Polar effects on iodine atom abstraction by charged phenyl radicals

Article abstract of DOI:10.1021/jo981911i

The ability of differently substituted charged phenyl radicals (a class of distonic radical cations) to abstract an iodine atom from allyl iodide was systematically examined in the gas phase by using Fourier transform ion cyclotron resonance mass spectrometry. The reaction products and second-order reaction rate constants were determined for several radicals that differ by the type and/or number of substituents located in the ortho- and/or meta- position with respect to the radical site. All the radicals also carry a para-pyridinium group needed for mass spectrometric manipulation. These electron-deficient phenyl radicals react with allyl iodide by predominant iodine atom abstraction. The reaction is facilitated by the presence of neutral electron-withdrawing substituents, such as F, CF₃, Cl, or CN. The extent of rate increase depends on the type and number of the substituents, as well as their location relative to the radical site. Based on molecular orbital calculations (PM3 and Becke3LYP/6-31G(d)+ZPVE), the indicated variations in the transition state energy are not related to enthalpic factors. Instead, the results are rationalized by polar effects arising from a variable contribution of a stabilizing charge transfer resonance structure to the transition state. A semiquantitative measure for the barrier-lowering effect of each substituent is provided by its influence on the electron affinity of the radical (the electron affinities were calculated by Becke3LYP/6-31+G(d) and AM1, which were found to produce similar values). Methyl substitution does not significantly affect the electron affinity, and accordingly, does not have a detectable effect on reactivity. Methyl groups located at ortho-positions are an exception, however. o-Methyl-substituted phenyl radicals undergo exothermic rearrangement to a benzyl radical in competition with iodine abstraction from allyl iodide.

Full text of DOI:10.1021/jo981911i

J . Org. Chem. 2000, 65, 645-651
645
P ola r Effects on Iod in e Atom Abstr a ction by Ch a r ged P h en yl
Ra d ica ls
J enny L. Heidbrink, Kami K. Thoen, and Hilkka I. Kentta¨maa*
Department of Chemistry, Purdue University, West Lafayette, Indiana 47907-1393
Received September 21, 1998
The ability of differently substituted charged phenyl radicals (a class of distonic radical cations) to
abstract an iodine atom from allyl iodide was systematically examined in the gas phase by using
Fourier transform ion cyclotron resonance mass spectrometry. The reaction products and second-
order reaction rate constants were determined for several radicals that differ by the type and/or
number of substituents located in the ortho- and/or meta-position with respect to the radical site.
All the radicals also carry a para-pyridinium group needed for mass spectrometric manipulation.
These electron-deficient phenyl radicals react with allyl iodide by predominant iodine atom
abstraction. The reaction is facilitated by the presence of neutral electron-withdrawing substituents,
such as F, CF3, Cl, or CN. The extent of rate increase depends on the type and number of the
substituents, as well as their location relative to the radical site. Based on molecular orbital
calculations (PM3 and Becke3LYP/6-31G(d)+ZPVE), the indicated variations in the transition state
energy are not related to enthalpic factors. Instead, the results are rationalized by polar effects
arising from a variable contribution of a stabilizing charge transfer resonance structure to the
transition state. A semiquantitative measure for the barrier-lowering effect of each substituent is
provided by its influence on the electron affinity of the radical (the electron affinities were calculated
by Becke3LYP/ 6-31+G(d) and AM1, which were found to produce similar values). Methyl
substitution does not significantly affect the electron affinity, and accordingly, does not have a
detectable effect on reactivity. Methyl groups located at ortho-positions are an exception, however.
o-Methyl-substituted phenyl radicals undergo exothermic rearrangement to a benzyl radical in
competition with iodine abstraction from allyl iodide.
In tr od u ction
phenyl radical is converted into an electrophilic radical
by Cl, Br, and NO2 substituents, i.e., the substituted
radicals react faster than the phenyl radical with atom
donors with additional electron-donating substituents
(e.g., benzyl bromide) and slower with those carrying
electron-withdrawing substituents (e.g., CBrCl3). On the
other hand, the apparent “polarity” of a phenyl radical
can be strongly dependent on the substrate, as demon-
strated by the finding that the p-Cl-, p-NO2-, and p-CH3-
substituted phenyl radicals behave as nucleophilic radi-
cals toward the 4-methylpyridinium ion but as electro-
philic radicals toward 4-methylpyridine.5h A more in-
depth understanding of these substituent effects is
desirable, considering the importance of substituted
phenyl radicals in reactions ranging from organic syn-
thesis to the action of antitumor and antiviral drugs.1
Gas-phase ion chemistry has proven to be a powerful
tool for the study of many ionic reaction intermediates
that are difficult to study in solution. This approach can
be extended to reactive neutral species via ions that
The reactions of the phenyl radical have been exten-
sively investigated in solution and in the gas phase.1-4
However, the knowledge on substituted phenyl radicals
appears to be limited to reactions of the p-Cl-, p-Br-,
p-NO2-, p-CH3-, and/or p-OCH3-substituted radicals with
bromine (benzyl bromides, CH2Br2, CBrCl3) and chlorine
atom donors (CCl4), various hydrogen atom donors,
disulfides, and substituted propenes, as well as 4-meth-
ylpyridine and its protonated form.5 These studies have
led to the general conclusion that the relatively nonpolar
(1) See for example: (a) Free Radicals; Kochi, J . K., Ed.; J ohn Wiley
& Sons: New York, 1973; Vol. 1. (b) Free Radicals in Organic
Chemistry; Fossey, J ., Lefort, D., Sorba, J ohn Wiley & Sons: New York,
1995. (c) Theoretical Aspects of Physical Organic Chemistry; Shaik, S.
S., Schlegel, H. B., Wolfe, S., Eds.; J ohn Wiley & Sons: New York,
1992. (d) Theoretical and Physical Principles of Organic Reactivity;
Pross, A., Ed.; J ohn Wiley & Sons: New York: 1995.
(2) See for example: (a) Bridger, R. F.; Russell, G. A. J . Am. Chem.
Soc. 1963, 85, 3754-3764. (b) Bridger, R. F.; Russell, G. S. J . Am.
Chem. Soc. 1963, 85, 3765-3766. (c) Pryor, W. A.; Guard, H. J . Am.
Chem. Soc. 1964, 86, 1150-1152. (d) Fu, J .-J . L.; Bentrude, W. G. J .
Am. Chem. Soc. 1972, 94, 7710-7722. (e) Levy, A.; Meyerstein, D.;
Ottolenghi, M. J . Phys. Chem. 1973, 77, 3044-3047. (f) Kryger, R. G.;
Lorand, J . P.; Stevens, N. R.; Herron, N. R. J . Am. Chem. Soc. 1977,
99, 7589-7600. (g) Scaiano, J . C.; Stewart, L. C., J . Am. Chem. Soc.
1983, 105, 3609-3614. (h) Yu, T.; Lin, M. C.; Melius, C. F. Int. J . Chem.
Kinet. 1994, 26, 1095-1104. (I) Radziszewski, J . G.; Nimlos, M. R.;
Winter, P. R.; Ellison, B. J . Am. Chem. Soc. 1996, 118, 7400-7401.
(3) Matimba, H. E. K.; Crabbendam, A. M.; Ingeman, S.; Nibbering,
N. M. M., J . Chem. Soc., Chem. Commun. 1991, 644-645.
(5) For substituted phenyl radicals, see: (a) Pryor, W. A.; Echols,
J . T., J r.; Smith, K. J . J . Am. Chem. Soc. 1966, 88, 1189-1199. (b)
Bunnett, J . F.; Wamser, C. C. J . Am. Chem. Soc. 1966, 88, 5534-5537.
(c) Pryor, W. A.; Smith, K. J . Am. Chem. Soc. 1970, 92, 2731-2738.
(d) Takayama, K.; Kosugi, M.; Migita, T. Chem. Lett. 1973, 193-195.
(e) Takayama, K.; Kosugi, M.; Migita, T. Chem. Lett. 1973, 215-218.
(f) Migita, T.; Nagai, T.; Abe, Y. Chem. Lett. 1975, 543-546. (g) Migita,
T.; Takayama, K.; Abe, Y.; Kosugi, M. J . Chem. Soc., Perkin Trans. 2
1979, 1137-1142. (h) Arnaud, R.; Bonnier, J . M.; Fossey, J . Nouv. J .
Chimie 1980, 4, 299-305. (i) Tilset, M.; Parker, V. Acta Chem. Scand.
B 1982, 36, 123-124. (J ) da Silva Correa, C. M. M.; Oliveira, M. A. B.
C. S.; de Lurdes S. Almeida, M.; Rodrigues, P. C. M.; Fonseca, M. M.
B.; Estrada, J . M. A. C. J . Chem. Soc., Perkin. Trans 2 1989, 1643-
1646. (k) Hazlewood, C.; Davies, M. J .; Gilbert, B. C.; Packer, J . E. J .
Chem. Soc., Perkin Trans. 2 1995, 2167-2174.
(4) For gas-phase studies, see for example: (a) Fahr, A.; Stein, S.
E. J . Phys. Chem. 1988, 92, 4951-4955. (b) Chen, R. H.; Kafafi, A.;
Stein, S. E. J . Am. Chem. Soc. 1989, 111, 1418-1423. (c) Kopinke,
F.-D.; Zimmerman, G.; Anders, K. J . Org. Chem. 1989, 54, 3571-3576.
(d) Yu, T.; Lin, M. C. 1994, 98, 9697-9699. (e) Yu, T.; Lin, M. C. J .
Phys. Chem. 1995, 99, 8599-8603.
10.1021/jo981911i CCC: $19.00 © 2000 American Chemical Society
Published on Web 01/20/2000

646 J . Org. Chem., Vol. 65, No. 3, 2000
Heidbrink et al.
same cell (nominal pressure of about 2 × 10-7 Torr) through
a batch inlet system containing an Andonian leak valve. The
mixture was subjected to electron ionization (typically 20 eV
electron energy, 8 µA emission current, 50 ms ionization time),
which resulted in an abundant signal for the halobenzene
radical cation. The radical cation was allowed to react with
pyridine or 3-fluoropyridine (for 11, ionized methanol and its
fragment ions were allowed to protonate 3-iodopyridine) to
produce the desired halogen displacement product (typical
reaction times were 500 ms-10 s).
The ions formed upon the halogen displacement reaction
were transferred into the other side of the dual cell by
grounding the conductance limit plate for approximately 150
µs and cooled for one second by allowing irradiative emission
and collisions with the neutral molecules present in this cell
(the reagent to be used in the final stage of the experiment).
The ions were isolated by ejecting all unwanted ions from the
cell through the application of a stored-waveform inverse
Fourier transform13 (SWIFT) excitation pulses to the excitation
plates of the cell (Extrel SWIFT module). Cleavage of the
remaining carbon-halogen or carbon-nitrogen bond was
accomplished by collisional activation with argon target (pulsed
into the cell at a peak nominal pressure of 6 × 10-6 to 1 ×
10-5 Torr). The sustained off-resonance irradiation14 (SORI)
technique was used to kinetically excite the ions (about one
second irradiation at a frequency 0.5-1.0 kHz higher than the
cyclotron frequency of the ions). The product ions (charged
phenyl radicals) were kinetically and internally cooled by
giving them time (g0.4 s) for radiative emission and for
collisions with the neutral molecules present in this cell. This
approach was found to yield the same results (reaction
products and rate constants) as obtained by using a high-
pressure burst of argon (pulsed into the cell at a peak nominal
pressure of about 10-5 Torr), which indicates sufficient cooling
of the ions.
contain the reactive group of interest, as well as a remote,
chemically inert charged group for mass spectrometric
manipulation. Various intermediates have been studied
this way, including the o- and m-benzynes,6 phenylcar-
benes,7 and phenylnitrene.7b We have applied this ap-
proach to phenyl radicals via distonic radical cations8,9
(ionized ylides, biradicals, and zwitterions) with a phenyl
radical site and a chemically inert, positively charged
substituent in the meta- or para-position.10,11 The reactiv-
ity of these species was found to parallel that of the
phenyl radical, i.e., the same reaction products and
similar reactivity trends were observed with different
substrates. The N-(4-dehydrophenyl)pyridinium ion, for
example, undergoes10b radical reactions characteristic of
the phenyl radical.
Inspired by the above results, we decided to examine
whether the N-(4-dehydrophenyl)pyridinium ion can be
employed to systematically study substituent effects on
phenyl radicals’ reactions. Iodine abstraction from allyl
iodide was selected as the first reaction for study since
the phenyl radical readily abstracts iodine atoms from
various aromatic and aliphatic substrates.12 Further, the
rate of iodine abstraction by the phenyl radical from
iodobenzenes has been found to depend on the substitu-
tion of the substrate, which indicates sensitivity to polar
effects.12 Hence, the iodine atom abstraction ability of the
charged phenyl radical may be sensitive to its substitu-
ents, and thus provide a means to improve our funda-
mental understanding of atom abstraction reactions.
Exp er im en ta l Section
The charged phenyl radicals were isolated by ejecting all
other ions from the cell, as described above, and allowed to
undergo reactions with allyl iodide (nominal pressure ap-
proximately 1.1 × 10-7 Torr; introduced by using a batch inlet
containing an Andonian leak valve) for a variable period of
time (typically 1-20 s). All ions were excited for detection by
using a “chirp” excitation sweep of 2.65 MHz bandwidth and
3200 Hz/µs sweep rate. All the measured spectra are the
average of at least 30 transients and were recorded as 64 K
data points subjected to one zero fill prior to Fourier trans-
formation.
All experiments were performed using a dual-cell Extrel
Model 2001 Fourier transform ion cyclotron resonance mass
spectrometer (FT/ICR) described previously.6,11 Allyl iodide and
halogenated benzenes with various additional substituents
(1,4-diiodobenzene, 1-chloro-3-(trifluoromethyl)-4-iodobenzene,
1-chloro-2-(trifluoromethyl)-4-iodobenzene, 5-bromo-2-iodo-
toluene, 2-bromo-5-iodotoluene, 1-bromo-3-chloro-4-iodoben-
zene, 5-bromo-2-iodo-m-xylene, and 1-chloro-2-cyano-4-ni-
trobenzene), as well as 3-iodopyridine, were obtained commer-
cially and used as received. Electron ionization mass spec-
trometry did not reveal impurities in these materials. The
halogenated neutral precursors were introduced at a nominal
pressure of (1.0-1.5) × 10-7 Torr into one side of the dual cell
by using a heated solids probe or a variable leak valve. The
appropriate nucleophile (pyridine or 3-fluoropyridine for 1-10)
or protonating agent (methanol for 11) was added into the
In the experiments described above, the neutral reagent is
present at a great excess relative to ions. This leads to pseudo-
first-order kinetics. The second-order rate constant (kreaction
)
of each ion-molecule reaction was obtained from a semiloga-
rithmic plot of the relative abundance of the reactant ion
versus time. The collision rate constants (kcoll) were calculated
using a parameterized trajectory theory.15 The reaction ef-
ficiencies are given by (kreaction/kcoll) × 100%. The accuracy of
the rate constant measurements is estimated to be (50%,
while the precision is usually better than (10%. The pressure
readings were corrected for the sensitivity of the ion gauge
toward allyl iodide16 and for the pressure gradient between
the dual cell and the ion gauge. Both correction factors were
obtained by measuring rates for electron-transfer reactions of
allyl iodide that can be expected to occur at collision rate due
to their high exothermicity.17 Internal consistency between
correction factors obtained by examining different radical
(6) Thoen, K. K.; Kentta¨maa, H. I. J . Am. Soc. 1999, 121, 800-805.
(7) (a) Hu, J .; Hill, B. T.; Squires, R. R. J . Am. Chem. Soc. 1997,
119, 11699-11700 and references therein. (b) Wenthold, P. G.; Hu, J .;
Hill, B. T.; Squires, R. R. Int. J . Mass Spectrom. Ion Processes 1998,
179/ 180, 173-183 and references therein.
(8) (a) Bouma, W. J .; Nobes, R. H.; Radom, L. J . Am. Chem. Soc.
1982, 104, 2929-2931. (b) Bouma, W. J .; McLeod, J . K.; Radom, L. J .
Am. Chem. Soc. 1982, 104, 2930-2931. (c) Holmes, J . L.; Lossing, F.
P.; Terlouw, J . K.; Burgers, P. C. J . Am. Chem. Soc. 1982, 104, 2931-
2932.
(9) (a) Hammerum, S. Mass Spectrom. Rev. 1988, 7, 123-202. (b)
Stirk, K. M.; Kiminkinen, M.; Kentta¨maa, H. I. Chem. Rev. 1992, 92,
1649-1665.
(10) (a) Smith, R. L.; Kentta¨maa, H. I. J . Am. Chem. Soc. 1995, 117,
1393-1396. (b) Thoen, K. K.; Smith, R. L.; Nousiainen, J . J .; Nelson,
E. D.; Kentta¨maa, H. I. J . Am. Chem. Soc. 1996, 118, 8669-8676. (c)
Leeck, D. T.; Stirk, K. M.; Zeller, L. C.; Kiminkinen, L. K. M.; Castro,
L. M.; Vainiotalo, P.; Kentta¨maa, H. I. J . Am. Chem. Soc. 1994, 116,
3028-3038.
(13) Chen, L.; Wang, T.-C. L.; Ricca, T. L.; Marshall, A. G. Anal.
Chem. 1987, 59, 449-454.
(14) Gauthier, J . W.; Trautman, T. R.; J acobson, D. B. Anal. Chim.
Acta 1991, 246, 211-225.
(15) Su, T.; Chesnavich, W. J . J . Chem. Phys. 1982, 76, 5183-5185.
(16) (a) Bartmess, J . E.; Georgiadis, R. M. Vacuum 1983, 33, 149-
153. (b) Miller, K. J .; Savchick, J . A. J . Am. Chem. Soc. 1979, 101,
7206-7213.
(11) Li, R.; Smith, R. L.; Kentta¨maa, H. I. J . Am. Chem. Soc. 1996,
118, 5056-5061.
(12) (a) Danen, W. C.; Saunders J . Am. Chem. Soc. 1969, 91, 5924-
5925. (b) Danen, W. C.; Tipton, T. J .; Saunders: D. G. J . Am. Chem.
Soc. 1971, 93, 5186-5189. (c) Danen, W. C.; Winter, R. L. J . Am. Chem.
Soc. 1971, 93, 716-720.
(17) (a) Han, C.-C.; Wilbur, J . L.; Brauman, J . I. J . Am. Chem. Soc.
1992, 114, 887-893 and references therein. (b) Richardson, D. E. J .
Phys. Chem. 1986, 90, 3697-3700. (d) Kebarle, P.; Chowdhury, S.
Chem. Rev. 1987, 87, 513.

Polar Effects on Iodine Atom Abstraction
J . Org. Chem., Vol. 65, No. 3, 2000 647
Ta ble 1. Efficien cy (kr ea ction /kcoll) of Rea ction of Allyl Iod id e w ith Differ en tly Su bstitu ted P h en yl Ra d ica ls
a
The reaction displayed nonpseudo-first-order kinetics. Efficiencies were obtained by fitting the data with a nonlinear curve-fitting
program that models the decay of the ion on the basis that the ion is rearranging within the collision complex.
Sch em e 1
cation reagents supports the validity of the measurements
(e.g., with CS2+• and the acetone radical cation). Each reaction
spectrum was background corrected by using a previously
described procedure.10,11
The Spartan and the Gaussian 92 Revision F suites of
programs18 were used to carry out the molecular orbital
calculations. Either a semiempirical method (usually AM1;
PM3 was used for calculations involving iodine) or the density
functional theory (Becke3LYP/6-31G(d)) was used to fully
optimize the geometries of all radicals. The optimized struc-
tures were verified to correspond to an energy minimum by
calculating the harmonic vibrational frequencies (no imaginary
frequencies). The relative energies obtained from the density
functional calculations were corrected by adding the zero point
vibrational energy calculated at the same level of theory. The
electron affinities calculated for the phenyl radicals are vertical
values, i.e., the relative energies of the reduced radicals
(carbanions or zwitterions) were calculated as single points
by using the optimized geometries of the corresponding
radicals (obtained either by AM1 or Becke3LYP6-31G+(d)).
tion10,11,19 of an iodine, chlorine, or bromine atom with
Resu lts a n d Discu ssion
pyridine or 3-fluoropyridine yields the desired substituted
iodo- or nitrobenzene ion. Collision-activated dissociation
induced by sustained off-resonance irradiation14 forms
the desired charged phenyl radical by cleavage of an
iodine atom or a nitro group (see Scheme 1). The radical
11 was formed by protonating 3-iodopyridine and then
collisionally cleaving the iodine atom. The charged
radicals were then isolated for reactivity studies. The
purpose of using the fluorinated nucleophile was to
enable distinction between the brominated radical cation
precursor and the substitution product (bromine and
Syn th esis of th e Ch a r ged Ra d ica ls. Various deriva-
tives of the N-(4-dehydrophenyl)pyridinium ion (Table 1)
were formed using a recently developed6,11 multistep
procedure in a dual-cell Fourier transform ion cyclotron
resonance mass spectrometer. The first step involves the
generation of a substituted haloiodo- or halonitrobenzene
radical cation by electron ionization. Ipso-substitu-
(18) Frisch, M. J .; Trucks, G. W.; Head-Gordon, M.; Gil, P. M. W.;
Wong, M. W.; Foresman, J . B.; J ohnson, B. G.; Schlegel, H. B.; Robb,
M. A.; Replogle, E. S.; Gomperts, R.; Andres, J . L.; Raghavachari, K.;
Binkley, J . S.; Gonzalez, C.; Martin, R. L.; Fox, D. J .; Defrees, D. J .;
Baker, J .; Stewart, J . J . P.; Pople, J . A. Gaussian, Inc., Pittsburgh,
PA 1992.
(19) (a) Tho¨lmann, D.; Gru¨tzmacher, H.-F. J . Am. Chem. Soc. 1991,
113, 3281-3287. (b) Tho¨lmann, D.; Gru¨tzmacher, H.-F. Org. Mass
Spectrom. 1989, 24, 439-441.

648 J . Org. Chem., Vol. 65, No. 3, 2000
Heidbrink et al.
pseudo-first-order kinetics (Figure 1b). This observation
is readily understood if the radicals, within the collision
complex, undergo partial rearrangement to more stable
benzyl radicals in direct competition with the iodine atom
abstraction. An incomplete rearrangement (resulting in
a mixture of isomeric radicals) would lead to deviation
from pseudo-first-order kinetics if the rearranged π-radi-
cal (Scheme 2) reacts with allyl iodide at a different rate
than the unrearranged σ-radical. The authentic charged
benzyl radical 13 (generated from the radical cation of
3-chlorobenzyl bromide by the same method as the
phenyl radicals) was found to be unreactive toward allyl
iodide, in support of the proposed rearrangement. The
observation of deviation from pseudo-first-order kinetics
for the o-CH3- substituted radicals but not for their meta-
isomers suggests that the isomerization occurs without
active participation by allyl iodide, e.g., it does not involve
reversible hydrogen atom transfer involving allyl iodide.
However, the solvation energy provided by allyl iodide
appears to be necessary to drive the reaction, since
complete isomerization was never observed, and the
population of reactive and unreactive isomers was con-
stant and reproducible from one experiment to another.
Gen er a l Con sid er a tion s on Ra d ica l Rea ction s of
Ga seou s Diston ic Ra d ica l Ca tion s. Only exothermic
reactions are observed under the conditions employed in
this study. Reaction efficiencies (fraction of collisions that
leads to reaction) of gas-phase ion-molecule reactions
are generally governed20a by the difference in energy
between the separated reactants and the transition state
(∆E in Figure 2a). Variations in reaction efficiencies due
to substituents can only be observed with this experi-
mental approach provided that this energy gap is small.
If the central barrier (transition state energy) is too low,
i.e., 5-10 kcal/mol below the energy of the separated
reactants, the efficiency will approach unity (every col-
lision leads to reaction).20a On the other hand, if the
central barrier is too high, the efficiency will become too
low to measure (a few kcal/mol above the energy of the
separated reactants is too high for the systems studied
here).
F igu r e 1. Temporal variation of the reactant and product ion
abundances for the reaction of allyl iodide with (a) the o-CF3-
substituted charged phenyl radical 5 (m/z 241; iodine abstrac-
tion: m/z 368; allyl abstraction: m/z 282), and (b) with the
o-CH3-substituted charged phenyl radical 2 (m/z 187; iodine
abstraction: m/z 314). The data points are connected with
arbitrary smooth lines except for b (see Table 1). For part a,
the decay of the reactant ion is presented with a linear least-
squares fit.
Sch em e 2
pyridine have the same MW, 79 amu). The fluorine
substituent does not affect the iodine atom abstraction
efficiency of these charged phenyl radicals, as indicated
by the same reaction efficiency (10%) measured for both
the N-(4-dehydrophenyl)pyridinium ion and the N-(4-
dehydrophenyl)-3-fluoropyridinium ion (1).
The time dependences of the relative abundances of
the reactant and product ions show that the reactions of
the N-(4-dehydrophenyl)-3-fluoropyridinium ion and most
of its derivatives obey pseudo-first-order kinetics, as
expected (excluding 2 and 4, Table 1; Figure 1a). This
finding, combined with prior evidence10 that the isomeric
ortho-, meta-, and para-substituted charged phenyl radi-
cals do not interconvert, suggests that the radical popu-
lations are likely to be isomerically pure.
Atom abstraction reactions of radicals are often as-
sociated with energy barriers that prevent them from
proceeding at collision rate.21 For these reactions to be
observable in high vacuum, they usually must be highly
exothermic.11,22 Iodine abstraction from allyl iodide by
phenyl radical is such a reaction (the enthalpy change23
is -23 kcal/mol). Hence, allyl iodide was selected as the
substrate for the present study. Since the same neutral
(20) (a) Barfknecht, A. T.; Dodd, J . A.; Salomon, K. E.; Tumas, W.;
Brauman, J . I. Pure Appl. Chem. 1984, 56, 1809-1818. (b) Gas-Phase
Ion Chemistry; Bowers, M. T., Ed.; Academic Press: London, 1979;
Vol. I.
(21) See, for example: (a) Tedder, J . M. Tetrahedron 1982, 38, 313-
329. (b) Sidebottom, H.; Treacy, J . Int. J . Chem. Kinet. 1984, 16, 579-
590. (c) Shonkhai, S.; Whittle, R. Int. J . Chem. Kinet. 1984, 16, 543-
558. (d) Fox, R. J .; Evans, F. W.; Szwarc, M. Trans. Faraday Soc. 1961,
57, 1915-1927. (e) Alcock, W. G.; Whittle, E. Trans. Faraday Soc. 1965,
61, 244-254.
However, reactions of the radicals 2 and 4, each with
methyl groups adjacent to the radical site, do not follow

Polar Effects on Iodine Atom Abstraction
J . Org. Chem., Vol. 65, No. 3, 2000 649
Ta ble 2. Com p u ta tion a l Exoth er m icities (∆Hr ea ction
(k ca l/m ol)) for th e Abstr a ction of a Ha logen Atom by
Mod el Ch a r ged P h en yl Ra d ica ls fr om Allyl Iod id e (P M3)
a n d Allyl Ch lor id e (DF T)
a
b
PM3 calculations. Becke3LYP/6-31G(d) + ZPVE. c Not cal-
culated.
(analogous C-H bond: ∼75 kcal/mol, PM3). Analogous
reactivity trends have been reported21b,d,e for simple
carbon-centered radicals toward abstraction of different
•
atoms (for CH3• and CF3 , the trend is Cl < H ∼ Br < I).
The N-(4-dehydrophenyl)pyridinium ion (1) abstracts
iodine atom from allyl iodide at a reaction efficiency of
10% (10 collisions from 100 lead to iodine abstraction;
Ta ble 1); this value is used as the reference in the
following discussion. The radicals 5-10, with either an
o-Cl, o-F, or o-CF3 substituent, a m-CF3 or m-CN sub-
stituent, or two o- and two m-F substituents, undergo
iodine abstraction at a greater efficiency than 1 (16-32%
vs 10%; Table 1). The m-CH3- substituted radical 3 reacts
at about the same efficiency as 1 (11%). The two o-CH3-
substituted radicals 2 and 4 react at a slightly lower
efficiency.
F igu r e 2. (a) A typical potential energy surface for an
exothermic gas-phase ion-molecule reaction. The reaction rate
is controlled by the energy difference between the isolated
reactants and the transition state. (b) A schematic potential
energy surface illustrating the lowering of the transition state
energy due to increasing reaction exothermicity (c) and due
to polar effects.
reagent was employed in all experiments and the only
variable was substitution of the charged radical, trends
in the measured reaction efficiencies can be expected to
reflect changes in the barrier height (Figures 2b and 2c)
rather than the stability20b of the initial gas-phase
collision complex.
Rea ction Efficien cies. All the phenyl radicals stud-
ied react with allyl iodide by predominant iodine atom
abstraction, presumably due to the high exothermicity
of the reaction (the homolytic C-I bond dissociation
energy of this substrate is only 44 kcal/mol).23 An
addition/elimination pathway leading to allyl abstraction
is slow (e 15%) or not detectable for all but two cases
(Table 1). Although hydrogen atom abstraction from allyl
iodide is slightly more exothermic (∼-28 kcal/mol) than
iodine abstraction (see discussion below; Table 2), hy-
drogen abstraction was not observed. This finding is
likely explained by the greater polarizability and the
lower homolytic bond dissociation energy of the C-I bond
Rea ction Exoth er m icities. The observation ofchanges
in the reaction efficiency upon introduction of additional
substituents into 1 suggests that the iodine atom ab-
straction barrier must be close to the energy level of the
isolated reactants and that electron-withdrawing sub-
stituents stabilize the transition state. Atom transfer
barriers are controlled by avoided curve crossings.1 The
barrier is usually expected to be found at or below the
crossing of two potential energy surfaces that correspond
to a breaking and a forming bond. Therefore, the factor
considered to be most important in determining the
barrier height is the strength of the breaking and the
forming bond (the Bell-Evans-Polanyi principle) (Figure
2b).1,21a A similar conclusion is reached by using the
widely used valence bond model,1c,d which is based on the
avoided crossing of the bonding and antibonding states
of the reactants and products, with a perturbation by
ionic excited states. Based on these theories, the most
likely explanation for the observed variations in reaction
efficiency is a varying strength for the newly formed
(22) Smith, R. L.; Chyall, L. J .; Stirk, K. M.; Kentta¨maa, H. I. Org.
Mass Spectrom. 1993, 28, 1623-1631.
(23) NIST Standard Reference Data Base Number 69; Mallard, W.
G., Lindstrom, P. J ., Eds.; August 1997, National Institute of Standards
and Technology: Gaithersburg MD, 20899 (http://webbook.nist.gov).

650 J . Org. Chem., Vol. 65, No. 3, 2000
Heidbrink et al.
Ta ble 3. Ver tica l Electr on Affin ities (eV) Ca lcu la ted for
Mod el Ch a r ged P h en yl Ra d ica ls
carbon-iodine bond (all reactions involve cleavage of the
same bond).
Molecular orbital calculations were carried out to
estimate the effects of substitution on the reaction
enthalpy. The calculations were performed on simple
model systems, the 4-dehydroanilinium ion analogues
(with an ammonium instead of a pyridinium charge site).
Analogous results were obtained for iodine and chlorine
abstraction reactions (Table 2). Abstraction of a chlorine
atom from allyl chloride by radicals with an additional
o-Cl, o-F, or o-CF3 or a m-CN, m-CH3, or m-CF3 substitu-
ent was calculated to be highly exothermic, and the
exothermicity was found to be nearly independent of
substitution (-25 to -27 kcal/mol; Becke3LYP/6-31G(d)
+ ZPVE). Similarly, the enthalpy change of iodine atom
abstraction from allyl iodide was calculated to be insensi-
tive to substitution. The reaction of the 4-dehydro-
anilinium ion and its o-F, o-Cl and m-CN, m-CH3, and
m-CF3 derivatives was estimated to be associated with
a large negative enthalphy change (-22 to -24 kcal/mol;
PM3) of the same magnitude as for the reaction of the
phenyl radical with allyl iodide (experimental value is
-23 kcal/mol;23 -22 kcal/mol was calculated by PM3).
Varying the substitution on the charged phenyl radical
does not appear to lead to such changes in reaction
enthalpy that would correlate with the differing reaction
efficiencies. This is perhaps not surprising, considering
the great exothermicity of the reaction in question.1b,d,21a
P ola r ity of th e Tr a n sition Sta te. The fact that the
observed substituent effects do not arise from variations
in reaction enthalpy suggests that polar effects1a,b,d,21a,24,25
play a critical role in controlling the reaction efficiency
of the charged radicals, i.e., the neutral substituents
lower the energy of the transition state by enhancing its
polar character (Figure 2c). Due to the polarizability of
halogen atoms, halogen atom transfer reactions have
been predicted to be more sensitive to polar effects than,
for example, hydrogen atom abstraction.21b,25 Examina-
tion of iodine abstraction by the nucleophilic methyl
radical from CF3I and CH3CH2I has demonstrated the
importance of polar effects in these iodine transfer
reactions: despite the same exothermicity, the reaction
with CF3I is associated with a 3 kcal/mol lower barrier
(4.3 kcal/mol) than the reaction with CH3CH2I.21b
According to Donahue et al.,26 the barrier formation
in many gas-phase radical abstraction reactions is best
described as an avoided crossing of the ground state and
an ionic excited state of the reactants and products. This
is even more likely to be the case for the charged radicals
studied here. The transition states of their reactions are
polarized by an ionic configuration that corresponds to
formal transfer of an electron from the neutral substrate
to the radical. Hence, the observed variations in barrier
heights are expected to arise primarily from differences
in the radicals’ electron affinities.26
a
Not calculated.
significantly greater than that of the phenyl radical
(experimental EA ) 1.1 eV;23 we calculate the same value
by Becke3LYP/6-31+G(d) + ZPVE). This finding suggests
that polar effects may play a significantly greater role
in controlling the reaction rates of the charged phenyl
radicals than those of the phenyl radical. An even greater
adiabatic EA of 6.53 eV is calculated for the 3-dehydro-
pyridinium ion (11) with the charge distributed over the
π-system of the same ring that contains the σ-radical site.
The high EA is likely to be the origin of the great iodine
abstraction efficiency measured for this radical upon
interaction with allyl iodide (59%).
Variation of EA with substitution of the 4-dehydro-
anilinium ion was examined by calculating vertical rather
than adiabatic electron affinities because this is not only
more straightforward but is also in better agreement with
the curve crossing model.1c,d,26 The vertical electron
affinities of the radicals studied follow the same trend
as the adiabatic ones, but their absolute value is smaller
by a constant amount (0.4 eV), e.g., the vertical EA’s
obtained for the phenyl radical, the 4-dehydroanilinium
ion and the 3-dehydropyridinium ion are 0.61, 4.67, and
6.12 eV, respectively (Becke3LYP/6-31+G(d)). Vertical
EA values were estimated for a series of substituted
4-dehydroanilinium ions by using the semiempirical AM1
method as well as Becke3LYP/6-31+G(d) (Table 3). These
two approaches give similar results that in most cases
agree within 0.1 eV (e.g., 4-dehydroanilinium ion: 4.66
and 4.67 eV; 3-dehydropyridinium: 6.21 and 6.12 eV).
The electron affinities (EA) of several charged phenyl
radicals were estimated computationally. The adiabatic
EA value of the 4-dehydroanilinium ion is calculated to
be 5.10 eV (Becke3LYP/6-31+G(d) + ZPVE) and hence
(24) Pross, A.; Yamataka, H.; Nagase, S. J . Phys. Org. Chem. 1991,
4, 135-140.
(25) Nieto, J . D.; Herrera, O. S.; Lane, S. I., Oexler, E. V. Ber.
Bunsen-Ges. Phys. Chem. 1998, 102, 821-825.
(26) (a) Donahue, N. M.; Clarke, J . S.; Anderson, J . G. J . Phys. Chem.
A, 1998, 102, 3923-3933. (b) Clarke, J . S.; Krill, J . H.; Donahue, N.
M.; Anderson, J . G. J . Phys. Chem. A 1998, 102, 9847-9857.

Polar Effects on Iodine Atom Abstraction
J . Org. Chem., Vol. 65, No. 3, 2000 651
+
The effect of replacing the pyridinium with the NH3
-
anilinium ions, respectively. These values are similar to
that of the unsubstituted system (4.66 eV). Hence, based
on the ionic curve crossing model, methyl substitution is
not expected to greatly affect the reactivity of the radical.
In agreement with this prediction, the radical with a
m-CH3 substituent (3) was found to react at the same
efficiency (within experimental precision) as the unsub-
stituted radical 1 (11% and 10%, respectively).
group in the model systems was probed by calculating
the vertical EA for one of the experimentally studied
radicals, radical 1. The value obtained (4.47 eV; AM1) is
0.2 eV smaller than that calculated for its model system,
the 4-dehydroanilinium ion (EA ) 4.66 eV). Hence, the
NH3+-group appears to increase the EA of the phenyl
radical more than a pyridinium charge site. However,
both groups cause about a 5-fold increase in the EA of
the phenyl radical and polarize the transition state for
iodine abstraction in the same way. Therefore, the model
system is expected to yield useful information on the
observed reactivity trends.
In general, the computational results show an increase
in the EA of the charged radical upon addition of electron-
withdrawing substituents, i.e., at least qualitatively, EA
appears to correlate with the iodine abstraction efficiency.
For example, the reactivity trend o-CF3 > o-Cl > o-H
(efficiencies are 24%, 16%, and 10%, respectively) follows
the trend obtained for the EA values of the analogous
model systems (5.22, 5.14, and 4.67 eV, respectively;
Becke3LYP/6-31+G(d)). Note that the increase in reac-
tion efficiency caused by addition of an o-Cl or o-CF3
substituent cannot be rationalized by enthalpic control
since this substitution is associated with a decrease in
reaction exothermicity.
To test the importance of EA in controlling the iodine
abstraction barrier height, the radical 12 was included
in this study. In this phenyl radical, the charged pyri-
dinium substituent is separated by a phenyl ring from
the radical-bearing phenyl. The EA of 12 is calculated to
be drastically smaller (3.07 eV at AM1) than that of any
of the other charged radicals studied. However, the
estimated exothermicity of iodine abstraction is compa-
rable to those calculated, for example, for the o-Cl- and
o-CF3-substituted 4-dehydroanilinium radicals (PM3:
-22.1 vs -21.8 and -18.4 kcal/mol, respectively). There-
fore, consideration of enthalphic factors alone suggests
similar reactivity for radical 12 and the other two
radicals. However, the ionic curve crossing model26
predicts radical 12 to be less reactive than the others.
This is what was observed, i.e., the radical 12 abstracts
iodine from allyl iodide at a fraction of the efficiency
measured for the o-Cl- and o-CF3- substituted radicals 5
and 7 (2% vs 24 and 17%, respectively).
Relative EA values can also be used to rationalize the
trends observed in reaction efficiencies upon variation
of the location and the number of the neutral substitu-
ents. For example, the observation of a greater iodine
atom abstraction efficiency for the o-CF3-substituted
radical 5 than for the isomeric meta-substituted radical
6 (24% and 17%, respectively) is readily explained by the
greater electron affinity of the ortho-substituted radical
(Becke3LYP/6-31+G(d) gives EA ) 5.22 eV for the o-CF3-
substituted 4-dehydroanilinium ion and 5.01 eV for the
meta-substituted isomer). An increase in the number of
the substituents also increases EA, as indicated by the
greater electron affinity (5.94 eV; AM1) calculated for the
2,3,5,6-tetrafluoro-4-dehydroanilinium ion than that of
the o-F analogue (5.03 eV; AM1). Indeed, the radical 9
with four fluorine substituents reacts at a significantly
greater efficiency (32% iodine abstraction efficiency) than
the radical 8 with only one o-fluorine substituent (25%
iodine abstraction efficiency).
However, it should be noted here that although CH3
groups adjacent to the radical site (2 and 4) do not
influence the polarity of the iodine abstraction transition
state, they affect the reactivity of the phenyl radical in
other ways. For example, steric hindrance is likely to
partially explain the lower reaction efficiencies of these
radicals (7% and 6%, respectively) as compared to 1
(10%). Further, as discussed above, the presence of a
nearby CH3 group apparently opens up a thermodynami-
cally favorable rearrangement channel that competes
with iodine abstraction within the collision complex with
allyl iodide.
Con clu sion s
The ability of the N-(4-dehydrophenyl)pyridinium ion
to abstract an iodine atom from allyl iodide was found to
be sensitive to substitution, and specifically, be enhanced
by the presence of electron-withdrawing substituents.
Qualitatively, the observations made here for iodine
abstraction by charged phenyl radicals under gas-phase
conditions parallel those made in solution for chlorine
and bromine atom abstraction from electron-donating
substrates (e.g., benzyl bromide) by the phenyl radical
and its electrophilic p-Cl, p-Br, and p-NO2 derivatives
(and are reverse to those observed with electron-accepting
polychlorinated substrates, e.g., CCl4, CBrCl3).5 Further,
in agreement with the slightly nucleophilic character
reported for the p-CH3-substituted phenyl radical in
solution,5a we find that introduction of a methyl sub-
stituent into the meta-position in the charged phenyl
radical does not change its iodine atom abstraction rate.
On the basis of our observations, the N-(4-dehydrophe-
nyl)pyridinium ion appears to provide a useful tool for
probing substituent effects. However, our findings also
suggest that polar effects may play a significantly greater
role in controlling the reaction rates of charged phenyl
radicals than those of the phenyl radical.
The results presented here underline the importance
of the ionic surface in controlling the barrier height in
radical atom abstraction reactions, as discussed by
Donahue et al. in several recent publications.26 The
observed reactivity trends, and hence barrier heights, do
not reflect changes in reaction enthalpy. Instead, the
observations are readily rationalized by considering an
avoided curve crossing that involves interaction of the
ground state and an ionic excited state of the reactants.
Accordingly, the electron affinity of the radicals is
demonstrated to play a crucial role in controlling their
iodine abstraction efficiency.
Ack n ow led gm en t. We thank the National Science
Foundation (CHE-9710456) for financial support of this
work. Mr. Hui Lui is thanked for performing some
experiments.
Calculations yield EA values of 4.59 and 4.61 eV (AM1)
for the 3-methyl- (ortho-) and 2-methyl-4-dehydro-
J O981911I