Can one predict changes from SN1 to SN2 mechanisms?

Article abstract of DOI:10.1021/ja903207b

The reactions of substituted benzhydryl bromides Ar₂CHBr with primary and secondary amines in DMSO yield benzhydryl amines Ar ₂CHNRR′, benzophenones Ar₂CdO, and benzhydrols Ar₂CHOH. Kinetic investigations at 20°C revealed the rate law -d[Ar₂CHBr]/dt = (k₁ + k₂[HNRR′])[Ar ₂CHBr], where the amine independent term k₁ gave rise to the formation of Ar₂CdO and Ar₂CHOH and the amine-dependent term k₂[HNRR′] was responsible for the formation of Ar₂CHNRR′. Clear evidence for concomitant S _N1 and S_N2 processes was obtained. While the rate constants of the S_N1 reactions correlate with Hammett's σ⁺ constants (ρ = -3.22), the second-order rate constants k₂ for the S_N2 reactions are not correlated with the electron releasing abilities of the substituents, indicating that the transition states of the S_N2 reactions do not merge with the transition states of the S_N1 reactions. The correlation equation log k _20°C = s(E + N), where nucleophiles are characterized by N and s and electrophiles are characterized by E (J. Am. Chem. Soc. 2001, 123, 9500-9512), was used to calculate the lifetimes of benzhydrylium ions in the presence of amines and DMSO. The change from S_N1 to S_N2 mechanism occurred close to the point where the calculated rate constant for the collapse of the benzhydrylium ions with the amines just reaches the vibrational limit; that is, the concerted S_N2 mechanism was only followed when it was enforced by the lifetime of the intermediate. The nucleophile-specific parameters N and s needed for this analysis were determined by studying the kinetics of the reactions of a variety of amines with amino-substituted benzhydrylium tetrafluoroborates (Ar₂CH⁺BF4^-) of known electrophilicity E in DMSO. Analogously, the rates of the reactions of laser flash photolytically generated benzhydrylium ions Ar₂CH ⁺ with DMSO in acetonitrile were employed to determine the nucleophile-specific parameters N and s of DMSO, and it is reported that DMSO is a significantly stronger O-nucleophile than water and ordinary alcohols.

Full text of DOI:10.1021/ja903207b

Published on Web 07/27/2009
Can One Predict Changes from SN1 to SN2 Mechanisms?
Thanh Binh Phan,† Christoph Nolte, Shinjiro Kobayashi, Armin R. Ofial, and
Herbert Mayr*
Department Chemie und Biochemie der Ludwig-Maximilians-UniVersita¨t Mu¨nchen,
Butenandtstrasse 5-13 (Haus F), 81377 Mu¨nchen, Germany
Received April 29, 2009; E-mail: herbert.mayr@cup.uni-muenchen.de
Abstract: The reactions of substituted benzhydryl bromides Ar2CHBr with primary and secondary amines
in DMSO yield benzhydryl amines Ar2CHNRR′, benzophenones Ar2CdO, and benzhydrols Ar2CHOH. Kinetic
investigations at 20 °C revealed the rate law -d[Ar2CHBr]/dt ) (k1 + k2[HNRR′])[Ar2CHBr], where the amine
independent term k1 gave rise to the formation of Ar2CdO and Ar2CHOH and the amine-dependent term
k2[HNRR′] was responsible for the formation of Ar2CHNRR′. Clear evidence for concomitant SN1 and SN2
processes was obtained. While the rate constants of the SN1 reactions correlate with Hammett’s σ+ constants
(F ) -3.22), the second-order rate constants k2 for the SN2 reactions are not correlated with the electron
releasing abilities of the substituents, indicating that the transition states of the SN2 reactions do not merge
with the transition states of the SN1 reactions. The correlation equation log k20°C ) s(E + N), where
nucleophiles are characterized by N and s and electrophiles are characterized by E (J. Am. Chem. Soc.
2001, 123, 9500-9512), was used to calculate the lifetimes of benzhydrylium ions in the presence of amines
and DMSO. The change from SN1 to SN2 mechanism occurred close to the point where the calculated rate
constant for the collapse of the benzhydrylium ions with the amines just reaches the vibrational limit; that
is, the concerted SN2 mechanism was only followed when it was enforced by the lifetime of the intermediate.
The nucleophile-specific parameters N and s needed for this analysis were determined by studying the
kinetics of the reactions of a variety of amines with amino-substituted benzhydrylium tetrafluoroborates
(Ar2CH+BF4-) of known electrophilicity E in DMSO. Analogously, the rates of the reactions of laser flash
photolytically generated benzhydrylium ions Ar2CH+ with DMSO in acetonitrile were employed to determine
the nucleophile-specific parameters N and s of DMSO, and it is reported that DMSO is a significantly stronger
O-nucleophile than water and ordinary alcohols.
character.3-6 Winstein’s concept of different types of ion pairs4
Introduction
was extended by Sneen who suggested that the entire SN1-SN2
mechanistic spectrum could be fitted into a simple scheme
involving ion-pair intermediates.5 Schleyer and Bentley criticized
this concept and suggested that there is a gradation of transition
states between the SN1 and SN2 extremes with varying degrees of
nucleophilic participation by the solvents.6,7 The intermediates in
the borderline region were considered as “nucleophilically solvated
ion pairs”6 which look like the transition states of SN2 reactions
Nucleophilic displacement reactions at C(sp3) centers1 proceed
either with simultaneous breaking and forming of the involved
bonds (SN2 or ANDN)2 or via a mechanism where breaking of the
old bond precedes formation of the new bond (SN1 or DN+AN).1
The borderline between these two mechanisms has been the subject
of considerable controversy. In contrast to Ingold who considered
SN1 and SN2 as discrete processes,1b it has been suggested that a
clear-cut distinction between these two mechanisms is impossible
because there is a gradual transformation of an SN2 into an SN1
mechanism as the transition state develops more carbocation
(3) Winstein, S.; Grunwald, E.; Jones, H. W. J. Am. Chem. Soc. 1951,
73, 2700–2707.
(4) (a) Winstein, S.; Clippinger, E.; Fainberg, A. H.; Heck, R.; Robinson,
G. C. J. Am. Chem. Soc. 1956, 78, 328–335. For a summary of
Winstein’s contribution, see: (b) Bartlett, P. D. J. Am. Chem. Soc.
1972, 94, 2161–2170. (c) Winstein, S.; Appel, B.; Baker, R.; Diaz,
A. In Symposium on Organic Reaction Mechanisms, Chemical Society
(London), Special Publication No. 19, 1965; p 109.
† Current address: Vietnam Institute of Industrial Chemistry, Hanoi,
Vietnam.
(1) (a) Streitwieser, A., Jr. SolVolytic Displacement Reactions; McGraw-
Hill: New York, 1962. (b) Ingold, C. K. Structure and Mechanism in
Organic Chemistry, 2nd ed.; Cornell University Press: Ithaca, NY,
1969. (c) Vogel, P. Carbocation Chemistry; Elsevier: Amsterdam,
1985. (d) AdVances in Carbocation Chemistry; Coxon, J. M., Ed.;
JAI Press: Greenwich, CT, 1995; Vol 2. (e) Raber, D. J.; Harris, J. M.;
Schleyer, P. v. R. In Ions and Ion Pairs in Organic Reactions; Szwarc,
M., Ed.; Wiley: New York, 1974; Vol. 2, pp 247-374. (f) Smith,
M. B.; March, J. March’s AdVanced Organic Chemistry, 6th ed.;
Wiley: New York, 2007; Chapter 10. (g) Vlasov, V. M. Russ. Chem.
ReV. 2006, 75, 765-796.
(5) (a) Sneen, R. A. Acc. Chem. Res. 1973, 6, 46–53. (b) Sneen, R. A.;
Larsen, J. W. J. Am. Chem. Soc. 1969, 91, 6031–6035. (c) Sneen,
R. A.; Larsen, J. W. J. Am. Chem. Soc. 1969, 91, 362–366.
(6) (a) Bentley, T. W.; Schleyer, P. v. R. J. Am. Chem. Soc. 1976, 98,
7658–7666. (b) Schadt, F. L.; Bentley, T. W.; Schleyer, P. v. R. J. Am.
Chem. Soc. 1976, 98, 7667–7675. (c) Bentley, T. W.; Bowen, C. T.;
Morten, D. H.; Schleyer, P. v. R. J. Am. Chem. Soc. 1981, 103, 5466–
5475.
(2) (a) Guthrie, R. D. Pure Appl. Chem. 1988, 61, 23–56. (b) Guthrie,
R. D.; Jencks, W. P. Acc. Chem. Res. 1989, 22, 343–349.
(7) Bentley, T. W.; Schleyer, P. v. R. AdV. Phys. Org. Chem. 1977, 14,
1–67.
9
11392 J. AM. CHEM. SOC. 2009, 131, 11392–11401
10.1021/ja903207b CCC: $40.75  2009 American Chemical Society

Can One Predict Changes from SN1 to SN2 Mechanisms?
A R T I C L E S
Scheme 1. Reactions of Benzhydryl Bromides with Amines in
DMSO
but are energy minima not maxima. They coined the term “SN2
intermediate” mechanism.6
Support for the operation of concurrent SN1 and SN2 reactions
in the borderline cases came from kinetic investigations of
nucleophilic substitutions under nonsolvolytic conditions, where
the concentration of the nucleophile could be varied.8 Nucleo-
philic displacement reactions of benzhydryl thiocyanates with
labeled *SCN- in acetonitrile and acetone,9 of benzhydryl
chlorides with labeled Cl- and Br-, and of benzhydryl bromide
-
with Br-, Cl-, and N3 as well as with amines followed the
rate law 1 with a nucleophile-independent term k1 and a
nucleophile-dependent term k2.10,11
-d[R-X]/dt ) [R-X](k1 + k2[Nu])
(1)
Yoh and Fujio et al. studied the kinetics of the reactions of benzyl
halides and tosylates with amines.12,13 While acceptor-substituted
benzyl derivatives reacted exclusively by the SN2 mechanism,
donor-substituted benzyl derivatives, such as p-methoxybenzyl
bromide, followed the rate law eq 1. This observation was
considered “convincing evidence for the occurrence of simultaneous
SN1 and SN2 mechanisms”.14a,b Concurrent stepwise and concerted
substitutions have also been reported by Amyes and Richard for
the reactions of azide ions with 4-methoxybenzyl derivatives in
trifluoroethanol/water mixtures.14c
Analogous rate laws have been observed by Katritzky for
alkyl and benzyl group transfers from N-alkyl and N-benzyl
pyridinium ions to various nucleophiles.15 Because of the
manifold of examples which demonstrate the duality of the two
mechanisms the question arises whether the change from one
to the other mechanism can be predicted.
ates.16 It has been argued that nucleophilic aliphatic substitutions
generally occur by the stepwise SN1 mechanism when the
intermediate carbocations exist in energy wells for at least
the time of a bond vibration (≈10-13 s) and that the change to
the SN2 mechanism is “enforced” when the energy well for the
intermediate disappears. Convincing support for this hypothesis
has been derived from the selectivities of carbocations (kazide
/
kROH), which were solvolytically generated in alcoholic solutions
of ionic azides.16c,d,17
We have reported that the rates of the reactions of carboca-
tions with nucleophiles can be calculated by eq 2, where E is a
carbocation-specific electrophilicity parameter and s and N are
solvent-dependent nucleophile specific parameters.18-20
log k20°C ) s(N + E)
(2)
Jencks and Richard based the differentiation of the mecha-
nistic alternatives on the lifetimes of the potential intermedi-
While the confidence limit of eq 2 is generally a factor of
10-100 in the presently covered reactivity range of 40 orders
of magnitude, the predictive power of eq 2 is much better for
reactions of benzhydrylium ions (factor 2-3) because benzhy-
drylium ions were used as reference electrophiles for deriving
the nucleophile-specific parameters s and N. We now set out to
examine whether the rate constants calculated by eq 2 can be
used to predict the change from SN1 to SN2 mechanism on the
basis of the lifetime criterion by Jencks and Richard. For that
purpose, we investigated rates and products of the reactions of
benzhydryl bromides with amines in DMSO, which yield
benzhydryl amines 4, benzophenones 5, and benzhydrols 6.
Scheme 1 shows that for each of the products 4-6 formation
via the SN1 process (k1) or the SN2 process (k2 and k1′) has to
be considered. In the following, it will be shown that the
pathways k1′ and kN can be excluded.
(8) (a) Gold, V. J. Chem. Soc. 1956, 4633–4637. (b) Gregory, B. J.;
Kohnstam, G.; Paddon-Row, M.; Queen, A. J. Chem. Soc. D 1970,
1032–1033. (c) Gregory, B. J.; Kohnstam, G.; Queen, A.; Reid, D. J.
J. Chem. Soc., Chem. Commun. 1971, 797–799. (d) Buckley, N.;
Oppenheimer, N. J. J. Org. Chem. 1997, 62, 540–551.
(9) (a) Ceccon, A.; Papa, I.; Fava, A. J. Am. Chem. Soc. 1966, 88, 4643–
4648. (b) Fava, A.; Iliceto, A.; Ceccon, A. Tetrahedron Lett. 1963,
685–692.
(10) (a) Diaz, A. F.; Winstein, S. J. Am. Chem. Soc. 1964, 86, 5010–5011.
(b) Casapieri, P.; Swart, E. R. J. Chem. Soc. 1963, 1254–1262. (c)
Casapieri, P.; Swart, E. R. J. Chem. Soc. 1961, 4342–4347.
(11) (a) Pocker, Y. J. Chem. Soc. 1959, 3939–3943. (b) Pocker, Y. J. Chem.
Soc. 1959, 3944–3949.
(12) (a) Lim, C.; Kim, S.-H.; Yoh, S.-D.; Fujio, M.; Tsuno, Y. Tetrahedron
Lett. 1997, 38, 3243–3246. (b) Tsuno, Y.; Fujio, M. AdV. Phys. Org.
Chem. 1999, 32, 267–385.
(13) (a) Kim, S. H.; Yoh, S.-D.; Lim, C.; Mishima, M.; Fujio, M.; Tsuno,
Y. J. Phys. Org. Chem. 1998, 11, 254–260. (b) Kim, S. H.; Yoh, S.-
D.; Fujio, M.; Imahori, H.; Mishima, M.; Tsuno, Y. Bull. Korean
Chem. Soc. 1995, 16, 760–764. (c) Yoh, S.-D.; Tsuno, Y.; Fujio, M.;
Sawada, M.; Yukawa, Y. J. Chem. Soc., Perkin Trans. 2 1989, 7–13.
(d) Yoh, S. D. J. Korean Chem. Soc. 1975, 19, 240–245.
Experimental Section
Conductimetric Measurements of Nucleophilic Substitu-
tions. Dissolution of the benzhydryl bromides 1-X,Y in DMSO or
in solutions of amines in DMSO led to an increase of conductivity
due to the generation of HBr, which reacted with excess amine to
give the hydrobromide salt. The rates of these reactions were
followed by conductimetry (conductimeters: Tacussel CD 810 or
Radiometer Analytical CDM 230; Pt electrode: WTW LTA 1/NS),
while the temperature of the solutions was kept constant (20.0 (
0.1 °C) by using a circulating bath thermostat. The correlation
between conductance and the concentration of liberated HBr was
determined by injecting 0.25 mL portions of 0.11 M acetonitrile
solutions of the rapidly ionizing benzhydryl bromide 1-Me,H into
30.0 mL of a 0.34 M solution of piperidine in DMSO. After the
(14) (a) Yoh, S.-D.; Cheong, D.-Y.; Lee, C.-H.; Kim, S.-H.; Park, J.-H.;
Fujio, M.; Tsuno, Y. J. Phys. Org. Chem. 2001, 14, 123–130. (b) Yoh,
S. D.; Lee, M.-K.; Son, K.-J.; Cheong, D.-Y.; Han, I.-S.; Shim, K.-T.
Bull. Korean Chem. Soc. 1999, 20, 466–468. (c) Amyes, T. L.;
Richard, J. P. J. Am. Chem. Soc. 1990, 112, 9507–9512.
(15) (a) Katritzky, A. R.; Brycki, B. E. Chem. Soc. ReV. 1990, 19, 83–
105. (b) Katritzky, A. R.; Brycki, B. E. J. Phys. Org. Chem. 1988, 1,
1–20. (c) Katritzky, A. R.; Musumarra, G. Chem. Soc. ReV. 1984, 13,
47–68. (d) Katritzky, A. R.; Sakizadeh, K.; Gabrielsen, B.; le Noble,
W. J. J. Am. Chem. Soc. 1984, 106, 1879–1880. (e) Katritzky, A. R.;
Musumarra, G.; Sakizadeh, K. J. Org. Chem. 1981, 46, 3831–3835.
(16) (a) Jencks, W. P. Acc. Chem. Res. 1980, 13, 161–169. (b) Jencks,
W. P. Chem. Soc. ReV. 1981, 10, 345–375. (c) Richard, J. P.; Jencks,
W. P. J. Am. Chem. Soc. 1984, 106, 1373–1383. (d) Richard, J. P.;
Jencks, W. P. J. Am. Chem. Soc. 1984, 106, 1383–1396.
9
J. AM. CHEM. SOC. VOL. 131, NO. 32, 2009 11393

A R T I C L E S
Phan et al.
treated with a high excess of amines (>10 equiv), the amine
concentrations remained almost constant during the reactions,
and the increase of conductivity followed the exponential
function 4, as illustrated in Figure 1.
dG/dt ) Gmax[1 - exp(-kobst)] + C
(4)
Plots of kobs versus the concentrations of the amines were
linear (Figure 2) but did not go through the origin. As expressed
by eq 5, the observed rate constants kobs (Supporting Information,
pp S16-S29) can be regarded as the sum of an amine-
independent term k1 and an amine-dependent term k2[amine],
which are collected in Table 1.
Figure 1. Exponential increase of conductivity during the reaction of
4-methylbenzhydryl bromide 1-Me,H with 0.2 M piperidine in DMSO.
Calibration in the inset: conductivity at t∞ is proportional to the concentration
of substrate.
kobs ) k1 + k2[amine]
(5)
conductivity had reached a constant value (typically 300 s), another
portion of benzhydryl bromide was added. As depicted in the inset
of Figure 1, the conductivity increased linearly with the concentra-
tion of released HBr, even at higher concentrations than used for
the kinetic experiments.
The second-order rate constants k2 can easily be assigned to
the SN2 reactions of the amines with the benzhydryl bromides.
The amine-independent term k1, which equals the directly
measured solvolysis rate constant in DMSO in the absence of
a nucleophilic amine, reflects either the rate of the SN1-type
process (k1, Scheme 1) or the rate of an SN2 reaction with DMSO
as the nucleophile (k1′, Scheme 1), or a combination of both
processes.
Comparison of the rate constants k1 in the first line of Table
1 with previously published solvolysis rates in alcohols21 shows
that the solvolysis rates in DMSO are comparable to those in
pure ethanol but considerably smaller than those in ethanol-water
mixtures and in 2,2,2-trifluoroethanol. Creary et al. published
kinetic data for the solvolysis of adamantyl mesylates and 3-aryl-
3-hydroxy-ꢀ-lactams in DMSO.22 In agreement with these data,
our measurements confirm that DMSO is a solvent with a
relatively high ionizing power.
The horizontal lines in Table 1 show that variation of the
substituents of the benzhydryl bromides from 1-CH3,H to
1-CF3,CF3 affects the second-order rate constants k2 for the
reactions with amines by less than a factor of 10. Accordingly,
plots of log k2 versus Σσ (Figure 3) or any other of Hammett’s
substituent constants (e.g., σ+) illustrate that variation of the
para-substituents in the benzhydryl bromides has only a marginal
effect on the rate constants of the SN2 reactions, indicating
transition states 7, where only little positive charge is developed
at the benzhydryl center.
Photometric Measurements of the Reactions of the Benzhydry-
lium Tetrafluoroborates with Amines. The rates of the reactions
of benzhydrylium tetrafluoroborates with amines were studied in
DMSO solutions. All amines were used as free bases. As the
reactions of the colored benzhydrylium ions with amines gave rise
to colorless products, the reactions could be followed by employing
UV-vis spectroscopy at the absorption maxima of the benzhydry-
lium ions (Supporting Information Table S1). The rates were
determined by using a Hi-Tech SF-61DX2 stopped-flow spectro-
photometer system (controlled by Hi-Tech KinetAsyst2 software).
Amine concentrations at least 10 times higher than the benzhy-
drylium ion concentrations were usually employed, resulting in
pseudo-first-order kinetics with an exponential decay of the
concentrations of the benzhydrylium ions Ar2CH+. First-order rate
constants kobs (s-1) were obtained by least-squares fitting of the
absorbance data (averaged from at least five kinetic runs at each
amine concentration) to the single-exponential eq 3.
dA/dt ) A0exp(-kobst) + C
(3)
Laser-Flash Photolysis. Laser-flash photolysis was employed
for determining the rates of the reactions of Ar2CH+ with DMSO
in acetonitrile. For that purpose, benzhydrylium ions were generated
by irradiation of Ar2CH-Cl in DMSO/acetonitrile with an Innolas
SpitLight 600 Nd:YAG laser (fourth harmonic at λ ) 266 nm;
power/pulse of 40-60 mJ, pulse length ) 6.5 ns) in a quartz cell.
The rate constants were determined by observing the time-dependent
decay of the UV-vis absorptions of the benzhydrylium ions. The
pseudo-first-order rate constants were obtained by fitting the decay
of the UV-vis absorptions to the exponential function of eq 3.
Product studies were carried out for several representative sys-
tems. For that purpose, 0.2 M solutions of the amines in DMSO
were combined with 0.1 equiv of benzhydryl bromide 1-X,Y. After
24 h, the reaction mixtures were quenched with water and extracted
with diethyl ether. After evaporation, the residue was diluted with
acetone containing a defined amount of n-hexadecane (internal
standard, ≈10-3 M). Aliquots of the solutions were analyzed with
a Thermo Focus GC equipped with a FID detector (column:
Macherey-Nagel Optima-17, 25 m, 0.25 mm i.d.; carrier gas: N2)
for the determination of the absolute yields. In addition, GC-MS
analysis (Agilent 6890 GC with an Agilent 5973 MS detector) was
used for identifying the individual peaks. For the calculation of
the absolute product concentrations, the products were synthesized
individually, and GC calibrations were carried out to obtain the
relative molar response factors (RMR).
Because of the small dependence of the second-order rate
constants on the nature of the substituents, the poor correlations
in Figure 3 are not surprising, particularly because substituent
effects in diarylmethyl compounds have been reported not to
be additive.23 The poor correlation in Figure 3 is also in line
(17) (a) Richard, J. P. In AdVances in Carbocation Chemistry; Creary, X.,
Ed.; JAI Press: Greenwich, CT, 1989; Vol. 1, pp 121-169. (b) Amyes,
T. L.; Toteva, M. M.; Richard, J. P. In ReactiVe Intermediate
Chemistry; Moss, R. A., Platz, M. S., Jones, M., Jr., Eds; Wiley-
Interscience: Hoboken, NJ, 2004; pp 41-69. (c) Richard, J. P.; Amyes,
T. L.; Toteva, M. M.; Tsuji, Y. AdV. Phys. Org. Chem. 2004, 39, 1–
26.
(18) (a) Database of reactivity parameters E, N, and s: http://www.cup.uni-
muenchen.de/oc/mayr. (b) Mayr, H.; Bug, T.; Gotta, M. F.; Hering,
N.; Irrgang, B.; Janker, B.; Kempf, B.; Loos, R.; Ofial, A. R.;
Remennikov, G.; Schimmel, H. J. Am. Chem. Soc. 2001, 123, 9500–
9512.
Results and Discussion
Kinetics of the Nucleophilic Substitutions in DMSO. When
solutions of the benzhydryl bromides 1-X,Y in DMSO were
9
11394 J. AM. CHEM. SOC. VOL. 131, NO. 32, 2009

Can One Predict Changes from SN1 to SN2 Mechanisms?
A R T I C L E S
Figure 2. Plots of kobs (s-1) of the reactions of different benzhydryl bromides with amines in DMSO vs the concentrations of the amines (mol L-1) (note
the different calibration of the various y-axes).
with previous findings by Baker,24 Jencks,25 and Bordwell26
that nucleophilic substitutions at substituted benzyl halides do
not follow simple Hammett correlations.
ing a continuous change of the transition state from very tight
for acceptor-substituted systems to loose transition states with
more positive charge on the benzylic carbon for the SN2
reactions of the p-methoxy-substituted systems.12
On the other hand, the SN2 reactions of substituted arylethyl
bromides show a continuous increase of the F value as the
electron-donating ability of the substituents is increased, indicat-
(21) (a) Denegri, B.; Ofial, A. R.; Juric, S.; Streiter, A.; Kronja, O.; Mayr,
H. Chem.sEur. J. 2006, 12, 1657-1666. (b) Denegri, B.; Streiter,
A.; Juric, S.; Ofial, A. R.; Kronja, O.; Mayr, H. Chem.sEur. J. 2006,
12, 1648–1656; Chem.sEur. J. 2006, 12, 5415.
(22) Creary, X.; Burtch, E. A. J. Org. Chem. 2004, 69, 1227–1234.
(23) Uddin, K.; Fujio, M.; Kim, H.-J.; Rappoport, Z.; Tsuno, Y. Bull. Chem.
Soc. Jpn. 2002, 75, 1371–1379.
(19) (a) Mayr, H.; Ofial, A. R. Pure Appl. Chem. 2005, 77, 1807–1821.
(b) Mayr, H.; Ofial, A. R. J. Phys. Org. Chem. 2008, 21, 584–595.
(c) Mayr, H.; Patz, M. Angew. Chem. 1994, 106, 990-1010; Angew.
Chem., Int. Ed. Engl. 1994, 33, 938-957.
(20) Mayr, H.; Kempf, B.; Ofial, A. R. Acc. Chem. Res. 2003, 36, 66–77.
9
J. AM. CHEM. SOC. VOL. 131, NO. 32, 2009 11395

A R T I C L E S
Phan et al.
Table 1. Rate Constants (at 20° C) for the Solvolysis Reactions of the Benzhydryl Bromides 1-X,Y in DMSO (s-1) and for Their Reactions
with Amines in DMSO (M-1 s-1
)
nucleophiles
1-Me,H
1-H,H
1-Cl,Cl
1-CF3,H
1-CF3,CF3
DMSO (k1)
DABCO (k2)
piperidine (k2)
morpholine (k2)
ethanolamine (k2)
1-aminopropan-2-ol (k2)
n-PrNH2 (k2)
benzylamine (k2)
diethanolamine (k2)
2-aminobutan-1-ol (k2)
6.71 × 10-3
1.92 × 10-1
3.57 × 10-2
2.16 × 10-2
a
5.45 × 10-4
5.45 × 10-2
1.69 × 10-2
7.30 × 10-3
1.54 × 10-3
1.45 × 10-3
1.33 × 10-3
6.76 × 10-4
6.37 × 10-4
b
1.36 × 10-4
a
1.25 × 10-5
a
2.76 × 10-6
a
2.33 × 10-2
9.51 × 10-3
2.37 × 10-3
2.23 × 10-3
2.19 × 10-3
1.35 × 10-3
7.46 × 10-4
3.13 × 10-4
9.36 × 10-3
3.29 × 10-3
1.13 × 10-3
8.92 × 10-4
1.13 × 10-3
6.30 × 10-4
2.55 × 10-4
1.77 × 10-4
6.66 × 10-3
2.17 × 10-3
1.25 × 10-3
7.96 × 10-4
1.17 × 10-3
5.50 × 10-4
1.19 × 10-4
a
4.93 × 10-3
3.98 × 10-3
1.90 × 10-3
a
b
a Not determined. b The kobs was independent of the amine concentration (see Figure 2).
Table 2. Products of the Reactions of the Benzhydryl Bromides
1-X,Y (c ) 0.02 mol L-1) with Amines (10 equiv) in DMSO (20 °C)
1-X,Y
amine
[5-X,Y]/% [6-X,Y]/% [4-X,Y]/% [4-X,Y]/([5-X,Y] + [6-X,Y])
1-CF3,CF3 piperidine
morpholine
propylamine
1-H,CF3 piperidine
morpholine
0
0
0
0
0
0
1.5
2.6
0
0
0
0
0
0
0.6
2.0
only
only
only
only
only
only
77.0
50.5
propylamine
1-Cl,Cl
1-H,H
piperidine
morpholine
36.7
11.0
propylamine alcohol cannot be separated from amine by GC
piperidine
morpholine
7.5
16.1
0.6
2.7
16.6
35.4a
49.8a
80.5a
70.0
70.0
48.4
41.8
27.1
14.6
8.6
3.7
1.1
1.2
0.5
0.2
propylamine 26.7
piperidine
1-Me,H
morpholine
propylamine
a As the ketone 5-Me,H and the alcohol 6-Me,H could not be
separated on the GC (see text), the yield refers to the sum of both
compounds.
Figure 3. Plot of log k2 of the reactions of the benzhydryl bromides with
amines vs Hammett’s substituent constants σ (from ref 27).
the benzhydryl bromides. The bis-trifluoromethyl-substituted
compound 1-CF3,CF3 deviates from this correlation, however,
and reacts approximately 40 times faster than extrapolated from
the linear log k1 versus σ+ correlation; we will discuss later
that this deviation may be due to an SN2-type reaction of
1-CF3,CF3 with DMSO.
The 4,4′-dimethyl-substituted benzhydryl bromide 1-Me,Me
reacted so fast that analogous experiments, as described in
Figure 2, could not be performed. From the Hammett correlation
given in Figure 4, one can extrapolate a first-order solvolysis
rate constant of 0.045 s-1 for the 4,4′-dimethyl-substituted
benzhydryl bromide (1-Me,Me).
Reaction Products. As summarized in Table 2, the reactions
of benzhydryl bromides 1-X,Y with 0.2 M amines in DMSO
give the benzhydryl amines 4-X,Y, accompanied by the ben-
zophenones 5-X,Y and the diarylmethanols 6-X,Y.
The exclusive formation of the benzhydryl amines 4-X,Y in
the reaction of 1-CF3,CF3 and 1-CF3,H with morpholine,
piperidine, and n-propylamine is in line with the kinetics
Figure 4. Plot of log k1 for the solvolysis reactions of the benzhydryl
bromides in DMSO vs Hammett’s substituent constants σ+ (σ+ from ref
27; k1 for 1-CF3,CF3 not used for the correlation; see text).
(24) Hammett plots of the second-order rate constants of the reactions of
substituted benzyl bromides with pyridine (20 °C, in acetone) are
shown in the Supporting Information: (a) Baker, J. W. J. Chem. Soc.
1936, 1448–1451. (b) Baker, J. W. Trans. Faraday Soc. 1941, 37,
632–644.
In contrast to the behavior of the second-order rate constants
in Figure 3, the first-order rate constants k1 (first line of Table
1) strongly depend on the para-substituents. From the plot of
log k1 versus Σσ+, one derives a Hammett reaction constant of
F ) -2.94 (Figure 4).
The magnitude of the reaction constant F suggests that the
amine-independent term k1 corresponds to the ionization step
of an SN1 reaction and not to an SN2-type attack of DMSO at
(25) Young, P. R.; Jencks, W. P. J. Am. Chem. Soc. 1979, 101, 3288–
3294.
(26) (a) Bordwell, F. G.; Hughes, D. L. J. Org. Chem. 1980, 45, 3320–
3325. (b) Bordwell, F. G.; Hughes, D. L. J. Am. Chem. Soc. 1986,
108, 7300–7309.
(27) Hansch, C.; Leo, A.; Taft, R. W. Chem. ReV. 1991, 91, 165–195.
9
11396 J. AM. CHEM. SOC. VOL. 131, NO. 32, 2009

Can One Predict Changes from SN1 to SN2 Mechanisms?
A R T I C L E S
Scheme 2
Figure 5. Plot of the ratios of [Ph2CdO]/[Ph2CH-OH] vs time for the
reaction of benzhydryl bromide 1-H,H with 0.2 M 2,6-lutidine in DMSO
after aqueous workup. Square indicates the workup with methanol (ratio
equals [Ph2CdO]/([Ph2CH-OH] + [Ph2CH-OMe])).
described in Figure 2: The amine-independent terms are
negligible in comparison to the amine-dependent terms. There-
fore, at amine concentrations of 0.2 M, SN2 reactions with
amines take place exclusively.
ions from alkyl halides22 and their subsequent reactions with
bases has been studied NMR spectroscopically by other
groups.29,30
Analogously, the predominant formation of amine 4-Cl,Cl
by reaction of 1-Cl,Cl with piperidine can be explained by the
high SN2 reactivity of the amine at a concentration of 0.2 M. In
the reactions with the less nucleophilic morpholine, the amounts
of benzhydrol 6-Cl,Cl and benzophenone 5-Cl,Cl rise. Unfor-
tunately, it was not possible to measure the product ratio
obtained by the reaction of 1-Cl,Cl with n-propylamine because
the GC signals of the benzhydrol and the amine overlapped.
In the reactions of 1-H,H with these amines, considerable
amounts of benzophenone 5-H,H and benzhydrol 6-H,H were
generated along with the benzhydryl amines 4-H,H, and their
quantities increase with decreasing nucleophilicities of the
amines.
In the reaction of the monomethyl-substituted benzhydryl
bromide 1-Me,H, an even larger amount of diarylmethanol
6-Me,H and benzophenone 5-Me,H was found, while less of
the benzhydryl amine 4-Me,H was formed. As with the other
substrates, the yield of the amine 4-Me,H decreased in the series
piperidine > morpholine > propylamine. While it was not
possible to distinguish between diarylmethanol 6-Me,H and
benzophenone 5-Me,H by our GC analysis because both
compounds had the same retention times, the GC-MS spectra
showed that benzophenones 5-Me,H are the major products.
Because the relative molar response (RMR) constant was nearly
the same for the benzophenone 5-Me,H and the benzhydrol
6-Me,H (experiment with pure compounds), their sum could
be determined.
As illustrated in Scheme 1, the benzophenones 5-X,Y as well
as the benzhydrols 6-X,Y are formed through the intermediacy
of the oxysulfonium ions 3-X,Y. In accordance with previous
reports on the mechanism of the Kornblum oxidation,28 we
assume that deprotonation of the oxysulfonium ion 3-X,Y at a
methyl group yields a sulfur ylide, which undergoes a proton
shift and cleavage of the O-S bond to yield the benzophenone
5-X,Y (Scheme 2). In line with this mechanism, benzhydrol
6-H,H was not oxidized when treated with equimolar amounts
of 2,6-lutidine and 2,6-lutidine hydrobromide under the condi-
tions of the solvolysis reactions. The formation of oxysulfonium
Benzhydrol 6-H,H and benzophenone 5-H,H were formed
exclusively when benzhydryl bromide 1-H,H (0.02 M) was
dissolved in a 0.2 M solution of the weakly nucleophilic 2,6-
lutidine in DMSO. Figure 5 illustrates that the ratio [5-H,H]/
[6-H,H] obtained after aqueous workup of the solvolysis
products from 1-H,H increases with reaction time. From the
observation that the increase of this ratio continues after
complete consumption of 1-H,H, one can derive that a precursor
of 5-H,H (e.g., the oxysulfonium ion 3-H,H) accumulates in
the reaction mixture before it is slowly converted into ben-
zophenone 5-H,H.
When the mixture obtained from 1-H,H (0.02 M) and 0.2 M
2,6-lutidine in DMSO was worked up with methanol, the
benzophenone 5-H,H was accompanied by benzhydryl methyl
ether, which may be formed by nucleophilic attack of methanol
at the oxysulfonium ion 3. Nucleophilic attack of impurities of
water, amine, or methanol at the sulfur atom of 3-H,H may
account for the small amount of benzhydrol 6-H,H (6.7%)
obtained under these conditions. The ratio [Ph2CdO]/
([Ph2CH-OH] + [Ph2CH-OMe]) was similar to the ratio
[Ph2CdO]/[[Ph2CH-OH] observed after aqueous workup at
comparable reaction times (0 in Figure 5).
Differentiation of SN1 and SN2 Processes. For each of the
products (4-6)-X,Y drawn in Scheme 1, formation through
an SN1 (k1) or SN2 (k2 and k1′) process has to be considered.
With the data presented so far, it is possible to eliminate
some of these reaction pathways. If the benzhydryl amines
4-X,Y would be formed by an SN1 reaction via the carbenium
ions 2-X,Y, which are subsequently trapped by the amines,
an amine-independent rate law would result because the
formation of the benzhydryl cations 2-X,Y would be rate-
determining. Pathway kN of Scheme 1 can, therefore, be
eliminated. This conclusion is confirmed by the comparison
of the kinetic data with the product ratios in Table 3. For
the reactions of different benzhydryl bromides 1-X,Y with
piperidine, morpholine, and propylamine, the product ratios
[4]/([5] + [6]) divided by the concentrations of the amines,
[4]/([amine]([5] + [6])), are almost equal to the ratios k2/k1.
(28) (a) Kornblum, N.; Jones, W. J.; Anderson, G. J. J. Am. Chem. Soc.
1959, 81, 4113–4114. (b) Kornblum, N.; Powers, J. W.; Anderson,
G. J.; Jones, W. J.; Larson, H. O.; Levand, O.; Weaver, W. M. J. Am.
Chem. Soc. 1957, 79, 6562–6562. (c) Dave, P.; Byun, H.-S.; Engel,
R. Synth. Commun. 1986, 16, 1343–1346.
(29) (a) Torssell, K. Tetrahedron Lett. 1966, 7, 4445–4451. (b) Torssell,
K. Acta Chem. Scand. 1967, 21, 1–14.
(30) Creary, X.; Burtch, E. A.; Jiang, Z. J. Org. Chem. 2003, 68, 1117–
1127.
9
J. AM. CHEM. SOC. VOL. 131, NO. 32, 2009 11397

A R T I C L E S
Phan et al.
Table 3. Comparison of Rate Constant Ratios with the Product
Ratios for the Reactions of Benzhydryl Bromides 1-X,Y with 0.2 M
Piperidine, Morpholine, and Propylamine in DMSO at 20 °C
Table 5. Eyring and Arrhenius Activation Parameters for the
Reaction of the Benzhydryl Bromide 1-Cl,Cl with Morpholine in
DMSO
1-Me,H
1-H,H
1-Cl,Cl
1-CF3,H
for k1
for k2
k1/s-1
6.71 × 10-3 5.45 × 10-4 1.36 × 10-4 1.25 × 10-5
3.57 × 10-2 1.69 × 10-2 2.33 × 10-2 9.36 × 10-3
∆Hq/kJ mol-1
∆Sq/J mol-1 K-1
Ea/kJ mol-1
ln A
59.4 ( 6.1
-116.7 ( 19.8
62.0 ( 6.1
36.5 ( 0.6
-159.0 ( 1.8
39.0 ( 0.5
reaction with piperidine (0.2 M)
k2/M-1 s-1
k2/k1
5.3
5.9
31
43
1.7 × 102
1.8 × 102
7.5 × 102
16.4 ( 2.4
11.4 ( 0.2
[4]/0.2([5] + [6])
only amine 4
reaction with morpholine (0.2 M)
k2/M-1 s-1
k2/k1
2.16 × 10-2 7.30 × 10-3 9.51 × 10-3 3.29 × 10-3
Table 6. Second-Order Rate Constants for the Reactions of
Amino-Substituted Benzhydrylium Ions with Amines in DMSO at
20 °C
3.2
2.7
13
19
70
55
2.6 × 102
[4]/0.2([5] + [6])
only amine 4
reaction with n-propylamine (0.2 M)
k2/M-1 s-1
k2/k1
3.98 × 10-3 1.33 × 10-3 2.19 × 10-3 1.13 × 10-3
amine
N/s
Ar2CH+
kN/M-1 s-1
0.59
0.91
2.4
5.6
16
-
90
2-aminobutan-1-ol
14.39/0.67
(ind)2CH+
(jul)2CH+
(lil)2CH+
(thq)2CH+
(jul)2CH+
(lil)2CH+
(ind)2CH+
(jul)2CH+
(lil)2CH+
(ind)2CH+
(jul)2CH+
(lil)2CH+
(ind)2CH+
(jul)2CH+
(lil)2CH+
6.08 × 103
2.23 × 103
8.33 × 102
3.91 × 104
6.60 × 103
2.51 × 103
2.27 × 104
9.31 × 103
3.29 × 103
4.83 × 104
1.74 × 104
6.19 × 103
2.87 × 104
1.19 × 104
4.71 × 103
[4]/0.2([5] + [6])
only amine 4
benzylamine
15.28/0.65
15.47/0.65
15.51/0.70
16.07/0.61
Table 4. Comparison of the Rate Constants and Product Ratios
for the Reaction of Benzhydryl Bromide 1-Cl,Cl (0.02 M) with
Morpholine (0.2 M) at Different Temperatures in DMSO
1-aminopropan-2-ol
diethanolamine
ethanolamine
T/°C
20
35
50
k1/s-1
1.36 × 10-4
9.51 × 10-3
1.01 × 10-2
5.26 × 10-4
3.95 × 10-4
55
3.85 × 10-4
2.10 × 10-2
1.07 × 10-2
1.27 × 10-3
2.26 × 10-4
36
1.45 × 10-3
4.20 × 10-2
8.52 × 10-3
1.83 × 10-3
2.33 × 10-4
21
k2/M-1 s-1
[4-Cl,Cl]/M
[5-Cl,Cl]/M
[6-Cl,Cl]/M
[4]/0.2([5]+[6])
k2/k1
70
55
29
If the amines 4-X,Y would be formed via the pathway kN in
addition to the SN2 pathway k2, a higher percentage of the
amines [4-X,Y] would be expected. We will demonstrate later
that the trapping of the benzhydrylium ions 2-X,Y by the
solvent DMSO is so fast that the pathway kN cannot compete
with ksolv at amine concentrations of 0.2 M.
Eyring and Arrhenius plots of high quality (r2 ) 0.9998,
Supporting Information) were obtained for the second-order rate
constants k2, from which the activation parameters listed in Table
5 were obtained. The highly negative activation entropy (-159
J mol-1 K-1) is in agreement with previous reports on alkylations
of amines.31
Formal kinetics do not allow differentiation between
pathways k1 and k1′ for the formation of 5-X,Y and 6-X,Y;
that is, the oxysulfonium ion 3-X,Y may be formed via either
an SN1 (k1) or an SN2 process (k1′) with the solvent DMSO.
The linear Hammett plot for log k1 (i.e., kobs in pure DMSO)
with a slope of -2.94 (Figure 4) indicates the operation of
the SN1 pathway for most systems. If the SN2 pathway
indicated by k1′ would be operating, a similar reactivity
pattern as shown in Figure 3 would be expected for the
different benzhydryl bromides. The significant deviation of
1-CF3,CF3 from the linear Hammett correlation in Figure 4
may be indicative of an SN2 participation in the reaction of
this acceptor-substituted benzhydryl bromide with DMSO
(k1′, nucleophilic solvent participation). Further support for
this interpretation will be given below.
Temperature Effect on Rate Constants and Product Ratios.
When the kinetics of the reaction of 1-Cl,Cl with morpholine
in DMSO were studied at variable temperature and evaluated
as described above, the rate constants summarized in Table 4
were obtained. Raising the temperature from 20 to 50 °C
increased the first-order rate constant k1 by a factor of 11, while
the second-order rate constant k2 increased only by a factor of
4. In accordance with the previous discussion, the stronger
increase of k1 compared with k2 resulted in a decrease of the
yield of the amine 4-X,Y (Table 4). The ratios of the products
[4]/([amine]([5] + [6])) and the ratios of the rate constants k2/
k1 again agreed within experimental error (Table 4), indicating
that also, at elevated temperatures, amines 4 are produced
through the SN2 pathway (k2) while 5 and 6 are formed via the
SN1 route (k1).
Because of the small contribution of the first-order term k1
to the overall rate constant, the rate constants k1 are less precise,
and the resulting Eyring and Arrhenius plots are of lower quality
(r2 ) 0.990). The calculated activation entropy (-117 J mol-1
K-1) is slightly more negative than typically observed for SN1
reactions in alcoholic and aqueous solutions.32
Nucleophilicity Parameters N and s for Amines in DMSO.
While N and s parameters for numerous amines have
previously been determined in aqueous33 and in methanolic
solution,34 only few amines have so far been characterized
in DMSO.35 Because amine nucleophilicities in DMSO will
be needed for the mechanistic analysis below, we have now
determined N and s values for the amines which were used
in this investigation in DMSO. For that purpose, we have
measured the rates of reactions of amino-substituted ben-
zhydrylium ions with amines in DMSO (eqs 6 and 7) under
pseudo-first-order conditions (excess of amine) using the
photometric method described previously.18-20 Comparison
of the rate constants listed in Table 6 with those reported in
water33 shows that the amines react roughly 100 times faster
(31) (a) Arnett, E. M.; Reich, R. J. Am. Chem. Soc. 1980, 102, 5892–
5902. (b) Duty, R. C.; Gurne, R. L. J. Org. Chem. 1970, 35, 1800–
1802.
(32) (a) Cowie, G. R.; Fitches, H. J. M.; Kohnstam, G. J. Chem. Soc. 1963,
1585–1593. (b) Fox, J. R.; Kohnstam, G. J. Chem. Soc. 1963, 1593–
1598.
(33) Brotzel, F.; Chu, Y. C.; Mayr, H. J. Org. Chem. 2007, 72, 3679–
3688.
(34) Phan, T. B.; Breugst, M.; Mayr, H. Angew. Chem. 2006, 118, 3954-
3959; Angew. Chem., Int. Ed. 2006, 45, 3869-3874.
(35) Minegishi, S.; Mayr, H. J. Am. Chem. Soc. 2003, 125, 286–295.
9
11398 J. AM. CHEM. SOC. VOL. 131, NO. 32, 2009

Can One Predict Changes from SN1 to SN2 Mechanisms?
A R T I C L E S
Scheme 3. Laser Flash Photolytic Generation of Benzhydrylium
Ions in MeCN/DMSO Mixtures
Figure 6. Plot of the rate constants kN for the reactions of amines with
benzhydrylium ions (DMSO, 20 °C) vs their electrophilicity parameters E
(E ) -10.04 for (lil)2CH+, -9.45 for (jul)2CH+, -8.76 for (ind)2CH+,
and -8.22 for (thq)2CH+; from ref 18).
in DMSO than in water due to the weaker solvation in the
nonprotic solvent.
Figure 7. Plot of kobs of the reactions of the benzhydrylium ions 2-X,Y
with DMSO in MeCN vs [DMSO].
Table 7. Second-Order Rate Constants for the Reactions of
Benzhydrylium Ions (2-X,Y) with DMSO (O-Attack) in Acetonitrile
2-X,Y
Ea
k2/M-1 s-1
2-MeO,MeO
2-MeO,PhO
2-MeO,H
2-PhO,Me
2-PhO,H
2-Me,Me
2-Me,H
2-F,H
2-H,H
2-Cl,Cl
0.00
0.61
2.11
2.16
2.90
3.63
4.59
5.60
5.90
6.02
1.69 × 107
5.00 × 107
6.13 × 108
7.04 × 108
8.60 × 108
2.63 × 109
3.50 × 109
4.78 × 109
3.34 × 109
4.79 × 109
Plots (Figure 6) of the second-order rate constants given
in Table 6 versus the electrophilicity parameters E of the
benzhydrylium ions were linear as required by eq 2, and
yielded the N and s parameters for amines in DMSO which
are given in Table 6.
a Empirical electrophilicity parameter from ref 18.
Solvent Nucleophilicity of DMSO. Dimethyl sulfoxide may
react with electrophiles either at sulfur or at oxygen.36 The
formation of benzophenones and benzhydrols reported above
indicates that the benzhydrylium ions employed in this work
react at oxygen to yield the oxysulfonium ions 3 (Scheme 3).
The rates of these reactions were determined by laser-flash
photolysis of solutions of benzhydryl chlorides in MeCN/DMSO
mixtures and UV-vis spectrometric monitoring of the decay
of the resulting benzhydrylium ions in the presence of variable
concentrations of DMSO. Plots of the observed rate constants
versus the concentrations of DMSO (Figure 7) give rise to the
second-order rate constants of the reactions (Table 7).
Figure 8 shows that the rate constants (Table 7) for the
reactions of DMSO with benzhydrylium ions increase with the
Figure 8. Plot of log k (second-order rate constants, M-1 s-1) for the
reactions of DMSO with the benzhydrylium ions 2-X,Y in MeCN at 20 °C
vs their electrophilicity parameters E.
electrophilicity parameters of the benzhydrylium ions and
become diffusion-controlled at E > 4. For that reason, all
benzhydrylium ions 2-X,Y generated in DMSO from benzhydryl
bromides 1-X,Y of Table 1 are immediately trapped by the
solvent DMSO. It can thus be explained that trapping of 2-X,Y
by amines (kN, Scheme 1) does not occur despite the higher
(36) (a) Ho, T.-L. Hard and Soft Acids and Bases Principle in Organic
Chemistry; Academic Press: New York, 1977. (b) Smith, S. G.;
Winstein, S. Tetrahedron 1958, 3, 317–318. (c) Rasul, G.; Prakash,
G. K. S.; Olah, G. A. J. Org. Chem. 2000, 65, 8786–8789. (d) da
Silva, R. R.; Santos, J. M.; Ramalho, T. C.; Figueroa-Villar, J. D. J.
Braz. Chem. Soc. 2006, 17, 223–226.
9
J. AM. CHEM. SOC. VOL. 131, NO. 32, 2009 11399

A R T I C L E S
Phan et al.
Table 8. Calculated Lifetimes τ (1/k, s) of Benzhydrylium Ions in DMSO and in 1 M Solutions of Various Amines in DMSO
2-Me,H
(E ) 4.59)
2-H,H
(E ) 5.90)
2-Cl,Cl
(E ) 6.02)
2-CF3,H
(E ∼ 9.7)a
2-CF3,CF3
(E ∼ 13.6)a
nucleophiles (N/s)
DMSO (N1 ) 11.3/0.74)b
2-aminobutan-1-ol (14.39/0.67)c
benzylamine (15.28/0.65)c
1-aminopropan-2-ol (15.46/0.65)c
diethanolamine (15.51/0.70)c
n-propylamine (15.70/0.64)d
ethanolamine (16.07/0.61)c
morpholine (16.96/0.67)d
piperidine (17.19/0.71)d
2 × 10-12
2 × 10-13
1 × 10-13
9 × 10-14
9 × 10-15
1 × 10-13
2 × 10-13
4 × 10-15
3 × 10-16
4 × 10-17
2 × 10-13
3 × 10-14
2 × 10-14
1 × 10-14
1 × 10-15
1 × 10-14
4 × 10-14
5 × 10-16
4 × 10-17
5 × 10-18
2 × 10-13
2 × 10-14
1 × 10-14
1 × 10-14
8 × 10-16
1 × 10-14
3 × 10-14
4 × 10-16
3 × 10-17
4 × 10-18
3 × 10-16
7 × 10-17
6 × 10-17
4 × 10-17
2 × 10-18
6 × 10-17
2 × 10-16
1 × 10-18
8 × 10-20
1 × 10-20
4 × 10-19
2 × 10-19
2 × 10-19
1 × 10-19
4 × 10-21
2 × 10-19
8 × 10-19
3 × 10-21
1 × 10-22
2 × 10-23
DABCO (18.80/0.70)e
a Extrapolated from σ constants (from ref 27) using the correlation σ+ ) 0.134E - 0.767 reported in ref 18b. b N1 for the calculation of first-order
rate constants with DMSO in DMSO (see text). c From Table 6. d N and s parameters were taken from ref 35. e N and s parameters in acetonitrile from
ref 40.
nucleophilicity of the amines. From the linear left part of Figure
8 (k2 < 8 × 108 M-1 s-1), one can derive the nucleophilicity
parameters N ) 9.75 and s ) 0.742 for the O-reactivity of
DMSO, showing that DMSO is considerably more nucleophilic
than water and ordinary alcohols.37,38 With the assumption that
the change of solvent polarity in MeCN/DMSO mixtures of
different compositions does not affect the rate constants
significantly, one can multiply the rate constants in Table 7 with
14.1 mol L-1, that is, the concentration of DMSO in 100%
DMSO to obtain the first-order rate constants of the decay in
100% DMSO. From the plot of the first-order rate constants
versus E, one derives the solvent nucleophilicity N1 ) 11.3 for
DMSO (N1 ) N + (log 14.1)/s).39
Calculation of Hypothetical Lifetimes of Benzhydrylium
Ions in DMSO Solution in the Presence of Amines. The N and
s parameters of the amines (Table 6 and refs 33, 35, and 40)
and N1 of DMSO (Figure 8) can now be combined with the
electrophilicity parameters E of the benzhydrylium ions to
calculate rate constants for the reactions of benzhydrylium ions
with these nucleophiles by eq 2. Many of the resulting rate
constants exceed the diffusion limit. In these cases, the values
1/k (s), which are listed in Table 8, have to be considered as
hypothetical lifetimes.
Only at morpholine concentrations > 0.3 M will the SN2 process
override the SN1 process. If ionization occurs in the absence of
a morpholine molecule (SN1), the intermediate p-methyl-
substituted benzhydrylium ion (2-Me,H) is rapidly trapped by
the solvent DMSO (lifetime ≈ 2 × 10-12 s), and the diffusion-
controlled reaction with morpholine cannot compete.
Piperidine (k2 ≈ 5k1) and DABCO (k2 ≈ 28k1) are stronger
nucleophiles and, therefore, provide a stronger nucleophilic
assistance for breaking the C-Br bond of 1-Me,H. As shown
in Figure 2, now the SN2 process overrides the SN1 process
already at low amine concentrations, and the calculated lifetimes
of 3 × 10-16 and 4 × 10-17 s are in line with Jencks’ enforced
concerted mechanism. Lifetimes τ > 10-14 s are calculated for
the p-methylbenzhydrylium ion 2-Me,H in 1 M solutions of the
other amines, and Figure 2 shows that, in the reactions with
benzylamine, 1-aminopropan-2-ol, and n-propylamine, the SN1
mechanism generally dominates.
For the unsubstituted benzhydrylium ion 2-H,H, 9-times
shorter lifetimes are calculated; as a consequence, the SN2
reactions gain more weight. Morpholine (k2 ) 13k1), piperidine
(k2 ) 31k1), and DABCO (k2 ) 100k1) prefer the SN2
mechanism already at low amine concentrations (> 0.08-0.01
M), in accord with calculated lifetimes of τ < 10-15 s. No SN2
contribution was found for the reaction of 1-H,H with 2-ami-
nobutan-1-ol (τ ) 3 × 10-14 s). For the reactions of 1-H,H
with diethanolamine (τ ) 1 × 10-15 s), ethanolamine (τ ) 4 ×
10-14 s), benzylamine (τ ) 2 × 10-14 s), 1-aminopropan-2-ol
(τ ) 1 × 10-14 s), and n-propylamine (τ ) 1 × 10-14 s),
lifetimes similar to the vibrational limit are calculated, and the
SN2 reactions overrated the SN1 process only at high amine
concentrations.
The upper diagram of Figure 2 shows that, at a concentration
of [morpholine] ) 0.3 M, the observed pseudo-first-order rate
constant is two times the magnitude of the intercept (kobs
≈
2k1); that is, at this concentration, the reaction of the methyl-
substituted benzhydrylium bromide (1-Me,H) with morpholine
follows the SN1 and the SN2 mechanisms to equal extent. The
calculated lifetime of 4 × 10-15 s for the reaction of the
benzhydrylium ion 2-Me,H with morpholine (Table 8) is shorter
than a bond vibration (≈ 10-13 s). According to Jencks and
Richard, this relationship implies that the SN2 mechanism will
be enforced; that is, the benzhydrylium ion 2-Me,H cannot exist
in an encounter complex with morpholine. From the relationship
k2 ≈ 3k1 (Table 1), one can derive that nucleophilic assistance
for breaking the C-Br bond (fSN2) is very weak and ionization
(k1) may also occur in the absence of a morpholine molecule.
Despite calculated lifetimes for the dichloro-substituted
benzhydrylium ion 2-Cl,Cl which closely resemble those of the
parent compound 2-H,H, Figure 2 shows that almost all amines
prefer the SN2 process at concentrations > 0.2 M. Only
2-aminobutan-1-ol (τ ) 2 × 10-14 s) allows the SN1 mechanism
to dominate at amine concentrations < 0.4 M.
In agreement with calculated lifetimes τ < 2 × 10-16 s, all
reactions of amines with the CF3-substituted benzhydryl bro-
mides 1-CF3, H and 1-CF3,CF3 studied in this work preferentially
follow the SN2 process, and the intercepts of the correlations in
Figure 2 are negligible compared with the pseudo-first-order
rate constants kobs in the presence of amines. The very short
lifetimes estimated for 2-CF3,H and 2-CF3,CF3 in DMSO suggest
that the first-order rate constants for the solvolyses of 1-CF3,H
and 1-CF3,CF3 in DMSO may not be due to SN1 reactions with
formation of the carbocations 2-CF3,H and 2-CF3,CF3 because
(37) McClelland, R.; Kanagasabapathy, V. M.; Banait, N. S.; Steenken, S.
J. Am. Chem. Soc. 1989, 111, 3966–3972.
(38) Minegishi, S.; Kobayashi, S.; Mayr, H. J. Am. Chem. Soc. 2004, 126,
5174–5181.
(39) Measurements of the nucleophilic reactivity of DMSO in neat DMSO
are not possible with our equipment because the laser radiation at 266
nm, which is needed for the photoionization of the benzhydryl
chlorides, is absorbed by DMSO.
(40) Baidya, M.; Kobayashi, S.; Brotzel, F.; Schmidhammer, U.; Riedle,
E.; Mayr, H. Angew. Chem. 2007, 119, 6288-6292; Angew. Chem.,
Int. Ed. 2007, 46, 6176-6179.
9
11400 J. AM. CHEM. SOC. VOL. 131, NO. 32, 2009

Can One Predict Changes from SN1 to SN2 Mechanisms?
A R T I C L E S
the direct nucleophilic attack of DMSO at these benzhydryl
bromides should be enforced. The positive deviation of
1-CF3,CF3 from the correlation of log k1 versus Σσ+ in Figure
4 is in line with a significant nucleophilic solvent participation
by DMSO (k1′, SN2). The fact that the first-order rate constant
for 1-CF3,H matches the correlation with σ+ in Figure 4 implies
that in this case nucleophilic solvent participation by DMSO
cannot be large.
tions by amines. As calculated from the nucleophilicity param-
eters N1 and s of DMSO, the intermediate benzhydrylium ions
Ar2CH+ (2-X,Y), formed by the SN1 process, are quantitatively
trapped by DMSO to give the benzhydryloxysulfonium ions
3-X,Y, the precursors of the benzhydryl alcohols 6-X,Y and
the benzophenones 5-X,Y. Because the change from SN1 to SN2
mechanisms was observed when the lifetimes of the carbocations
in the presence of amines (1 M) were calculated to be ap-
proximately 10-14 s by eq 2, the E, N, and s parameters proved
to be suitable for predicting the preferred mechanism of the
nucleophilic substitutions of benzhydryl bromides. So far, our
analysis, based on Jencks’ lifetime criterion, did not include
the role of the leaving groups. One can expect, however, that
the SN1/SN2 ratio for a certain substrate R-X will also depend
on the leaving group and will increase with increasing nucle-
ofugality of X. Systematic investigations of these effects are in
progress.
Conclusion
In their seminal 1984 paper,16d Richard and Jencks concluded
that a reaction can proceed concurrently through stepwise,
monomolecular and concerted, bimolecular reaction mechanisms
when the intermediate has a long lifetime in the solvent, but no
lifetime when it is in contact with an added nucleophilic agent.
This situation has now been found when benzhydryl bromides
1-X,Y were treated with amines in DMSO. In several cases,
first-order rate constants k1 (s-1) for the formation of the
carbocations are of similar magnitude as the second-order rate
constants k2 (L mol-1 s-1) for the concerted SN2 reactions of
the benzhydryl bromides with amines. For that reason, carboca-
tions with short lifetimes were generated when amine molecules
were not present in the vicinity, while in the same solution,
concerted SN2 reactions were enforced when amine mole-
cules were present. The relationship k2/k1 ) [Ar2CHNRR′]/
([amine]([Ar2CdO] + [Ar2CHOH])) implies that the benzhydryl
amines Ar2CHNRR′ are formed exclusively through the SN2
process and not through trapping of the intermediate carboca-
Acknowledgment. Dedicated to Professor Johann Mulzer on
the occasion of his 65th birthday. We thank Profs. T. W. Bentley
and D. N. Kevill for helpful discussions, and the Deutsche
Forschungsgemeinschaft (Ma 673-20) and the Fonds der Chemi-
schen Industrie for financial support.
Supporting Information Available: Details of kinetic experi-
ments and product studies. This material is available free of
charge via the Internet at http://pubs.acs.org.
JA903207B
9
J. AM. CHEM. SOC. VOL. 131, NO. 32, 2009 11401