Hindered inversion of chiral ion-dipole pairs

Article abstract of DOI:10.1021/ja001960o

O-Protonated S-(-)-1-phenyl-1-methoxyethane (I_s) has been generated in the gas phase by (CH₃)₂Cl⁺ methylation of S-(-)-1-phenylethanol (1_s). Detailed information on the reorganization dynamics of the

Full text of DOI:10.1021/ja001960o

J. Am. Chem. Soc. 2001, 123, 2251-2254
2251
Hindered Inversion of Chiral Ion-Dipole Pairs
Antonello Filippi, Francesco Gasparrini, and Maurizio Speranza*
Contribution from the Dipartimento di Studi di Chimica e Tecnologia delle Sostanze Biologicamente
AttiVe, UniVersita` degli Studi di Roma “La Sapienza”, P.le A. Moro 5, I-00185 Roma, Italy
ReceiVed June 6, 2000
Abstract: O-Protonated S-(-)-1-phenyl-1-methoxyethane (IS) has been generated in the gas phase by (CH3)2-
Cl+ methylation of S-(-)-1-phenylethanol (1S). Detailed information on the reorganization dynamics of the
intimate ion-dipole pair (IIS), arising from IS by C-O bond dissociation, is inferred from the kinetic study
of the intramolecular inversion of configuration of IS vs its dissociation to R-methylbenzyl cation (III) and
CH3OH. The behavior of IIS in the gas phase is compared to that observed in aqueous solutions, where the
loss of optical activity of IS is prevented by exchange of the leaving CH3OH with the solvent shell. Hindered
inversion of IS in solution is attributed to the operation of attractive interactions between the moving CH3OH
moiety and the solvent cage which inhibit internal return in the intimate ion-dipole pair IIS. Similar interactions
do not operate in the solvolysis of 18O-labeled 1S in aqueous acids, whose loss of optical activity efficiently
competes with exchange of the leaving H218O with the solvent shell.
Introduction
Scheme 1
Despite the great interest on the role of intimate ion-dipole
intermediates in solvolytic reactions, the dynamics of their
reorganization in a solvent cage remains largely unknown.1
Recent progress in this area is mainly due to detailed kinetic
and mechanistic studies of the acid-catalyzed solvolysis of
optically active 1-phenylalkanols and their methyl ethers.2,3 The
investigations were based on the competition between the loss
of optical activity of the chiral substrate and the exchange of
the leaving moiety (XOH in Scheme 1) with a molecule of
solvent (for simplicity, the solvent cage is represented by H2O
in Scheme 1).
The rate of 18O exchange between water and the chiral-labeled
alcohols as a function of racemization has been extensively used
as a criterion for discriminating the SN2 from the SN1 mecha-
nisms of solvolysis. The expected ratio of exchange vs racem-
ization rate is 0.5 for the SN2 mechanism and 1.0 for a pure
SN1 process.4 With chiral 18O-enriched 1-phenylethanol in
aqueous acids, this ratio is found to be equal to 0.84 ( 0.05.
This value has been interpreted in terms of the kinetic pattern
of Scheme 1 involving the reversible dissociation of the oxonium
ion IS (XOH ) H218O) to the chiral intimate ion-dipole pair
IIS (k-1 > kinv). In IIS, the leaving H218O molecule does not
equilibrate immediately with the solvent (i.e. H216O), but remains
closely associated with the ion. This means that kinv is of the
same order of magnitude as kdiss.2 In contrast, the rate constant
ratio of exchange vs racemization of chiral 1-phenyl-1-meth-
oxyethane in acidic acetonitrile-water solutions is as large as
0.99. The closeness of this value to that of a pure SN1
mechanism indicates that, in Scheme 1 (XOH ) CH3OH), either
k
inv is many orders of magnitude lower than kdiss or, if not, that
internal return is negligible (k-1 , kinv).3 This kinetic ambiguity
prevents identification of the actual factors hindering inversion
in IIS (XOH ) CH3OH).5
This paper is aimed at removing this ambiguity by comparing
the behavior of optically active 1-phenylethanol and its methyl
ethers in aqueous acids with that of their protonated derivatives
in the gas phase. The absence of the solvent and the possibility
of modulating in the gas phase the concentration of the added
nucleophiles allow us to assess kinv and kdiss without any
perturbation from the reaction medium and incursion of spurious
reaction pathways.5 In addition, the comparison allows the
determination of the effects of the aqueous solvent on the
kinetics of Scheme 1.
(1) See, for instance: Richard, J. P.; Tsuji, Y. J. Am. Chem. Soc. 2000,
122, 3963 and references therein.
(2) (a) Merritt, M. V.; Bell, S. J.; Cheon, H. J.; Darlington, J. A.; Dugger,
T. L.; Elliott, N. B.; Fairbrother, G. L.; Melendez, C. S.; Smith, E. V.;
Schwartz, P. L. J. Am. Chem. Soc. 1990, 112, 3560 and references therein.
(b) Merritt, M. V.; Anderson, D. B.; Basu, K. A.; Chang, I. W.; Cheon, H.
J.; Mukundan, N. E.; Flannery, C. A.; Kim, A. Y.; Vaishampayan, A.; Yens,
D. A. J. Am. Chem. Soc. 1994, 116, 5551 and references therein.
(3) Thibblin, A. J. Phys. Org. Chem. 1993, 6, 287 and references therein.
(4) For a review, see: Samuel, D.; Silver, B. AdV. Phys. Org. Chem.
1965, 3, 128.
The chiral oxonium ion IS (XOH ) CH3OH) is conveniently
produced in the gas phase by methylating S-(-)-1-phenylethanol
(5) The different behavior of chiral 1-phenylalkanols and their methyl
ethers in acids has been attributed to the different lifetime of the relevant
intermediates IIS and their sensitivity to solvation effects. Incursion of
bimolecular substitutions and elimination-addition pathways may play a
role as well (refs 2 and 3).
10.1021/ja001960o CCC: $20.00 © 2001 American Chemical Society
Published on Web 02/15/2001

2252 J. Am. Chem. Soc., Vol. 123, No. 10, 2001
Filippi et al.
ments were carried out to exclude the occurrence of thermal decom-
position and racemization of the starting alcohol as well as of its ethereal
products 2S and 2R within the temperature range investigated.
The extent of 18O incorporation into the radiolytic products was
determined by GLC-MS, setting the mass analyzer in the selected ion
mode (SIM). The ion fragments at m/z 121 (16O - [M - CH3]+) and
123 (18O - [M - CH3]+) were monitored to analyze the 2S and 2R
ethers. The corresponding alcohols 1S and 1R were examined by using
the fragments at m/z 107 (16O - [M - CH3]+ content) and 109 (18O -
[M - CH3]+).
Scheme 2
Computational Details. Quantum chemical calculations were
performed with use of an IBM RISC/6000 version of the GAUSSIAN
94 set of programs.8 The 6-31G* basis set was employed for all the
atoms to optimize the geometries of the investigated species at the
density functional level of theory, using the B3LYP functional which
combines Becke’s three-parameter hybrid description of exchange and
the correlation functional of Lee, Yang, and Parr.9 At the same level
of theory, frequency calculations were performed for all the optimized
structure to ascertain their minimum or transition state nature. Thermal
contribution to enthalpy at 298 K and 1 atm, which include the effects
of translation, rotation, and vibration, was evaluated by classical
statistical thermodynamics within the approximation of ideal gas, rigid
rotor, and harmonic oscillator behavior and using the recommended
scale factor (0.994) for frequencies and zero-point energy correction.
The intrinsic reaction coordinate (IRC) procedure10 has been used to
ascertain that the transition structures, identified on the potential energy
hypersurface, are directly and continuously linked to the corresponding
energy minima.
(1S) with (CH3)2Cl+ ions (Scheme 2). The latter ions are
generated by γ-radiolysis of CH3Cl, present as a bulk component
(720 Torr) of gaseous mixtures containing traces of the alcoholic
substrate, of H218O, of a radical scavenger (i.e. O2), and of a
powerful base (i.e. (C2H5)3N). This procedure allows formation
of IS (XOH ) CH3OH) in a gaseous inert medium (CH3Cl) at
pressures high enough to ensure its complete thermalization.
Experimental Section
Materials. Methyl chloride and oxygen were high-purity gases from
UCAR Specialty Gases N. V., used without further purification. H218O
(18O-content > 97%) and (C2H5)3N were purchased from ICON
Services. S-(-)-1-Phenylethanol (1S), its R-enantiomer (1R), and styrene
(3) were research grade chemicals from Aldrich Co. Alcohol 1S, used
as starting substrate, was purified by enantioselective semipreparative
HPLC on a chiral column of (R,R)-Ulmo (5 µm, 250 × 4.6 mm i.d.),
Results and Discussion
The main products from γ-radiolysis of the gaseous CH3Cl/
S-(-)-1-phenylethanol (1S) systems are S-(-)-1-phenyl-1-meth-
oxyethane (2S), R-(+)-1-phenyl-1-methoxyethane (2R), styrene
(3), and the 1-phenylethanol racemate (rac-1).11 Their relative
eluent 99/1 (v/v) n-hexane/propan-2-ol, flow rate 1.0 mL min-1
;
12
yields are listed in Table 1 under the Y2S, Y2R, Y3, and Yrac-1
detection by UV (254 nm) and ORD (polarimeter) in series [k1′(-) )
3.64; R)1.09; T 25 °C] and by enantioselective HRGC: (i) MEGADEX
DACTBS-â (30% 2,3-di-O-acetyl-6-O-tert-butyldimethylsilyl-â-cyclo-
dextrin in OV 1701; 25 m long, 0.25 mm i.d, df 0.25 µm) fused silica
column, at 60 < T <170 °C, 4 °C min-1; (ii) MEGADEX 5 (30%
2,3-di-O-methyl-6-O-pentyl-â-cyclodextrin in OV 1701; 25 m long,
0.25 mm i.d, df 0.25 µm) fused silica column at T ) 125 °C. S-(-)-
1-Phenyl-1-methoxyethane (2S) and its R-enantiomer (2R) were syn-
thesized from the corresponding alcohols by the dimethyl sulfate
method.6 Their identity was verified by classical spectroscopic methods.
Procedure. The gaseous mixtures were prepared by conventional
techniques, with the use of a greaseless vacuum line. Alcohol 1S (0.5-
0.6 Torr), H218O (2-3 Torr), the radical scavenger O2 (4 Torr), and
the powerful base B ) (C2H5)3N (1.2 Torr; proton affinity (PA) )
234.7 kcal mol-1)7 were introduced into carefully outgassed 130 mL
Pyrex bulbs, each equipped with a break-seal tip. The bulbs were filled
with CH3Cl (720 Torr), cooled to the liquid-nitrogen temperature, and
sealed off. The irradiation were carried out at constant temperatures
ranging from 25 to 160 °C with a 60Co source to a dose of 2 × 104 Gy
at a rate of 1 × 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,
with a Perkin-Elmer 8700 gas chromatograph equipped with a flame
ionization detector, on the same columns used to analyze the starting
alcohol 1S. The products were identified by comparison of their retention
volumes with those of authentic standard compounds and their identity
confirmed by GLC-MS, using a Hewlett-Packard 5890 A gas chro-
matograph in line with a HP 5970 B mass spectrometer. Their yields
were determined from the areas of the corresponding eluted peaks, using
the internal standard (i.e. benzyl alcohol) method and individual
calibration factors to correct for the detector response. Blank experi-
headings, respectively. The figures in the table represent the
mean yield factors of the products, as obtained from several
separate irradiations carried out under the same experimental
conditions and whose reproducibility is expressed by the
uncertainty level quoted. The ionic origin of the products is
demonstrated by the sharp decrease (over 80%) of their
abundance as the (C2H5)3N concentration is raised from ca. 0.1
to ca. 0.5 mol %.
No appreciable incorporation of the 18O label is observed in
the ethereal products 2S and 2R, whereas the 18O-content in
racemate rac-1 amounts to ∼40%. The lack of any significant
incorporation of the 18O label into the ethereal products 2S and
2R excludes the involvement of water at any stage of their
formation and points to the radiolytic (CH3)2Cl+ ions as their
exclusive precursors. The predominance of ether 2S over its
enantiomer 2R under all conditions indicates that (CH3)2Cl+
(8) Frish, M. J.; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.; Johnson,
B. G.; Robb, M. A.; Cheeseman, J. R.; Keith, T. A.; Petersson, G. A.;
Montgomery, J. A.; Raghavachari, K.; Al-Laham, M. A.; Zakrzewski, V.
G.; Ortiz, J. V.; Foresman, J. B.; Cioslowski, J.; Stefanov, B. B.;
Nanayakkara, A.; Challacombe, M.; Peng, C. Y.; Ayala, P. Y.; Chen, W.;
Wong, M. W.; Andres, J. L.; Repogle, E. S.; Gomperts, R.; Martin, R. L.;
Fox, D. J.; Binkley, J. S.; Defrees, D. J.; Baker, J.; Stewart, J. P.; Head-
Gordon, M.; Gonzales, C.; Pople, J. A. Gaussian 94, ReVision C. 2;
Gaussian, Inc.: Pittsburgh, PA, 1995.
(9) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 1372, 5648. (b) Lee, C.;
Yang, W.; Parr, R. G. Phys. ReV. B 1988, 37, 785.
(10) Gonzales, C.; Schlegel, H. B. J. Phys. Chem. 1990, 94, 5523.
(11) The irradiated systems invariably contain H216O, as ubiquitous
impurity either initially introduced in the mixture together with its bulk
component or formed from its radiolysis. As pointed out previously (Troiani,
A.; Gasparrini, F.; Grandinetti, F.; Speranza, M. J. Am. Chem. Soc. 1997,
119, 4525. Speranza, M.; Troiani, A. J. Org. Chem. 1998, 63, 1020), the
average stationary concentration of H216O in the radiolytic systems is
estimated to approach that of the added H218O (ca. 2-3 Torr).
(6) Achet, D.; Rocrelle, D.; Murengezi, I.; Delmas, A.; Gaset, A.
Synthesis 1986, 643.
(7) Lias, S. G.; Hunter, E. P. L. J. Phys. Chem. Ref. Data 1998, 27, 413.

Hindered InVersion of Chiral Ion-Dipole Pairs
J. Am. Chem. Soc., Vol. 123, No. 10, 2001 2253
Table 1. Gas-Phase Inversion vs Dissociation of O-Protonated
a
(S)-(-)-1-Phenyl-1-methoxyethane IS
rate constants,
c
yield factorb
(×10-6 s-1
)
reactn
temp (°C)
Y2R
Y2S
Y3
Yrac-1
kinv
kdiss
25
60
85
100
120
140
140
160
0.015
0.047
0.096
0.085
0.082
0.079
0.061
0.030
0.960
0.837
0.716
0.556
0.335
0.191
0.156
0.054
0.005
0.042
0.068
0.116
0.299
0.421
0.468
0.615
0.020
0.074
0.120
0.243
0.284
0.309
0.315
0.301
0.7
2.4
5.4
5.9
8.9
15.2
14.2
20.2
1.2
5.4
8.3
17.1
31.2
45.1
52.7
79.9
Figure 1. Arrhenius plots for the IS / IR intracomplex rearrangement
(circles) and the IS f III + CH3OH dissociation (diamonds).
a CH3Cl: 720 Torr. O2: 4 Torr. H218O: 2-3 Torr. (C2H5)3N: 1.2
Torr. Radiation dose: 2 × 104 Gy (dose rate: 1 × 104 Gy h-1). b Each
value is the average of several determinations, with an uncertainty level
of ca. 5%; Yrac-1 estimated by doubling the yield of (R)-(+)-1-
phenylethanol (ref 12). c See text.
log kdiss ) (11.9 ( 0.3) - [(7.9 ( 0.2) × 103]/2.303RT
(r2 ) 0.992) (7)
attacks the O-center of alcohol 1S yielding primarily the oxonium
intermediate IS with the same configuration of the starting
substrate. Formation of alkene 3 and of the partially labeled
rac-1 mixture13 is attributed to the partial unimolecular14
dissociation of IS into the R-methylbenzyl cation (III) and CH3-
OH prior to its neutralization by the strong base B ) (C2H5)3N.
Accordingly, formation of the products of Table 1 conforms
to the reaction network of Scheme 2. Its kinetic treatment leads
The relevant activation parameters, calculated from the
transition-state theory equation, are the following: ∆H‡
)
inv
5.4 ( 0.3 kcal mol-1 and ∆S‡inv ) -13.3 ( 1.0 cal mol-1 K-1
;
∆H‡ ) 7.1 ( 0.3 kcal mol-1 and ∆S‡ ) -6.7 ( 1.2 cal
diss
diss
mol-1 K-1
.
The values in Table 1 indicate that, at all temperatures
investigated, kinv is anything but negligible relative to kdiss, being
just 2-4 times lower. This implies that, in acidic media, the
hindered inversion of IS (XOH ) CH3OH) has to be ascribed
to the lack of appreciable IIS f IS (and IIR f IR) internal return
(k-1 , kinv; Scheme 1), rather than to kinv negligible relative to
kdiss.3 Accordingly, the difference in the behavior of IIS (and
IIR) in acidic solution essentially reduces to k-1 > kdiss, when
XOH ) H218O, and k-1 , kdiss, when XOH ) CH3OH.
A reason for such a difference has to be sought in the relevant
IS f IR transition structures and their position along the reaction
coordinate.
15
to the following equations:
inv + kdiss)τ
Y2S ) 0.5[e-k τ + e-(2k
]
(1)
(2)
(3)
diss
inv + kdiss)τ
Y2R ) 0.5[e-k τ - e-(2k
]
diss
dissτ
Y2 + Yrac-1 ) 1 - e-k
The kinv and kdiss rate constants, derived from eqs 1-3, are
expressed as:
Quantum-chemical calculations at the B3LYP/6-31G* level
have been employed to gather information on this point.17 The
structures and 298 K relative enthalpies of the critical points
identified on the [C6H5CH+CH3,CH3OH] hypersurface are
illustrated in Figure 2. The first-order critical points TS1 and
TS2 represent two transition structures of the IS f IR inversion
reaction. Taking into account the intrinsic limitations of the
B3LYP functional in describing noncovalent interactions and
the necessarily limited dimensions of the 6-31G* basis set
employed, the IS f IR (XOH ) CH3OH) activation enthalpies
computed at 298 K (2 (TS1) and 4 kcal mol-1 (TS2)) are not
kinv ) 0.5τ-1{ln[(Y2S + Y2R)/(Y2S - Y2R)]}
kdiss ) τ-1{ln[1 -(Y3 + Yrac-1)]-1}
(4)
(5)
The τ term represents the lifetime of ions IS and IR prior to
deprotonation by the base B ) (C2H5)3N. Taking equal to unit
the efficiency of the ion deprotonation by B (e.g. kb in Scheme
2), τ is expressed by (kb[B])-1 16
The Arrhenius plots of kinv
.
inconsistent with that obtained experimentally (∆H‡
)
and kdiss over the 25-160 °C temperature range are reported in
Figure 1. The linear curves obey the following equations:
inv
5.4 ( 0.3 kcal mol-1). According to Figure 2, the TS1 and TS2
geometries show a pronounced CR-O bond distance between
the leaving CH3OH molecule and the planar carbocation III
(2.86 and 3.55 Å, respectively). The moving CH3OH moiety is
much closer to the acidic hydrogens of III than to its CR center.
This suggests that hydrogen bond interactions between CH3-
OH and the acidic hydrogens of III play a major role in TS1
and TS2, whose structure resembles more those of the inter-
mediates IV and V, respectively, than that of the starting
oxonium ion IS. This view is corroborated by the relevant
log kinv ) (10.4 ( 0.1) - [(6.2 ( 0.2) × 103]/2.303RT
(r2 ) 0.994) (6)
(12) Owing to the presence of the starting alcohol 1S, the extent of
IIS f IS association cannot be directly determined. However, GLC-MS
analysis reveals the formation of 18O-labeled 1S in concentrations equal to
that of the 18O-labeled enantiomer 1R. On these grounds, the overall
abundance of 1S can be taken as equal to that of 1R and, therefore, the
yield factor of rac-1 can be estimated by doubling that of 1R.
(13) The conceivable bimolecular H218O-to-CH3OH displacement on IS
as a route to racemate rac-1 is ruled out on both stereochemical and
thermochemical grounds. For instance, H2O-to-CH3OH substitution on
O-protonated benzyl methyl ether is 14.4 kcal mol-1 endothermic.
(14) Kinetic predominance of proton transfers vs â-eliminative processes
allows us to assign the formation of 3 to deprotonation of III by (C2H5)3N.
(15) Andraos, J. J. Chem. Educ. 1999, 76, 1578.
experimental data. The negative ∆S‡ value is a symptom of
inv
the stiffness of the transition structures TS1 and TS2 due to the
coordination of the CH3OH molecule between two hydrogens
of the planar carbocation III and to the restricted rotation of its
(17) Owing to the molecular complexity of the species involved, any
description of the [C6H5CH+CH3, CH3OH] hypersurface at a more
sophisticated level of theory is computationally unfeasible.
(16) The collision constant kb between IIS and (C2H5)3N is calculated
according to: Su, T.; Chesnavitch, W. J. J. Chem. Phys. 1982, 76, 5183.

2254 J. Am. Chem. Soc., Vol. 123, No. 10, 2001
Filippi et al.
Figure 2. B3LYP/6-31G*-optimized geometries and 298 K relative enthalpies (in kcal mol-1) of the critical structures on the [C6H5CH+CH3,CH3-
OH] hypersurface.
alkyl groups. Furthermore, the late character of the TS1 and
TS2 transition structures is supported by the observation that
the measured ∆H‡ value is not much lower than the ∆H‡
hydrogens. A surplus of energy is needed to remove the H218O
moiety far enough to establish appreciable interactions with the
solvent cage and to promote IIS f IIW dissociation. As a
consequence, IIS f IS internal return can efficiently compete
with H218O diffusion to the aqueous cage (k-1 > kdiss). Besides,
the shielding effect of the H218O leaving group accounts for
the observed prevalence of the inversion of configuration in
the H2O-to-H218O exchange in solution.2
inv
diss
one.
In the solvolysis of 1S in aqueous acids (Scheme 1), the
removed CH3OH molecule in TS1 and TS2 is close enough to
the solvent cage to feel the effects of electrostatic interactions
with the H2O molecules. These attractive electrostatic forces
are expected to lower the IIS f IIW dissociation barrier relative
to that of the IIS f IS internal return which should feel much
less the effects of the solvent shell (k-1 < kdiss in Scheme 1).3
Taking into account the factors governing inversion of
configuration in the strictly related chiral allylic alcohols in the
gas phase18,19 and in solution,2,3 the IS f IR (XOH ) H218O)
transition structure is instead expected to be placed much earlier
along the reaction coordinate so as to resemble the starting IS
ion more than the hydrogen-bonded intermediates corresponding
to IV and V of Figure 2. In it, the moving H218O, less basic
than CH3OH,7 sits nearby the departure face of the still flexible
benzylic residue and does not appreciably interact with its acidic
In conclusion, hindered inversion of IS (XOH ) CH3OH) in
solution is attributed to the operation of attractive interactions
between the moving CH3OH moiety and the solvent cage which
inhibit internal return in the intimate ion-dipole pair IIS
(XOH ) CH3OH). Similar interactions do not operate in the
solvolysis of 18O-labeled 1S in aqueous acids, whose loss of
optical activity efficiently competes with exchange of the leaving
H218O with the solvent shell.
Acknowledgment. This work was supported by the Minis-
tero della Universita` e della Ricerca Scientifica e Tecnologica
(MURST) and the Consiglio Nazionale delle Ricerche (CNR).
(18) Troiani, A.; F.Gasparrini, F.; Grandinetti, F.; Speranza, M. J. Am.
Chem. Soc. 1997, 119, 4525.
(19) Troiani, A.; Speranza, M. J. Org. Chem. 1998, 63, 1012.
JA001960O