Synthesis, characterization, and reactivity of the m-xylylene anion in the gas phase. The enthalpy of formation of m-xylylene

Article abstract of DOI:10.1021/ja002138n

The enthalpy of formation of m-quinodimethane (m-xylylene) has been determined using two different gas-phase approaches. The first involves combining the electron affinity of the biradical (0.919 ± 0.008 eV, Wenthold, P. G.; Kim, J. B.; Lineberger, W. C. J. Am. Chem. Soc. 1997, 119, 1354) with the acidity of the 3-methylbenzyl radical. The acidity of the 3-methylbenzyl radical was determined by bracketing the proton affinity of m-xylylene ion, prepared from the reaction of m-xylene and atomic oxygen ions in a flowing afterglow triple quadrupole apparatus. Deuterium labeling and reactivity studies show that 75% of the H₂⁺ abstraction product formed in the reaction is the m-xylylene ion, with the rest being 3-methyl-α,n-dehydrotoluene ions resulting from [α, ring] hydrogen abstraction. The m-xylylene negative ion underwent reactions similar to those observed for 3-methylbenzyl anion and for other open-shell ions. The m-xylylene ion also reacted with mesitylene by H₂⁺ transfer. The acidity of the 3-methylbenzyl radical was 382.5 ± 2.1 kcal/mol, which leads to an enthalpy of formation of m-xylylene of 80.1 ± 3.8 kcal/mol. The enthalpy of formation of m-xylylene was also determined to be 81.2 ± 3.0 kcal/mol using collision-induced dissociation (CID) threshold energy measurements with the 3-(chloromethyl)benzyl ion. The measured enthalpies of formation indicate a second C-H bond dissociation energy in m-xylene of 90.7 ± 2.9 kcal/mol, only slightly higher than the first (90.1 ± 1.7 kcal/mol). The fact that the strength of the second C - H bond in the 3-methylbenzyl radical is essentially the same as that in m-xylene indicates that the interaction between the two unpaired electrons is negligible, as would be expected for a triplet biradical.

Full text of DOI:10.1021/ja002138n

J. Am. Chem. Soc. 2000, 122, 11203-11211
11203
Synthesis, Characterization, and Reactivity of the m-Xylylene Anion
in the Gas Phase. The Enthalpy of Formation of m-Xylylene
Loubna A. Hammad and Paul G. Wenthold*
Contribution from the The Department of Chemistry, Purdue UniVersity,
West Lafayette, Indiana 47907-1393
ReceiVed June 14, 2000
Abstract: The enthalpy of formation of m-quinodimethane (m-xylylene) has been determined using two different
gas-phase approaches. The first involves combining the electron affinity of the biradical (0.919 ( 0.008 eV,
Wenthold, P. G.; Kim, J. B.; Lineberger, W. C. J. Am. Chem. Soc. 1997, 119, 1354) with the acidity of the
3-methylbenzyl radical. The acidity of the 3-methylbenzyl radical was determined by bracketing the proton
affinity of m-xylylene ion, prepared from the reaction of m-xylene and atomic oxygen ions in a flowing afterglow
+
triple quadrupole apparatus. Deuterium labeling and reactivity studies show that 75% of the H2 abstraction
product formed in the reaction is the m-xylylene ion, with the rest being 3-methyl-R,n-dehydrotoluene ions
resulting from [R, ring] hydrogen abstraction. The m-xylylene negative ion underwent reactions similar to
those observed for 3-methylbenzyl anion and for other open-shell ions. The m-xylylene ion also reacted with
mesitylene by H2+ transfer. The acidity of the 3-methylbenzyl radical was 382.5 ( 2.1 kcal/mol, which leads
to an enthalpy of formation of m-xylylene of 80.1 ( 3.8 kcal/mol. The enthalpy of formation of m-xylylene
was also determined to be 81.2 ( 3.0 kcal/mol using collision-induced dissociation (CID) threshold energy
measurements with the 3-(chloromethyl)benzyl ion. The measured enthalpies of formation indicate a second
C-H bond dissociation energy in m-xylene of 90.7 ( 2.9 kcal/mol, only slightly higher than the first (90.1 (
1.7 kcal/mol). The fact that the strength of the second C-H bond in the 3-methylbenzyl radical is essentially
the same as that in m-xylene indicates that the interaction between the two unpaired electrons is negligible, as
would be expected for a triplet biradical.
m-Xylylene (m-quinodimethane, 1)1 has been the subject of
considerable interest for several years. Because it is a non-
Kekule molecule with a triplet ground state, 1 has attracted
significant attention as a building block for paramagnetic
materials. Subsequently, there have been many experimental
and theoretical studies into the structure and reactivity of this
biradical and its derivatives. In 1915, Schlenk2 reported the first
derivative of m-xylylene, 2, as an isolable crystalline substance.
The electron spin resonance (ESR) spectrum of Schlenk’s
biradical was determined in 1970,3,4 and indicated a ground-
state triplet. The ESR spectrum of 1, obtained in 1983 by Platz
and co-workers,5 confirmed the triplet ground state in the
unsubstituted biradical. These observations agree with the
theoretical predictions based on a simple electron-repulsion
model6 and high-level molecular orbital calculations.7,8
biradical. Goodman and Berson examined the reactions of 1
with organic reagents.9-11 They observed that 1 reacts with
conjugated dienes to give indanes and m-cyclophenes,11 and
reactions with deuterium-labeled starting materials indicated a
symmetric intermediate.9 Kentta¨maa and co-workers have
described the gas-phase reactivity of charged derivatives of
m-xylylene, including the 3,5- and 2,6-bis-methylenepyridinium
ions (3a and 3b, respectively).12 They found contrasting
reactivity for the two isomers, and attributed it to different spin
states of the ions. Whereas ion 3a undergoes hydrogen abstrac-
tion reactions, consistent with the behavior expected for a triplet,
3b yields reaction products similar to those obtained for the
related closed-shell singlet cation, the 2-methyl-6-methylenepy-
ridine cation, as would be expected for a singlet biradical. The
inferred spin states are consistent with those predicted for 3a
and 3b by Dougherty and co-workers.13 The reactivity of other
m-xylylene derivatives has also been investigated.14
Recently, Wenthold et al.15 reported the photoelectron
spectrum of the m-xylylene negative ion, 1-. In the spectrum,
3
they observed the B2 ground state and singlet states, corre-
1
1
sponding to the A1 and B2 states, which were 9.6 ( 0.2 and
<21.5 kcal/mol higher in energy, respectively. Moreover, they
found the electron affinity (EA) of 1 to be 0.919 ( 0.008 eV.
Studies have been carried out to probe the reactivity of the
(1) Platz, M. S. Quinodimethanes and Related Diradicals. In Diradicals;
Borden, W. T., Ed.; John Wiley & Sons: New York, 1982; Vol. 1, p 195.
(2) Schlenk, W.; Brauns, M. Ber. Dtsch. Chem. Ges. 1915, 48, 661.
(3) Kothe, G.; Denkel, K. H.; Summermann, W. Angew. Chem., Int. Ed.
Engl. 1970, 9, 906.
(4) Luckhurst, G. R.; Pedulli, G. F. J. Chem. Soc. B 1971, 1971, 1971.
(5) Wright, B. B.; Platz, M. S. J. Am. Chem. Soc. 1983, 105, 628.
(6) Borden, W. T.; Davidson, E. R. J. Am. Chem. Soc. 1977, 99, 4587.
(7) Fort, R. C., Jr.; Getty, S. J.; Hrovat, D. A.; Lahti, P. M.; Borden, W.
T. J. Am. Chem. Soc. 1992, 114, 7549.
(9) Goodman, J. L.; Berson, J. A. J. Am. Chem. Soc. 1984, 106, 1867.
(10) Goodman, J. L.; Berson, J. A. J. Am. Chem. Soc. 1985, 107, 5409.
(11) Goodman, J. L.; Berson, J. A. J. Am. Chem. Soc. 1985, 107, 5424.
(12) Nelson, E. D.; Thoen, K. K.; Kentta¨maa, H. I. J. Am. Chem. Soc.
1998, 120, 3792.
(13) West. A. P., J.; Silverman, S. K.; Dougherty, D. A. J. Am. Chem.
Soc. 1996, 118, 1452.
(14) Gajewski, J. J.; Paul, G. C.; Chang, M. J.; Gortva, A. M. J. Am.
Chem. Soc. 1994, 116, 5150.
(8) Kato, S.; Morokuma, K.; Feller, D.; Davidson, E. R.; Borden, W. T.
J. Am. Chem. Soc. 1983, 105, 1791.
(15) Wenthold, P. G.; Kim, J. B.; Lineberger, W. C. J. Am. Chem. Soc.
1997, 119, 1354.
10.1021/ja002138n CCC: $19.00 © 2000 American Chemical Society
Published on Web 11/01/2000

11204 J. Am. Chem. Soc., Vol. 122, No. 45, 2000
Hammad and Wenthold
reaction using d3-labeled m-xylene, as shown in eq 4.
This value can be used to calculate the C-H bond dissociation
energy (BDE) in 3-methylbenzyl radical (i.e., the second C-H
bond dissociation energy in m-xylene) using the relationship
shown in eq 1. However, the quantity ∆Hacid(3-CH2C6H4CH3)
has not been measured, so the second BDE in m-xylene is not
known.
For this reaction, they observed the transfer of HD+ but did
not see any D2+, and concluded that the abstraction takes place
from each of the two methyl groups, forming the m-xylylene
ion. Surprisingly, although it has been almost 25 years since
Bruins et al.22 first reported the formation of 1-, the ion has
not been fully characterized or studied. Among the information
that could be obtained from such a study is the proton affinity
of the ion [i.e., ∆Hacid(3-CH2C6H4CH3)], which can be used
along with the recently determined EA of 1 to calculate the
second C-H BDE in m-xylene.
In this work, we describe the measurement of the enthalpy
of formation of m-xylylene from which the second C-H BDE
in m-xylene can be calculated. Two approaches are used to
determine the enthalpy of formation of 1. First, we use proton-
transfer reactions to bracket the proton affinity of 1-. However,
because these experiments are complicated by the presence of
competing reaction pathways, we also measure ∆Hf,298(1) using
the collision-induced dissociation (CID) threshold approach
developed by Squires and co-workers.16,23 We find that the
second C-H BDE in m-xylene agrees with the first, within error,
indicating little interaction between the electrons, as expected
for a triplet biradical.24,25
BDE(3-CH2C6H4CH2sH) ) EA(1) +
∆Hacid(3-CH2C6H4CH3) - IP(H) (1)
The second C-H BDEs for systems such as these are
important because the difference between it and the first BDE
provides a measure of the interaction between the unpaired
electrons in the biradical. For example, the C-H BDEs in the
phenyl radical (i.e., the second C-H BDEs in benzene) in the
ortho, meta, and para positions to form the singlet benzynes
have recently been measured to be lower than that in benzene
by 36.0, 20.7, and 4.8 kcal/mol, respectively.16,17 The difference
between the first and second bond dissociation enthalpies in
benzene reflect the extent of interaction between the unpaired
electrons in the biradicals through either direct overlap of the
nonbonding atomic orbitals (ortho, meta) or indirect interaction
via through-bond coupling (para).17
The BDEs also reflect the electronic structure in triplet
biradicals. For example, the experimental BDEs for the forma-
tion of the triplet benzynes are all ∼113 kcal/mol,17 the same
as that in benzene, indicating little interaction between the two
unpaired electrons. Similarly, the recently reported enthalpy of
formation of triplet trimethylenemethane agrees with the value
predicted using simple bond additivity.18,19 However, Hill and
Squires have recently found that the second C-H BDE in 1,3-
bis-methylenecyclobutane, which leads to the formation of the
triplet biradical (eq 2), is 16 kcal/mol higher than the first.20
Zhang and Borden21 have shown that the second BDE is higher
because, unlike the radical, the triplet is not stabilized by
delocalization throughout the cyclobutane system. Delocalization
does not occur in the biradical because it would involve a
nonfavorable, cyclobutadiene-like configuration. Therefore, the
enthalpies of formation for triplet biradicals are not always equal
to bond additivity.
Experimental Section
All the gas-phase experiments described in this paper were carried
out in a flowing afterglow, triple quadrupole instrument described
elsewhere.26,27 For the present studies, helium buffer gas was maintained
in the 1 m × 7.3 cm flow reactor at a total pressure of 0.4 Torr, with
a flow rate of 200 STP cm3/s and bulk flow velocity of 9700 cm/s.
The primary reactant ions O•-, and OH-, are produced by electron
ionization (EI) of N2O and a N2O/CH4 mixture, respectively, in the
upstream end of the flow tube. We find that the yield of O•- is
significantly enhanced if a small electrostatic potential (-5 - -10 V)
is placed on the cross region, which is electrically isolated from the
rest of the flowing afterglow. Once formed, the ions are transported
down the tube by the flowing helium, where they are allowed to react
with neutral reagent vapors introduced through leak valves. Reaction
rates can be determined by monitoring signal depletion as a function
of neutral reagent flow, where the neutral reagent vapors are added
through a ring inlet at a fixed distance from the flow reactor. The ions
in the flow tube, thermalized to ambient temperature by ∼105 collisions
with the helium buffer gas, are extracted from the flow tube through a
1-mm orifice and then focused into an EXTREL triple quadrupole
analyzer.
For the photoelectron experiments described above, 1- was
+
generated by H2 abstraction from m-xylene by the atomic
oxygen ion, O•- (eq 3), a reaction first reported by Jennings
and co-workers.16
The CID studies are carried out by selecting the ion with the desired
mass-to-charge ratio using the first quadrupole (Q1), and then injecting
them into the second quadrupole (Q2, radio frequency only), where
they undergo collision with an argon target. For energy-resolved CID
studies, the cross sections for product formation are measured while
the Q2 rod offset is scanned. The reactant and product ions are analyzed
with the third quadrupole (Q3) and are detected with an electron
multiplier operating in pulse-counting mode. Absolute cross sections
are calculated using σp ) Ip/INl, where Ip and I are the intensities of
To verify the structure of the ion, Bruins et al.22 carried out the
(16) Wenthold, P. G.; Squires, R. R. J. Am. Chem. Soc. 1994, 116, 6401.
(17) Wenthold, P. G.; Squires, R. R.; Lineberger, W. C. J. Am. Chem.
Soc. 1998, 120, 5279.
(18) Wenthold, P. G.; Hu, J.; Squires, R. R.; Lineberger, W. C. J. Am.
Chem. Soc. 1996, 118, 475.
(19) Wenthold, P. G.; Hu, J.; Squires, R. R.; Lineberger, W. C. J. Am.
Soc. Mass Spectrom. 1999, 10, 800.
(23) Wenthold, P. G.; Wierschke, S. G.; Nash, J. J.; Squires, R. R. J.
Am. Chem. Soc. 1994, 116, 7378.
(24) Clauberg, H.; Minsek, D. W.; Chen, P. J. Am. Chem. Soc. 1992,
114, 99.
(20) Hill, B. T.; Squires, R. R. J. Chem. Soc., Perkin Trans. 2 1998,
1027.
(21) Zhang, D. Y.; Borden, W. T. J. Mol. Struct., in press.
(22) Bruins, A. P.; Ferrer-Correia, A. J.; Harrison, A. G.; Jennings, K.
R.; Mitchum, R. K. AdV. Mass Spectrom. 1978, 7, 355.
(25) Zhang, X.; Chen, P. J. Am. Chem. Soc. 1992, 114, 3147.
(26) Marinelli, P. J.; Paulino, J. A.; Sunderlin, L. S.; Wenthold, P. G.;
Poutsma, J. C.; Squires, R. R. Int. J. Mass Spectrom. Ion Processes 1994,
130, 89.
(27) Graul, S. T.; Squires, R. R. Mass Spectrom. ReV. 1988, 7, 1.

The Enthalpy of Formation of m-Xylylene
J. Am. Chem. Soc., Vol. 122, No. 45, 2000 11205
involves measuring the proton affinity of the negative ion, 1-,
we first discuss its formation and characterization. We next
describe proton-transfer experiments used to determine the
proton affinity of 1- and the subsequent calculation of ∆Hf,298(1).
In the last part of this section, we describe an alternate
measurement of the enthalpy of formation using CID threshold
measurements. Finally, the measured value for the enthalpy of
formation is discussed in light of the electron - electron
interaction within the biradical.
the product and reactant ions, respectively, N is the number density of
the target, and l is the effective length of the collision cell, calibrated
to be 24 ( 4 cm. The CID cross sections are measured at different
pressures and extrapolated to p ) 0, such that they correspond to single
collision conditions.
The center of mass collision energies are calculated using ECM
lab[m/(M + m)], where Elab is the collision energy in the laboratory
)
E
frame of reference, and m and M are the masses of the target and the
ion, respectively. Determination of the ion kinetic energy origin and
beam energy spread is accomplished by retarding potential analysis,
with Q2 serving as the retarding field element. Ion beam energy
distributions are found to be Gaussian in shape, with a typical full width
at half-height (fwhh) of 0.5-1.5 eV (laboratory frame).
Energy-resolved cross sections are fit using the assumed model
shown in eq 5,28-30 where E is the energy of the ion, Ei is the vibrational
energy, ET is the dissociation energy, n is an adjustable parameter, and
σo is a scaling factor.
Ion Formation. The reaction of O•- with m-xylene (eq 3) in
the flowing afterglow gives m/z 104 (H2+ transfer) as the major
product.22 The 3-methylbenzyl anion is also observed, with the
intensity of the m/z 105 peak ∼40% of that for m/z 104. The
3-methylbenzyl anion can be formed directly by deprotonation
of m-xylene with O•-, or it can be the result of proton transfer
between m-xylylene and hydroxide, which is formed by hydro-
gen atom abstraction.
To characterize the products of eq 3, we have first repeated
the experiments of Bruins et al.16 using m-xylene-d3. Of the
possible products, we only observe those at m/z 106, 107, and
108. Whereas m/z 107 and 108 can result from D+ and H+
transfer, respectively, the peak at m/z 106 indicates the transfer
of HD+. More importantly, the absence of a peak at m/z 105
rules out formation of 3-methylphenylcarbene radical anion, 4-,
the result of D2+ abstraction. Therefore, our results agree with
those of Jennings and co-workers22 in that the reaction of
m-xylene-d3 does not produce the phenylcarbene radical anion.
giPD(E,Ei,τ)(E + Ei - ET)n
σ(E) ) σ0
(5)
∑
[
]
E
i
Modeling is carried out by minimizing the deviation between the model
function and the steeply rising portion of the appearance curve just
above threshold. The modeling also accounts for the ion energy
distribution and Doppler broadening due to thermal motion of the target.
Also incorporated into the analysis are the dissociation rates of the
ions, PD, calculated using Rice-Ramsperger-Kassel-Marcus (RRKM)
theory, to account for potential kinetic shifts that result from slow
dissociation on the instrumental time scale (τ ) ∼30 µs). The
dissociation reactions are assumed to have loose, product-like transition
states that correspond to the “phase-space limit,” and are calculated
using the approach described by Rodgers et al.31 Physical parameters
for the reactants and products, including vibrational frequencies,
rotational constants, and polarizabilities, are calculated at the Becke3LYP/
6-31+G* level of theory.32 Vibrational frequencies are scaled by 0.965
to account for anharmonicities. Dissociation energies obtained from
the fitting procedures correspond to the 0 K energies. These values are
converted to the 298 K bond dissociation enthalpies, DH298, using the
integrated heat capacities. All analyses are carried out using the
CRUNCH program developed by Armentrout and co-workers.28-31
Materials. The labeled m-xylenes, d3 and d6, were prepared from
the reaction of CD3I with the Grignard reagents prepared from
m-bromotoluene and m-dibromobenzene, respectively, in tetrahydro-
furan at room temperature, and were purified by column chromatog-
raphy. The d2-chloro-m-xylene was prepared by the reduction of the
corresponding methyl carboxylate with lithium aluminum deuteride,33
followed by chlorination with SOCl2.34 All other reagents were obtained
from commercial sources and were used as supplied. Gas purities were
as follows: He (99.995%), NO2 (99.5%), and CH4 (99%).
Although the reaction of O•- with m-xylene-d3 rules out R,R-
abstraction to form the carbene ion, the results do not require
the product to be the m-xylylene ion. For example, the HD+
transfer product observed in the reaction may result from [R,
ring] abstraction to form a 3-methyl-R,n-dehydrotoluene ion,
as shown for the R,2 isomer in eq 6. However, if [R, ring]
+
abstraction is occurring, then we also expect to observe the H2
transfer product (m/z 107). Unfortunately, this ion has the same
mass as that resulting from D+ transfer, and we cannot rule out
the formation of the ring-abstracted ion using the d3-labeled
material.
Results and Discussion
Here we describe the determination of the enthalpy of
formation of m-xylylene, 1. Given that our first approach
(28) Ervin, K. M.; Armentrout, P. B. J. Chem. Phys. 1985, 83, 166.
(29) Schultz, R. H.; Crellin, K. C.; Armentrout, P. B. J. Am. Chem. Soc.
1991, 113, 8590.
(30) Dalleska, N. F.; Honma, K.; Sunderlin, L. S.; Armentrout, P. B. J.
Am. Chem. Soc. 1994, 116, 3519.
(31) Rodgers, M. T.; Ervin, K. M.; Armentrout, P. B. J. Chem. Phys.
1997, 106, 4499.
(32) Gaussian 98, Revision C.3; Frisch, M. J.; Trucks, G. W.; Schlegal,
H. B.; Gill, P. M. W.; Johnson, B. G.; Robb, M. A.; Cheeseman, J. R.;
Kieth, T.; Petersson, G. A.; Montgomery, J. A.; Raghavachari, K.;
Al-Laham, M. A.; Zakrewski, V. G.; Ortiz, J. V.; Foresman, J. B.;
Cioslowski, J.; Stefanov, B. B.; Nanayakkara, A.; Challacombe, M.; Peng,
C. Y.; Ayala, P. Y.; Chen, W.; Wong, M. W.; Andres, J. L.; Replogle, E.
S.; Gomperts, R.; Martin, R. L.; Fox, D. J.; Binkley, J. S.; Defrees, D. J.;
J.Baker; Stewart, J. J. P.; Head-Gordon, M.; Gonzalez, C.; Pople, J. A.,
Eds. Gaussian, Inc., Pittsburgh, PA, 1998.
Additional insight is gained from the reaction with m-xylene-
d6. Assuming that R,R-abstraction does not occur, potential
products of the reaction are shown in Scheme 1. Formation of
1- is indicated by the formation of m/z 108, whereas 3-methyl-
R,n-dehydrotoluene ion (shown as the R,2 isomer in Scheme
1) has m/z 109.
The mass spectrum of the region m/z 100-115 is shown in
Figure 1. The main product of the reaction is the m-xylylene
ion, with m/z 108. However, we also observe significant yields
at m/z 109 and 110, resulting from HD+ and D+ transfer,
respectively. From the relative yields of the m/z 108 and 109
(33) Nystrom, R. F.; Brown, W. G. J. Am. Chem. Soc. 1962, 69, 1197.
(34) Lee, C. C.; Clayton, J. W.; Finlayson, A. J.; Lee, D. G. Tetrahedron
1962, 18, 1395.
+
ions, we estimate that the 75% of the H2 transfer product in

11206 J. Am. Chem. Soc., Vol. 122, No. 45, 2000
Hammad and Wenthold
Scheme 1
Scheme 2
to the flowing afterglow, but the only product observed at low
NO flows is an adduct ion at m/z 134. This result verifies that
only the open-shell M-2H ions give ionic products with nitric
oxide. In addition, at higher flows of NO, a second NO molecule
adds to the m/z 104 ion, but the adduct of m/z 105 is not
observed.
This result is further illustrated by the reactions of the ions
derived from the labeled m-xylenes as shown in Scheme 2. For
example, with m-xylene-d6, only the ions that result from HD+
+
or D2 transfer form adducts with nitric oxide, in a ratio that
reflects the relative abundances of the reactant ions, whereas
the NO adduct of the [M-D]- ion is not observed at all. This
result indicates that the 1- and the 3-methyl-R,n-dehydrotoluene
ions form adducts with NO at comparable rates. For the
m-xylene-d3 system, both the m/z 106 and 107 ions form adducts
with NO. However, whereas the ratio of the m/z 107 to 106
ions is 0.78, the ratio of the corresponding adduct ions is 0.19.
This result reflects the fact that the m/z 107 signal consists of
a mixture of 3-methyl-R,n-dehydrotoluene ion resulting from
Figure 1. Mass spectrum of the products of the reaction of m-xylylene-
d6 with atomic oxygen ion from m/z 100 - 115.
the reaction of O•- with m-xylene is 1-, with the rest
corresponding to unspecified 3-methyl-R,n-dehydrotoluene ions
resulting from ring abstraction. The small (<2%) signal of m/z
107 observed in the spectrum is a result of less than complete
deuterium labeling in the m-xylene-d6 and/or H/D exchange with
adventitious water in the flowing afterglow (vide infra).
Reactivity and Characterization. The products of the
reaction of m-xylene with O•- can also be investigated using
ion/molecule reactions. We have examined the reactivity of the
ions with standard reagents to determine their structures. For
example, given that the m-xylylene and 3-methyl-R,n-dehydro-
toluene ions have radical character, they might be expected to
undergo radical reactions. Similarly, 1- is expected to undergo
reactions similar to those of the 3-methylbenzyl anion, whereas
the R,n-dehydrotoluene ions are expected to exhibit phenyl
anion-like reactivity.35 Therefore, reactivity studies, especially
when carried out on labeled systems, can be used to characterize
the products of eq 3.
The first reaction examined was that with nitric oxide. Nitric
oxide has been shown to react with closed-shell ions by addition
followed by electron detachment,36 such that no ionic products
are observed, whereas adducts36 and other products37,38 have
been observed with open-shell ions. Therefore, we might expect
adduct formation with the m-xylylene and R,n-dehydrotoluene
ions but not with the 3-methylbenzyl anion. In agreement with
this expectation, both the m/z 104 and 105 ions formed from
unlabeled m-xylene react away completely when NO is added
+
H2 transfer and the 3-methylbenzyl anion resulting from
transfer of D+. The ratio of the adduct ions indicates that 70%
of the open-shell ions is the m-xylylene ion, in agreement with
the amount estimated from the reaction of O•- with m-xylene-
d6.
Like most organic anions,39 the ions derived from m-xylene
undergo addition with CO2 to form carboxylate ions, and the
reactivity of the carboxylates reflects the ion structure. The
carboxylates of the open-shell ions have radical sites and
therefore undergo addition with NO, but the carboxylated
3-methylbenzyl anion does not react with nitric oxide. In this
case, the reactions of the ions obtained from m-xylene-d3
indicate that 70% of the open-shell ions have the m-xylylene
structure.
Reaction of the ions with CS2 also shows that a mixture of
open-shell ions is formed. By utilizing m-xylene-d6, we are able
to distinguish the reactivity of the m-xylylene ion (m/z 108)
from that of the 3-methyl-R,n-dehydrotoluene ion (m/z 109).
The m/z 108 ion is found to react exclusively by CS2 addition,
consistent with what is expected for a benzylic ion. On the other
hand, the m/z 109 ion reacts by both CS2 addition and by S
atom abstraction. We interpret this to mean that the dehydro-
toluene ion is much more basic than the m-xylylene,40-43
suggesting that the dehydrotoluene is a phenyl-like anion with
(39) Bierbaum, V. M.; DePuy, C. H.; Shapiro, R. H. J. Am. Chem. Soc.
1977, 99, 5800.
(40) DePuy, C. H.; Bierbaum, V. M. Tetrahedron Lett. 1981, 22, 5129.
(41) DePuy, C. H.; Bierbaum, V. M.; Damrauer, R.; Soderquist, J. A. J.
Am. Chem. Soc. 1985, 107, 3385.
(42) DePuy, C. H.; Bierbaum, V. M.; Robinson, M. S.; Davico, G. E.;
Gareyev, R. Tetrahedron 1997, 53, 9847.
(43) Hare, M.; Emrick, T.; Eaton, P. E.; Kass, S. R. J. Am. Chem. Soc.
1997, 119, 237.
(35) Hu, J. Ph.D. Thesis, Purdue University, West Lafayette, IN, 1996.
(36) Wenthold, P. G.; Hu, J.; Squires, R. R. J. Mass Spectrom. 1998,
33, 796.
(37) Wenthold, P. G.; Hu, J.; Squires, R. R. J. Am. Chem. Soc. 1994,
116, 6961.
(38) Wenthold, P. G.; Squires, R. R. J. Am. Chem. Soc. 1994, 116, 11890.

The Enthalpy of Formation of m-Xylylene
J. Am. Chem. Soc., Vol. 122, No. 45, 2000 11207
Scheme 3
formation of the conjugate base anion, suggesting that proton
transfer has occurred. However, all the reagents that appear to
react with the m-xylylene anion are also found in control
experiments to undergo proton transfer with the 3-methylbenzyl
anion. Therefore, given that m/z 105 is always formed along
with the m-xylylene ion, we are not able to unequivocally assign
the formation of the conjugate base anion to deprotonation by
1-. From these results, we conclude that the acid/base properties
of 3-methylbenzyl radical are similar to those of m-xylene (vide
infra). This conclusion is supported by the measured rates of
the reactions. The kinetics plots of signal versus reagent flow26
are reasonably linear, as would be expected if each ion were
undergoing simple proton transfer. For the acids with which
reaction is observed, we find that the 1- ion reacts at rates
similar to those for the reaction of the 3-methylbenzyl anion,
m/z 105, suggesting comparable thermodynamic driving forces
for the two reactions.
a benzylic radical. This electronic structure is consistent with
what is expected given that the EA of a phenyl radical is ∼0.1
eV higher than that of benzyl.44
The M-2H ions formed in the reaction of O•- with unlabeled
m-xylene react with CS2 by addition and by S atom abstraction.
The sulfur atom abstraction pathway indicates that the dehy-
drotoluene ion is formed in the reaction and the m/z 104 signal
is likely a mixture of isomers. However, we have found that
the dehydrotoluenes rearrange to 1- in the presence of weak
acids. For example, although there is no apparent reaction
between the m/z 104 ion and water, the reactivity of the ion
changes in that it no longer gives a sulfur atom abstraction
product with CS2. We attribute this to water-catalyzed isomer-
ization of the dehydrotoluene ions to the m-xylylene ion.16 The
mechanism for the isomerization is likely similar to that
proposed for H/D exchange of carbanions,45-47 as shown in
Scheme 3. Most importantly, the results with CS2 indicate that
the dehydrotoluene can be removed from the reaction mixture
by the addition of water. Therefore, it is possible to examine
the reactivity of the pure m-xylylene ion without having to use
deuterium-labeled reagents.
Acid/Base Reactions. Given that the EA of 1 has been
measured,15 the enthalpy of formation can be determined if the
proton affinity of the ion (i.e., the gas-phase acidity of
3-methylbenzyl radical) is known. In light of this possibility,
we have carried out reactions of the ion with acids to bracket
the gas-phase acidity. For these experiments, we used the M-2H
ion (m/z 104) derived from unlabeled m-xylene. The ions are
first allowed to react with water introduced directly downstream
from the electron gun to isomerize the dehydrotoluene compo-
nent to 1-, as already described. Acids are added ∼50 cm farther
downstream to allow the reaction with water to occur to
completion.
One reagent that gives a product attributable to reaction with
m-xylylene ion is 1,3,5-trimethylbenzene (mesitylene), which
reacts by H2+ transfer to give, presumably, a methyl-substituted
m-xylylene ion (eq 7). Although H2+ transfer has been proposed
previously to occur in the reaction between o-benzyne ion and
+
methanol,54 this is the first instance in which H2 transfer has
been observed between an organic ion and organic reagent to
form an open-shell ion.55
Enthalpy of Formation of m-Xylylene. The goal of the
bracketing experiments just described was to determine the gas-
phase acidity of 3-methylbenzyl radical so that the enthalpy of
formation could be determined using eq 1. Although we are
not able to assign the acidity of 3-methylbenzyl radical directly,
we estimate from the reactivity studies that it is indistinguishable
from that of m-xylene. The acidity of m-xylene was determined
in this work from the proton-transfer equilibrium data for the
reaction shown in eq 8a. The average pseudo-first-order rate
constant, kf, for the forward reaction obtained from replicate
measurements is (1.7 ( 0.6) × 10-10 mol/s. Similarly, the rate
constant for the reverse reaction, kr, is (1.6 ( 0.5) × 10-10 mol/
s, which means ∆G(eq 8a) ) -0.06 ( 0.65 kcal/mol, where
the uncertainty is conservatively assigned to allow for up to a
factor of 3 error in the measured equilibrium constant. Our value
of ∆G(eq 8a) is in good agreement with that obtained by
Caldwell and Bartmess for this reaction.56,57
The m/z 104 ion prepared in this fashion does not appear to
react with fluorobenzene (∆Gacid ) 379.1 ( 2.0 kcal/mol).48,49
Reaction is observed with methanol, toluene, cumene, and
ethanol (∆Gacid ) 375.2 ( 0.2, 374.9 ( 0.2, 372.9 ( 2.0, and
371.7 ( 1.1 kcal/mol, respectively).50-53 Moreover, for the
reagents with which reaction occurs, we also observe the
(44) Gunion, R. F.; Gilles, M. K.; Polak, M. L.; Lineberger, W. C. Int.
J. Mass Spectrom. Ion Processes 1992, 117, 601.
(45) Stewart, J. H.; Shapiro, R. H.; DePuy, C. H.; Bierbaum, V. M. J.
Am. Chem. Soc. 1977, 99, 7650.
(46) DePuy, C. H.; Bierbaum, V. M.; King, G. K.; Shapiro, R. H. J.
Am. Chem. Soc. 1978, 100, 2921.
Neumark and co-workers50 have derived ∆Gacid(CH3OH) )
(47) Squires, R. R.; Bierbaum, V. M.; Grabowski, J. J.; DePuy, C. H. J.
Am. Chem. Soc. 1983, 105, 5185.
(48) Wenthold, P. G.; Squires, R. R. J. Mass Spectrom. 1995, 30, 17.
(49) Andrade, P. B. M.; Riveros, J. M. J. Mass Spectrom. 1996, 31, 767.
(50) Osborn, D. L.; Leahy, D. J.; Kim, E. H.; deBeer, E.; Neumark, D.
M. Chem. Phys. Lett. 1998, 292, 651.
(53) Dang, T. T.; Motell, E. L.; Travers, M. J.; Clifford, E. P.; Ellison,
G. B.; Depuy, C. H.; Bierbaum, V. M. Int. J. Mass Spectrom. Ion Proc.
1993, 123, 171.
(54) Guo, Y.; Grabowski, J. J. J. Am. Chem. Soc. 1991, 113, 5923.
(55) Lee, J.; Grabowski, J. J. Chem. ReV. 1992, 92, 1611.
(56) Bartmess, J. E., personal communication.
(51) Ellison, G. B.; Davico, G. E.; Bierbaum, V. M.; DePuy, C. H. Int.
J. Mass Spectrom. Ion Processes 1996, 156, 109.
(52) Bartmess, J. E.; Scott, J. A.; McIver, R. T. J. J. Am. Chem. Soc.
1979, 101, 6047.
(57) Bartmess, J. E. Negative Ion Energetics Data. In NIST Chemistry
WebBook, NIST Standard Reference Database Number 69, February 2000
ed.; Mallard, W. G., Linstrom, P. J., Eds.; National Institute of Standards
and Technology: Gaithersburg, MD, 20899, 2000 (http://webbook.nist.gov).

11208 J. Am. Chem. Soc., Vol. 122, No. 45, 2000
Hammad and Wenthold
375.2 ( 1.1 kcal/mol using the homolytic O-H BDE58,59 and
the EA of the methoxy radical. Combining this value with
∆G(eq 8a) gives ∆Gacid(m-xylene) ) 375.1 ( 1.3 kcal/mol. The
quantity ∆Hacid(m-xylene) can be determined using ∆H ) ∆G
+ T∆S, where the entropy term was calculated using eq 8b.
Ellison and co-workers51 have reported the ∆Sacid(toluene) )
24.7 ( 1.6 cal/mol K. The entropy change for the reaction in
eq 8b is calculated to be 1.29 cal/mol K, using frequencies
calculated at the B3LYP/6-31+G* level of theory,60 which gives
∆Sacid(m-xylene) ) 26.0 ( 2.0 cal/mol K, and ∆Hacid(m-xylene)
) 382.8 ( 1.6 kcal/mol.
On the basis of the reactivity of the m-xylylene negative ion,
∆Gacid(3-methylbenzyl) is 375.1 ( 2.0 kcal/mol. Again, the
quantity ∆Hacid(3-methylbenzyl) is calculated using ∆H ) ∆G
+ T∆S. The entropy of deprotonation for 3-methylbenzyl radical
was determined by calculating the entropy change for the proton-
transfer reaction shown in eq 9, which has ∆S ) -0.03 eu at
the B3LYP/6-31+G* level of theory.
Figure 2. Cross sections for collision-induced dissociation of Cl- from
3-(chloromethyl)benzyl and R-chloro-3-methylbenzyl anions as a
function of translational energy in the center-of-mass frame. The solid
lines are the model appearance curves calculated using eq 5.
10b.
Therefore, ∆Sacid(3-methylbenzyl) is predicted to be the same
as ∆Sacid(toluene) ) 24.7 ( 1.6 cal/mol K, which gives ∆Hacid
-
(3-methylbenzyl) ) 382.5 ( 2.1 kcal/mol. Combining this
acidity with the EA of m-xylylene (0.919 ( 0.008 eV)15 and
the ionization potential (IP) of hydrogen atom according to eq
1 gives BDE(3-CH2C6H4CH2-H) ) 90.1 ( 1.7 kcal/mol.61
Similarly, the C-H BDE in m-xylene is calculated from the
acidity and EA (0.905 ( 0.008 eV)15 to be 90.1 ( 1.7 kcal/
mol,61 such that the second C-H bond in m-xylene is as strong
as the first. This result is a consequence of the fact that the
acidity and EA of 3-methylbenzyl radical are essentially the
same as those for m-xylene and m-xylylene, respectively. From
the BDEs and the enthalpy of formation of m-xylene (4.12 kcal/
mol),62 the enthalpy of formation of m-xylylene is calculated
to be 80.1 ( 3.8 kcal/mol, where the uncertainty reflects the
fact that both BDE values depend on the acidity of m-xylene.
Collision-Induced Dissociation Threshold Measurements.
Wenthold and Squires16,23 have shown that biradical enthalpies
of formation can also be obtained using CID threshold energy
measurements with halogen-substituted ions. Therefore, it is
possible to obtain a second value for the enthalpy of formation
of m-xylylene by measuring the dissociation energy for 3-(chlo-
romethyl)benzyl anion, as shown in eq 10a. The enthalpy of
formation of the biradical is calculated from the bond dissocia-
tion enthalpy, DH298, using eq 11.
∆H°f,298(C8H8) ) DH298(C8H8Cl) + ∆Hacid(C8H9Cl) +
∆H°f,298(C8H9Cl) - ∆Hacid(HCl) - ∆H°f,298(HCl) (11)
The CID of the mass-selected 6- (m/z 141) gives chloride as
the only ionic fragment. The cross sections as a function of
center-of-mass collision energy are shown in Figure 2. Also
shown in Figure 2 are the cross sections for dissociation of the
R-chloro-3-methylbenzyl anion, 7- (m/z 140). Average values
of DH298 and n obtained by modeling the cross sections for the
dissociations of ions 6- and 7- are listed at the bottom of Table
1. The chloride binding enthalpies (298 K) of the 3-(chloro-
methyl)benzyl and R-chloro-3-methylbenzyl anions, obtained
from replicate measurements, are 0.87 ( 0.07 and 1.70 ( 0.08
eV, respectively, where the uncertainty includes the deviation
of the data, a 0.15 eV (lab) uncertainty in the absolute energy
scale, and a contribution due to uncertainty in the choice of the
transition state. Because of the low dissociation energy, the
kinetic shift for the formation of m-xylylene from ion 6- is
<0.01 eV. The threshold for the dissociation of 7- is similar to
that reported previously by Poutsma and Squires (1.67 ( 0.10
eV)63 for the dissociation of R-chlorobenzyl anion (eq 12).
The 3-(chloromethyl)benzyl anion can be generated by
deprotonation of R-chloro-m-xylene using hydroxide. Although
OH- is capable of deprotonating both methyl sites in R-chloro-
m-xylene, the use of deuterium-labeled reagent allows the
3-(chloromethyl)benzyl anion that results from proton transfer,
6-, to be distinguished from R-chloro-3-methylbenzyl anion,
7-, resulting from deuterium cation transfer, as shown in eq
Calculation of the enthalpy of formation of m-xylylene using
eq 11 requires the gas-phase acidity of R-chloro-m-xylene in
the nonchlorinated position. Standard approaches for determin-
ing gas-phase acidities, such as equilibrium measurements,64
are not applicable in this case because the methyl position is
not the most acidic site in the molecule. However, for the
(58) Engelking, P. C.; Ellison, G. B.; Lineberger, W. C. J. Chem. Phys.
1978, 69, 1826.
(59) Meot-Ner, M.; Sieck, L. W. J. Phys. Chem. 1986, 90, 6687.
(60) Entropies for vibrations are calculated from unscaled frequencies.
Unsubstituted methyl groups are assumed to be hindered rotors with entropy
contributions similar to that for the methyl group in toluene (ref 51).
(61) The small temperature correction term is ignored.
(62) Afeefy, H. Y.; Liebman, J. F.; Stein, S. E. Neutral Thermochemical
Data. In NIST Chemistry WebBook, NIST Standard Reference Database
Number 69, February 2000 ed.; Mallard, W. G., Linstrom, P. J., Eds.;
National Institute of Standards and Technology: Gaithersburg, MD, 20899,
2000 (http://webbook.nist.gov).
(63) Poutsma, J. C.; Nash, J. J.; Paulino, J. A.; Squires, R. R. J. Am.
Chem. Soc. 1997, 119, 4686.
(64) Berkowitz, J.; Ellison, G. B.; Gutman, D. J. Phys. Chem. 1994, 98,
2744.

The Enthalpy of Formation of m-Xylylene
J. Am. Chem. Soc., Vol. 122, No. 45, 2000 11209
Table 1. Supplemental and Derived Thermochemical Dataa
parameter
value
reference
298 K enthalpies of formation
m-xylene
toluene
HCl
1-(chloromethyl)-3-
methylbenzene
m-xylylene, 1
from eq 1
4.12 ( 0.18 62
11.95 ( 0.15 62
-22.06 ( 0.02 62
-3.3 ( 1
bond additivity
80.1 ( 3.8 this work
81.2 ( 3.0 this work
80.8 ( 2.4 this work
from eq 11
average
gas-phase acidities
∆Gacid ∆Hacid
374.9 ( 0.2 383.2 ( 0.5 51
375.1 ( 1.3 382.8 ( 1.6 this work
375.1 ( 2.0 382.5 ( 2.1 this work
333.4 ( 0.1 57
toluene
m-xylene
3-methylbenzyl radical
HCl
electron affinities
21.03 ( 0.14 44, 65
20.87 ( 0.14 15
Figure 3. Plot of EA versus ln r for reference radicals, along with a
linear fit to the data. The center of mass collision energy corresponds
to 4.0 eV: (a) benzyl radical; (b) p-fluorobenzyl radical; (c) o-
fluorobenzyl radical; (d) m-fluorobenzyl radical; (e) p-chlorobenzyl
radical; and (f) p-bromobenzyl radical. The result for the 3-(chloro-
methyl)benzyl radical, 6, is also indicated on the plot.
benzyl
3-methylbenzyl
m-xylylene, 1
21.19 ( 0.18 15
3-(chloromethyl)benzyl, 6
27.5 ( 1.2 this work
chloride dissociation energies
DH298(3-CH2C6H4CD2-Cl-)
20.1 ( 1.6 this work
nb
1.6 ( 0.1
4-fluoro-, 4-chloro-, and 4-bromobenzyl radicals65 as calibrants
is shown in Figure 3. The regression line in Figure 3 has the
form shown in eq 14, where r ) I(SO2-)/I(R-).
DH298(3-CH3C6H4CD-Cl-)
39.2 ( 1.8 this work
1.7 ( 0.1
nb
a Values in kcal/mol unless otherwise noted. b Exponent in eq 5.
EA(R) ) 24.75-1.248 * ln r
(14)
unsubstituted methyl group, an accurate value for the acidity
can likely be obtained indirectly using the familiar relationship
shown in eq 13,64 where DH298(R-H) is the homolytic C-H
bond dissociation enthalpy, EA(R) is the electron affinity of
the 3-(chloromethyl)benzyl radical, 6, and IP(H) is the ionization
potential of hydrogen.
From the CID branching ratio for the SO2 adduct of 6-, r )
0.107, and the EA of radical 6 is 1.194 ( 0.053 eV. As usual,
certain assumptions are involved in this measurement. First, it
is assumed that the effective temperature70 for the dissociation
of the SO2 adduct of 6- is the same as those of the calibrant
ions. As Ervin has shown,71 this assumption may introduce an
error of 0.5-1.0 kcal/mol into the measurement. In addition,
the relationship in eq 15 assumes that the entropy for the
dissociation of the SO2 adduct of 6- is the same as the entropies
for the dissociation of the calibrant cluster ions. To test this
assumption, we carried out a theoretical study of the measure-
ment of the EA of 6 using the kinetic method. Using energies
and geometries calculated at the B3LYP/6-31+G* level of
theory and RRKM theory, we find that the difference in ∆S,
the entropy of activation, among the different cluster ions to be
0.77 cal/mol K. Although full details of this study will be
provided elsewhere,72 the results indicate that neglecting the
entropy contribution may lead to an error of up to 0.5 kcal/mol
(using an effective temperature of 628 K at 4 eV collision
energy). Therefore, the uncertainty assigned to EA(6) includes
a 0.5 kcal/mol component, to account for potential errors that
can arise by neglecting the entropy differences, in addition to a
1 kcal/mol contribution due to the uncertainty in the effective
temperature, a 0.25 kcal/mol contribution to account for
uncertainty in the ratio measurements (which corresponds to a
(20% uncertainty in the relative ratios), and a 0.38 kcal/mol
contribution due to the statistical error in the regression plot
(the uncertainties in the EAs of the references are negligible).
The branching ratio for the dissociation of the SO2 adduct of
ion 7- was also measured and found to be r ) 0.336, implying
an EA of 1.132 eV. However, as an R-substituted benzyl anion,
7- is structurally different from the ring-substituted anions used
as calibrants. Therefore, the entropy for the dissociation of the
SO2 adduct of 7- is not expected to be the same as that for
DH298(RsH) ) EA(R) + ∆Hacid(RsH) s IP(H) (13)
The key to this approach is that Kim et al.65 have shown that
the homolytic bond dissociation enthalpies of halogen-substi-
tuted toluenes are all ∼89 kcal/mol. Therefore, the gas-phase
acidities of ring-substituted toluenes can be reliably estimated
by assuming a bond dissociation enthalpy of 89 ( 2 kcal/mol
and measuring the EA.
The EA of the 3-(chloromethyl)benzyl radical, 6, can be
determined using the kinetic method developed by Cooks and
co-workers.66,67 The procedures for obtaining EAs using this
approach have been described elsewhere.20,68,69 The branching
ratios for the CID of SO2 adducts of a series of substituted
benzyl anions with known EAs are measured. A semilog
relationship between the branching ratios and the EAs is
assumed, and a calibration plot is constructed. In the simplest
version of the kinetic method, the branching ratio for the SO2
adduct of the unknown is measured and the EA is determined
from the calibration. This approach has been utilized by Squires
and co-workers for the determination of m- and p-benzyne,68
1,3-bis-methylenecyclobutane-2,4-diyl,20 and cyclopropyl radi-
cal.69 The calibration curve obtained at a center-of-mass collision
energy of 4 eV using benzyl and the 2-fluoro-, 3-fluoro-,
(65) Kim, J. B.; Wenthold, P. G.; Lineberger, W. C. J. Phys. Chem. A
1999, 103, 19833.
(66) Cooks, R. G.; Patrick, J. S.; Kotiaho, T.; McCluckey, S. A. Mass
Spectrom. ReV. 1994, 13, 287.
(67) Cooks, R. G.; Wong, P. S. H. Acc. Chem. Res. 1998, 31, 379.
(68) Wenthold, P. G.; Hu, J.; Squires, R. R. J. Am. Chem. Soc. 1996,
118, 11865.
(70) Drahos, L.; Ve´key, K. J. Mass Spectrom. 1999, 34, 79.
(71) Ervin, K. M. Int. J. Mass Spectrom. 2000, 195/196, 271.
(72) Liu, X.; Hammad, L.; Wenthold, P. G., manuscript in preparation.
(69) Seburg, R. A.; Squires, R. R. Int. J. Mass Spectrom. Ion Processes
1997, 167/168, 541.

11210 J. Am. Chem. Soc., Vol. 122, No. 45, 2000
Hammad and Wenthold
dissociation of the calibration ions, leading to even larger errors
than those already discussed. An accurate determination of the
EA of 7- using the kinetic method would require full entropy
analysis.73,74 These experiments are currently underway and will
be reported elsewhere.72
the chlorinated position and the bond dissociation enthalpy in
7- to be 98.3 ( 3.0 kcal/mol.
Although the neutral products are not detected, the enthalpies
of formation determined from CID of 6- and 7- are assigned
to m-xylylene and 3-methylphenylcarbene, respectively, on the
basis of the following observations. First, the CID experiments
with 7- are similar to those reported previously by Poutsma
and Squires for deprotonated benzyl chloride.63 In that study,
they ruled out formation of noncarbene products on the basis
of dynamical and energy barrier arguments. For example,
dissociation to form the carbene is kinetically favorable because
it occurs by a simple bond cleavage through a loose transition
state as opposed to a rearrangement through a tight transition
state. Moreover, rearrangements to lower energy products are
accompanied by high-energy barriers that would be manifested
in the experiment. Therefore, the CID of 7- is expected to give
3-methylphenylcarbene as the neutral product.
Similarly, the CID of 6- should give m-xylylene. Again, direct
dissociation is dynamically favored over a rearrangement
process. Moreover, skeletal rearrangement will be accompanied
by large barriers, in excess of that measured in the present
experiment. In fact, the low dissociation threshold measured
for 6- indicates that if rearrangement were to occur, it must
have a barrier lower than ∼0.9 eV. Therefore, the low
dissociation energy strongly suggests that 6- dissociates by
simple dissociation to form m-xylylene.
Using a homolytic C-H bond dissociation of enthalpy for
the unsubstituted position of R-chloro-m-xylene of 89.0 ( 2.0
kcal/mol and an EA of 1.194 ( 0.053 eV for the 3-(chloro-
methyl)benzyl radical, the gas-phase acidity of the unsubstituted
methyl position in R-chloro-m-xylene is estimated to be 375.1
( 2.3 kcal/mol. Therefore, this position is more acidic than
toluene or m-xylene (∆Hacid ) 382.3 ( 0.651 and 382.8 ( 1.6
kcal/mol, respectively) but not quite as acidic as m- or
p-chlorotoluene (∆Hacid ) 374 ( 3 kcal/mol),23 in agreement
with what is expected on the basis of substituent effects.75 The
enthalpy of formation of m-xylylene obtained from the gas-
phase acidity, chloride binding energy, and the auxiliary data62
listed in Table 1 is ∆Hf,298(m-xylylene) ) 81.2 ( 3.0 kcal/mol,
which agrees within error with the value estimated from the
reactivity of the m-xylylene ion.
As part of this work, the acidity of the substituted position
in R-chloro-m-xylene was determined using the bracketing
approach. For example, we find that OH- and CH3O- abstract
both D+ and H+ from 5, indicating that both methyl positions
are more acidic than water or methanol (∆Gacid(H2O) ) 384.10
( 0.05 kcal/mol, ∆Gacid(CH3OH) ) 375.2 ( 1.1 kcal/mol).50,64
The observation of H+ transfer with these bases is consistent
with the estimated gas-phase acidity at the nonchlorinated
position already described. Reaction of 5 with the conjugate
base anions of toluene (∆Gacid ) 374.9 ( 0.2 kcal/mol),51
ethanol (371.7 ( 1.1),53 dichloromethane (369.0 ( 0.7),76 tert-
butyl alcohol (367.7 ( 2.1),77 propionitrile (367.4 ( 2.0),52 and
dimethyl sulfoxide (366.4 ( 2.0)52 results in transfer of D+,
leading to formation of 7-. Although deprotonation at the
nonsubstituted methyl position to form 6- is thermodynamically
allowed for benzyl and ethoxide ions, it is likely much slower
than that at the substituted site and is therefore not observed to
an appreciable extent. Slow D+ transfer is observed in the
reaction of 5 with deprotonated methyl acetate (∆Gacid ) 365.1
( 2.0),52 whereas neither H+ nor D+ transfer is observed in
the reaction between 5 and the conjugate base anions of
acetonitrile (365.2 ( 2.0),52 benzyl alcohol (363.4 ( 2.0),52
acetone (361.9 ( 2.0),52 or weaker bases. From these results,
we conclude that the acidity of 5 is between that of acetonitrile
and dimethyl sulfoxide and assign ∆Gacid ) 365.5 ( 2.0 kcal/
mol. The quantity ∆Hacid ) 373.8 ( 2.1 kcal/mol is calculated
using ∆H ) ∆G + T∆S, where ∆S ) 28.0 cal/mol K is
calculated using the same procedure already described for
m-xylene. Therefore, from the bracketing experiments we find
that the gas-phase acidity value at the chlorinated position is
slightly higher than that reported previously for benzyl chloride,
372.0 ( 2.0 kcal/mol.78 The enthalpy of formation of 3-meth-
ylphenyl carbene is calculated from the bracketed acidity of
Dissociation of the singlet benzyl anion to form the triplet
biradical or carbene is formally spin-forbidden, and is therefore
potentially slow on the experimental time scale. Wenthold et
al.23 have argued previously that dissociation of o- or p-
chlorobenzyl anions to form dehydrotoluene biradicals suffers
from a kinetic shift because of slow intersystem crossing (ISC).
Given the biradical character of m-xylylene, it is important to
recognize the possibility of a kinetic shift in this system as well.
However, as discussed by Squires and co-workers,23,63 slow ISC
is expected only for those systems where the singlet state has
a large degree of open-shell character. Intersystem crossing will
be more favorable for systems such as carbenes where the lowest
energy singlet states have closed-shell character because the
accompanying change in orbital angular momentum compen-
sates for the change in spin angular momentum of the electron.79
Consequently, Poutsma and Squires did not see evidence for
slow ISC in their studies of methylene, vinyl-, and phenylcar-
bene,63 and we would not expect a kinetic shift in the
dissociation of 7-. Moreover, the lowest energy singlet state of
1
m-xylylene is the closed-shell, A1 which is state, ∼10 kcal/
mol lower in energy than the open-shell 1B2 state.15 Thus, ISC
should be favorable for the dissociation of 6-. In previous
studies, the possibility of a kinetic shift has been examined by
utilizing bromo- and iodo- substituted ion precursors.23,63
Unfortunately, we were unable to generate the 3-(bromomethyl)-
benzyl anion by deprotonation of the deuterated bromo-m-
xylene. Therefore, although the dissociation 6- is not expected
to suffer from a kinetic shift due to slow ISC, the possibility
cannot be ruled out. If there is a kinetic shift, then the measured
enthalpy of formation is formally an upper limit.
(73) Armentrout, P. B. J. Am. Soc. Mass Spec. 2000, 11, 371.
(74) Wenthold, P. G. J. Am. Soc. Mass Spec. 2000, 11, 601.
(75) Taft, R. W.; Topsom, R. D. The Nature and Analysis of Substituent
Effects. In Progress in Physical Organic Chemistry; Taft, R. W., Ed.; John
Wiley & Sons: New York, 1987; vol 16, p 1.
(76) Poutsma, J. C.; Paulino, J. A.; Squires, R. R. J. Phys. Chem. A 1997,
101, 5327.
Thermochemical Results. A weighted average of the two
measured enthalpies of formation gives ∆Hf,298(1) ) 80.8 (
2.4 kcal/mol. The final enthalpy of formation gives a second
C-H bond dissociation energy in m-xylene of 90.7 ( 2.9 kcal/
mol, only slightly higher than the first (90.1 ( 1.7 kcal/mol).
The fact that the strength of the second C-H bond in
3-methylbenzyl radical is the same as that in m-xylene indicates
(77) Haas, M. J.; Harrison, A. G. Int. J. Mass Spectrom. Ion Proc. 1993,
124, 115.
(78) Most of the difference between the acidities of benzyl chloride and
5 is in the ∆Sacid term. Poutsma and Squires used ∆Sacid(benzyl chloride)
) 24.3 cal/mol K. Recalculation of this quantity using the approach
described in this work gives ∆Sacid(benzyl chloride) ) 27.4 cal/mol K, and
∆Hacid(benzyl chloride) ) 373.1 kcal/mol.
(79) Salem, L.; Rowland, C. Angew. Chem., Int. Ed. Eng. 1972, 11, 92.

The Enthalpy of Formation of m-Xylylene
J. Am. Chem. Soc., Vol. 122, No. 45, 2000 11211
that the interaction between the two unpaired electrons is
negligible, as expected for a triplet biradical.24,25 Calculations
at the CASSCF, CASPT2, and B3LYP levels of theory with
6-31G* basis sets predict that the second C-H bond in m-xylene
should be weaker than the first by 2.5, 1.6, and 2.0 kcal/mol,
respectively,80 in reasonable agreement with the results obtained
in this work.
the R C-H bond dissociation energy in the 3-methylbenzyl
radical is 108.3 ( 3.4 kcal/mol. This result is similar to the
dissociation energy of the R C-H in benzyl radical reported
by Poutsma and Squires.63 Therefore, as expected, methyl
substitution does not affect the stability of phenyl carbene.
Conclusions
For comparison, Hehre and co-workers81 have determined the
enthalpies of formation of o- and p-xylylene, 8 and 9,
respectively, by measuring their proton affinities. Both isomers
Isotopic labeling and chemical reactivity studies indicate that
the reaction of m-xylene with atomic oxygen ion proceeds by
H2+ abstraction to form 75% m-xylylene radical anion and 25%
unspecified ions corresponding to [R,ring] abstraction. Simple
proton transfer is also observed. The open-shell anions generated
in the reaction are characterized by their reaction with nitric
oxide. Whereas the m-xylylene ion undergoes reactions char-
acteristic of a benzylic ion, the M-2 ion resulting from R,ring
abstraction shows phenyl anion-like reactivity.
The thermochemical properties of the 3-methylbenzyl radical
are very similar to those for m-xylene, indicating that the meta-
methylene substituent does not have a significantly different
effect on the system than does a methyl substituent in the same
position. The enthalpies of formation of meta-quinodimethane
(m-xylylene, 1) obtained by two different experimental ap-
proaches are in excellent agreement with each other, giving
∆Hf,298(1) ) 80.8 ( 2.4 kcal/mol. This value is consistent with
what is predicted using a simple bond additivity approach and
similar to what was found for trimethylenemethane18,19 and the
triplet benzynes.17 The results indicate that m-xylylene can be
thought of as a biradical with noninteracting radical centers.
were found to have an enthalpy of formation of 48 kcal/mol,
significantly lower than that obtained for the meta isomer. The
difference is because 8 and 9 have stable Kekule structures with
singlet ground states, whereas 1 is a triplet, non-Kekule biradical.
From their proton affinity measurements, Pollack et al.81 were
not able to determine the enthalpy of formation of 1, but did
provide a lower limit of 76 kcal/mol, in agreement with the
results obtained here.
The enthalpy of formation of 3-methylphenylcarbene is found
in this work to be ∼18 kcal/mol higher than that of m-xylylene.
The difference between the energies of these two isomers can
be attributed to the benzylic delocalization of both electrons in
m-xylylene, whereas one electron in 3-methylphenylcarbene is
localized in an in-plane sp2 orbital on the R carbon.82-84 The
enthalpy of formation of 3-methylphenylcarbene indicates that
Acknowledgment. This work was supported by the Ameri-
can Society for Mass Spectrometry and the Purdue Research
Foundation. Thanks also go to the donors of the Petroleum
Research Fund, as administered by the American Chemical
Society, for partial support of this work. Special thanks go to
Profs. Hilkka Kentta¨maa, Weston Thatcher Borden, John
Bartmess, Peter Armentrout, and Dr. David A. Hrovat for
providing supporting information and helpful discussions. We
also thank Rhonda Chyall and Great Lakes Chemical for the
donation of some of the equipment used in this work.
(80) Hrovat, D. A.; Borden, W. T., personal communication of unpub-
lished work.
(81) Pollack, S. K.; Raine, B. C.; Hehre, W. J. J. Am. Chem. Soc. 1981,
103, 6308.
(82) Schreiner, P. R.; Karney, W. L.; Schleyer, P. V. R.; Borden, W. T.;
Hamilton, T. P.; Schaefer, H. F. J. Org. Chem. 1996, 61, 7030.
(83) Wong, M. W.; Wentrup, C. J. Org. Chem. 1996, 61, 7022.
(84) Matzinger, S.; Bally, T.; Patterson, E. V.; McMahon, R. J. J. Am.
Chem. Soc. 1996, 118, 1535 and references therein.
JA002138N