The gas-phase reactivity of p-Me3Si-substituted 1,3-diphenylpropane towards charged electrophiles: Intra- and interannular hydrogen migrations

Article abstract of DOI:10.1002/(SICI)1521-3765(19980615)4:6<993::AID-CHEM993>3.0.CO;2-I

The gas-phase reaction of p-Me₃SiC₆H₄(CH₂)₃C₆H₅ (p-TSDPP) with gaseous cations, including C₂H₅⁺, Me₂Cl⁺ and DCO⁺, has been studied in the pressure range from 10^-8 to 10³ Torr by Fourier-transform ion cyclotron resonance (FT-ICR) and by the radiolytic technique. The protonated or alkylated intermediates undergo intramolecular migration and intermolecular transfer of protons and/or Me₃Si⁺. The results underline the role of the spectator ring in providing internal solvation to an arenium moiety, as evidenced by the noticeable stability towards Me₃Si loss with respect to a single-ring model substrate, p-Me₃SiC₆H₄Me (p-TST), upon reaction with the same gaseous ions. The extent of the alkylation route relative to the alkyldesilylation processes, measured as a function of the arenium ion lifetime, permits derivation of the rate constant for the conversion by proton transfer of the originally formed arenium ions to ipso-silylated isomers (k(i)). The estimated values of k(i(p-TST)) = 5 x 10⁹ s^-1 and k(i(p-TSDPP)) = 2 x 10⁸ s^-1 at 120°C suggest that interannular H shifts are faster than ring-to-ring H transfer, in agreement with previous evidence from tert-butylated arenium ions. The reactivity of [Me₃Si=arene]⁺ adducts, adequately described by the Wheland σ-complex model, does not exclude the intermediacy of an ion-neutral noncovalent complex.

Full text of DOI:10.1002/(SICI)1521-3765(19980615)4:6<993::AID-CHEM993>3.0.CO;2-I

FULL PAPER
The Gas-Phase Reactivity of p-Me3Si-Substituted 1,3-Diphenylpropane
Towards Charged Electrophiles: Intra- and Interannular Hydrogen Migrations
Maria Elisa Crestoni*
Abstract: The gas-phase reaction of p-
Me3SiC6H4(CH2)3C6H5 (p-TSDPP) with
noticeable stability towards Me3Si loss
with respect to a single-ring model
substrate, p-Me3SiC6H4Me (p-TST),
upon reaction with the same gaseous
ions. The extent of the alkylation route
relative to the alkyldesilylation process-
es, measured as a function of the are-
nium ion lifetime, permits derivation of
the rate constant for the conversion by
proton transfer of the originally formed
arenium ions to ipso-silylated isomers
‡
‡
gaseous cations, including C2H5 , Me2Cl
(ki). The estimated values of ki(p-TST)
ˆ
‡
1
1
and DCO , has been studied in the
5  109 s and ki(p-TSDPP) ˆ 2  108 s at
1208C suggest that intraannular H shifts
are faster than ring-to-ring H transfer, in
agreement with previous evidence from
tert-butylated arenium ions. The reac-
8
pressure range from 10 to 103 Torr by
Fourier-transform ion cyclotron reso-
nance (FT-ICR) and by the radiolytic
technique. The protonated or alkylated
intermediates undergo intramolecular
migration and intermolecular transfer
‡
tivity of [Me3Si ± arene] adducts, ade-
quately described by the Wheland s-
complex model, does not exclude the
intermediacy of an ion ± neutral nonco-
valent complex.
‡
of protons and/or Me3Si . The results
Keywords: arenes ´ FT-ICR ´ gas-
phase chemistry ´ hydrogen transfer
´ ion ± molecule reactions
underline the role of the spectator ring
in providing internal solvation to an
arenium moiety, as evidenced by the
metastable decomposition studies. Fulfilling both conditions,
the high-pressure radiolytic technique[1] has allowed a de-
tailed kinetic study of the interannular proton transfer in
protonated diarylalkanes,[6] disclosing the effect of the spec-
tator ring on the stability of the charged intermediates.
The peculiar features of the trimethylsilyl substituent,
namely its effect of increasing the basicity of the ipso carbon[7]
and the ease of protolytic cleavage,[7b,8] have made the Me3Si
group a useful probe for elucidation of the proton motion
within selected arenium ions.[9,3a] This probe is used in the
present study to provide further evidence for the occurrence
of both intraannular and interannular proton shifts, with the
aim of assessing the relative rates of the two processes. At the
same time the reactivity of simple silylated arenium ions is
compared to that of the same species bound to a remote,
neutral aryl group.
Introduction
Arenium ions have long been recognized as key intermediates
in gas-phase aromatic substitution by charged electrophiles.[1]
A salient mechanistic feature is the tendency to isomerise by
1,2-shifts of hydrogen or other groups. A related interannular
hydrogen migration takes place following electrophilic sub-
stitution of a,w-diphenylalkanes,[2] along pathways whose gas-
phase kinetics have been the topic of several experimental[2,3]
and theoretical investigations.[4] Mass-spectrometric investi-
gations have revealed extensive hydrogen scrambling in
protonated a,w-diphenylalkanes that may lead to complete
equilibration in the 10 msec time frame, typical of metastable
loss of benzene.[5] However, in order to confer a general
validity to gas-phase kinetic studies and allow their compar-
ison with condensed-phase results, the experimental techni-
que adopted should ensure thermal equilibration of the
reactant species and allow a meaningful definition of the
reaction temperature. At the same time, it is desirable to work
with a much higher time resolution than is possible in
Results
‡
FT-ICR reactions: The reaction of Me3Si ions with toluene
(C7H8) introduced by a pulsed valve into the ICR cell yields
[Me3Si ± C7H8] ions (1), whose stability was checked follow-
ing their decay in the presence of C7D8 at 3.0 Â 10 8 Torr.
Although the admission of C7H8 by the pulsed valve ensures a
partial collisional stabilization of the excited adduct ions such
as 1, allowing their detection, the only reaction pathway
‡
[*] Dr. M. E. Crestoni
Dipartimento di Chimica e Tecnologia delle Sostanze Biologicamente
Attive
Á
Universita di Roma La Sapienza
P.le A. Moro 5, I-00185 Roma (Italy)
Fax : ( ‡ 39)6-4991-3602
Chem. Eur. J. 1998, 4, No. 6
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
0947-6539/98/0406-0993 $ 17.50+.25/0
993

Intra-/Interannular H Migrations
FULL PAPER
‡
observed was the dissociation of 1 into free Me3Si ions with a
1
‡
‡
232 kcalmol ]. Efficient H /D transfer to the base (100%
of the ADO collision rate)[13] was observed, only minor
1
unimolecular decay rate constant of 8 s . This result points to
the facile desilylation of 1 under the low-pressure conditions
prevailing in the FT-ICR experiments, which clearly do not
allow silylation of C7D8, neither by Me3Si addition nor by
Me3Si transfer from 1. By contrast, efficient Me3Si transfer
from 1 takes place when 1,3-diphenylpropane (DPP) at 4.2 Â
10 8 Torr is added into the cell. Remarkably, the ionic product
‡
amounts of Et3NSiMe3 (m/z ˆ 174) being formed. This
finding, in particular the observation of abundant ions at m/
‡
‡
z ˆ 103 corresponding to Et3ND , provides strong evidence
‡
‡
that the silyl group is covalently bound[14] within the [Me3Si ±
‡
[D5]DPP] intermediate.
‡
2, corresponding to [Me3Si ± DPP] (m/z ˆ 269), becomes the
Radiolytic experiments: The radiolytic reactions were carried
out at 1208C at a pressure of ca. 700 Torr of bulk gas (CH4 or
MeCl) in the presence of 10 Torr of O2 as a radical scavenger.
The absolute radiolytic yields of the products account for a
major portion of the primary reactant ions formed by the
ionization of the bulk gas and ensuing ion ± molecule reac-
tions. Furthermore, the depression of the yields upon addition
of a base/nucleophile at increasing concentration, which
competes with the substrate for the electrophile, confirms
the ionic origin of the process under study. Conversion of the
starting substrate was kept below 1% in order to minimize the
extent of further secondary processes that would complicate
the primary product pattern.[1]
predominant species at long storage time, showing a notice-
able reluctance both to dissociate unimolecularly and to
‡
‡
transfer Me3Si to yield [Me3Si ± C7D8] . Protonation of para-
trimethylsilyldiphenylpropane (p-TSDPP) provides an alter-
‡
native route to 2. iC3H7CNH ions [proton affinity
PA(iC3H7CN) ˆ 194.3 kcalmol ][10] were chosen as a mild
1
protonating agent for p-TSDPP, yielding 2, which, when
formed by this alternative route, does not undergo fragmen-
tation and reacts only with adventitious water present in the
‡
‡
cell, giving Me3SiOH2 . This reaction of [Me3Si ± DPP] has
been verified by deliberately admitting H2O into the ICR cell
at 2 Â 10 8 Torr. As shown in Figure 1, Me3SiOH2 and
‡
Table 1 summarizes the distribution of the major products
(species 3 ± 6, where R ˆ Et, (SiMe)3) from the radiolytic
‡
‡
&
Figure 1. Normalized ion abundances of [Me3Si ± DPP] ( ), Me3SiOH2
‡
!
*
(
) and DPPH
(
) as a function of time. The pressure of H2O was 2 Â
10 8 Torr.
methylation of p-TST and p-TSDPP at nearly constant
concentration, in the presence of MeCl as bulk gas and of
varying amounts of basic additives. As expected from the use
‡
DPPH ions are formed, which may arise by partition of a
common ion ± neutral intermediate complex [DPP ´
‡
of a purely methylating agent, Me2Cl from the radiolysis of
‡
Me3SiOH2 ; Eq. (1)]. Process (1c) involves an intracomplex
MeCl,[15] as reactant ion, only methylated and methyldesily-
lated products are detected for both substrates. The isomeric
distribution of the desilylated products from p-TSDPP and p-
TST shows that only a fraction (50% and 30%, respectively)
of electrophilic attack is directed to the (silylated) para
position. The significant yields of o- and m-methylated
isomers formed together with the p-derivative indicate that,
besides the direct nuclear substitution at the silylated position,
alkyldesilylation can be traced to proton migration from the
alkylated ring positions to the silylated site, followed by
release of Me3Si. The abundance of such a process, even if not
completely suppressed, strongly declines in the presence of
‡
2 ‡ H2O ! [2 ´ H2O] ! [DPP ´ Me3SiOH2 ]
(1a)
(1b)
‡
‡
[DPP ´ Me3SiOH2 ] ! DPP ‡ Me3SiOH2
‡
‡
[DPP ´ Me3SiOH2 ] ! [DPPH ´ Me3SiOH]
! DPPH ‡ Me3SiOH
(1c)
‡
proton transfer, energetically favoured by the higher PA of
DPP[11] with respect to Me3SiOH [PA(Me3SiOH) ˆ
1
194 kcalmol ], whereas process (1a) is thermodynamically
driven by the higher energy of the Si ± O bond than of a typical
Si ± C bond. Accordingly, process (1a) reflects the larger
Lewis basicity of oxygenated compounds towards Me3Si
than towards aromatics.[12]
Structural insight was sought by forming silylated adducts
from the Me3Si reaction with C6H5(CH2)3C6D5 ([D5]DPP).
The ensuing [Me3Si ± [D5]DPP] ions were isolated and
allowed to react with triethylamine [PA(Et3N) ˆ
‡
1
pyridine [PA(cC5H5N) ˆ 226 kcalmol ] and at higher con-
1
centrations of (MeO)3PO [PA(MeO)3PO ˆ 212 kcalmol ],
as inferred from the noticeable increase of the yield of
alkylated products with respect to the alkyldesilylated ones.
This trend is paralleled by an increasing fraction of the
products bearing a methyl group ortho to the Me3Si group of
‡
‡
994
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
0947-6539/98/0406-0994 $ 17.50+.50/0
Chem. Eur. J. 1998, 4, No. 6

M. E. Crestoni
993±999
‡
Table 1. Gas-phase reaction of p-Me3Si ± C6H4Me (p-TST) and p-Me3Si ± C6H4(CH2)3C6H5 (p-TSDPP) with radiolytically formed Me2Cl ions.
System composition (Torr)[a]
p-TSDPP Additives
Relative yields [%]
p-TST
3
4
5
6
(1,2:1,3)
(o:m:p)
(X:Y)[b]
(o:m:p)
0.36
0.41
0.42
0.30
0.53
±
±
0.38
±
0.15
0.1
0.16
0.12
±
0.24
0.17
0.19
0.24
0.30
(MeO)3PO
(MeO)3PO
(MeO)3PO
(MeO)3PO
pyridine
pyridine
pyridine
NEt3
N(iC3H7)2Et
N(CH3)2(CH2)2OH
(0.49)
(1.26)
(2.44)
(4.10)
(0.51)
(0.56)
(0.61)
(0.47)
(0.41)
(0.50)
10 (85:15)
15 (77:23)
23 (45:55)
33 (45:55)
15 (48:52)
±
90 (45:25:30)
85 (52:23:25)
77 (48:23:25)
67 (40:23:25)
85 (41:32:27)
±
17 (36:64)
68 (43:57)
83 (48:52)
90 (60:40)
±
71 (35:65)
69 (38:62)
47 (43:57)
30 (43:57)
33 (37:63)
83 (21:27:52)
32 (25:31:44)
17 (25:28:47)
10 (27:30:43)
±
29 (21:26:53)
31 (20:28:52)
53 (20:23:57)
70 (20:28:52)
67 (19:23:58)
±
±
10 (60:40)
±
11 (44:56)
90 (46:27:28)
±
89 (44:26:30)
0.52
[a] All gaseous systems contained MeCl (700 Torr) and O2 (10 Torr). Reactions carried out at 1208C. [b] X-type isomers have the methyl group on the
substituted phenyl ring, while Y-type isomers result from alkylation at the unsubstituted ring.
‡
p-TST. In particular, it may be noticed that the ratio of the
yields of the isomeric products 1,2-dimethyl-4-trimethylsilyl-
benzene and 1,3-dimethyl-4-trimethylsilylbenzene changes by
a factor of four with an almost tenfold increase of the
(MeO)3PO concentration. In the case of p-TSDPP, two types
of methylation products are identified, deriving from the
attack of the cationic reagent either at the silylated ring
(henceforth denoted as products of type X) or at the
unsubstituted one (henceforth called products of type Y).
Increasing amounts of (MeO)3PO favour the formation of X-
type products over the Y type. A remarkable feature of the
product pattern of Table 1 concerns the stronger positive
dependence of alkylation/alkyldesilylation yield ratio on
increasing (MeO)3PO concentration in the p-TSDPP reaction
compared to p-TST.
the attempt to silylate DPP by Me3Si ions formed in a
gaseous mixture of CH4/SiMe4/O2 (70:2:1)[12,16] in the presence
of 1.5 Torr of (MeO)3PO failed, in spite of the estimated
comparable basicity of (MeO)3PO and p-TSDPP that could
have ensured the neutralization of the intermediate [Me3Si ±
‡
DPP] ion by proton transfer. In contrast, when Et3N replaces
(MeO)3PO in the Me3Si reaction with DPP, high yields of
‡
silylated products are obtained.
The composition of the gaseous systems containing CH4 as
the bulk component and the relative yields of the products
(products 7 ± 10, R ˆ Et, (SiMe)3) are reported in Table 2.
‡
Attack of CnH5 (n ˆ 1,2) ions from the radiation-induced
reaction of methane on mixtures of trimethylsilylbenzene
(TSB) and toluene or DPP yields three types of products: a)
alkylation products retaining the Me3Si group; b) proto- and
ethyldesilylation products, where the Me3Si group has been
replaced by H or Et; c) silylation products. The last products
are found only in the presence of a strong nitrogen base, such
as pyridine or Et3N, since no silylation products (i.e. trime-
thylsilyltoluenes or 1-(trimethylsilylphenyl)-3-phenylpro-
panes) could be detected with (MeO)3PO. As in previous
mass spectrometric and radiolytic results, alkyldesilylation
represents the major reaction channel.[9b,12]
Among the bases used at comparable concentrations,
pyridine is most effective in promoting alkylation products,
whereas alkyldesilylation is the dominant channel when
(MeO)3PO is added. An intermediate behavior is shown by
Et3N and diisopropylethylamine, a sterically hindered base
1
[PA(iPr2NEt) ˆ 235 kcalmol ] and by 2-(dimethylamino)-
ethanol, a strong nitrogen base containing an OH group
1
[PA(Me2N(CH2)2OH) ˆ 236 kcalmol ]. It is noteworthy that
‡
Table 2. Gas-phase reaction of p-Me3SiC6H5 (TSB), o-Me3SiC6H4Me (p-TST), PhMe and 1,3-diphenylpropane (DPP) with CnH5 (n ˆ 1,2) ions.
System composition (Torr)[a]
Relative yields of products [%][b]
Apparent
kS1/kS2
[c]
Substrates
S2
Additives
7
8
9
10
S1
(o:m:p)
(o:m:p)
(o:m:p)
TSB (0.48)
TSB (0.43)
TSB (0.54)
o-TST (1.5)[d]
PhMe (0.64)
DPP (0.20)
DPP (0.29)
±
pyridine (0.36)
pyridine (0.39)
Et3N (0.20)
R ˆ Et
2
53
37
±
±
±
(0:40:60)
5
(0:51:49)
4
(0:40:60)
3
(0:45:55)
3
(41:37:22)
3
(0:24:76)
±
R ˆ SiMe3
R ˆ Et
2.2
48
48
(24:36:39)
2
(0:26:74)
40
(24:33:43)
8
(0:17:83)
±
R ˆ SiMe3
R ˆ Et
±
0.70
±
39
R ˆ SiMe3
R ˆ SiMe3
10
±
(0:58:42)
±
0.74
±
Et3N (0.29)
(0:24:76)
[a] All gaseous systems contained CH4 (700 Torr) and O2 (10 Torr) and were submitted to g radiolysis at 1208C. [b] Standard deviation ca. 3%. [c] kS1/kS2
[P1][S2]/[P2][S1], where P1 and P2 are the products of the attack at the S1 and S2 substrates respectively. [d] The system contained CO/D2 (20:600) and O2
ˆ
(10 Torr) at 378C.
Chem. Eur. J. 1998, 4, No. 6
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
0947-6539/98/0406-0995 $ 17.50+.25/0
995

Intra-/Interannular H Migrations
FULL PAPER
The kSubstrate1/kSubstrate2 (kS1/kS2) values were obtained from
the relative yields of the silylated products of TSB and toluene
or DPP, corrected for the molar ratio of the respective
substrates. The kTSB/kC7H8 ratio is 2.2 and the kTSB/kDPP ratio is
0.7 in the presence of comparable amounts of base. A salient
feature of the isomeric distribution of the silylated products is
the lack of ortho-substitution and a noticeable 2para/meta
selectivity, which ranges
the most basic site even in the presence of other activating
ring substituents[18]. The overall reaction sequence is depicted
in Equation (2), where R ˆ H, p-CH3, (CH2)3Ph.
Previous mass spectrometric and radiolytic studies have
demonstrated that also in the gas phase positively charged or
positively polarised silicon also displays a noticeable oxophi-
licity,[19] which accounts for the pronounced tendency of
from approximately 2.0
for TSB to approximately
6.0 for both toluene and
DPP. Finally, it may be
noted that the alkylation
products appear to be fav-
oured over the alkyldesily-
lation products when Et3N
rather than pyridine is
used.
‡
The reaction promoted by the attack of DCO on o-
Me3Si ± C6H4CH3 (o-TST), a deuterated Brùnsted acid ob-
tained from the ionization of a CO/D2 mixture, was examined
for information on the protonation-induced isomerization of
this substrate. The meta and para isomers (24:76) are formed,
‡
suggesting the occurrence of intramolecular Me3Si migration
following protonation.
oxygenated compounds to desilylate ipso-silylated arenium
ions.
Discussion
Whereas all the bases employed in this study can deprot-
onate 11 and 12 exothermically, the branching between
routes (2c) and (2d) depends on the specific features of the
base. Nitrogen bases, especially when sterically crowded,
behave as efficient deprotonating agents, leading to silylated
products, whereas oxygen bases react as nucleophiles, selec-
tively removing the Me3Si group whenever it is bound to a
tetracoordinate carbon.[16]
‡
The g-radiolysis of CH4 produces known yields of CnH5 (n ˆ
1,2) ions, which are powerful acids [PA(CH4) ˆ
131.6 kcalmol 1; PA(C2H4) ˆ 162.6 kcalmol ]. Ionization
1
‡
of MeCl eventually yields Me2Cl ions, whose alkylation by
methyl cation transfer may involve a sizeable activation
barrier. Me3Si ions[17] are obtained from the reaction of CnH5
ions with SiMe4, according to extensive mass-spectrometric[12]
and radiolytic studies.[16] The pronounced positive charge
‡
‡
Equation (2) provides the framework to discuss the results
of the present study. Methylation (E ˆ Me) occurs when
‡
located on the silicon atom of Me3Si combined with the
‡
Me2Cl reacts with the aromatic substrate by an irreversible
pronounced basicity of its conjugate base [PA(CH3)2Si ˆ
1
MeCl displacement. Probably as a result of steric constraints,
CH2) ˆ 226 kcalmol ] confers upon this ion an exclusive
‡
Me2Cl ions attack the para-silylated carbon to only a minor
Lewis acid character toward the selected substrates. The
gaseous precursors used in this work generate the charged
electrophiles according to a well-established sequence of
ionization, fragmentation and ion ± molecule reactions. In the
radiolytic systems all ions undergo multiple unreactive
collisions with the bulk gas, approaching thermal equilibra-
tion before interacting with the aromatic substrate.[1]
extent, leading to a mixture of isomeric ions 11, where the
methyl group remains at the original site of attack, owing to
the poor migrating ability of a methyl group within an
arenium ring. Other factors being equal, the presence of
(MeO)3PO in the place of a nitrogen base enhances the
methyldesilylation route [Eq. (2c)] at the expense of the
methylation route [Eq. (2d)], in that an oxygenated base
promotes exclusive desilylation of ion 12. Accordingly, DPP
did not give silylation products in the presence of (MeO)3PO,
in spite of its high PA. Increasing (MeO)3PO concentrations
enhance the rate of process (2b), allowing a faster sampling of
11 before its conversion to 12. It follows that the isomeric
composition of the methylation products is not representative
of the positional selectivity of the electrophile, but rather
reflects the average lifetime of the intermediate ions 11.
This view is confirmed by the increasing fraction of ortho-
alkylated isomer in the presence of increasing amounts of
(MeO)3PO.
Alkylation vs alkyldesilylation: Several experimental[12,14,16]
and theoretical investigations[7] have addressed the problem
of the structure of Me3Si-substituted arenium ions, pointing to
the enhanced basicity of the ring carbon bearing the Me3Si
group of a silylated arene. This factor accounts for the easy
cleavage of the Me3Si group following protonation or
alkylation of silylated arenes, such as TSB, p-TST, and p-
TSDPP, directing the proton to the silylated position. The
desilylation process is induced by a) direct electrophilic attack
at the silylated carbon and b) electrophilic attack at any other
aromatic carbon followed by an H shift to the silylated carbon,
996
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
0947-6539/98/0406-0996 $ 17.50+.50/0
Chem. Eur. J. 1998, 4, No. 6

M. E. Crestoni
993±999
The dependence of the alkylation/al-
kyldesilylation ratio on the structural
features of highly basic nitrogen com-
pounds can be taken as an indication for
the role of steric factors,[20] decreasing the
rate of deprotonation of ions 11 and 12
and indirectly favouring desilylation of 12.
In fact, the ratio decreases along the series
pyridine, Et3N, iPr2NEt. Although it is
highly basic, Me2N(CH2)2OH can react
either by proton transfer to the nitrogen
atom or by nucleophilic attack of the OH
group at silicon. Accordingly, the alkyla-
tion/alkyldesilylation ratio is intermediate
between those obtained with (MeO)3PO
and with the nitrogen bases.
An additional process comes into play
when a second phenyl ring is covalently
linked to the silylated arene by a methyl-
ene chain. The spectator ring is known to
‡
engage in proton[6] or Me3C transfer[21]
with the arenium moiety, with efficiencies
that depend, inter alia, on the features of
the substituents[22] in the arenium ion and
the spectator ring and on the length of the
polymethylene chain[23,24] joining the two
rings. Accordingly, the reactivity of p-
TSDPP is expected to differ significantly
from that of p-TST, a simple aromatic
model. In the former case, the electro-
philic attack, occurring either at the
unsubstituted or at the silylated ring,
may be followed by both interannular
and intraannular proton migration, de-
picted in Scheme 1.
Scheme 1. Electrophilic attack on p-TSDPP occurs either at the unsubstituted or at the silylated ring,
and may be followed by both interannular and intraannular H migration.
The silyl group is expected to direct the
‡
electrophilic attack by Me2Cl preferen-
tially towards the silylated ring of p-TSDPP, leading to
arenium ions that are prone to efficient intraannular H shift to
the silylated position. Interannular H migration is instead
predicted to follow methylation at the unsubstituted ring of p-
TSDPP, finally ending in desilylation products through a
sequence of inter- and intraannular H shifts.
latter process involves a combination of inter- and intra-
annular proton transfer steps. A linear relationship is
predicted between the ratio (r) of the yields of alkylated
and alkyldesilylated products and the (MeO)3PO concentra-
tion, shown by the plots of Figure 2. The slopes, which
correspond to kB/ki, allow evaluation of the ki values, rely-
ing on the assumption that the highly exothermic deproto-
Me3Si loss from p-TSDPP is much less pronounced than
from p-TST. This conspicuous difference is verified both by
the alkylation/alkyldesilylation ratio of the radiolytic methyl-
‡
ation and by the tendency towards Me3Si loss from ions 1 and
2 in the ICR cell. A major role may be ascribed to the
favourable influence exerted by the spectator ring on the
stability of the silylated arenium ion, which underlines the
demand of charge delocalization and ion solvation in the gas
phase.
By application of the steady-state treatment to Equa-
tion (2), the relative decrease of the alkyldesilylation products
in Table 1 with increasing (MeO)3PO concentration can be
utilized to evaluate the ki rate constants that refer to the
11 !12 isomerization in the case of p-TST and the overall
isomerization rate of a mixed 13a/13b population to a mixed
14a/14b population (Scheme 1) in the case of p-TSDPP. The
Figure 2. Plots of ratios rˆ [Me ± p-TSDPP]/[Me ± DPP] (.) and rˆ [Me ±
&
p-TST]/[xylenes] ( ) vs. [(MeO)3PO] concentration from the radiolytic
‡
reaction of Me2Cl with p-TST/p-TSDPP mixtures.
Chem. Eur. J. 1998, 4, No. 6
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
0947-6539/98/0406-0997 $ 17.50+.25/0
997

Intra-/Interannular H Migrations
FULL PAPER
nation by the base B proceeds at the collision frequency
(kB ˆ kcollision), which can be calculated by the ADO theory.
At 393 K, in MeCl at ca. 1 atm, the rate constant for the
1
process 11 !12 (ki(p-TST) ˆ 5  109 s ) is considerably higher
than that for the 13a/13b !14a/14b process (ki(p-TSDPP) ˆ 2 Â
1
108 s ). From these data a valuable insight can be gained on
the overall proton shift sequences. In fact, the lower rate for
the 13a/13b !14a/14b process can be ascribed to the lower
rate of a (possibly) rate-determining interannular proton shift
and/or to a decreased rate of intraannular proton shift when
compared to the same process occurring within the single-ring
arenium ion. The data in Table 1 show that, whereas higher
amounts of the base enhance the relative yield of both X and
Y products with respect to methyldesilylation products, the
effect is more pronounced for X products. This finding
supports the hypothesis that 1,2-H shifts within the trimethyl-
silylated arenium ion are faster than ring-to-ring migrations,
in agreement with previous findings on tert-butylated arenium
ions, for which the different rates have been traced to the
lower preexponential factor of interannular compared with
1,2-H transfer.[6]
‡
as well as the pronounced D transfer from [Me3Si ±
‡
[D5]DPP] to Et3N in the FT-ICR cell, speak in favour of a
s complex with a covalent C ± Si bond. The existence of both p
and s complexes cannot be excluded, however. The inter-
vention of a p complex with some degree of orientation along
‡
the reaction coordinate to silylated products from the Me3Si
reaction with arenes appears reasonable and may also be
involved in the protonation-induced isomerization of o-
Me3Si ± toluene. The isomerization to the meta and especially
to the para isomers in a ratio close to that obtained from the
direct silylation of toluene by Me3Si ions suggests that 16
dissociates to an intermediate ion ± neutral complex, corre-
‡
‡
sponding to the species obtained from reaction of free Me3Si
with toluene, probably resembling 15, before undergoing
‡
Me3Si attack at an aromatic carbon.
Intermolecular Me3Si group transfer: The exceptional mobi-
lity of Me3Si , especially when it is originally bound to carbon
‡
and directed to an oxygen or halogen atom as the migration
‡
terminus, prompted the study of Me3Si transfer processes in
Conclusions
parallel with the observed proton migration processes.
Evidence for Me3Si transfer from [Me3Si ± arene] to aro-
matic compounds is provided by mass spectrometry and
radiolysis. The dependence of the efficiences of Me3Si
transfer on energetic factors emerges clearly from the relative
reactivity scale (DPP > TSB > toluene), which follows the
relative PA scale. Me3Si transfer is observed only when
thermodynamically favourable. No thermoneutral or endo-
thermic silyl migration occurs, for example, from [Me3Si ±
‡
‡
The gas-phase reactions of charged electrophiles, including
Me2Cl and C2H5 , with p-TSDPP and with p-TST as model
substrates demonstrate the occurrence of proton and Me3Si
migration processes. Furthermore, the efficiency of the
processes is affected by the presence of an additional aryl
ring providing internal solvation to the arenium ring. Whereas
the [Me3Si ± DPP] ion is adequately described by a s-
complex structure, by analogy to the [Me3Si ± toluene] ion,
‡
‡
‡
‡
‡
‡
‡
‡
arene] (arene ˆ [D8]toluene, DPP) to toluene. The reluc-
it displays an enhanced stability with regard to unimolecular
loss of Me3Si . The Me3Si group has been successfully used to
‡
‡
tance of [Me3Si ± DPP] to undergo arene exchange is
consistent with the stability of adduct 2 with respect to
unimolecular dissociation, in marked contrast to [Me3Si ±
C7H8] . These results point to the role played by the spectator
phenyl ring of DPP in stabilizing adduct 2, probably through a
chelate-type interaction. In fact, when the conformation of
the linking chain allows, as in this case, the parallel arrange-
ment of the two ring planes, a substantial intramolecular
solvation effect has been found to provide a most effective
electrostatic stabilization. Examples where this effect was
probe the intra- and interannular proton migrations within
alkylated p-TSDPP, pointing to a higher rate for the intra-
annular proton shift. This result confirms previous views
regarding the relative rates of the two processes based on the
comparison with single-ring arenium ions.
‡
Experimental Section
‡ [25]
‡ [24]
found include the DPP adduct ions DPPCr , DPPH ,
Materials: The gases were research-grade products from Matheson, with a
stated purity exceeding 99.98 mol%. Most other chemicals used were
purchased from commercial sources. p-Me3SiC6H4Me (p-TST) and o-
Me3SiC6H4Me (o-TST) were prepared from the reaction of trimethylsilyl
methanesulfonate with the Grignard reagent of the aromatic precursors, p-
BrC6H4Me and o-BrC6H4Me, respectively. o- and p-TST were purified by
preparative GLC in a 1.5-m stainless steel column packed with SE 30 (20%
w/w). p-Me3SiC6H4(CH2)3C6H5 (p-TSDPP) was a kind gift from Prof. D.
Kuck.[26] The identity of the products was confirmed by NMR and mass
spectrometric analysis.
‡ [21]
and DPPCMe3 .
‡
Structure of [Me3Si ± DPP] complexes: The study of the
reactivity of [Me3Si ± DPP] complexes provides further
valuable structural insight. Both a p-complex (15) or a s-
complex (16) structure can account for the observed Me3Si
transfer to H2O. In fact, while 15 can react only by ligand
exchange, 16 can in principle undergo both proton transfer
and Me3Si transfer, which is especially favoured when an
‡
‡
Radiolytic experiments: The gaseous systems were prepared by standard
vacuum procedures in sealed 135-mL Pyrex vessels. The competition
experiments between p-TST and p-TSDPP required a long equilibration
time (at least 2 h at 1208C) owing to the low vapour pressure of the less
volatile substrate. The resultant gaseous mixtures were submitted to g
‡
oxygenated partner is involved. However, the isolation of
‡
silylated neutral products when Me3Si is allowed to react
with DPP in the presence of Et3N under radiolytic conditions,
998
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
0947-6539/98/0406-0998 $ 17.50+.50/0
Chem. Eur. J. 1998, 4, No. 6

M. E. Crestoni
993±999
irradiation at 1208C in a 220 Gammacell (Nuclear Canada) to a total dose
of 104 Gy, at a dose rate of 104 Gyh 1. A solution of the radiolytic products
in cyclohexane was recovered after repeated freeze ± thaw cycles and
analyzed by GLC-MS using a Hewlett Packard 5890A gas chromatograph
equipped with a Model 5970B mass-selective detector or by a GC-AED
System (HP 5890 gas chromatograph, in line with an AED HP 5921A
detector under the control of a HP 5895A chemstation). The following
columns were used: a) a 100-m, 0.32-mm i.d. Petrocol DH fused silica
capillary column; b) a 50-m, 0.25-mm i.d. poly(ethyleneglycol) (Supecol-
wax 10) bonded-phase column (0.25 mm film thickness). The identity of the
products was verified by comparison of their retention volumes and mass
spectra with those of authentic standard samples.
[4] a) S. Sieber, P. von R. Schleyer, J. Am. Chem. Soc. 1993, 115, 6987;
b) M. N. Glukhovtsev, A. Pross, A. Nicolaides, L. Radom, J. Chem.
Soc. Chem. Commun. 1995, 2347.
[5] a) D. Kuck, W. Bäther, H.-F. Grützmacher, J. Am. Chem. Soc. 1979,
101, 7154; b) D. Kuck, W. Bäther, H.-F. Grützmacher, Int. J. Mass
Spectrom. Ion Processes 1985, 67, 75.
[6] F. Cacace, M. E. Crestoni, S. Fornarini, D. Kuck, J. Am. Chem. Soc.
1993, 115, 1024.
[7] a) P. von R. Schleyer, P. Buzek, T. Müller, Y. Apeloig, H.-U. Siehl,
Angew. Chem. 1993, 105, 1558; Angew. Chem. Int. Ed. Engl. 1993, 32,
1471; b) F. Cacace, M. E. Crestoni, G. dePetris, S. Fornarini, F.
Grandinetti, Can. J. Chem. 1988, 66, 3099.
[8] C. Eaborn, J. Organom. Chem. 1975, 100, 43.
FT-ICR experiments: The experiments were performed on a Bruker
Spectrospin Apex TM47e FT-ICR spectrometer with an external ion
source and a cylindrical Infinityꢁ cell situated between the poles of a 4.7 T
Á
[9] a) M. Attina, F. Cacace, A. Ricci, J. Am. Chem. Soc. 1991, 113, 5937;
Á
b) M. Attina, F. Cacace, A. Ricci, J. Phys Chem. 1996, 100, 4424.
[10] Unless otherwise stated, all thermodynamic data are taken from the
following: S. G. Lias, J. E. Bartmess, J. F. Liebman, J. L. Holmes, R. D.
Levin, N. G. Mallard, J. Phys. Chem. Ref. Data, Suppl. 1 1988, 17.
‡
superconducting magnet. The reactant Me3Si ion, generated in the
external ion source operated at 1608C by 70 eV electron impact on Me4Si
(nominal pressure 5 Â 10 6 Torr), was transferred into the cell and reacted
with C6X5R (X ˆ H, D; R ˆ H, D, CX3), introduced by a pulse valve up to
peak values of about 10 5 Torr for 10 2 s. In this way a partial cooling of the
1
[11] PA(DPP) ˆ 195.5 kcalmol estimated from ref. [24] on the basis of
the recent equilibrium constant measurements by FT-ICR.
[12] A. C. M. Wojtyniak, J. A. Stone, Int. J. Mass Spectrom. Ion Processes
‡
[Me3SiC6H5R] adduct was achieved by collisional stabilization. A C3H8/
1986, 74, 59.
iC3H7CN (90:10 mol%) mixture at a total pressure of 5 Â 10 5 Torr was
[13] T. Su, M. T. Bowers, Int. J. Mass Spectrom. Ion Phys. 1975, 17,
‡
used to prepare iC3H7CNH ions, which were allowed to protonate p-
211.
TSDPP present in the cell at the stationary pressure of 3.0 Â 10 8 Torr at
300 K. After isolation of the parent ion of interest by soft ejection
techniques, the progress of the reaction with the neutrals was followed. The
pressure readings of the ion gauge were calibrated by means of a standard
Á
[14] F. Cacace, M. Attina, S. Fornarini, Angew. Chem. 1995, 107, 754;
Angew. Chem. Int. Ed. Engl. 1995, 34, 654.
[15] M. Speranza, N. Pepe, R. Cipollini, J. Chem. Soc. Perkin Trans. II 1979,
1179.
[16] F. Cacace, M. E. Crestoni, S. Fornarini, R. Gabrielli, Int. J. Mass
Spectrom. Ion Processes 1988, 84, 17.
.
.
‡
‡
1 [27]
reaction CH4 ‡ CH4 ! CH5 ‡ CH3 (k ˆ 1.5  10 9 cm3 molecule 1s
)
and corrected for the response factor of each neutral.[28] All parent and
product ions were characterized by exact mass measurements. The gases,
the inlets and the pulsed valves were at room temperature, except in the
case of DPP and p-TSDPP, loaded in reservoir chambers kept at a constant
1208C because of their low volatility. The pseudo-first-order rate constants
were determined from the exponential decay rate of the reactant ion
intensity and converted into second-order rate constants from the known
value of the neutral pressure.
[17] H. Schwarz, The Chemistry of Organic Silicon Compounds (Eds: S.
Patai, Z. Rappoport), Wiley, Chichester, 1989, p. 445.
[18] M. E. Crestoni, S. Fornarini, unpublished results.
[19] I. Fleming in Comprehensive Organic Chemistry, Vol. 3 (Eds.: D.
Barton, W. D. Ollis), Pergamon, Oxford, 1979.
[20] M. E. Crestoni, S. Fornarini, Angew. Chem. Int. Ed. Engl. 1994, 106,
1157; Angew. Chem. Int. Ed. Engl. 1994, 33, 1094.
[21] B. Chiavarino, M. E. Crestoni, S. Fornarini, D. Kuck, Int. J. Mass
Spectrom. Ion Processes 1995, 148, 215.
[22] M. E. Crestoni, J. Phys. Chem. 1993, 97, 6197.
[23] M. E. Crestoni, S. Fornarini, D. Kuck, J. Phys. Chem. 1995, 99,
3144.
[24] M. E. Crestoni, S. Fornarini, D. Kuck, J. Phys. Chem. 1995, 99,
3150.
Acknowledgements: This research was supported by the Italian Ministry
for University and Scientific and Technological Research (MURST). The
author is grateful to Prof. Fulvio Cacace and to Prof. Simonetta Fornarini
for valuable discussions and to Prof. Dietmar Kuck for supplying a sample
of p-TSDPP.
Received: July 21, 1997 [F775]
[25] M. E. Crestoni, S. Fornarini, Organometallics 1996, 15, 5695.
[26] a) C. Matthias, Doctoral Thesis, University of Bielefeld, 1996; b) C.
Matthias, D. Kuck, unpublished results.
[27] M. Meot-Ner in Gas Phase Ion Chemistry, Vol. 1 (Ed.: M. T. Bowers),
Academic Press, New York, 1979.
[1] F. Cacace, Acc. Chem. Res. 1988, 21, 215.
[2] D. Kuck, Mass Spectrom. Rev. 1990, 9, 583.
[3] a) F. Cacace, M. E. Crestoni, S. Fornarini, J. Am. Chem. Soc. 1992,
114, 6776; b) G. Cerichelli, M. E. Crestoni, S. Fornarini, ibid. 1992,
114, 2002.
[28] J. E. Bartmess, R. M. Georgiadis, Vacuum 1983, 33, 149.
Chem. Eur. J. 1998, 4, No. 6
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
0947-6539/98/0406-0999 $ 17.50+.25/0
999