Hydride-Transfer Reactions in the Gas Phase. 2. Anchimeric Assistance in the H(-) Transfer from 1,1-Dimethylcyclopentane to Alkyl Cations

Article abstract of DOI:10.1021/jp952728l

The gas-phase hydride transfer from the title compound to several ionic acceptors GA(+) = CH5(+), C2H5(+), s-C3H7(+), and t-C4H9(+) was studied by mass spectrometric and radiolytic methods in the pressure range 1.1E-8 - 700 Torr.Analysis of the irradiated mixtures points to the exclusive formation of rearranged products, with those with methyl migration prevailing over those with ring expansion.Thermochemical considerations, isotope labeling experiments, and deuterium isotope effect measurements concur in defining the detailed mechanism of the gas-phase hydride transfer, which involves significant anchimeric assistance of the methyl and methylene groups adjacent to the reaction center.The primary and the secondary α-D kinetic isotope effects, measured in the hydride transfer from the methylene moieties adjacent to the quaternary carbon of tthe title compound to either s-C3H7(+) or t-C4H9(+), amount to 2.70 (s-C3H7(+)) and 1.89 (t-C4H9(+)) and to 1.77 (s-C3H7(+)) and 1.15 (t-C4H9(+)), respectively.These values are interpreted in terms of a transition state, which is placed rather late along the reaction coordinate, and wherein the assistance of the neighboring (methyl or methylene) group to the departing hydride increases by decreasing the strength of the GA(+) acceptor.The results obtained from the present gas-phase investigation are discussed and compared with those of related gas-phase and solution data.

Full text of DOI:10.1021/jp952728l

J. Phys. Chem. 1996, 100, 8285-8294
8285
Hydride-Transfer Reactions in the Gas Phase. 2.1 Anchimeric Assistance in the H-
Transfer from 1,1-Dimethylcyclopentane to Alkyl Cations
Maria Elisa Crestoni, Simonetta Fornarini, Massimo Lentini, and Maurizio Speranza*
Dipartimento di Studi di Chimica e Tecnologia delle Sostanze Biologicamente AttiVe, UniVersita` di Roma
“La Sapienza”, Pl.e A. Moro 5, 00185 Rome, Italy
ReceiVed: September 15, 1995; In Final Form: February 19, 1996X
+
+
The gas-phase hydride transfer from the title compound to several ionic acceptors GA+ ) CH5 , C2H5 ,
+
s-C3H7 , and t-C4H9+ was studied by mass spectrometric and radiolytic methods in the pressure range 1.1 ×
10-8-700 Torr. Analysis of the irradiated mixtures points to the exclusive formation of rearranged products,
with those with methyl migration prevailing over those with ring expansion. Thermochemical considerations,
isotope labeling experiments, and deuterium isotope effect measurements concur in defining the detailed
mechanism of the gas-phase hydride transfer, which involves significant anchimeric assistance of the methyl
and methylene groups adjacent to the reaction center. The primary and the secondary R-D kinetic isotope
effects, measured in the hydride transfer from the methylene moieties adjacent to the quaternary carbon of
+
+
+
+
the title compound to either s-C3H7 or t-C4H9 , amount to 2.70 (s-C3H7 ) and 1.89 (t-C4H9 ) and to 1.77
+
+
(s-C3H7 ) and 1.15 (t-C4H9 ), respectively. These values are interpreted in terms of a transition state, which
is placed rather late along the reaction coordinate, and wherein the assistance of the neighboring (methyl or
methylene) group to the departing hydride increases by decreasing the strength of the GA+ acceptor. The
results obtained from the present gas-phase investigation are discussed and compared with those of related
gas-phase and solution data.
Introduction
time scale and, thus, reacting at an undefined “temperature”.
Besides, the inherently limited structural resolution of the mass
As pointed out by Hine,2 when a reaction would require the
intermediacy of a highly unstable species, stepwise bond
makings and breakings may be avoided by combining such bond
changes into a concerted step. In intramolecular processes, this
involves the through-space interaction of neighboring or remote
groups to the reaction center and the effect is said to be
“neighboring group participation”.3 When neighboring group
participation affects the reaction rate by lowering the activation
barrier of the rate-limiting concerted step, the group is said to
provide anchimeric assistance.3 Thus, according to Capon and
McManus,3 anchimeric assistance is a subdivision of the more
general neighboring group participation concept.
Owing to their relevance in many aspects of gas-phase ion
chemistry, including stereoisomeric discrimination, many ex-
amples of neighboring group participation have been reported
in the mass spectrometry literature.4,5 Isotope effect, selective
labeling, and stereospecificity studies indicate that the majority
of these processes proceeds via stepwise mechanisms, involving
more than one stable intermediate.6
No unambiguous examples of anchimeric assistance in gas-
phase ionic reactions are instead reported in these studies,
although misuses of the “anchimeric assistance” term just to
denote neighboring group participation sometimes appear in the
relevant publications. This failure is caused by the fact that
positive recognition of anchimeric assistance in an ionic reaction
demands evaluation of the relevant kinetic parameters at well-
defined reaction temperatures, as well as assessment of the
mechanistic details and of their dependence upon the nature of
the reactants.3
spectrometric methods often hampers the positive assessment
of the structural and stereochemical course of the ionic
processes.7 These limitations can be overcome by resorting to
the high-pressure radiolytic methodology (g760 Torr), where
the excitation energy of the ionic species can be fully relaxed
by fast collisional quenching with a buffer gas and their
reactivity determined by standard kinetic procedures, including
pressure- and temperature-dependence studies and competition
experiments. The actual isolation and characterization of the
neutral reaction products provide the necessary information on
structure and stereochemistry of the ionic intermediates in-
volved.7 A demonstration of this capability is provided by the
observation that few unambiguous pieces of evidence of gas-
phase anchimerically assisted ionic processes are reported in
the literature and most of them based upon high-pressure kinetic
investigations using radiolytic methods.1,8 Among these, the
most recent one provided the first unambiguous evidence of
methyl-group anchimeric assistance in the gas-phase hydride
transfer from the CH2 of 2,2-dimethylbutane to tert-butyl cation
(eq 1a; R ) CH3), on the grounds of the over 100-fold increase
of the efficiency of the H- transfer by increasing the system
pressure from 4.0 × 10-8 Torr (FT-ICR) to ca. 1 atm (stationary
radiolysis).1
These essential kinetic and mechanistic pieces of information
are seldom accessible at the low-pressure regimes operating in
most classical mass spectrometric techniques (e10 Torr), where
ionic species are formed with substantial internal energy hardly
relaxed to the rarefied reaction environment on the available
X Abstract published in AdVance ACS Abstracts, April 15, 1996.
S0022-3654(95)02728-6 CCC: $12.00 © 1996 American Chemical Society

8286 J. Phys. Chem., Vol. 100, No. 20, 1996
Crestoni et al.
This pressure-dependence criterion is based upon the con-
sideration that the efficiency of a gas-phase ion-molecule
reaction ()kf/(kf + kr)), involving the formation of a stable,
electrostatically bonded ion-neutral complex (INC) is deter-
mined by the competition between the evolution of INC to
products (kf), through an internal activation barrier E*, and its
back-dissociation to reactants (kr), through the pseudobarrier E°.
This competition depends on both the features of the free energy
profile and the experimental conditions. The complex at the
top of the internal barrier E* is a tight complex (low density of
states, negative activation entropy), whereas that involved in
the back-dissociation is a loose complex (high density of states,
positive activation entropy).7
SCHEME 1
At the collision-free limit (approached in the FT-ICR kinetic
experiments),1 the INC is characterized by a constant total
energy and angular momentum. Its evolution to reactants or
products is essentially governed by entropic, rather than by
enthalpic factors. Under such conditions, its crossing of the
internal barrier E* is rather unlikely and thus, the reaction
efficiency is generally low. An increase in the reaction pressure
(radiolytic experiments)1 favors the disposal of the excess energy
E° of INC by unreactive collisions with the buffer gas (present
at 160-700 Torr).1 Its lifetime increases and its evolution tends
to be progressively governed by enthalpic, rather than by
entropic factors and, therefore, to follow the lowest activation
energy pathway. For exothermic processes with E* > E°, this
favors back-dissociation of INC over its conversion to products
and, therefore, a negative pressure dependence of the reaction
efficiency is observed. For exothermic processes with E* <
E°, instead, an increase in the reaction pressure favors conver-
sion of INC to products over its back-dissociation to reactants
and, therefore, a positive pressure dependence of the reaction
efficiency is observed. Thus, the pressure dependence of the
reaction efficiency can be used as a criterion for establishing
whether the internal barrier E* of a gas-phase ion-molecule
reaction protrudes above the energy level E° of reactants and,
therefore, for throwing light on its detailed mechanism.
conceivable concerted interaction causes an enhancement of the
hydride-transfer rate, by lowering of the corresponding activation
barrier relative to the stepwise process, the reaction is “anchi-
merically assisted”.
It is expected that this study will allow the positive recognition
of anchimeric assistance in the various hydride transfers from
1,1-dimethylcyclopentane and discernment of the intrinsic
factors determining its occurrence. Additional interest in this
gas-phase study arises from the fact that the selected hydride-
transfer reaction 1 (R-R ) -(CH2)3-) can represent the gas-
phase counterpart of the extensively investigated solvolysis of
gem-dimethylcyclopentyl arene sulfonate esters, wherein iden-
tification of anchimeric assistance is severely hampered by
solvation and ion-pairing phenomena.9
In the gas-phase hydride transfer from the CH2 of 2,2-
dimethylbutane to tert-butyl cation (eq 1a; R ) CH3),1 the
positive pressure dependence of the reaction efficiency excludes
the stepwise mechanism 1b, involving formation of an INC
between isobutane and the 3,3-dimethyl-2-butyl cation at energy
levels protruding above that of the starting reactants (E* - E°
> ca. 1-2 kcal mol-1). The positive pressure trend is rather
consistent with a potential energy profile characterized by E*
- E° < 0 and, thus, necessarily proceeding through a concerted,
anchimerically assisted mechanism.
The kinetic approach adopted in this study has been already
presented.1 It is based upon the preparation of stationary
To enlarge the experimental kinetic evidence of alkyl group
anchimeric assistance in gas-phase hydride-transfer reactions,
the investigation has been now extended to another H- donor,
i.e., 1,1-dimethylcyclopentane (1H, eq 1, R-R ) -(CH2)3-),
using different H- acceptors, such as the gaseous acids GA+ )
concentrations of the selected hydride acceptor (GA+ ) CH5
,
+
C2H5+, s-C3H7+, or t-C4H9+) from γ-radiolysis of the corre-
sponding precursor (CH4, C3H8, or i-C4H10), containing traces
(4 Torr) of an effective thermal radical scavenger, i.e., O2. Its
attack on 1,1-dimethylcyclopentane is investigated at 640-700
Torr in the presence of variable concentrations of a powerful
nucleophile, i.e., MeOH (0.4-5.1 Torr). Under these condi-
tions, the high collision frequency (ca. 1010 s-1) with the bath
gas molecules (CH4, C3H8, or i-C4H10) ensures efficient colli-
sional thermalization of the intermediates of sequence 1, whose
structure and isomeric distribution can be inferred from the
relative abundance of their neutral derivatives after trapping by
MeOH.
+
CH5+, C2H5+, s-C3H7+, and t-C4H9 ions. In 1,1-dimethylcy-
clopentane, H- may be released from three chemically different
sites, namely (Scheme 1): (i) the methyl groups (path a); (ii)
the CRH2 and CR′H2 moieties (path b); and (iii) the CâH2 and
Câ′H2 moieties (path c). In principle, the actual occurrence of
the corresponding high-energy intermediates I, IV, and VI might
be avoided by concerted through-space interaction of vicinal
groups with the reaction center. In path a, this would involve
the adjacent CRH2 and CR′H2 groups (yielding directly inter-
mediate II) or the methyl moiety (yielding directly intermediate
III); in path b, the vicinal methyl groups (yielding directly
intermediate V); in path c, the neighboring R and R′ hydrogens
(yielding directly intermediate IV and eventually V). If this
Experimental Section
Materials. Methane, propane, isobutane, and oxygen were
high-purity gases from Matheson Co., used without further

Hydride-Transfer Reactions in the Gas Phase
J. Phys. Chem., Vol. 100, No. 20, 1996 8287
purification. 1,1-Dimethylcyclopentane (1H) was synthesized
by reduction of the tosylhydrazone of 2,2-dimethylcyclopen-
tanone (Aldrich Chemical Co.) with sodium cyanoborohydride.10
The same procedure was employed to obtain exclusively
2-deutero-1,1-dimethylcyclopentane (deuterium content > 98%,
1D) from sodium cyanoborodeuteride (Aldrich Chemical Co.).
1,2-Dimethylcyclopentene was prepared by dehydration of trans-
and cis-1,2-dimethylcyclopentanol (K&K Co.) with phosphoric
acid and purified by distillation at 104 °C, 760 Torr.11
1-Methylcyclohexanol was purchased from Aldrich Chemical
Co. 2,2-Dimethylcyclopentanol was synthesized by reduction
of 2,2-dimethylcyclopentanone with lithium aluminum hy-
dride.12 3,3-Dimethylcyclopentanol was obtained from reduc-
tion of 4,4-dimethyl-2-cyclopenten-1-one (Aldrich Chemical
Co.) with sodium borohydride.13 trans- (2) and cis-1,2-
dimethyl-1-methoxycyclopentane (3), 1-methyl-1-methoxycy-
clohexane (4), 2,2-dimethyl-1-methoxycyclopentane, and 3,3-
dimethyl-1-methoxycyclopentane were obtained from the
Williamson reaction on the corresponding alcohols. 1,1-
Dimethylcyclopentane (1H), 2-deutero-1,1-dimethylcyclopen-
tane (1D), and 1,2-dimethylcyclopentene were purified by
preparative GLC on a 4-m long, 4-mm i.d. stainless steel
column, packed with 5% Carbowax 20M-KOH (2%) on
Supelcoport, at 80 °C. Their final chemical purity exceeded
99.95%. Their identity, as well as that of the above alcohols
and ethers, was verified by NMR spectroscopy and their
chemical and isotopic purity assayed by GLC and GLC-MS on
the same columns employed for the analysis of the irradiated
mixtures.
+
Figure 1. Time dependence of the ln[(s-C3H7+)t/(s-C3H7
)
t)0] ratio
+
following attack of thermalized s-C3H7 ions on 1,1-dimethylcyclo-
pentane 1H (1H partial pressure ) 1.1 × 10-8 Torr (open circles); 1.3
× 10-8 Torr (full circles)).
pentane at (1.1-1.3) × 10-8 Torr. Propane was introduced into
the FT-ICR cell by a pulsed magnetic valve to collisionally
quench any vibrationally and translationally excited ionic
intermediate. After a half second pumping time, the alkyl cation
reactant was isolated from the plasma by sequential broad-band
ejection and “single shots”15 and allowed to react with 1,1-
dimethylcyclopentane. In this way the decrease of the logarithm
of the relative abundance of the starting alkyl cation with the
reaction time appears approximately linear, as required for
thermally equilibrated reactants. The concentration of 1,1-
dimethylcyclopentane in the FT-ICR cell was determined from
the pressure measured by the ionization gauge located in the
high-vacuum pumping line of the FT-ICR cell housing, corrected
by using independently measured calibration factors.16
Radiolytic Experiments. The gaseous mixtures were pre-
pared by conventional techniques, using a greaseless vacuum
line. The reagents and the additives were introduced into
carefully outgassed 250-mL Pyrex bulbs, each equipped with a
break-seal tip. The bulbs were filled with the required mixture
of gases, cooled to the liquid nitrogen temperature, and sealed
off. The irradiations were carried out at 37.5 °C in a 220
Gammacell from Nuclear Canada Ltd to a dose of 2 × 104 Gy
at a rate of 104 Gy h-1, as determined by a neopentane
dosimeter. Control experiments, carried out at doses ranging
from 1 × 104 to 1 × 105 Gy, showed that the relative yields of
products are largely independent of the dose. The radiolytic
products were analyzed by GLC, using a HP 5890 A gas
chromatograph from Hewlett-Packard, equipped with a flame
ionization detector. The following columns were employed:
(i) a 100-m long, 0.32-mm i.d. Petrocol DH fused silica capillary
column, operated at temperatures ranging from 60 to 120 °C, 3
°C min-1; (ii) a 50-m long, 0.32-mm i.d. PONA fused silica
capillary column, operated at temperatures ranging from 60 to
Results
+
FT-ICR Experiments. No detectable formation of C7H13
products was observed after 30 s reaction time, when t-C4H9
ions are introduced and thermalized into a FT-ICR cell
containing 1.3 × 10-8 Torr of 1,1-dimethylcyclopentane 1H.
This implies that, after ca. 17 collisions,17a the extent of the
conceivable hydride transfer from 1H to t-C4H9+ is well below
the mass spectrometric detection limit (ca. 1% of the t-C4H9
peak). Thus, the upper limit of the efficiency of the process,
expressed by the ratio between the observed bimolecular rate
+
+
constant (kobs) and the calculated collision rate constant (kcoll
)
14.0 × 10-10 cm-3 molecule-1 s-1),17a can be placed around 6
× 10-4
.
On the contrary, thermal s-C3H7+ ions react efficiently with
1H under FT-ICR conditions, yielding exclusively the C7H13
120 °C, 3 °C min-1
.
The products were identified by
+
comparison of their retention volumes with those of authentic
standard compounds, and their identity was confirmed by GLC-
MS, using a Hewlett-Packard HP 5970 B mass spectrometer.
Their yields were determined from the areas of the correspond-
ing eluted peaks, using individual calibration curves.
product. The kinetic curves shown in Figure 1 refer to
experiments carried out at 1H partial pressures of 1.3 × 10-8
Torr (full circles) and 1.1 × 10-8 Torr (open circles). The
slopes of the curves, referring to the logarithm of the fraction
+
of starting s-C3H7 ion as a function of time, over the
Fourier Transform Mass Spectrometric (FT-ICR) Experi-
ments. The FT-ICR kinetic experiments were performed with
a Bruker Spectrospin Apex TM 47e FT-MS equipped with an
external source and a cylindrical (60-mm length, 60-mm
diameter) “infinity cell”,14 situated between the poles of a
superconducting magnet operated at 4.7 T. The kinetic experi-
ments were performed at the nominal temperature of 300 K.
The alkyl cation reactant was generated in the external ion
source by 70 eV electron impact on the corresponding neutral
precursor at the nominal source pressure of 1 × 10-5 Torr and
transferred into the FT-ICR cell containing 1,1-dimethylcyclo-
corresponding [1H] concentration, provide directly the bimo-
lecular rate constant of the hydride transfer process (kobs) which
amounts to 9.9 × 10-10 cm-3 molecule-1 s-1. The hydride-
transfer efficiency can be estimated from the relevant collision
rate constant (kcoll ) 15.4 × 10-10 cm-3 molecule-1 s-1) as
amounting to 0.64.
Radiolytic Experiments. Radiolytic experiments were car-
ried out to investigate the kinetics and the mechanism of the
hydride-transfer 1 (R-R ) -(CH2)3-) in the high-pressure
range, by applying traditional procedures such as the trapping
of the ionic intermediates by a powerful nucleophile (MeOH)

8288 J. Phys. Chem., Vol. 100, No. 20, 1996
Crestoni et al.
TABLE 1: Product Yields from the Gas-Phase Attack of GA+ Ions on 1,1-Dimethylcyclopentane
a O2: 4 Torr. Radiation dose: 2 × 104 Gy (dose rate: 1 × 104 Gy h-1). b Percent values as the ratios between the absolute yield of each product
and the combined yields of all products identified. Each value is the average of several determinations, with an uncertainty range of ca. 5% of the
value. c The absolute yields are estimated from the percent ratio of the products, expressed as their G(M) values (the number of molecules M produced
per 100 eV of absorbed energy) and the literature G(GA values.18 d n ) 1, 2. e n.d. ) below detection limit, ca. 0.2%.
+
)
after a given reaction time and the isolation and structural
discrimination of their neutral end products. The composition
of the reaction systems and the absolute yields of products,
expressed as the ratio between their G(M) values (number of
molecules formed/100 eV absorbed energy) and that of their
GA+ precursors,18 as well as their isomeric distribution are
reported in Table 1. The ionic nature of the radiolytic products
of Table 1 is ensured by addition to the gaseous mixtures of
ca. 0.5 mol % of O2, a powerful thermal radical scavenger, and
it is demonstrated by the sharp decrease of the overall product
yields (over 80%) caused by addition to the gaseous mixture
of 0.4 mol % of NMe3, an efficient positive ion interceptor.
An outstanding feature of the product patterns of Table 1 is
the complete absence of both 2,2-dimethyl-1-methoxycyclo-
pentane and 3,3-dimethyl-1-methoxycyclopentane among the
radiolytic products, irrespective of the concentration of MeOH
(0.4-5.1 Torr). Analysis of the GLC-MS fragmentation pattern
of the radiolytic products excludes also the occurrence of methyl
(1-methylcyclopentyl)methyl ether, as well as of any cyclopen-
tane derivatives containing the ethyl moiety, including 1-ethyl-
1-methoxycyclopentane. Instead, nearly equal amounts of both
trans- (2H) and cis-1,2-dimethyl-1-methoxycyclopentane (3H)
are invariably formed under all conditions, together with minor
yields of 1-methyl-1-methoxycyclohexane (4H, 8.0-21.6%).
The relative abundance of 4H does not depend much on the
MeOH partial pressure (entries ii-iv of Table 1), but it rather
increases with the strength of the GA+ acid, i.e., in the order
and cis-1,2-dimethyl-1-methoxycyclopentane (3D), accompanied
by smaller equimolar yields of their unlabeled analogues 2H
(4.4-5.9% (values corrected for the initial ca. 2% H content in
1D)) and 3H (5.4-8.1% (values corrected for the initial ca.
2% H content in 1D)). Deuterated 1-methyl-1-methoxycyclo-
hexane (4D) is formed as well (8.0-17.4%), whereas its
unlabeled analogue 4H is barely detectable (0.2-1.8% (values
corrected for the initial 2% H content in 1D)). The observation
that in the majority of the experiments, the [4D]/[4H] yield ratios
largely exceed that expected for statistical uptake of a ring
hydrogen by GA+ (i.e., [4D]/[4H] ) 7), suggests that 1-methyl-
1-methoxycylohexane arises predominantly from the loss of a
hydride ion originally belonging to a methyl group of 1,1-
dimethylcyclopentane, while any conceivable contribution from
the loss of one of its ring hydrogens seems to play a negligible
role.
Discussion
Nature of the GA+ Acceptors. The nature and distribution
of the stable GA+ ions from ionization of each individual bulk
component of the gaseous mixture have been intensely inves-
tigated in a variety of mass spectrometric and radiolytic
studies.18,19 γ-Radiolysis of CH4 produces known yields of both
CH5+ (∆Hf° (standard formation enthalpy) ) 216 kcal mol-1 20
)
and C2H5 ions (∆Hf° ) 215.6 kcal mol-1),20 which are
+
powerful Brønsted (proton affinity (PA) ) 131.6 kcal mol-1
(CH4); 162.6 kcal mol-1 (C2H4))20 and Lewis acids. In 1 atm
of CH4 and at 298 K, the CH5+ ions are present predominantly
in the monosolvated [CH5+•CH4] form (ca. 85%, ∆Hf° ) ca.
191 kcal mol-1), amounting the free CH5+ ions to only 15% of
t-C4H9 < s-C3H7 < CnH5+. Replacement of s-C3H7 with
s-C3D7+, as the hydride ion acceptor, in the experiments with
1H does not lead to any significant incorporation of the D label
into the ethereal products (e0.8%, entries v and vi of Table 1).
Equal proportions of trans- (2H) and cis-1,2-dimethyl-1-
methoxycyclopentane (3H) are obtained as well from CnH5+(n
) 1,2)-protonation of 1,2-dimethylcyclopentene. No detectable
formation of 1-methyl-1-methoxycyclohexane (4H) was ob-
served from this starting substrate.
The same product pattern observed with 1H is obtained when
using 2-deuterio-1,1-dimethylcyclopentane (1D), as the hydride
(or deuteride) donor. Analysis of the GLC-MS fragmentation
pattern of the radiolytic products reveals the predominant
formation of almost equimolar amounts of deuterated trans- (2D)
+
+
+
the ionic species.21 In the same systems, C2H5 ions are
+
obtained mostly in the free state, together with ca. 13% in the
monosolvated [C2H5+•CH4] form (∆Hf° ) ca. 192 kcal mol-1).22
When traces of C3H8 are introduced in the CH4 mixtures at
atmospheric pressure, both CH5+ and C2H5+ ions rapidly abstract
a hydride ion from C3H8 yielding quantitative amounts of
s-C3H7 (∆Hf° ) 190.9 kcal mol-1).20 In 1 atm of C3H8 and
+
+
at 298 K, complete clustering between the s-C3H7 ions and
their parent C3H8 molecules leads to the [s-C3H7+•C3H8] adduct
(∆Hf° ) ca. 152 kcal mol-1).23 No significant clustering
+
between the s-C3H7 ions and the CH4 molecules takes place

Hydride-Transfer Reactions in the Gas Phase
J. Phys. Chem., Vol. 100, No. 20, 1996 8289
TABLE 2: Thermochemical Data (kcal/mol-1) (Estimated
Values in Italic)
in the methane/propane mixtures.22 γ-Radiolysis of i-C4H10
produces known yields of t-C4H9 ions (∆Hf° ) 165.8 kcal
+
mol-1).20 In 1 atm of i-C4H10 and at 298 K, nearly one-half of
the t-C4H9+ ions remains in the free state (ca. 48%), while the
other half adds to an i-C4H10 molecule forming a loosely bonded
[t-C4H9+•i-C4H10] adduct (∆Hf° ) ca. 126 kcal mol-1).23 Later
on, thermochemical considerations will indicate that almost all
the free and the solvated ions described above may actually act
as hydride acceptor toward 1H.
Reaction Kinetics. In both the radiolytic and the FT-ICR
experiments, the GA+ ions undergo many unreactive collisions
with their bulk precursor and, therefore, are thermally equili-
brated with the reaction medium before interacting with 1,1-
dimethylcyclopentane.
Nevertheless, under collision-free conditions (FT-ICR; pres-
sure e 1.3 × 10-8 Torr), the electrostatic interaction energy
liberated in the formation of the INC between GA+ and 1,1-
dimethylcyclopentane cannot be dissipated to the reaction
medium. Most of it is stored as excess internal energy among
the INC degrees of freedom. It follows that the phenomeno-
logical kinetic parameters measured in the FT-ICR experiments
do not refer to a defined reaction temperature (thermal kinetics)
but rather represent microscopic rate constants averaged over a
non-Boltzmann excitation energy distribution.7 Besides, a
portion of the electrostatic interaction energy liberated in the
formation of the INC is employed to counterbalance the increase
of the rotational energy of the complex, due to the conservation
of the total angular momentum.24 Thus, not all the INC
interaction energy is spendable to overcome E*, and hence the
reaction efficiency may be further lowered. For the systems
investigated, it is estimated that, at room temperature, the effect
of conservation of total angular momentum on the reaction
efficiency never exceeds 1 order of magnitude.24,25
no change in the heat of formation apart from that associated
to the inherent stability of the two radicals, it results a ∆H°f
value of 177 kcal mol-1 for IV. A correction for the stabilizing
effect of two electron-releasing methyl groups of CMe2 on the
positive charge of 6 kcal mol-1 is made, consistent with
calculation on similar systems. This leads to a corrected value
of 171 kcal mol-1 for ∆H°f of IV. Recourse to the same
procedure provides the quoted formation enthalpies for 3,3-
dimethylcyclopentyl cation VI (∆H°f ) 175 kcal mol-1),
1-ethylcyclopentyl cation III (∆H°f ) 159 kcal mol-1), and for
the hypothetical (1-methylcyclopentyl)methyl cation I (∆H°f )
197 kcal mol-1). The formation enthalpies used for evaluating
the thermochemistry of the reactions of Scheme 1 are listed in
Table 2.
Instead, at the high pressures typical of the radiolytic
experiments (640-700 Torr), the high collisional frequency (ca.
1010 s-1) with the bulk gas allows fast disposal of the excess
translational, vibrational, and rotational energy of the INC before
its conversion to products. It follows that the kinetic and
mechanistic information obtained under these conditions refers
to thermally equilibrated systems and can be adequately
correlated with similar processes occurring in solution.7 The
efficiencies of the thermal processes occurring in the radiolytic
experiments are represented by the total absolute yields of
products of Table 1, which reflect the conversion extent of their
ionic precursors in the time interval between the generation of
the ions and their trapping by MeOH (ca. 3 × 10-9 s).17b The
absolute yield values, reported in Table 1, indicate that, while
with s-C3H7+, as the hydride acceptor, the reaction efficiency
slightly decreases (from 0.6 to 0.2-0.4) by increasing the buffer
gas pressure from ca. 10-8 to 700 Torr, with t-C4H9+ it increases
by over 3 orders of magnitude, i.e., by a factor which cannot
be attributed exclusively to the effect of conservation of total
angular momentum under the FT-ICR conditions (ca. 10-8 Torr),
but which rather depends on the specific mechanistic path
followed by the hydride-transfer reaction under largely different
pressure conditions.
According to the reported values, all pathways of Scheme 1,
involving the CnH5+ (n ) 1, 2) acceptor, are thermochemically
allowed (∆H° (standard reaction enthalpy) e -6 kcal mol-1),
except perhaps that involving hydride abstraction from a methyl
group of 1H by [CH5+•CH4] (path a) (∆H° ) 3 kcal mol-1).27
If this latter pathway is actually energetically precluded, the
[CH5+•CH4] cluster (G(M) ) 0.85 × 1.918a ) ca. 1.62) is able
to abstract only the ring hydride ions (∆H°(kcal mol-1) ) -23
+
(path b); -19 (paths c)), whereas the residual free CH5 ions
(G(M) ) 0.15 × 1.9 ) ca. 0.28) are able to abstract all the
hydrogens of 1H (∆H°(kcal mol-1) ) -4 (path a); -30 (path
+
b); -26 (paths c)). C2H5 ions, either free (G(M) ) 0.87 ×
0.918a ) ca. 0.8) and in the CH4-solvated form (G(M) ) 0.13 ×
0.9 ) ca. 0.1), are able to abstract all the hydrogens of 1H
(∆H°(kcal mol-1) ) -6 (free) and 0 (solv) (path a); -32 (free)
and -26 (solv) (path b); -28 (free) and -22 (solv) (paths c)).
Within the hypothesis that the exothermic attack on 1,1-
dimethylcyclopentane by the methane ions is totally indiscrimi-
nate, the expected [4]/([2] + [3]) product ratios would range
around 0.22, a value which is close to the measured one (0.27
(1H), Table 1). These findings are consistent with a rather
unselective hydride-transfer pattern from 1,1-dimethylcyclo-
pentane to the methane ions, although its detailed mechanism
remains still undefined.27 The rather stable intermediate III does
not seem to be kinetically accessible on the potential energy
surface describing path a, as demonstrated by the complete
absence of its methoxy derivative among the reaction products.28
Formation of the methoxy derivatives of the much less stable
species I, IV, and VI is not observed either, thus suggesting
Ion Termochemistry and Reaction Mechanism. In view
of the unavailability of experimental thermochemical data
concerning the conceivable cyclopentyl ion intermediates I, III,
IV, and VI of Scheme 1, their formation enthalpies are estimated
using a method based on the isodesmic substitution26 approach.
Thus, for IV, the heats of formation of 1,1-dimethylcyclopen-
tane, cyclopentyl cation, and cyclopentane are known.20 As-
suming that replacement of the CH2 moiety adjacent to the
vacant orbital in cyclopentyl cation by the CMe2 group produces

8290 J. Phys. Chem., Vol. 100, No. 20, 1996
Crestoni et al.
the degenerate rearrangement of unsubstituted cyclopentyl cation
(e2.8 kcal mol-1),31 whereas that of the IV f V and I f II
1,2-alkyl (or alkenyl) group shifts as equal to that measured in
the degenerate rearrangement of 2,3,3-trimethylbutyl-2 cation
(ca. 6 kcal mol-1).32 The activation barrier of gas-phase
intracluster hydride-ion transfer [GAH•I (or VI) (or IV)] f
[GA+•1H] has been found to be strictly correlated to the
corresponding reaction enthalpy.18i For exothermic processes
(-∆H° > 3 kcal mol-1), involving unencumbered H- donors,
the activation barrier, if existing at all,23 rarely exceeds 2 kcal
mol-1 18i,33
.
According to the relevant energy profiles, neither
[s-C3H7+•C3H8] (Figure 2) nor the t-C4H9+/[t-C4H9+•i-C4H10]
pair (Figure 3) are energetically able to abstract a hydride ion
from the methyl groups of 1H via the hypothetical stepwise
mechanism involving the high-energy primary species I. This
is because the relevant energy barriers protrude by no less than
14 and 25 kcal mol-1, respectively, above that of back-
dissociation of starting reagents. Nonetheless, substantial
amounts of 4H, whose formation implies necessarily the hydride
abstraction from the methyl groups of 1H (path a of Scheme
1), were recovered in both systems (Table 1). In analogy with
the conclusions reached in related investigations,1 formation of
4H can be only accounted for by anchimeric assistance of the
CRH2 group (or the CR′H2 one) of 1H to the departure of the
hydride ion from a CH3 moiety of the donor to the GA+
acceptor. This implies that interaction of either [s-C3H7+•C3H8]
or the t-C4H9+/[t-C4H9+•i-C4H10] pair with the methyl groups
of 1H leads directly to II via the concerted mechanism 2a,
whose activation free energy must therefore lie not only below
that of their hypothetical stepwise process but also below the
corresponding back-dissociation barrier. CRH2-group anchi-
meric assistance to the hydride loss from the CH3 of 1H (eq
2a) supersedes that involving the other methyl group (eq 2b)
and leading directly to intermediate III, as demonstrated by the
absence of its ethereal derivative among the radiolytic products.
Figure 2. Potential energy profile for the gas-phase reaction between
s-C3H7 ions and 1,1-dimethylcyclopentane 1H. The figures in italic
refer to the total standard formation enthalpy of the corresponding
species.
+
Figure 3. Potential energy profile for the gas-phase reaction between
+
t-C4H9 ions and 1,1-dimethylcyclopentane 1H. The figures in italic
refer to the total standard formation enthalpy of the corresponding
species.
A similar conclusion is reached as regards the mechanism of
formation of the ionic precursor V of ethers 2H and 3H from
attack of the t-C4H9+/[t-C4H9+•i-C4H10] pair on 1H (Figure 3).
Here, the stepwise mechanisms involving the intermediacy of
either the [i-C4H10•VI] adduct (followed by rapid intracluster
VI f IV isomerization) (path c of Scheme 1) or, directly, the
[i-C4H10•IV] adduct (path b of Scheme 1) would require
overcoming an activation barrier of at least 5 kcal mol-1 above
that of back-dissociation to the starting reagents. The corre-
sponding free energy gap is even larger since stepwise conver-
sion of [t-C4H9+•1H] to [i-C4H10•V] via either [i-C4H10•VI] or
[i-C4H10•IV] proceeds through a “tight” transition state (negative
activation entropy), whereas its back-dissociation involves a
“loose” transition structure (positive activation entropy).34 In
this case, the efficiency of the stepwise process should follow
a negative pressure dependence, rather than the experimentally
observed positive pressure dependence (reaction efficiency: ca.
6 × 10-4 at 10-8 Torr; 0.07-0.10 at ca. 700 Torr). Thus, it
can be concluded that attack of the t-C4H9+/[t-C4H9+•i-C4H10]
pair on 1H leads directly to the intermediate [i-C4H10•V] through
a concerted mechanism (eq 3), involving an activation barrier
that either structures I, IV, and VI are not accessible along the
CnH5+(n ) 1, 2)-1H reaction coordinate or, if they are, they
are completely converted into the more stable forms II and V
within the time of their interception by the MeOH nucleophile
(ca. 3 × 10-9 s).17b
A more informative picture is obtained from analysis of the
potential energy profiles reported in Figures 2 and 3, concerning
+
respectively the attack of [s-C3H7+•C3H8] and of the t-C4H9
/
[t-C4H9+•i-C4H10] pair on 1H. The profiles refer to the
hypothetical stepwise reaction mechanism depicted in Scheme
1. The depths of the potential wells corresponding to INCs,
e.g., [s-C3H7+•1H] and [t-C4H9+•1H], were obtained from the
classical ion-molecule interaction theory,29 after calibration of
the calculation method using the experimentally measured
potential well depths of the [s-C3H7+•C3H8] and [t-C4H9+•i-
C4H10] adducts.23,30 No precise measurements of the activation
barriers of Figures 2 and 3 are available. Nevertheless, the
activation barrier for the conceivable VI f IV 1,2-hydrogen
shift can be taken as approximately equal to that measured in

Hydride-Transfer Reactions in the Gas Phase
J. Phys. Chem., Vol. 100, No. 20, 1996 8291
CHART 1
much lower than those of the conceivable stepwise pathways,
via the higher energy [i-C4H10•VI] and [i-C4H10•IV] intermedi-
ates. Since the concerted pathway 3 inVolVes a rate-determining
step with an actiVation barrier significantly lower than those
of the conceiVable stepwise routes, the process can be rightfully
labeled as anchimerically assisted.
Analysis of the potential energy profiles involving the
[s-C3H7+•C3H8] species (Figure 2), as the hydride acceptor, while
suggesting a similar anchimerically assisted mechanism 2a for
the hydride transfer involving the methyl group of the donor,
does not allow us to reach any straightforward conclusions as
to whether the process involving the ring hydrogens is concerted
or not. In fact, for these reactions, the internal energy barriers
•
of the conceivable stepwise paths, involving either the [C3H8 VI]
adduct (followed by rapid intracluster VI f IV isomerization)
•
(path c of Scheme 1) or, directly, the [C3H8 IV] one (path b of
Scheme 1), do not protrude above the reactants energy, and
therefore the observed pressure dependence of the reaction
efficiency can satisfy both hypotheses. Nevertheless, discrimi-
nation between the concerted and the stepwise mechanism can
be obtained from the results of Table 1 (entries v and vi)
concerning the attack of [s-C3D7+•C3D8] on 1H. These are
characterized by the almost exclusive formation of unlabeled
ethers 2H and 3H, whereas their deuterated analogues 2D and
3D are barely detectable. On the grounds of the energy profiles
shown in Figure 2, in fact, any conceivable [C3D7H•IV], formed
either directly, by hydride abstraction from the CRH2 moiety of
1H (path b of Scheme 1), or via rapid intracluster VI f IV
isomerization (path c of Scheme 1), is expected to rapidly
undergo a sort of internal return to the adduct [s-C3D7+•1H]
(and, of course, its [s-C3D6H+•1D] isotopomer),35 before its rate-
limiting conversion to the [propane•V] isomer. It follows that
a significant incorporation of the deuterium label should be
observed in both the ethereal products and the starting 1,1-
dimethylcyclopentane. The lack of any significant D incorpora-
tion in these compounds after irradiation speaks against the
occurrence of the [propane•IV] structure, formed either directly
from hydride transfer from the CRH2 moiety of 1H to the
s-propyl ion, or indirectly through the [propane•VI] structure.
(OMe)dCHD]+ from 2′′D-3′′D of Chart 1). Both ethers 2H
and 3H from 1H display identical fragmentation patterns,
coinciding, mutatis mutandis, within ca. 2% with those of 2D
and 3D from 1D. Thus, for instance, the [m/z ) 72]/[m/z )
128] and the [m/z ) 72]/[m/z ) 113] intensity ratios from ethers
2H and 3H are nearly identical to the ([m/z ) 72] + [m/z )
73])/[m/z ) 129] and the ([m/z ) 72] + [m/z ) 73])/[m/z )
114] ones from 2D and 3D, respectively. It is concluded that
the presence and the position of a deuterium label in 2D and
3D do not influence appreciably their fragmentation patterns
and, therefore, the relative abundance of the m/z ) 72 and m/z
) 73 peaks closely reflects the relative distribution of the ethers
of Chart 1.
From the reaction pattern shown in Scheme 2, and within
the very reasonable assumption that trapping of intermediates
VH, V′D, and VD by MeOH and neutralization of the ensuing
derivatives are relatively fast steps of the reaction sequences, it
results that:
Θ ) [MeC(OMe)dCH2]+/[MeC(OMe)dCHD]+ )
•
On these grounds, formation of intermediate [C3H8 V] is
R′
believed to proceed through the concerted mechanism 3.
Kinetic Isotope Effects (KIEs). An additional piece of
evidence in favor of the intervention of anchimeric assistance
in these gas-phase hydride-transfer reactions arises from the
evaluation of the kinetic isotope effects (KIEs), using 2-deutero-
1,1-dimethylcyclopentane as the starting substrate. The evalu-
ation of the primary (PKIE) and secondary kinetic isotope effects
(SKIE) is based upon the mass spectrometric fragmentation
pattern of trans- and cis-1,2-dimethyl-1-methoxycyclopentanes
arising from attack of GA+ on 1D in the presence of MeOH. A
significant fragment from 70 eV electron impact on the
unlabeled ethers 2H and 3H is [MeC(OMe)dCH2]+ (m/z ) 72)
Chart 1.36 The same peak is observed in the mass fragmentation
of the cis- and trans-2-deutero-1,2-dimethyl-1-methoxycyclo-
(kD + kHR + kHR′)/kH
R
The kD/(kH + 2kHR′) ratio is expressed by the Φ ) ([2H] +
R
[3H])/([2D] + [3D]) fraction. Therefore, since kD ) Φ(kH
2kHR′), Θ can be expressed by the following equation:
+
Θ ) (1 + Φ)(kHR/kHR′) - (1 + 2Φ)
which, if rearranged, gives
kHR′/kHR ) (1 + Φ)/[Θ - (1 + 2Φ)]
In the same way, the kHR′/kD can be expressed by the following
equation:
+
pentane (2′D-3′D in Chart 1), obtained from gas-phase CnD5
(n ) 1,2)-deuteronation of 1,2-dimethyl-cyclopentene in the
presence of methanol. Mass spectra of trans- and cis-1,2-
dimethyl-1-methoxycyclopentanes from attack of GA+ on 1D
in the presence of MeOH are instead characterized by the
presence of the base peak at m/z ) 72 ([MeC(OMe)dCH2]+
from 2H-3H, 2′D-3′D, and 2′′′D-3′′′D of Chart 1), ac-
companied by a less abundant fragment at m/z ) 73 ([MeC-
kHR′/kD ) (1 + Φ)/[Φ(Θ + 1)]
Now, in consideration of the negligible CR′HD kinetic effect (a
γ-deuterium effect)37 expected on the kHR′ (Scheme 2), the kH
R′
can be taken as equal to the rate constant for the hydride transfer
from the CRH2 moiety of 1H to GA+. In this way, the kHR′/kD
and the kHR′/kHR ratios express respectively the primary (PKIE)

8292 J. Phys. Chem., Vol. 100, No. 20, 1996
Crestoni et al.
SCHEME 2
TABLE 3: First-Order Rate Constants (Θ)a and Product
Yield (Φ)a Ratios and Deuterium Isotope Effects in the
Hydride Transfer from the CrH2 of
radiolytic experiments (kHR′/kD ) 1.89-2.70, Table 3) excludes
that the occurrence of the hypothetical stepwise route b in favor
of a concerted, anchimerically assisted mechanism with a
transition state involving a significant CR-H bond rupture in
1,1-dimethylcyclopentane.
1,1-Dimethylcyclopentane to the GA+ Acceptors (37.5 °C)
run no.
GA+
(Θ)
(Φ)
′kRH/kD
′kRH/kRH
+
+
+
ix
x
xi
s-C3H7
s-C3H7
s-C3H7
1.970 0.143
1.919 0.139
1.919 0.152
av
2.69
2.81
2.60
1.67
1.78
1.87
Unlike PKIE, inference of transition-state geometries from
SKIEs is more complicated. The magnitude of SKIEs may be
partitioned into variable translational, rotational, and vibrational
contributions from all the transition-state moieties, although,
for those involving CR-H bonds, vibrational motion plays a
predominant role.38 Accordingly, following Streitwieser’s
formulation,41a a normal SKIE mainly reflects the out-of-plane
bending (wagging) motion of the CR-H bonds, as a response
of the extent of sp3-sp2 CR rehybridization at the transition state.
2.70 ( 0.11 1.77 ( 0.10
+
xii
t-C4H9
2.400 0.184
1.89 1.15
a See text.
and the secondary R-D kinetic isotope effect (SKIE) in reaction
3. The relevant values are summarized in Table 3, together
with the corresponding Φ and Θ factors.
R
In this framework, the relatively large SKIE (kHR′/kH ) 1.77)
The significant PKIE observed in the hydride transfer from
1,1-dimethylcyclopentane to the s-C3H7+ and t-C4H9+ acceptors,
essentially related to the virtual hydride stretching motion,38
provides further evidence in favor of a concerted, anchimerically
assisted mechanism. In fact, on the grounds of the energy
profiles of Figures 2 and 3, the hypothetical stepwise sequence
(b) of Scheme 1 would proceed via a fast [GA+•1H] T
[GAH•IV] preequilibrium step, followed by the rate-limiting
[GAH•IV] f [GAH•V] intracluster isomerization. Within this
hypothesis, the overall rate constant of the stepwise hydride
transfer would be directly proportional to the kiKeq product,
where ki is the [GAH•IV] f [GAH•V] isomerization rate
constant, and Keq is the [GA+•1H] T [GAH•IV] pseudoequi-
librium constant. The ki constant is unaffected by replacement
of the leaving hydride by a deuteride ion in the stepwise
sequence b of Scheme 1. The effect might be sizable on Keq,
depending upon the force field of the s-C3H7-H and t-C4H9-H
bonds relative to that of the CR-H bonds of 1,1-dimethylcy-
clopentane.39 Since the force field of homologue bonds, e.g.,
the C-H bonds, is proportional to the corresponding dissociation
energy,40 that of the CR-H bonds of 1,1-dimethylcyclopentane
is just below that of the GA-H bonds (GA ) s-C3H7 and
t-C4H9). On these grounds, Keq would display only a limited
shown in entries ix-xi of Table 3 indicates that, in the transition
state of the concerted hydride transfer from the CRH2 moiety
of 1H to s-C3H7+, the CR-H wagging motion is nearly
completely free.41 This implies that the relevant transition
structure is characterized by a limited anchimeric assistance of
the methyl groups of 1H to the leaving hydride ion. The lower
kHR′/kHR ) 1.15, measured with t-C4H9+, indicates that the same
interaction is much more developed in the corresponding
transition state. Here, the accompanying decrease of the PKIE
(kHR′/kD from 2.70 (s-C3H7+) to 1.89 (t-C4H9+)) reflects a parallel
effect (Hammond postulate effect) on the reaction surface, which
is consistent with a transition structure resembling the [i-C4H10•V]
one and placed later along the reaction coordinate relative to
•
that leading to the [C3H8 V] adduct.
In conclusion, both PKIE and the SKIE in the hydride transfer
reaction 3 with GA+ ) s-C3H7 and t-C4H9 indicate that the
reaction is anchimerically assisted by the adjacent methyl
groups. In the departure of the CR-H hydride ion, the
anchimeric assistance of the vicinal methyl groups, although
only partially developed, plays a decisive role. In fact, it
provides that limited energy push necessary to break completely
the residual CR‚‚‚H‚‚‚GA+ bonding in the transition state,
without which the reaction would not take place.
+
+
H
inVerse isotope effect (Keq e KeqD), if any. It follows that,
within the hypothesis of a hydride-transfer reaction proceeding
via the stepwise sequence (b), the measured rate constant would
be affected by a small inVerse isotope effect as well (kiKeqH e
kiKeqD). On the contrary, the normal primary isotope effect,
measured for the hydride-transfer rate constant from the
An evaluation of the efficiency of the hydride transfer
between 1H and GA+ ) s-C3H7+ or t-C4H9+, measured in the
radiolytic experiments at 640-700 Torr, with that measured at
ca. 1 × 10-8 Torr in the FT-ICR cell can elucidate this latter
point. Apart from the limited effect of conservation of total

Hydride-Transfer Reactions in the Gas Phase
J. Phys. Chem., Vol. 100, No. 20, 1996 8293
angular momentum,24,25 the unreactivity of t-C4H9+ ions toward
1H, observed in the FT-ICR experiments, is due to the inability
of the [t-C4H9+•1H] adduct to dispose readily of its excitation
energy by collisions with the rarefied gaseous medium (ca. 1
× 10-8 Torr). Under such high-energy conditions, entropy-
favored back-dissociation of the excited [t-C4H9+•1H] adduct
by far predominates over the anchimerically assisted hydride
transfer, whose transition state requires a rigorous antiperiplanar
alignment between the leaving moiety and the assisting group.
In dense gaseous media (640-700 Torr), efficient collisional
quenching of the excited [t-C4H9+•1H] adduct takes place in a
time (ca. 10-10 s) short relative to its back-dissociation, which
suddenly becomes no longer energetically feasible. This
enhances the lifetime of the [t-C4H9+•1H] adduct so as to allow
a methyl group of 1H vicinal to the reaction center to attain
the proper alignment relative to the leaving hydride ion and to
anchimerically assist its departure.
route, through a sort of internally solVated transition state
characterized by the anchimeric assistance of the methyl groups
adjacent to the reaction center.
Aqueous ethanolysis of (1-methylcyclopentyl)methyl sul-
fonate esters involves extensive anchimeric assistance by the
ring methylene moieties adjacent to the reaction center, in strict
analogy with the concerted mechanism observed in the gas-
phase hydride transfer from the methyl group of 1H to the GA+
acceptor.9b In both systems, anchimeric assistance partially
compensates for the methyl carbon-leaving group bond dis-
sociation energy (eq 2). Reaction 2 demands a higher extent
of anchimeric assistance than that involved in reaction 3, owing
to the greater energy of the unrearranged primary cation I
relative to that of the unrearranged secondary cation IV.
Conclusions
The present gas-phase investigation, based upon structural
discrimination of the neutral products, deuterium labeling
experiments, and kinetic isotope and GA+-acceptor effects,
demonstrates that, in the gas phase, the hydride-transfer reaction
from 1,1-dimethylcyclopentane to several ionic acceptors in-
volves significant anchimeric assistance by neighboring alkyl
and alkylene groups. This conclusion, which reinforces previous
incidental pieces of evidence,1 outlines the general view of gas-
phase hydride-transfer processes, whose occurrence is governed
by internal solvation from suitably located groups. This implies
that the intention of generating in the gas phase a specific
carbocationic structure, especially if secondary or primary, by
loss of a hydride ion from suitable hydrocarbons may prove
utopian, if the hydride transfer requires the anchimeric assistance
from a neighboring group and, therefore, a concerted structural
rearrangements.
Irrespective of the GA+ acceptor employed, the transition state
of the concerted hydride transfer reaction from the CRH2 moiety
of 1,1-dimethylcyclopentane is characterized by the extensive
cleavage of the CR-H bond, anchimerically assisted by an
adjacent methyl group. The extent of both the CR-H bond
cleavage and the methyl-group anchimeric assistance increases
by decreasing the strength of the GA+ acceptor. Comparatively
more pronounced anchimeric assistance by the CRH2 moiety is
observed in the gas-phase hydride-transfer reaction from the
CH3 groups of 1,1-dimethylcyclopentane to GA+.
According to the relevant PKIE and SKIE, a similar pattern
+
is operative in the hydride transfer from 1H to the s-C3H7
acceptor under high-pressure radiolytic conditions. The fact
that, at the low-pressures of the FT-ICR experiments (ca. 1 ×
10-8 Torr), the hydride transfer from 1H to s-C3H7+ is a rather
efficient process as well suggests that the excited [s-C3H7+•1H]
adduct can efficiently overcome the limited activation-barrier
•
leading to [C3H8 V] prior to its back-dissociation to reactants.
Under these conditions, the specific mechanism for the excited
•
[s-C3H7+•1H] conversion to [C3H8 V], whether concerted or
involving the entropy-favored stepwise pathway, remains un-
defined.
Comparison with Related Solvolytic Studies. Comparison
of the present gas-phase results with those concerning strictly
related solvolytic studies allows us to discern the intrinsic
structural and electronic factors determining the anchimerically
assisted mechanism from the perturbing influence of solvation
and ion-pairing phenomena.
Solvolysis of 2,2-dimethylcyclopentyl p-bromobenzene-
sulfonate (BsO-1H) in media of different nucleophilicity
follows a stepwise pattern, whose energy profile is qualitatively
similar to those reported in Figures 2 and 3.9a In nucleophilic
solvents, BsO-1H (corresponding to [GA+•1H] in Figures 2
and 3) ionizes to form the intimate ion pair [BsO-•IV]
(corresponding to [GAH•IV] in Figures 2 and 3), which
undergoes internal return (analogous to the fast [GA+•1H] T
[GAH•IV] preequilibrium). Because it is sterically hindered
from nucleophilic attack by the solvent, [BsO-•IV] undergoes
rate-determining rearrangement to [BsO-•V] (corresponding to
[GAH•V] in Figures 2 and 3). The solvent is not involved
nucleophilically before and during the rate-determining methyl-
migration step, so that the reaction rate is influenced primarily
by changes in the ionizing strength of the solvent. Accordingly,
almost completely rearranged products (analogous to ethers 2H
and 3H) are formed from solvolysis of BsO-1H, irrespective
of the nucleophilicity of the reaction medium. The large SKIE
of 1.20, measured in these systems, is consistent with a transition
state involving no significant covalent bonding by nucleophile
(either the solvent or the migrating methyl moiety) to the
reaction center and therefore, according to Hammond postulate,
its structure resembles that of the [BsO-•IV] ion pair.
Comparison of the gas-phase results with those of strictly
related solvolytic reactions, while revealing a substantial
mechanistic uniformity, points to some discrepancies between
the stepwise mechanism involved in the solvolysis of 2,2-
dimethylcyclopentyl brosylate and the concerted one governing
the gas-phase hydride-ion transfer from the CRH2 moiety of 1,1-
dimethylcyclopentane, which are attributed to solvation and ion-
pairing phenomena.
References and Notes
(1) Part 1: Crestoni, M. E.; Fornarini, S.; Lentini, M.; Speranza, M.
J. Chem. Soc., Chem. Commun. 1995, 121.
(2) Hine, J. J. Am. Chem. Soc. 1972, 94, 5766.
(3) Capon, B.; McManus, S. Neighboring Group Participation; Ple-
num: New York, 1976.
(4) See, for instance: (a) Turecek, F. Int. J. Mass Spectrom. Ion Proc.
1991, 108, 137. (b) Kingston, E. E.; Shannon, J. S.; Lacey, M. J. Org.
Mass Spectrom. 1983, 18, 183. (c) Mandelbaum, A. Mass Spectrom. ReV.
1983, 2, 223. (d) Morton, T. H. Tetrahedron 1982, 38, 3195.
(5) Splitter, J. S.; Turecek, F. Applications of Mass Spectrometry to
Organic Stereochemistry; VCH Publishers: New York, 1994.
(6) (a) McAdoo, D. J.; Hudson, C. E. Int. J. Mass Spectrom. Ion Proc.
1984, 62, 269 and references therein. (b) Stringer, M. B.; Underwood, D.
J.; Bowie, J. H.; Allison, C. E.; Donchi, K. F.; Derrick, P. J. Org. Mass
Spectrom. 1992, 27, 270 and references therein.
In the gas phase, the occurrence of the [GAH•IV] intermediate
from intracluster hydride transfer in the [GA+•1H] adduct cannot
benefit by stabilizing interaction with the medium and, espe-
cially, with the counterion, whose role in these systems is played
by the far removed nonnucleophilic neutral hydrocarbon GAH
(the C+‚‚‚GAH distance in the transition state may largely
exceed 3 Å).24 Therefore, the system evolves via a concerted

8294 J. Phys. Chem., Vol. 100, No. 20, 1996
Crestoni et al.
(7) (a) Cacace, F. Acc. Chem. Res. 1988, 21, 215. (b) Speranza, M.
Mass Spectrom. ReV. 1992, 11, 73. (c) Speranza, M. Int. J. Mass Spectrom.
Ion Proc. 1992, 118/119, 395.
(29) (a) Biermann, H. W.; Freeman, W. P.; Morton, T. H. J. Am. Chem.
Soc. 1982, 104, 2307. (b) Hall, D. G.; Morton, T. H. J. Am. Chem. Soc.
1980, 102, 5688. (c) Morton, T. H. J. Am. Chem. Soc. 1980, 102, 1596.
+
(8) (a) Angelini, G.; Speranza, M. J. Am. Chem. Soc. 1981, 103, 3800.
(b) DePetris, G.; Giacomello, P.; Picotti, T.; Pizzabiocca, A.; Renzi, G.;
Speranza, M. J. Am. Chem. Soc. 1986, 108, 7491. (c) Fornarini, S.;
Sparapani, C.; Speranza, M. J. Am. Chem. Soc. 1988, 110, 34. (d) Ibid.
1988, 110, 42. (e) DePetris, G.; Giacomello, P.; Pizzabiocca, A.; Renzi,
G.; Speranza, M. J. Am. Chem. Soc. 1988, 110, 1098. (f) Cecchi, P.;
Cipollini, R.; Pizzabiocca, A.; Renzi, G.; Speranza, M. Tetrahedron 1988,
44, 4847. (g) Fornarini, S.; Muraglia, V. J. Am. Chem. Soc. 1989, 111,
873. See also ref 19a.
(9) (a) Shiner, Jr., V. J.; Imhoff, M. A. J. Am. Chem. Soc. 1985, 107,
2121. (b) Shiner, Jr., V. J.; Tai, J. J. J. Am. Chem. Soc. 1981, 103, 436. (c)
Imhoff, M. A.; Ragain, R. M.; Moore, K.; Shiner, Jr., V. J. J. Org. Chem.
1991, 56, 3542.
(30) In Figure 2, the association enthalpy of any C7H13 ion with a
propane molecule is taken as equal to that measured for s-C3H7+ (13.6 kcal
mol-1).23 The ligand exchange leading to [s-C3H7+•1H] + C3H8 is estimated
to release ca. 13 kcal mol-1, namely, the difference between the association
+
enthalpies of s-C3H7 with 1H (ca. 26.4 kcal mol-1) (see text) and with
23
propane (13.6 kcal mol-1
)
In Figure 3, the association enthalpy of any
C7H13+ isomer with an isobutane molecule is taken as equal to that measured
for t-C4H9+ (7.2 kcal mol-1 23 The ligand exchange leading to [t-C4H9+•1H]
)
+ i-C4H10 is estimated to release ca. 3 kcal mol-1, namely, the difference
between the association enthalpies of t-C4H9+ with 1H (ca. 10 kcal mol-1
,
see text) and with isobutane (7.2 kcal mol-1).23
(31) Quoted in ref 75 of Hehre, W. J.; Radom, L.; Schleyer; P. v. R.;
Pople, J. A. Ab Initio Molecular Orbital Theory; Wiley: New York, 1986;
p 395.
(32) (a) Saunders, M.; Kates, M. R. J. Am. Chem. Soc. 1978, 100, 7082.
(b) Frenking, G. Tetrahedron 1984, 40, 377. (c) Saunders, M.; Vogel, P.;
Hagen, E. L.; Rosenfeld, J. Acc. Chem. Res. 1973, 6, 53. (d) Myhre, P. C.;
Yannoni, C. S. J. Am. Chem. Soc. 1981, 103, 230.
(33) Solomon, J. J.; Meot-Ner, M.; Field, F. H. J. Am. Chem. Soc. 1974,
96, 3727.
(34) See ref 7b and references therein.
(10) Hutchins, R. O.; Milewski, C. A.; Maryanoff, B. E. J. Am. Chem.
Soc. 1973, 95, 3662.
(11) Bulgrin, V. C.; Dahlgren, G. J. Am. Chem. Soc. 1958, 80, 3883.
(12) Brown, H. C.; Deck, H. R. J. Am. Chem. Soc. 1965, 87, 5620.
(13) Brown, H. C.; Hess, H. M. J. Org. Chem. 1969, 34, 2206.
(14) Caravatti, P.; Alleman, M. Org. Mass Spectrom. 1991, 26, 514.
(15) de Koning, L. J.; Fokkens, R. H.; Pinkse, F. A.; Nibbering, N. M.
M. Int. J. Mass Spectrom. Ion Proc. 1987, 77, 95.
(16) Bartmess, J. E. Vacuum 1983, 33, 149.
(17) (a) Langevin, P. Ann. Chim. Phys. Ser. 8 1905, 5, 245. (b) Su, T.;
Chesnavitch, W. J. J. Chem. Phys. 1982, 76, 5183.
(35) If the volume V of the electrostatically bonded [s-C3D7+•1H] is
taken as that of a sphere of 10 Å in diameter (V ) ca. 4 × 10-21 cm3), the
concentration C of 1H in [sC3D7+•1H] and, thus, of C3D7H in the
conceivable [C3D7H•IV] adduct would amount to ca. 2 × 1020 molecules
cm-3. The constant ()k1/k-1) of the hypothetical [s-C3D7+•1H] T
[C3D7H•IV] equilibrium can be approximately estimated as large as ca. 0.2
from the ∆H° ) +1 kcal mol-1, if differential entropy contributions can
be neglected. The first-order k1 constant ranges around 2 × 1011 s-1, if
10-9 cm3 molecule-1 s-1 is taken for the corresponding second-order
constant (see the present FT-ICR kinetic results; see also ref 25). It follows
that the first-order rate constant k-1 ) ca. 5k1 ) ca. 1012 s-1. Even
considering a pronounced deuterium isotope effect, extensive [C3D7H•IV]
(18) (a) Ausloos, P.; Lias, S. G.; Gorden, Jr., R. J. Chem. Phys. 1963,
39, 3341. (b) Ausloos, P. Ion-Molecule Reactions; Franklin, J. L., Ed.;
Plenum: New York, 1970. (c) Ausloos, P.; Lias, S. G. J. Chem. Phys. 1962,
36, 3163. (d) Lias, S. G.; Ausloos, P. J. Chem. Phys. 1962, 37, 877. (e)
Sandoval, I. B.; Ausloos, P. J. Chem. Phys. 1963, 38, 452. (f) Freeman, G.
R. Radiat. Res. ReV. 1968, 1, 1. (g) Ausloos, P. Progr. React. Kinet. 1969,
5, 113. (h) Ausloos, P.; Lias, S. G. J. Chem. Phys. 1965, 43, 2748. (i)
Ausloos, P.; Lias, S. G. J. Am. Chem. Soc. 1970, 92, 5037. (j) Rebbert, R.
E.; Ausloos, P. J. Res. Natl. Bur. Stand., Sect. A 1972, 76, 329. (k) Lias, S.
G.; Rebbert, R. E.; Ausloos, P. J. Chem. Phys. 1972, 57, 2080.
(19) (a) Kim, J. K.; Findlay, M. C.; Henderson, W. G.; Caserio, M. C.
J. Am. Chem. Soc. 1973, 95, 2184. (b) Jardine, I.; Fenselau, C. J. Am. Chem.
Soc. 1976, 98, 5086. (c) Luczynski, Z.; Herman, J. A. J. Phys. Chem. 1978,
82, 1679. (d) Sen Sharma, D. K.; Kebarle, P. J. Am. Chem. Soc. 1978, 100,
5826. (e) Richter, W. J.; Schwarz, H. Angew. Chem., Int. Ed. Engl. 1978,
17, 424.
(20) (a) Lias, S. G.; Bartmess, J. E.; Liebman, J. F.; Holmes, J. L.; Levin,
R. D.; Mallard, W. G. J. Phys. Chem. Ref. Data 1988, 17, Suppl. 1. (b)
Stull, D. R.; Westrum, Jr., E. F.; Sinke, G. C. The Chemical Thermodynamics
of Organic Compounds; Wiley: New York, 1969.
(21) Hiraoka, K.; Kebarle, P. J. Am. Chem. Soc. 1975, 97, 4179.
(22) Hiraoka, K.; Mori, T.; Yamabe, S. Chem. Phys. Lett. 1993, 207,
178.
•
T [sC3D6H+•1D] T [C3D6H2 IV′] (IV′ ) 2,2-dimethyl-1-D-cyclopentyl
cation) interchange is expected to occur prior to stepwise intracluster
conversion of IV (or IV′) to V.
(36) Isomeric 1,2-dimethylcyclopentanols and 1-methylcyclohexanol
display the same mass spectroscopic fragmentation pattern. It is characterized
by a distribution of fragment ions paralleling that from ethers 2-4, but
shifted by -14 amu (the CH2 moiety). Thus, the [MeC(OMe)dCH2]+
fragment (m/z ) 72) of ethers 2-4 corresponds in the spectra of isomeric
1,2-dimethylcyclopentanols and 1-methylcyclohexanol to a significant m/z
) 58 peak, due to the [MeC(OH)dCH2]+ fragment. A pronounced m/z )
58 peak, but this time due to the [HC(OMe)dCH2]+ fragment, characterizes
the mass spectra of isomers of ethers 2-4, not containing a methyl group
geminal to the MeO moiety, i.e., 2,2-dimethyl-1-methoxycyclopentanols.
(37) The deuterium isotope effect of a nonmigrating CHD moiety in
the γ position, such as the CR′HD one in 1D (third equation of Scheme 2),
in the rate determining formation of a carbocation has been measured to
fall slightly below 1 (ca. 0.98), see ref 9 and Schubert, W. M.; LeFevre, P.
H. J. Am. Chem. Soc. 1969, 91, 7746).
(23) Sunner, J. A.; Hirao, K.; Kebarle, P. J. Phys. Chem. 1989, 93, 4010.
(24) Magnera, T. F.; Kebarle, P. Ionic Processes in the Gas Phase;
Almoster-Ferreira, M. A., Ed.; Plenum: New York, 1984.
(25) Cf. ref 18i and: Lias, S. G.; Eyler, J. R.; Ausloos, P. Int. J. Mass
Spectrom. Ion Phys. 1976, 19, 219.
(38) Schrøder Glad, S.; Jensen, F. J. Am. Chem. Soc. 1994, 116, 9302.
(39) Hartshorn, S. R.; Shiner, Jr., V. J. J. Am. Chem. Soc. 1972, 94,
9002.
(26) Radom, L.; Pople, J. A.; Schleyer, P. v. R. J. Am. Chem. Soc. 1972,
94, 5935.
(27) Owing to the pronounced Brønsted acidity of CnH5+ (n ) 1, 2),20
occurrence of the relevant ionic intermediates I-VI in the CH4 systems
may as well involve preliminary protonation of 1H followed by H2
elimination from the protonated form.
(40) The CR-H bond dissociation energy (D°(CR-H)) of 1,1-dimeth-
ylcyclopentane is unknown. If cyclopentane can be regarded as a suitable
model of 1,1-dimethylcyclopentane, its D°(CR-H) can be estimated as large
as 94.8 kcal mol-1. The D°(t-Bu-H) is 95.2 kcal mol-1, while D°(i-Pr-
H) is 99.4 kcal mol-1 20
.
(28) The hypothesis of complete III T II rearrangement before their
trapping by MeOH is rather unlikely owing to the pronounced energy barrier
associated to the III f II ring-expansion process (15 kcal mol-1) ( Kirchen,
R. P.; Sorensen, T. S.; Wagstaff, K. M. J. Am. Chem. Soc. 1978, 100, 5134.
Viruela-Martin, P. M.; Nebot-Gil, I.; Viruela-Martin, R.; Planelles, J. J.
Chem. Soc., Perkin Trans. 2 1987, 307.
(41) (a) Streitwieser, Jr., A.; Jagow, R. H.; Fahey, R. C.; Suzuki, S. J.
Am. Chem. Soc. 1958, 80, 2326. (b) Scheppele, S. E. Chem. ReV. 1972, 72,
511.
JP952728L