Solvent effects on methyl transfer reactions. 2. The reaction of amines with trimethylsulfonium salts

Article abstract of DOI:10.1021/ja0042940

The reaction of ammonia and pyridine with trimethylsulfonium ion has been studied in gas phase and solution. Density functional theory at the B3LYP/6-31+G* level was used to describe the energy changes along the reaction coordinate in the gas phase, and t

Full text of DOI:10.1021/ja0042940

6092
J. Am. Chem. Soc. 2001, 123, 6092-6097
Solvent Effects on Methyl Transfer Reactions. 2. The Reaction of
Amines with Trimethylsulfonium Salts
Henry Castejon,*,† Kenneth B. Wiberg,*,‡ Stepan Sklenak,‡ and Wolfgang Hinz‡
Contribution from Science Computing, UniVersity of Notre Dame, Notre Dame, Indiana 46556, and the
Department of Chemistry, Yale UniVersity, New HaVen, Connecticut 06520-8107
ReceiVed December 19, 2000
Abstract: The reaction of ammonia and pyridine with trimethylsulfonium ion has been studied in gas phase
and solution. Density functional theory at the B3LYP/6-31+G* level was used to describe the energy changes
along the reaction coordinate in the gas phase, and the self-consistent isodensity polarizable continuum model
(SCI-PCM) was used to calculate the effect of cyclohexane and dimethyl sulfoxide as the solvent on the
energy changes. The effect of water as the solvent was studied using the Monte Carlo free energy perturbation
method. The reaction with both ammonia and pyridine follows a similar rather convoluted path in gas phase,
with the formation of several reaction complexes before and after the formation of the transition state. All the
species found in gas phase persist in cyclohexane, yielding a reaction path very similar to that in gas phase but
with significant differences in the relative energy of the critical points. In DMSO, the energy profile is greatly
simplified by the disappearance of several of the species found in gas phase and in cyclohexane. The activation
free energy increases with the polarity of the solvent in both reactions. Increasing the polarity of the solvent
also increases the exothermicity of the reaction of trimethylsulfonium ion with ammonia and reduces it in the
reaction with pyridine. In water, the free energy profile follows the same trend as found for DMSO, and free
energy of activation is calculated to be larger by about 2-3 kcal/mol. This is in good agreement with an
experimental measurement of the effect of solvent on the rate of reaction.
Introduction
We have previously reported the results of an investigation
of solvent effects on Menshutkin reactions using the self-
consistent isodensity polarizable continuum model (SCI-PCM)
for aprotic solvents and Monte Carlo free energy perturbation
for water as the solvent.5 The rate of reaction of pyridine with
methyl bromide in several solvents was measured, and the
calculated free energies of activation were in very good
agreement with the experimental values.
This study has now been extended to an examination of the
reactions of amines with trimethylsulfonium salts. These reac-
tions are of some special interest in that the sulfonium salt is a
model for S-adenosylmethionene (1), an important biological
methyl transfer reagent,2 in which the methyl group is activated
by the positive adjacent sulfur atom. It is responsible for the
formation of phosphatidylcholine,6 a phosphoglyceride found
in most membranes of higher organisms, as well as the
conversion of homocysteine to methionine2 and other reactions.
The reaction of the trimethylsulfonium cation with ammonia
and with pyridine has now been studied in the gas phase via
density functional theory calculations. The effect of aprotic
solvents was examined making use of the SCI-PCM,7 and the
Monte Carlo free energy perturbation8 method was used to study
the reaction in aqueous solution.
Methyl transfer reactions are of importance in both synthesis1
and biochemical processes.2 The reactions are often reversible,
which is of importance in processes such as gene regulation.3
Here, methylation may be used to inactivate a gene, and
subsequently, demethylation may be used to reactivate it.
There are two basic types of methyl transfer reactions
involving neutral acceptors. In the first, the methyl transfer
reagent is neutral, as in the Menshutkin reaction. Here, the
transition state is dipolar and will be strongly stabilized by polar
solvents leading to the well-known rate acceleration in these
solvents.4
R3N: + MeX f [R3N-Me-X] f R3NMe+ + X-
In the second type, the methyl transfer reagent bears a positive
charge. Here, the transition state for the reaction is less polar
than the reactant because the charge is spread out over a large
volume. Now, polar solvents should decrease the rate of reaction.
R3N: + Me3S+ f [R3N-Me-SMe2]+ f R3NMe+ +
Me2S
Calculations
† University of Notre Dame..
The reactions in the gas phase were studied at the B3LYP/6-31+G*
level, which has generally been found to give a good representation of
reaction energies. Diffuse functions (+) were included in order to
‡ Yale University.
(1) Abboud, J.-L. M.; Notario, R.; Bertran, J.; Sola, M. Prog. Phys. Org.
Chem. 1993, 19, 1.
(2) Stryer, L. Biochemistry, 4th ed.; Freeman: New York, 1995; p 721.
(3) Finnegan, E. J.; Peacock, W. J.; Dennis, E. S. Curr. Opin. Genet.
DeV. 2000, 10, 217.
(4) Menshutkin, N. Z. Phys. Chem. 1890, 5, 589; 1890, 6, 41. Tommila,
E.; Kauranen, P. Acta Chem. Scand. 1954, 8, 1152. Arnett, E. M.; Reich,
R. J. Am. Chem. Soc. 1980, 102, 5982.
(5) Castejon, H.; Wiberg, K. B. J. Am. Chem. Soc. 1999, 121, 2139.
(6) Vance, D. E. Biochem. Cell Biol. 1990, 68, 1151.
(7) Foresman, J. B.; Keith, T. A.; Wiberg, K. B.; Snoonian, J.; Frisch,
M. J. J. Phys. Chem. 1996, 100, 16098.
(8) Jorgensen, W. L. Acc. Chem. Res. 1989, 22, 84.
10.1021/ja0042940 CCC: $20.00 © 2001 American Chemical Society
Published on Web 05/26/2001

SolVent Effects on Methyl Transfer Reactions
J. Am. Chem. Soc., Vol. 123, No. 25, 2001 6093
Figure 1. Reaction path for the reaction of ammonia and trimethylsulfonium ion in the gas phase. Relative energies in kilocalories per mole.
correctly represent the lone pairs9 on nitrogen. The calculations were
carried out using Gaussian-99.10
The reaction in water was studied using the Monte Carlo free energy
perturbation method of Jorgensen, making use of the BOSS program.11
In the present case, an NTP ensemble at 1.0 atm and 25 °C was used.
The solvated system consisted of a periodic cube containing the solute
system and 512 water molecules. Preferential sampling was used in
the Metrolpolis algorithm, and perturbations were carried out using
double-wide sampling windows. Equilibration was carried out over 3
million configurations, followed by averaging over 5 million configura-
tions for each window.
Pairwise additive intramolecular potential functions were used for
the solvent-solvent and solvent-solute interactions, with a nonbonded
solvent-solvent and solute-solvent cutoff distances set at 10 Å, and
the value of WKC parameter of 200. No intramolecular terms were
included. TIP4P water was used with Lennard-Jones parameters taken
directly from the BOSS database. CHELPG charges12 were calculated
at each point along the reaction coordinate and used in the coulomb
term of the intermolecular potentials.
The solvent effects of cyclohexane and dimethyl sulfoxide were
studied using the SCI-PCM, which has been found to be quite successful
in reproducing the effects of aprotic solvents such as cyclohexane,
ethers, acetone, acetonitrile, and dimethyl sulfoxide.7 This reaction field
method requires only two parameters: the dielectric constant of the
solvent and the value of the electron density used to describe the size
and shape of the cavity into which the solute will be placed. The latter
is chosen to be 0.0004 e/au3 since it leads to molar volumes that are in
good accord with experimental data. The method cannot be used for
protic solvents such as water because the model assumes that the solvent
molecules will adopt all random orientations with respect to the solute.
This is not true if there is hydrogen bonding between the solvent and
the solute.
In the calculations, dielectric constant values of 2.02 and 46.7 for
cyclohexane and dimethyl sulfoxide, respectively, were used. Surface
integration was done using both single origin and multiple origins.
Reactions in Gas Phase
(9) Clark, T.; Chandrasekhar, J.; Spitznagel, G. W.; Schleyer, P. v. R. J.
Comput. Chem. 1983, 4, 294.
The energy profile for the reaction of ammonia with trim-
ethylsulfonium ion, as calculated at the B3LYP/6-31+G*
theoretical level, is shown in Figure 1. When approaching each
other, the reacting molecules initially form an ion-dipole
complex. The strong electrostatic interaction between the ion
and the dipole of the ammonia molecule stabilizes the complex
by -9.4 kcal/mol, with respect to the separated reactants. The
structure of this complex corresponds to the most stable
arrangement of the two interacting species in gas phase.
(10) 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.; Ortis, J. V.;
Baboul, A. G.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.;
Piskorz, P.; Komaromi, I.;. Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith,
T. A.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.;
Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Andres, J. L.;
Gonzalez, C.; Head-Gordon, M.; Replogle, E. S.; Pople, J. A.. Gaussian
99, Development Version (Rev. B) Gaussian, Inc., Pittsburgh, PA, 1998.
(11) Jorgensen, W. L. Boss v. 4.1, Yale University.
(12) Breneman, C. M.; Wiberg, K. B. J. Comput. Chem. 1990, 11, 361.

6094 J. Am. Chem. Soc., Vol. 123, No. 25, 2001
Castejon et al.
Table 1. Critical Points in the Gas Phase Energy Profile for the
Reaction of Trimethylsulfonium with Ammonia and with Pyridinea
entity
r(C-N) r(S-C)
NH3-(CH3)3S+
HCS) Erel % BSb CTc
reactants
ion-dipole complex 3.624
second complex
transition state
third complex
products
1.824
1.824
1.831
2.402
3.916
108
108
109
90
0.0
-9.4
-6.9
0.0 0.000
0.018
0.4 0.047
3.074
2.065
1.506
1.516
9.8 31.7 0.442
-16.3
0.667
0.778
0.4
Pyridine-(CH3)3S+
reactants
1.824
1.824
1.831
2.321
3.938
108
108
109
92
0.0
-11.3
-8.0
0.0 0.000
0.025
0.4 0.050
ion-dipole complex 3.583
second complex
transition state
third complex
products
3.043
2.102
1.487
1.486
4.9 27.2 0.408
-24.1
-16.9
0.739
0.781
a Bond distances in angstroms, bond angles in degrees, and energy
in kilocalories per mole. The Erel include the zero point energies.
b Percent of the S-C bond stretching. c Percent of charge transferred
to the B:CH3 moiety for B ) ammonia, pyridine.
Figure 2. Reaction paths for the reactions of trimethylsulfonium ion
with ammonia in the gas phase (a), in cyclohexane (b), and in DMSO
(c). The energies are given relative to the reactants in each of the media
and do not include the zero point energies. The reation coordinate is
the difference between the C-N and S-C bond lengths. The
discontinuity following the transition state represents the rotation of
the methylammonium fragment in order to form a hydrogen bond with
dimethyl sulfide, a process in which the above reaction coordinate
changes little.
However, in order for the methyl transfer reaction to occur,
the leaving methyl group must be aligned with the lone pair in
the nitrogen atom. This aligning process destabilizes the ion-
dipole complex and leads to a complex having a slightly higher
energy, -6.9 kcal/mol. A transition state between these
complexes was not located, but the activation energy is
presumably very low.
From this point up to the formation of the transition state for
the overall reaction, a good description of the reaction path can
be made by using the difference between the C-N and S-C
bond distances as the reaction coordinate. The transfer of the
methyl group to ammonia leads to a transition state at the
reaction coordinate value of 0.337 Å with an activation energy
of 9.8 kcal/mol. The value of 90° for the HCS bond angle
indicates that the methyl group is halfway in the inversion of
conformation that it undergoes upon transfer to the ammonia.
After the transfer of the methyl group is completed, the C-S
bond initially continues to lengthen, but then the strong
electrostatic interaction between the hydrogen atoms bonded
to the nitrogen of the methylammonium and the lone pair of
the sulfur atom causes the methylammonium to rotate to form
a hydrogen-bonded complex as depicted in Figure 1. This
interaction strongly stabilizes the complex by -16.3 kcal/mol
with respect to the reactants. A further separation of the products
brings the energy of the system up to 0.4 kcal/mol, indicating
a very similar methyl affinity for ammonia and dimethyl sulfide.
When ammonia is replaced with pyridine, the overall energy
profile is shifted down in energy. Table 1 shows the energy of
the critical points in the reaction path for the reactions with
ammonia and with pyridine. The changes in relative energies
between ammonia and pyridine as the base are largest at and
after the transition state.
is also observed in the relative stability of the products: those
in the pyridine reaction are 16.9 kcal/mol below the reactants,
while in the ammonia reaction, they are 0.4 kcal/mol above the
reactants. Similar trends were found in the Menshutkin reaction.5
The last two columns in Table 1 show the reaction coordinate
as a percent of the S-C bond stretching and the amount of
charge transferred at each critical point on the reaction path. It
can be seen that although the transition state for the pyridine
appears at an earlier stage on the reaction path, the amount of
charge transferred is almost the same as in the ammonia reaction.
Solvent Effect Calculations
(a) Aprotic Solvents. The SCI-PCM reaction field model
has been found to be quite successful in modeling solvent effects
for aprotic solvents.13 It is capable of representing the significant
effect of solvents with low polarity such as cyclohexane, long
thought to be equivalent to gas phase as a reaction medium.14
As an example, the calculated barrier for the Menshutkin
reaction of methyl bromide with pyridine in cyclohexane agreed
remarkably well with the experimental value.5 The polarizability
of cyclohexane, which gives it a dielectric constant of 2, is
responsible for its small solvent effect. Figures 2 and 3 and
Table 2 show the reaction path and the relative energy of the
critical points along the path for the reaction of trimethylsul-
fonium with ammonia and with pyridine in cyclohexane.
Even though the relative energies of the critical points in
cyclohexane solution are different, the reaction path is essentially
the same as in gas phase since all of the species found in gas
phase persist in solution. The energies of all of the initial species
and that of the transition state have been raised. It is only with
the separated product species that solvent stabilization is found.
The value of the reaction coordinate for the pyridine reaction
transition state, taken again as the difference between the C-N
and C-S bond distances, is 0.219 Å as compared to 0.337 Å
for the transition state in the reaction with ammonia. The C-S
bond is stretched half as much as in the ammonia reaction, and
the H-C-S angle is 92°, indicating that the methyl group has
not passed the middle point of the inversion umbrella mode.
No hydrogen-bonded complex is found after the transition state,
since methylpyridinium ion has no protons attached to the
nitrogen.
(13) Wiberg, K. B.; Castejon, H.; Keith, T. A. J. Comput. Chem. 1996,
17, 185. Wong, M. W.; Frisch, M. J.; Wiberg, K. B. J. Am. Chem. Soc.
1996, 100, 16162. Wiberg, K. B.; Rablen, P. R.; Rush, D. J.; Keith, T. A.
J. Am. Chem. Soc. 1995, 117, 4261.
(14) Jorgensen, W. L.; McDonald, N. A.; Selmi, M.; Rablen, P. R. J.
Am. Chem. Soc. 1995, 117, 11809.
The larger basicity of pyridine in the gas phase stabilizes the
ion-dipole complex of methylpyridinium ion and dimethyl
sulfide with respect to that in the ammonia reaction. This effect

SolVent Effects on Methyl Transfer Reactions
J. Am. Chem. Soc., Vol. 123, No. 25, 2001 6095
Table 3. Critical Points in the Energy Profile for the Reaction of
Trimethylsulfonium with Ammonia and with Pyridine in Dimethyl
Sulfoxidea
entity
r(C-N) r(S-C) (HCS) Erel % BSb CTc
NH3-(CH3)3S+
reactants
ion-dipole complex 3.571
transition state
third complex
products
1.821
1.822
2.297
3.959
108
108
93
0.0
0.0 0.000
0.016
-2.1
2.130
1.504
1.501
15.7 26.1 0.422
-17.6
0.685
0.779
-13.4
Pyridine-(CH3)3S+
reactants
transition state
products
1.821
2.331
108
92
0.0
0.0 0.000
2.102
1.484
16.0 28.0 0.431
-12.8 0.783
a Bond distances in angstroms, bond angles in degrees, and energy
in kilocalories per mole. The Erel include the zero point energies.
b Percent of the S-C bond stretching. c Charge transferred to the B:CH3
moiety for B ) ammonia, pyridine.
complex) for the reaction with ammonia survive with energies
-2.1 and -17.6 kcal/mol, respectively, below the reactants.
Figure 3. Reaction paths for the reactions of trimethylsulfonium ion
with pyridine in the gas phase (a), in cyclohexane (b), and in DMSO
(c).
The transition states for both reactions have been destabilized
with respect to the reactants. The destabilization is again larger
in the pyridine reaction. The energies of the transition states
have been increased by 11.1 and 5.4 kcal/mol with respect to
that in the gas phase and cyclohexane, respectively. This is larger
than the increments of 5.9 and 3.1 kcal/mol for the transition
states for the reaction of ammonia in the same solvents.
Table 2. Critical Points in the Energy Profile for the Reaction of
Trimethylsulfonium with Ammonia and with Pyridine in
Cyclohexanea
entity
r(C-N) r(S-C) (HCS) Erel % BSb CTc
NH3-(CH3)3S+
reactants
1.822
1.823
1.827
2.351
3.959
108
108
109
91
0.0
-6.1
-3.7
0.0 0.000
0.015
The energy barrier is the same for the two reactions in
dimethyl sulfoxide. This is again a result of the differential
solvation of the two transition states.
ion-dipole complex 3.595
second complex
transition state
third complex
products
3.104
2.097
1.504
1.510
0.3 0.040
12.6 29.0 0.431
-16.3
0.682
0.778
-6.6
(b) Water as the Solvent. The reaction field model would
lead to the expectation that the solvent effects of DMSO and
water would be essentially the same since in both cases (ꢀ -
1)/(2ꢀ + 1) is close to the maximum value of 0.5. However, in
many cases, these solvents give different solvent effects.15 One
of the more satisfactory ways of treating aqueous solutions is
to make use of an explicit representation of the solvent that
will better account for the interaction of water with the solute.
Thus, we have chosen to use the free energy perturbation method
of Jorgensen, which has been shown to reproduce experimental
results for many reactions.16
Pyridine-(CH3)3S+
reactants
1.822
1.824
1.826
2.319
4.009
108
108
109
92
0.0
-6.8
-4.9
0.0 0.000
0.011
ion-dipole complex 3.627
second complex
transition state
third complex
products
3.144
2.090
1.486
1.485
0.2 0.032
10.6 27.3 0.421
-19.5
0.737
0.781
-15.2
a Bond distances in angstroms, bond angles in degrees, and energy
in kilocalories per mole. The Erel include the zero point energies.
b Percent of the S-C bond stretching. c Charge transferred to the B:CH3
moiety for B ) ammonia, pyridine.
Here, the concentration of the charge onto a smaller ion (i.e.,
methylammmonium ion) facilitates the solvation process.
In the case of pyridine, the solvation of the products is not
as effective because of the larger size of the pyridinium ion
and its more diffuse charge density. Thus, the whole energy
profile is shifted upward. The differential solvation between
ionic species of different sizes is even reflected in the energies
of the transition states for the two reactions. In the reaction with
ammonia, the energy of the transition state is raised by 2.8 kcal/
mol on going to cyclohexane solution, while in the reaction
with pyridine, whose transition state is larger with a more diffuse
charge, the energy increases by 5.7 kcal/mol.
The dielectric constant of cyclohexane is not large enough
to either alter the path of the reaction or overcome the strong
stabilizing effect of pyridine on the products, which is the reason
the pyridine reaction exhibits the same exothermicity in the gas
phase and in cyclohexane.
Figures 2 and 3 show the energy profile for the reaction with
ammonia and pyridine in dimethyl sulfoxide. The relative
energies of the critical points along the reaction path are shown
in Table 3. The polarity of the solvent is now large enough to
alter the reaction path. Almost all the intermediates found in
cyclohexane solution have disappeared. Only the initial ion-
dipole complex and the hydrogen-bonded complex (third
A full characterization of the energy profile in the gas phase
was done using the B3LYP/6-31+G* level of theory. The
reaction path was traced from the transition state to the reactants
in intervals of 0.05 Å in the C-N distance. This procedure
produces essentially a movie of the reaction between the two
critical points on the potential energy surface. ChelpG charges17
were obtained for each of the structures along the path. The
Monte Carlo simulation makes use of the gas-phase reaction
coordinate. Petersson has shown that because the reaction
coordinate is a relatively “soft” mode in contrast to the other
vibrational modes, reaction coordinates calculated at a lower
theoretical level can be used as the basis for calculations at a
higher theoretical level.18 The location of the transition state
will move, but to a good approximation, it will remain on the
reaction coordinate. The same should hold true for the Monte
Carlo calculations using the gas-phase reaction coordinate.
Starting at the gas-phase transition-state geometry and moving
(15) Reichardt, C. SolVents and SolVent Effects in Organic Chemistry;
VCH: Weinheim, Germany, 1988.
(16) Blake, J. F.; Jorgensen, W. L. J. Am. Chem. Soc. 1991, 113, 7430.
Severance, D. L.; Jorgensen, W. L. J. Am. Chem. Soc. 1992, 114, 10966.
(17) Breneman, C. M.; Wiberg, K. B. J. Comput. Chem. 1990, 11, 361.
(18) Malik, D. K.; Petersson, G. A.; Montgomery, J. A., Jr. J. Chem.
Phys. 1998, 108, 5704.

6096 J. Am. Chem. Soc., Vol. 123, No. 25, 2001
Castejon et al.
Table 4. Reaction and Activation Free Energies in Gas Phase and
Solution, 25 °C
Table 7. Kinetics of the Reaction of Pyridine with
Trimethylsulfonium Triflate at 55.0 °C
∆G‡
∆Gr
solvent
acetone
[pyridine] (M)
k2 (L mol-1 s-1
)
∆G‡ (kcal/mol)
ammonia
calc
pyridine
calc
obsa
ammonia pyridine
0.03-0.06
(9.2 ( 0.1) × 10-7
(1.19 ( 0.03) × 10-6
(7.06 ( 0.03) × 10-8
(3.64 ( 0.03) × 10-7
28.3
28.1
30.0
28.9
acetonitrile 0.03-0.07
phase
calc
calc
D2O
0.03-0.07
0.03-0.07
D2Oa
gas phase
cyclohexane
DMSO
17.1
19.9
23.0
26.5
13.3
19.0
-1.0
-8.0
-14.8
-12.1
-18.0
-16.3
-13.9
-13.4
a Using trimethylsulfonium iodide.
24.4 (27.8)b
25.9 29.6
water
Tables 2 and 3 derived from the SCI-PCM calculations contain
the energies of the species and the free energies of solvation.
To convert these values to free energies, it was assumed that
the entropy change would be the same in the gas phase and in
solution. In at least one well-studied case, this has been shown
to be the case.20 It also was assumed that the changes in the
cavity formation and dispersion terms along the reaction path
will approximately cancel. There is evidence that this is an
appropriate approximation.21
a Corrected to 25 °C using the calculated entropies of activation.
b Measured value in acetone or acetonitrile at 55 °C.. The DMSO value
would be expected to be the same.
Table 5. Position of the Transition States
base
medium
NH3
pyridine
gas phase
cyclohexane
DMSO
0.337
0.254
0.167
0.188
0.219
0.229
0.229
0.219
water
Experimental Studies
The rate of reaction between pyridine and trimethylsulfonium
triflate was studied in acetone, acetonitrile, and aqueous
solutions at 55 °C. Pseudo-first-order kinetics were studied using
pyridine as the limiting reagent, and the concentration of
Table 6. Transition-State Structures from Aqueous Solution BOSS
Simulations
base
NH3
pyridine
r(C-N)
r(C-S)
(HCS) (deg)
2.193
2.102
2.381
2.321
92.2
91.6
1
pyridine as a function of time was determined via H NMR
spectroscopy. The reaction was studied through one half-life,
and linear plots of ln[pyridine] vs time were obtained. The
resulting second-order rate constants and the free energies of
activation are shown in Table 7. The free energies of activation
in acetone and acetonitrile were essentially the same, as expected
from the reaction field model, and they are in satisfactory
agreement with the calculated value for DMSO, which also
should give the same activation parameter. The small deviation
between experiment and theory is not unexpected. The assump-
tion that the change in activation entropy is the same in the gas
phase and in solution is one potential source of the difference.
toward reactants, the free energy first increases, and after the
new transition-state location is reached, it begins to decrease.
Table 4 shows the activation free energy and the reaction
free energy for both reactions in all media, including water.
The calculated activation free energies of the two reactions in
water were only slightly larger than those in dimethyl sulfoxide.
One might have expected a larger effect since the amines are
stabilized by hydrogen bonding in aqueous solution, and this
will be lost during the reaction. The destabilization trend as the
solvent polarity increases is, however, observed, as the activation
free energies are higher in this solvent. The reaction free energies
here also follow the same trend as in the other solvents. Namely,
it becomes more negative than in cyclohexane and gas phase
for the reaction of ammonia and less negative for the pyridine
reaction.
Table 5 shows the position of the transition state along the
reaction path using the difference between the C-N and S-C
bond distances as the reaction coordinate. For the ammonia
reaction, the transition state is shifted to an earlier stage of the
reaction as the polarity of the solvent increases. In contrast, no
significant shift is observed in the position of the transition states
for the pyridine reaction. Table 6 shows the structural parameters
for the transition states obtained in the BOSS simulations.
Despite the small effect of solvent polarity on the position of
the transition state for the pyridine reaction, the values of the
H-C-S angles suggest that both reactions have early transition
states in water.
Although the calculated difference in activation free energy
on going from DMSO to water as the solvent is relatively small
(1.5 kcal/mol), it does agree very well with the experimental
observation (1.8 kcal/mol).
When the reaction with pyridine in water was carried out
using trimethylsulfonium iodide, the reaction occurred at a
significantly faster rate. It seems likely that iodide ion initially
reacts with the sulfonium salt to form methyl iodide,22 which
would rapidly react with pyridine to form the product.
Conclusions
The rather convoluted reaction path for the reaction of
trimethylsulfonium with ammonia and with pyridine in gas
phase is greatly simplified on going to a polar solvent. The
activation free energy increases with the increasing polarity of
the solvent as a consequence of the differential solvation
between the compact ionic reactants and the transition states
with a diffuse charge distribution. The spreading of the charge
that stabilizes the transition state for pyridine, compared to
ammonia, hinders its solvation leading to a larger increase of
the activation free energy with the increase of the solvent’s
polarity. Differential solvation is also responsible for the
The relative energies given in Table 1 must be converted to
free energies in order to compare them with the experimental
data given below. This was done using the calculated entropy
terms and zero-point energies derived from vibrational frequen-
cies calculated at the B3LYP/6-31+G* level.5 It has been found
that vibrational frequencies are changed relatively little on going
from the gas phase to a solution.19 The relative energies in
(20) Wasserman, A. Monatsh. Chem. 1952, 83, 543.
(21) Amovilli, C.; Mennucci, B.; Floris, F. M. J. Phys. Chem. B 1998,
102, 3023, Table 2.
(19) Wong, M. W.; Wiberg, K. B.; Frisch, M. J. Chem. Phys. 1991, 95,
8991.
(22) Hughes, E. D.; Ingold, C.; Pocker, Y. Chem. Ind. 1959, 1282.

SolVent Effects on Methyl Transfer Reactions
J. Am. Chem. Soc., Vol. 123, No. 25, 2001 6097
opposite effect of the polarity of the solvent on the reaction
free energy of both reactions. In the case of ammonia, the
products are stabilized with respect to the reactants, increasing
the exothermicity of the reaction, while with pyridine, the better
solvation of the reactants leads to a much less exothermic
reaction in solution than in gas phase.
The small change in the position of the transition state for
the reaction with pyridine on going from one medium to another
is remarkable and probably results from the dispersal of charge
into the pyridine ring, making the medium effect on the position
of the transition state relatively small.
Acknowledgment. We thanks the National Institutes of
Health for their support of this study, and we thank Prof. W.
Jorgensen for providing a copy of his BOSS program.
Supporting Information Available: A table of point groups,
energies, zero-point energies, and Cartesian coordinates of the
compounds in this study. This information is available free of
charge via the Internet at http://pubs.acs.org. See any current
masthead page for ordering information and Web access
instructions.
JA0042940