Isomerization of neopentyl chloride and neopentyl bromide by a 1,2-Interchange of a halogen atom and a methyl group

Article abstract of DOI:10.1021/jp1047166

The recombination of chloromethyl and t-butyl radicals at room temperature was used to generate neopentyl chloride molecules with 89 kcal mol^-1 of internal energy. The observed unimolecular reactions, which give 2-methyl-2-butene and 2-methyl-1-butene plus HCl, as products, are explained by a mechanism that involves the interchange of a methyl group and the chlorine atom to yield 2-chloro-2-methylbutane, which subsequently eliminates hydrogen chloride by the usual four-centered mechanism to give the observed products. The interchange isomerization process is the rate-limiting step. Similar experiments were done with CD₂Cl and C(CH₃)₃ radicals to measure the kinetic-isotope effect to help corroborate the proposed mechanism. Density functional theory was employed at the B3PW91/6-31G(d',p') level to verify the Cl/CH₃ interchange mechanism and to characterize the interchange transition state. These calculations, which provide vibrational frequencies and moments of inertia of the molecule and transition state, were used to evaluate the statistical unimolecular rate constants. Matching the calculated and experimental rate constants, gave 62 ± 2 kcal mol ^-1 as the threshold energy for interchange of the Cl atom and a methyl group. The calculated models also were used to reinterpret the thermal unimolecular reactions of neopentyl chloride and neopentyl bromide. The previously assumed Wagner-Meerwein rearrangement mechanism for these reactions can be replaced by a mechanism that involves the interchange of the halogen atom and a methyl group followed by HCl or HBr elimination from 2-chloro- 2-methylbutane and 2-bromo-2-methylbutane. Electronic structure calculations also were done to find threshold energies for several related molecules, including 2-chloro-3,3-dimethylbutane, 1-chloro-2-methyl-2-phenylpropane, and 1-chloro-2-methyl-2-vinylpropane, to demonstrate the generality of the interchange reaction involving a methyl, or other hydrocarbon groups, and a chlorine atom. The interchange of a halogen atom and a methyl group located on adjacent carbon atoms can be viewed as an extension of the halogen atom interchange mechanisms that is common in 1,2-dihaloalkanes.

Full text of DOI:10.1021/jp1047166

J. Phys. Chem. A 2010, 114, 10395–10402
10395
Isomerization of Neopentyl Chloride and Neopentyl Bromide by a 1,2-Interchange of a
Halogen Atom and a Methyl Group
Carmen E. Lisowski,† Juliana R. Duncan,† Anthony J. Ranieri,† George L. Heard,†
D. W. Setser,‡ and Bert E. Holmes*,†
Department of Chemistry, UniVersity of North Carolina-AsheVille, One UniVersity Heights, AsheVille, North
Carolina 28804-8511, and Department of Chemistry, Kansas State UniVersity, Manhattan, Kansas 66506
ReceiVed: May 22, 2010; ReVised Manuscript ReceiVed: August 14, 2010
The recombination of chloromethyl and t-butyl radicals at room temperature was used to generate neopentyl
chloride molecules with 89 kcal mol-1 of internal energy. The observed unimolecular reactions, which give
2-methyl-2-butene and 2-methyl-1-butene plus HCl, as products, are explained by a mechanism that involves
the interchange of a methyl group and the chlorine atom to yield 2-chloro-2-methylbutane, which subsequently
eliminates hydrogen chloride by the usual four-centered mechanism to give the observed products. The
interchange isomerization process is the rate-limiting step. Similar experiments were done with CD2Cl and
C(CH3)3 radicals to measure the kinetic-isotope effect to help corroborate the proposed mechanism. Density
functional theory was employed at the B3PW91/6-31G(d′,p′) level to verify the Cl/CH3 interchange mechanism
and to characterize the interchange transition state. These calculations, which provide vibrational frequencies
and moments of inertia of the molecule and transition state, were used to evaluate the statistical unimolecular
rate constants. Matching the calculated and experimental rate constants, gave 62 ( 2 kcal mol-1 as the threshold
energy for interchange of the Cl atom and a methyl group. The calculated models also were used to reinterpret
the thermal unimolecular reactions of neopentyl chloride and neopentyl bromide. The previously assumed
Wagner-Meerwein rearrangement mechanism for these reactions can be replaced by a mechanism that involves
the interchange of the halogen atom and a methyl group followed by HCl or HBr elimination from 2-chloro-
2-methylbutane and 2-bromo-2-methylbutane. Electronic structure calculations also were done to find threshold
energies for several related molecules, including 2-chloro-3,3-dimethylbutane, 1-chloro-2-methyl-2-phenyl-
propane, and 1-chloro-2-methyl-2-vinylpropane, to demonstrate the generality of the interchange reaction
involving a methyl, or other hydrocarbon groups, and a chlorine atom. The interchange of a halogen atom
and a methyl group located on adjacent carbon atoms can be viewed as an extension of the halogen atom
interchange mechanisms that is common in 1,2-dihaloalkanes.
I. Introduction
transfers a CH3 group and eliminates HCl (or HBr) by 1,1- and
1,3- processes, see reaction 2.
The thermal unimolecular reactions of neopentyl chloride and
neopentyl bromide can be summarized by eq 1 for experimental
conditions such that radical reactions are suppressed.1-4
(CH3)3CCH2Cl (or Br) f HCl (or HBr) +
CH2dC(CH3)C2H5 (1a)
f HCl (or HBr) +
(CH3)2CdCHCH3 (1b)
The thermal pyrolysis studies agree that 2-methyl-1-butene and
2-methyl-2-butene are the major products in an approximate
ratio of 2:1, and that the Arrhenius parameters for the neopentyl
Experiments3 with (CH3)3CCD2Cl were consistent with the
essential aspects of the mechanism, since the products were
2-methyl-1-butene-d2 and 2-methyl-2-butene-d1, see reaction 2.
This ionic mechanism of reaction 2 was proposed1,2 at the time
when the HX elimination mechanism also was considered to
proceed by a highly polar transition state. According to modern
views5-7 of the transition state for HX elimination, the partial
charges on the carbon atoms are not large. We wish to propose
an alternative mechanism for neopentyl chloride (or bromide)
that involves the unimolecular interchange of the chlorine atom
(or bromine atom) and a methyl group to give 2-chloro-2-
methylbutane (or 2-bromo-2-methylbutane), which subsequently
chloride reaction are 1013.8(0.6 s-1 and Ea ) 62 ( 2 kcal mol-1
.
Since elimination of HCl or HBr by a conventional four-centered
mechanism is not possible from neopentyl chloride or bromide,
investigators adopted a Wagner-Meerwein rearrangement
mechanism that involves formation of an ionic (or, at least, a
very polar) intermediate that subsequently (or simultaneously)
* To whom correspondence should be addressed. E-mail: bholmes@
unca.edu.
† University of North Carolina-Asheville.
‡ Kansas State University.
10.1021/jp1047166  2010 American Chemical Society
Published on Web 09/01/2010

10396 J. Phys. Chem. A, Vol. 114, No. 38, 2010
Lisowski et al.
eliminates HCl (or HBr) by the conventional four-centered
mechanism. Reactions 3 and 4 summarize the mechanism for
neopentyl chloride-d2. The polar nature of the interchange
transition state will be considered in the Discussion section.
Reaction 3 is exothermic by about 2 kcal mol-1; thus,
2-methyl-2-chlorobutane acquires an additional 2 kcal mol-1
of energy during the interchange in reaction 3. The threshold
energy, E0, for the interchange reaction is ≈60 kcal mol-1, based
on our electronic structure calculations and the thermal activation
studies of neopentyl chloride, whereas those for HCl elimination,
reactions 4a and 4b, are ≈45 kcal mol-1 according8 to pyrolysis
studies of 2-methyl-2-chlorobutane. Thus, the rearrangement
process in 3 is the rate-limiting step for all forms of activation
of neopentyl chloride (or bromide) molecules. Supporting
Information has a Figure illustrating the energy profile for the
formation and decomposition of neopentyl chloride via reactions
3 and 4. According to the interchange mechanism, the kinetic
isotope-effect for reaction 3 would just be the statistical
secondary effect associated with the two D atoms in neopentyl
chloride-d2. The product branching ratios in reaction 4 would
have different isotope effects, but they would be small. In
contrast, the single-step Wagner-Meerwein mechanism of
reaction 2 would have different kinetic-isotope effects for the
formation of 2-methylbutene-1 and 2-methylbutene-2, with the
latter being larger by a factor of 1.5, at least, since it is a primary
isotope effect.15-17 Measuring the experimental H/D kinetic-
isotope effect for the 2-methylbutene products would be an
additional test of whether the mechanism consisted of reaction
2 or reactions 3 and 4.
In the present work we provide chemical activation data for
neopentyl chloride and neopentyl chloride-d2 formed by recom-
bination of t-butyl and chloromethyl (CH2Cl and CD2Cl) radicals
plus electronic structure calculations using density functional
theory (DFT) in support of reaction 3. The experiments consist
of measurement of the ratios of collisionally stabilized product,
neopentyl chloride, to the decomposition products, 2-methyl-
1-butene and 2-methyl-2-butene, as a function of pressure. The
2-methyl-2-chlorobutane molecules with 91 kcal mol-1 of
vibrational energy are not stabilized at the pressures of these
experiments. The radicals were generated by the cophotolysis
of (CH3)3CI and CH2ClI (or CD2ClI).
Figure 1. Diagram of the transition state for interchange of a methyl
group and a chlorine atom from neopentyl chloride. The distances are
in Å; the angles, 16.2° and 13.5°, denote the inclination of the triangular
planes of the CH2 and C(CH3)2 groups with the C-C axis. Calculations
were done by the B3PW91 method with the 6-31G(d′,p′) basis set.
The partial charges on each atom, calculated by the atoms-in-molecules
method, are listed. The large negative charge on the Cl atom should
be noted. The transition state for neopentyl bromide is very similar to
the diagram for neopentyl chloride; the C-Br distances increase to
2.83 and 3.10 Å for C6 and C5, but the C6-C1 and C5-C1 distances
remain the same, that is, 1.80 and 1.95 Å.
43 (Br/Cl)7 to 60 (Cl/F)8-17 kcal mol-1. The Cl/methyl group
and Br/methyl group interchange reactions are members of this
genre of unimolecular reaction. The presence of two methyl
groups on one carbon of the transition state in neopentyl chloride
(bromide) serves to lower the E0 of reaction 3, and the
interchange reaction becomes viable at modest temperatures.
Electronic structure calculations were done using DFT at the
B3PW91 level with the 6-31G(d′,p′) basis set. Some calculations
also were done with the 6-311+G(2d,p) basis set, as well as
with the B3LYP method, for comparison. The structures of the
molecule and the transition state were very similar for all the
calculations. However, the E0 for the rearrangement does depend
somewhat on the method and basis set. Our collective experience
from studies of numerous HX elimination and X/Y interchange
reactions (X, Y are halogen atoms) is that the 6-31G(d′,p′) basis
set with the B3PW91 method most closely matches experimen-
tally based threshold energies. The structure of the Cl/CH3
interchange transition state, E0 ) 62 kcal mol-1, is shown in
Figure 1. The nearly planar nature of the two carbon atoms in
the bridged structure is evident. This also is a characteristic
feature of the transition states for halogen atom interchange.7,13-17
Another noteworthy feature is the very low torsional frequency,
73 cm-1, of the CH3 group in the bridged structure, which
implies that the bonding is with the π-orbital between the two
carbon atoms. The calculated frequencies and moments of inertia
of neopentyl chloride and its transition state that were used to
evaluate the statistical rate constants for the interchange reactions
of neopentyl chloride-d0 and -d2 are given in the Supporting
Information. Matching the calculated and experimental rate
The Cl/F, Br/Cl, and Br/F interchange reactions for haloal-
kanes with halogen atoms located on adjacent carbon atoms
have been demonstrated and characterized in a series of recent
studies from this laboratory.9-17 These interchange reactions,
which often are in competition with hydrogen halide elimination
reactions, have threshold energies for dihaloethanes ranging from

Isomerization of Neopentyl Chloride and Bromide
J. Phys. Chem. A, Vol. 114, No. 38, 2010 10397
constants using E0 as a variable permitted assignment of an
experimentally based threshold energy for reaction 3. The
models of the transition states were also used to obtain Arrhenius
pre-exponential factors for comparison to thermal activation
studies1-4 of neopentyl chloride and bromide. Exploratory
calculations also were done for other molecules to identify and
illustrate the class of molecules for which reactions are expected
to interchange Cl or Br atoms with a methyl or other organic
group.
II. Experimental Methods
A given experiment consisted of photolysis at room temper-
ature of measured quantities of CH2ClI (or CD2ClI) and
(CH3)3CI followed by gas chromatographic (GC) analysis for
the ratios of 2-methyl-1-butene/neopentyl chloride and 2-methyl-
2-butene/neopentyl chloride. Pyrex vessels with volumes ranging
from 34.7 to 4900 cm3 containing 0.540 µmol of CH2ClI (or
CD2ClI) and 0.180 µmol of (CH3)3CI were irradiated with the
output of a high pressure Oriel 6137 arc lamp. A small amount
of mercury(I) iodide was added to the vessels to aid in the
formation of the radicals and to remove the HCl product.18 The
use of Pyrex vessels limits the photolysis to wavelengths longer
than 290 nm. Irradiation times of 2-15 min gave approximately
10% conversion of the reactants to products. Gas samples were
measured on grease-free vacuum lines and pressures were
measured with a MKS 270 Signal Conditioner. The measured
samples of CH2ClI and (CH3)3CI were quantitatively transferred
from standard volumes, that were attached to the vacuum line,
to the photolysis vessels by cryogenic pumping. The entire
photolyzed sample was transferred to the inlet vacuum system
of the gas chromatographs for analysis of the products.
Products were identified from their retention times and mass
spectra by comparison to authentic samples from analysis with
a Shimadzu QP5000 gas chromatograph with a mass-spectrom-
eter detector. A Rtx-VMS 120 m column with 0.25 mm diameter
was used in the analysis. The temperature program began by
holding the column at 30 °C for 20 min, followed by heating at
a rate of 2 °C per min until the column reached 100 °C, then
the rate was increased to 4 °C per min until the temperature
was 200 °C. The main products were 2-methyl-1-butene,
2-methyl-2-butene, and neopentyl chloride. No evidence was
found for 1,1-dimethylcyclopropane or 3-methyl-1-butene,
which were minor products in some of the thermal pyrolysis
studies. An authentic sample of 1,1-dimethylcyclopropane was
available for confirmation. Photolysis of t-butyl iodide and
CD2ClI (CDN Isotopes) gave CH2dC(CH3)CD2CH3, (CH3)2Cd
CDCH3, and (CH3)3CCD2Cl as products.
Analysis of the reaction mixtures to obtain decomposition to
stabilization (D/S) product ratios (2-methyl-1-butene/neopentyl
chloride and 2-methyl-2-butene/neopentyl chloride) were done
with a Shimadzu GC-14A gas chromatograph with a flame
ionization detector and a Shimadzu CR5A Chromatopac inte-
grator. A Rtx-VGC column of 105 m length and 0.53 mm
diameter was used. The temperature program consisted of 20
min at 35 °C, followed by a heating rate of 12 °C per min until
a final temperature of 190 °C was reached. The typical retention
times were 2-methyl-1-butene (15.6 min), 2-methyl-2-butene
(17.4 min), neopentyl chloride (32.5 min), chloroiodomethane
(36.7 min), and t-butyl iodide (45.1 min). Deuterated compounds
had similar retention times as nondeuterated compounds.
Response factors for the products were measured by comparing
four samples from each of three independently prepared mixtures
containing CH2dC(CH3)CH2CH3, (CH3)2CdCHCH3, (CH3)3-
CCH2Cl, and CH2ClI. The response factors were 0.92 ( 0.05
Figure 2. Plot of the decomposition to stabilization ratios vs pressure-1
for neopentyl chloride-d0: O CH2dC(CH3)C2H5/(CH3)3CCH2Cl with
slope and intercept of 0.000 516 ( 0.000 081 and 0.0355 ( 0.0122
respectively, 0 (CH3)2CdCHCH3/(CH3)3CCH2Cl with slope and in-
tercept of 0.000 249 ( 0.000 033 and 0.0141 ( 0.0051, respectively,
[ [(CH3)2CdCHCH3 + CH2dC(CH3)C2H5]/(CH3)3CCH2Cl with slope
and intercept of 0.000 765 ( 0.000 112 and 0.0496 ( 0.0171
respectively.
for (CH3)2CdCHCH3/CH2dC(CH3)CH2CH3, 0.90 ( 0.06 for
CH2dC(CH3)CH2CH3/(CH3)3CCH2Cl, and 0.98 ( 0.11 for
(CH3)2CdCCHCH3/(CH3)3CCH2Cl.
III. Experimental Results
A. Rate Constants. The chemical activation rate constants
for the interchange reaction were obtained from the usual
measurement of the decomposition (D) to stabilization (S)
product ratio at various pressures; D/S ) kexp/kM[M] where kM
is the collision rate constant for neopentyl chloride with the
bath gases. In this simple formulation we are assuming strong
collisions, that is, sufficient internal energy is removed from
neopentyl chloride per collision such that further reaction has
negligible probability. The experimental results for neopentyl
chloride-d0 and -d2 are shown in Figures 2 and 3; the pressure
range is from 70 to 3.4 mTorr. The D/S range is limited to
e0.3 because of the small rate constants and the low-pressure
limit of our technique. The appropriate decomposition product
for obtaining kexpt for the interchange reaction is the sum of
2-methyl-1-butene and 2-methyl-2-butene, since all of the
(CH3)2C(Cl)CH2CH3* molecules will decompose at the pres-
sures of these experiments because of the low E0 of 45 kcal
mol-1 for HCl elimination. The slopes of the pressure-1 plots
in Figures 2 and 3 give the average high pressure rate constant,
kexpt, which is equivalent to k〈E〉. The values of the slopes for
the interchange reaction are (7.65 ( 0.41) × 10-4 and (5.76 (
0.69) × 10-4 Torr for neopentyl chloride-d0 and -d2, respectively.
The rate constants in pressure units of Torr were converted to
units of s-1, see Table 1, using the factor of 1.80 × 107 s-1
/
Torr, which was obtained using collision diameters and (ꢀ/k)19
values of 5.8 Å (425°), 5.1 Å (400°), and 5.3 Å (400°) for
neopentyl chloride, chloroiodomethane, and t-butyl iodide,
respectively in the standard formula kM ) πd2(8kT/πµ)1/2
Ω
(2,2)(T*). The resulting rate constants are (1.4 ( 0.2) × 104

10398 J. Phys. Chem. A, Vol. 114, No. 38, 2010
Lisowski et al.
B. Average Energy of Neopentyl Chloride. Under our
experimental conditions, the CH2Cl and C(CH3)3 radicals will
collide with bath molecules before recombining; thus, the
radicals are in thermal equilibrium. The average energy of the
molecules formed by the recombination of CH2Cl and t-butyl
radicals is given by the C-C bond dissociation energy at 0 K
plus 3RT plus the thermal vibrational energy of the radicals.
The 3RT arises from the 3 translational and 3 rotational motions
that become vibrations in the neopentyl chloride molecule. Since
the enthalpy of formation for neopentyl chloride does not seem
to be readily available, we decided to use the bond dissociation
energy of neopentane, (CH3)3C-CH3, as a model. The enthalpies
of formation at 298 K of CH3, C(CH3)3, and (CH3)4C are 35.0,20
12.3,20,21 and 39.722 kcal mol-1, which give D(CH3-C(CH3)3)
) 87.1 kcal mol-1 at 298 K. Converting to 0 K gives 85.5 kcal
mol-1. Adding the thermal energy of the radicals that contribute
to the internal energy of the neopentane molecules raises the
energy to 89.6 kcal mol-1. The bond dissociation energy of
propyl chloride is about 0.3 kcal mol-1 lower than that of
propane. Therefore, we adopted 89 kcal mol-1 as the average
thermal energy of neopentyl chloride formed by the recombina-
tion of (CH3)3C and CH2Cl radicals at room temperature. The
uncertainty is (3 kcal mol-1. The 2-methyl-2-chlorobutane
molecules will contain 91 kcal mol-1 of internal energy
following reaction 3.
Figure 3. Plot of decomposition to stabilization ratios vs pressure-1
for neopentyl chloride-d2; O CH2dC(CH3)CD2CH3/(CH3)3CCD2Cl with
slope and intercept of 0.000 419 ( 0.000 046 and 0.0174 ( 0.0075,
respectively, 0 (CH3)2CdCD(CH3)/(CH3)3CCD2Cl with slope and
intercept of 0.000 157 ( 0.000 022 and 0.006 79 ( 0.003 77, respec-
tively, [ [(CH3)2CdCDCH3 + CH2dC(CH3)CD2CH3]/(CH3)3CCD2Cl
with slope and intercept of 0.000 576 ( 0.000 066 and 0.0242 ( 0.0109,
respectively.
IV. Computational Results
and (1.0 ( 0.2) × 104 s-1 for neopentyl chloride-d0 and -d2.
The slopes of the plots in Figures 2 and 3 give an overall kinetic-
isotope effect of 1.33 ( 0.20, which is a statistical secondary
effect since the D atoms are not involved in the reaction
coordinate of the interchange reaction.
A. Models of Molecules and Transition States. Models for
the vibrational frequencies and moments of inertia of the
molecules and transition states are needed to calculate statistical
(RRKM) rate constants at a specific energy and Arrhenius pre-
exponential factors at a given temperature. The computed
threshold energies are of general interest, but they are not
sufficiently reliable to be used in the RRKM rate constant
calculations. Electronic structure calculations were done with
the suite of codes in the Gaussian package.23 We have continued
to use DFT with the B3PW91 method and the 6-31G(d′,p′) basis
set. Exploratory calculations24 also were done with the B3LYP
method and the 6-311+G(2d,p) and aug-cc-pVTZ basis sets.
However, the pre-exponential factors for the interchange reaction
from these calculations24 were nearly identical, as is usually
the case,25-27 and we adopted the 6-31G(d′,p′) basis set for this
work. The neopentyl chloride (or bromide) molecules have four
torsional modes and the transition states have three torsional
modes. We treated these modes as hindered internal rotations.
Our assignments of the reduced moments of inertia and the
barrier heights are given below. The computed vibrational
frequencies and moments of inertia are given in the Supporting
Information.
The barriers to internal rotation for the CH2Cl and CH2Br
groups in neopentyl chloride (bromide) were calculated with
the 6-31G(d′,p′) basis set as 5.8 and 6.1 kcal mol-1, and the
reduced moments were 25.5 and 31.8 amu Å2, respectively. We
have continued26 to employ the method of Pitzer to calculate
the reduced moments for internal rotation. The barriers for
methyl rotation in the t-butyl group were assigned as 4.8 kcal
mol-1 from the calculation for neopentane by Baudry;28 the
reduced moment was 3.1 amu Å2.
The (CH3)2C(Cl)CH2CH3 molecules decompose by 1,2- and
3,2-HCl elimination pathways with reaction path degeneracies
of 6 and 2, respectively. RRKM calculations for E0 ) 45 kcal
mol-1 were done for 1,2-HCl elimination from 2-methyl-2-
chlorobutane with 91 kcal mol-1 of energy. The rate constant
was 1.5 × 108 s-1 and that for 3,2-HCl elimination would be
similar for the same E0. The unimolecular reactions of 2-methyl-
2-chlorobutane are 4 orders of magnitude faster than the
interchange rate, and the 2-methyl-2-chlorobutane molecules will
decompose at all feasible pressures employed in experiments
for the study of neopentyl chloride. The branching between 1,2-
and 3,2-HCl elimination in reaction 4 depends on the relative
rate constants. The experimental branching ratios can be
obtained from the average ratio of 2-methylbutene-1 to 2-me-
thylbutene-2 from each experiment or from the ratio of the
slopes of the plots in Figure 2 and 3. Of course, they are very
similar, but we have chosen the former and they are 2.3 and
2.6 with the 1,2-HCl channel being favored for neopentyl
chloride-d0 and -d2, respectively. The branching ratio in the
pyrolysis experiments2 of neopentyl chloride was similar. The
reaction path degeneracies favor 1,2-HCl by a factor of 3, but
the 3,2-HCl elimination channel must have a slightly lower E0.
In fact, 3,2-HCl elimination to give 2-methyl-2-butene was
favored in pyrolysis of 2-chloro-2-methyl-butane,8 which also
suggests that the E0 is lower for 3,2-HCl elimination than for
1,2-HCl elimination. This is supported by the DFT calculations
of the threshold energies that give E0(1,2-HCl) ) 41.4 kcal
mol-1 and E0(3,2-HCl) ) 40.4 kcal mol-1, although the absolute
values are too low, see the Supporting Information. The kinetic-
isotope effect is larger for 3,2-DCl elimination than for 1,2-
HCl elimination from (CH3)3C(Cl)CD2CH3, and that is why the
branching ratio increased to 2.6, see Figure 3.
The structure and the charge distribution for the transition
state for neopentyl chloride are shown in Figure 1. An intrinsic
reaction coordinate calculation confirmed that Figure 1 is the
transition state for the interchange reaction 3. The carbon atoms
in the backbone of the transition state have nearly sp2 geometry
and the structure of the CH3 groups at the (CH3)2C end should

Isomerization of Neopentyl Chloride and Bromide
J. Phys. Chem. A, Vol. 114, No. 38, 2010 10399
(CH3)3CCH2Cl-d0 (-d2) (CH3)3CCH2Br-d0
TABLE 1: Comparison of Experimental and Calculated Results
property
kexpt; s-1
1.38 ( 0.2 × 104 (1.04 ( 0.2 × 104)
k〈E〉; s-1 E0 ) 62 kcal mol-1
E0 ) 61 kcal mol-1
1.51 × 104 (0.94 × 104)
2.68 × 104 (1.70 × 104)
1013.8(0.60
AArra; s-1
1014.2(0.3
1014.87
59 ( 1
(c)
Acalcb; s-1 T ) 700 K
Ea (expt)a; kcal mol-1
Ea (calc from best E0); kcal mol-1
1014.83
62 ( 2
64 ( 2
a The Arrhenius pre-exponential factors from pyrolysis experiments, refs 3 and 4. The Arrhenius parameters from ref 2 for neopentyl
chloride are 1013.26(0.36 s-1 and Ea ) 60.0 ( 1.6 kcal mol-1
.
b The DFT calculated pre-exponential factor at 700 K are from our models. c If the
calculated E0 of 61.1 kcal mol-1 is accepted, the Ea would be 63.3 kcal mol-1
.
TABLE 2: Computed Threshold Energies for the
Halogen-Methyl or Phenyl (Ph) Interchange Reaction using
B3PW91/6-31G(d′,p′)a
resemble that of (CH3)2CdCH2. The reported29 barrier for
internal rotation in isobutene is 2.1 kcal mol-1. We suspect that
this barrier may be a lower limit for the transition state, and we
selected 3.1 kcal mol-1 as the barrier for the two methyl groups;
the reduced moments are 3.18 amu Å2. The calculated torsional
frequency for the CH3 group in the bridge is 73 cm-1. This low
value indicates that the barrier to internal rotation is very low.
We assigned a barrier of 1.0 kcal mol-1; the reduced moment
is 3.03 amu Å2. The description of the internal rotations for the
neopentyl bromide transition state is the same as for neopentyl
chloride. The lower barriers for internal rotation of the CH3
groups in the transition state relative to those in the molecule
should be noted.
threshold
energy
reaction
(kcal/mol)
70.0
66.7
62.0
61.1
87.2
58.1
57.8
58.0
61.0
53.3
Me-Cl interchange in (CH3)3CCCl3
Me-Cl interchange in (CH3)3CCHCl2
Me-Cl interchange in (CH3)3CCH2Cl
Me-Br interchange in (CH3)3CCH2Br
Me-F interchange in (CH3)3CCH2F
Me-Cl interchange of (CH3)3CCH(CH3)Cl
Me-Cl interchange of (CH3)2(OCH3)CCH2Cl
Me-Cl interchange in (CH3)2PhCCCH2Cl
Me-Cl interchange of (CH3)2(CH)CH2)CCH2Cl
Ph-Cl interchange in (CH3)2PhCCHCl2
Statistical rate constants for neopentyl chloride-d0 and -d2
were calculated using the standard equation given below.
kE ) s†/h(I†/I)1/2
P†(E - E )/N*(E)
(5)
∑
0
a The DFT calculated E0 values for these reactions should be
similar if the Cl-atom is replaced by a Br-atom based on the
comparison between neopentyl chloride and neopentyl bromide.
The density of states, N*(E), for the molecules and the sums of
states for the transition state, ΣP†(E - E0) were obtained from
the Multiwell code of Barker.30 The reaction path degeneracy
is s†, and I†/I is the ratio of the three overall moments of inertia.
The reaction path degeneracy arises naturally from the symmetry
of the internal rotation of the t-butyl group. For neopentyl
chloride s† ) 3 and the square root of the ratio of the moments
of inertia was 1.22. The threshold energy was varied to obtain
agreement with k〈E〉 and kexpt. Due to the rather small 〈E〉 - E0,
about 28 kcal mol-1, and the large number of vibrational
frequencies, the dependence of kE on 〈E〉 and E0 is strong; a 2
kcal mol-1 increase in 〈E〉 gives a factor of 1.8 increase in kE,
and a 2 kcal mol-1 decrease in E0 gives a factor of 3.1 increase.
The models also can be used to calculate the Arrhenius pre-
exponential factor (s†kT/h exp(1 + ∆S†/R)) or the pre-
exponential factor in partition function form (s†kT/h) (Q†/Q).
B. Comparison of Experimental and Calculated Results
for Neopentyl Chloride. The experimental rate constants for
the chemical activation experiments of neopentyl chloride-d0
and -d2 and the experimental pre-exponential factors of neo-
pentyl chloride (and bromide) are compared in Table 1. The E0
needed to match k〈E〉 and kexpt for 〈E〉 ) 89 kcal mol-1 is 62 (
1 kcal mol-1 for both neopentyl chloride-d0 and neopentyl
chloride-d2. At 91 kcal mol-1 the calculated isotope effect (Table
2) is 1.55, which agrees with the upper range of the experimental
result of 1.33 ( 0.20. The threshold energy is invariant with
isotopic substitution. Considering the uncertainties in 〈E〉, in
the experimental measurement of kexpt, and in the harmonic
models for the calculations, the actual uncertainty in E0 is
probably (2 kcal mol-1. The E0 from the thermal activation
experiments is 60 ( 2 kcal mol-1. Thus, the thermal and
chemical activation experiments are in fundamental agreement
for the threshold energy of the CH3/Cl interchange reaction in
neopentyl chloride. The calculated pre-exponential factor in
Arrhenius form is 6.7 × 1014 s-1, which is 2.5 times larger than
the upper limit of the experimental result,1-4,8 A ) 2.5 × 1014
s-1. The large pre-exponential factor is a consequence of the
low bending vibrational frequencies associated with the weakly
bound Cl and CH3 groups of the transition state. In this context
it should be noted that the transition state still has three methyl
hindered internal rotors.
The electronic structure calculations gave E0 values of 62.0
kcal mol-1 at the B3PW91/6-31G(d′,p′) level and 59.8 kcal
mol-1 at the B3LYP/6-31G(d′,p′) level for the CH3/Cl inter-
change reaction of neopentyl chloride. The larger basis set,
6-311+G(2d,p), gave a somewhat lower, ∼ 3 kcal mol-1
,
threshold energy for each method of calculation. The aug-cc-
pVTZ basis set gave the same result as the 6-311+G(2d,p) basis
set. Thus, the calculations also support a threshold energy of
60-62 kcal mol-1
.
V. Discussion
Chemical activation studies of neopentyl bromide have not
been done, but comparison can be made with a thermal
activation study,4 which reported Arrhenius parameters of
1014.2(0.3 s-1 and Ea ) 59.0 ( 1.2 kcal mol-1. The model, which
is presented in the Supporting Information, from the 6-31G(d′,p′)
calculations gave a threshold energy of 61.1 kcal mol-1 and a
pre-exponential factor of 1014.9 s-1. The calculated and experi-
mental threshold energies and pre-exponential factors have the
same degree of agreement as for neopentyl chloride, and

10400 J. Phys. Chem. A, Vol. 114, No. 38, 2010
Lisowski et al.
Figure 4. (A) Diagrams of the transition states for interchange of a phenyl group and a chlorine atom and (B) interchange of a methyl group and
a chlorine atom from 1-chloro-2-methyl-2-phenylpropane. The distances are in Å; the two identified angles are the angles of inclination of the
planes of the end groups, e.g., the CH2 group, with the C-C axis. Calculations were with the B3PW91 method and the 6-31G(d′,p′) basis set. The
partial charges calculated by the atoms-in-molecules method33 are given for each atom in the tables. The similarity to the transition state for
neopentyl chloride in Figure 1 should be noted.
substituting a bromide atom for a chloride atom does not
seriously affect the unimolecular interchange reaction.
CCl3 groups in place of the CH2Cl group. The threshold energies
were raised by about 4 kcal mol-1 for each additional chlorine
atom. However, the interchange reaction still should be the
unimolecular reaction with the lowest E0 for these two mol-
ecules. Calculations for CH3/F interchange in neopentyl fluoride
gave a threshold energy of 87.2 kcal mol-1, and the interchange
reaction of methyl or other groups with a fluorine atom will
have very high activation energies and interchange will not be
important. Chuchani and co-workers31 have suggested that the
small (10%) yield of 2,3-dimethyl-1-butene and 2,3-dimethyl-
2-butene from pyrolysis of 2-chloro-3,3-dimethylbutane (pina-
colyl chloride) arises from a Wagner-Meerwein rearrangement.
However, a CH3/Cl interchange mechanism would be preferred
by analogy to neopentyl chloride. Adding a third methyl group
to the interchange transition state shown in Figure 1 should
lower the threshold energy, and the interchange pathway
becomes competitive with 1,2-HCl elimination, which gives
(CH3)3CCHdCH2 from (CH3)3CCHCl(CH3). The reported31b
activation energy based on experiments over only a 55 K range
was 47.2 kcal mol-1 for the total decomposition rate constant
of pinacolyl chloride; however, the pre-exponential factor seems
to be too small and the experimental Ea probably is a lower
limit.
Our DFT calculations confirm the effect of the third CH3
group and the calculated threshold energies for 1,2-HCl elimina-
tion and interchange were 44.3 and 58.1 kcal mol-1, respec-
tively; however, calculated E0 values for HCl elimination by
this method frequently are too low.
How general is the interchange reaction between a Cl or Br
atom and an adjacent methyl (or other organic group)? Chuchani
and Dominguez32 have studied the unimolecular reactions of
1-chloro-2-methyl-2-phenylpropane. The principal products
(96%) were 2-methyl-3-phenyl-1-propene and 2-methyl-3-
The methyl/halogen interchange mechanism is consistent with
all the kinetic data for neopentyl chloride and bromide. In
particular the experimental kinetic-isotope effect of 1.33 for
neopentyl chloride-d0 and -d2 is the expected statistical effect
for the interchange rate constant and agrees with the compu-
tational findings. However, it is not the result expected for a
one-step Wagner-Meerewein mechanism. For the later, a
small statistical effect would hold for the formation of
CH2dC(CH3)CD2CH3; however, a primary effect would aug-
ment the statistical effect for formation of (CH3)2CdC(D)CH3
and the combined isotope effect would be g2 for this channel.
Such a difference was not experimentally observed, since the
ratios are 1.23 (1,2-HCl) and 1.58 (2,3-HCl), based upon the
slopes of the plots in Figures 2 and 3. In the interchange
mechanism, the intermolecular isotope effect for the two
competing HCl elimination channels is given by the ratio of
interchange rate constants for the d0 and d2 molecules multiplied
by the ratios of branching fractions for the formation of
CH2dC(CH3)C2H5 and (CH3)2CdCHCH3. The isotope effect
on the branching fractions is small as illustrated by the
intramolecular branching ratios of 2.3 and 2.6 for d0 and d2,
respectively. In addition to the experimental results, the
electronic structure calculations gave plausible transition states
for structures representing CH3/Cl or Br interchange, whereas
no structures could be identified that would support a
Wagner-Meerwein type transition state. We conclude that the
CH3/halogen atom interchange mechanism adequately explains
the unimolecular reactions of neopentyl chloride and neopenthyl
bromide.
Exploratory calculations, see the summary of threshold
energies in Table 2, were done for molecules with CHCl2 and

Isomerization of Neopentyl Chloride and Bromide
J. Phys. Chem. A, Vol. 114, No. 38, 2010 10401
VI. Conclusions
phenyl-1-propene. The authors32 proposed a Wagner-Meerwein
rearrangement mechanism to explain the products. However,
the interchange of a phenyl group and a Cl atom to give
1-phenyl-2-chloro-2-methylpropane followed by HCl elimina-
tion is a more likely mechanism. Our DFT calculations using
B3PW91/6-31G(d′,p′) identified a transition state, see Figure
4A, for the interchange of the phenyl group and the Cl atom
with a threshold energy of 53 kcal mol-1, which is in excellent
agreement with the experimental activation energy of 54 ( 2
kcal mol-1. In addition, the calculations identified a transition
state for CH3/Cl interchange, see Figure 4B, to give 2-chloro-
2-phenylbutane with a threshold energy of 58 kcal mol-1, which
is consistent with the products (4%) from this pathway. Similar
calculations also were done for 1-chloro-2-methyl-2-vinylpro-
pane, and the threshold energy for methyl interchange with the
Cl atom was 61.0 kcal mol-1. Electron donating groups are
expected to lower the threshold energy for the interchange
reaction, and this possibility was tested by calculations for
1-chloro-2-methoxy-2-methylpropane; the threshold energy for
CH3/Cl interchange was 57.8 kcal mol-1, which is 3 kcal mol-1
lower than for neopentyl chloride. Electron attracting groups,
such as halogen atoms, increase the threshold energy for CH3/
Cl interchange reactions as demonstrated by the (CH3)3CCHCl2
and (CH3)3CCCl3 examples in Table 2. On the basis of the
examples in Table 2, there should be a sizable group of
molecules for which interchange of methyl (or other) groups
and Cl or Br atoms is an important unimolecular gas-phase
reaction, provided that the competing HCl or HBr elimination
reactions are either blocked or have high threshold energies.
As a final part of the Discussion, we wish to consider the
structure of the CH3/Cl (or CH3/Br) interchange transition state
and its relation to the transition states for halogen atom
interchange reactions.7,11-17 One clear difference is the larger
entropy of activation for the CH3 case, because of the weakly
bound CH3 group with low torsion and bending frequencies.
The Cl (or Br) atom is further from the carbon atoms for the
methyl-interchange transition state than for the Cl/Br-interchange
transition state,7 and the carbon backbone is less planar in the
Cl/CH3 transition state. A second major difference is the
asymmetric charge distribution for the CH3 case, as illustrated
by the very negative Cl (or Br) atom in Figures 1 and 4, whereas,
the charge distribution for halogen atom interchange tends to
be quite symmetric7 with both halogen atoms sharing the
negative charge. The negative Cl atom causes a strong polariza-
tion of the C-H bonds of the two CH3 groups that are directed
toward the Cl atom as shown in Figures 1 and 4. In fact, these
H-atoms in the CH3 groups are the most positive atoms of the
structures of Figures 1 and 4A. In summary, the negative charge
is mainly localized on the halogen atom, but the positive charge
is dispersed over all the other atoms, except the carbon atom in
the CH3 group of the bridge. The charge distributions were
obtained by the atoms-in-molecule approach.33 The structures
in Figures 1 and 4 have some interesting features. The methyl
and phenyl groups, as well as the Cl atom, are nearer to the
CH2 end than to the C(CH3)2 ends of the transition state. The
carbon atom of the phenyl group in the phenyl/Cl transition
state is closer to the bridgehead carbon atom than is the carbon
atom of the CH3 group for either CH3/Cl transition state. Despite
lowering E0 by 4 kcal mol-1 for CH3 interchange, the structures
in Figures 1 and 4B are very similar; however, the positive
charges on the bridgehead carbon atoms (C5 and C6) are slightly
lower with the phenyl substituent.
The interchange of a methyl group and the Cl (or Br) atom
in neopentyl chloride (and bromide) has been demonstrated for
chemically activated neopentyl chloride-d0 and -d2 and for
thermally activated neopentyl chloride and bromide. The
threshold energies are 60-62 kcal mol-1. This interchange
mechanism followed by HCl (or HBr) elimination from 2-chloro
(or bromo)-2-methylbutane replaces the previously proposed
Wagner-Meerwein rearrangement mechanism. A small kinetic
H/D isotope effect of 1.33 ( 0.2 measured for (CH3)3CCH2Cl/
(CH3)3CCD2Cl substantiates the interchange-elimination mech-
anism. The CH3/Cl interchange reaction seems to compete with
HCl elimination in the thermal reaction of 2-chloro-3,3-
dimethylbutane. Other examples of interchange reactions of a
methyl group and a Cl atom were discussed using DFT
calculations to find the transition states and threshold energies.
The interchange of a phenyl group and a Cl atom probably
occurs in the thermal reaction of 1-chloro-2-methyl-2-phenyl-
propane. Electronic structure calculations predict that CH3/Cl
interchange should be observable for (CH3)2(OCH3)CCH2Cl and
(CH3)2(CHdCH2)CCH2Cl. The halogen atom/methyl group
interchange reactions resemble the interchange reactions involv-
ing two adjacent halogen atoms of dihaloalkane molecules.
Acknowledgment. Financial support from the National
Science Foundation (CHE-0647582) and (MRI-0320795) is
acknowledged.
Supporting Information Available: Tables of the molecular
and transition state structure vibrational frequencies, overall
moments of inertia, and the reduced moments of inertia for the
internal rotors calculated using B3PW91/6-31G(d′,p′) for
CH2BrC(CH3)3 and CH2ClC(CH3)3. A figure of an energy profile
for the formation and decay of chemically activated neopentyl
chloride calculated with the same DFT method. This information
is available free of charge via the Internet at http://pubs.acs.org.
References and Notes
(1) Maccoll, A.; Swinbourne, E. S. J. Chem. Soc. 1964, 149.
(2) Shapiro, J. S.; Swinbourne, E. S. Can. J. Chem. 1968, 46, 1341–
1351.
(3) Failes, R. L.; Mollah, Y. M. A.; Shapiro, J. S. Int. J. Chem. Kinet.
1979, 11, 1271.
(4) Failes, R. L.; Mollah, Y. M. A. Int. J. Chem. Kinet. 1981, 13, 7.
(5) Toto, J. L.; Pritchard, G. O.; Kirtman, B. J. Phys. Chem. 1994, 98,
8359 This reference gives a good summary of the various models for the
transition states for HX elimination from alkyl halides.
(6) Duncan, J. R.; Solaka, S. A.; Setser, D. W.; Holmes, B. E. J. Phys.
Chem. A 2010, 114, 794.
(7) Friederich, L.; Duncan, J. R.; Heard, G. L.; Setser, D. W.; Holmes,
B. E. J. Phys. Chem. A 2010, 114, 4138.
(8) Maccoll, A.; Wong, S. C. J. Chem. Soc., B 1968, 1492.
(9) Beaver, M. R.; Heard, G. L.; Holmes, B. E. Tetrahedrom Lett. 2003,
44, 7265.
(10) Dolbier, W. R., Jr.; Romelaeer, R.; Baker, J. M. Tetrahedron. Lett.
2002, 43, 8075.
(11) Burgin, M. O.; Heard, G. L.; Martell, J. M.; Holmes, B. E. J. Phys.
Chem. A 2001, 10, 615.
(12) Heard, G. L.; Holmes, B. E. J. Phys. Chem. A 2001, 105, 1622.
(13) Burgin, M. O.; Simmons, J. G., Jr.; Heard, G. L.; Setser, D. W.;
Holmes, B. E. J. Phys. Chem. A 2007, 111, 2283.
(14) (a) Zaluzhna, O.; Simmons, J. G., Jr.; Heard, G. L.; Setser, D. W.;
Holmes, B. E. J. Phys. Chem. A 2008, 112, 6090. (b) Zaluzhna, O.;
Simmons, J. G., Jr.; Setser, D. W.; Holmes, B. E. J. Phys. Chem. A 2008,
112, 12117.
(15) (a) Holmes, D. A.; Holmes, B. E. J. Phys. Chem. A 2005, 109,
10726. (b) Zhu, L.; Simmons, J. G., Jr.; Burgin, M. O.; Setser, D. W.;
Holmes, B. E. J. Phys. Chem. A 2006, 110, 1506.
(16) Beaver, M. R.; Simmons, J. G., Jr.; Heard, G. L.; Setser, D. W.;
Holmes, B. E. J. Phys. Chem. A 2007, 111, 8445.

10402 J. Phys. Chem. A, Vol. 114, No. 38, 2010
Lisowski et al.
(17) Lisowski, C. E.; Duncan, J. R.; Heard, G. L.; Setser, D. W.; Holmes,
B. E. J. Phys. Chem. A 2008, 112, 441.
(18) Holmes, B. E.; Paisley, S. D.; Rakestraw, D.; King, E. E. J. Int.
J. Chem. Kinet. 1986, 18, 639.
(19) Hippler, H.; Troe, J.; Wendelken, H. J. J. Chem. Phys. 1983, 78,
6709.
(20) Berkowitz, J.; Ellison, G. B.; Gutman, D. J. Phys. Chem. 1994,
98, 2744.
(21) Kynazev, V.; Dubinsky, I. A.; Slagle, I. R.; Gutman, D. J. Phys.
Chem. 1994, 98, 5279.
(22) Kerr, J. A.; Stocker, D. W. Standard Thermodynamic Properties
of Chemical Substances, In Handbook of Chemistry and Physics; Linde,
D. R., Jr., Ed.; CRC Press; Boca Raton, FL, 2007.
(23) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kuden, K. N.;
Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.;
Mennucci, B.; Cossi, M.; Scalmani, G; Bega, N; Petersson, G. A.; Nakatsuji,
H.; Hada, M.; Ehara, M.; Toyota, K. ; Fukuda, R.; Hasegawa, J.; Ishida,
M.; Nakajima, T; Honda, Y.; Kitao, O.; Adamo, C.; Jaramillo, J.; Gomperts,
R.; Stratman, R. E.; Yazyev, O.; Austen, A. J.; Cammi, R.; Pomelli, C.;
Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.;
Dannenberg, J. J.; Zakrzewksi, V. G.; Dapprich, S.; Daniels, A. D.; Strain,
M. C.; Farkas, O.; Malik, D. K.; Rabuck, A. D.; Raghavachari, K.;
Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski,
J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.;
Martin, R. L.; Fox, D. J.; Keith, T. Al-Laham, M. A.; Peng, C. Y.;
Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen,
W.; Wong, M. W.; Gonzalez, C. Pople, J. A. Gaussian 03, ReVision B.04;
Gaussian, Inc.: Pittsburg, PA, 2003.
(24) Ranieri, A. J.; Heard, G. L.; Holmes, B. E. Unpublished Results,
2009.
(25) Duncan, J. R.; Roach, M. S.; Stiles, B. S.; Holmes, B. E. J. Phys.
Chem. A 2010, 114, 6996.
(26) Ferguson, J. D.; Johnson, N. L.; Kekenes-Huskey, P. M.; Everett,
W. C.; Heard, G. L.; Setser, D. W.; Holmes, B. E. J. Phys. Chem. A 2005,
109, 4540.
(27) Rajakumar, B.; Aruman, E. E. Phys. Chem. Chem. Phys. 2003, 5,
3897.
(28) Baudry, J. J. Am. Chem. Soc. 2006, 128, 11088.
(29) (a) Durig, J. R.; Natter, J.; Groner, P. J. Chem. Phys. 1977, 67,
4948. (b) Bruhn, T.; Stahl, W. J. Mol. Spectrosc. 2000, 202, 272.
(30) Barker, J. R. Int. J. Chem. Kinet. 2001, 33, 232.
(31) (a) Dominguez, R. M.; Rotinov, A.; Chuchani, G. J. Phys. Chem.
1986, 90, 6277. (b) Chuchani, G.; Martin, I.; Alonso, M. E. Int. J. Chem.
Kinet. 1977, 9, 819.
(32) Chuchani, G.; Dominguez, R. M. J. Chem. Soc. Perkins Trans.
1994, 2, 1499.
(33) Bader, F. R. Atoms in Molecules s A Quantum Theory; Clarendon
Press: Oxford, 1990.
JP1047166