Pyrolysis and Oxidation of Anisole near 1000 K
J. Phys. Chem. A, Vol. 101, No. 18, 1997 3313
TABLE 4: CHEMACT Input Parameters for Analysis of
decomposes readily via cleavage of the C5H5-CH3 bond (Ea
) 67.5 kcal). In contrast, the activation energy for analogous
decomposition of 1- or 2-CH3C5H5, estimated from thermody-
namics, is approximately 95 kcal, rendering these reactions
negligible. The pyrolysis data in Figure 9 suggest that at the
later residence times methylcyclopentadiene persists in a pseudo-
steady-state. In order to reproduce the observed stability of
methylcyclopentadiene in the anisole pyrolysis system, it is
necessary to include CH3C5H5 isomerization reactions in the
model. Rate parameters for the two isomerization reactions (5-
to 1-CH3C5H5 and 1- to 2-CH3C5H5) are derived from ∆Hq and
∆Sq, estimated from semiempirical molecular orbital calculations
using MOPAC40 (Table 6). The results are consistent with the
observations of McClean and Haynes,39 who showed that
5-CH3C5H5 rearranged “very rapidly” to 1-CH3C5H5, which then
rearranged “more slowly” to 2-CH3C5H5.
Prediction of total phenolics (Figure 12) agrees well with
experiment though the split between phenol and cresols
demonstrates poor agreement with cresols overpredicted by a
factor of 2; at 98 ms total phenolics are comprised of
approximately equal parts phenol and cresols while the model
predicts essentially only cresols. An accompanying underpre-
diction of methane and ethane (Figure 13) is consistent with
this result; methyl groups are trapped in excess cresols,
unavailable to reactions forming methane and ethane. All
attempts to improve prediction of these species by opti-
mizationswithin realistic limitssof rate parameters were unsuc-
cessful.
C6H5O + CH3
k
Aa
Ea (kcal/mol)
source
1
-1
2
3
4
-4
5
1.04E+14
5.66E+17
7.40E+11
3.26E+15
5.67E+13
1.75E+12
1.40E+16
0.0
57.4
32.5
67.9
39.0
55.6
99.8
b
c
d
e
f
c
g
ν ) 1141 cm-1
σ ) 5.10 Å
ꢀ/k ) 595 K
h
i
i
a Units: s-1 (except A1 ∼ cm3 mol-1 s-1). b A1 set equal to 10 times
A factor for C6H5O + CH3 f C6H5OCH3 according to electron density
arguments (see text). c Calculated from forward rate parameters on the
basis of thermodynamics. d A2 given as A factor for C6H5O f C5H5 +
CO. Ea,2 equal to ∆HR (17 kcal) plus intrinsic activation energy of
phenoxy reaction (16 kcal), reduced by 0.5 kcal/mol. e A3 and Ea,3
calculated from thermodynamics with A-3 ) 1.0E+14. f A4 determined
from transition state theory, A ) (ekT/h) exp(∆Sq/R). Transition state
was assumed to be tight, i.e., ∆Sq ≈ 0. Ea,4 estimated by analogy with
g
tautomerization reaction 2-pyridone f 2-hydroxypyridine.56 Ea,5 and
A5 by analogy with C6H5CH3 f C6H5 + CH3.54 h Vibrational
frequencies estimated by Lin and Lin37 on the basis of the spectra of
p-benzoquinone and toluene. i Lennard-Jones parameters estimated by
methods detailed in Reid et al.57
CH3C6H4OH, must be considered as well. The multichannel
unimolecular reaction of (H)(CH3)C6H4O was treated using the
CHEMACT companion code DISSOC. The relative importance
of the direct (prior to stabilization of the adduct) product
formation routes with respect to the indirect (following stabiliza-
tion) routes was determined by flux analysis, discussed below.
Flux analysis is simply the determination of the “flux” or
rate (mol cm-3 s-1) of each reaction at each time step. A flux
diagram (Figure 11) is then a convenient pictorial representation
of the flow of reaction intermediates. In Figure 11 each arrow
is representative of a given reaction in the indicated direction;
the forward and reverse of a reaction are depicted individually.
The magnitude of the flux of a reaction is indicated by the
weight of its arrow. This analysis was performed for t ) 15
ms, the residence time corresponding to ≈50% anisole conver-
sion and steep gradients in species profiles. Note that every
reaction in Table 2 is not represented in Figure 11. Only the
most significant reactions have been included.
In particular, reactions of phenoxy and methyl radicals with
the intermediate o- and p-methylcyclohexadienones were in-
vestigated. These molecules possess a weak, abstractable
hydrogen atom and are also polar, so abstraction could be a
fast (low Ea) process. Mulcahy and Williams,24 in their study
of methyl radical reactions with phenol, observed larger yields
of CH4 than could be accounted for by abstraction from phenol.
They attributed their surplus CH4 to an unusually rapid reaction
between CH3 and methylcyclohexadienone intermediates. Nev-
ertheless, adoption of large rate constants (as high as 1012 cm3
mol-1 s-1) for the reactions
C6H5O + (H)(CH3)C6H4O f C6H5OH + CH3C6H4O
CH3 + (H)(CH3)C6H4O f CH4 + CH3C6H4O
Immediately, Figure 11 reveals that at t ) 15 ms the indirect
routes (i.e., via stabilized (H)(CH3)C6H4O) are more significant
sources of 5-CH3C5H5 and CH3C6H4OH than the direct reac-
tions. The flux of cresol through isomerization of the stabilized
adduct is, on average over the 100 ms reaction time, roughly 5
times the flux by way of the direct reaction. The flux of
methylcyclopentadiene through decomposition of (H)(CH3)-
C6H4O is on average an order of magnitude greater than the
flux from phenoxy and methyl directly. Together, these two
reaction pathways account for roughly 90% of the total
methylcyclopentadiene formed. Recombination of methyl and
cyclopentadienyl radicals accounts for the remainder.
Formation of methylcyclopentadiene through addition of CH3
to either C5H5 or C6H5O necessarily yields the 5-isomer. The
1- and 2-forms are derived via sigmatropic rearrangement(s).38,39
A [1,5] sigmatropic hydrogen shift yields 1-CH3C5H5, which
may itself undergo a [1,5] shift to yield 2-CH3C5H5. Two GC
peaks have been identified by MS as isomers of methylcyclo-
pentadiene.21 On the basis of equilibrium considerations,39 these
peaks are assigned to 1- and 2-CH3C5H5. While thermodynamic
properties do not differ vastly among isomers, the kinetic
stability of the 5-isomer does contrast significantly with that of
the 1- and 2-forms. Specifically, at 1000 K 5-CH3C5H5
did not reproduce the present experimental phenol and methane
profiles. In the present system, [(H)(CH3)C6H4O] is at best 2
orders of magnitude less than [CH3] or [C6H5O], and thus the
above reactions cannot compete with reactions of phenoxy and
methyl with one another.
As shown in Figure 11, recombination of phenoxy with H is
virtually the sole source of phenol in the present model. The
rate constant for the reaction (2 × 1014 s-1, He et al.41) is large.
However, formation of phenol via this recombination is limited
by the deficiency of atomic hydrogen in this system.
An attempt was made to model the production of phenol
directly from anisole. Arends18 proposed the addition of H to
anisole at the ortho or para position of the aromatic nucleus
followed by methyl elimination to yield (keto)phenol. In the
current study a phenol formation pathway not involving atomic
hydrogen was sought. Unimolecular elimination of the meth-
ylene singlet from anisole was postulated. Insertion of :CH2
into O-H bonds to form methyl ethers is known to occur.42 In
accordance with microreversibility the reverse reaction, expul-
sion from O-CH3, must be possible. The reaction
C6H5OH + :CH2 f C6H5OCH3
(8)