Dynamics of the gas-phase reactions of chloride ion with fluoromethane: High excess translational activation energy for an endothermic SN2 reaction

Article abstract of DOI:10.1021/ja012031f

Guided ion beam tandem mass spectrometry techniques are used to examine the competing product channels in the reaction of Cl^- with CH₃F in the center-of-mass collision energy range 0.05-27 eV. Four anionic reaction products are detected: F^-, CH₂Cl^-, FCl^- and CHCl^-. The endothermic S_N2 reaction Cl^- + CH₃F → CH₃Cl + F^- has an energy threshold of E₀ = 181 ± 14 kJ/mol, exhibiting a 52 ± 16 kJ/mol effective barrier in excess of the reaction endothermicity. The potential energy of the S_N2 transition state is well below the energy of the products. Dynamical impedances to the activation of the S_N2 reaction are discussed, including angular momentum constraints, orientational effects, and the inefficiency of translational energy in promoting the reaction. The fluorine abstraction reaction to form CH₃ + FCl^- exhibits a 146 ± 33 kJ/mol effective barrier above the reaction endothermicity. Direct proton transfer to form HCl is highly inefficient, but HF elimination is observed above 268 ± 95 kJ/mol. Potential energy surfaces for the reactions are calculated using the CCSD(T)/aug-cc-pVDZ and HF/6-31+G(d) methods and used to interpret the dynamics.

Full text of DOI:10.1021/ja012031f

Published on Web 12/15/2001
Dynamics of the Gas-Phase Reactions of Chloride Ion with
Fluoromethane: High Excess Translational Activation Energy
for an Endothermic SN2 Reaction
Laurence A. Angel, Sandra P. Garcia, and Kent M. Ervin*
Contribution from the Department of Chemistry and Chemical Physics Program,
UniVersity of NeVada, Reno, NeVada 89557
Received August 22, 2001
Abstract: Guided ion beam tandem mass spectrometry techniques are used to examine the competing
product channels in the reaction of Cl- with CH3F in the center-of-mass collision energy range 0.05-27
eV. Four anionic reaction products are detected: F-, CH2Cl-, FCl-, and CHCl-. The endothermic SN2
reaction Cl- + CH3F f CH3Cl + F- has an energy threshold of E0 ) 181 ( 14 kJ/mol, exhibiting a 52 (
16 kJ/mol effective barrier in excess of the reaction endothermicity. The potential energy of the SN2 transition
state is well below the energy of the products. Dynamical impedances to the activation of the SN2 reaction
are discussed, including angular momentum constraints, orientational effects, and the inefficiency of
translational energy in promoting the reaction. The fluorine abstraction reaction to form CH3 + FCl- exhibits
a 146 ( 33 kJ/mol effective barrier above the reaction endothermicity. Direct proton transfer to form HCl
is highly inefficient, but HF elimination is observed above 268 ( 95 kJ/mol. Potential energy surfaces for
the reactions are calculated using the CCSD(T)/aug-cc-pVDZ and HF/6-31+G(d) methods and used to
interpret the dynamics.
Introduction
independent endothermic reactions, proton transfer and halide
Experimental1 and theoretical2,3 studies of gas-phase bimolecular
nucleophilic substitution (SN2) in halomethanes have provided
detailed information on the dynamics, mechanisms, and energy
dependence of reaction 1, where X and Y are halogen atoms.
abstraction, contributing at collision energies between 1 and 20
eV. Here we present complementary work on the reverse
endothermic reaction 3. This investigation is one of the first
Cl- + CH3F f ClCH3 + F-
(3)
X- + CH3Y f XCH3 + Y-
(1)
detailed energy-resolved studies of an endothermic SN2 reaction.
The only other report of a strongly endothermic SN2 reaction,
to our knowledge, is that of Zellermann and Vietzke,8 who used
an ion beam/gas cell technique without initial mass selection
to measure the collision energy dependence of the products H-,
F-, FBr-, and CH2Br- from the reaction Br- + CH3F, and the
products F-, FI-, and CH2I- from the reaction I- + CH3F. The
halogen abstraction reaction, resulting in the XY- ion, has also
been observed in three previous guided ion beam experimentss
The strong attraction between the halide ion and the dipolar
halomethane results in a double-well potential energy surface.4-6
The two potential energy minima correspond to the formation
of ion-molecule intermediates on both the reactant and the
product sides of the reaction. The two minima are separated by
a central energy barrier corresponding to the five-coordinate
[X-CH3-Y]- transition state.
We recently applied guided ion beam mass spectrometry
techniques to investigate the cross section behavior of reaction
2 in the collision energy range 0.05-30 eV.7 The exothermic
the ClBr- ion from the reaction Cl- + CH3Br,9 the Cl2 ion
-
from the reaction Cl- + CH3Cl,10 and the FCl- ion in our
previous work on reaction 2.7
Here we show that in the collision energy range 2-27 eV
the endothermic reactions 4-6 can be driven by translational
energy in competition with reaction 3.
F- + CH3Cl f FCH3 + Cl-
(2)
reaction 2 dominates at the lowest collision energies with two
* Corresponding author. E-mail: ervin@chem.unr.edu.
(1) Chabinyc, M. L.; Craig, S. L.; Regan, C. K.; Brauman, J. I. Science 1998,
279, 1882-1886.
Cl- + CH3F f HF + CH2Cl-
(4)
(2) Shaik, S. S.; Schlegel, H. B.; Wolfe, S. Theoretical Aspects of Physical
Organic Chemistry: The SN2 Mechanism; John Wiley and Sons: New York,
1992.
(8) Zellermann, G.; Vietzke, E. Radiochem. Acta 1990, 50, 107-115.
(3) Hase, W. L. Science 1994, 266, 998-1002.
(9) Cyr, D. A.; Scarton, M. G.; Wiberg, K. B.; Johnson, M. A.; Nonose, S.;
Hirokawa, J.; Tanaka, H.; Kondow, T.; Morris, R. A.; Viggiano, A. A. J.
Am. Chem. Soc. 1995, 117, 1828-1832.
(4) Olmstead, W. N.; Brauman, J. I. J. Am. Chem. Soc. 1977, 99, 4219-4228.
(5) Dodd, J. A.; Brauman, J. I. J. Phys. Chem. 1986, 90, 3559-3562.
(6) Brauman, J. I. J. Mass Spectrom. 1995, 30, 1649-1650.
(7) Angel, L.; Ervin, K. J. Phys. Chem. A 2001, 105, 4042-4051.
(10) DeTuri, V. F.; Hintz, P. A.; Ervin, K. M. J. Phys. Chem. A 1997, 101,
5969-5986.
9
336 VOL. 124, NO. 2, 2002 J. AM. CHEM. SOC.
10.1021/ja012031f CCC: $22.00 © 2002 American Chemical Society

Gas-Phase Reactions of Chloride Ion with Fluoromethane
Cl- + CH3F f CH3 + FCl-
A R T I C L E S
(5)
(6)
Cl- + CH3F f [HF + H] + CHCl-
The possible mechanisms of reactions 3-6 can be broadly
divided into the two categories of back-side and front-side attack.
For the back-side attack, the path of the approaching Cl- lies
on the methyl side of CH3F. The SN2 mechanism for reaction
1 proceeds through a back-side attack at carbon, resulting in
inversion of the methyl group. In the front-side attack mecha-
nism, the Cl- approaches the CH3F molecule from the C-F
bond side and may lead to displacement or abstraction. To help
investigate the microscopic reaction mechanisms for reactions
3-6, the coupled cluster CCSD(T)/aug-cc-pVDZ and HF/6-
31+G(d) methods are used to calculate potential energy surfaces
(PES).
Experimental Methods
Chloride anions are produced in a microwave discharge source from
tetrachloromethane, CCl4, added in trace amounts in a flowing buffer
gas of helium. The Cl- anions pass along a flow tube and are sampled
through a nose cone into the guided ion beam tandem mass spectrom-
eter.10 The anions are then shaped, focused, and accelerated by a series
of lenses to a magnetic mass spectrometer, which mass selects the 35Cl-
ions before they are injected into an octopole radio frequency ion trap.
Situated at the center of the octopole is a reaction cell where the
fluoromethane (Matheson Tri-Gas, 99%) reactant gas is introduced.
The collision energy between the chloride ions and the neutral
fluoromethane is controlled by the dc potential difference between the
flow tube ion source and the octopole. The anionic reactants and
products are then extracted from the octopole and injected into a
quadrupole mass spectrometer where they are mass analyzed. Ion
intensities are detected by a collision dynode/channeltron multiplier
operated in negative-ion pulse counting mode.
Absolute reaction cross sections are determined as a function of
collision energy by scanning the octopole dc potential and counting
the reactant and product ions for predetermined dwell times.10 The
laboratory ion energy is measured using retarding potential analysis,
confirmed by time-of-flight measurements, and converted to relative
collision energy, E, in the center-of-mass (c.m.) frame.10,11 Background
ion counts occurring outside the reaction cell are also collected and
subtracted from the total. All cross sections are measured at three
pressures in the range (5-20) × 10-5 mbar. The results are extrapolated
to zero pressure by a least-squares linear regression, ensuring that the
reported cross sections are in the single-collision limit.10,11
Figure 1. Cross sections for the product ions F-, CH2Cl-, FCl-, and CHCl-
from the reaction of Cl- + CH3F as a function of relative collision energy
in the center-of-mass frame.
The reported error limits are propagated from individual sources of
uncertainty (assuming they are independent of each other) and represent
(2 combined standard uncertainties19 or an approximate 95% confi-
dence level. Uncertainties are included for the determination of the
ion beam energy zero, the reproducibility of data taken on separate
occasions, the vibrational frequencies, the least-squares fit of eq 7, and
the consistency of the fits using different energy ranges.
Results
The experimental reaction cross sections from 0.05 to 27 eV
c.m. are shown in Figure 1. Four products ions, F-, CH2Cl-,
FCl-, and CHCl-, are observed from reactions 3-6. Table 1
lists experimental endothermicities of possible product chan-
nels.20-29 The mass ranges of the octopole and quadrupole
prevent detection of light products, H- or e-. In survey mass
spectra, a few ion counts per second of CH2F- were sometimes
observed, but the signal was near the detection limit of the GIB
apparatus and the channel was too weak to obtain cross sections
as a function of energy. The SN2 product ion F- from reaction
3 is detected at the lowest collision energies. The cross section
rises from approximately 1.5 eV and exhibits a maximum of
The threshold behavior of the cross section, σ(E), is modeled by
using an empirical threshold law,10-14
σ(E) ) σ0
gi [E + Ei - E0]N/E
(7)
∑
i
where Ei is the internal energy of reactant state i with fractional thermal
population gi corresponding to a Boltzmann distribution at 300 K, σ0
and N are adjustable parameters, and E0 is the 0 K reaction threshold
energy. Experimental15 vibrational frequencies and rotational constants
of CH3F are used for the sum over the reactant internal energy density
of states.13,14 Finally, eq 7 is convoluted over the experimental collision
energy distributions16,17 as described previously.11 These calculations
are performed using the CRUNCH data analysis program.18
(15) Herzberg, G. Molecular Spectra and Molecular Structure II. Infrared and
Raman Spectra of Polyatomic Molecules; Van Nostrand Reinhold: New
York, 1945.
(16) Chantry, P. J. J. Chem. Phys. 1971, 55, 2746-2759.
(17) Lifshitz, C.; Wu, R.; Tiernan, T. O.; Terwilleger, D. T. J. Chem. Phys.
1978, 68, 247-260.
(18) Armentrout, P. B.; Ervin, K. M. CRUNCH; Fortran program.
(19) Taylor, B. N.; Kuyatt, C. Guidelines for EValuating and Expressing the
Uncertainty of NIST Measurement Results; NIST Technical Note 1297;
National Institute of Standards and Technology: Washington, DC, 1994
(http://physics.nist/gov/Document/tn1297.pdf).
(11) Ervin, K. M.; Armentrout, P. B. J. Chem. Phys. 1985, 83, 166-189.
(12) Armentrout, P. B. Int. J. Mass Spectrom. 2000, 200, 219-241.
(13) Schultz, R. H.; Crellin, K. C.; Armentrout, P. B. J. Am. Chem. Soc. 1991,
113, 8590-8601.
(14) Dalleska, N. F.; Honma, K.; Sunderlin, L. S.; Armentrout, P. B. J. Am.
Chem. Soc. 1994, 116, 3519-3528.
9
J. AM. CHEM. SOC. VOL. 124, NO. 2, 2002 337

A R T I C L E S
Angel et al.
Table 1. Threshold Energies and Enthalpies for the Reaction Cl- + CH3F f Products (kJ mol-1
)
a
reaction
products
E (this work)
0
N
∆H (exptl)b
∆H (G3)c
∆H (CCSD(T)/aug-cc-pVDZ)d
0
0
0
Cl- + CH3F
0
0
0
134
237
304
320
325
338
370
342
348
417
417
452
462
472
551
604
690
737
766
766
783
800
804
0
133
238
3
4
F- + CH3Cl
181 ( 14
268 ( 95
2.5
1.0
129 ( 8e
223 ( 20f
300 ( 13
309 ( 21g
319 ( 22h
333 ( 13
334 ( 22
341 ( 9
CH2Cl- + HF
CH2Cl + HF + e-
CH2F- + HCl
5
FCl- + CH3
465 ( 24
2.5
314
CH2F + HCl + e-
H- + CH2ClF
Cl- + CH2 + HF
Cl + CH3F + e-
H + CH2ClF + e-
Cl- + CH2F + H
Cl- + F + CH3
F + CH3Cl + e-
F- + Cl + CH3
FCl + CH3 + e-
CHCl- + HF + H
Cl + CH2 + HF + e-
CHCl- + H2 + F
Cl + CH2F + H + e-
FCl-+ CH + H2
FCl- + CH2 + H
Cl + F + CH3 + e-
CH2Cl- + H + F
348 ( 11
407 ( 22
434 ( 18
452 ( 8
457 ( 8
8
6
9
472 ( 8
547 ( 8
653 ( 26
0.5
588 ( 31i
690 ( 9
585
722 ( 31i
761 ( 13
763 ( 22h
775 ( 22h
800 ( 8
790 ( 20g
a Fitting parameter in eq 7. b Enthalpies of reactions calculated with enthalpies of formation values cited in Gurvich et al.20,21 except as noted. c Calculated
here or obtained from http://chemistry.anl.gov/compmat/comptherm.htm.22 d Corrected for zero-point energy from frequencies calculated at the B3LYP/
aug-cc-pVDZ level. e From ∆fH0(CH3F) ) -225 ( 8 kJ/mol estimated by Kolesov,23 which agrees with theoretical calculations by Berry et al.24 f Calculated
using EA(CH2Cl) ) 0.80 ( 0.16 eV determined by Bartmess from work by Ingemann and Nibbering.25 g Calculated using ∆fH0(CH2F-) ) -53 ( 19
kJ/mol determined by Bartmess26 from work by Graul and Squires.27 h Calculated using EA(FCl) ) 2.37 ( 0.21 eV published by Dudin et al.28 i Calculated
using EA(CHCl) ) 1.210 ( 0.005 eV published by Gilles et al.29
0.6 × 10-16 cm2 at about 5 eV, before declining at energies
above 6 eV. The cross section for methylene abstraction, reaction
4, exhibits a dual rising feature. An initial small rise at 2.5 eV
to 0.004 × 10-16 cm2 is followed by a second onset originating
at about 4 eV. The second feature rises to a maximum of 0.07
× 10-16 cm2 at about 8 eV before declining to the experimental
detection limit by 16 eV. The FCl- ion is the product of halide
abstraction reaction 5 and exhibits a rising cross section from
about 4.5 eV. At 9 eV, the FCl- cross section exhibits a
maximum of 0.3 × 10-16 cm2. As the FCl- cross section starts
to decline at energies above 9 eV, there is a small but observable
increase in the F- cross section, evidence that there is a small
contribution from reaction 8 to F- formation. Finally, originating
reaction 6 is detected. After an initial steep rise, the CHCl-
cross section continues to rise more slowly and forms a peak at
about 17 eV, suggesting that reaction 9, a more endothermic
process (Table 1), is also contributing at these higher energies.
Cl- + CH3F f [F + H2] + CHCl-
(9)
Threshold energies, E0, for the four endothermic reactions
3-6 are obtained by fits to eq 7 and are compared with
established thermochemical values in Table 1. The fits of the
empirical threshold law to the rising experimental cross section
data are shown in Figure 2. Reactions 3, 5, and 6 all exhibit
steep cross-section rises, allowing for reasonable empirical
threshold fits. For reactions 3 and 5, good fits using eq 7 can
be achieved up to only about 2 eV above the apparent threshold.
Reaction 4 exhibits a small initial rise in the cross section,
resulting in a large uncertainty in the threshold energy.
Cl- + CH3F f F- + [Cl + CH3]
(8)
at apparent collision energies of 6 eV, the CHCl- ion from
(20) Gurvich, L. V.; Veyts, I. V.; Alcock, C. B. Thermodynamic Properties of
IndiVidual Substances, 4th ed.; Hemisphere Publishing Corp.: New York,
1989; Vol. 1 (Elements O, H (D, T), F, Cl, Br, I, He, Ne, Ar, Kr, Xe, Rn,
S, N, P and Their Compounds), Parts 1-2.
(21) Gurvich, L. V.; Veyts, I. V.; Alcock, C. B. Thermodynamic Properties of
IndiVidual Substances, 4th ed.; Hemisphere: New York, 1991; Vol. 2
(Elements C, Si, Ge, Sn, Pb, and Their Compounds), Parts 1-2.
(22) Curtiss, L. A.; Raghavachari, K.; Redfern, P. C.; Rassolov, V.; Pople, J.
A. J. Chem. Phys. 1998, 109, 7764-7776. (Supplementary material at http://
chemistry.anl.gov/compmat/comptherm.htm).
Theoretical Methods and Potential Energy Surfaces
Coupled cluster (CCSD(T)),30 density functional theory (B3LYP),31
Hartree-Fock (HF), and Gaussian-3 (G3)22 calculations have been
performed using Gaussian 98.32 The potential energy surfaces (PES)
(30) Pople, J. A.; Head-Gordon, M.; Raghavachari, K. J. Chem. Phys. 1987,
87, 5968-5975.
(31) Becke, A. D. J. Chem. Phys. 1993, 98, 5648-5652.
(23) Kolesov, V. Russ. Chem. ReV. 1978, 47, 1145-1168.
(24) Berry, R.; Burgess, D.; Nyden, M.; Zachariah, M.; Schwartz, M. J. Phys.
Chem. 1995, 99, 17145-17150.
(32) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M.
A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A., Jr.; Stratmann,
R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A. D.; Kudin,
K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi, M.; Cammi,
R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.; Ochterski, J.;
Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick, D. K.;
Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Cioslowski, J.; Ortiz,
J. V.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.;
Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng,
C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill, P. M. W.;
Johnson, B. G.; Chen, W.; Wong, M. W.; Andres, J. L.; Head-Gordon,
M.; Replogle, E. S.; Pople, J. A. Gaussian 98, revision A.6; Gaussian,
Inc.: Pittsburgh, PA, 1998.
(25) Ingemann, S.; Nibbering N. M. M. J. Chem. Soc., Perkin Trans. 2 1985,
837, 837-840.
(26) Bartmess, J. E. Negative Ion Energetics Data. In NIST Chemistry WebBook,
NIST Standard Reference Database Number 69; Mallard, W. G., Linstrom,
P. J., Eds.; National Institute of Standards and Technology: Gaithersburg,
MD, Feb 2000 (http://webbook.nist.gov).
(27) Graul, S. T.; Squires, R. R. J. Am. Chem. Soc. 1990, 112, 2517-2529.
(28) Dudin, A. V.; Gorokhov, L. N.; Baluev, A. V. IzV. Akad. Nauk SSR Ser.
Khim. 1979, 2408.
(29) Gilles, M.; Ervin, K.; Ho, J.; Lineberger, W. J. Phys. Chem. 1992, 96,
1130-1141.
9
338 J. AM. CHEM. SOC. VOL. 124, NO. 2, 2002

Gas-Phase Reactions of Chloride Ion with Fluoromethane
A R T I C L E S
Figure 3. Potential energy surface for reaction 3 in C3V symmetry. The
energy relative to reactants calculated at the CCSD(T)/aug-cc-pVDZ level
without ZPE correction is plotted versus the difference between the C-Cl
and C-F bond lengths.
The authors37 of the calculations at the highest level of theory,
CCSD(T)/aug-cc-pVQZ, recommend 13.8 ( 1.3 kJ/mol after
considering basis set superposition error and additional electron
correlation and basis set effects but without correction for ZPE.
Our previous work7 on reaction 2 found that a restricted C3V
ion-dipole potential energy surface may be insufficient for
describing the SN2 reaction because of the strong hydrogen-
bonding propensity of the fluoride ion. The Cs F-‚‚‚H-CH2Cl
hydrogen-bonded complex is 5 kJ/mol more stable than the C3V
F-‚‚‚CH3Cl ion-dipole complex at the CCSD(T)/aug-cc-pVDZ
level (Table 2). Figure 4 shows the PES of reaction 3
recalculated in Cs symmetry and including the hydrogen-bonded
F-‚‚‚H-CH2Cl complex 4b. The depth of the exit well is now
increased from 8 kJ/mol for 3c to 13.2 kJ/mol for 4b (Table 2).
The PES exhibits a second transition state, 4a, separating the
two ion-molecule minima 3c and 4b. Figure 4 is constructed
from CCSD(T)/aug-cc-pVDZ energy minimizations at all
stationary points and along the curves with fixed r(C-F) or
r(C-Cl), except for the points that join the two minima to the
transition state 4a. These points are interpolated from a
Synchronous Transit-Guided Quasi-Newton (QST3)40,41 calcula-
tion of the reaction path at the B3LYP/aug-cc-pVDZ level,
which confirms that the two minima (3c and 4b) are connected
by the transition state 4a. (QST3 calculations at the CCSD(T)
level were not feasible computationally.) Starting at the transition
state 3b, we can see that as the fluoride ion starts to exit along
the C3V axis, the PES shows an initial relaxation into the C3V
ion-dipole minimum 3c. The ion-dipole minimum 3c, how-
ever, is not the lowest energy ion-molecule intermediate in Cs
Figure 2. Cross sections for the threshold regions for reactions 3-6 as a
function of relative collision energy in the center-of-mass frame. The circles
are experimental data, and the solid lines are the best fits of eq 7 convoluted
as described in the text.
for reactions 3 and 5 are calculated at the CCSD(T) level with the
aug-cc-pVDZ basis set33 and are shown in Figures 3-5. To avoid the
high computational cost of CCSD(T) frequencies, the stationary points
are confirmed by auxiliary geometry and frequency calculations at the
B3LYP/aug-cc-pVDZ level,31 which is also used for vibrational zero-
point energy (ZPE) corrections. The CCSD(T)/aug-cc-pVDZ reaction
enthalpies and calculated SN2 stationary point energies for reactions
3-6 are shown in Tables 1 and 2, respectively. Comparison of the
CCSD(T)/aug-cc-pVDZ reaction enthalpies with the experimental
thermochemical values show good correlation for reactions 3-6. G3
reaction enthalpies of a range of possible reactions resulting from Cl-
+ CH3F are also included in Table 1. The G3 reaction enthalpies show
overall good agreement with the experimental reaction enthalpies.
Results
The PES for reaction 3 has been calculated at the CCSD-
(T)/aug-cc-pVDZ level in C3V symmetry and is shown in Figure
3, which includes the geometries of the entrance ion-dipole
complex (3a), the SN2 transition-state central barrier (3b), and
the ion-dipole exit complex (3c). The C3V PES exhibits a central
barrier height of 113.2 kJ/mol relative to 3a and a small well
for the exit C3V F-‚‚‚CH3Cl ion-dipole complex (3c) of 8 kJ/
mol relative to 3b, with corrections for ZPE as shown in Table
2. Previous studies using HF, MP2, B3LYP, QCISD, CCSD-
(T), and G2(+) have shown a wide variation of results for the
depth of the C3V 3c well, with values of 0-26 kJ/mol.7,34-39
(37) Botschwina, P.; Horn, M.; Seeger, S.; Oswald, R. Ber. Bunsen-Ges. Phys.
Chem. 1997, 101, 387-390.
(38) Wang, H.; Hase, W. L. J. Am. Chem. Soc. 1997, 119, 3093-3102.
(39) Su, T.; Wang, H.; Hase, W. L. J. Phys. Chem. A 1998, 102, 9819-9828.
(40) Peng, C.; Schlegel, H. B. Isr. J. Chem. 1994, 33, 449.
(41) Peng, C.; Ayala, P. Y.; Schlegel, H. B.; Frisch, M. J. J. Comput. Chem.
1996, 17, 49-56.
(33) Woon, D. E.; Dunning, T. H., Jr. J. Chem. Phys. 1993, 98, 1358-1371.
(34) Wolfe, S. Can. J. Chem. 1984, 62, 1465.
(35) Shi, Z.; Boyd, R. J. J. Am. Chem. Soc. 1990, 112, 6789-6796.
(36) Glukhovtsev, M. N.; Pross, A.; Radom, L. J. Am. Chem. Soc. 1996, 118,
6273-6284.
9
J. AM. CHEM. SOC. VOL. 124, NO. 2, 2002 339

A R T I C L E S
Angel et al.
Figure 5. Potential energy surface for reaction 5 with Cs symmetry. The
energy relative to reactants calculated at the CCSD(T)/aug-cc-pVDZ level
without ZPE correction is plotted versus the difference between the C-Cl
and Cl-F bond lengths.
Table 2. Stationary State Energies (kJ mol-1
)
CCSD(T)/
aug-cc-pVDZ
CCSD(T)/
aug-cc-pVQZ
a
b
complex or transition state
3a C3V Cl-‚‚‚CH3F ion-dipole complex
3b C3V [ClsCH3sF]- transition state
3c C3V ClH3C‚‚‚F- ion-dipole complex
4a Cs [ClsCH3sF]- transition state
4b Cs ClCH2sH‚‚‚F- hydrogen-
bonded complex
-41.7
71.5
63.5
65.3
58.3
-39.3 (-40.2)
80.0 (84.4)
67.7 (71.6)
a Energies at 0 K relative to Cl- + CH3F reactants and corrected for
zero-point energy calculated at the B3LYP/aug-cc-pVDZ level. b Botschwina
et al.37 Values include B3LYP/aug-cc-pVDZ zero-point energies with the
original reported values reported in parentheses.
Figure 4. Potential energy surface for reaction 3 in Cs symmetry. The
energy relative to reactants calculated at the CCSD(T)/aug-cc-pVDZ level
without ZPE correction is plotted versus the difference between the C-Cl
and C-F bond lengths and includes the Cs hydrogen-bonded minimum
energy complex. The lower plot is a detail of the area on the Cs potential
energy surface, where the two ion-molecule minima are separated by a Cs
transition state.
between 0 and 1 Å is the result of partial inversion of CH3 due
to the attraction of the departing FCl-, followed by reversion
to the asymptotic planar geometry.
symmetry, and the complex may further relax to the hydrogen-
bonded complex 4b if it first passes over the Cs transition state
structure 4a, which is 1.8 kJ/mol higher in energy than 3c (Table
2). The energy needed to pass over the lower transition state
corresponds to approximately one quantum of Cl-C-F bend
vibrational energy (1.6 kJ/mol). The three-fold degenerate
complex 4b on the PES effectively deepens the well of the exit
ion-molecule intermediate and in turn may influence the
reactivity and final products of the Cl- + CH3F reaction, as
will be discussed in more detail below.
The PES for the halophilic attack reaction (eq 5) in C3V
symmetry is shown in Figure 5. The PES exhibits a repulsive
entrance surface with a steep exit channel out to the products
FCl- and CH3. There is no minimum potential energy structure
along the repulsive surface. To model the exit channel out to
the correct products, the PES in that region is calculated using
the unrestricted method UCCSD(T)/aug-cc-pVDZ. This is
necessary because the restricted CCSD(T)/aug-cc-pVDZ method
dissociates into FCl2- and CH3+. The unrestricted wave function
exhibits a considerable amount of spin contamination, so the
overall results may have only qualitative significance. However,
the calculation implies that there is no significant intrinsic barrier
in excess of the endoergicity for reaction 5. A small feature in
the exit surface of the PES at r(C-F)-r(Cl-F) distances
We have used HF/6-31+G(d) calculations to illustrate the
topology of the PES for reactions 2 and 3 in two dimensions.
The pseudo-collinear PES in C3V symmetry is shown in Figure
6. The contour plot is mapped on a mass-weighted coordinate
system42 with the axes skewed at an angle of 51.4°. The
positions of the hydrogen atoms are optimized for each pair of
r(C-Cl) and r(C-F) values. Figure 6 shows that the energy
barrier corresponding to 3b lies nearer to F- + CH3Cl than to
Cl- + CH3F. The stationary point geometries and relative
energies at the HF/6-31+G(d) level are in qualitative agreement
with the CCSD(T)/aug-cc-pVDZ calculations. The main dif-
ference is that the distances between the ion and the molecular
species in 3a and 3c are 0.2 and 0.1 Å longer at the HF/6-
31+G(d) level, respectively.
Discussion
SN2 Reaction Cross Section Behavior. The threshold energy
from the fit of eq 7 to the cross section for reaction 3 shown in
Figure 2 is E0 ) 181 ( 14 kJ/mol, or 52 ( 16 kJ/mol in excess
of the established experimental endothermicity (Table 1). Thus,
the system exhibits a large excess barrier for translational
(42) Levine, R. D.; Bernstein, R. B. Molecular Reaction Dynamics and Chemical
ReactiVity; Oxford University Press: New York, 1987.
9
340 J. AM. CHEM. SOC. VOL. 124, NO. 2, 2002

Gas-Phase Reactions of Chloride Ion with Fluoromethane
A R T I C L E S
Figure 6. Contour plot of the potential energy surface of reaction 3 treated
as a pseudo-collinear C3V reaction. The contours are mapped on the mass-
skewed coordinate system (see text). The potential energies are calculated
at the HF/6-31+g(d) level. The contour intervals are 20 kJ/mol, with
supplementary contours (dashed lines) at 130 and -30 kJ/mol relative to
Cl- + CH3F.
activation. The observed threshold energy does not necessarily
represent an energy where the SN2 reaction suddenly turns on,
but rather reflects the inefficiency of translational activation and
the sensitivity threshold of the experiment (about 0.0005 Å2).
We previously calculated the energy of the Cs transition state
(CH3ClF-) for a front-side nucleophilic exchange mechanism.7
The potential energy barrier for that process is about 99 kJ/mol
above the products of reaction 3, which is too high to explain
the observed reaction at its onset. Therefore, the system must
pass over the SN2 central barrier, albeit at elevated energies.
The SN2 potential energy surface for reaction 3 (Figures 3
and 4) exhibits no energy barrier above the product energy. The
excess translational activation energy for reaction 3, therefore,
results from some constrictions that are not due to an overall
energy barrier. This contrasts with the general behavior for
endothermic ion-molecule reactions, which often yield thresh-
old energies equal to the thermochemical endoergicity of
reaction.12 Among the known exceptions are cases with an actual
potential energy barrier above the products, a change of spin
or electronic state symmetry between reactants and products,
or competition with a lower-energy channel, none of which
applies here. For exothermic SN2 reactions, the double-well
potential is known to cause inefficient reactivity even when the
central barrier at the SN2 transition state lies well below the
product energies. For instance, reaction 2, the reverse of reaction
3 in the exothermic direction, has a somewhat low reaction
efficiency (56-61%) compared with the collision rate at thermal
energies,7,43 but it certainly does not show an activation energy
approaching the 52 ( 16 kJ/mol excess barrier that we observe
for reaction 3. This behavior shows that the effective excess
barrier for reaction 3 is dynamic in origin. Apparently, the
influence of the SN2 central barrier is greatly magnified for the
endothermic reaction. Here we discuss three effects that may
contribute to the observed high threshold energy: angular
momentum barriers, orientational or steric constraints, and
translational versus vibrational energy requirements.
Figure 7. (a) Effective potential energy for reaction 3 in C3V symmetry
including the centrifugal potential energy term for various values of Eb2 as
labeled in units of kilojoules per mole per square angstrom, where E is the
relative collision energy and b is the impact parameter. (b) Collision theory
cross sections for reaction 3 for crossing the central barrier and the exit
barrier. The solid lines correspond to fixed C3V geometry, i.e., constrained
to the most favorable ion-dipole orientation. The dotted line for the exit
barrier is based on parametrized trajectory calculations50 for a CH3Cl
rotational temperature of 300 K.
(i) Angular Momentum Barriers. Angular momentum
conservation has been previously discussed as a dynamical
impediment to translational activation of reaction 1.10,38,44 High
orbital angular momenta can be generated by collisions at high
translation energy and nonzero impact parameters. The angular
momentum must be conserved as the system passes through
the transition-state region. Because the moment of inertia at the
tight transition state is smaller than that of the colliding reactants
or the ion-dipole complexes, high rotational energies are
required to conserve angular momentum, which in turn reduces
the energy available along the reaction coordinate. Figure 7a
illustrates the effective potential energy including the angular
momentum term (centrifugal energy) along the C3V reaction
coordinate. The effective potential energy in the spherical
(43) Su, T.; Morris, R. A.; Viggiano, A. A.; Paulson, J. F. J. Phys. Chem. 1990,
94, 8426-8430.
(44) Mann, D. J.; Hase, W. L. J. Phys. Chem. A 1998, 102, 6208-6214.
9
J. AM. CHEM. SOC. VOL. 124, NO. 2, 2002 341

A R T I C L E S
Angel et al.
approximation is given by eq 10, where V(s) is the real potential
collinear collision model), the lower of the two at a given energy
represents the probability for reaching products in this model.
The central angular momentum barrier is more limiting than
the exit channel barrier for a range of collision energies above
the threshold, E ) 140-190 kJ/mol. The first crossing point at
E ) 140 kJ/mol corresponds roughly to the effective potential
energy curve labeled Eb2 ) 800 (kJ/mol) Å2 in Figure 7a, for
which the central and exit-channel effective barriers are about
the same height. Although the details likely would change with
a more realistic multidimensional treatment, this behavior
indicates that the central angular momentum barrier can reflect
reactants before they reach the exit-channel complex under some
conditions. HoweVer, both calculated cross sections are much
larger than the observed cross sections (comparing the cross
section scales in Figures 1 and 7), implying that other constraints
besides angular momentum conservation are more important
overall. Because the C3V and locked-dipole geometries are the
most favorable reaction orientation for reaction 3, an overesti-
mate of the cross section magnitude on the basis of this simple
collision model is not too surprising. This conclusion leads to
the following consideration of orientational effects.
(ii) Orientational Constraints. Passage over the SN2 central
barrier in Figure 3 requires, at low energies, that the Cl- be
aligned for back-side attack along the CF bond axis. The height
of the SN2 barrier increases as the Cl-C-F angle deviates from
180°. For the exothermic reaction 2, orientational effects have
been invoked to explain the rapid decrease in the reaction
efficiency with increasing collision energy.7,39,43 The endother-
mic reaction is likely to have a direct reaction mechanism
already at the threshold collision energy E > ∆3E ) 129 kJ/
mol because there is not enough time for reorientation during
a collision event. Trajectory studies of reaction 1 where X )
Cl or F and Y ) Cl also show direct reactions.3,39,48,49 Therefore,
reaction 3 should have a strong orientation dependence.
Orientational effects at the exit-channel effective barrier can
be estimated by relaxing the locked-dipole model discussed
above. For the ion-dipole potential, Su50 has used classical
trajectory calculations to parametrize the capture rate, applicable
to exothermic ion-molecule reactions. As discussed by DeTuri
and Ervin,51 the corresponding cross section for the reverse
endothermic reaction can be calculated by microscopic revers-
ibility.52 Specifically, the microscopic reversibility relation-
ship42,52 between the cross sections for reactions 2 and 3 is given
by eq 14, where σ2(E) and σ3(E) are the cross sections of
L2
2I(s)
µEb2
I(s)
Veff(s) ) V(s) +
) V(s) +
(10)
energy along the (generic) reaction coordinate s, L ) µVb is
the orbital angular momentum, µ is the reduced mass of
reactants, V and E ) µV2/2 are the initial relative collision
velocity and energy, b is the impact parameter, and I is the
moment of inertia of the complex. The moment of inertia is
specified as a function of s to emphasize that it varies along
the reaction path. At long range, I ) µr2, where r is the distance
between the reactants. In Figure 7, we have calculated the
moment of inertia from the CCSD(T) C3V geometries by treating
Cl-CH3-F as a linear pseudo-triatomic with central atom mass
15 amu (in error by less than 3% compared with calculated
moments of inertia at the SN2 transition state). Figure 7 shows
that the angular momentum barrier at the SN2 transition state
becomes dominant for high values of E and b.
Thus, angular momentum constraints likely play a role in the
reaction dynamics. However, their importance depends on
whether the necessary high values of the impact parameters are
actually reached in reactive collisions. Figure 7b shows the
calculated cross sections from simple collision theory42,45 for
crossing the SN2 barrier and the exit channel centrifugal barrier.
Specifically, we require that E g Veff(max), which leads to the
maximum impact parameter as a function of collision energy,
b
max(E). This yields eq 11 for the cross section for crossing the
central barrier,44 where ITS ) 228 (g/mol) Å2 is the moment of
πITS E - ∆TSE
2
σTS(E) ) πbmax
)
(11)
(
)
µ
E
inertia at the CCSD(T) SN2 transition-state geometry and ∆TSE
) 71.5 kJ/mol is the central barrier height relative to reactants.
For crossing the exit channel centrifugal barrier, we use the
locked-dipole approximation46 because that is consistent with
the C3V reaction path. The locked-dipole effective potential
energy is given by eq 12, where r is the distance between
R′e2
µEb2
µ′r2
µD′e
Veff(r) ) -
-
+
+ ∆3E (12)
4πꢀ0r2 2(4πꢀ0)2r4
departing F- + CH3Cl products and primed quantities refer to
products of reaction 3, µD′ ) 1.896 D and R′/4πꢀ0 ) 5.35 Å3
are the dipole moment and polarizability of CH3Cl,47 and ∆3E
is the endoergicity of reaction 3. Solving for bmax(E) in the usual
way,42,45 we obtain the cross section for crossing the exit-channel
centrifugal barrier at the orbiting transition state, eq 13.
E - ∆3E
µ′
σ3(E) ) σ (E - ∆ E)
(14)
(
)
2
3
µ
E
reactions 2 and 3 as a function of collision energy E. The result
is compared with the locked-dipole approximation in Figure 7
for a CH3Cl rotational temperature of 300 K. The actual
rotational energy distribution for the products of reaction 3 is
unknown, but the comparison shows that the cross section for
crossing the exit-channel centrifugal barrier decreases when
product rotation is included. Because the short-range orienta-
tional dependence near the central barrier is stronger than the
πµ′e
(4πꢀ0)µE
[µD′ + (2R′)1/2(E - ∆3E)1/2]
2
σ
exit(E) ) πbmax
)
(13)
Equations 11 and 13 are compared in Figure 7b. Since each
represents an upper limit (within the approximations of the
(45) Steinfeld, J. I.; Francisco, J. S.; Hase, W. L. Chemical Kinetics and Reaction
Dynamics, 2nd ed.; Prentice Hall: Upper Saddle River, NJ, 1998.
(46) Su, T.; Bowers, M. T. Classical Ion-Molecule Collision Theory. In Gas-
Phase Ion Chemistry; Bowers, M. T., Ed.; Academic: New York, 1979;
pp 83-118.
(47) Lide, D. R., Ed. CRC Handbook of Chemistry and Physics, 75th ed.; CRC
Press: Boca Raton, FL, 1994.
(48) Hase, W. L.; Cho, Y. J. J. Chem. Phys. 1993, 98, 8626-8639.
(49) Tachikawa, H.; Igarashi, M. Chem. Phys. Lett. 1999, 303, 81-86.
(50) Su, T. J. Chem. Phys. 1994, 100, 4703.
(51) DeTuri, V. F.; Ervin, K. M. J. Phys. Chem. A, submitted for publication.
(52) Levine, R. D.; Bernstein, R. B. J. Chem. Phys. 1972, 56, 2281-2287.
9
342 J. AM. CHEM. SOC. VOL. 124, NO. 2, 2002

Gas-Phase Reactions of Chloride Ion with Fluoromethane
A R T I C L E S
(or at least the same state distribution) as the reactants for
reaction 3. In our guided ion beam experiment on reaction 3,
the state distribution is a 300 K thermal distribution for CH3F.
The large deviation between the microscopic reversibility model
and the experimental cross section means, therefore, that the
products from reaction 2 must actually be highly internally
excited relative to a 300 K distribution. High internal excitation
of products is consistent with the theoretical studies of reaction
2 predicting high vibrational excitation.39,49 Kinetic energy
release studies of reaction 1 with (X,Y) ) (Cl,Br) and (Br,I)
and other SN2 reactions of alkyl halides all show higher internal
energies in products than predicted statistically,54-56 although
an ion cyclotron resonance study of reaction 2 reports high
translational energies of products.57
The pseudo-collinear PES in Figure 6 can be interpreted in
terms of typical Polanyi behavior for a collinear reaction.3,58
Translational energy promotes passage over a relatively early
transition state as in the case for the exoergic direction, reaction
2. After translational passage over the transition state, the
reaction in the F- + CH3Cl direction must pass the “bend” along
the collinear path, which will result in much of the exoergicity
of the reaction being converted into vibrational energy.42
Conversely, for the endoergic reaction 3 to occur, vibrational
energy is needed in the C-F stretching mode of CH3F for
passing around the initial tight bend in the reaction path before
surmounting the late transition state. The contour plot, therefore,
illustrates a dynamical impediment resulting from the lack of
the vibrational energy needed for reaction 3 to proceed ef-
ficiently. The methyl group enforces the collinearity of the SN2
reaction in the back-side attack mechanism, so the vibrational
effect and the orientational effect are coupled. Trajectory
simulations by Vande Linde and Hase59 predicted a direct SN2
mechanism promoted by vibrational excitation for reaction 1
where X ) Y ) Cl.
Figure 8. (a) Experimental cross section for reaction 2 from ref 7. (b) The
cross section for reaction 3 calculated from the cross section for reaction 2
using the simplified microscopic reversibility relationship of eq 14 (circles)
and the actual experimental reaction cross section for reaction 3 (squares).
The initial rising of the microscopic reversibility cross section (dashed line)
was calculated from cross sections for reaction 2 extrapolated to lower
energies.
long-range dipole orientation potential, the reduction in σTS(E)
should be greater.
The mass-skewed PES in Figure 6 illustrates another dynamic
difference between the endothermic and exothermic directions.
Consider a trajectory of Cl- + CH3F for reaction 3 with 180
kJ/mol or less of translational energy and no vibrational energy.
Specular reflections off the cirque on the repulsive wall at the
top of the entrance valley will preferentially send the trajectories
back to Cl- + CH3F reactants. In contrast, a trajectory of F- +
CH3Cl in the exothermic direction with a total energy of 180
kJ/mol on the scale of Figure 6 (relative to Cl- + CH3F
products) will hit the repulsive wall at a slant and tend to reflect
the system toward Cl- + CH3F products. These dynamics are
consistent with the observed efficiency of the exothermic
A collision theory model with an orientation-dependent
energy barrier yields a high value for the adjustable exponent
in eq 7, N g 2 compared with N ) 1 predicted for the line-
of-centers hard-sphere collision model.42,53 The empirical fit to
the cross section gives N ) 2.5 using eq 7, consistent with
orientational constraints. However, although this effect explains
a slow-rising cross section above the threshold, it cannot
completely explain a large shift in the threshold energy.
(iii) Translational versus Vibrational Energy Activation.
Direct ab initio dynamics calculations by Tachikawa and
Igarashi49 and trajectory studies by Su, Wang, and Hase39 found
that for the exothermic reaction 2, the reaction exoergicity is
preferentially partitioned into the C-F stretching vibrations in
the product. Considering microscopic reversibility,42 therefore,
the energy needed to promote reaction 3 should be supplied in
the C-F stretching mode of the CH3F reactant. Figure 8 shows
the result of the microscopic reversibility relationship of eq 14
applied to our previously reported experimental cross section
for reaction 2. The microscopic reversibility model predicts that
σ3(E) would have a magnitude 600 times greater than the
experimental cross section at E ) 1.5 eV, in obvious disagree-
ment. Equation 14 is valid for a single state-to-state process,
i.e., only if the products from reaction 2 are in the same state
reaction (56-61% of the collision rate at thermal energies7,43
)
and our observed effective threshold of 181 ( 14 kJ/mol for
reaction 3. At higher energies for reaction 3, the inner repulsive
wall becomes more favorable for reflecting trajectories toward
the F- + CH3Cl product valley.
In summary, the observed elevated effective barrier for
reaction 3 can be explained by a combination of angular
(54) Graul, S. T.; Bowers, M. T. J. Am. Chem. Soc. 1991, 113, 9696-9697.
(55) Graul, S. T.; Bowers, M. T. J. Am. Chem. Soc. 1994, 116, 3875-3883.
(56) Graul, S. T.; Carpenter, C. J.; Bushnell, J. E.; van Koppen, P. A. M.; Bowers,
M. T. J. Am. Chem. Soc. 1998, 120, 6785-6796.
(57) VanOrden, S. L.; Pope, R. M.; Buckner, S. W. Org. Mass Spectrom. 1991,
26, 1003-1007.
(58) Polanyi, J. C.; Wong, W. H. J. Chem. Phys. 1969, 51, 1439-1449.
(59) Vande Linde, S. R.; Hase, W. L. J. Am. Chem. Soc. 1989, 111, 2349-
2351.
(53) Ervin, K. M. Int. J. Mass Spectrom. 1999, 185/186/187, 343-350.
9
J. AM. CHEM. SOC. VOL. 124, NO. 2, 2002 343

A R T I C L E S
Angel et al.
momentum barriers, orientational effects, and the inefficiency
of activation by translational energy. At higher energies, the
orientational acceptance angle of the potential energy surface
for back-side attack opens up, and the dynamic constraints on
translational energy activation become less severe, allowing the
reaction to proceed. Classical trajectory calculations on a realistic
multidimensional potential energy surface could test these
conclusions. Detailed experimental information on the partition-
ing of product energies into vibrational and translational modes
for reactions 2 and 3 would also be revealing.
Methylene Abstraction and Formation of CH2Cl-. Figure
1 shows that above collision energies of about 2 eV, methylene
abstraction and HF elimination (reaction 4) occur with the
reaction cross section exhibiting a dual rising feature. The
empirical threshold fit to the initial small rise in the cross section
is shown in Figure 2 and exhibits a threshold energy of E0 )
268 ( 95 kJ/mol. The large error bars prevent determination
of whether there is an excess barrier relative to the experimental
endothermicity of 223 ( 20 kJ/mol for reaction 4 (Table 1),
but HF must be the neutral product on thermochemical grounds.
transition state 3b followed by internal proton transfer via 4b
and the elimination of HF.
The dual feature in the rising CH2Cl- cross section suggests
that either a second reaction or a more efficient mechanism is
contributing at collision energies above 4 eV. A switching on
of the CH2Cl- + H + F product channel is excluded by the
high endothermicity of that reaction, 790 ( 20 kJ/mol (Table
1). The reaction forming Cl- + CH2 + HF can explain the initial
decline at 3.5 eV in the CH2Cl- reaction cross section (Figure
2). This reaction represents a loss mechanism for CH2Cl-
through its dissociation to Cl- + CH2. Table 1 shows that the
reaction has an endothermicity of 341 ( 9 kJ/mol, in agreement
with the energy where the initial CH2Cl- cross section rise is
interrupted. At energies above 400 kJ/mol, however, the cross
section evidently recovers and starts to increase again. The
mechanism for this second higher rising feature is not certain,
but the threshold energy has an approximate correlation with
the rising cross section of the FCl- ion. The reaction, therefore,
may be promoted when the energy is available to dissociate
the CH3-F bond and the HF formation reaction can proceed
through an impulsive mechanism (see following section).
The detection of the CH2Cl- ion, formed by reaction 4,
coincides with the formation of HF and provides experimental
evidence of the displaced fluoride ion reacting with one of the
hydrogens of chloromethane. In the present experiment, only
extremely weak signals near the detection limit were observed
for the simpler, direct proton-transfer reaction Cl- + CH3F f
CH2F- + HCl, even though its endothermicity is only ∆rH0 )
309 ( 21 kJ/mol (Table 1). In our previous work7 on the F- +
CH3Cl reaction, we detected the ion CH2Cl- from the direct
proton-transfer reaction F- + CH3Cl f HF + CH2Cl-, but no
signal was detected for the CH2F- ion, which would be the
product channel analogous to reaction 4 in the reverse direction.
An explanation for this difference in behavior is that hydrogen-
bonding interactions for Cl- are less favorable than for F-.
While our calculations (Table 2 and Figure 4) show that the Cs
hydrogen-bonded F-‚‚‚HsCH2Cl complex is about 5 kJ/mol
more stable than the C3V F-‚‚‚CH3Cl ion-dipole complex, we
were unable to find a true local minimum for a Cl-‚‚‚HsCH2F
hydrogen-bonded complex by either the B3LYP/aug-cc-pVDZ
or CCSD(T)/aug-cc-pVDZ methods. Constraining the Cl-H-C
angle to 180° in Cl-‚‚‚HsCH2F in Cs symmetry gives a
complex that is 8 kJ/mol less stable than 3c.
The CH2X- or CH2Y- ion is observed as a major product
channel from the reaction X- + CH3Y (where X and Y are
both halogens) only when the neutral product (HY or HX) is
hydrogen fluoride. All such reactions are endothermic.26 The
beam/gas experiments by Zellermann and Vietzke8 detected the
products CH2Br- + HF and CH2I- + HF from the reactions
Br- + CH3F and I- + CH3F, respectively. The fluoride ion is
a special case, because of its high affinity for hydrogen-bonding
with carbon acids.60 HF elimination has been observed61 as a
facile product channel in collision-induced dissociation of
C6H4F- and C6H3F2-, with C6Hn‚‚‚F- complexes postulated as
intermediates. Figure 4 suggests the influence of the Cs
hydrogen-bonded complex 4b and the potential for the displaced
F- to react with one of the hydrogens of CH3Cl. Reaction 4,
therefore, could be the result of back-side attack via the
Fluorine Abstraction and Formation of FCl-. At apparent
collision energies above 4 eV, fluorine abstraction from CH3F
and the formation of the dihalide ion FCl- (reaction 5) proceed
as exhibited in Figures 1 and 2. The halide abstraction reaction
implies that reactive collisions attack in a front-side attack
mechanism at the fluorine (halophilic attack) of the CH3F
molecule. The threshold value for the reaction is 465 ( 24 kJ/
mol, exhibiting an excess of 146 ( 33 kJ/mol over the reaction
endothermicity (Table 1). The PES shown in Figure 5 was
calculated to investigate the front-side attack mechanism. No
minimum is located along the surface, which exhibits a repulsive
surface with a steep exit channel to the products FCl- and CH3.
The PES suggests that only a direct, impulsive front-side
halophilic attack mechanism will be successful.
The late rise in the FCl- cross section might be explained
by the impulsive rupture of the CH3-F bond. A rupture model
is consistent with the energy from a direct collision between
Cl- and CH3F partitioned into the vibrational mode of the
CH3-F bond. When the bond strength of CH3-F is exceeded,
it will dissociate into CH3 + F. From the initial collision the
products Cl- + F + CH3 are formed, which by the association
of the two halide species may be responsible for the detected
FCl-. The bond energy,20,21,23 D0(CH3-F) ) 452 ( 8 kJ/mol,
is in agreement within the error bars with the threshold energy,
E0 ) 465 ( 24 kJ/mol, for reaction 5. The excess 146 ( 33
kJ/mol energy above the endothermicity must be distributed into
kinetic or internal energy. The analogous mechanism was
postulated for a double feature exhibited by the FCl- cross
section in our previous work on the F- + CH3Cl f FCl-
+
CH3 reaction.7
Formation of CHCl-. Figures 1 and 2 show that at apparent
collision energies above 6 eV, the CHCl- ion is detected. The
threshold fit to the rising cross section results in E0 ) 653 (
26 kJ/mol, as shown in Table 1. Reactions 6 and 9 become
energetically possible above 588 ( 31 and 722 ( 31 kJ/mol,
respectively,20,21,23 with the CHCl- reaction cross section in
Figure 1 exhibiting possible contributions from both of these
reactions. The initial increase in the cross section of CHCl-
coincides with reaction 6, although the threshold measurement
(60) Roy, M.; McMahon, T. B. Can. J. Chem. 1985, 63, 708-715.
(61) Gronert, S.; DePuy, C. H. J. Am. Chem. Soc. 1989, 111, 9253-9254.
9
344 J. AM. CHEM. SOC. VOL. 124, NO. 2, 2002

Gas-Phase Reactions of Chloride Ion with Fluoromethane
A R T I C L E S
exhibits a possible barrier of 65 ( 40 kJ/mol. Reaction 6 may
be related to reaction 4, with the additional energy causing the
further dissociation of the CHCl--H bond. The increase in the
CHCl- cross section coincides with the leveling off of the
CH2Cl- cross section and the start of the steep decrease in the
SN2 cross section observed at collision energies above 6 eV.
The cross section behavior is evidence that reactions 3, 4, and
6 are all connected mechanistically via the back-side attack
mechanism, with passage over the SN2 transition state as the
common bottleneck. A further rise in the CHCl- cross section
at higher energies is consistent with contributions from reaction
9.
The HF elimination reaction is detected at collision energies
E0 ) 268 ( 95 kJ/mol and may result from a back-side attack
mechanism with the displaced F- ion reacting with a hydrogen
of CH3Cl via the Cs hydrogen-bonded exit intermediate 4b. At
energies between 3 and 4 eV, the methylene transfer reaction
is disrupted by dissociation of the CH2Cl- ion into CH2 + Cl-,
as observed by the dual rising feature. Significantly, the direct
proton-transfer reaction to form HCl + CH2F- is highly
inefficient, even when sufficient energy is available. The
inefficiency of proton transfer emphasizes that only the smallest
halogen, F-, strongly interacts with hydrogens of the halo-
methanes. The heavier halogens predominantly form ion-dipole
interactions with halomethanes, resulting in SN2 or halogen
abstraction at higher energy.
Fluorine abstraction, reaction 5, has a threshold energy of E0
) 465 ( 24 kJ/mol, exhibiting an excess to the reaction
endothermicity of 146 ( 33 kJ/mol for translational activation.
The late rise may be explained by a direct mechanism with
impulsive rupture of the CH3-F bond and association of F +
Cl-.
The experimental and theoretical results support the view that
the gas-phase SN2 reactions 2 and 3 occur predominantly
through a back-side attack mechanism with inversion of the
methyl group. The front-side attack mechanism is unfavorable
due to the repulsive nature of Cl- approaching the fluorine on
the C-F side, but it results in fluorine abstraction at high
energies. A hydrogen-bonded complex 4b is found to be the
true minimum on the exit path of the back-side attack potential
energy surface for reaction 3. The HF elimination reaction 4,
which occurs at collision energies above 2 eV, is experimental
evidence of the importance of hydrogen-bonding interactions
of the fluoride anion.
Other Dissociation Channels. Table 1 lists the endother-
micities of a large number of dissociative channels that could
occur at the observed collision energies. Many of these channels
involve charged products that we cannot detect (e- and H-) or
that are identical to observed lower energy channels. Although
some of these dissociative reactions are unlikely on mechanistic
grounds, a complete description of the high-energy reaction
dynamics should include all possibilities. As discussed above,
it is noteworthy that we observe insignificant direct proton
transfer, even though it is energetically allowed.
Conclusions
Four independent reactions (reactions 3-6) have been
detected by guided ion beam techniques at collision energies
of 0.05-27 eV. The SN2 threshold energy is measured at E0 )
181 ( 14 kJ/mol, exhibiting a 52 ( 16 kJ/mol excess barrier
to the reaction endothermicity by translational activation. The
high effective barrier may be explained by dynamical con-
straints, including angular momentum barriers, orientational
constraints, and inefficiency of promotion of the reaction by
translational energy versus vibrational energy. The double-well
potential for SN2 reactions has long been known to suppress
the efficiency of exothermic reactions. This work shows how
the influence of the double-well potential may be even greater
for an endoergic SN2 reaction, even when the central barrier is
well below the energy of products. At energies near the reaction
endothermicity for reaction 3, the shape of the pseudo-collinear
potential energy surface (Figure 6) tends to inhibit trajectories
with only translational energy from reaching products. Orien-
tational constraints tend to restrict the trajectories to this near-
collinear path.
Acknowledgment. We gratefully acknowledge the donors of
the Petroleum Research Fund, administered by the American
Chemical Society, for partial support of this research. This work
is also supported in part by the U.S. Department of Energy,
Office of Science, Office of Basic Energy Sciences, Chemical
Sciences, Geosciences and Biosciences Division. S.P.G. was
supported by the Research Experiences for Undergraduates
program of the National Science Foundation. We thank Veronica
M. Bierbaum, Vincent F. DeTuri, and William L. Hase for
helpful discussions.
JA012031F
9
J. AM. CHEM. SOC. VOL. 124, NO. 2, 2002 345