Electron-induced dissociation of chlorosilanes: Role of aromatic side groups in gas phase and solution chemistry

Article abstract of DOI:10.1016/j.ijms.2013.12.011

Dissociative electron attachment (DEA) to chlorotrimethylsilane (Me ₃SiCl) and chlorodimethylphenylsilane (PhMe₂SiCl) was studied to reveal a possible enhancing effect of the aromatic side group on the SiCl bond dissociation. Calculations point to a relevant interaction between the lowest unoccupied π* orbital of the phenyl ring and the σ*(SiCl) orbital and a consequent lowering of the lowest-energy electron attachment channel. It was previously reported that such a situation can lead to a significant enhancement of DEA. However, the lowest energy DEA channel in the investigated chlorosilanes lies below the thermodynamic limit for dissociation of the SiCl bond in the radical anion and thus cannot contribute to the Cl^- yield. In consequence, release of chloride ions is in fact the dominant fragmentation channel but the Cl^- yield is not noticeably enhanced in PhMe₂SiCl. In contrast to the DEA results, comparative experiments on the reductive coupling in THF of Me₃SiCl, PhMe₂SiCl, and benzyldimethylchlorosilane (BzMe₂SiCl) under identical conditions reveal a significant enhancement of the reactivity by aromatic side groups. As this finding does not correlate with the DEA results, an alternative explanation for the different reactivity of Me₃SiCl and PhMe₂SiCl in solution is suggested based on their vertical electron attachment energies and considering the impact of these quantities on the equilibrium between solvated electrons and neutral reactants on one side and the solvated radical anion on the other.

Full text of DOI:10.1016/j.ijms.2013.12.011

Accepted Manuscript
Title: Electron-induced dissociation of chlorosilanes: Role of
aromatic side groups in gas phase and solution chemistry
¨
Author: E. Bohler J. Postler D. Gschliesser T. Borrmann P.
Scheier S. Denifl J. Beckmann P. Swiderek
PII:
S1387-3806(13)00447-8
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.ijms.2013.12.011
MASPEC 15067
To appear in:
International Journal of Mass Spectrometry
Received date:
Revised date:
Accepted date:
9-10-2013
11-12-2013
13-12-2013
¨
Please cite this article as: E. Bohler, J. Postler, D. Gschliesser, T. Borrmann, P. Scheier,
S. Denifl, J. Beckmann, P. Swiderek, Electron-induced dissociation of chlorosilanes:
Role of aromatic side groups in gas phase and solution chemistry, International Journal
of Mass Spectrometry (2013), http://dx.doi.org/10.1016/j.ijms.2013.12.011
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

1
Graphical abstract
The role of aromatic side groups in dissociative electron attachment to chlorosilanes is
investigated and the relation of this process to reductive coupling reactions is discussed.
Page 1 of 34

2
Highlights
• Dissociative electron attachment (DEA) to chlorosilanes in the gas phase is studied.
• An enhancing effect of an aromatic side group on DEA has not been observed.
• The rate of reductive coupling of chlorosilanes in THF using Li is investigated.
• Reductive coupling is accelerated by an aromatic side group.
• The influence of an aromatic side group on DEA and reductive coupling is discussed.
Page 2 of 34

3
Electron-induced dissociation of chlorosilanes:
Role of aromatic side groups in gas phase and solution chemistry
E. Böhlera, J. Postlerb, D. Gschliesserb, T. Borrmanna, P. Scheierb, S. Deniflb, J. Beckmannc,
P. Swidereka
a Universität Bremen, Institut für Angewandte und Physikalische Chemie,
Fachbereich 2 (Chemie/Biologie), Leobener Straße / NW 2,
Postfach 330440, D-28334 Bremen, Germany
b Leopold-Franzens-Universität Innsbruck, Institut für Ionenphysik und Angewandte Physik,
Technikerstraße 25, A-6020, Innsbruck, Austria
c Universität Bremen, Institut für Anorganische Chemie,
Fachbereich 2 (Chemie/Biologie), Leobener Straße / NW 2,
Postfach 330440, D-28334 Bremen, Germany
Abstract
Dissociative electron attachment (DEA) to chlorotrimethylsilane (Me3SiCl) and
chlorodimethylphenylsilane (PhMe2SiCl) was studied to reveal a possible enhancing effect of
the aromatic side group on the Si-Cl bond dissociation. Calculations point to a relevant
interaction between the lowest unoccupied π* orbital of the phenyl ring and the σ*(Si-Cl)
orbital and a consequent lowering of the lowest-energy electron attachment channel. It was
previously reported that such a situation can lead to a significant enhancement of DEA.
However, the lowest energy DEA channel in the investigated chlorosilanes lies below the
thermodynamic limit for dissociation of the Si-Cl bond in the radical anion and thus cannot
contribute to the Cl- yield. In consequence, release of chloride ions is in fact the dominant
fragmentation channel but the Cl- yield is not noticeably enhanced in PhMe2SiCl. In contrast
to the DEA results, comparative experiments on the reductive coupling in THF of Me3SiCl,
PhMe2SiCl, and benzyldimethylchlorosilane (BzMe2SiCl) under identical conditions reveal a
Page 3 of 34

4
significant enhancement of the reactivity by aromatic side groups. As this finding does not
correlate with the DEA results, an alternative explanation for the different reactivity of
Me3SiCl and PhMe2SiCl in solution is suggested based on their vertical electron attachment
energies and considering the impact of these quantities on the equilibrium between solvated
electrons and neutral reactants on one side and the solvated radical anion on the other.
Keywords: Chlorosilanes, dissociative electron attachment, reductive coupling
Page 4 of 34

5
1. Introduction
Controlling a reaction by oxidation or reduction, i.e., by electron transfer from or to a
molecule is a fundamental concept in chemistry. Experiments on electron-molecule
interactions in the gas phase monitor the initiating step of such reactions /1/ and can reveal
influences of the molecular structure on their efficiency. In particular, information on bond
cleavage following electron transfer to a molecule can be obtained from experiments on
dissociative electron attachment (DEA). In fact, recent results on DEA to exo-5-chloro-2-
norbornene and its saturated analogue, exo-2-chloro-norbornane, have shown that dissociation
of the C-Cl bond is enhanced by the presence of a double bond if an antibonding π* orbital
can mix with the antibonding σ*(C-Cl) orbital /2/. Electron attachment into σ*(C-Cl) of alkyl
halides such as exo-2-chloro-norbornane induces dissociation of this bond releasing Cl-
without a further intermediate:
R3C-Cl + e- → R3C• + Cl-
(1)
In the unsaturated analogue, exo-5-chloro-2-norbornene, this process occurs with higher cross
section than DEA to the saturated analogue, exo-2-chloro-norbornane, suggesting that the
double bond enhances the reaction by acting as electron capturing site which then transfers
the electron to the C-Cl bond /2/. A similar intramolecular electron transfer has also been
described in the case of DEA to thymidine /3/ and an enhancing effect of aromatic side groups
was also observed in differently substituted triflates /4/.
An enhancing effect of aromatic side groups has also been observed in the case of reactions in
solution. The reductive coupling of dichlorosilanes using alkali metals that is broadly applied
in the synthesis of oligo- and polysilanes /5,6/ is an example of such a reaction. As a simpler
case, chlorosilanes can be transformed to disilanes /7,8/ by the following reaction sequence
Page 5 of 34

6
R3SiCl + Li → R3Si• + Cl- + Li+
R3Si• + Li → R3Si- + Li+
(2)
(3)
(4)
R3Si- + R3SiCl → R3Si-SiR3
This Wurtz-type reaction is typically performed in a dispersion or suspension of an alkali
metal, e.g. lithium, in an inert solvent like toluene or tetrahydrofuran (THF). The initiating
step of reactions such as (2) is an electron transfer from the metal to the reactant molecule,
Li(s) + Reactant(org) → Li+(org) + Reactant•-(org),
(5)
which, in the case of chlorosilanes, leads to dissociation of the Si-Cl bond /7/. However, the
reaction proceeds considerably faster when at least one of the substituents R is an aryl group
/8,9/ but can also be accelerated by sonic waves /10,11/ or, more importantly, by adding either
catalytic or stochiometric amounts of aromatic compounds /10,12/ or potassium graphite
/13,14/.
These examples raise the question if an enhancement of DEA correlates with an increased rate
of dissociative electron transfer in solution reactions of the same compound. We have thus
investigated, as a simple example, if results on DEA correlate with the difference in the
reactivity towards reductive coupling of two chlorosilanes (Fig. 1), namely
chlorotrimethylsilane (Me3SiCl) and chlorodimethylphenylsilane (PhMe2SiCl). As reference,
an NMR study on the reactivity towards reductive coupling of Me3SiCl, PhMe2SiCl, and
benzyldimethylchlorosilane (BzMe2SiCl) under exactly comparable condition is described
first. To search for evidence on intramolecular electron transfer in DEA to PhMe2SiCl via
electron attachment to a low-lying π*orbital and thus a mediating effect of the phenyl group
in cleavage of the Si-Cl bond, we present gas phase DEA experiments on Me3SiCl and
PhMe2SiCl. Finally, the relevance of these DEA results for reductive coupling in solution is
discussed leading to a novel explanation for the observed difference in reactivity of Me3SiCl
and PhMe2SiCl.
Page 6 of 34

7
Figure 1: Molecular structures of the investigated chlorosilanes.
2. Experimental
2.1. Chemicals
Chlorotrimethylsilane (Me3SiCl) was purchased from Fluka at a stated purity of 99%,
chlorodimethylphenylsilane (PhMe2SiCl, 98%) from Aldrich, and benzyldimethylchlorosilane
(BzMe2SiCl, 97%) from ABCR. All samples were used as received.
2.2. Reaction of chlorosilanes with lithium and NMR spectroscopy
0.5 g (0.072 mol) granulated lithium (1-3 mm) was added in a glove box under argon
atmosphere to a 100 ml Schlenk flask that was previously evacuated and flushed with dry
argon. 10 ml dry THF (HPLC grade) and 7.2 mmol chlorosilane were added under argon
atmosphere. The flask was cooled to -18°C using an ethanol/liquid nitrogen mixture. At begin
of the reaction and after 5, 10, 30, 60, and 120 minutes 0.4 ml of the reaction mixtures were
removed from the flask and the remaining solid lithium and added to an NMR tube. 0.6 ml
deuterated chloroform (99.8% D) were added to the samples for 1H-NMR analysis (AVANCE
NB-360).
Page 7 of 34

8
2.3. Dissociative electron attachment experiments
DEA to Me3SiCl and PhMe2SiCl at incident electron energies between 0 and 20 eV was
investigated using a setup in which molecular beams of the compounds were crossed under
single collision conditions with a beam of low-energy electrons having an energy resolution
of approximately 1 eV at an electron current of 10 µA. The setup with a base pressure of
2×10-7 mbar was described in detail previously /15/. Anions formed in the ion source were
extracted by a weak electric field and accelerated into the mass spectrometer by a potential
drop of several keV before being mass-selected using a high resolution double focusing two
sector mass spectrometer (VG-ZAB) with reversed Nier-Johnson-type geometry (magnetic
sector field followed by the electric sector field) and detected by a channeltron-type secondary
electron multiplier, operated in the pulse counting regime.
The chlorosilanes were filled in a small metal tube and connected to the vacuum chamber.
The samples were degassed by repeated freeze-pump-thaw cycles prior to the experiments
and introduced into the collision chamber at room temperature through a gas inlet. The
pressure in the ion source was comparable for the two chlorosilanes (10-5 mbar for Me3SiCl
and 8*10-6 mbar for PhMe2SiCl). Complete negative ion mass spectra were recorded for
several electron energies to identify relevant negative ion fragments for which ion yield
curves were then measured. The electron energy (E0) scale was calibrated using several
resonances as observed in DEA to SF6 /16,17/.
3. Calculations
Molecular geometries and orbitals of Me3SiCl, PhMe2SiCl, and benzene were calculated on
the B3LYP/6-31G* level using Gaussian98 /18/. Following the approach of Modelli /19/, the
virtual orbital energies were empirically corrected to yield predictions for the vertical
Page 8 of 34

9
attachment energies (VAEs). The nature of the relevant orbitals was determined visually from
the plots.
4. Results
4.1. Reactivity of different chlorosilanes in reductive coupling with Li
Although previous work has shown that the reductive coupling of chlorosilanes to disilanes
proceeds much faster in the presence of at least one aryl group on the silicon /8,9/, the studies
were often not performed under exactly comparable conditions or have reported only yields
after some arbitrarily chosen reaction time. Here, the consumption of the chlorosilanes is
quantitatively investigated by monitoring the methyl group 1H NMR signals that appear at
slightly different chemical shifts for the parent compounds and the resulting products. Fig. 2
thus shows the percentage of the methyl group signals ascribed to the parent compounds as a
function of reaction time. It is obvious that both compounds containing an aryl group undergo
a rapid reaction in the presence of Li and are completely consumed after 30 min while
Me3SiCl does not react at all within the same time and under the same conditions. The
presence of an aryl group thus in fact enhances the reaction rate.
Page 9 of 34

10
1.0
0.8
0.6
0.4
0.2
0.0
Me3SiCl
PhMe2SiCl
BzMe2SiCl
0
5
10
15
20
25
30
35
t (min)
Figure 2: Fraction of methyl group 1H NMR intensities in the reaction mixture ascribed to the
parent compounds Me3SiCl (-0.06 ppm), PhMe2SiCl (0.57 ppm), and BzMe2SiCl (0.27 ppm)
as function of reaction time.
4.2. Dissociative electron attachment to trimethylchlorosilane and
phenyldimethylchlorosilane
Negative ion mass spectra recorded at several electron energies (E0) reveal clearly that Cl- is
in fact produced as the dominant negative fragment from both Me3SiCl and PhMe2SiCl under
electron exposure. The ion yield curves for Cl- are presented in Fig. 3. However, DEA also
induces cleavage of Si-C bonds leading to loss of neutral or anionic organic side groups CH3
and C6H5 (Fig. 4) as well as cleavage of C-H bonds leading to loss of neutral H (Fig. 5).
Furthermore, for Me3SiCl, the formation of CH3Cl- gives evidence of bond reorganisation. As
the pressure within the ion source was similar for the two compounds we assume that the gas
density in the molecular beam was also of the same order of magnitude. The observed ion
yield values should thus be comparable within an order of magnitude.
Page 10 of 34

11
DEA to Me3SiCl and PhMe2SiCl is studied for the first time. Therefore, previous results on
resonant electron attachment to other silanes and further related molecules must serve as
reference here. These data include measurements of total scattering cross sections in SiHnCl4-n
revealing resonances, however, with results on SiH3Cl missing /20/. A related work has
reported the total anion yield by DEA to SiHCl3 /21/. Resonances in the electron-molecule
interaction for tetramethylsilane, SiMe4, were observed in the energy dependence of
vibrational excitation /22/. Furthermore, results exist on DEA to SiH4 and CH4 /23/ as well as
to benzene /24,25/. Finally, Table I summarizes bond dissociation energies and electron
affinities that are needed to estimate thermodynamic low-energy limits for the thresholds of
formation of the different anions in DEA.
4.2.1 Formation of Cl-
The energy dependent ion yield curves also show that Cl- formation is the dominant electron-
induced fragmentation channel for both investigated chlorosilanes (Fig. 3). Cl- formation from
Me3SiCl is pronounced between 5.5 eV and 12 eV with maximum at 7 eV. DEA to
PhMe2SiCl yields Cl- within the same energy range but the maximum yield is shifted to
higher energy and the yield remains at a high level above 12 eV. In addition, signals are
observed at or below 1 eV for both compounds. However, taking BDE(Si-Cl) for Me3SiCl as
5.1 eV and EA(Cl) as 3.6 eV (Table I), the dissociation limit for the reaction
Me3Si-Cl + e- → Me3Si• + Cl-
(6)
results as 1.5 eV. In consequence, the Cl- signals below 1 eV must be ascribed to an impurity.
The low-energy signal compares closely with that observed previously for Cl- production by
DEA to HCl /28,29/. HCl is likely to be released from chlorosilanes by hydrolysis in the
presence of small impurities of water.
Page 11 of 34

12
Release of Cl- near threshold has been observed previously in the case of SiHCl3 but the
yields were low and a similar signal was even missing from SiH2Cl2 and SiCl4 /21/.
Unfortunately, measurements on SiH3Cl that would be the closest analogue to the compounds
investigated here were not reported at all. The impurity signal in the present data still levels
off at the position of the estimated threshold of 1.5 eV for release of Cl- from the
chlorosilanes and thus partly overlaps a small but visible signal extending up to 3.5 eV in the
case of Me3SiCl (Fig. 3, lower curve). However, it is obvious that the intensity in the same
range is even smaller in the case of PhMe2SiCl, contrary to the expected mediating effect of
the phenyl group but in agreement with the low Cl- yield for SiHCl3. On the other hand, the
maximum yields in the higher energy range are roughly comparable for Me3SiCl and
PhMe2SiCl. In conclusion, we do not find evidence for a significant enhancement of Si-Cl
bond cleavage by the presence of the aromatic substituent.
3x104
2x104
m/z 35
Cl-
PhMe2SiCl
1x104
0
3x104 Me3SiCl
2x104
m/z 35
Cl-
1x104
0
0
2
4
6
8
10
12
14
Electron energy (eV)
Fig. 3:Cl- formation by DEA to Me3SiCl and PhMe2SiCl.
Page 12 of 34

13
4.2.2 Products of Si-C cleavage
The ion yield curves of negative fragments resulting from cleavage of Si-C bonds in Me3SiCl
-
([M-CH3]-, [M-2CH3]-, and CH3 , M denoting the intact parent molecule) are very similar and
show only one relatively narrow peak with maximum again at 7 eV (Fig. 4a). In addition, a
small signal of CH3Cl- was observed with a comparable peak position and shape (not shown).
While we may only speculate on the nature of the resonance underlying this dissociation
process, it most likely involves a delocalized anion state because it also contributes to Si-Cl
bond cleavage. The pattern becomes more complex in PhMe2SiCl (Fig. 4b). Here,
fragmentation appears to proceed via several resonances. [M-CH3]- is formed with maximum
yield at again 7 eV and with smaller intensity around 10.5 eV, while [M-C6H5]- is observed
-
with a broader signal between 5 eV and 9 eV and C6H5 shows a maximum at 7.5 eV and a
broad overlapping feature around 10 eV. Again, information from other sources on the nature
of these resonances is so far non-existent. However, it is well conceivable that coupling of
orbitals of the phenyl ring with those of the saturated part of the molecule produces a
sequence of anion states, each contributing to DEA. All observed Si-C dissociation processes
appear well above their thermodynamic limits as can be estimated from the values listed in
Table I. This is even the case for formation of [M-2CH3]- taking into account that the
necessary bond dissociation energy of 7.8 eV can be compensated for by formation of a new
C-C bond through recombination of the two CH3 radicals. By this, the threshold decreases to a
maximum value of 3.9 eV, the electron affinity of MeSiCl being unknown.
The ion yields for analogous dissociation processes are again of the same order of magnitude
for Me3SiCl and PhMe2SiCl. This is most obvious for loss of CH3 leading to [M-CH3]-, were
the maximum yields differ by only a factor of 2.5 between the two parent compounds. Also,
-
-
the ion yields for the concurrent formation of CH3 from Me3SiCl and of C6H5 from
Page 13 of 34

14
PhMe2SiCl are both small. In conclusion, DEA leading to dissociation of Si-C bonds is also
not enhanced by the presence of the phenyl ring.
400
120
(a)
Me3SiCl
m/z 93
[M-CH3]-
m/z 155
[M-CH3]-
(b)
PhMe2SiCl
100
80
60
40
20
0
300
200
100
0
8
m/z 78
[M-2CH3]-
m/z 93
[M-C6H5]-
30
20
10
0
6
4
2
0
20
m/z 15
CH3-
m/z 77
C6H5-
40
30
20
10
0
15
10
5
0
0
2
4
6
8
10
12
14
0
2
4
6
8
10
12
14
Electron energy (eV)
Electron energy (eV)
Fig. 4: Si-C dissociation by DEA to Me3SiCl (a) and PhMe2SiCl (b).
4.2.3 Products of C-H cleavage
Loss of neutral H occurs through at least two resonances. The lowest one has a maximum
yield at 3.9 eV for Me3SiCl and at 3.5 eV for PhMe2SiCl (Fig. 5). A similar resonance has
been observed previously in SiMe4 by vibrational electron energy loss spectroscopy /22/. This
study reported a resonant enhancement of the first CH stretching overtone with maximum
intensity around 4 eV. The C-H dissociation through DEA in the chlorosilanes (Fig. 5) is in
accord with this result. In contrast, SiH4 and CH4 reveal anion formation by DEA only around
roughly 9 eV and 10 eV, respectively /23/. A very small signal above 10 eV in the case of
Me3SiCl (Fig. 5) may correlate with C-H bond dissociation channels known from alkanes like
CH4. However, the presence of a Si-C bond clearly leads to an additional low-energy DEA
Page 14 of 34

15
channel. This is also supported by previous calculations assigning the 4 eV resonance
observed in vibrational electron energy loss spectra to electron attachment into a σ*(Si-C)
orbital /22/. In addition, gas phase electron transmission experiments locate the unoccupied π*
orbitals of benzene at 1.15eV (2E2g) and 4.8eV (2B2g) /24/. This may suggest that electron
attachment into the b2g orbital also contributes to the dissociation channel around 3.5 eV of
PhMe2SiCl.
A second major DEA signal is observed around 7.5 eV for Me3SiCl and around 8 eV for
PhMe2SiCl. The latter in addition shows a small signal around 6 eV. The yield of [M-H]-, M
being again the parent compound, is clearly enhanced by the presence of the phenyl ring. In
fact, the higher energy signal in PhMe2SiCl is about two orders of magnitude more intense
that in Me3SiCl. However, this is not surprising because the 8 eV peak in the latter correlates
-
well a known DEA process in benzene that leads to formation of C6H5 /25/. In conclusion,
this enhancement does not give evidence of an electron transfer process but can be traced
back to a specific dissociation process localized on the phenyl ring.
PhMe2SiCl
Me3SiCl
m/z 169
[M-H]-
1500
1000
500
0
50
40
30
20
10
0
m/z 107
[M-H]-
0
2
4
6
8
10
12
14
Electron energy (eV)
Fig. 5: C-H dissociation by DEA to Me3SiCl and PhMe2SiCl.
Page 15 of 34

16
4. Discussion
The results reveal in fact that Me3SiCl is much less reactive towards reductive coupling by Li
in THF than its analogues carrying phenyl side groups, namely, PhMe2SiCl and BzMe2SiCl.
In fact, aromatic compounds have been described to act as mediators in organic electron
transfer reactions /30-33/. This effect has been ascribed to formation of an intermediate
radical anion of the mediator according to reaction sequence (7) and (8) which reduces the
overall activation barrier for electron transfer from the metal to the reactant molecule /30/.
Li(s) + Mediator(org) → Li+(org) + Mediator •-(org)
(7)
(8)
Mediator•-(org) + Reactant(org) → Mediator(org) + Reactant•-(org)
This is based on the assumption that the total activation barrier for the electron transfer
depends on the three different activation energies EA, namely, for dissolution of lithium,
Li(s) → Li+(org) + e-,
(9)
for electron attachment to the mediator
Mediator(org) + e- → Mediator•-(org),
and formation of the radical anion of the reactant,
Reactant(org) + e- → Reactant •-(org).
(10)
(11)
According to this model /30/, the activation energy of the direct electron transfer (5) may be
estimated from the sum of EA for reactions (9) and (11). Analogous relations hold for the two
steps (7) and (8) of the mediated electron transfer. A key property of an efficient mediator is a
particularly low value of EA for electron attachment (10) so that the overall barrier of the
mediated electron transfer is considerably reduced as compared to the direct transfer. This
mediator model suggests that aromatic side groups of chlorosilanes may act as an
Page 16 of 34

17
intramolecular mediator and should thus result in a lower EA for electron attachment (10) as
compared to aliphatic chlorosilanes.
In contrast, the DEA data on Me3SiCl and PhMe2SiCl in the gas phase do not show a
noticeable difference in release of Cl- between the two compounds. This also appears to
contradict the previous DEA study on exo-5-chloro-2-norbornene and its saturated analogue
exo-2-chloro-norbornane, which has in fact clearly revealed that dissociation yielding Cl- is
enhanced by nearly two orders of magnitude and occurs at lower E0 when the molecule
contains a π* orbital that can interact with σ*(C-Cl) /2/. It is therefore at first sight surprising
that a similar mediating effect of the phenyl group is not observed in the chlorosilanes.
However, this can be explained by considering the molecular orbitals involved in the electron
attachment and dissociation processes and the energetics of C-Cl and Si-Cl bond dissociation.
Electron attachment occurs at a shorter time scale than nuclear motion at E0 corresponding to
a vertical transition between the neutral ground state and a particular anion potential energy
curve. Electron transmission spectroscopy (ETS) reveals such vertical attachment energies
(VAEs) which can, in the simplest case, be approximately correlated with the energy of a
particular orbital into which the electron is attached /19/. In the case of exo-2-chloro-
norbornane a VAE of 2.3 eV was observed and assigned to the σ*(C-Cl) orbital /2/. We note
that the maximum DEA yield typically occurs at somewhat lower E0, in this case 1.65 eV /2/
because the probability of the anion state to survive beyond the crossing point with the ground
state potential energy curve upon bond elongation is higher for a transition at lower energy
/34/. Two VAEs of 1.10 eV and 2.78 eV were measured for exo-5-chloro-2-norbornene /2/.
These two states result from the interaction between σ*(C-Cl) and the π* orbital of the double
bond, the latter having a VAE of 1.78 eV in the case of ethylene and at somewhat higher
values in alkyl substituted ethylenes /35/. Calculations have shown that the orbital at 1.10 eV
has mostly π* character while the orbital at 2.78 eV corresponds mainly to σ*(C-Cl) /2/
Page 17 of 34

18
(Fig. 6a). Following attachment into the lower-lying orbital, the electron is, due to the
coupling, transferred to σ*(C-Cl), leading to efficient DEA.
Experimental VAEs are so far unknown for Me3SiCl and PhMe2SiCl. Fig. 6b thus shows the
orbital correlation this case as derived from orbital energies calculated on the B3LYP/6-31G*
level and using the empirical correlation between VAEs and orbital energies /19/. Here, the
σ*(Si-Cl) located at 2.3 eV interacts with one component of the degenerate lowest π* orbital
of a benzene ring, the latter having an experimental VAE of 1.15 eV /24,35/, to yield a
predominantly π* type orbital at 0.7 eV and a mainly σ*(Si-Cl) orbital at 2.49 eV. However,
and in contrast to the case of C-Cl bond cleavage, dissociation of the Si-Cl bond is
energetically not accessible upon electron attachment into the lower orbital. This can be
traced back to the different bond dissociation energies of C-Cl and Si-Cl bonds (Table I).
While the former amounts to roughly 3.7 eV as determined for (C2H5)(CH3)2C-Cl, the BDE
for (CH3)3Si-Cl has been reported as 4.9 eV. Considering again EA(Cl) = 3.6 eV, this leads to
the picture shown in Fig. 7a. The dissociation threshold for reaction
Me3C-Cl + e- → Me3C• + Cl-
(12)
results as BDE(C-Cl) – EA(Cl). As EA(Cl) outbalances BDE(C-Cl), DEA to halocarbons can
proceed already at E0 near 0 eV given that suitable unoccupied orbitals exist. DEA via the π*
orbital at 1.10 eV in exo-5-chloro-2-norbornene can thus proceed readily. In contrast, as
shown in Section 4.2.1, the thermodynamic low-energy limit for release of Cl- from a
chlorosilane is 1.5 eV (Fig. 7b). Assuming the energetic situation to be similar for
PhMe2Si-Cl, this suggests that electron attachment into the predominantly π* type orbital at
0.7 eV cannot induce dissociation of a Si-Cl bond according to
PhMe2Si-Cl + e- → PhMe2Si• + Cl-
(13).
Page 18 of 34

19
Dissociation can thus only proceed via the higher lying predominantly σ* type orbital
predicted at 2.5 eV in PhMe2Si-Cl (Fig. 6) but this process is obviously also negligible in exo-
5-chloro-2-norbornene /2/ where it would be expected below the VAE of 2.3 eV.
Fig. 6: Correlation of the vertical attachment energies for the low-lying unoccupied orbitals
of (a) exo-5-chloro-2-norbornene with those of exo-2-chloronorbornane and 2-butene
according to Refs. 2,29 and of (b) PhMe2SiCl with those of Me3SiCl and benzene derived
from present calculations. The experimental value for benzene was taken from Refs. 24,35.
M-
E
M
M-
(a) M = R3CCl
(b) M = R3SiCl
E
M
M-solv
R3C + Cl
R3Si + Cl
BDE EA
BDE
EA
R3C + Cl-
R3Si + Cl-
EA
0
0
R3Si solv+ Cl-solv
rSi-Cl
rC-Cl
Fig. 7: Schematic potential energy curves for DEA processes yielding Cl- in from C-Cl and
Si-Cl bond dissociation.
Page 19 of 34

20
The preceding discussion provides an explanation why gas phase DEA measurements do not
give insight into a possible mediator effect of the phenyl side groups in chlorosilanes. A
mediator in a solution electron transfer reaction has a low EA for electron attachment and thus
can act as an intermediate electron carrier in the electron transfer between the alkali metal and
the reactant /30/. As a typical example, EA for electron transfer to a typical mediator 4,4’-di-
tert-butyl-1,1`-biphenyl (DBB) (reaction (10)) was reported as only 8.2 kJ/mol (0.08 eV) in
THF /30/. In contrast, EA for the Li/Li+ couple in THF (reaction (11) was determined as
80 kJ/mol (0.83 eV) /36/. The overall EA for a redox couple can be estimated from the sum of
the two half reactions involved and consequently amounts to roughly 88 kJ/mol (0.91 eV) for
reaction (7). It was thus concluded that the mediated reaction should proceed faster than the
direct reaction given that EA for reaction (11) is larger than about 10 kJ/mol (0.1 eV) /30/.
In principle, EA values for DEA as identified with reaction (1) are defined by the height of the
crossing point between the potential energy curves along the C-Cl or Si-Cl bond distances of
the neutral reactant in its ground state and the radical anion (Fig. 7) /37/. A similar model has
been put forward by Savéant as an extension of Marcus theory for electron transfer to describe
dissociative electron transfer in solution. Using again alkyl halides as example, it was shown
that the crossing point of the two potential energy curves defines the activation energy for
DEA /38/. However, this gas phase EA is not applicable to the situation in solution as the
potential energy curve for the anion state shifts more strongly towards lower energy by
polarization of the surrounding medium than that of the neutral state /34/. This would lower
the activation barrier but a reliable estimate requires calculations including a solvation model.
A recent experimental study on equilibrium of the benzene radical anion with neutral benzene
and solvated electrons in THF gives valuable insight from a thermodynamic perspective /39/.
At room temperature, the equilibrium constant for the reaction
Page 20 of 34

21
•-
e-(THF) + C6H6(THF) ' C6H6 (THF)
(14)
was determined as 0.22 M-1 indicating that Benzene dissolved in THF is mainly present as a
neutral species and that the solvated benzene radical anion is about 40 meV less stable than
the solvated electron /39/. Compared to the gas phase where the EA for benzene is -1.15 eV,
corresponding to a VAE of 1.15 eV /39/, and the energy of the electron at rest is 0 eV, this
implies that the molecular radical anion is stabilized more strongly than the electron when
dissolved in THF. Calculations have yielded a difference in solvation energy of 2.45 eV
between neutral benzene and its radical anion /39/. A similar value was obtained for the
solvation energy of the electron but here an additional destabilizing confinement energy
brings its total energy to the value of 40 meV below that of the benzene radical anion /39/.
The described equilibrium implies that benzene radical anions formed upon contact with the
Li metal would preferably dissociate to neutral benzene and a solvated electron. In contrast,
the equilibrium constant for fluorobenzene under the same conditions was determined as
175 M-1 /39/. This is consistent with a lower VAE of 0.89 eV into the lowest π* orbital as
compared to the value of 1.15 eV in benzene /35/ under the assumption that the solvation
energies of benzene and fluorobenzene are similar. This assumption is reasonable based on
the finding that even a single electron has a comparable solvation energy /39/. The
equilibrium constant also implies that the fluorobenzene radical anion is more stable by
130 meV in THF compared to the solvated electron and would thus probably be available in a
higher concentration to transfer its excess electron onto a reactant if used as mediator. In
contrast, electron transfer to THF has not been observed and this negative result has been
ascribed to its much higher LUMO energy /39/.
Considering the calculated VAEs for the two chlorosilanes (Fig. 6b) and assuming again a
similar solvation energy as for benzene, it is obvious that the energy of the radical anion of
PhMe2SiCl should be even lower than that of fluorobenzene implying an even higher
Page 21 of 34

22
equilibrium constant, i.e., a higher concentration of radical anions. A high concentration of
the intermediate is in favor of subsequent dissociation even if this process is still subject to an
activation barrier. On the contrary, the VAE of Me3SiCl is predicted to be 1 eV higher than
that of benzene suggesting that the equilibrium in THF should be much in favor of the
solvated electron, i.e., only a very small concentration of radical anions is to be expected in
this case. This difference is clearly consistent with the observed differences in reactivity of
Me3SiCl and PhMe2SiCl. VAEs measured in the gas phase thus appear to be a useful quantity
to rationalize the different reactivity in solution phase electron transfer reactions given that all
other conditions are comparable.
5. Conclusions
Chlorosilanes carrying aromatic substituents are reduced much faster by metallic lithium than
their aliphatic analogues. The aromatic side group thus appears to act as mediator in reactions
such as the reductive coupling of chlorosilanes which are typically performed using metallic
lithium dispersed in THF. However, the yield of Cl- formed by dissociative electron
attachment (DEA) in the gas phase is not enhanced in PhMe2SiCl as compared to Me3SiCl.
This is explained by the fact that the lowest unoccupied orbital of PhMe2SiCl which results
from mixing of the lowest unoccupied π* orbital of the phenyl ring with the antibonding
σ*(Si-Cl) orbital is located below the thermodynamic threshold for dissociation of the Si-Cl
bond of the radical anion of PhMe2SiCl. The lowest DEA channel is thus closed, i.e.,
dissociation following electron attachment into the lowest unoccupied orbital is not
energetically accessible in PhMe2SiCl.
The origin of the mediating action of the aromatic side groups in solution can be explained by
considering the equilibrium between solvated electrons and neutral reactants on one side and
the solvated radical anion on the other. Using a previous detailed study on the benzene radical
Page 22 of 34

23
anion in THF as reference and assuming that the solvation energies are similar for molecules
of similar molecular structure, the relative energies of different radical anions can be related
to the vertical electron attachment energies (VAEs) in the gas phase. In consequence, VAEs
can be correlated with the expected radical anion concentrations and thus the rate of
subsequent reactions of these radical anions in solution, given that all other parameters
(temperature, solvent, concentrations) applied are identical. This yields a novel explanation
for the reactivities of different chlorosilanes in reductive coupling reactions.
The present results on DEA are also of more general relevance, as interactions with electrons
play an important role in the chemistry of silicon compounds and have applications in
different technical processes. This includes electron- and plasma-initiated crosslinking
reactions for the fabrication of thin coatings /40/ and electron beam induced deposition
(EBID) to produce nanoscale silicon oxide structures by decomposition of volatile silicon
compound like silanes and alkoxysilanes /41/. In addition, possible contributions of electron-
driven processes to astrochemistry have been discussed /42,43/ and may also be relevant to
silicon chemistry in cosmic environments where low-valent molecules such as SiC, SiH, and
SiO have been observed /44/.
Acknowledgement
Support provided by the Cost Action CM0805 ‘‘The chemical cosmos: Understanding
chemistry in astrochemical environments’’ is gratefully acknowledged.
References
/1/
/2/
E. Böhler, J. Warneke, P. Swiderek, Chem. Soc. Rev. 42 (2013) 9219-9231.
D.M. Pearl, P.D. Burrow, J.J. Nash, H. Morrison, K.D. Jordan, J. Am. Chem. Soc. 115
(1993) 9876-9877.
Page 23 of 34

24
/3/
/4/
S. Ptasińska, S. Denifl, S. Gohlke, P. Scheier, E. Illenberger, T.D. Märk, Angew.
Chem. Int. Ed. 45 (2006) 1893-1896.
S. Ptasińska, D. Gschliesser, P. Bartl, I. Janik, P. Scheier, S. Denifl, J. Chem. Phys.
135 (2011) 214309.
/5/
/6/
/7/
/8/
E.D. Miller, J. Michl, Chem. Rev. 89 (1989) 1359-1410.
J. Koe, Polym. Int. 58 (2009) 255-260.
K.E. Ruehl, M.E. Davis, K. Matyjaszewski, Organomet. 11 (1992) 788-794.
H. Gilman, K. Shiina, D. Aoki, B.J. Gaj, D. Wittenberg, T. Brennan, J. Organometal.
Chem. 13 (1968) 323-328.
/9/
M. Oestreich, G. Auer, M. Keller, Eur. J. Org. Chem. 2005, 184-195.
/10/ P. Boudjouk, B.H. Han, Tetrahedron Lett. 22 (1981) 3813-3814.
/11/ P.D. Lickiss, R. Lucas, J. Organomet. Chem. 444 (1993) 25-28.
/12/ S. M. Chen, A. Katti, T. A. Blinka, R. West, Synthesis 1985, 684-686.
/13/ A. Fürstner, H. Weidmann, J. Organomet. Chem. 354 (1988) 15-21.
/14/ B. Lacave-Goffin, L Hevesi, J. Devaux, J. Chem. Soc., Chem. Commun. 1995, 769-
770.
/15/ D. Huber, M. Beikircher, S. Denifl, F. Zappa, S. Matejcik, A. Bacher, V. Grill, T. D.
Märk and P. Scheier, J. Chem. Phys. 125 (2006) 084304.
/16/ E. Illenberger and J. Momigny (Eds.), Gaseous Molecular Ions, An Introduction to
Elementary Processes Induced by Ionization, Steinkopff/Springer, Darmstadt/ New
York, 1992.
/17/ M. Fenzlaff, R. Gerhard, E. Illenberger, J. Chem. Phys. 88 (1988) 149-155.
/18/ Gaussian 98, Revision A.6, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, V. G. Zakrzewski, J. A. Montgomery, Jr., R. E.
Stratmann, J. C. Burant, S. Dapprich, J. M. Millam, A. D. Daniels, K. N. Kudin, M. C.
Strain, O. Farkas, J. Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci, C.
Pomelli, C. Adamo, S. Clifford, J. Ochterski, G. A. Petersson, P. Y. Ayala, Q. Cui, K.
Morokuma, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman, J.
Cioslowski, J. V. Ortiz, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I.
Komaromi, R. Gomperts, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y.
Peng, A. Nanayakkara, C. Gonzalez, M. Challacombe, P. M. W. Gill, B. Johnson, W.
Chen, M. W. Wong, J. L. Andres, C. Gonzalez, M. Head-Gordon, E. S. Replogle, and
J. A. Pople, Gaussian, Inc., Pittsburgh PA, 1998.
/19/ A. Modelli, Phys. Chem. Chem. Phys. 5 (2003) 2923-2930.
/20/ H.-X. Wan, J.H. Moore, J.A. Tossell, J. Chem. Phys. 91 (1989) 7340-7347.
/21/ H.-X. Wan, J.H. Moore, J.A. Tossell, J. Chem. Phys. 94 (1991) 1868-1874.
/22/ V. Huber, K.R. Asmis, A.-C. Sergenton, M.Allan, S. Grimme, J. Phys. Chem. A 102
(1998) 3524-3531.
/23/ M. Hoshino, Š. Matejčík, Y. Nunes, F. Ferreira da Silva, P. Limão-Viera, H. Tanaka,
Int. J. Mass Spectrom. 306 (2011) 51-56.
/24/ M. Allan, J. Electron Spectrosc. Relat. Phenom. 48 (1989) 219-351.
/25/ H.-P. Fenzlaff, E. Illenberger, Int. J. Mass Spectrom. 59 (1984) 185-202.
Page 24 of 34

25
/26/ Y.-R. Luo, Comprehensive Handbook of Chemical Bond Energies, CRC Press (Boca
Raton, FL), 2007.
/27/ D.R. Lide, Handbook of Chemistry and Physics, Taylor & Francis, Boca Raton,
Florida, 2005.
/28/ L. Feketeova, D.J. Skalny, G. Hanel, B. Gstir, M. Francis, T.D. Märk, Int. J. Mass
Spectrom. 223-224 (2003) 661-667.
/29/ J. Fedor, O. May, M. Allan, Phys. Rev. A 78 (2008) 032701.
/30/ N.V. Rees, R. Baron, N.M. Kershaw, T.J. Donohoe, R.G. Compton, J. Am. Chem.
Soc. 130 (2008) 12256-12257.
/31/ N.C.L. Wood, C.A. Paddon, F.L. Bhatti, T.J. Donohoe, R.G. Compton, J. Phys. Org.
Chem. 20 (2007) 732-742.
/32/ M. Yus, R.P. Herrera, A. Guijarro, Tetrahedron Letters 42 (2001) 3455-3458.
/33/ A. Houmam, Chem. Rev. 108 (2008) 2180-2237.
/34/ I. Bald, J. Langer, P. Tegeder and O. Ingólfsson, Int. J. Mass Spectrom. 277 (2008) 4-
25.
/35/ K.D. Jordan, P.D. Burrow, Acc. Chem. Res. 11 (1978) 341-348.
/36/ C.A. Paddon, S.E. Ward Jones, F.L. Bhatti, T.J. Donohoe, R.G. Compton, J. Phys.
Org. Chem. 20 (2007) 677-684.
/37/ W.E. Wentworth, R. George, H. Keith, J. Chem. Phys. 51 (1969) 1791-1801.
/38/ J.-M. Savéant, J. Am. Chem. Soc. 109 (1987) 6788-6795.
/39/ R.A. Marasas, T. Iyoda, J.R. Miller, J. Phys. Chem. A 107 (2003) 2033-2038.
/40/ P. Swiderek, E. Jolondz, J.H. Bredehöft, T. Borrmann, C. Dölle, M. Ott, C. Schmüser,
A. Hartwig, V. Danilov, H.-E. Wagner, J. Meichsner, Macromol.Mat.Eng. 297 (2012)
1091-1101.
/41/ I. Utke, P. Hoffmann, J. Melngailis, J.Vac.Sci.Technol. B 26 (2008) 1197-1276.
/42/ M.A. Huels, L. Parenteau, A.D. Bass, L. Sanche, Int. J. Mass Spectrom. 277 (2008)
256-261.
/43/ A. Lafosse, M. Bertin, A. Domaracka, D. Pliszka, E. Illenberger, R. Azria, Phys.
Chem. Chem. Phys. 8 (2006) 5564-5568.
/44/ S.K. Mandal, H.W. Roesky, Chem. Commun. 46 (2010) 6016-6041.
Page 25 of 34

26
Captions to figures
Fig. 1 Molecular structures of the investigated chlorosilanes.
Fig. 2 Fraction of methyl group 1H NMR intensities in the reaction mixture ascribed to the
parent compounds Me3SiCl (-0.06 ppm), PhMe2SiCl (0.57 ppm), and BzMe2SiCl
(0.27 ppm) as function of reaction time.
Fig. 3 Cl- formation by DEA to Me3SiCl and PhMe2SiCl.
Fig. 4 Si-C dissociation by DEA to Me3SiCl (a) and PhMe2SiCl (b).
Fig. 5 C-H dissociation by DEA to Me3SiCl and PhMe2SiCl.
Fig. 6 Correlation of the vertical attachment energies for the low-lying unoccupied orbitals of
(a) exo-5-chloro-2-norbornene with those of exo-2-chloronorbornane and 2-butene
according to Refs. 2,29 and of (b) PhMe2SiCl with those of Me3SiCl and benzene
derived from present calculations. The experimental value for benzene was taken from
Refs. 24,35.
Fig. 7 Schematic potential energy curves for DEA processes yielding Cl- in from C-Cl and
Si-Cl bond dissociation.
Page 26 of 34

27
Fig. 1
Page 27 of 34

28
1.0
0.8
0.6
0.4
0.2
0.0
Me3SiCl
PhMe2SiCl
BzMe2SiCl
0
5
10
15
20
25
30
35
t (min)
Fig. 2
Page 28 of 34

29
3x104
2x104
m/z 35
Cl-
PhMe2SiCl
1x104
0
3x104 Me3SiCl
2x104
m/z 35
Cl-
1x104
0
0
2
4
6
8
10
12
14
Electron energy (eV)
Fig. 3
Page 29 of 34