors are employed. In this
sense, the present work was initiated to determine a joint exper-
imental and theoretical study on the homogeneous, molecular
gas-phase elimination kinetics of methyl benzoylformate. In or-
der to formulate a reasonable mechanism in the decomposition
of this substrate, a theoretical study at DFT computational levels
has been included and also to compare the kinetic and thermo-
dynamic parameters of activation with those values obtained in
the experiments. Another purpose of this research is whether to
support or modify the mechanism formulated in the above-
mentioned work.[3]
(6)
The stoichiometry of reaction (6) requires that the final pres-
sure, Pf, be twice the initial pressure, Po. Experimentally, the aver-
age Pf/Po value at three different temperatures and ten half-lives
was 1.95 (Table 1).
Further examination of the stoichiometry of reaction (6) is
shown in Table 2. The methyl benzoylformate, up to 55% reac-
tion, gave a reasonable good agreement between the extent of
decomposition of the substrate as predicted from pressure mea-
surements and the chromatographic analyses of the product
methyl benzoate.
To examine the effect of surface upon the rate of elimination,
several runs under the presence of propene inhibitor were car-
ried out in a vessel with a surface-to volume ratio of 6 relative
to the normal vessel which equal to one (Table 3). The packed
and unpacked clean Pyrex vessel showed a small effect on the
rates. However, when the packed and unpacked are deactivated
with the product decomposition with allyl bromide, no effect on
the rate coefficients was obtained.
Kinetic runs were performed in the presence of propene to in-
hibit any possible radical reaction (Table 4). Consequently, the
elimination reaction in deactivated vessel had to be carried out
in the presence of at least twice the amount of the inhibitor.
No induction period was observed.
The rate coefficients determined in deactivated vessels and in
the presence of the inhibitor propene are independent of the ini-
tial pressure of the substrate, and the first-order plot is satisfacto-
rily linear to at least 55% decomposition (Table 5). The rate
coefficients are reproducible with a standard deviation not
greater that 5% at a given temperature.
The variations of the rate coefficient with temperature are
listed in Table 6. The results shown in Table 6 lead, by using
the least squares procedure and 90% confidence coefficients,
to the shown Arrhenius expression [Fig. 1, in white circle (o)].
The theoretical plot in black squares (■) is also depicted in Fig. 1,
where a reasonable agreement was obtained between experi-
mental and theoretical plots with nearly similar slope and inter-
cept. This result suggests a good description of the elimination
process by the PBE1PBE/6-311++G(d,p) level of theory explained
in the methodology section. In this sense, the data of the kinetic
and thermodynamic activations parameters described in Table 7
COMPUTATIONAL METHOD AND MODEL
The kinetic and mechanism of the gas-phase elimination of methyl
benzoylformate have been examined as implemented in GAUSS-
IAN 09 package[12] and by means of the electronic structure calcu-
lations using density functional theory (DFT) methods. The
calculations at this level of theory were PBE1PBE,[13]
MPW1PW91,[14,15] M06,[16] M062X[16] and WB97XD[17] with 6-
311++G(d,p) and 6-311G++(2d,2p) basis sets. For the optimization
process, the Berny analytical gradient was employed. The
convergence on the density matrix was 10ꢀ9 atomic units, the
threshold value for maximum displacement was 0.0018 Å and
maximum force was 0.00045 Hartree/Bohr. The optimized reactant
and product structures in the lowest energy conformation were
used to find the Transition State (TS) structure, using Quadratic
Synchronous Transit (QST) protocol. The nature of stationary points
was recognized by frequency calculations. The TS was identified
and characterized by a unique imaginary frequency associated
with the transition vector (TV). Intrinsic reaction coordinate (IRC)
calculations were carried to verify that the transition state con-
nected reactant and products.
Frequency calculations allow estimation of the thermody-
namic quantities, such as zero point vibrational energy (ZPVE),
temperature corrections (E(T)) and absolute entropies (S(T)) at
temperatures of 440.9, 450.4, 459.8, 470.2 and 481.4 °C and pres-
sure values within the experimental range. Scale factors and hin-
dered rotors were not used. These frequency calculations at
different temperatures allow constructing a theoretical Arrhenius
plot in order to compare with the experimental Arrhenius plot.
The activation Gibbs free (ΔG‡) and enthalpy (ΔH‡) energy
were assessed using the following equations 1 and 2.
Table 1. Ratio of final (Pf) to initial pressure (P0)
ΔG‡ ¼ ΔH‡ ꢀ TΔS‡
(1)
Temperature (°C)
P0 (Torr)
Pf (Torr)
Pf/P0
ΔH‡ ¼ V‡ þ ΔZPVE þ ΔEðTÞ
(2)
460.5
470.5
480.4
65.4
64.5
64.7
125.5
127.0
127.5
1.92
1.97
1.97
V‡ is the potential energy barrier, while ΔZPVE and ΔE(T) are the
differences of ZPVE and temperature corrections between the TS
and the reactant, respectively.
Po = initial pressure, Pf = final pressure.
wileyonlinelibrary.com/journal/poc
Copyright © 2014 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2015, 28 40–46

GAS-PHASE THERMAL DECOMPOSITION
Table 2. Stoichiometry of the reaction at 459 °C (732 K)
Parameter
Value
7
Time (min)
3
4
9
13
Reaction (%) (pressure)
Ester product (%) GLC)
15.8 23.1 35.5 45.0 55.5
18.4 25.5 37.1 48.5 55.8
Table 3. Homogeneity of the reaction at 459.6 °C (732 K)
S/V (cmꢀ1
)
104 k (sꢀ1 (a)
)
104 k (sꢀ1 (b)
)
1
6
12.85 0.49
11.65 0.21
10.02 0.28
10.45 0.49
aClean Pyrex vessel.
bDeactivated Pyrex vessel.
Figure 1. Arrhenius plots of gas-phase elimination of methyl
benzoylformate, in white circles (o) experimental and in black squares
(■) theoretical at PBE1PBE/6-311++G(d,p) level of theory
suggest an elimination process through a three-membered tran-
sition state (Scheme 3). These results allow the present work to
approach to quantum chemical calculations for the consider-
ation and interpretation of a rational mechanism of elimination
of methyl benzoylformate depicted as in reaction (6).
Table 4. Effect of propene inhibitor on rates at 460.1 °C
(733 K)
Po (Torr)
Pi (Torr)
Pi/Po
104 k1 (sꢀ1
)
THEORETICAL RESULTS
79.7
70.4
70.6
68.2
42.1
30.0
0
66
131
136
120
130
0
22.60
10.50
9.69
10.10
10.40
10.30
1.1
1.9
2.0
2.9
4.3
Kinetic and thermodynamic parameters
According to the experimental results of the gas phase elimina-
tion of methyl benzoylformate producing methyl benzoate and
CO gas, this work aimed at examining by quantum chemical cal-
culations the mechanism of this process. Based on reaction (7)
the proposed mechanism through a three-membered cyclic
Po = pressure of substrate. Pi = pressure inhibitor.
Table 5. Invariability of rate coefficient with initial pressure at 459.7 °C (732.9 K)
Parameters
P0 (Torr)
Value
55.4
32.6
9.65 0.19
46.2
9.81 0.32
70.6
9.69 0.23
80.1
10.10 0.50
104 k1 (sꢀ1
)
10.10 0.38
Table 6. Experimental and theoretical temperatures dependence of the rate coefficients
Parameters
Value
Temp. (°C)
440.9
3.53 0.06
2.44
450.4
5.79 0.21
3.97
459.8
10.02 0.28
6.74
470.2
15.90 0.65
11.27
481.4
29.20 0.71
19.54
(a) 104 k1 (sꢀ1
(b) 104 k1 (sꢀ1
)
)
Rate equation: (a) log k1 (sꢀ1) = (13.56 0.31) ꢀ (232.64 4.38) kJ molꢀ1(2.303RT)ꢀ1 (r = 0.9994).
(b) log k1 (sꢀ1) = (13.30 0.14) ꢀ (231.32 2.01) kJ molꢀ1(2.303RT)ꢀ1 (r = 0.9998).
(a) Experimental; (b) theoretical by using the frequentials of PBE1PBE/6-311++G(d,p) level of theory.
J. Phys. Org. Chem. 2015, 28 40–46
Copyright © 2014 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/poc

Y. MONASCAL ET AL.
Table 7. Kinetic and thermodynamic parameters
Ea (kJ/mol)
232.64
Log A (sꢀ1
)
ΔH‡ (kJ/mol)
ΔS‡ (J/mol K)
ꢀ1.07
ΔG‡ (kJ/mol)
13.56
226.50
227.30
ΔH‡, ΔS‡ and ΔG‡ were estimated at 459.8 °C.
transition state structure was studied. Table 8 describes the ther-
modynamic and kinetic parameters that have been determined
theoretically together with the experimental results.
theoretical Arrhenius (Fig. 1, black squares) is parallel to the experi-
mental line, resulting in the same energy of activation.
Transition state and mechanism
(7)
Structural parameters for R, TS and P are given in Table 9; atom
numbers for discussion are shown in Scheme 3. Geometrical pa-
rameters demonstrate the breaking of O1―C2 bond is the most
advanced coordinate in the elimination process, with change in
atom distance from 1.325 Å to 1.870 Å from R to TS. The C2―C3
and C3―C4 bond distances show small increase from R to TS in
the reaction path. The distance C3―O1 diminishes from 2.362 Å
to 1.677 Å from R the TS. The imaginary frequency associate with
the TS structure is 507.64 cmꢀ1, thus confirming the observation
of a saddle point with the movement in the reaction coordinate.
We have obtained a reasonable agreement between experi-
mental and theoretical results for calculated enthalpies of activa-
tion, and consequently the energies of activation with all DFT
methods used. The calculated entropy of activation from
M062x/6-311++G(2d,2p) was closer to the experimental value;
however, the enthalpy of activation of PBE1PBE/6-311++G(d,p)
gave better results when compared to experimental. For
PBE1PBE/6-311++G(d,p), the departure in the calculated entropy
of 2.41 J/mol K is an acceptable agreement considering the ex-
perimental error. PBE1PBE/6-311++G(d,p) was chosen for the
subsequent analysis. Figure 2 shows the optimized structures
for the Reactant (R), Transition State (TS) and Products (P) at
PBE1PBE/6-311++G(d,p) level of theory.
Intrinsic reaction coordinate (IRC) calculations were carried out
to verify the Transition State structures. Figure 3 shows the IRC
reaction profile, illustrating how the transition state connects
with reactant and products.
A theoretical Arrhenius was generated from the frequency calcu-
lations at PBE1PBE/6-311++G(d,p) level of theory at the five different
temperature reported in the experimental work (Table 6). The
NBO charges
The changes in electronic structures through the reaction path were
examined using natural bond orbital (NBO) analysis. NBO analysis is
based on a method for optimally transforming a given wave func-
tion into localized form. The NBO charges of atoms involved in the
mechanisms are reported in Table 10. Atom numbers and the per-
cent change in charges are given in Scheme 4.
The most important change in the charge distributions is ob-
served for the methoxy oxygen atom O1, from ꢀ0.516 to
ꢀ0.624 from R to TS, respectively. This result implies an increase
in the negative charge by 20.9% in the TS. At carbon C3 there is
an increase in the positive charge of 11.0%. No significant
change was observed in the atomic charge at C2 (2.4%)
(Scheme 4).
Natural bond orbital (NBO) analysis
NBO bond order calculations have been used to describe the reaction
coordinate of the gas-phase thermal decomposition of methyl
benzoylformate (Table 11).[18–20] Wiberg bond indexes Bi[21] were
Scheme 3. Transition state of CO elimination
Table 8. Thermodynamic and Arrhenius parameters for the Gas phase elimination reaction of methyl benzoylformate at 460 °C
Method
Ea (kJ/mol)
log A (sꢀ1
)
ΔH≠ (kJ/mol)
ΔS≠ (J/mol K)
ΔG≠ (kJ/mol)
Experimental
232.64
233.09
233.93
233.83
234.84
229.96
229.77
235.87
237.63
238.33
239.23
13.56
13.44
13.44
13.47
13.49
13.18
13.04
13.39
13.55
13.05
12.97
226.50
227.00
227.84
227.73
228.75
223.87
223.67
229.78
231.53
232.24
233.14
ꢀ1.07
ꢀ3.48
ꢀ3.45
ꢀ2.93
ꢀ2.42
ꢀ8.47
ꢀ10.99
ꢀ4.42
ꢀ1.38
ꢀ10.96
ꢀ12.34
227.30
229.55
230.36
229.88
230.52
230.08
231.73
233.02
232.54
240.27
242.18
PBE1PBE/6-311++G(d,p)
PBE1PBE/6-311++G (2d,2p)
MPW1PW91/6-311++G(d,p)
MPW1PW91/6-311++G(2d,2p)
M06/6-311++G(d,p)
M06/6-311++G(2d,2p)
M062x/6-311++G(d,p)
M062x/6-311++G(2d,2p)
WB97xd/6-311++G(d,p)
WB97xd/6-311++G(2d,2p)
wileyonlinelibrary.com/journal/poc
Copyright © 2014 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2015, 28 40–46

GAS-PHASE THERMAL DECOMPOSITION
Figure 2. Optimized structures for reactant (R), transition state (TS) and products (P) in the gas phase elimination of C6H5COCOOCH3 at PBE1PBE/6-311
++G(d,p) level of theory
Scheme 4. Charge percentage of atoms in methyl benzoylformate
Table 11. Wiberg bond index, Bi, of reactant (R), transition
state (TS) and products (P) and synchronicity (Sy) of CO elim-
ination from methyl benzoylformate at PBE1PBE/6-311++G(d,
p). T = 460 °C
Figure 3. IRC profile for CO formation from methyl benzoylformate at
PBE1PBE/6-311++G(d,p) level of theory. T = 460 °C
O1―C2
C2―C3
C3―O1
Sy
BiR
1.079
0.359
0.000
66.7
0.907
0.791
0.001
12.8
0.016
0.500
1.032
0.651
BTi S
BiP
Table 9. Structural parameters of reactant (R), transition
state (TS) and products (P) for CO elimination from methyl
benzoylformate in gas phase at 460 °C, obtained from
PBE1PBE/6-311++G(d,p) calculations
%Ev
47.7
proposed by Moyano et al.[23]; the equations involved are described in
the literature.[5] The evaluated bonds were O1―C2, O1―C2 and
O1―C2.
Atomic length (Å)
O1―C2
1.325
1.870
5.043
C2―C3
1.534
1.596
4.400
C3―O1
2.362
1.677
1.341
C3―C4
1.481
1.493
1.485
The most advanced reaction coordinate is the O1―C2 bond break-
ing (66.7%), followed by the C3―O1 bond formation (47.7%). The syn-
chronicity parameter defined by Moyano for concerted reactions
varies from Sy = 1 for synchronous reactions to Sy =0 for non-
synchronous processes. In this reaction, the Sy value of 0.651 implies
a non-synchronous process for the mechanism describing the gas-
phase elimination process of methyl benzoylformate shown in reac-
tion (7). These results are consonant with the previously reported ther-
mal decomposition of methyl trifluoropyruvate[9] where the
decomposition products are methyl trifluoroacetate and CO gas
(Scheme 2).
Our findings contrast with a previous study on the thermal decom-
position of methyl benzoylformate reported in the literature,[3] regard-
ing product formation and the stoichiometry. The cited work reports
three decomposition products, e.g. benzaldehyde, formaldehyde
and carbon monoxide, while in this work two products are obtained:
methyl benzoate and CO. To further investigate this issue, we carried
out theoretical calculations of their above-reported mechanism (3) at
PBE1PBE/6-311++G(d,p) level of theory. Compared to the reported ex-
perimental barrier 222.9 kJ/mol,[3] the calculated parameter for reac-
tion (2) was significantly higher, 308.0 kJ/mol. The discrepancy
between the results described in (3)[3] with ours may probably be
due to the different conditions of experimental works. In the work re-
ported in ref. 3, methyl benzoylformate was in a very dilute solution of
acetonitrile in a 9-ml vacuum sealed uncoated tube, whereas our
R
TS
P
Imaginary frequency (cmꢀ1): 507.64.
Table 10. NBO charges of reactant (R), transition state (TS)
and products (P) for CO elimination from methyl
benzoylformate at PBE1PBE/6-311++G(d,p). T = 460 °C
NBO charges
O1
C2
C3
C4
R
TS
P
ꢀ0.516
ꢀ0.624
ꢀ0.545
0.739
0.757
0.488
0.480
0.533
0.799
ꢀ0.167
ꢀ0.130
ꢀ0.167
estimated for the reactant (R), transition state (TS) and products (P) by
using the natural bond orbital (NBO) program[22] as implemented in
Gaussian 09.[12] The synchronous character of concerted reaction
mechanism can be accounted through the synchronicity (Sy) concept
J. Phys. Org. Chem. 2015, 28 40–46
Copyright © 2014 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/poc

Y. MONASCAL ET AL.
experiments were carried out using the substrate pure, under homog-
enous condition, in a static system at sub-atmospheric pressure, and
which has been depicted in figures as reported in previous
papers.[26,27]
suggests
a
concerted non-synchronous semi-polar three-
membered cyclic transition state type of mechanism. The most ad-
vanced coordinate is the bond breaking Cδ+---δ-OCH3 with an evolu-
tion of 66.7%, implying this as the limiting factor of the elimination
process.
Our initiative to estimate the theoretical Arrhenius by means
of calculating the frequencies from the chosen PBE1PBE/6-
311++G(d,p) level of theory at the five different temperature re-
ported in the experimental work is described.
EXPERIMENTAL SECTION
The substrate methyl benzoylformate (98%) was bought from Al-
drich. The purity of the substrate and product was determined by
GC/MS/MS (Saturn 2000, Varian), capillary column DB—5MS,
30 mm × 0.250 mm. id. 0.25 μm. The quantitative chromatographic
analysis of methyl benzoate was determined by using a Gas Chro-
matograph Hewlett Packard 5710-A with a column 3% Versamid
900/Chromosorb G, 3.5 mts. The identification of the product was
made by comparing chromatogram of the reaction product with
true authentic sample of methyl benzoate bought from Aldrich
and in a GC–MS (Saturn 2000, Varian 3600X with a DB-5MS capillary
column 30 m × 0.25 mm. i.d., and 0.25 μm).
REFERENCES
[1] R. Taylor, The Chemistry of Functional Group. Suplementary Volume B,
Acid Derivatives (ed: S. Patai), Wiley, Chichester, 1979, Chapter 15,
859–914.
[2] K. A. Holbrook, The Chemistry of Acid Derivatives. Volume 2. Vapor
and Gas Phase Reaction of Carboxylic Acids and their Derivatives
(ed: S. Patai), Wiley, Chichester, 1992, Chapter 12, 703–746.
[3] N. A. Al-Awadi, O. M. E. El-Dusouqui, Int. J. Chem. Kinet. 1997, 29,
295–298.
Kinetics
[4] T. Cordova, A. Rotinov, G. Chuchani, J. Phys. Org. Chem. 2004, 17,
148–151.
The experiments were carried out in vessel seasoned or deactivated
by the decomposition products of allyl bromide, in a static sys-
tem[24–27] and in the presence of the free radical suppressor propene
(100%, Air Product Co.). The rate coefficients were determined man-
ometrically (mercury manometer). The temperature was controlled
within 0.2 °C with a SHINKO DIC-PS 23TR resistance thermometer
controller with a calibrated Iron Constantan thermocouple. The read-
ing temperature was within 0.1 °C with the Iron-Constantan thermo-
couple attached to a Digital Multimeter Omega 3465B. The reaction
vessel showed no temperature gradient at different points, and the
substrate was injected directly (≈0.05–0.1 ml) into the reaction vessel
through a silicone rubber septum.
[5] E. V. Chacin, M. Tosta, A. Herize, R. M. Domínguez, Y. Alvarado, G.
Chuchani, J. Phys. Org. Chem. 2005, 18, 539–545.
[6] J. R. Mora, M. Loroño, T. Cordova, G. Chuchani, J. Phys. Org. Chem.
2006, 19, 503–511.
[7] A. Reyes, R. M. Dominguez, M. Tosta, A. Herize, G. Chuchani, Int. J.
Chem. Kinet. 2007, 39, 268–275.
[8] R. Taylor, Int. J. Chem. Kinet. 1987, 19, 709–713.
[9] M. M. Tosta, J. R. Mora, T. Córdova, G. Chuchani, J. Phys. Chem. A
2010, 114, 789–797.
[10] A. Reyes, R. M. Dominguez, M. Tosta, A. Herize, G. Chuchani, J. Phys.
Org. Chem. 2011, 24, 74–82.
[11] D. H. Wurtz, N. L. Allinger, J. Org. Chem. 1977, 42, 698–703.
[12] 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, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao,
H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro,
M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R.
Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S.
S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E.
Knox, 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, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth,
P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas,
J. B. Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox, Gaussian 09, Revi-
sion A.02, Gaussian, Inc., Wallingford CT, 2009.
[13] C. Adamo, V. Barone, J. Chem. Phys. 1999, 110, 6158–6169
[14] Y. Zhao, O. Tishchenko, D. G. Truhlar, J. Phys. Chem. B, 2005, 109,
19046–19051
[15] C. Adamo, V. Barone, J. Chem. Phys. 1998, 108, 664–675
[16] Y. Zhao, D. G. Truhlar, Acc. Chem. Res. 2008, 41(2), 157–167.
[17] J.-D. Chai, M. Head-Gordon, Phys. Chem. Chem. Phys. 2008, 10, 6615–
6620.
[18] G. Lendvay, J. Phys. Chem. 1989, 93, 4422–4429.
[19] A. E. Reed, R. B. Weinstock, F. Weinhold, J. Chem. Phys. 1985, 83,
735–746.
[20] A. E. Reed, L. A. Curtiss, F. Weinhold, Chem. Rev. 1988, 88, 899–926.
[21] K. B. Wiberg, Tetrahedron 1968, 24, 1083–1095.
[22] E. D. Glendening, A. E. Reed, J. E. Carpenter, F. Weinhold, NBO Ver-
sion 3.1.
[23] A. Moyano, M. A. Pericás, E. Valenti, J. Org. Chem. 1989, 54, 573–582.
[24] A. Maccoll, J. Chem. Soc. 1955, 965–973.
[25] E. S. Swinbourne, Aust. J. Chem. 1958, 11, 314–330.
[26] R. M. Dominguez, A. Herize, A. Rotinov, A. Alvarez-Aular, G. Visbal, G.
Chuchani, J. Phys. Org. Chem. 2004, 17, 399–408.
CONCLUSIONS
The kinetics and mechanism for the thermal decomposition of
methyl benzoylformate in the gas phase were investigated both ex-
perimental and theoretically. The reaction was found to be homog-
enous, unimolecular and obeys a first-order rate law with Arrhenius
equation log10 k (sꢀ1) = 13.56 0.31 and Ea (kJ molꢀ1) = 232.6 4.4.
The products of methyl benzoylformate thermal decomposition
are methyl benzoate and carbon monoxide. This observation di-
verges from previous work findings reporting benzaldehyde, form-
aldehyde and carbon monoxide product formation,[3] and it may be
attributed to the differences in the experimental conditions. To ex-
plain the product formation and to propose a reasonable mecha-
nism for this reaction, we carried out theoretical calculations of
the kinetics and thermodynamics parameters at several DFT levels.
We obtained a reasonable agreement between experimental and
theoretical results for calculated enthalpies of activation and conse-
quently, energies of activation with all DFT methods used. With
M062x/6-311++G(2d,2p) calculated entropy of activation was closer
to the experimental value; however, for the enthalpy of activation
PBE1PBE/6-311++G(d,p) gave better results, compared to experi-
mental, and the departure in the calculated entropy of 2.41 J/mol K,
which is an acceptable agreement taking into account the experi-
mental error. A theoretical Arrhenius plot was generated at
PBE1PBE/6-311++G(d,p) using the five temperatures reported in
the experimental work, showing the calculated and experimental
lines nearly parallel, with the result in the same energy of activation.
Analysis of NBO charges, bond orders and synchronicity parameters
[27] L. Espitia, R. Meneses, R. M. Dominguez, M. Tosta, A. Herize, J. Lezama, J.
Lafont, G. Chuchani, Int. J. Chem. Kinet. 2009, 41, 145–152.
wileyonlinelibrary.com/journal/poc
Copyright © 2014 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2015, 28 40–46