Comparison between SiMe2 and CMe2 spacers as σ-bridges for photoinduced charge transfer

Article abstract of DOI:10.1021/ja960546e

The potential of dimethylsilylene and isopropylidene σ-spacers as bridges for photoinduced charge transfer (CT) in 4-cyano-4'-(dimethylamino)- and 4-cyano-4'-methoxy-substituted diphenyldimethylsilanes and 2,2-diphenylpropanes was studied. Fluorescence solvatochromism and time-resolved microwave conductivity measurements show that upon photoexcitation a charge separated state (D^.+σA^.-)* is populated in all compounds. Excited state dipole moments for a given donor-acceptor combination are, irrespective of the bridge, equal. The CT states of the silanes are however lying at lower energies, implying that the presence of silicon thermodynamically facilitates the CT process. Cyclic voltammetry data of model compounds show that this is a consequence of the lowering of the acceptor reduction potential by the silicon bridge. It was however inferred from radiative decay rates that the electronic coupling between the CT and locally excited states as well as the coupling between the ground and CT state is larger for the carbon-bridged compounds. As shown by both solution and solid state electronic spectra and radiative decay rates, the photophysics of the DσA compounds are dominated by intensity borrowing of the CT transitions from transitions localized in the Dσ and σA chromophores.

Full text of DOI:10.1021/ja960546e

J. Am. Chem. Soc. 1996, 118, 8395-8407
8395
Comparison between SiMe2 and CMe2 Spacers as σ-Bridges for
Photoinduced Charge Transfer
Cornelis A. van Walree,† Martin R. Roest,‡ Wouter Schuddeboom,§
Leonardus W. Jenneskens,*,† Jan W. Verhoeven,‡ John M. Warman,§
Huub Kooijman, and Anthony L. Spek
Contribution from the Debye Institute, Department of Physical Organic Chemistry, Utrecht
UniVersity, Padualaan 8, 3584 CH Utrecht, The Netherlands, Laboratory of Organic Chemistry,
UniVersity of Amsterdam, Nieuwe Achtergracht 129, 1018 WS Amsterdam, The Netherlands,
Radiation Chemistry Department, IRI, Delft UniVersity of Technology, Mekelweg 15, 2629 JB
Delft, The Netherlands, and BijVoet Center for Biomolecular Research, Department of Crystal
and Structural Chemistry, Utrecht UniVersity, Padualaan 8, 3584 CH Utrecht, The Netherlands
ReceiVed February 20, 1996X
Abstract: The potential of dimethylsilylene and isopropylidene σ-spacers as bridges for photoinduced charge transfer
(CT) in 4-cyano-4′-(dimethylamino)- and 4-cyano-4′-methoxy-substituted diphenyldimethylsilanes and 2,2-diphen-
ylpropanes was studied. Fluorescence solvatochromism and time-resolved microwave conductivity measurements
show that upon photoexcitation a charge separated state (D•+σA•-)* is populated in all compounds. Excited state
dipole moments for a given donor-acceptor combination are, irrespective of the bridge, equal. The CT states of the
silanes are however lying at lower energies, implying that the presence of silicon thermodynamically facilitates the
CT process. Cyclic voltammetry data of model compounds show that this is a consequence of the lowering of the
acceptor reduction potential by the silicon bridge. It was however inferred from radiative decay rates that the electronic
coupling between the CT and locally excited states as well as the coupling between the ground and CT state is larger
for the carbon-bridged compounds. As shown by both solution and solid state electronic spectra and radiative decay
rates, the photophysics of the DσA compounds are dominated by intensity borrowing of the CT transitions from
transitions localized in the Dσ and σA chromophores.
Introduction
by silicon catenates was proposed.13 These properties find their
origin in the occurrence of σ-conjugation along the silicon
backbone,1,14,15 the presence of polarizable valence electrons,16
low ionization potentials,3,14 and σ-π interactions.17-19 Because
of these features Zyss and co-workers20-25 investigated the
second-order nonlinear optical activity, which is strongly related
to the presence of intramolecular CT interactions, of donor-
acceptor substituted silanes. Although moderate second-order
nonlinear optical responses were found, it remained obscure
whether CT between donors and acceptors takes place in oligo-
and especially monosilylene-bridged compounds. In a prelimi-
nary account we have shown that photoinduced charge separa-
tion occurs in donor-acceptor substituted diphenyldimethyl-
silanes after initial excitation to a locally excited D*σA or DσA*
state (Figure 1).26
Oligosilylenes, linear or cyclic σ-bonded silicon compounds,
have received considerable attention owing to their unusual
electronic and photophysical properties. They exhibit intense
σ-σ* electronic transitions in the near-UV region,1-5 inter-
molecular charge transfer (CT) complexes involving permethy-
lated oligosilanes have been reported,6,7 and in photoexcited
arylsilanes intramolecular CT emission originating from a
1(σ,π*) charge transfer state has been observed.8-12 Further-
more, rapid electron transfer between chromophores connected
† Department of Physical Organic Chemistry, Utrecht University.
‡ Laboratory of Organic Chemistry, University of Amsterdam.
§ Radiation Chemistry Department, IRI, Delft University of Technology.
Department of Crystal and Structural Chemistry, Utrecht University.
X Abstract published in AdVance ACS Abstracts, August 1, 1996.
(1) Miller, R. D.; Michl, J. Chem. ReV. 1989, 89, 1359-1410.
(2) Sakurai, H. J. Organomet. Chem. 1980, 200, 261-286.
(3) Pitt, C. G.; Bursey, M. M.; Rogerson, P. F. J. Am. Chem. Soc. 1970,
92, 519-522.
(4) Brough, L. F.; West, R. J. Am. Chem. Soc. 1981, 103, 3049-3056.
(5) Sun, Y.-P.; Hamada, Y.; Huang, L.-M.; Maxka, J.; Hsiao, J.-S.; West,
R.; Michl, J. J. Am. Chem. Soc. 1992, 114, 6301-6310.
(6) Traven, V. F.; West, R. J. Am. Chem. Soc. 1973, 95, 6824-6826.
(7) Sakurai, H.; Kira, M.; Uchida, T. J. Am. Chem. Soc. 1973, 95, 6826-
6827.
(13) Paddon-Row, M. N.; Verhoeven, J. W. New J. Chem. 1991, 15,
107-116.
(14) West, R. Pure Appl. Chem. 1982, 54, 1041-1050.
(15) Plitt, H. S.; Downing, J. W.; Raymond, M. K.; Balaji, V.; Michl, J.
J. Chem. Soc., Faraday Trans. 1994, 90, 1653-1662.
(16) Bigelow, R. W.; McGrane, K. M. J. Polym. Sci., Part B: Polym.
Phys. 1986, 24, 1233-1245.
(17) Pitt, C. G.; Bock, H. J. Chem. Soc., Chem. Commun. 1972, 28-29.
(18) Sakurai, H. Pure Appl. Chem. 1987, 59, 1637-1646.
(19) Pitt, C. G.; Carey, R. N.; Toren, E. C., Jr. J. Am. Chem. Soc. 1972,
94, 3806-3811.
(8) (a) Shizuka, H.; Obuchi, H.; Ishikawa, M.; Kumada, M. J. Chem.
Soc., Chem. Commun. 1981, 405-406. (b) Shizuka, H. Pure Appl. Chem.
1993, 65, 1635-1640.
(9) Sakurai, H.; Sugiyama, H.; Kira, M. J. Phys. Chem. 1990, 94, 1837-
1842.
(20) Mignani, G.; Kra¨mer, A.; Pucetti, G.; Ledoux, I.; Soula, G.; Zyss,
J.; Meyrueix, R. Organometallics 1990, 9, 2640-2643.
(21) Mignani, G.; Kra¨mer, A.; Pucetti, G.; Ledoux, I.; Soula, G.; Zyss,
J. Mol. Eng. 1991, 1, 11-21.
(10) Kira, M.; Miyazawa, T.; Sugiyama, H.; Yamaguchi, M.; Sakurai,
H. J. Am. Chem. Soc. 1993, 115, 3116-3124.
(22) Mignani, G.; Kra¨mer, A.; Pucetti, G.; Ledoux, I.; Zyss, J.; Soula,
G. Organometallics 1991, 10, 3656-3659.
(11) Horn, K. A.; Grossman, R. B.; Thorne, J. R. G.; Whitenack, A. A.
J. Am. Chem. Soc 1989, 111, 4809-4821.
(23) Mignani, G.; Barzoukas, M.; Zyss, J.; Soula, G.; Balegroune, F.;
Grandjean, D.; Josse, D. Organometallics 1991, 10, 3660-3668.
(24) Ledoux, I. Synth. Met. 1993, 54, 123-137.
(12) Declercq, D.; De Schryver, F. C.; Miller, R. D. Chem. Phys. Lett.
1991, 186, 467-473.
(25) Morley, J. O. J. Chem. Soc., Faraday Trans. 1991, 87, 3015-3019.
S0002-7863(96)00546-X CCC: $12.00 © 1996 American Chemical Society

8396 J. Am. Chem. Soc., Vol. 118, No. 35, 1996
Van Walree et al.
silane (11), obtained by reaction of dichlorodimethylsilane and
4-bromophenyllithium (Scheme 1, reaction 1). Compound 11
was reacted with either 4-(dimethylamino)phenyllithium or
4-methoxyphenyllithium to give 12Si and 13Si, respectively.
In this reaction, the use of phenyllithium instead of Grignard
reagents20 gives significantly improved yields. Finally, bromides
12Si and 13Si were converted into the corresponding cyanides
1Si and 2Si with copper(I) cyanide in DMF (reaction 9).28 The
syntheses of monosubstituted silanes 3Si-8Si and bissubstituted
silanes 9Si and 10Si have been reported previously.29
4-Methoxy-substituted 2,2-diphenylpropanes 4C and 13C
were synthesized by Friedel-Crafts alkylation of anisole with
a 2-phenyl-2-propanol in the presence of either phosphoric acid
or methanesulfonic acid (reaction 3).30,31 Preparation of 4C was
conducted with phosphoric acid in 31% yield; however, use of
methanesulfonic acid32 as applied in the synthesis of 13C gave
improved results (yield 63%).
Figure 1. Schematic representation of (zero order) states in DσA
compounds, showing the ground state DσA, a locally excited state
(DσA)*, and the charge transfer state D•+σA•-. The energy of the latter
state is strongly dependent on the medium polarity. LE denotes a local
excitation and CT a charge transfer transition.
Chart 1
Neither a protic acid nor an aluminium chloride mediated
Friedel-Crafts route was accessible for (dimethylamino)-
diphenylpropanes. Therefore the synthetic pathway started with
nitration of 2,2-diphenylpropane with nitronium tetrafluoro-
borate33 to give 2-(4-nitrophenyl)-2-phenylpropane (14) (reaction
4); the bisnitro-substituted compound 15 was also obtained using
this method. Bromination of 14 afforded 16, while bromination
of 2,2-diphenylpropane gave 17 (reaction 5). Compounds 14,
15, and 16 were reduced to the corresponding amines 18, 19,
and 20, respectively. Reduction of 14 and 15 was executed
with hydrogen with palladium as catalyst. Bromide 16 was
reduced with hydrazine in the presence of a ruthenium catalyst
in order to circumvent dehalogenation of the bromophenyl
ring.34 Subsequently amines 18, 19, and 20 were methylated
to their quaternary ammonium salts, which were subsequently
cleaved by use of ethanolamine (reaction 7).35 2-(4-Bromophen-
yl)-2-phenylpropane (21) was obtained by reaction of the
diazonium salt of amine 20 with hypophosporous acid (H3PO2)
(8).36 Finally, the 4-bromo-substituted diphenylpropanes 12C,
13C, 17, and 21, as well as 4-bromo-tert-butylbenzene, were
converted into the corresponding cyanides 1C, 2C, 10C, 5C,
and 8C, respectively, using copper(I) cyanide in DMF. 4-tert-
Butylanisole (7C) was prepared by methylation of 4-tert-
butylphenol with iodomethane in the presence of potassium
carbonate.
In the present study the potential of the SiMe2 spacer as a
bridge for photoinduced CT is further elaborated and its
properties are compared to those of the CMe2 bridge.27 To this
end we investigated the photophysical and electrochemical
properties of 4-donor-4′-acceptor substituted diphenyldimeth-
ylsilanes and 2,2-diphenylpropanes and related compounds
(Chart 1). It is shown that the charge transporting properties
of the two bridges are strongly related to the interaction of these
bridges with the donor and acceptor moieties, which can be
traced back to the behavior of trimethylsilyl and tert-butyl
groups as substituents. In addition, the electronic coupling
between involved photophysical states is evaluated. Besides
providing insight into the charge transporting capability of
monoatomic bridges the present study also contributes to the
understanding of photoinduced charge separation processes.
Substituent Effects. In Table 1 UV absorption maxima and
intensities of monosubstituted trimethylsilylbenzenes, tert-
butylbenzenes, diphenyldimethylsilanes, and 2,2-diphenylpro-
panes are collected. Inspection of the data reveals that the
spectra of 3-10Si and 3-10C are all related to the spectra of
N,N-dimethylaniline, anisole, or benzonitrile. The effect of
4-positioned tert-butyl (tBu) and trimethylsilyl (TMS) substit-
uents on the energy levels and electronic spectra of N,N-
dimethylaniline and anisole has been discussed previously by
Alt and Bock.37 They pointed out that these substituents do
(28) Friedman, L.; Shechter, H. J. Org. Chem. 1961, 26, 2522-2524.
(29) van Walree, C. A.; Lauteslager, X. Y.; van Wageningen, A. M. A.;
Zwikker, J. W.; Jenneskens, L. W. J. Organomet. Chem. 1995, 496, 117-
125.
(30) Tambotseva, V.; Tsukervanik, I. P. J. Gen. Chem. USSR 1945, 15,
820-824; Chem. Abstr. 1947, 41, 732.
(31) Tsukervanik, I. P. J. Gen. Chem. USSR 1945, 15, 699-703; Chem.
Abstr. 1946, 40, 5707.
(32) Dominianni, S. J.; Ryan, C. W.; DeArmitt, C. W. J. Org. Chem.
1977, 42, 344-346.
(33) Kuhn, S. J.; Olah, G. A. J. Am. Chem. Soc. 1961, 83, 4564-4571.
(34) Mosby, W. L. J. Org. Chem. 1959, 24, 421-423.
(35) Hu¨nig, S.; Baron, W. Chem. Ber. 1957, 90, 395-402.
(36) Vogel’s Textbook of Practical Organic Chemistry; Longman:
London, 1978; p 711.
Results
Synthesis. Donor-acceptor substituted silanes 1Si and 2Si
were prepared starting from (4-bromophenyl)chlorodimethyl-
(26) van Walree, C. A.; Kooijman, H.; Spek, A. L.; Zwikker, J. W.;
Jenneskens, L. W. J. Chem. Soc., Chem. Commun. 1995, 35-36.
(27) Methylene-bridged donor-acceptor systems are already known to
exhibit CT (intramolecular exciplex) emission. See for example: De
Schryver, F. C.; Boens, N.; Put, J. AdV. Photochem. 1977, 10, 359. Okada,
T.; Fujita, T.; Kubota, M.; Masaki, S.; Mataga, N.; Ide, R.; Sakata, Y.;
Misumi, S. Chem. Phys. Lett. 1972, 14, 563-568. Ide, R.; Sakata, Y.;
Misumi, S.; Okada, T.; Mataga, N. J. Chem. Soc., Chem. Commun. 1972,
1009-1011. CT emission of a methylene-bridged donor-donor′-acceptor
compound has been reported as well: Masaki, Y.; Uehara, Y.; Yanagida,
S.; Pac, C. Chem. Lett. 1992, 315-318.
(37) Alt, H.; Bock, H. Tetrahedron 1971, 27, 4965-4975.

Comparison between SiMe2 and CMe2 Spacers as σ-Bridges
J. Am. Chem. Soc., Vol. 118, No. 35, 1996 8397
Scheme 1
Table 1. UV and Fluorescence Data, in Cyclohexane, of Donor or
Acceptor Substituted Diphenyldimethylsilanes, 2,2-Diphenyl-
propanes, Trimethylsilylbenzenes, tert-Butylbenzenes, and Benzenes
N,N-dimethylaniline, whereas the potential of 6Si is somewhat
more positive (Table 2). The latter observation suggests that
the electron accepting effect of the TMS group is somewhat
stronger than the electron donating effect. In contrast, the
oxidation potential of anisole seems to be hardly affected by
the TMS substituent, whereas for 7C the potential is reduced
by some 0.2 V. However, owing to the irreversible nature of
the electrode process electrochemical oxidation potentials of
anisoles do not adequately reflect their ionization potentials.40
The 1Lb bands of anilines and anisoles correspond predomi-
nantly to a HOMO-LUMO transition. Consequently, they are
not shifted significantly by tBu and TMS substitution,37 as the
orbitals involved are either not (LUMO) or only slightly
(HOMO) affected by substitution. The tBu group, being an
inductive donor, raises both the HOMO and SLUMO, in about
1La
1Lb
fluorescence
a
b
a
b
a
c
compd
3Si
νmax
ꢀmax
νmax
ꢀmax
νfl
29.9
29.8
Φfl
37.3
27.7 32.8
19.2 33.0
sh
2.4
0.13
0.24
0.09
0.41
3C
4Si
4C
5Si
5C
6Si
6C
7Si
7C
8Si
8C
9Si
9C
10Si
10C
39.1
43.5
44.2
7.9 36.5/35.6d 0.70 33.2
13.2 36.1/35.2d 2.1
33.3
42.4/41.3d 20.1 36.5/35.5d 1.0
33.7/32.9 0.28
42.4
37.9
39.7
43.5
44.8
15.9 37.3/35.8d 0.95 34.1/33.1 0.26
26.6 33.6
14.5 33.1
sh
2.1
30.2
30.0
33.3
33.6
0.20
0.19
0.19
0.48
12.1 36.5/35.7d 1.4
10.3 36.2/35.5d 2.0
42.9/41.7d 17.9 36.6/35.5d 0.80 33.8/33.0 0.33
43.5/42.0d 15.9 37.6/36.1d 0.50 34.1/33.3 0.18
1
equal amounts.41 The La bands of 6C and 7C, which are
37.0
38.6
44.6 33.6
35.6 33.1
sh
4.8
30.2
29.9
0.22
0.19
predominantly based on promotion of an electron from the
HOMO to the SLUMO, are therefore hardly shifted with respect
to those of N,N-dimethylaniline and anisole. The 1La bands are
however substantially affected by the TMS group. The HOMO
was seen to be only weakly sensitive to this group, but the TMS
substituent substantially lowers the energy of the SLUMO,
presumably as a result of (p-d)π bonding.19,37,42 This results
in a bathochromic shift of ca. 2000 cm-1 for the 1La transitions,
which can be regarded as involving CT from the dimethylamino
42.4/41.3d 54.5 36.4/35.3d 1.8
33.7/32.9 0.43
34.1/33.3 0.22
42.4/41.3d 28.8 37.3/35.8d 1.1
C6H5NMe2 39.8
C6H5OMe 45.5
C6H5CN
14.9 33.6
2.3
30.3
34.0
0.19
0.45
7.8 36.9/36.0d 2.1
45.2/43.5d 12.0 37.0/36.2d 0.62 34.6/33.9 0.23
a Units 103 cm-1 b Units 103 M-1 cm-1 c Fluorescence quantum
. .
yield. d Main vibrational components; the ꢀmax value refers to the most
intensive component given in italics.
1
toward the TMS moiety. The shift of the La band is more
not affect the SHOMO (HOMO-1) and LUMO owing to their
nodal character at the 1 and 4 positions (Figure 2). The HOMO
energies of N,N-dimethylaniline and anisole are slightly raised
by the inductively donating tBu substituent. The TMS group
is inductively donating as well, but this effect is opposed by its
mesomeric electron accepting behavior,19,25,37-39 so that the net
perturbation of the HOMO should be small. These substituent
effects are more or less in line with cyclic voltammetry data,
which reveal that 6C has a comparable oxidation potential as
pronounced for 6Si than for 7Si because the CT character of
the transition is larger for the stronger donating dimethylamino
group.37
The SLUMO and SHOMO of benzonitrile are insensitive to
substitution, again as a consequence of nodes at the 1 and 4
positions. The tBu group raises the energy of both the HOMO
(40) Zweig, A.; Hodgson, W. G.; Jura, W. H. J. Am. Chem. Soc. 1964,
86, 4124-4129.
(41) Jaffe´, H. H.; Orchin, M. Theory and Applications of UltraViolet
(38) Bock, H.; Seidl, H.; Fochler, M. Chem. Ber. 1968, 101, 2815-
Spectroscopy; Wiley: New York, 1962.
2822.
(42) Weeks, G. H.; Adcock, W.; Klingensmith, K. A.; Waluk, J. W.;
West, R.; Vasak, M.; Downing, J.; Michl, J. Pure Appl. Chem. 1986, 58,
39-53.
(39) Bock, H.; Alt, H. J. Am. Chem. Soc. 1970, 92, 1569-1570 and
references cited.

8398 J. Am. Chem. Soc., Vol. 118, No. 35, 1996
Van Walree et al.
Oxidation potentials of donor substituted diphenyldimethyl-
silanes 3Si and 4Si and 2,2-diphenylpropanes 3C and 4C all
are slightly more positive than those of analogously substituted
(trimethylsilyl)- and tert-butylbenzenes. Replacement of a
methyl group of the SiMe3 and CMe3 groups by a phenyl group
apparently leads to a somewhat modified substituent behavior.
The lowering of the cyanophenyl reduction potential by the
dimethylphenylsilyl substituent of 5Si is stronger than that by
the TMS group of 8Si. The reduction potential of 5C, on the
other hand, is higher than that of 8C. Consequently, a difference
of 0.22 eV is found between the reduction potentials of 8Si
and 8C.
Despite the changes in redox potentials the UV spectra of
the monosubstituted diphenyldimethylsilanes and diphenylpro-
panes hardly differ from the corresponding (trimethylsilyl)- and
1
tert-butylbenzenes. Generally, the positions of Lb bands are
not shifted, while shifts of the 1La bands are small, ca. 600 cm-1
.
It can thus be concluded that excitations are localized in the
anilino, anisyl, and cyanophenyl moieties. The bissubstituted
compounds 9Si, 9C, 10Si, and 10C display the same UV
maxima as their analogues of type 3 and 5 as well. Hence, no
substantial electronic interaction between two donors or accep-
tors across silicon and carbon can be inferred. Intensities of
the 1La and 1Lb bands of the bissubstituted compounds are to a
first approximation twice as large as those of the compounds
containing only one dimethylamino or cyano group, reflecting
the presence of two identical chromophores in one molecule.
Fluorescence emission maxima and quantum yields are also
recorded in Table 1. For a given donor hardly any shift in
emission maximum upon substitution is observed whereas
distinct differences were found in the UV spectra. Conse-
quently, the type of donor determines the emission wavelength;
in the lowest singlet excited state the silicon or carbon containing
substituent is not contributing. This compares favorably with
the nature of the lowest lying (1Lb) excited state as the position
of the LUMO is independent of substituents (Figure 2). This
also rationalizes that for donor substituted compounds differ-
ences in fluorescence maxima between compounds containing
SiMe3 and SiMe2C6H5 groups or CMe3 and CMe2C6H5 groups,
respectively, are marginal. For the cyanides a small difference
of some 400 cm-1 between the emission maxima of the carbon
and silicon based substituents is observed; the latter are situated
at lower energies. Moreover, there is a clear bathochromic shift
with respect to benzonitrile. Fluorescence quantum yields of
the donor substituted silicon containing compounds 3Si, 4Si,
6Si, and 7Si are generally lower than or equal to the values of
their analogues with carbon-based substituents. For the cyanides
quantum yields of the silanes are generally higher.
Figure 2. Qualitative MO diagram of donor benzenes (D ) Me2N,
MeO) and benzonitrile and their 4-tert-butyl- and 4-trimethylsilyl
substituted derivatives. Orbital coefficients were obtained from extended
Hu¨ckel calculations (see Experimental Section).
Table 2. Cyclic Voltammetry Data of Donor or Acceptor
Substituted Diphenyldimethylsilanes, 2,2-Diphenylpropanes,
Trimethylsilylbenzenes, tert-Butylbenzenes, and Benzenesa
compd
1Si
1C
2Si
2C
3Si
3C
4Si
4C
Eox
Ered
compd
5Si
5C
6Si
6C
7Si
7C
8Si
8C
Eox
Ered
+0.79
+0.70
+1.75
+1.75
+0.74
+0.69
+1.78
+1.57
+0.64
+1.74
-2.30
-2.51
-2.29
-2.52
-2.28
-2.52
+0.71
+0.63
+1.74
+1.52
-2.34
-2.56
-2.46
Donor-Acceptor Substituted Compounds. The UV spec-
tra (Figure 3) of the donor-acceptor substituted compounds 1Si,
2Si, 1C, and 2C are composed of the transitions appearing in
reference compounds containing only a donor or acceptor
chromophore (e.g. 3Si and 5Si for 1Si). It is however seen in
C6H5NMe2
C6H5OMe
C6H5CN
a Halve-wave oxidation (Eox) and reduction potentials (Ered) are given
in V relative to SCE.
Figure 3 that enhanced absorption occurs below 35 000 cm-1 44
,
and LUMO, while the TMS group hardly affects the HOMO
energy and stabilizes the LUMO. The cyclic voltammetry data
show that the reduction potential of 8Si is 0.12 V lower than
that of benzonitrile and 0.22 V lower than that of 8C. The 1La
band, mainly associated with the HOMO-LUMO transition,43
is shifted by introduction of both the tBu and TMS groups. In
the case of 8C this shift is caused by an increase of the HOMO
energy, while in the case of 8Si the lowering of the LUMO is
responsible. The shift of the benzonitrile 1Lb band upon
substitution is small.
which arises from a transition involving the anilino or anisyl
donor and cyanophenyl acceptor orbitals and can be designated
as a CT absorption. No concentration dependence on the UV
spectra was observed, implying that the CT is intramolecular
in nature. The presence of discrete intramolecular CT bands
(44) Difference spectroscopy revealed that the CT bands of both 1Si
and 1C are situated at 34 800 cm-1, while those of 2Si and 2C are difficult
to determine reliably. According to the relation νCT ) Eox(D) - Ered(A) +
C (compare with eq 3) the CT bands of the silanes should be red shifted to
the CT bands of the diphenylpropanes. This discrepancy is in line with an
earlier report (ref 47) on the unreliability of determining CT maxima by
difference spectroscopy.
(43) Del Bene, J.; Jaffe´, H. H. J. Chem. Phys. 1968, 49, 1221-1229.

Comparison between SiMe2 and CMe2 Spacers as σ-Bridges
J. Am. Chem. Soc., Vol. 118, No. 35, 1996 8399
1
maxima. The Lb bands are not involved in the CT process
because of the presence of nodes at the substitution positions
of the orbitals involved in these transitions.
The oxidation and reduction potentials of the donor-acceptor
substituted compounds (Table 2) are generally within experi-
mental error equal to those of the separate chromophores. The
only redox potential which deviates from that of its separate
chromophores is the oxidation potential of 2C. As already
discussed above the oxidation potentials of the methoxy
compounds are not too reliable. Ground state interactions can
thus not be inferred from the cyclic voltammetry data.48
Fluorescence spectra of 1C, 2C, 1Si, and 2Si are highly
dependent on the solvent polarity (Table 4). The bathochromic
shift with increasing solvent polarity is typical of CT fluores-
cence originating from a charge separated state D•+σA•-. The
spectra of all compounds consist of a single emission band in
all solvents, excepted 1Si in acetonitrile where the CT emission
is so weak that another band centered at 25 640 cm-1 becomes
visible. This is probably a local emission. Time resolved
fluorescence measurements showed that the two emission bands
behave independently, indicating that they are related to separate
photophysical phenomena.
A quantitative relationship between the CT fluorescence
maxima and the solvent polarity is provided by the Lippert-
Mataga equation:49
2µ2
hcF3
νCT ) νCT(0) -
e ∆f
(1a)
with
Figure 3. UV spectra in cyclohexane of donor-acceptor compounds
1Si (top) and 1C (bottom) and their separate chromophores.
n2 - 1
4n2 + 2
ꢀs - 1
2ꢀs + 1
Table 3. UV Transitions of Donor-Acceptor Substituted
∆f )
-
(1b)
Diphenyldimethylsilanes and 2,2-Diphenylpropanes in Cyclohexane
1La
1Lb
where νCT denotes the emission wavenumber, νCT(0) the
(hypothetical) gas-phase emission wavenumber, µe the excited
state dipole moment (the ground state dipole moment is
neglected in eq 1a), h the Planck constant, c the light velocity,
and F the Onsager radius of the solute. The solvent polarity
parameter ∆f is related to the solvent dielectric constant ꢀs and
refractive index n (eq 1b).50 Plotting of νCT Vs ∆f enables the
calculation of the excited state dipole moment µe from the slope
of the linear fit, provided that the Onsager radius F of the solute
is known.
The solvatochromic relationships are shown in Figure 4, and
parameters of these relationships are given in Table 4. For all
compounds a satisfactory linear relationship is found. However,
the solvatochromic behavior of 2C is more appropriately
described by two relationships, one for solvents of low polarity
and one for solvents of high polarity, suggesting that as a
function of solvent polarity different emitting species are present.
This can be rationalized by taking the driving force ∆G0 for
photoinduced CT processes into account, which is given
by:51
a
b
a
b
compd
νmax
ꢀmax
νmax
ꢀmax
1Si
1C
2Si
2C
37.6c/42.7d
38.9c/42.9d
42.9f
25.7c/19.0d
18.4c/20.3d
18.8
e
e
e
e
36.5/35.6f
2.0
2.4
44.2f
23.2
35.1f
a Units 103 cm-1 b Units 103 M-1 cm-1 c Donor band. d Acceptor
. .
band. e Submerged under other bands. f Superposition of donor and
acceptor bands.
indicates that a ground state donor-acceptor interaction exists
and that a charge separated state D•+σA•- can be populated
directly from the ground state.
1
As the intensity of the La transitions in the spectra of the
donor-acceptor compounds is lower than the summation of the
1La transitions of the model compounds, it appears that the CT
1
transitions borrow intensity from the donor and acceptor La
transitions. I.e., mixing of the CT state with locally excited
1La states (D*σA and DσA*) occurs.45-47 Furthermore, the
1
small but distinct hypsochromic shift of the La bands of the
donor-acceptor substituted compounds (Table 3) with respect
to the La maxima of the reference compounds suggests that
the local transitions mix with the CT transition Via configuration
interaction.46 A hypsochromic shift for the 1La bands of
methoxy compounds 2Si and 2C is not revealed by the data in
Table 3 because overlapping of bands obscures the actual
1
∆G0 ) Eox(D) - Ered(A) - E00 + C
(2a)
with
(48) Fourmigue´, M.; Huang, Y.-S. Organometallics 1993, 12, 797-802.
(49) Beens, H.; Knibbe, H.; Weller, A. J. Chem. Phys. 1967, 47, 1183-
1184.
(45) Murrell, J. N. The Theory of the Electronic Spectra of Organic
Molecules; Methuen: London, 1963; p 271.
(46) Pasman, P.; Rob, F.; Verhoeven, J. W. J. Am. Chem. Soc. 1982,
104, 5127-5133.
(47) Verhoeven, J. W.; Dirkx, I. P.; de Boer, T. J. Tetrahedron 1969,
25, 4037-4055.
(50) Solvent dielectric constants and refractive indices were taken from
the following: Reichardt, C. SolVents and SolVent Effects in Organic
Chemistry, 2nd ed.; VCH: Weinheim, 1988.
(51) Weller, A. Z. Phys. Chem. 1982, 133, 93-98.

8400 J. Am. Chem. Soc., Vol. 118, No. 35, 1996
Van Walree et al.
Table 4. Fluorescence Emission Maxima (νCT) and Quantum Yields (Φfl) of Dimethylsilylene- and Isopropylidene-Bridged Donor-Acceptor
Compounds in Various Solvents and Results of Solvatochromic Fits
1Si
2Si
1C
2C
a
a
a
a
solvent
∆f
νct
Φfl
νct
Φfl
νct
Φfl
νct
Φfl
cyclohexane
benzene
pentyl ether
butyl ether
diethyl ether
ethyl acetate
THF
0.100
0.115
0.173
0.193
0.251
0.292
0.308
0.376
0.392
28.17
23.31b,c
23.39
23.98
22.60
19.81
19.77
17.67
16.67e
0.04
0.12
d
33.67b
0.11
29.76
25.00b
25.64
26.00
24.39
21.60
21.65
19.42
18.28
0.22
0.15
d
33.35
0.48
30.12
29.76
28.17
26.30
26.21
24.36
23.27
d
d
0.04
0.05
0.04
d
32.05
32.26
31.09
27.93
27.78
26.39
25.00
d
d
0.58
0.25
0.26
d
d
d
0.10
0.08
0.07
d
0.24
0.37
0.44
d
butyronitrile
acetonitrile
0.0015
0.13
0.03
0.31
36.51
31.03
0.983
30.78
35.61
0.995
37.18
33.03
0.989
26.50f
14.51g
-2µ2e/hcF3 a
a
νct(0)
35.84f
34.79g
corr coeff
0.954f
0.975g
a Units 103 cm-1 b Not used in solvatochromic fit. c Excitation wavenumber 35 090 cm-1 d Not determined because of solvent absorption at
. .
37 740 cm-1. Excitation wavenumber 33 330 cm-1 for 1Si and 1C and 35 090 cm-1 for 2Si and 2C, respectively. e Another band observed at
25 640 cm-1 f Polar solvents. g Nonpolar solvents.
.
Table 5. Thermodynamic Data for Photoinduced Charge Transfer
in Donor-Acceptor Compounds
0
∆G0∞
(eV)
Κ1
Κ2
∆Gchx
a
b
b
b
c
E00
b
compd
(eV)
(eV)
(eV)
ꢀs(0)
(eV)
1Si
1C
2Si
2C
3.71
3.76
4.19
4.25
-0.69
-0.55
-0.13
+0.02
0.09
0.09
0.09
0.09
0.53
0.31
0.53
0.31
0.7
0.5
2.4
4.4
-0.52
-0.49
+0.04
+0.08
a Zero-zero transition energy, taken as the onset of the UV spectrum.
b See text for explanation. c Driving force in cyclohexane (ꢀs ) 2.015).
+
diphenylpropanes (RDA ) 4.9 Å). The effective radii rD and
rA- were estimated from the molecular volumes Vm of 1Si (392
Å3) and 1C (364 Å3) in their crystal structures by rD ) rA- )
+
(3Vm/4π)1/3, giving 4.5 Å for 1Si and 2Si and 4.4 Å for 1C and
2C. K1 and K2 values obtained with these parameters are
collected in Table 5.
Equation 2c implies that upon going to a less polar solvent
∆G0 becomes more positive; at a certain medium polarity CT
is thermodynamically not allowed anymore. The dielectric
constants at which ∆G0 is zero, ꢀs(0), and the driving force in
cyclohexane (ꢀs ) 2.015) are given in Table 5. These data show
that in the dimethylamino compounds 1Si and 1C photoinduced
CT is feasible even in cyclohexane, the most nonpolar solvent.
In contrast, CT is forbidden for 2C in solvents with ꢀs < 4.4.
This result corresponds nicely to the data shown in Figure 4,
which reveal that in diethyl ether (ꢀs ) 4.2) and solvents of
lower polarity another emitting species is present than in high-
polarity solvents. Since the emission in the low-polarity solvents
is still moderately solvatochromic, it is anticipated that a partially
charge separated state Dδ•+σAδ•- is present. The ꢀs(0) value
for 2Si, 2.4, indicates that this is also the case for this compound
in cyclohexane. Indeed the emission maximum for 2Si in
cyclohexane deviates from the solvatochromic fit and coincides
with the emission maxima of its separate chromophores 4Si and
5Si.26
Excited state dipole moments µe were obtained from the
slopes of the solvatochromic fits (eq 1a). Radii F were estimated
by F ) (3Vm/4π)1/3, giving 4.5 Å for the diphenyldimethylsilanes
and 4.4 Å for the diphenylpropanes (Vide supra). With these
values excited state dipole moments of 18.2 (1Si), 17.7 (1C),
16.7 (2Si), 15.0 (2C, polar solvents), and 11.1 D (2C, nonpolar
solvents), respectively, were obtained. The µe obtained for 2C
in polar solvents seems to be rather small in comparison with
the excited state dipole moment of the other compounds, but
inspection of Figure 4 reveals that this is caused by a somewhat
Figure 4. Fluorescence solvatochromism of donor-acceptor substituted
compounds 1Si (open circles), 1C (filled circles), 2Si (open squares),
and 2C (filled squares).
e2
ꢀsRDA
e2
1
1
1
1
C ) -
-
+
-
(2b)
( )(
)(
)
2
rD
+
rA 35.9 ꢀs
-
In these equations E00 represents the zero-zero excitation
energy, e the elementary charge, RDA the donor-acceptor center-
to-center distance, and rD and rA- the effective radii of the
+
radical cations and anions, respectively. Eox(D) is the oxidation
potential of the donor and Ered(A) the reduction potential of the
acceptor as measured in acetonitrile (ꢀs ) 35.9); the term C
corrects for the solvent effect by taking into account the
Coulombic attraction energy between radical cations and anions
and a solvation free energy term. Equations 2a and 2b can be
conveniently combined to
∆G0 ) ∆G∞0 - K1 + K2/ꢀs
(2c)
in which ∆G0 ) Eox(D) - Ered(A) - E00 (ꢀs being infinitely
52
∞
large) and K1 and K2 constants are determined by the type of
bridge Via Rda, rD , and rA-. RDA was taken as the distance
between the centers of the aromatic rings in the crystal structure
+
of 1Si for the silanes (RDA ) 5.4 Å) and 1C for the
(52) The redox potentials of the reference chromophores (3Si, 5Si, and
so on) were used to obtain ∆G0∞. However, for the uncertain oxidation
potential of 2C the value of the compound itself, +1.75 V, was used.

Comparison between SiMe2 and CMe2 Spacers as σ-Bridges
J. Am. Chem. Soc., Vol. 118, No. 35, 1996 8401
deviating emission maximum in butyronitrile. With the excep-
tion of the value for 2C in nonpolar solvents, the dipole
moments correspond to a unit charge separation over 3.1-3.8
Å. Since the distance between the centers of the anilino or
anisyl and cyanophenyl moieties is estimated from X-ray
structures to be 5.4 Å for the silanes and 4.9 Å for the carbon
compounds, 63-70% of an elementary charge is transferred
over the donor-acceptor distance. Although this suggests at
first sight that full CT does not take place, obtained µe values
are very sensitive to the magnitude of F. In view of the
uncertainty in both F and the donor-acceptor distance it is
therefore difficult to decide whether full or partial CT occurs.
Excited state dipole moments are thus virtually independent
of the bridge and are determined by the donor-acceptor
combination only. For a given donor-acceptor combination
the almost parallel solvatochromic plots show that the energies
of the CT states of the silicon-bridged compounds are a
systematic amount lower than those of their carbon-bridged
counterparts. This is understood by considering the solvent
dependence of the CT state energies ECT, given by53
ꢀs - 1
3 2ꢀs + 1
2
ECT ) Eox(D) - Ered(A) + C′ - µe2
(3)
F
Taking µe, F, and the Coulomb term C′ equal for a given donor-
acceptor pair, differences in ECT appear to be determined only
by differences in oxidation and reduction potentials. This
applies nicely to the dimethylsilylene- and isopropylidene-
bridged donor-acceptor compounds. The difference in
Eox(D) - Ered(A) for the pair 1Si and 1C is 0.19 eV; for 2Si
and 2C the value is 0.21 eV (Table 2). The emission
wavenumber of 1Si is on average 0.23 eV (1820 cm-1) lower
than that of 1C, while the emission energy difference between
2Si and 2C is on average 0.22 eV (1740 cm-1) in the polar
solvents. Consequently, a dimethylsilylene bridge provides an
energetically more favorable pathway for CT to occur than a
isopropylidene bridge by relatively lowering the CT state by
some 0.2 eV. Full CT probably occurs for instance in 2Si in
etheral solvents, whereas only partial CT is observed in its
analogue 2C in these solvents. As was seen above the 0.2-eV
difference is predominantly caused by a lowering of the
cyanophenyl acceptor reduction potential by the dimethylsilylene
bridge.
Time-Resolved Microwave Conductivity Studies. The
nature of the excited state of 1Si and 1C was further elucidated
using the flash-photolysis time-resolved microwave conductivity
(TRMC) technique.54-56 The formation of dipolar excited states
upon pulsed laser photoexcitation (flash-photolysis) leads to an
increase in the high-frequency dielectric loss of a solution. Using
the TRMC technique this increase is monitored by time-resolved
measurement of the change in microwave conductivity ∆σ
which is related to the difference in dipole moment of the excited
state µe and ground state µg according to:
Figure 5. Time-resolved microwave conductivity transients of 1Si (top)
and 1C (bottom) in benzene. Apart from the measured signals the singlet
and triplet components of the fits to the experimental data are shown.
i.e. the mean time required for randomization of the dipole
orientation, depends on the solute geometry and the solvent
viscosity.
TRMC transients of 1Si and 1C in benzene are shown in
Figure 5. In both cases, an initial short-lived component with
a lifetime of 18 ns for 1Si and 7 ns for 1C, followed by a long-
lived (hundreds of nanoseconds) component, is discernible.
These components are attributed to the singlet and triplet excited
states, respectively. Both excited states possess a dipole moment
substantially larger than that of the ground state, confirming
the formation of charge separated states. It is conspicuous that
the triplet state contribution to the TRMC transient is larger
for 1C than for 1Si. On the basis of the heavy atom effect57
the opposite behavior would have been expected. The larger
triplet component found for 1C may be related to the occurrence
of exciton coupling between the two phenyl moieties, which is
known to be present in diphenylmethanes and related com-
pounds and favors intersystem crossing.58 Owing to their closer
proximity, the degree of exciton coupling between the phenyl
moieties will be larger for 1C.
Excited state dipole moments can be derived from the
absolute magnitude of the TRMC transients if the dipole
relaxation time is known. For the present molecules Θ may
be taken to be controlled by rotational diffusion of the molecules
and therefore can be equated with the rotational diffusion time,
ΘR. A method for calculating ΘR for molecules with spherical,
cylinder-like and disk-like geometries has been described.56
Since the geometries of 1Si and 1C are intermediate between
that of a sphere and that of a cylinder, we have calculated ΘR
for these two extremes. This results in 70 ps < ΘR < 189 ps
for 1Si and 66 ps < ΘR < 165 ps for 1C. Using these limiting
∆σ ) (ꢀs + 2)2N(µe2 - µg2)FωΘ/27kBTΘ
(4)
with N being the solute concentration, kb the Boltzmann
constant, T the temperature, and FωΘ a factor containing the
applied microwave frequency ω and the dipole relaxation time
Θ: FωΘ ) (ωΘ)2/[1 + (ωΘ)2]. The dipole relaxation time Θ,
(53) Rettig, W. Angew. Chem. 1986, 98, 969-986.
(54) de Haas, M. P.; Warman, J. M. Chem. Phys. 1982, 73, 35-53.
(55) Infelta, P. P.; de Haas, M. P.; Warman, J. M. Radiat. Phys. Chem.
1980, 15, 273.
(57) Turro, N. J. Modern Molecular Photochemistry; Benjamin/Cum-
mings: Menlo Park, 1978; p 125.
(56) Schuddeboom, W. Ph.D. Thesis, Delft University of Technology,
1994.
(58) Kasha, M.; Rawls, H. R.; El-Bayoumi, M. A. Pure Appl. Chem.
1965, 11, 371-392.

8402 J. Am. Chem. Soc., Vol. 118, No. 35, 1996
Van Walree et al.
Table 6. Solvent Dependence of Fluorescence Lifetimes and
Radiative Decay Rates
1Si
1C
a
b
a
b
solvent
τfl
1.7
18c
krad
g*
τfl
krad
g*
cyclohexane
benzene
diethyl ether
ethyl acetate
acetonitrile
23.5
6.7
5.1
2.7
0.83
0.058
0.025
0.022
0.016
0.011
2.8
7c
8.4
34.0
5.0
78.6
21.4
28.5
10.9
6.0
0.19
0.084
0.077
0.055
0.038
19.6
29.6
1.8
a Units ns. b Units 106 s-1
ment.
.
c Lifetime obtained with TRMC experi-
values of Θ together with PM3 calculated ground state dipole
moments of 5.0 (1Si) and 4.1 D (1C) results in estimates of the
excited state dipole moments within the range 16 D < µe < 25
D and 15 D < µe < 23 D for 1Si and 1C, respectively. The
actual dipole moments would be expected to lie close to the
means of these extremes, i.e. 20.5 and 19 D. The values of
18.2 and 17.7 D determined from solvatochromic shifts are seen
to be in reasonably good agreement with these estimates from
the TRMC results.
Figure 6. Linear fits of krad/n3 Vs νC3 T/(ELE - νCT
)
for 1Si (open
2
circles) and 1C (filled circles). Fitting results are y ) (2.11 × 105) +
29.51x, r ) 0.998 for 1Si and y ) (2.08 × 106) + 83.3x, r ) 0.973 for
1C.
64π4n3
krad
)
[(V∆µ)2νCT + 2VV*(∆µ‚µLE)ν2CT/(ELE
-
3h
νCT) + (V*µLE)2νC3 T/(ELE - νCT)2] (5)
Radiative Decay Rates and Interstate Electronic Cou-
plings. Fluorescence quantum yields of the donor-acceptor
compounds display a general trend of maximizing at medium
solvent polarity (Table 4). Often a monotonous decrease with
increasing medium polarity is found for donor-acceptor
systems. The trend in quantum yields found here has been
explained previously by the occurrence of intersystem crossing
to locally excited triplet states in nonpolar solvents.59 Fluo-
rescence lifetimes τfl of 1Si and 1C, presented in Table 6, follow
a similar trend as the quantum yields: they reach a maximum
at medium solvent polarity (ethyl acetate). It is conspicuous
that lifetimes are quite long for compounds with such small
donor-acceptor separations.
In eq 5 ∆µ ) µe - µg and the couplings between states are
defined by V ) D•+σA•-|Hˆ |DσA and V* ) D•+σA•-|Hˆ |D*σA .
When the mixing of the CT state with the LE state dominates
the decay process only the last term of eq 5 has to be taken
into account, leading to
64π4n3
3h
krad
)
(V*µLE)2νC3 T/(ELE - νCT)2
(6)
This relationship predicts a linear relationship between krad/n3
and ν3CT/(ELE - νCT)2. Taking ELE ) 37 300 cm-1 for 1Si and
1
ELE ) 39 100 cm-1 for 1C (the maxima of the La bands, the
Most remarkably the fluorescence lifetime of 1C is longer
in polar media, while in less polar solvents such as diethyl ether
and benzene it is shorter. In cyclohexane the lifetime of 1C is
again longer. Intuitively it is expected that, since the CT state
of 1Si is systematically situated at lower energies, charge
recombination in 1Si is faster and hence its fluorescence
lifetimes will be shorter. The situation becomes more clear
when the radiative decay rates, defined by krad ) Φfl/τfl, are
taken into consideration (Table 6); in a given solvent the decay
of the CT state of 1C is faster than that of 1Si. The still
remaining marked solvent dependence of the radiative rates of
both compounds strongly suggests that locally excited (LE)
states are involved in the decay process, Via intensity
borrowing.60-62 The involvement of LE states can qualitatively
be understood by the fact that owing to the small donor-
acceptor separation and hence small excited state dipole
moments the solvent stabilization on the CT states of 1Si and
1C is relatively small (eq 3). The CT states are thus lying close
to the LE states, promoting their mutual electronic coupling.
When LE states are involved in the radiative decay process
the decay rate depends, in addition to the electronic coupling
between the ground and CT state V, on the coupling between
nearest lying bands for which mixing with the CT state is
symmetry allowed) indeed linear relations are found between
these quantities (Figure 6). The linear fit for 1Si is excellent,
whereas that for 1C is somewhat less satisfactory. Nevertheless,
the coupling between the CT and LE states in both 1Si and 1C
is convincingly demonstrated. The coupling parameter V* can
be obtained from the slope of the fits, provided that µLE is
known. This transition dipole moment is obtained using
3he2fLE
8π2meωLE
2
|µLE| )
(7)
with ωLE the absorption frequency (s-1) and 3he2/8π2me ) 7.095
× 10-43 m2 s-1 C2. The oscillator strength of the local excitation
fle is given by
fLE ) (4.32 × 10-9)ꢀmax∆ν1/2
(8)
where ∆ν1/2 is the band width at half height in cm-1. The values
of interest for 3Si are ∆ν1/2 ) 2890 cm-1, fLE ) 0.35, and µLE
) 4.4 D. For 3C they are ∆ν1/2 ) 3540 cm-1, fLE ) 0.29, and
µLE ) 4.0 D. It follows that V* ) 2190 cm-1 for 1Si and V*
) 4080 cm-1 for 1C. The CT state of 1C is thus stronger
coupled to the LE state than is the case for 1Si.
the LE and CT state V* and the transition dipole moment µle of
60
the local excitation of energy ELE
:
(59) Hermant, R. M.; Bakker, N. A. C.; Scherer, T.; Krijnen, B.;
Verhoeven, J. W. J. Am. Chem. Soc. 1990, 112, 1214-1221.
(60) Bixon, M.; Jortner, J.; Verhoeven, J. W. J. Am. Chem. Soc. 1994,
116, 7349-7355.
According to eq 6 the fits in Figure 6 should have an intercept
of zero. This applies for 1Si but not for 1C. Consequently,
lower order terms in eq 5 including V should be taken into
consideration to describe the behavior of 1C, whereas they can
be neglected for 1Si. It can thus be concluded that the electronic
coupling between the CT and ground state V is larger for the
carbon compound as well.63
(61) Gould, I. R.; Young, R. H.; Mueller, L. J.; Albrecht, A. C.; Farid,
S. J. Am. Chem. Soc. 1994, 116, 3147-3148.
(62) Verhoeven, J. W.; Scherer, T.; Wegewijs, B.; Hermant, R. M.;
Jortner, J.; Bixon, M.; Depaemelaere, S.; De Schryver, F. C. Recl. TraV.
Chim. Pays-Bas 1995, 114, 443-448.

Comparison between SiMe2 and CMe2 Spacers as σ-Bridges
J. Am. Chem. Soc., Vol. 118, No. 35, 1996 8403
The foregoing analysis is valid on the conditions that the
degree of admixture of the LE state into the CT state g*, given
by
g* ) [V*/(ELE - νCT)]2
(9)
is small: g* < 0.1 and V* , ELE - νCT. These conditions are
fulfilled quite well with the exception of 1C in cyclohexane
(g* ) 0.19, Table 6). It must however be realized that the
obtained V* values are strongly associated with the magnitude
of the local excitation transition dipole moment µLE as the
product V*µLE is obtained from the slopes of the fits. It has
been shown62 that application of µLE of only the first local
transition often leads to an underestimated value. The effective
magnitude of µLE, which incorporates contributions of several
local excitations, may be a factor of 3 to 4 larger. The
consequence is that V* (and hence g*) is the same factor smaller,
and in retrospect the applied analysis is justified.
The g* data show that the admixture of the LE into the CT
state decreases with increasing solvent polarity. The conse-
quence is that CT state lifetimes are short in nonpolar solvents
because decay Via the short lived LE state is prominent. The
shorter lifetimes of 1C in nonpolar media are related to the larger
extent of mixing of the CT and LE states in this compound. In
polar solvents the lifetimes are more an intrinsic property of
the CT states themselves. In these solvents the lifetimes of 1Si
are shorter since its CT state is situated at lower energies.
Furthermore the small g* value in all solvents indicates that
CT state dipole moments are hardly affected by the admixture
of LE states, and can be considered to be virtually independent
of the solvent polarity.
Figure 7. ORTEP65 (30% probability level) drawing of 1Si (top) and
1C (bottom).
Solid State Structures and Properties of 1Si and 1C. The
molecular structures of 1Si26 and 1C have been determined by
single-crystal X-ray diffraction (Figure 7, Tables 7-9). Both
compounds crystallize in the same space group (Pnma), are
located on a crystallographic mirror plane, and possess similar
structures. In both compounds the geometry around the bridging
atom is essentially tetrahedral, with bond angles ranging from
107.00(6)° to 112.11(8)° in 1Si and from 107.7(2)° to 111.2(2)
in 1C (Table 9). The geometry around the dimethylamino
nitrogen atoms is nearly planar, indicating that these nitrogen
atoms are to a large extent sp2 hybridized. In 1C the sum of
bond angles around N(2) is 356.3(4)°, and the C(13)-N(2)-
C(12)-C(11) dihedral angle is 12.7(5)°. In the silane the
corresponding parameters are 357.09(18)° and 10.8(2)°. The
sp2 hybridization of nitrogen atoms, allowing maximal overlap
between their lone pairs and the benzene π-systems, is usual
for N,N-dimethylanilines.64
Table 7. Crystallographic Data for 1Si and 1C
1Si
C17H20N2Si
280.44
orthorhombic
Pnma (No. 62)
11.7053(5)
9.7903(10)
13.6840(14)
1568.2(2)
1.188
1C
formula
mol wt
cryst system
space group
a, Å
C18H20N2
264.37
orthorhombic
Pnma (No. 62)
11.4551(12)
9.6451(12)
13.1648(13)
1454.5(3)
1.207
b, Å
c, Å
V, Å3
D
Z
calc, g‚cm-3
4
4
F(000)
600
568
0.7
µ[Mo KR], cm-1 1.4
cryst size, mm
T, K
0.5 × 0.4 × 0.4
0.4 × 0.2 × 0.2
150
150
final Ra
0.042 [1582Fo > 4σ(Fo)] 0.066 [799Fo > 4σ(Fo)]
0.098
1.06
final wR2b
goodness of fit
0.138
0.99
Interestingly, an increased quinoid character of the ground
state of the (dimethylamino)phenyl group in 1Si can be
distinguished. This quinoid contribution can be expressed by
a parameter Q,66 which is in this case
a R
∑[w(F2o)2]]1/2
)
∑||Fo|
-
|Fc||/∑|Fo|. b wR2
) -
[∑[w(F2o F2c)2]/
.
quinoid structure with a single bond length of 1.455 Å and a
double bond length of 1.317 Å.66 For the (dimethylamino)-
phenyl ring in 1Si Q is 0.018(2) Å, while it is 0.000(5) Å in
1C. The latter result is somewhat surprising, since a small
quinoid contribution is expected to be present in dimethyla-
nilines. We attribute the apparent absence of quinoid contribu-
tions in 1C to the unexpected short C(9)-C(10) distance of
1.383(3) Å. Even though the increased quinoid character in
1Si is indicative of a (ground state) CT interaction between the
dimethylamino group and the silicon bridge, reflecting the
latter’s electron accepting properties. The presence of the CT
interaction in 1Si is also manifested by the shorter N(2)-C(12)
distance (1.377(2) Å in 1Si, 1.393(5) Å in 1C) and the larger
difference between the M-C(9) bond length (1.8576(19) Å in
Q ) 1/2(d9-10 + d11-12) - d10-11
(10)
where d denotes the distance between the indicated carbon
atoms. Q is zero for D6h benzene, and 0.138 for a perfect
(63) Very recently fitting procedures for a quantitative evaluation of V
have been developed (ref 62). Unfortunately they cannot be reliably applied
here because of the limited number of available data points. Application of
the procedures, however, seems to establish that the coupling between
ground and CT state is larger for 1C.
(64) Maier, J. P.; Turner, D. W. J. Chem. Soc., Faraday Trans. 2 1973,
69, 521-531.
(65) Spek, A. L. Acta Crystallogr. 1990, A46, C34.
(66) Bertolasi, V.; Feretti, V.; Gilli, P.; de Benedetti, P. G. J. Chem.
Soc., Perkin Trans. 2 1993, 213-219.

8404 J. Am. Chem. Soc., Vol. 118, No. 35, 1996
Van Walree et al.
Table 8. Bond Lengths (Å) in 1Si and 1C with esd’s in
Parentheses
bond
1Si
1C
bond
1Si
1C
N(1)-C(7) 1.122(3)
N(2)-C(12) 1.377(2)
N(2)-C(13) 1.448(2)
C(1)-C(2) 1.402(3)
C(1)-C(6) 1.398(3)
1.140(6) C(4)-C(5)
1.393(5) C(4)-C(7)
1.448(4) C(5)-C(6)
1.391(6) C(8)-M
1.396(3)
1.462(3)
1.384(3)
1.400(6)
1.453(6)
1.380(7)
1.8611(19) 1.541(4)
1.407(6) C(9)-C(10) 1.3973(16) 1.383(3)
C(1)-M a
1.8872(18) 1.542(5) C(9)-M 1.8576(19) 1.531(6)
C(2)-C(3) 1.379(3)
C(3)-C(4) 1.393(3)
1.387(6) C(10)-C(11) 1.3835(19) 1.390(4)
1.381(6) C(11)-C(12) 1.4051(16) 1.397(3)
a M: Si(1) in 1Si, C(14) in 1C.
1Si, 1.531(6) Å in 1C) and the M-C(1) bond length (1.8872-
(18) Å in 1Si, 1.542(5) Å in 1C). Thus, the difference between
these bond lengths is 0.030(3) Å for 1Si and 0.011(8) for 1C.
Quinoid contributions in the cyanophenyl moieties of 1Si (Q
) 0.016(7) Å) and 1C (Q ) 0.011(14) Å) are of similar
magnitude.
The most remarkable feature in the X-ray structures is the
orthogonality of the donor and acceptor chromophores, with
the cyanophenyl moieties lying in the plane bisecting the
(dimethylamino)phenyl groups. Concerning 1Si this geometry
differs markedly from the X-ray structure of dimethylbis-
(tetrathiafulvanelyl)silane, which possesses a butterfly confor-
mation (see Figure 9) with an angle of 88.92° between the
normals to the tetrathiafulvanelyl planes.48 Other 2,2-di-
phenylpropanes in which the phenyl groups are arranged
orthogonally are known, but scarce.67 It would be expected
that attractive forces between the donor and acceptor moieties
force 1Si and 1C in particular into a spatial arrangement in
which a through-space donor-acceptor interaction is possible.
Apparently these forces are weak. It is not so that the orthogonal
geometry around the bridging atoms prevents steric repulsions
between the methyl carbon atoms and phenyl protons positioned
ortho to the bridge. A close contact of 2.57(3) Å between C(8)a
and H(10) is found in the structure of 1C. This distance, which
is 0.33 Å less than the sum of contact radii of carbon and
hydrogen (2.90 Å), has also been reported to be present in
another conformation of the 2,2-diphenylpropane skeleton.68 The
steric crowding leads to a nonequivalence of the C(1)-C(14)-
C(8) and C(8)-C(14)-C(9) bond angles.68 Although the
C(8)a-H(10) distance in 1Si, 2.95(3) Å, is larger a similar
difference between the bond angles is found. The occurrence
of the orthogonal geometry in 1C is even more surprising as it
leads to the extreme short C(9)-H(2) distance of 2.40(4) Å.
The respective distance in 1Si is 2.86(2) Å, close to the sum of
contact radii. In the crystal packing structures of 1Si and 1C,
donor and acceptor sites on neighboring molecules are also
oriented orthogonally, leading to an edge to face packing motif.69
Figure 8. Solid state UV spectra (KBr matrix) of 1Si and 1C and
their bissubstituted reference chromophores.29
are found.70 Although a quantitative comparison is not feasible
it is evident that even in the solid state UV spectra of 1Si and
1C are more than a superposition of reference chromophore
bands. In both cases an additional (CT) absorption at ap-
proximately 31 000 cm-1 is distinguishable. The orthogonality
of donor and acceptor orbitals at both the molecular and
supramolecular level excludes both an intramolecular through-
bond and an intermolecular through-space ground state interac-
tion. Although an NMR study on 4-substituted diphenyldi-
methylsilanes and 2,2-diphenylpropanes provided evidence that
substituent effects extend beyond the bridging atoms,29 this
transmission of substituent effects, indicative of a ground state
interaction, is operative Via σ-orbitals (bond polarization mech-
anism), which prevents an interaction with orbitals of π-sym-
metry. An intramolecular through-space donor-acceptor in-
teraction in the solid state structures of 1Si and 1C is however
feasible. All π-type orbitals of the cyanophenyl moiety are
antisymmetric with respect to the crystallographic mirror plane,
and can overlap with (dimethylamino)phenyl orbitals such as
the SHOMO and the LUMO (Figure 2) which are antisymmetric
with respect to this plane as well. In particular p-type AO’s of
atom C(2), which is nearest to the dimethylamino ring, are
involved. The overlap must however be very small, suggesting
once more that intensity borrowing contributes significantly to
the CT transitions.
The orthogonality of donor and acceptor orbitals in 1Si and
1C renders it interesting to investigate their solid state UV
spectra, which are depicted in Figure 8. The spectra are
compared with those of the (solid) bissubstituted reference
compounds 9Si, 9C, 10Si, and 10C since compounds 3Si, 3C,
5Si, and 5C were liquids. Inspection of data in Table 1 reveals
that these bissubstituted compounds exhibit their UV maxima
at the same positions as the monosubstituted compounds. The
solid state UV spectra of the donor-acceptor compounds are
in close agreement with their solution analogues; similar maxima
Discussion and Conclusions
The apparently conflicting situation has arisen that the
electronic interactions of the bridge with the donor and acceptor
(67) Cambridge Structural Database, Version 5.10, October 1995. Allen,
F. H.; Kennard, O. Chem. Design Automation News 1993, 8, 31-37.
(68) Perez, S.; Scaringe, R. P. Macromolecules 1987, 20, 68-77.
(69) Jorgensen, W. L.; Severance, D. L. J. Am. Chem. Soc. 1990, 112,
4768-4774.
(70) The intensities of the 1La bands appear to be relatively small, which
is attributed to poor reflectance of samples at wavenumbers higher than
42 000 cm-1
.

Comparison between SiMe2 and CMe2 Spacers as σ-Bridges
J. Am. Chem. Soc., Vol. 118, No. 35, 1996 8405
Table 9. Selected Bond Angles (deg) in 1Si and 1C with esd’s in Parentheses
bond anglea
1Si
1C
bond angle
1Si
1C
C(12)-N(2)-C(13)
C(13)-N(2)-C(13)a
C(2)-C(1)-Mb
119.69(9)
117.71(14)
123.47(14)
119.51(14)
119.54(18)
120.10(18)
177.9(2)
119.3(2)
117.7(3)
122.9(3)
119.4(3)
120.3(4)
119.4(4)
179.7(5)
C(10)-C(9)-M
N(2)-C(12)-C(11)
C(1)-M-C(8)
C(1)-M-C(9)
C(8)-M-C(9)
C(8)-M-C(8)a
122.25(8)
121.50(8)
107.00(6)
110.18(8)
110.22(6)
112.11(8)
121.94(18)
121.58(17)
107.7(2)
111.2(3)
110.6(2)
108.8(3)
C(6)-C(1)-M
C(3)-C(4)-C(7)
C(5)-C(4)-C(7)
N(1)-C(7)-C(4)
1
3
a Suffix a denotes symmetry operation x, /2-y, z for 1Si and x, /2-y, z for 1C. b M: Si(1) in 1Si, C(14) in 1C.
recognized previously that the first hyperpolarizability of
donor-acceptor substituted diphenyldimethylsilanes can be
described by a vector summation of the hyperpolarizabilities
of the Dσ and σA chromophores.23 The same should be true
for the donor-acceptor substituted 2,2-diphenylpropanes.
Figure 9. Through-space overlap of p atomic orbitals of diphenyldi-
methylsilanes and 2,2-diphenylpropanes in a butterfly conformation.
Experimental Section
General. Reactions involving organolithium reagents and/or silyl
chlorides were conducted in a nitrogen atmosphere using standard
Schlenk techniques. Starting materials and reagents were obtained from
commercial sources and used as received. Acetone, DMF, DMSO,
and methanol were stored on 4 Å or 3 Å molecular sieves prior to use.
Other solvents generally were distilled before use; diethyl ether was
distilled from sodium-benzophenone. Column chromatography was
performed on Merck kieselgel 60 silica (230-400 ASTM).
NMR spectra were obtained on a Bruker AC 300 apparatus, operating
at 300 MHz for 1H, 75 MHz for 13C, and 60 MHz for 29Si NMR. 29Si
NMR spectra were recorded with a proton decoupled inverse gated
pulse sequence; samples contained chromium(III) acetylacetonate as
relaxation reagent. Infrared spectra were recorded using a Perkin Elmer
283 infrared spectrometer. Solids were measured as KBr pellets and
liquids between NaCl plates. GC-MS spectra were collected on a ATI
Unicam Automass System 2 quadrupole mass spectrometer. Melting
points were determined using a Mettler FP5/FP51 photoelectric
apparatus. For gas chromatography a Varian 3700 chromatograph
equipped with a DB5 capillary column was used. Elemental analyses
were carried out at Kolbe Microanalytisches Laboratorium, Mu¨lheim
an der Ruhr, Germany.
moieties are more pronounced for 1Si, whereas the electronic
coupling of the CT state with both the LE and the ground state
is larger for 1C. The larger couplings in 1C can best be
explained by assuming that a through-space mechanism is
operative. Although the X-ray structures show that the donor
and acceptor moieties are orthogonally positioned, this molecular
conformation is in equilibrium with other forms in solution.
Other conformations involved in the equilibrium in these types
of compounds, i.e. monoatomically bridged diphenyls, are the
twist (propeller) and butterfly forms.66,71,72 Especially in the
latter conformation spatial overlap between p atomic orbitals
on phenyl carbon atoms adjacent to the bridge is possible (Figure
9).47,73 Since the C-C bond length is shorter than the Si-C
bond length (Table 8), the spatial overlap is larger for the
carbon-bridged compounds, which explains the stronger inter-
state couplings. This way of reasoning is supported by the
supposed larger rate of intersystem crossing Via exciton coupling
in the diphenylpropanes. Another contemplation which suggests
the involvement of a through-space mechanism is that, even
when in solution the orthogonality of the donors and acceptors
exists, the geometry of excited states can be quite different.
In summary, the nature of the monoatomic bridges under
investigation has a profound effect on the CT process. Ther-
modynamically charge transfer is more favorable for the SiMe2
bridge, which can be traced back to a lowering of the
cyanophenyl reduction potential by the silicon bridge. On the
other hand, the electronic coupling of the CT state with both
the LE state and the ground state is larger for the CMe2 bridge,
which is primarily associated to its smaller size. The photo-
physics of 1Si and 1C, in which the donor-acceptor distance
is very short, are dominated by intensity borrowing, which has
been demonstrated by both the solution and solid state electronic
spectra and the radiative decay behavior.
Cyclic voltammetry measurements were performed with a Heka PG
287 potentiostat/galvanostat in acetonitrile (Janssen p.A. grade, freshly
distilled from calcium hydride) containing 0.1 M tetrabutylammonium
hexafluorophosphate (Fluka, electrochemical grade) as supporting
electrolyte at scanning rates of 0.1 V‚s-1
. Redox potentials were
determined relative to Ag/(0.1 M AgNO3 in acetonitrile) and were
referenced to SCE by regularly measuring the oxidation potential of
the FeCp/FeCp•+ couple (E1/2 Vs SCE ) 0.31 V74 ) in this system.
Solution UV spectra were measured using a Cary 1 spectrometer in
spectrophotometric grade solvents (Janssen). Solid state UV spectra
were collected with a Cary 5 UV-vis-NIR spectrophotometer using
a Harrick praying mantis diffuse reflectance accessory on powdered
dispersions (approximately 1% w/w) in KBr. Reflectance ordinates R
were transformed into Kubelka-Munk units F(R) by F(R) ) (1 - R)2/
2R. Fluorescence spectra were obtained on a Spex Fluorolog instru-
ment, equipped with a Spex 1680 double excitation monochromator, a
Spex 1681 emission monochromator, and a Spex 1911F detector.
Solvents (Janssen) generally were of spectrophotometric grade and dried
on molecular sieves prior to use. Butyronitrile (Janssen, p.A.), di-n-
butyl ether (Janssen, 99%), and di-n-pentyl ether (Aldrich, 99%) were
distilled before use. Fluorescence emission spectra were corrected for
the detector spectral response with the aid of a correction file provided
by the manufacturer. Since this file does not cover the spectral range
above 33300 cm-1, fluorescence maxima in this region are uncorrected.
Fluorescence quantum yields were determined relative to naphthalene
(Φfl ) 0.23)75 at an excitation wavelength of 265 nm. Solutions were
degassed by purging with argon during 15 min. Fluorescence lifetimes
were determined using a Lumonix EX700 XeCl excimer laser (308
Furthermore, we have shown that an intramolecular CT
absorption occurs in a donor-acceptor substituted monosilane,
while other workers could not detect such an absorption.23,24
The CT absorption is however likely to be based on a through-
space interaction, by which it is unfortunately not possible to
evaluate the intrinsic ability of silicon, as compared to carbon,
to transmit electronic effects.
The intensity of the CT absorption is too small to expect
interesting second-order nonlinear optical responses for donor-
acceptor substituted diphenyldimethylsilanes, despite the quite
large excited state dipole moments. It has indeed been
(71) Bothorel, P. Annu. Chim. (Paris) 1959, 13, 669-712.
(72) Edlund, U.; Norstro¨m, Å. Org. Magn. Reson. 1977, 9, 196-202.
(73) Craenen, H. A. H.; Verhoeven, J. W.; de Boer, T. J. Recl. TraV.
Chim. Pays-Bas 1972, 91, 405-416.
(74) Dachbach, J.; Blackwood, D.; Pons, J. W.; Pons, S. J. Electroanal.
Chem. 1987, 237, 269-273.
(75) Eaton, D. F. Pure Appl. Chem. 1988, 60, 1107-1114.

8406 J. Am. Chem. Soc., Vol. 118, No. 35, 1996
Van Walree et al.
nm, 8.4 ns fwhm) as excitation source, monitoring the fluorescence
signal at a right angle by a RCA C-31025C photodiode Via a Zeiss
MQ II monochromator. The signal was fed into a Tektronix TDS 684A
oscilloscope triggered by a photodiode that detects the laser pulse.
Samples (A < 0.2 at 308 nm) were degassed by purging with argon. A
description of the equipment and experimental details of the flash
photolysis TRMC technique can be found elsewhere.55,56
Extended Hu¨ckel (EH) calculations were performed with the
CACAO program76,77 on PM378,79 optimized (precise option) geometries.
Silicon d orbital parameters, not implemented in CACAO, were taken
from the Forticon8 program.80 Other parameters used in both programs
are identical. All calculations were run on a 486/33 MHz personal
computer.
X-ray Structure Determinations. Details of the structure deter-
mination of 1Si have been reported elsewhere;26 pertinent data have
been included in Table 7. A colorless crystal of 1C suitable for X-ray
diffraction was glued to the tip of a Lindemann-glass capillary and
transferred into the cold-nitrogen stream of an Enraf-Nonius CAD4-T
diffractometer on rotating anode. Accurate unit-cell parameters and
an orientation matrix were determined by least-squares fitting of the
setting angles of 25 well-centered reflections (SET481 ) in the range
11.40° < θ < 14.05°. The unit-cell parameters were checked for the
presence of higher lattice symmetry.82 Crystal data and some details
on data collection are presented in Table 7. Reflection data were
collected in ω scan mode with scan angle ∆ω ) 0.50 + 0.35 tan θ°.
Intensity data of 3869 reflections were collected in the range 1.55° <
θ < 27.50°, 1768 of which are independent. Data were corrected for
Lp effects and for a linear decay of 6% of the three periodically
measured reference reflections (2h 2 2, 2 2 2, 0 2 2) during 10 h of
X-ray exposure time, but not for absorption. The structure was solved
by automated direct methods (SHELXS8683 ). Refinement on F2 was
carried out by full-matrix least-squares techniques (SHELXL-9384 );
no observance criterion was applied during refinement. Hydrogen
atoms were located on a difference Fourier map and their coordinates
were included in the refinement. Non-hydrogen atoms were refined
with anisotropic thermal parameters. The hydrogen atoms were refined
with a fixed isotropic thermal parameter related to the value of the
equivalent isotropic displacement parameter of their carrier atoms by
a factor of 1.5 for the methyl hydrogen atoms and 1.2 for the other
hydrogen atoms. In the final refinement cycles 141 parameters were
to a solution of dichlorodimethylsilane (92 mL, 670 mmol) in 75 mL
of diethyl ether at such a rate that the mixture refluxed gently. A white
solid formed immediately. After finishing the addition the mixture
was refluxed for an additional 2 h and stirred overnight at room
temperature. The etheral solution was decanted under nitrogen and
subsequently volatile components were distilled off at atmospheric
pressure (35-80 °C). The residue was fractionally distilled under
reduced pressure (0.2 mmHg). The fraction boiling at 92-96 °C proved
to be 11. Yield 24.40 g (98 mmol, 40%) of a colorless liquid. Purity
85% by GC. 1H NMR (CDCl3): δ 7.57 and 7.51 (AA′BB′, 2 × 2H,
Ar-H), 0.71 (s, 6H, SiMe). GC-MS: m/e 248 (M•+).
(4-Bromophenyl)(4′-(dimethylamino0phenyl)dimethylsilane (12Si).
A solution of 4-bromo-N,N-dimethylaniline (6.70 g, 33.5 mmol) in 50
mL of diethyl ether was added dropwise to finely cut lithium pieces
(0.64 g, 92.2 mmol) in 20 mL of diethyl ether. The mixture was stirred
overnight at room temperature; GC showed that at that time the starting
compound was not present anymore. The turbid solution was filtered
over quartz wool into a dropping funnel and added to a solution of 11
(8.01 g, purity 85%, 27.3 mmol) in 20 mL of diethyl ether, after which
the mixture was refluxed overnight. Water (25 mL) was added
cautiously, and the layers were separated. Subsequently the organic
layer was washed with 3 × 25 mL of a 10% ammonium chloride
solution, dried on magnesium sulfate, and filtered. After concentration
under reduced pressure a purple liquid was obtained, which solidified
on standing. 12Si (7.32 g, 21.9 mmol, 80%) was obtained in the form
of white crystals after flash chromatography (silica, chloroform) and
recrystallization from pentane. Mp 58 °C. 1H NMR (CDCl3): δ 7.50
and 7.41 (AA′BB′, 2 × 2H, Ar-H), 7.39 and 6.77 (2 × AA′XX′, 2H,
Ar-H), 2.99 (s, 6H, NMe), 0.53 (s, 6H, SiMe). 13C NMR (CDCl3): δ
151.2, 138.4, 135.8, 135.3, 130.8, 123.7, 122.2, 112.0 (all aromatic
C), 40.2 (NMe), -2.2 (SiMe). 29Si NMR (CDCl3): δ -8.4. IR
(KBr): 2815 (NMe2), 1250, 835-800 (SiMe), 1115 (Si-Ar) cm-1
.
(4-Cyanophenyl)(4′-(dimethylamino)phenyl)dimethylsilane (1Si).
In a nitrogen atmosphere, a mixture of 12Si (2.52 g, 7.54 mmol),
copper(I) cyanide (0.82 g, 9.15 mmol), and 5 mL of DMF was refluxed
for 5 h. The cooled brownish mixture was then poured into 100 mL
of 25% ammonia. After passing air through this suspension for 3.5 h
under vigorous stirring, it was extracted with 3 × 30 mL of chloroform.
The combined organic layers were concentrated to a volume of ca. 10
mL, and 75 mL of hexane was added. This solution was washed with
water (3 × 50 mL), dried on magnesium sulfate, filtered, and evaporated
under reduced pressure. The resulting brownish solid was purified by
flash chromatography (eluent chloroform) and recrystallization from
methanol. Yield 1.58 g (56.3 mmol, 75%) of block shaped crystals
with a greenish glance. Mp 128 °C. Anal. Calcd for C17H20N2Si: C,
72.81; H, 7.19; N, 9.99; Si, 10.01. Found: C, 72.66; H, 7.25; N, 9.95;
Si, 9.88. 1H NMR (CDCl3): δ 7.66 and 7.61 (AA′BB′, 2 × 2H, Ar-
H), 7.42 and 6.79 (AA′XX′, 2 × 2H, Ar-H), 3.01 (s, 6H, NMe), 0.58
(s, 6H, SiMe). 13C NMR (CDCl3): δ 151.4, 146.7, 135.3, 134.6, 130.9,
121.0, 112.4, 112.0 (all aromatic C), 119.2 (CN), 40.1 (NMe), -2.4
(SiMe). 29Si NMR (CDCl3): δ -7.9. IR (KBr): 2820 (NMe2), 2230
optimized using the weighting scheme w ) [σ2(F2) + (0.0303P)2]-1
,
with P ) (Max(F2o,0) + 2Fc2)/3. A final difference Fourier map
showed no residual density outside -0.20 and 0.24 e‚Å-3. Positional
parameters are included in the supporting information. Neutral atom
scattering factors and anomalous dispersion corrections were taken from
the International Tables for Crystallography.85 Geometrical calcula-
tions and illustrations were performed with PLATON;65 all calculations
were performed on a DECstation 5000 cluster.
Syntheses. The preparation of the compounds 3Si, 4Si, 5Si, 6Si,
7Si, and 8Si has been reported previously.29 The preparation and
characterization of synthetic intermediates 13Si, 2-(4-bromophenyl)-
2-propanol, 13C, 14, 15, 16, 17, 18, 19, 20, 12C, and 21 is described
in the supporting information. 4-tert-Butyl-N,N-dimethylaniline (6C)
was purchased from Aldrich and used as received.
(CN), 1260, 835-800 (SiMe), 1115 (Si-Ar) cm-1
.
(4-Cyanophenyl)(4′-methoxyphenyl)dimethylsilane (2Si). This
compound was prepared as described for its dimethylamino analogue
1Si. It was purified by column chromatography (eluent chloroform)
and obtained as white crystals upon trituration with methanol. Yield
76%. Mp 41 °C. Anal. Calcd for C16H17NOSi: C, 71.87; H, 6.41;
(4-Bromophenyl)chlorodimethylsilane (11). A solution of 4-bro-
mophenyllithium86 (250 mmol) in 250 mL of diethyl ether was added
(76) Mealli, C.; Proserpio, D. M. CACAO, Version 3.0: Florence, 1992.
(77) Mealli, C.; Proserpio, D. M. J. Chem. Educ. 1990, 67, 399-402.
(78) Stewart, J. J. P. J. Comp. Chem. 1989, 10, 209-220.
(79) Stewart, J. J. P., MOPAC, Version 6.0; Serena Software: Bloom-
ington, 1990; QCPE Program 504.
(80) Howell, J.; Rossi, A.; Wallace, D.; Haraki, K.; Hoffmann, R.,
Forticon8, QCPE QCMP011.
1
N, 5.24; Si, 10.50. Found: C, 70.21; H, 6.50; N, 5.11; Si, 10.15. H
NMR (CDCl3): δ 7.60 (AA′BB′, 4H, Ar-H), 7.42 and 6.94 (AA′XX′,
2H, Ar-H), 3.83 (s, 3H, OMe), 0.56 (s, 6H, SiMe). 13C NMR
(CDCl3): δ 160.9, 145.9, 135.6, 134.6, 131.0, 127.2, 113.9, 112.6 (all
aromatic C), 119.0 (CN), 55.1 (OMe), -2.5 (SiMe). 29Si NMR
(CDCl3): δ -7.4. IR (KBr): 2830, 1250, 1030 (Ar-O-Me), 2235 (CN),
(81) de Boer, J. L.; Duisenberg, A. J. M. Acta Crystallogr. 1984, A40,
C410.
1255, 840-800 (SiMe), 1115 (Si-Ar) cm-1
.
2-(4-Methoxyphenyl)-2-phenylpropane (4C). In a 1-necked flask
2-phenylpropanol (3.78 g, 27.7 mmol) and 85% phosphoric acid (3
mL, 43.8 mmol) were mixed. To the white suspension was added
anisole (3.00 g, 27.7 mmol) and the mixture was stirred overnight at
90 °C. After addition of 25 mL of diethyl ether the solution was washed
three times with 15 mL of water, dried on magnesium sulfate, and
filtered. The diethyl ether was removed in a rotary evaporator and the
mixture was kugelrohr distilled, collecting the fraction boiling in the
(82) Spek, A. L. J. Appl. Crystallogr. 1988, 21, 578.
(83) Sheldrick, G. M., SHELXS86 Program for crystal structure deter-
mination, University of Go¨ttingen, Germany, 1986.
(84) Sheldrick, G. M., SHELXL93 Program for crystal structure refine-
ment, University of Go¨ttingen, Germany, 1993.
(85) International Tables for Crystallography; Wilson, A. J. C., Ed.;
Kluwer Academic: Dordrecht, 1992; Vol. C.
(86) Eisch, J. J. Non-Transition Metal Compounds; Organometallic
Synthesis. Vol. 2; Academic: New York, 1981; p 93.

Comparison between SiMe2 and CMe2 Spacers as σ-Bridges
J. Am. Chem. Soc., Vol. 118, No. 35, 1996 8407
range 120-135 °C (0.1 mmHg). This fraction was separated into its
components with column chromatography. The first component (2.04
g, 8.63 mmol, 31%) was eluted with hexane and was identified as 2,4-
diphenyl-4-methyl-1-pentene by 1H NMR, IR, and GC-MS (2,4-
diphenyl-4-methyl-1-pentene is one of the dimers of the reaction
intermediate R-methylstyrene87-89). The second component was iso-
lated after elution with ethyl acetate-hexane 1:1 (v/v) and proved to
be 4C. Yield 1.32 g (5.83 mmol, 21%) of a colorless oil. Anal. Calcd
for C16H18O: C, 84.91; H, 8.02. Found: C, 84.82; H, 7.96. 1H NMR
(CDCl3): δ 7.37 (m, 4H, Ar-H), 7.24 (m, 1H, Ar-H), 7.22 and 6.90
(AA′XX′, 2 × 2H, Ar-H), 3.85 (s, 3H, OMe), 1.75 (s, 6H, CMe). 13C
NMR (CDCl3): δ 157.4, 150.8, 142.7, 127.9, 127.7, 126.6, 125.4, 113.2
(all aromatic C), 54.9 (OMe), 42.2 (CMe), 30.8 (CMe). IR (NaCl):
7.14 and 6.68 (AA′XX′, 2 × 4H, Ar-H), 2.95 (s, 12H, NMe2), 1.64 (s,
6H, CMe). 13C NMR (CDCl3): δ 148.5, 139.5, 127.4, 112.3 (all
aromatic C), 41.1 (CMe), 40.7 (NMe), 31.0 (CMe). IR (KBr): 2810
(NMe2) cm-1
.
2-(4-Cyanophenyl)-2-(4′-(dimethylamino)phenyl)propane (1C).
This compound was prepared as described for 1Si. Purification was
carried out with column chromatography (eluent chloroform) and
recrystallization from methanol. Yield 0.41 g (1.55 mmol, 30%) of
crystals with a greenish glance. Mp 136 °C. Anal. Calcd for
C17H14N2: C, 81.78; H, 7.63; N, 10.59. Found: C, 81.66; H, 7.58; N,
10.52. 1H NMR (CDCl3): δ 7.56 and 7.37 (AA′BB′, 2 × 2H, Ar-H),
7.09 and 6.71 (AA′XX′, 2 × 2H, Ar-H), 2.95 (s, 6H, NMe), 1.69 (s,
6H, CMe). 13C NMR (CDCl3): δ 157.2, 148.9, 136.9, 131.8, 127.6,
127.4, 112.4, 109.3 (all aromatic C), 119.2 (CN), 42.6 (CMe), 40.6
2830, 1250, 1030 (Ar-O-Me) cm-1
.
(NMe), 30.5 (CMe). IR (KBr): 2820 (NMe2), 2230 (CN) cm-1
.
2-(4-Cyanophenyl)-2-(4′-methoxyphenyl)propane (2C). This com-
pound was prepared from 13C (1.81 g, 5.93 mmol) and copper(I)
cyanide (0.63 g, 7.08 mmol) as described for 1Si. The crude product
was purified by column chromatography (eluent chloroform-hexane
2:1 (v/v)) to give 1.07 g (4.26 mmol, 72%) of a colorless liquid. Anal.
Calcd for C17H17NO: C, 81.24; H, 6.82; N, 5.57; O, 6.37. Found: C,
81.31; H, 6.78; N, 5.49; O, 6.45. 1H NMR (CDCl3): δ 7.58 and 7.37
(AA′BB′, 2 × 2H, Ar-H), 7.15 and 6.87 (AA′XX′, 2 × 2H, Ar-H),
3.82 (s, 3H, OMe), 1.71 (s, 6H, CMe). 13C NMR (CDCl3): δ 157.9,
156.6, 141.2, 132.0, 127.7, 127.6, 113.6, 109.4 (all aromatic C), 119.1
(CN), 54.9 (OMe), 42.8 (CMe), 30.5 (CMe). IR (NaCl): 2830, 1250,
2-(4-Cyanophenyl)-2-phenylpropane (5C). 2-(4-Bromophenyl)-
2-phenylpropane (21, 0.54 g, 1.96 mmol) was converted into the
corresponding cyanide following the approach as described for 1Si.
After purification by column chromatography (eluent chloroform) a
colorless oil (0.17 g, 0.76 mmol, 41%) was obtained. Anal. Calcd
for C16H15N: C, 86.84; H, 6.83; N, 6.33. Found: C, 86.94; H, 6.83;
N, 6.36. 1H NMR (CDCl3): δ 7.56 and 7.35 (AA′BB′, 2 × 2H, Ar-
H), 7.32-7.21 (m, 5H, Ar-H), 1.72 (s, 6H, CMe). 13C NMR (CDCl3):
δ 156.2, 148.9, 131.8, 128.2, 127.5, 126.5, 126.1, 109.4 (all aromatic
C), 119.0 (CN), 43.3 (CMe), 30.2 (CMe). IR (NaCl): 2230 (CN) cm-1
.
1040 (Ar-O-Me), 2240 (CN) cm-1
.
4-tert-Butylbenzonitrile (8C). This compound was synthesized
from 4-bromo-tert-butylbenzene as described for the synthesis of 1Si.
After distillation (boiling range 67-69 °C, 0.1 mmHg) a colorless liquid
was obtained in 71% yield. 1H NMR (CDCl3): δ 7.57 and 7.46
(AA′BB′, 2 × 2H, Ar-H), 1.31 (s, 9H, CMe). 13C NMR (CDCl3): δ
156.6, 131.9, 126.2, 109.3 (all aromatic C), 119.1 (CN), 35.3 (CMe),
2,2-Bis(4-cyanophenyl)propane (10C). Synthesis of this compound
from 17 was performed following the same procedure as employed
for the other cyanides. After purification by flash chromatography
(eluent chloroform) and recrystallization from methanol 10C was
isolated in the form of white crystals. Yield 71%. Mp 143 °C. Anal.
Calcd for C17H14N2: C, 82.90; H, 5.73; N, 11.37. Found: C, 82.76;
H, 5.71; N, 11.31. 1H NMR (CDCl3): δ 7.57 and 7.30 (AA′BB′, 2 ×
4H, Ar-H), 1.69 (s, 6H, CMe). 13C NMR (CDCl3): δ 154.5, 132.1,
127.5, 110.2 (all aromatic C), 118.6 (CN), 43.9 (CMe), 30.0 (CMe).
30.9 (CMe). IR (NaCl): 2230 (CN), 1370 (CMe3) cm-1
.
4-tert-Butylanisole (7C). Potassium carbonate (27.20 g, 196.8
mmol) was added in one portion to a solution of 4-tert-butylphenol
(9.95 g, 66.24 mmol) in 100 mL of DMSO. After the mixture was
stirred for 50 min methyl iodide (5 mL, 80.3 mmol) was added
dropwise, keeping the temperature at 18 °C by cooling with an ice-
bath. Subsequently the mixture was stirred 1 h at 45 °C and overnight
at room temperature. Since GC analysis showed the presence of some
starting material another 3 mL of iodomethane (48 mmol) were added,
and stirring was continued for 2 h at 50 °C. Water (500 mL) was then
added and the mixture was extracted with 3 × 50 mL of methylene
chloride. The combined organic layers were concentrated under
diminished pressure, after which the residual yellow liquid was
dissolved in 60 mL of hexane and washed with 50 mL of 5% potassium
hydroxide solution and 3 × 50 mL of water. After drying on
magnesium sulfate, filtering, and evaporation of the solvent the residual
liquid was vacuum distilled over a vigreux column (bp 58 °C at 0.1
mmHg). Yield 7.83 g (47.7 mmol, 72%) of a colorless liquid. 1H
NMR (CDCl3): δ 7.38 and 6.92 (AA′XX′, 2 × 2H, Ar-H), 3.85 (OMe),
1.38 (s, 9H, CMe). 13C NMR (CDCl3): δ 157.4, 143.4, 126.2, 113.4
(all aromatic C), 55.2 (OMe), 34.1 (CMe), 31.6 (CMe). IR (NaCl):
IR (KBr): 2220 (CN) cm-1
.
2-(4-(Dimethylamino)phenyl)-2-phenylpropane (3C). A solution
of amine 18 (0.49 g, 2.32 mmol) and methyl iodide (1.2 mL, 19.3 mmol)
in 10 mL of ethanol was refluxed overnight in the presence of potassium
carbonate (1.60 g, 11.6 mmol). After cooling to room temperature
solids were filtered off and washed thoroughly with water. The filtrate,
in which another crop of solid had precipitated, was filtered again, and
the residue was washed as well. The combined residues were dried in
a vacuum desiccator over potassium hydroxide. The quaternary amine
salt (0.79 g) was subsequently boiled for 5 min in 5 mL of freshly
distilled ethanolamine. Water (10 mL) was added, and the resulting
brown liquid precipitate was isolated by extraction with 3 × 10 mL of
methylene chloride, drying on calcium chloride, filtering, and evapora-
tion of the solvent. Pure 3C was obtained as a light tanned solid after
column chromatography with chloroform as eluent. Yield 0.34 g (1.42
mmol, 61% with respect to amine 18). Mp 36 °C. Anal. Calcd for
C17H21N: C, 85.31; H, 8.84; N, 5.85. Found: C, 85.21; H, 8.81; N,
5.79. 1H NMR (CDCl3): δ 7.36 (m, 4H, Ar-H), 7.23 (m, 1H, Ar-H),
7.21 and 6.76 (AA′XX′, 2H, Ar-H), 3.00 (s, 6H, NMe), 1.76 (s, 6H,
CMe). 13C NMR (CDCl3): δ 151.3, 148.5, 138.7, 127.8, 127.4, 126.7,
125.3, 112.2 (all aromatic C), 42.0 (CMe), 40.6 (NMe), 30.8 (CMe).
2820, 1250, 1040 (Ar-O-Me), 1370 (CMe3) cm-1
.
Acknowledgment. This work was supported in part (A.L.S)
by the Netherlands Foundation of Chemical Research (SON)
with financial aid from the Netherlands Organization for
Scientific Research (NWO).
IR (KBr): 2830 (NMe2) cm-1
.
2,2-Bis(4-(dimethylamino)phenyl)propane (9C). Bisamine 19 was
methylated using the same method as employed for the methylation of
18. Yield 61% of light pink crystals after recrystallization from
methanol. Mp 82 °C. Anal. Calcd for C19H26N2: C, 80.80; H, 9.28;
N, 9.92. Found: C, 80.68; H, 9.22; N, 9.86. 1H NMR (CDCl3): δ
Supporting Information Available: Preparation and char-
acterization of synthetic intermediates (13Si, 2-(4-bromophenyl)-
2-propanol, 13C, 14, 15, 16, 17, 18, 19, 20, 12C and 21) and
further details of the structure determination, including atomic
coordinates, bond lengths and angles and thermal parameters
for 1C (Tables S1-7) (13 pages). See any current masthead
page for ordering and Internet access instructions.
(87) Welsh, L. H.; Drake, N. L. J. Am. Chem. Soc. 1938, 60, 59-62.
(88) Kawakami, Y.; Toyoshima, N.; Yamashita, Y. Chem. Lett. 1980,
13-16.
(89) Higashimura, T.; Nishii, H. J. Polym. Sci., Polym. Chem. Ed. 1977,
15, 329-339.
JA960546E