Wagner-Meerwein rearrangements in the gas phase: Anchimeric assistance to acid-induced dissociation of optically active phenylpropanols

Article abstract of DOI:10.1002/(SICI)1521-3765(19990301)5:3<834::AID-CHEM834>3.0.CO;2-G

The kinetics and the stereochemistry of Wagner-Meerwein rearrangements of O-protonated and O-methylated (S)-1-phenyl-2-propanol (1s(A)+; A=H or Me) and (S)-2-phenyl-1-propanol (2S(A)+; A=H or Me) have been investigated in the gas phase at 750 Torr and in the 25-140°C temperature range. The 1S(A) and 2S(A)+ intermediates were generated in the gas phase by reaction of the C(n)H₅+(n=1, 2; A= H) and (CH3)₂F+ ions (A - Me), formed by stationary γ radiolysis of bulk CH₄ and CH₃F, respectively, with the corresponding optically active alcohols. The results are consistent with unimolecular H₂O loss from both 1s(H)+ and 2s(H); this is anchimerically assisted by all the groups adjacent to the leaving moiety. Anchimeric assistance appears much less efficient in both 1S(Me)+ and 2S(Me)+. Analysis of the activation parameters indicates that competing neighboring-group participation in C-O bond fission in 1s(H)+ and 2s(H) respond essentially to entropic rather than enthalpic factors. The stereochemical distribution of the reaction products allowed us to discern between backside and frontside phenyl-group participation in 1s(H+). The counterintuitive observation of a frontside Ph participation, with an activation energy 1.3±0.5 kcal mol^-1 lower than that of the accompanying backside assistance, is attributed to conformational factors and to the stabilizing electrostatic interactions between the phenonium ion and the leaving H₂O complex that is spatially allowed only in the frontside participation and forbidden in the backside one.

Full text of DOI:10.1002/(SICI)1521-3765(19990301)5:3<834::AID-CHEM834>3.0.CO;2-G

FULL PAPER
Chiral Ions in the Gas Phase, Part 6 [= ]
Wagner± Meerwein Rearrangements in the Gas Phase: Anchimeric Assis-
tance to Acid-Induced Dissociation of Optically Active Phenylpropanols
Maurizio Speranza* and Antonello Filippi[a]
Abstract: The kinetics and the stereo-
chemistry of Wagner± Meerwein re-
arrangements of O-protonated and
O-methylated (S)-1-phenyl-2-propanol
results are consistent with unimolecular
H O loss from both 1s and 2sH ; this is
anchimerically assisted by all the groups
adjacent to the leaving moiety. Anchi-
meric assistance appears much less effi-
rather than enthalpic factors. The ster-
eochemical distribution of the reaction
products allowed us to discern between
backside and frontside phenyl-group
‡
H
‡
2
(
1
1sA
-propanol (2sA
been investigated in the gas phase at
50 Torr and in the 25 ± 1408C temper-
ature range. The 1sA and 2sA inter-
‡
; A ˆH or Me) and (S)-2-phenyl-
participation in 1sH . The counterintui-
‡
‡
; A ˆH or Me) have
cient in both 1sMe
‡
and 2sMe
‡
. Analysis of
tive observation of a frontside Ph par-
ticipation, with an activation energy
the activation parameters indicates that
competing neighboring-group participa-
�
1
7
1.3 Æ0.5 kcalmol lower than that of
the accompanying backside assistance, is
attributed to conformational factors and
to the stabilizing electrostatic interac-
tions between the phenonium ion and
‡
‡
tion in C�O bond fission in 1s
‡
and
respond essentially to entropic
H
mediates were generated in the gas phase
2s
‡
H
‡
by reaction of the C H (n ˆ1, 2; A ˆ
n
5
‡
H) and (CH ) F ions (A ˆMe), formed
3
2
Keywords: anchimeric assistance ´
chirality gas-phase chemistry
phenonium ions ´ stereochemistry
by stationary g radiolysis of bulk CH
the leaving H O complex that is spatially
4
2
´
´
and CH F, respectively, with the corre-
allowed only in the frontside participa-
tion and forbidden in the backside one.
3
sponding optically active alcohols. The
Introduction
Stable p-bridged alkenearenium ions I (Scheme 1) were first
proposed in 1942 by Cram as intermediates in the solvolysis of
[
1, 2]
optically active b-arylalkyl tosylates.
This hypothesis was
questioned in 1962 by Brown, who instead interpreted Cramꢀs
observations in terms of the weakly p-bridged, rapidly
[
3]
equilibrating open ions II. Later on, evidence was brought
forth for a continuous spectrum of intermediates in the
solvolysis of b-arylalkyl systems, spanning structures from I
through to III, depending upon the nature of the solvent and
the degree of substitution in the precursor.[4] The same factors
intervene as well in the sensitive balance between anchimeric-
assisted (k ), solvent-assisted (k ), and unassisted (k ) path-
D
S
C
[
5]
ways in b-arylalkyl solvolysis.
[
a] Prof. M. Speranza, Dr. A. Filippi
Facolt aÁ di Farmacia, Dipartimento No. 64
Chimica e Tecnologia delle Sostanze Biologicamente Attive)
(
Universit aÁ di Roma La Sapienza, P. le A. Moro 5, I-00185 Rome (Italy)
Fax: (‡39)6-49913602
Scheme 1. Structures of phenonium ions and other isomers.
E-mail: speranza@axrma.uniroma1.it
The unsubstituted phenonium ion Ia (Scheme 1) was
[=] Part 5: A Filippi, M. Speranza, Int. J. Mass Spectrom. Ion Proc., in
press.
directly observed by Olah and co-workers in SbF /SO ClF at
5
2
8
34
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0503-0834 $ 17.50+.50/0
Chem. Eur. J. 1999, 5, No. 3

8
34±844
Tꢀ�308C and spectroscopically characterized as the classi-
and mechanistic investigation about the gas-phase unimolec-
[
6, 7]
[11, 12]
cal spiro[5.2]octa-5,7-diene-4-ylium ion.
Above this tem-
ular dissociation of cationized b-arylalkyl systems.
The
perature, quantitative Ia !IVa isomerization via the open-
results point to gaseous alkenearenium ions as stable inter-
mediates that do not readily isomerize to the more stable
[
6c]
chain structure IIIa was observed. In these media, however,
the kinetics and mechanism of formation of Ia escaped
determination owing to the elusive nature of its precursors.
Any attempt to detect the methyl-substituted homologues of
Ia, namely the Is ‡Ir racemate, in SbF /SO ClF solutions
open-chain structures IV.[
11, 12]
These open-chain isomers were
found to arise from independent dissociation pathways,[
whose extent and mechanism remain obscure. Besides, no
information is presently available as to the static (I) or rapidly
interconverting (II) character of the alkenearenium ions
intermediates in the gas phase .
11]
5
2
invariably failed. The only observable species were the benzyl
[
8]
ion IVr in rapid equilibrium with IIr. Such a failure is
attributable to the combined effects of side-chain substitution
and of specific ion solvation that stabilize IIIr and IVr, with a
large fraction of the charge adjacent to the substituent, more
than Is ‡Ir, with most of the charge dispersed over the
aromatic ring and far from the substituent. As a result, a
decrease of the energy gap between Is ‡Ir and IVr is
expected in the absence of solvent stabilization, together with
a parallel increase of the activation barrier for their inter-
conversion. The lower activation energy of the Ia !IVa
To clear up these aspects, we decided to undertake a
detailed kinetic and stereochemical study of the acid-cata-
lyzed rearrangement in some chiral b-phenyl propanols, that
is (S)-1-phenyl-2-propanol (1s) and (S)-2-phenyl-1-propanol
(2s), and the racemates of 1-phenyl-1-D-2-propanol (1d) and
2-phenyl-2-D-1-propanol (2d), in gaseous inert media con-
16
18
taining an external nucleophile (Nu OH or Nu OH; Nu ˆH,
CH ; Scheme 2). The kinetic approach adopted, which has
3
�
1
[6c]
rearrangement in superacidic media (E* 13 kcalmol ),
�
1
relative to the estimated about 20 ± 25 kcalmol in the
isolated state on the grounds of ab initio calculations, supports
[
9]
this expectation.
In this context, and in view of the considerable interest in
the role of ion ± molecule complexes involved in gas-phase
[
10]
analogues of solvolysis reactions, a sustained research effort
was directed in the last decades to a comprehensive kinetic
Abstract in Italian: La cinetica e la stereochimica del
riarrangiamento di Wagner ± Meerwein in (S)-1-fenil-2-propa-
nolo (1sA
‡
; A ˆH o Me) e (S)-2-fenil-1-propanolo (2sA ; A ˆ
‡
H o Me) protonati e metilati allꢀatomo di ossigeno sono state
studiate in fase gassosa a 750 Torr nellꢀintervallo di tempera-
tura da 25 a 1408C. Gli intermedi 1sA
‡
e 2s
‡
sono stati generati
A
‡
in fase gassosa per attacco di ioni C H (n ˆ1, 2) (A ˆH) e
n
5
‡
(
CH ) F (A ˆMe), generati rispettivamente dalla g-radiolisi
3
2
stazionaria dei bulk gas CH e CH F, sui corrispondenti alcooli
4
3
enantiomericamente puri. I risultati sono in accordo con un
meccanismo unimoleculare di rilascio di H O da 1s e 2sH
anchimericamente assistito da tutti i gruppi adiacenti al gruppo
uscente. Lꢀassistenza anchimerica appare molto meno efficace
Scheme 2. Reaction patterns of O-protonated (S)-1-phenyl-2-propanol,
‡
H
‡
2
1
-phenyl-1-deutero-2-propanol, (S)-2-phenyl-1-propanol, and 2-phenyl-2-
deutero-1-propanol in the presence of methanol.
nel rilascio di MeOH da 1sMe
di attivazione indica che lꢀassistenza anchimerica in 1sH
eÁ governata da fattori entropici, piuttosto che da fattori
entalpici. La distribuzione stereochimica dei prodotti di
‡
e 2sMe
‡
. Lꢀanalisi dei parametri
e 2sH
been recently reviewed,[13] is based upon the preparation of
‡
‡
‡
‡
stationary concentrations of gaseous Brùnsted [GA ˆC H
n
5
‡
‡
2
(
(
n ˆ1,2)] and Lewis acids [GA ˆ(CH ) F ] by g radiolysis
0
3
6
Co source) of bulk gases, such as CH and CH F (750 Torr).
reazione da 1sH
‡
ci permette di discernere fra lꢀassistenza del
4
3
‡
Attack of GA on the oxygen atom of the alcoholic substrate,
for example, 1s present in traces (0.3 ± 0.5 Torr) in the gaseous
mixture, is expected to generate the corresponding oxonium
gruppo fenile vicinale con inversione del centro di reazione e
quella che avviene con ritenzione del centro di reazione.
Lꢀosservazione controintuitiva che questꢀultimo tipo di parte-
ion, that is, 1sA
‡
(A ˆH or Me), wherein the loss of the leaving
cipazione del Ph avvenga con una energia di attivazione
minore di 1.3 Æ0.5 kcal mole�1 rispetto a quella riguardante la
moiety (AOH) may be assisted by the participation of the
adjacent groups.
partecipazione con inversione di configurazione eÁ attribuita a
fattori conformazionali ed alle interazioni elettrostatiche
The aim of this study is to gather more information on the
gas-phase Wagner± Meerwein rearrangements in the chiral
stabilizzanti fra il gruppo uscente (H O) e la carica positiva
dello ione fenonio, essenzialmente localizzata sullꢀanello a 6
termini. Tali interazioni sono spazialmente favorite quando la
partecipazione del gruppo Ph in 1sH avviene con ritenzione
del centro di reazione.
2
oxonium ions 1sA and 2sA
‡ ‡
, and on the relevant activation
parameters. It is hoped thereby to elucidate the nature of the
intermediates involved in the rearrangements, as well as the
role of the leaving AOH group in determining the kinetics and
‡
the dynamics of the 1sA
‡
and 2sA rearrangements.
‡
Chem. Eur. J. 1999, 5, No. 3
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0503-0835 $ 17.50+.50/0
835

M. Speranza, A. Filippi
FULL PAPER
‡
18
Table 1. Product yield and distribution from the gas-phase attack of C
n
H
5
‡(n ˆ1, 2) and (CH
3
)
2
F
ions on 1s in the presence of Nu OH (Nu ˆH, CH
3
).
[
a]
18
[b]
System composition
Temperature
Relative product yields [%] ( O%)
Total absolute
1
8
[c]
Bulk gas
Substrate [Torr] Nu OH [Nu, Torr] [8C]
3s
3r
5s
5r
yield G(M)
CH
CH
CH
CH
CH
CH
4
4
4
4
4
4
1s, 0.35
1s, 0.38
1s, 0.37
1s, 0.47
1s, 0.45
1s, 0.32
CH
CH
CH
CH
CH
CH
3
3
3
3
3
3
, 2.50
, 1.54
, 1.46
, 1.42
, 1.34
, 1.40
25
60
60
100
120
140
48.0(90)
60.4(93)
62.0(93)
75.7(94)
86.3(94)
89.8(93)
7.0(90)
8.8(94)
8.0(94)
8.5(94)
7.9(94)
7.0(93)
22.5(94)
15.3(94)
15.0(94)
8.0(94)
2.9(91)
1.6(94)
22.5(94)
15.5(94)
15.0(94)
7.8(94)
2.9(90)
1.6(93)
0.08 (3)
0.12 (4)
0.22 (8)
0.20 (7)
0.18 (7)
0.20 (7)
CH
CH
CH
CH
3
3
3
3
F
F
F
F
1s, 0.34
1s, 0.38
1s, 0.38
1s, 0.44
H, 1.34
H, 3.44
H, 2.61
H, 3.18
25
60
85
98.0(<1)
98.5(<1)
100.0(<1)
100.0(<1)
nd
nd
nd
nd
1.0(36)
0.7(41)
nd
1.0(36)
0.8(41)
nd
0.47
0.45
0.48
0.40
120
nd
nd
4
4
�1
[
a] Bulk gas: 750 Torr; O
2
: 4 Torr. Radiation dose 2 Â10 Gy (dose rate: 1 Â10 Gyh ); [b] Expressed as the percent ratio of the methylated products.
18
Formation of alcohols 2r, 2s, and 1r and ethers 4r, 4s, and 6 were below the detection limit (nd < 0.2%; nd ˆnot detected). The O content is given in
parentheses; [c] G(M) as the number of molecules M produced per 100 eV of absorbed energy. Each value is the average of several determinations, with an
uncertainty level of approximately 5%. The percent ratios of the overall G(M) values to the G(GA‡) of the acid precursors are given in parentheses.
‡
18
Table 2. Product yield and distribution from the gas-phase attack of C
n
H
5
‡(n ˆ1, 2) and (CH
3
)
2
F
ions on 2s in the presence of Nu OH (Nu ˆH, CH
3
).
[
a]
18
[b]
System composition
Temperature
Relative product yields [%] ( O%)
Total absolute
1
8
yield G(M) [%][c]
Bulk gas
Substrate [Torr] Nu OH [Nu, Torr] [8C]
1s
3s
4s
5s
5r
6
CH
CH
CH
CH
CH
4
4
4
4
4
2s, 0.41
2s, 0.40
2s, 0.41
2s, 0.42
2s, 0.51
CH
CH
CH
CH
CH
3
3
3
3
3
, 1.50
, 1.63
, 1.56
, 1.61
, 1.71
25
60
85
85
120
1.3(<1) 53.6(93)
4.6(<1) 72.1(94)
8.4(<1) 78.8(96)
7.9(<1) 78.5(95)
10.4(<1) 81.5(94)
nd
nd
nd
nd
nd
16.8(93)
9.8(94)
5.5(96)
5.8(95)
3.6(94)
16.8(94)
9.7(94)
5.5(96)
5.8(95)
3.6(94)
11.5(94)
3.7(94)
1.8(94)
2.0(94)
0.9(90)
0.10 (4)
0.19 (7)
0.18 (7)
0.10 (4)
0.18 (7)
CH
CH
CH
CH
3
3
3
3
F
F
F
F
2s, 0.34
2s, 0.34
2s, 0.38
2s, 0.35
H, 2.55
H, 3.31
H, 2.61
H, 2.77
25
60
85
nd
nd
nd
nd
nd
nd
4.0(7)
10.0(6)
100.0(<1) nd
100.0(<1) nd
96.0(<1) nd
90.0(<1) nd
nd
nd
nd
nd
nd
nd
nd
nd
0.30
0.22
0.17
0.10
140
4
4
�1
[
a] Bulk gas: 750 Torr; O
2
: 4 Torr. Radiation dose 2 Â10 Gy (dose rate: 1 Â10 Gyh ); [b] Expressed as the percent ratio of the products. Formation of
18
alcohols 1r and 2r and ethers 3r and 4r were below the detection limit (nd < 0.2%; nd ˆnot detected). The O content is given in parentheses; [c] see
footnote [c] of Table 1.
The irradiations were carried out at constant temperatures ranging from 25
Experimental Section
60
4
to 1408C with a Co g source to a dose of 2 Â10 Gy at a rate of 1 Â
4
�1
1
0 Gyh , as determined by a neopentane dosimeter. Control experiments,
Materials: Methane, methyl fluoride, oxygen, and trimethylamine were
4
5
carried out at doses ranging from 1 Â10 to 1 Â10 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 a 25 m long,
high purity gases from Matheson, and were used without further
18
18
18
18
purification. H
2
O ( O > 97%) and CH
3
OH ( O ˆ95%) were pur-
chased from ICON Services. (S)-1-phenyl-2-propanol (1s) and its R
enantiomer (1r), (S)-2-phenyl-1-propanol (2s) and its R enantiomer (2r),
0
.25 mm inner diameter, MEGADEX DACTBS-b (30% diacetyl-tert-
(
S)- and (R)-1-phenyl-1-propanol, and 2-phenyl-2-propanol were research
butylsilyl-b-cyclodextrin in OV 1701) fused silica column operated at
grade chemicals from Aldrich. The (R,R)/(S,S) racemate of 1-phenyl-1-
deutero-2-propanol (1d) was synthesized, together with the (R,S)/(S,R)
racemate of 1-phenyl-2-deutero-1-propanol, by the reaction of trans-b-
�1
temperatures ranging from 40 to 1708C, 48C min . The products were
identified by comparison of their retention volumes with those of authentic
standard compounds, and their identity was confirmed by GLC ± MS, with a
Hewlett ± Packard 5890A gas chromatograph in line with a HP 5970B mass
spectrometer (GC ± MS). Their yields were determined from the areas of
the corresponding eluted peaks, with the internal standard (i.e., benzyl
alcohol) method, and individual calibration factors to correct for the
detector response. Blank experiments were carried out to exclude the
occurrence of thermal isomerization and racemization of the starting
alcohols, as well as of the corresponding methyl ethers within the
temperature range investigated.
4
methyl-styrene with LiAlD and subsequent oxidation with hydrogen
peroxide.[ The same procedure was employed to prepare the (R)/(S)-
racemate of 2-phenyl-2-deutero-1-propanol (2d) from a-methyl-styrene.
The deuterium content of 1d and 2d (>98%) was determined by GLC ±
MS analysis. The methyl ethers of the above alcohols, namely, (S)- (3s) and
14]
(
R)-1-phenyl-2-methoxypropane (3r), (S)- (4s) and (R)-2-phenyl-1-me-
thoxypropane (4r), (S)- (5s) and (R)-1-phenyl-1-methoxypropane (5r),
and 2-phenyl-2-methoxypropane (6) were synthesized by the dimethyl
sulfate method, and their configuration assigned according to the starting
alcohol.[ The alcoholic substrates 1s, 2s, 1d, and 2d were purified by
preparative GLC on a 2 m long, 4 mm inner diameter, stainless steel
column, packed with 10% SP-1000 on 100 ± 120 Supelcoport, at 1808C.
Their final purity exceeded 99.95%. The identity of the above alcohols and
of their methyl ethers was verified by NMR spectroscopy and their purity
assayed by GLC and GLC ± MS on the same columns employed for the
analysis of the irradiated mixtures.
15]
Results
Tables 1 and 2 report the absolute and relative yields of the
products formed from alcohols 1s and 2s, respectively,
undergoing gas-phase attack from the radiolytically generated
‡
GA acids in the presence of O , as a thermal radical
2
Procedure: The gaseous mixtures were prepared by conventional techni-
ques, with the use of a greaseless vacuum line. The reagents and the
additives were introduced into carefully outgassed 130 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.
18
18
scavenger, and of either H2 O or CH3 OH, as the nucleo-
phile. The figures in the tables represent the mean percent
distribution of the products, as obtained from several separate
irradiations carried out under the same experimental con-
8
36
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0503-0836 $ 17.50+.50/0
Chem. Eur. J. 1999, 5, No. 3

Wagner ± Meerwein Rearrangements
834±844
Table 3. Characteristic mass spectrometric peaks of the radiolytic products.
1
6
18
‡
‡
Radiolytic products
O/ O-[F]
m/z
H/D-[F]
m/z
‡
16
18
‡[c]
1
-Phenyl-2-propanols, rac-1
[CH
3
CHOH]
45 ( O) 47 ( O)
[PhCH
2
]
91 (H)
92 (H)
45 (H)
136 (H) 137 (D)
105 (H) 106 (D)
150 (H) 151 (D)
59 (H)
150 (H) 151 (D)
45 (H) 46 (D)
92 (D)
93 (D)
46 (D)
.
‡[a]
16
18
‡[b][c]
[
PhCH
2
CH(Me)OH]
136 ( O) 138 ( O)
[C
CH
[PhCH(Me)CH
7
H
8
]
‡
[
3
CHOH]
.
‡[a]
16
18
.‡[a]
2
1
2
1
2
-Phenyl-1-propanols, rac-2
[PhCH(Me)CH
2
OH]
136 ( O) 138 ( O)
2
OH]
‡
3
PhCHCH ]
[d]
[
‡[b]
16
18
.‡[a]
.‡[a]
.‡[a]
-Phenyl-2-methoxypropanes, rac-3
-Phenyl-1-methoxypropanes, rac-4
-Phenyl-1-methoxypropanes, rac-5
-Phenyl-2-methoxypropane, 6
[CH
3
OCHCH
3
]
59 ( O) 61 ( O)
[PhCH
2
CH(Me)OCH
OCHCH
[PhCH(Me)CH OCH
2
3 2
[CH OCH ]
[PhCH(OMe)CH
[PhCHOCH
3
]
3
]
3
]
‡[d]
[CH
3
3
]
60 (D)
.
.
‡[a]
16
18
[PhCH(Me)CH
2
OCH
CH
3
]
]
150 ( O) 152 ( O)
‡
16
18
‡[d]
[
CH
3
OCH
2
]
45 ( O) 47 ( O)
‡[a]
16
18
[PhCH(OMe)CH
2
3
150 ( O) 152 ( O)
2
CH
150 (H) 151 (D)
121 (H) 122 (D)
135 (H) 136 (D)
‡
[b]
16
18
‡[b][d]
[
PhCHOCH
3
]
121 ( O) 123 ( O)
3
]
‡
[b]
16
18
‡[b][d]
3
[PhC(Me)OCH
3
]
135 ( O) 137 ( O)
[PhC(Me)OCH ]
[
a] Molecular ion. [b] Base peak. [c] Deconvolution of the m/z ˆ91 ± 93 triplet allows determination of the extent of deuteration at the C1 center of the
product. [d] The heavier fragment from the heavier molecular ion denotes location of the label at its C center. The lighter fragment from the heavier
molecular ion denotes location of the label at the other C center.
ditions, and whose reproducibility is expressed by the
uncertainty level quoted. The absolute yields of products
are expressed as the number of molecules of the product M
formed per 100 eV of energy absorbed by the gaseous mixture
[rac-5] yield ratios are found to increase from ꢂ7 to ꢂ13 and
from ꢂ1 to ꢂ28, respectively, by increasing the temperature
from 25 to 1408C.
The 18O-labeled ether 3s ( O content > 93%) predomi-
nates (53.6 ± 81.5%) among the products from the irradiated
systems with 2s (Table 2). In this case, the R enantiomer 3r is
never observed despite an accurate search. Instead, ether 3s is
always accompanied by minor amounts of 2-phenyl-2-me-
thoxypropane 6 (0.9 ± 11.5%), of the racemate of 1-phenyl-1-
methoxypropane rac-5 (7.2 ± 33.6%), and of (S)-1-phenyl-2-
propanol 1s (1.3 ± 10.4%). Also in this case, the product
distribution depends upon the reaction temperature, with the
[3s]/[rac-5] and [3s]/[6] yield ratios significantly increasing
from 1.6 to 11.3 and from 4.7 to approximately 90.5,
respectively, by increasing the temperature from 25 to
1208C. In contrast, the [3s]/[1s] yield ratios decrease from
41.2 to 7.8 by the same temperature increase.
18
(
G(M) values). The percent ratios of the overall G(M) values to
the G(GA of their gaseous acid precursor are quoted in
‡
)
[
16]
parentheses.
The ionic character of these reactions is
demonstrated by the sharp decrease of the overall product
yields (over 80%) caused by the addition to the gaseous
mixture of 0.5 mol% of NMe , an efficient positive ion
3
interceptor.
The presence of the 18O label in the radiolytic products
reported in Tables 1 and 2 is determined from 70 eV mass
spectra. The 18O content is measured from the intensity of the
signals of their molecular ion, when present, and of the
‡
O-containing fragments (henceforth both denoted as [F] ).
‡
18
The [F] peaks used for evaluating the extent of O-labelling
of the radiolytic products are listed in the first half of Table 3.
Since the mass spectra of the unlabeled products in Table 3 do
not display any detectable signals at masses 2 amu above that
Ancillary experiments were carried out to gain information
on the origin of the inverted (S)-1-phenyl-2-propanol (1s)
recovered among the products from (S)-2-phenyl-1-propanol
‡
‡
18
of [F] (i.e., [F‡2] ) the percent of O incorporation into the
products is simply calculated from the corresponding 100 Â
(2s; Table 2). Thus, mixtures containing 2s in bulk CH were
4
irradiated at room temperature in the presence of added
‡
‡
‡
18
18
[
F‡2] /([F] ‡[F ‡2] ) ratio. The relevant O-incorporation
H2 O, as the external nucleophile. The significant incorpo-
18
values are reported in Tables 1 and 2 in parentheses.
ration of the O label in the 1s product (ꢀ40%) points to a
prevailing acid-induced 2s !1s intermolecular isomerization
Inspection of Tables 1 and 2 reveals that, in the CH F/H218O
3
pathway involving uptake of the external H218O. The [F]
[17]
‡
mixtures, the predominant product is always the unlabeled
1
8
methyl ether of the starting alcohol ( O content < 1%),
sometimes accompanied by very minor amounts of other
isomers. At temperatures ꢀ608C, ether 3s is obtained from
the starting alcohol 1s together with ꢀ2% of the racemate of
peaks used for evaluating the extent and the specific position
of the deuterium label in the 1-phenyl-1-methoxypropanes
rac-5, obtained from gas-phase protonation of the 1-phenyl-1-
deutero-2-propanol (1d) and 2-phenyl-2-deutero-1-propanol
(2d) racemates, are shown in the second half of Table 3.
Unlabeled rac-5 and their D-labeled counterparts, that is,
1-phenyl-1-deutero-1- and 1-phenyl-2-deutero-1-methoxy-
propanes, display very similar mass spectrometric patterns
that exclude any significant isotope effect on their fragmen-
tation. On these grounds, the D content in these products has
been calculated from the molecular ion and the
1
8
1
-phenyl-1-methoxypropane (rac-5; O-content ˆ36 ± 41%;
Table 1). In contrast, at temperatures ꢁ858C, ether 4s is
formed from 2s together with ꢀ10% of (S)-1-phenyl-2-
1
8
methoxypropane (3s; O content < 7%; Table 2).
More complex product patterns arise from the irradiated
1
8
18
CH /CH3 OH systems. Thus, abundant formation of the O-
4
1
8
methylated substrate (3s; O content > 90%) is observed
from the starting alcohol 1s, accompanied by minor amounts
of its enantiomer (3r; 7.0 ± 8.8%) and of the racemate of
‡
[CH OCHCH ] peak intensities after correction for the
3
3
natural isotopic contributions. The relevant results are
reported in Table 4.
1-phenyl-1-methoxypropane (rac-5; 3.2 ± 45.0%; Table 1).
The relative distribution of these products depends upon the
reaction temperature. In fact, both the [3s]/[3r] and the [3s]/
Analysis of Table 4 indicates that the 1-phenyl-1-methoxy-
propanes rac-5 formed from 1d and 2d retain most of the D
Chem. Eur. J. 1999, 5, No. 3
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0503-0837 $ 17.50+.50/0
837

M. Speranza, A. Filippi
FULL PAPER
Table 4. Extent and position of labelling in the rac-5 from deuterated starting CH 18OH /CH 16OH2‡pair, in proportions depending upon
‡
3
2
3
16
alcohols 1d and 2d.[
a]
18
the [H2 O]/[H2 O] ratio. Therefore, the nature and the
Deuterated products [%][b] D content relative distribution of the acidic species, generated in the
Substrate Temperature Nucleophile
[
Torr]
[8C]
CH
3
16OH [Torr] 2-D-rac-5 1-D-rac-5
[%]
irradiated samples, are determined by the nature of the bulk
gas and by presence and the relative concentration of the
nucleophiles present in the mixture. All these acidic species
can eventually attack the selected substrates, provided that
the process is thermochemically allowed.
1
2
2
2
2
d, 0.25
d, 0.27
d, 0.28
d, 0.26
d, 0.29
25
25
60
85
120
1.27
1.23
1.06
1.03
1.25
41.8
40.0
44.6
50.5
59.9
58.2
60.0
55.4
49.5
40.1
87
98
98
93
96
The evaluation of the thermochemistry of the reaction
sequences of Scheme 2 meets with some difficulty owing to
the lack of experimental thermochemical data for the
4
[
1
a] Bulk gas: 750 Torr; O
2
: 4 Torr. Radiation dose 2 Â10 Gy (dose rate: 1 Â
4
�1
0 Gyh ). Each value is the average of several determinations, with an uncertainty
level of ca.10%; [b] 2-D-rac-5 denotes the mixture of the two racemates of
diastereomeric 1-phenyl-2-deutero-1-methoxypropanes; 1-D-rac-5 denotes the
racemate of 1-phenyl-1-deutero-1-methoxypropane.
involved ionic species, for example, 1sH (or 2sH ) and 1sMe
‡ ‡ ‡
(or 2sMe ), and neutral substrates, that is, 1s and 2s. However,
‡
approximate values of the relevant reaction enthalpies can be
inferred from the thermochemical data reported in Table 5,
label present in the starting alcohol, thus, ruling out any
significant D exchange with the gaseous reaction medium
within the 25 ± 1208C temperature range. Labeled rac-5
exhibit the deuterium label at both C1 (1-D-rac-5) and C2
centers (2-D-rac-5). While this is not unexpected when arising
from 1d, the formation of 1-phenyl-2-deutero-1-methoxypro-
pane (2-D-rac-5) from 2d, in yields comparable with those of
Table 5. Thermochemical data [kcalmol 1; estimated values in italics].
�
o
f
o
f
Species
CH
DH
Species
DH
[
a]
[b]
[b]
4
�17.8
1s
2s
�33
�33
[a]
C
2
H
4
12.5
[
[
[
a]
a]
a]
[c]
[c]
[d]
[d]
[e]
[a]
[f]
H O
2
�58
�48
�59
1sH
‡
136
139
130
133
199
186
191
1-phenyl-1-deutero-1-methoxypropane (1-D-rac-5) and sen-
CH
CH
CH
3
3
5
OH
2sH‡
1sMe‡
2sMe‡
Ir
IVr
IVs
F
sitive to the reaction temperature, is rather surprising.
‡
[a]
216
‡
[a]
C
2
H
5
215.6
‡
[a]
(
CH
3
)
2
F
147
136
‡
[a]
CH OH
3
2
Discussion
[
a] S. G. Lias, J. E. Bartmess, J. F. Liebman, J. L. Holmes, R. D. Levin, W. G.
‡
Gas-phase GA attack on the alcoholic substrates: The very
Mallard, J. Phys. Chem. Ref. Data 1988, 17, Suppl. 1. [b] Estimated by the
group additivity method (S. W. Benson, Thermochemical Kinetics, Wiley,
New York, 1968. [c] Estimated from the proton affinity (PA) limits of
primary and secondary alcohols (194 and 197 kcalmol , respectively); J.
Long, B. Munson, J. Am. Chem. Soc. 1973, 95, 2427. [d] Estimated by
low concentrations of alcohols 1s and 2s (<0.1 mol%) and of
the nucleophile (0.2 ± 0.5 mol%) present in the gaseous
mixtures exclude their direct radiolysis as a significant route
to the products listed in Tables 1 and 2. Addition of an
�
1
�1
DPA ˆPA(methyl ether) �PA(alcohol) ˆ10 kcalmol
(R. D. Bowen,
efficient thermal radical scavenger (i.e., O ) in about tenfold
D. H. Williams, J. Am. Chem. Soc. 1978, 100, 7545; R. D. Bowen, D. H.
Williams, G. Hvistendahl, J. R. Kalman, Org. Mass Spectrom. 1978, 13,
2
excess over the substrate inhibits possible free-radical reac-
tion pathways in favor of the ionic ones, whose large
predominance is testified by the marked effect of an ion trap,
7
21). The heats of formation of the neutral methyl ethers have been
calculated as in footnote [b]. [e] Estimated by correcting the enthalpy of
formation of Ia (204 kcalmol ; M. Mishima, Y. Tsuno, M. Fujio, Chem.
�1
such as NMe , on the overall product yield. The trace
Lett. 1990, 2277) for the stabilizing effect of the methyl substituent on the
3
concentration of the added nucleophile implies that all the
ionic species generated from the attack of the GA acids on
the substrates must undergo many unreactive collisions with
the bulk gas and, therefore, be thermalized prior to reaction
with the neutral species present.
cyclopropyl ring, taken as equal to that on the cyclopropylcarbinyl cation
�
1
‡
(ꢂ5 kcalmol ; W. J. Hehre, P. Hiberty, J. Am. Chem. Soc. 1974, 96, 302).
f] Estimated by correcting the enthalpy of formation of a-phenylethyl
[
�
1
cation (199 kcalmol ; ref. [a]) for the stabilizing effect of the methyl
substituent at the C
�1
b
(ꢂ1 kcalmol ).
g Radiolysis of the bulk gas, either CH or CH F, generates
some of which were derived from the well-established
estimation procedures outlined in the footnotes of the table.
4
3
‡
‡
known yields of the C H (n ˆ1, 2) and (CH ) F acids,
n
5
3 2
o
respectively. Once collisionally thermalized, these ions effi-
ciently attack all the nucleophiles present in the mixture,
including H216O, which is a ubiquitous impurity present in the
From the relevant DH
f
values, the first steps of Scheme 2 with
‡
‡
‡
OH , and (CH ) F
3 2 3 2
‡
GA ˆC
H
5
(n ˆ1, 2), CH
are all
n
�1
thermochemically
allowed
[�DH8 ˆ12 ± 65 kcalmol ;
irradiated systems. As a consequence, in the CH /1s (or 2s)/
Eqs. (1) ± (8) in Table 6].
4
CH318OH mixtures, the initially formed C H (n ˆ1, 2)
‡
Unimolecular C�O bond cleavage in the primary oxonium
n
5
Brùnsted acids can attack either the alcoholic substrate,
yielding inter alia the corresponding O-protonated derivative
intermediates 1sH‡ (or 2sH‡) to give either IVr or IVs is
�
1
thermochemically allowed as well [�DH8 ˆ3 ± 11 kcalmol ;
[
henceforth denoted as 1s
‡
(or 2sH
‡
)], and the added
Eqs. (10), (11), (13), and (14) in Table 6], whereas the
H
18
16
CH3 OH (or the ubiquitous H O impurity), yielding even-
hypothetical one yielding separated Ir and H
2
O is slightly
2
1
8
‡
�1
tually the CH OH2 Brùnsted acid. Similarly, in the CH F/1s
endothermic [DH8 ˆ2 ± 5 kcalmol ; Eqs. (9) and (12) in
Table 6]. Nevertheless, taking into account the favorable
entropic factors, this latter process can also be regarded as
thermodynamically accessible. Another factor, particularly
relevant in gaseous systems at high pressures, concerns the
possibility that C�O bond cleavage in the oxonium inter-
3
3
‡
(
or 2s)/H218O systems, the initially formed (CH ) F Lewis
3 2
acid can attack either the alcoholic substrate, yielding the
corresponding O-methylated derivative [henceforth denoted
1
8
as 1sMe
‡
(or 2sMe )], or the added H2 O (and the ubiqui-
‡
tous H216O isotopomer), giving rise to the corresponding
8
38
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0503-0838 $ 17.50+.50/0
Chem. Eur. J. 1999, 5, No. 3

Wagner ± Meerwein Rearrangements
834±844
Table 6. Reaction enthalpies [kcalmol�1].
mechanism of Scheme 3 to any significant extent. A rationale
for such an inefficiency can be found in the likely formation of
a stable proton-bonded adduct V, whose rearrangement or
fragmentation before neutralization appears to be inhibited at
750 Torr by the rapid collisional quenching with the bulk gas
Neutral reactant Ionic reactant
Reaction products
DH8
‡
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
1s
1s
1s
1s
2s
2s
2s
2s
‡
‡
‡
‡
‡
‡
‡
‡
CH
5
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
1sH‡
1sH‡
1sH‡
1sMe‡
2sH‡
2sH‡
2sH‡
2sMe‡
Ir
‡
‡
‡
‡
‡
‡
‡
‡
‡
‡
‡
‡
‡
‡
‡
‡
‡
‡
‡
‡
‡
‡
CH
4
�65
�34
�15
�43
�62
�31
�12
�40
‡5
‡
C
2
H
5
C
2
H
4
‡
CH
3
OH
2
CH
CH
CH
3
3
4
OH
F
‡
(CH
3
‡
)
2
F
CH
5
‡
C
2
H
5
C
2
H
4
‡
CH
3
OH
2
CH
CH
3
OH
F
‡
(CH
1sH‡
1sH‡
1sH‡
2sH‡
2sH‡
2sH‡
3
)
2
F
3
H
H
H
H
H
H
2
2
2
2
2
2
O
O
O
O
O
O
1
1
1
1
1
1
1
1
1
1
2
2
2
IVs
IVr
Ir
�3
�8
‡2
IVs
IVr
Ir
IVs
IVr
Ir
IVs
IVr
1rMe‡
2sMe‡
�6
�11
‡21
‡13
‡8
1sMe‡
1sMe‡
1sMe‡
2sMe‡
2sMe‡
2sMe‡
1sH‡
CH
CH
CH
CH
CH
CH
3
3
3
3
3
3
OH
OH
OH
OH
OH
OH
‡18
‡10
‡5
CH
CH
3
3
OH ‡
OH ‡
H
H
2
O
O
�16
�16
2sH‡
2
Scheme 3. Hypothetical CH 1
reaction.
8OH
‡
-catalyzed intracomplex displacement
3
2
mediates gives rise to stable, electrostatically bonded adducts
‡
between the C H
and AOH fragments. In this way, the
9
11
enthalpy of 1sH
‡
(or 2sH
‡
) dissociation is lowered by a quantity
molecules. On these grounds, it can be concluded that the acid
‡
18
18
corresponding to the [C H ´ H O] binding energy (easily
exceeding 10 kcalmol ). In this case, fragmentation of 1sH
catalysts active in the CH /CH OH and CH F/H2 O mix-
4 3 3
9
11
2
�
1
[18]
‡
‡
‡
tures are the C H5 (n ˆ1, 2) and the (CH ) F ions,
n
3 2
(
or 2sH
accessible. The same considerations apply to the unimolecular
C�O bond cleavage in 1s
(or 2sMe ) [Eqs. (15) ± (20) in
Table 6]. Here, unimolecular conversion of 1s (or 2sMe
‡
) into the [Ir´ H O] complex becomes energetically
respectively.
2
In this context, assuming equal the collision frequencies
‡
‡
‡
between the radiolytic C H5 (n ˆ1, 2) ions and the neutral
Me
n
18
‡
‡
)
nucleophiles present in the CH /CH OH systems, that is,
4 3
Me
CH318OH, H216O, and the alcoholic substrate 1s (or 2s), the
into the [IVr´ CH OH] and [IVs ´ CH OH] complexes are
3
3
energetically feasible [Eqs. (16), (17), (19), and (20) in
theoretical overall absolute yield of products can be approx-
[
18]
‡
Table 6], whereas the 1sMe
appears thermodynamically forbidden [Eqs. (15) and (18) in
Table 6]. A CH OH-to-H O bimolecular pathway can be also
‡
(or 2sMe
‡
) ![Ir´ CH OH] one
imately expressed by {G(C H (n ˆ1,2)) Â[1s (or 2s)]}/
3
n
5
{[CH318OH] ‡[H O] ‡[1s (or 2s)]} ˆ0.2 ± 0.3.
16
[16, 17]
Thus,
2
3
2
the closeness of these estimates with the G values given
(M)
conceived to account for the formation of the ethereal
in Tables 1 and 2 indicates that trapping of the 1s
intermediates (and of their derivatives) by CH3 OH is a very
efficient process.
‡
(or 2sH )
‡
H
�
1
18
products in Tables 1 and 2 [�DH8 ˆ16 kcalmol ; Eqs. (21)
and (22) in Table 6].
Besides the oxygen atom, another basic site is present in the
selected substrates, namely, the phenyl ring, which may
‡
compete with the oxygen center for the gaseous GA acids.
The reaction pattern: The almost exclusive recovery of
18
For the purposes of the present work, ring protonation of the
aromatic substrate by Brùnsted GA acids is a parasitic,
unlabeled 3s from the CH F/1s/H O mixtures and of
3 2
‡
18
unlabeled 4s from CH F/2s/H2 O clearly demonstrates that
3
unproductive process. Occurrence of this blind reaction may
account for the limited absolute yields of the ethereal
products reported in Tables 1 and 2. Another conceivable
reason resides in the diverse aptitude of the various Brùnsted
the 1sMe
‡
and 2s
‡
intermediates, once formed from the
Me
‡
attack of (CH ) F at the oxygen atom of 1s and 2s, are very
stable gaseous species with no propensity towards rearrange-
ment or fragmentation. Only in the CH F/2s/H2 O systems at
the highest temperatures (Tꢁ858C; Table 2), it is possible to
observe the formation of small amounts (4 ± 10%) of mostly
unlabeled 3s. Taking into account that its formation from 2s
requires complete inversion of configuration of the chiral
center, it is suggested that the reaction proceeds through the
3
2
18
3
‡
GA acids, formed in the irradiated mixtures, to promote the
reaction patterns of Scheme 2. Indeed, in agreement with
[
11]
18
‡
previous evidence, the CH3 OH2 Brùnsted acid, which is
abundantly generated in both the CH and CH F mixtures,
4
3
appears to be rather ineffective in catalyzing any substrate
rearrangement, although able to O-protonate 1s (or 2s)
irreversible 2sA
rearrangement followed by proton loss to a suitable base
[sequence ii) !i) of Scheme 4]. At first glance, an analo-
gous intracomplex process with A ˆH could be taken as being
responsible for the formation of the unlabeled inverted 1s
‡
![Ir´ AOH] !1s
‡
(A ˆCH ) intracomplex
A
3
[
Eqs. (3) and (7) in Table 6]. A clear symptom of its
[19]
inefficiency is provided by the exceedingly low amounts of
1
8
the O-labeled products (ꢀ0.7%) formed in the CH F/1s/
3
1
8
18
H2 O (Table 1) and CH F/2s/H O systems at any temper-
3
2
18
ature (Table 2). This excludes occurrence of the intracomplex
product in the CH /2s/CH3 OH systems at any temperature
4
Chem. Eur. J. 1999, 5, No. 3
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0503-0839 $ 17.50+.50/0
839

M. Speranza, A. Filippi
FULL PAPER
Hence, the exclusive formation of the inverted ether 3s from
2
s points to the irreversible backside attack of the external
18
CH3 OH at the methyl-substituted center of Ir. In this
context, formation of the retained ether 3s from 1s (48.0 ±
8
9.8%; Table 1) must proceed through the two step sequence
i) !iv) of Scheme 4, each step involving inversion of the
configuration of the chiral center. In view of the complete
backside attack of the external CH318OH in step iv), formation
of the inverted ether 3r from 1s (7.0 ± 8.8%; Table 1)
necessarily involves retention of the configuration of the
chiral center in step i) of the sequence (i.e. 1sH
‡
!Is ‡H O).
2
In other words, the 3s versus 3r pattern from 1s (Table 1)
reflects the competition between backside versus frontside
participation of the phenyl group in the unimolecular C�O
bond cleavage in 1s
‡
to give the corresponding 1,2-propene-
H
benzenium enantiomers Ir and Is.
Concerning the genesis of the 1-phenyl-1-methoxypropane
racemate (rac-5) and 2-phenyl-2-methoxypropane (6) prod-
ucts reported in Tables 1 and 2, the question arises as to
whether it involves the participation of adjacent groups other
than phenyl, that is the hydrogen atom or the methyl group, in
the unimolecular C�O bond fission in 1s
‡
and 2sH , or
‡
H
whether it is due to the partial unimolecular isomerization of
the phenonium ion structure. Unequivocal assessment of this
point is crucial for rationalizing the factors affecting the
relative distribution of the ethereal products in Tables 1 and 2,
and their dependence upon the reaction temperature. Several
pieces of evidence rule out the latter hypothesis in favor of the
first one, including:
1) The evident difference in the product patterns from 1s
(Table 1) and 2s (Table 2) under identical experimental
conditions, which is not compatible with the intermediacy
of a single phenonium ion structure. Besides, the sizable
yields of rac-5 from 1s and the complete absence of 6
(Table 1), which is an appreciable product from 2s
(Table 2), points to a rather inefficient IVs $IVr inter-
conversion under the conditions used.
Scheme 4. Inter- and intramolecular reaction patterns.
(
Table 2). However, the observation that 1s produced in the
18
CH /2s/H O systems incorporates an appreciable amount of
4
2
1
8
the O label (ꢀ40%; see Results section) suggests that the
1
8
[Ir´ H O] complex may undergo extensive H O-to-H O
2
2
2
[
17]
ligand exchange [path iii) of Scheme 4]. The absence of
labeled 1s from the CH F/2s/H O systems indicates that an
analogous H218O-to-CH OH ligand exchange does not take
1
8
3
2
3
place in the [Ir´ CH OH] complex prior to intracomplex
3
covalent rebonding to 1sMe .
‡
Ether 3s is the major product from both the CH /1s/
4
CH 18OH (Table 1) and the CH /2s/CH OH systems (Ta-
18
3
4
3
ble 2) under all conditions. Its formation from 1s requires
retention of the configuration of the chiral center, whereas its
formation from 2s requires inversion of the configuration of
the chiral center. This observation rules out any mechanism
other than the unimolecular sequences i) !iv) and ii) !iv) of
2) The negative temperature dependence of the [rac-5]/[3r]
and [rac-5]/[3s] from 1s (Table 1) and of the [rac-5]/[3s]
and [6]/[3s] ratios from 2s (Table 2). Raising the temper-
ature would in fact favor rearrangement of the phenonium
ion structure to the thermodynamically more stable ones
IVr and IVs (Table 3).
Scheme 4 (A ˆH). In fact, the conceivable S 2 processes
N
[
Eqs. (21) and (22) in Table 6] would yield preferentially the
inverted 3r from 1s and the retained 4s from 2s. Furthermore,
the pronounced 3s versus 3r unbalance from 1s speaks
against the intermediacy of the open 1-phenyl-2-propyl cation
On these grounds, and in agreement with previous indirect
evidence,[ we conclude that the products of Table 1 are
generated by the reaction network of Scheme 5, involving
competing neighboring-group participation in the unimolec-
ular C�O bond cleavage in the 1sH‡ precursor.
11]
from simple C�O bond fission in 1s
‡
. In this case, trapping of
H
the free 1-phenyl-2-propyl cation by the present nucleophiles
would give rise eventually to the racemate of the correspond-
ing products (e.g., rac-3), in contrast with the experimental
evidence. The intermediacy of the primary 2-phenyl-1-propyl
A similar competition network (Scheme 6) accounts for the
formation of the products of Table 2. Thus, alcohol 1s and
ether 3s arise from phenyl-group participation in the uni-
molecular C�O bond cleavage in 2sH‡. This process competes
cation from simple C�O bond cleavage in 2s
‡
is excluded a
H
fortiori. Therefore, in agreement with previous indications,[
11]
with hydrogen-atom participation leading directly to ether 6.
Concerning the origin of ethers rac-5, methyl-group partic-
ipation in 2sH‡ does not explain by itself the observations
reported in Table 4. Indeed, methyl-group participation in the
racemate of O-protonated 2-phenyl-2-deutero-1-propanol
(2dH‡) would generate ethers rac-5 with the deuterium label
exclusively located at the C1 center, that is, 1-D-rac-5. The
the oxonium intermediates 1sH
‡
and 2sH
‡
undergo fast
unimolecular C�O bond fission assisted by the participation
of the vicinal groups. If the participating moiety is the phenyl
group, the process leads to a relatively stable phenonium ion.
Irrespective of the stereochemistry of the process, the 2s
alcohol only give rise to the retained phenonium ion Ir.
8
40
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0503-0840 $ 17.50+.50/0
Chem. Eur. J. 1999, 5, No. 3

Wagner ± Meerwein Rearrangements
834±844
of the products of Table 2 follow the reaction network shown
in Scheme 6. Accordingly, the 1-D-rac-5 versus 2-D-rac-5
relative distributions, reported in Table 4, can be taken in the
first approximation as reflecting the relative contribution at
any given temperature of the channels involving intermedi-
ates IVs and 1s
(
‡
in the formation of ethers rac-5 from 2sH
‡
H
Table 2).[
20]
Neighboring-group participation and anchimeric assistance:
In the context of the reaction networks of Schemes 5 and 6,
the relative rate constants of the competing neighboring-
group participation in 1sH
‡
and 2sH can be simply inferred
‡
from the yield ratios of the relevant labeled ethereal products.
Thus, the ratios [3s]/[rac-5], [3r]/[rac-5], and [3s]/[3r] from
back
front
back
front
Ph
Table 1 express the kPh /k , k /k , and k /k
rate
H
Ph
H
Ph
constant ratios, respectively, shown in Scheme 5. The corre-
Table 7. Rate constant ratios of intramolecular isomerization of 1sH‡ in
[
a]
CH
4
at 750 Torr.
back
front
front
back
Temperature [8C]
k
Ph /kPh
k
Ph /k
H
k
Ph /k
H
2
5
6.86 (0.863)
6.86 (0.863)
7.75 (0.889)
8.91 (0.950)
10.92 (1.038)
12.83 (1.108)
0.16 (�0.808)
0.29 (�0.544)
0.27 (�0.574)
0.54 (�0.269)
1.36 (0.134)
1.07 (0.028)
1.96 (0.292)
2.07 (0.315)
4.79 (0.680)
14.88 (1.173)
28.06 (1.448)
Scheme 5. Reaction pattern from gas-phase O-protonation of (S)-1-
phenyl-2-propanol in the presence of methanol.
60
60
00
20
40
1
1
1
2.19 (0.340)
[
a] Each value is the average of several determinations, with an uncertainty
back
front
front
back
level of ca.10%. kPh /kPh ˆ[3s]/[3r]; kPh /k
H
ˆ[3r]/[rac-5]; kPh /k ˆ
H
[
3s]/[ rac-5]. Logarithm of the ratios in parentheses.
sponding values are reported in Table 7 and their dependence
upon the reaction temperature (25 ± 1408C) is shown in
Figure 1. In the same way, the {[3s] ‡[1s] ‡[2-D-rac-5]}/[1-
D-rac-5] ratios from Tables 2 and 4, reflect the k /k rate-
Ph Me
Scheme 6. Reaction pattern from gas-phase O-protonation of (S)-2-
phenyl-1-propanol in the presence of methanol.
Figure 1. Arrhenius plots for the competing backside phenyl versus
hydride (circles), frontside phenyl versus hydride (diamonds), and backside
versus frontside phenyl participation (triangles) to the C�O bond cleavage
in 1sH‡
.
observation that these labeled ethers are accompanied by
those with the deuterium label at the C2 center (i.e., 2-D-rac-
5
) points to an additional route to ethers rac-5 from 2dH
position of the label in the 2-D-rac-5 from 2dH , coupled with
the isolation of appreciable amounts of the rearranged alcohol
s from 2sH , suggests that formation of both these products
involves the 1sH
‡
. The
constant ratio (Scheme 6), while the {[3s] ‡[1s] ‡[2-D-rac-
‡
5]}/[6] and [1-D-rac-5]/[6] ratios provide a measure of the k /
Ph
kH and kMe/kH rate-constant ratios, respectively. The corre-
sponding values are listed in Table 8 and their dependence
upon the reaction temperature (25 ± 1208C) is illustrated in
1
‡
‡
intermediate in the sequence ii) !iii) [A ˆ
back
front
�1
H] of Scheme 4, followed by its neutralization to 1s in
Figure 2. The linearity of the log(kPh /k ) versus T
Ph
competition with the hydrogen-atom-assisted H O loss shown
dependence of Figure 1 is consistent with independent back-
2
in Scheme 5 (k ). From this, we can conclude that formation
side versus frontside phenyl-group participation in C�O bond
H
Chem. Eur. J. 1999, 5, No. 3
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0503-0841 $ 17.50+.50/0
841

M. Speranza, A. Filippi
FULL PAPER
interval of 1sH
‡
(or 2sH
‡
) with CH 18OH (t ˆ2 Â10 s).[21] It
�8
Table 8. Rate constant ratios of intramolecular isomerization of 2sH‡ in
3
CH
4
at 750 Torr.[
a]
follows that the absolute rate constant of the fastest re-
Temperature [8C]
k
Ph/kMe
k
Ph/k
H
k
Me/k
H
arrangement in 1sH
‡
and 2sH , namely, that involving phenyl-
‡
7
�1
group participation, should greatly exceed 1/t ˆ5 Â10 s .
25
60
85
85
3.39 (0.530)
7.90 (0.898)
17.03 (1.231)
16.07 (1.206)
33.32 (1.523)
5.94 (0.774)
1.75 (0.244)
2.92 (0.465)
3.02 (0.481)
2.87 (0.458)
3.21 (0.506)
23.08 (1.363)
51.53 (1.721)
46.13 (1.664)
106.90 (2.029)
On the one hand, the differential activation parameters of
Table 9, in particular the DE*(Ir 1sH
‡
!IVs) ˆ7.0 Æ
�1
�1
0
.3 kcalmol and DE*(Ir 2sH
‡
!IVr) ˆ7.1Æ0.3 kcal mol ,
120
demand that this lower rate limit should correspond to an
[
a] Each value is the average of several determinations, with an uncertainty
level of ca.10%. kPh/kMe ˆ{[3s] ‡[1s] ‡[2-D-rac-5]}/[1-D-rac-5]; kPh/k
[3s] ‡[1s] ‡[2-D-rac-5]}/[6]; kMe/k ˆ[1-D-rac-5]/[6]. Logarithm of the
ratios in parentheses.
activation energy for the phenyl-group participation
H
ˆ
�1
1
sH
‡
!Ir 2sH
‡
of at least 7 ± 8 kcalmol , which in turn
{
H
corresponds to the negligible activation energy for the
competing hydrogen participation 1sH !IVr.
!IVs and 2s
On the other hand, the same lower rate limit excludes
‡
‡
H
�
1
activation barriers in excess of 8 ± 10 kcalmol
sH process, since otherwise the corresponding
!Ir 2sH
preexponential factors would exceed the maximum theoretical
for the
1
‡
‡
13
14 �1
value of 10 ± 10 s . On these grounds, we calculate that the
absolute activation energy for the phenyl-group participation
�
1
in 1sH
‡
and 2sH
‡
(E* ) amounts to 9 Æ2 kcalmol . This means
Ph
that the absolute activation energy for the hydride participation
�
1
in 1sH
‡
and 2sH
‡
(E*) is around 2 Æ2 kcalmol , and that for
H
the methyl-group participation in 2S
‡
(E* ) is around 4 Æ
H
Me
�
1
2
kcalmol . Comparison of these activation barriers with the
Figure 2. Arrhenius plots for the competing phenyl versus methyl (tri-
angles), methyl versus hydride (circles), and phenyl versus hydride
�
much larger energy required for the simple C O bond cleavage
in 1sH
‡
and 2sH
‡
to give the corresponding open-chain
participation (diamonds) to the C�O bond cleavage in 2sH‡
.
secondary and primary carbocations demonstrates that release
of the H O molecule from these systems is anchimerically
assisted by all the adjacent groups.
The preexponential factor for the hydride participation in
2
fission in 1sH
‡
and excludes any significant Ir$Is intercon-
version even at the highest temperatures. Given the relatively
long lifetime of 1,2-propenebenzenium intermediate (ca. 2 Â
1
sH
‡
(A ) is five orders of magnitude lower than that for the
H
�
8
[21]
10
s),
corresponding to many thousands C�C bond
competing phenyl-group participation (A ). For 2s
six orders of magnitude lower than A , whereas the
preexponential factor for the methyl-group participation
‡
, AH is
H
Ph
rotations, its reluctance to rearrange or racemize unimolec-
ularly in the gas phase corresponds to a static cyclic structure
I, rather than to the rapidly equilibrating open ones II
Ph
(
AMe) is over four orders of magnitude lower than A . These
Ph
(Scheme 1).
large differences indicate that, under the experimental con-
ditions used, neighboring-group assistance for unimolecular
Regression analysis of the linear curves of Figures 1 and 2
leads to the differential activation parameters listed in
Table 9. The relevant absolute values can be rationalized as
�
C O bond fission in 1s
entropic, rather than enthalpic factors. Activation entropies
‡
and 2sH is essentially governed by
‡
H
for C�O bond cleavage in 1s
‡
and 2sH cannot be associated
exclusively with conformational equilibria or with statistical
‡
H
Table 9. Differential Arrhenius parameters for the competing neighbor-
ing-group participation to the C�O bond cleavage in 1sH‡ and 2sH‡
.
factors (two migrating b-hydrogens vs. a single phenyl group
Competing
reactions[
Arrhenius equation[b]
Correlation
coefficient, r
in 1sH
‡
). Indeed, spectroscopic studies indicate that alcohols
a]
1
s and 2s in apolar solvents at room temperature preferen-
back
Ir 1sH‡!IVs log(kPh /k
H
ˆ(5.0 Æ0.3) �(7.0 Æ0.3)x
ˆ(3.2 Æ0.3) �(5.6 Æ0.3)x
front
0.960
0.964
tially acquire a gauche ± anti conformation stabilized by
intramolecular OH ´´ p-ring H-bonding. This preference is
front
[22]
Is 1sH‡!IVs log(kPh /k
H
back
Ir 1sH‡!Is
log(kPh /kPh ˆ(1.7 Æ0.5) �(1.3 Æ0.5)x 0.927
further exalted in the isolated state and for the O-protonated
Ir 2sH‡!IVs log(kPh/kMe) ˆ(4.7 Æ0.2) �(5.7 Æ0.2)x
Ir 2sH‡!IVr log(kPh/k ) ˆ(6.0 Æ0.2) �(7.1 Æ0.2)x
IVs 2sH‡!IVr log(kMe/k ) ˆ(1.3 Æ0.5) �(1.4 Æ0.5)x
0.997
0.997
0.906
[22, 23]
derivatives 1sH
‡
and 2sH
‡
.
Scheme 7 shows the stability
H
order of the rotamers of 1sH and 2sH
‡ ‡
.
H
As a result, backside Ph participation in 1sH
would be the entropically least favored process, in contrast
‡
and 2sH
‡
[
a] H O as byproduct; [b] x ˆ1000/2.303RT.
2
to its largest APh experimental value. This excludes conforma-
5
6
follows. In view of the inertness of the encounter complex V
Scheme 3) towards rearrangement and dissociation, the
closeness of the G(M) values in Tables 1 and 2 with the
tional factors as being the cause of the large A /A ˆ10 ± 10
Ph
H
4
(
and A /A ˆ5 Â10 ratios of Table 9. The hypothesis of a
Ph
Me
low-energy low-entropy hydrogen migration in the collision
1
8
maximum theoretical yields can be explained only by
complexes between 1sH
‡
or 2sH and an external CH3 OH
‡
extensive structural rearrangement of 1sH
before their trapping by the external CH3 OH. Therefore,
‡
and 2sH
‡
well
molecule is also unlikely, with the latter acting as a transducer
of the migrating hydrogen moiety. Indeed, in this case, a
pronounced loss of the deuterium label would be observable
1
8
timing of the neighboring-group participation in the C�O
bond fission in 1sH
‡
and 2s
‡
must fall well within the collision
in the rac-5 products from 1dH
‡
and 2dH , in contrast with
‡
H
8
42
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0503-0842 $ 17.50+.50/0
Chem. Eur. J. 1999, 5, No. 3

Wagner ± Meerwein Rearrangements
834±844
promoting CH OH loss from free O-methylated (S)-1-phenyl-
3
2
-propanol 1sMe
b) Analysis of the activation parameters points to the
neighboring-group assistance in 1sH and 2sH being regulated
‡
and (S)-2-phenyl-1-propanols 2sMe
‡
.
‡
‡
more by entropic than by enthalpic factors. Thus, at variance
with previous indications,[
11c]
the more energy-demanding
phenyl-group participation prevails over hydrogen participa-
tion under the conditions used only by virtue of its much less
unfavorable activation entropy.
c) In 1sH
‡
, competing phenyl-group (k ) versus hydrogen
Ph
(
kH) anchimeric assistance is observed. The same groups
compete with methyl (kMe) in anchimerically assisting C�O
bond fission in 2sH
allowed us to discern between frontside (k ) and backside
phenyl-group participation (kPh ).
‡
. The use of enatiomerically pure 1s
‡
H
front
Ph
Scheme 7. Order of stability of the conformers of O-protonated (S)-1-
phenyl-2-propanol and (S)-2-phenyl-1-propanol.
back
d) The observation of frontside phenyl-group participation in
front
1
sH
‡
(k ), with an activation energy lower than that of the
Ph
their high D retention (Table 4). Therefore, the large differ-
ences in the activation parameters of Table 9 can only be
ascribed to the different position of the corresponding
transition structures along the reaction coordinate. A com-
prehensive discussion of these aspects is reported in the
following paper.
back
competing backside phenyl-group participation (kPh ), is
rationalized in terms of the favored gauche conformations
in the chiral oxonium ion, and of the stabilizing interactions
between the leaving H O and the ring of the phenonium ion.
2
e) In agreement with spectroscopic evidence in superacidic
[7]
[25]
solutions and in contrast with theoretical predictions,
phenyl-group participation in C�O bond cleavage in 1s
Inspection of Table 9 reveals that, in the gas phase, the
‡
H
back
activation energy for the backside 1s
‡
!Ir (E* ) is 1.3 Æ
H
Ph
and 2s gives rise to static, optically active 1,2-propeneben-
H
‡
�
1
0
.5 kcalmol greater than that of the frontside 1s
reaction (E*Ph ). Such a counterintuitive observation for the
solution phase standards can be explained by considering that
‡
!Is
H
zenium intermediates that do not display any propensity for
unimolecular ring opening and racemization through high-
energy open-chain structures.
front
the process takes place in the absence of solvation and ion
front
pairing. Indeed, while the experimental E*
concerns the
Ph
back
(
(
1sH
‡
)gauche !Is step, the E*
refers to the overall
Ph
1sH
‡
)gauche !(1s )anti !Ir sequence, in which the first step
‡
H
Acknowledgements
[24]
may cost several kilocaleries per mole. Besides, the energy
requirements for the gas-phase frontside 1sH
‡
!Is process
This research was supported by the Italian Ministero dellꢀUniversita e della
Ricerca Scientifica e Tecnologica (MURST) and by the Italian National
Research Council (CNR). We thank Anna Troiani for help with some of the
early experimental work.
Á
may be mitigated by attractive electrostatic interactions in the
transition structure between the leaving moiety H O and the
2
phenyl ring of the cation. Similar electrostatic interactions are
spatially prevented in the gas-phase backside 1s
‡
!Ir
H
back
front
Ph
process. This also explains the APh /A
ˆꢂ50 ratio in
[
1] a) D. J. Cram, J. Am. Chem. Soc. 1949, 71, 3863; b) D. J. Cram, R.
Davis, J. Am. Chem. Soc. 1949, 71, 3875; c) D. J. Cram, J. Am. Chem.
Soc. 1964, 86, 3767; d) S. Winstein, B. K. Morse, E. Grunwald, K. C.
Schreiber, J. Corse, J. Am. Chem. Soc. 1952, 74, 1113.
Table 9, which reflects the more restricted H O rotations and
2
translations in the gas-phase frontside 1sH
‡
!Is process
relative to the backside 1sH
‡
!Ir one. Of course, as pointed
[
1]
out in related gas-phase studies, electrostatic interactions
play only a minor role in solvolytic nucleophilic displacements
owing to interference from the reaction medium (nucleophi-
licity, dielectric properties, etc.). In these media, conforma-
tional equilibria in ionic species can be strongly altered and
electrostatic interactions between the nucleophile and the
leaving moiety minimized, so that solvolysis is usually
governed by stereoelectronic factors favoring backside par-
ticipation.
[2] C. J. Lancelot, D. J. Cram, P. von R. Schleyer, Carbonium Ions, Vol. III
Eds.: G. A. Olah, P. von R. Schleyer) Wiley, New York, 1972,
(
chapter 27.
[
3] a) H. C. Brown, The Transition State, Special Publ. No. 16, The
Chemical Society, London, 1962, p. 149; b) H. C. Brown, K. J. Morgan,
F. J. Cloupek, J. Am. Chem. Soc. 1965, 87, 2137.
[4] a) H. C. Brown, R. Bernheimer, C. J. Kim, S. E. Scheppele, J. Am.
Chem. Soc. 1967, 89, 370; b) H. C. Brown, C. J. Kim, J. Am. Chem. Soc.
1
968, 90, 2080; c) W. H. Saunders, Jr., S. Asperger, D. H. Edison, J. Am.
Chem. Soc. 1958, 80, 2421; d) S. L. Loukas, M. R. Velkou, G. A.
Gregoriou, J. Chem. Soc. Chem. Commun. 1969, 1199; e) S. L. Loukas,
F. S. Varveri, M. R. Velkou, G. A. Gregoriou, Tetrahedron Lett. 1971,
1
803.
Conclusions
[
5] B. Capon, S. P. McManus, Neighboring Group Participation, Vol. 1,
Plenum, New York, 1976.
a) The present gas phase results provide an experimental
insight into vicinal-group anchimeric assistance in the uni-
[6] a) G. A. Olah, R. D. Porter, J. Am. Chem. Soc. 1970, 92, 7627; b) G. A.
Olah, R. D. Porter, J. Am. Chem. Soc. 1971, 93, 6877; c) G. A. Olah,
R. J. Spear, D. A. Forsyth, J. Am. Chem. Soc. 1976, 98, 6284.
molecular loss of H O from free O-protonated (S)-1-phenyl-
2
[
7] G. A. Olah, N. J. Head, G. Rasul, G. K. Surya Prakash, J. Am. Chem.
Soc. 1995, 117, 875 (see also: S. Sieber, P. von R. Schleyer, J. Gauss, J.
Am. Chem. Soc. 1993, 115, 6987).
2
-propanol 1sH
‡
and (S)-2-phenyl-1-propanols 2sH
‡
. Neigh-
boring-group participation appears much less effective in
Chem. Eur. J. 1999, 5, No. 3
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0503-0843 $ 17.50+.50/0
843

M. Speranza, A. Filippi
FULL PAPER
[
8] G. A. Olah, R. J. Spear, D. A. Forsyth, J. Am. Chem. Soc. 1977, 99,
615.
[18] The binding energy in the cluster between tert-butyl cation and water
�
1
2
is estimated as large as about 11 kcalmol by high-level ab initio
calculations (D. Berthomieu, H. E. Audier, Eur. Mass Spectrom. 1997,
3, 19). The interaction energy in electrostatically bonded clusters
between carbocations and n-type donors is rather insensitive of the
[
9] a) W. J. Hehre, J. Am. Chem. Soc. 1972, 94, 5919; b) S. Yamabe, T.
Tanaka, Nippon Kagaku Kaishi 1986, 11, 1388. See also ref. [6c].
[
[
10] a) T. H. Morton, Tetrahedron 1982, 38, 3195; b) D. J. McAdoo, T. H.
Morton, Acc. Chem. Res. 1993, 26, 295.
2 3
nature of both the ion and the donor, whether H O or CH OH (see,
11] a) S. Fornarini, C. Sparapani, M. Speranza, J. Am. Chem. Soc. 1988,
for instance: a) M. Meot-Ner (Mautner), Z. Karpas, C. Deakyne, J.
Am. Chem. Soc. 1986, 108, 3913; b) H. E. Audier, G. K. Koyanagi,
T. B. McMahon, D. Thölmann, J. Phys. Chem. 1996, 100, 8220; c) E.
Uggerud, J. Am. Chem. Soc. 1994, 116, 6873; d) M. Meot-Ner (Maut-
ner), M. M. Ross, J. E. Campana, J. Am. Chem. Soc. 1985, 107, 4839).
1
1
8
10, 34; b) S. Fornarini, C. Sparapani, M. Speranza, J. Am. Chem. Soc.
988, 110, 42; c) S. Fornarini, V. Muraglia, J. Am. Chem. Soc. 1989, 111,
73.
[
12] a) C. Koppel, C. C. van der Sande, N. M. M. Nibbering, T. Nishishita,
F. W. McLafferty, J. Am. Chem. Soc. 1977, 99, 2883; b) C. N. McEwen,
M. A. Rudat, J. Am. Chem. Soc. 1981, 103, 4355; c) M. Carsten, D.
Kuck, Org. Mass Spectrom. 1993, 28, 1073; d) W. Bäther, H. F.
Grützmacher, Int. J. Mass Spectrom. Ion Proc. 1985, 64, 193: e) D.
Kuck, Mass Spectrom. Rev. 1990, 9, 583.
3
[19] In the absence of NMe , neutralization of the O-protonated ethers,
arising from addition of Ir, IVr, and IVs to methanol, may proceed
through several pathways, including proton transfer to the substrate
itself, to the walls of the reaction vessel, or to adventitious bases, either
initially added to the gaseous mixture or formed from its radiolysis.
[20] This extension is valid only if the secondary kinetic b-deuterium
effects on the phenyl and methyl-group participation in 2sH‡are taken
equal to unity. The validity of this assumption is demonstrated in the
following paper.
[
13] a) F. Cacace, Acc. Chem. Res. 1988, 21, 215; b) M. Speranza, Mass
Spectrom. Rev. 1992, 11, 73.
[
[
14] H. C. Brown, G. Zweifel, J. Am. Chem. Soc. 1960, 82, 4708.
15] D. Achet, D. Rocrelle, I. Murengezi, M. Delmas, A. Gaset, Synthesis
1
986, 643.
[21] The reaction time is taken as the inverse of the pseudo first-order
‡
18
[
16] G(C
n
H
5
(n ˆ1,2)) ˆ2.8 : a) P. Ausloos, S. G. Lias, R. Gorden, Jr., J.
collision rate constant k ˆkcoll[CH
the CH
calculated according to T. Su, W. J. Chesnavitch, J. Chem. Phys.
3
OH] between 1sH‡ (or 2sH‡) and
3
OH nucleophile, with the collision rate constant kcoll
18
Chem. Phys. 1963, 39, 3341; b) P. Ausloos, Ion-Molecule Reactions
Ed.: J. L. Franklin), Plenum, New York, 1970; c) P. Ausloos, S. G.
(
9
3
�1 �1
Lias, J. Chem. Phys. 1962, 36, 3163; d) S. G. Lias, P. Ausloos, J. Chem.
Phys. 1962, 37, 877.
1982, 76, 5183; [kcoll (Â10 cm molecule
s
) ˆ1.81 (258C); 1.74
(608C); 1.70 (858C); 1.67 (1008C); 1.64 (1208C); 1.62 (1408C)].
[22] a) M. J. Cook, T. A. Khan, K. Nasri, Tetrahedron Lett. 1984, 5129;
b) M. J. Cook, T. A. Khan, K. Nasri, Tetrahedron 1986, 42, 249.
[23] J. J. Urban, C. W. Cronin, R. R. Roberts, G. R. Famini, J. Am. Chem.
Soc. 1997, 119, 12292.
[
17] The irradiated systems invariably contain
H
2
16O, as ubiquitous
impurity, either initially introduced in the mixture together with its
bulk component or formed from its radiolysis. As pointed out
previously (A. Troiani, F. Gasparrini, F. Grandinetti, M. Speranza, J.
Am. Chem. Soc. 1997, 119, 4525; M. Speranza, A. Troiani, J. Org.
Chem. 1998, 63, 1020.), the average stationary concentration of H O
2
in the radiolytic systems is estimated to approach that of the added
[24] S. Usui, M. Okamura, A. Ohno, Proceedings of the VII Kyushu
International Symposium on Physical Organic Chemistry, Kyushu
University, December 2 ± 5, 1997.
16
18
18
H
2
O (ca. 2 ± 3 Torr). Therefore, the ꢀ40% O incorporation in 1s is
[25] H. Griengl, P. Schuster, Tetrahedron 1974, 30, 117.
interpreted in terms of the intermolecular 2s !1s isomerization
mechanism, shown in path ii) !iii) of Scheme 4.
Received: June 16, 1998 [F1211]
8
44
ꢁ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0503-0844 $ 17.50+.50/0
Chem. Eur. J. 1999, 5, No. 3