Photochemistry of group 14 1,1,1-trimethyl-2,2,2-triphenyldimetallanes (Ph3MM′Me3; M, M′ = Si, Ge). Direct detection and characterization of silene and germene reactive intermediates

Article abstract of DOI:10.1021/om000174p

The photochemistry of trimethylsilyltriphenylgermane (Ph₃GeSiMe₃), triphenylsilyltrimethylgermane (Ph₃SiGeMe₃), and 1,1,1-trimethyl-2,2,2-triphenyldigermane (Ph₃GeGeMe₃) has been studied in hydrocarbon solution by steady state and laser flash photolysis methods and is compared to previously reported data for the homologous disilane Ph₃SiSiMe₃. A variety of products are formed upon photolysis of the three compounds in the presence of 2,3-dimethyl-1,3-butadiene or chloroform, but in each case the major ones are derived from M-M′ bond homolysis and dimethyl- or diphenylgermylene extrusion. The trapping products of the 1,3,5-(1-metalla)hexatriene derivatives formed by [1,3]-MMe₃ migration into the ortho-position of one of the phenyl rings are formed as well, in yields of 9-30%. While these experiments indicate that germylenes are formed in at least twice the yield of the 1,3,5-(1-metalla)hexatrienes, only the latter and triphenylsilyl or triphenylgermyl radicals can be detected by laser flash photolysis techniques. The metallaenes have been identified on the basis of their time-resolved UV absorption spectra and absolute rate constants for reaction with 2,3-dimethylbutadiene, methanol, acetone, acetic acid, oxygen, and carbon tetrachloride and can be distinguished from germylenes by their lack of reactivity toward triethylsilane and chloroform. Radical formation is shown to result from reaction of the triplet states of these compounds, and a triplet lifetime is estimated for Ph₃GeSiMe₃ and compared to that of the disilane homologue. The results of time-resolved experiments on other, related compounds are discussed in light of these results.

Full text of DOI:10.1021/om000174p

3232
Organometallics 2000, 19, 3232-3241
P h otoch em istr y of Gr ou p 14
1,1,1-Tr im eth yl-2,2,2-tr ip h en yld im eta lla n es (P h 3MM′Me3;
M, M′ ) Si, Ge). Dir ect Detection a n d Ch a r a cter iza tion of
Silen e a n d Ger m en e Rea ctive In ter m ed ia tes
William J . Leigh* and Nicholas P. Toltl
Department of Chemistry, McMaster University, 1280 Main Street West,
Hamilton, Ontario, Canada L8S 4M1
Paula Apodaca, Madeli Castruita, and Keith H. Pannell
Department of Chemistry, University of Texas at El Paso, El Paso, Texas 79968-0513
Received February 24, 2000
The photochemistry of trimethylsilyltriphenylgermane (Ph3GeSiMe3), triphenylsilyltri-
methylgermane (Ph3SiGeMe3), and 1,1,1-trimethyl-2,2,2-triphenyldigermane (Ph3GeGeMe3)
has been studied in hydrocarbon solution by steady state and laser flash photolysis methods
and is compared to previously reported data for the homologous disilane Ph3SiSiMe3. A
variety of products are formed upon photolysis of the three compounds in the presence of
2,3-dimethyl-1,3-butadiene or chloroform, but in each case the major ones are derived from
M-M′ bond homolysis and dimethyl- or diphenylgermylene extrusion. The trapping products
of the 1,3,5-(1-metalla)hexatriene derivatives formed by [1,3]-MMe3 migration into the ortho-
position of one of the phenyl rings are formed as well, in yields of 9-30%. While these
experiments indicate that germylenes are formed in at least twice the yield of the 1,3,5-(1-
metalla)hexatrienes, only the latter and triphenylsilyl or triphenylgermyl radicals can be
detected by laser flash photolysis techniques. The metallaenes have been identified on the
basis of their time-resolved UV absorption spectra and absolute rate constants for reaction
with 2,3-dimethylbutadiene, methanol, acetone, acetic acid, oxygen, and carbon tetrachloride
and can be distinguished from germylenes by their lack of reactivity toward triethylsilane
and chloroform. Radical formation is shown to result from reaction of the triplet states of
these compounds, and a triplet lifetime is estimated for Ph3GeSiMe3 and compared to that
of the disilane homologue. The results of time-resolved experiments on other, related
compounds are discussed in light of these results.
Sch em e 1
In tr od u ction
The photochemistry of aryldisilanes has been exten-
sively investigated over the past 25 years.1-5 In general,
photolysis of these compounds leads to the formation
of three distinct types of reactive intermediates, in
relative yields that vary markedly depending on the
degree of aryl substitution in the disilane and on the
solvent. This complexity is most aptly demonstrated by
the photochemistry of methylpentaphenyldisilane (1),
which yields silenes 2 and 3 by [1,3]-SiPh2Me migration
and dehydrosilylation (elimination of H-SiPh3 (4)),
respectively, and the silyl free radicals 5 and 6 by
Si-Si bond homolysis (Scheme 1).6-8 The formation of
the (1-sila)hexatriene 2 by concerted9,10 [1,3]-silyl mi-
gration is relatively minor in 1, but is quite frequently
the major photoreaction of aryldisilanes in solution.7,11-19
(1) Shizuka, H.; Hiratsuka, H. Res. Chem. Intermed. 1992, 18, 131.
(2) Steinmetz, M. G. Chem. Rev. 1995, 95, 1527.
(3) Brook, A. G.; Brook, M. A. Adv. Organomet. Chem. 1996, 39, 71.
(4) Brook, A. G. In The chemistry of organic silicon compounds;
Rappoport, Z., Apeloig, Y., Eds.; J ohn Wiley and Sons: New York, 1998;
pp 1233-1310.
(8) Sluggett, G. W.; Leigh, W. J . Organometallics 1994, 13, 1005.
(9) Ohshita, J .; Niwa, H.; Ishikawa, M. Organometallics 1996, 15,
4632.
(10) Ohshita, J .; Niwa, H.; Ishikawa, M.; Yamabe, T.; Yoshii, T.;
Nakamura, K. J . Am. Chem. Soc. 1996, 118, 6853.
(11) Ishikawa, M.; Fuchikami, T.; Sugaya, T.; Kumada, M. J . Am.
Chem. Soc. 1975, 97, 5923.
(5) Kira, M.; Miyazawa, T. In The chemistry of organic silicon
compounds; Rappoport, Z., Apeloig, Y., Eds.; J ohn Wiley & Sons: New
York, 1998; pp 1311-1337.
(6) Boudjouk, P.; Roberts, J . R.; Golino, C. M.; Sommer, L. H. J .
Am. Chem. Soc. 1972, 94, 7926.
(7) Sluggett, G. W.; Leigh, W. J . J . Am. Chem. Soc. 1992, 114, 1195.
(12) Ishikawa, M.; Fuchikami, T.; Kumada, M. J . Organomet. Chem.
1976, 118, 139.
10.1021/om000174p CCC: $19.00 © 2000 American Chemical Society
Publication on Web 07/14/2000

Group 14 1,1,1-Trimethyl-2,2,2-triphenyldimetallanes
Organometallics, Vol. 19, No. 16, 2000 3233
The opposite is true of the dehydrosilylation reaction
to produce silene 3. Both these reaction pathways are
singlet-derived; silyl free radical formation results from
the triplet state of the disilane and hence can be made
to predominate either by enhancing intersystem cross-
ing through adjustment of the solvent polarity17 or by
slowing down the singlet-derived reaction pathways by
attaching sterically bulky substituents.20
Disilanes are the simplest congeners of higher oligo-
and polysilanes, whose photochemical and photophysical
behavior has also been of considerable interest.21-23 The
photochemistry of trisilanes and higher oligosilanes is
usually characterized by predominant extrusion of si-
lylene reactive intermediates,24 a process of relatively
minor importance in aryldisilanes except in the presence
of special structural constraints.25
Laser flash photolysis studies of aryldisilanes have
provided a wealth of mechanistic information on the
reactions of transient silenes with various “sileno-
philes” such as alcohols, ketones and aldehydes, car-
boxylic acids, dienes, and oxygen, as well as important
insight into various aspects of disilane photophysics and
the effects of substituents on disilane photochemis-
try.7,8,17,19,20,26,27 1,1,1-Trimethyl-2,2,2-triphenyldisilane
(7a ) has proven to be particularly useful in this re-
gard.17,19,26,27 In hydrocarbon solvents, photolysis of 7a
yields cleanly and in high yield the (transient) 1,1-
diphenyl-(1-sila)hexatriene derivative 8a , which can
easily be detected and studied by such methods. In
contrast, photolysis in more polar solvents leads to high
yields of the corresponding silyl free radicals, with 8a
being formed in low (but still detectable) amounts. The
remarkable sensitivity of the photochemistry of this
compound to solvent polarity provided the impetus for
the discovery that silyl free radicals are formed via the
triplet states of aryldisilanes and are not precursors to
silene products (at least on the >10-8 s time scale).17
The reactivity of 8a also proved to be unusual, owing
to unique steric constraints which serve to direct the
chemistry of the SidC bond along pathways that are
often only minor in other aryldisilane-derived silenes.
This allowed, for example, the isolation of stable 1,2-
siloxetanes from reaction with acetone and other ke-
tones and provided much detail on the mechanism of
this well-known (but seldom fruitful) reaction of tran-
sient silenes.17,26,27
derivatives 7b,c and of 1,1,1-trimethyl-2,2,2-triphenyl-
digermane (7d ) by steady state and laser flash photoly-
sis techniques. Our intent was to examine, in a system-
atic way, the changes induced in the photochemistry of
7a as a result of replacement of first one and then both
of the silicon atoms with germanium. We anticipated
(based on literature precedent; vide infra) that photoly-
sis of these compounds might afford the corresponding
(1-metalla)hexatriene reactive intermediates 8b-d in
high enough yields that they might be detected by flash
photolysis methods, thus affording the opportunity to
broaden the scope of our studies of the differences in
the reactivity of silicon- and germanium-carbon double
bonds.28,29
Previously reported work on 7d 30 and a related series
of compounds (9b-d ) by Mochida and Gaspar and their
co-workers31,32 suggests that reactive intermediates of
this type are formed in significant yields in the pho-
tolysis of arylgermasilanes, but in minor amounts
relative to silyl and/or germyl radicals and dimethylger-
mylene. Germylene formation is even more dominant
in the photochemistry of FeII-germasilane and -silager-
mane complexes, where the efficiency of the reaction
appears to not depend greatly on which of the two group
14 atoms is attached to the metal center. For example,
irradiation of (η5-C5H5)Fe(CO)2SiMe2GeMe3 and its
isomer (η5-C5H5)Fe(CO)2GeMe2SiMe3 each results mainly
in elimination of Me2Ge with concomitant formation of
(η5-C5H5)Fe(CO)2SiMe3.33
Laser flash photolysis experiments on 9b-d verified
that free radical formation dominates the photochem-
istry of these compounds and allowed the detection of
an additional transient product absorbing in the 400-
460 nm region of the spectrum, which was assigned to
dimethylgermylene.31,32 On the basis of our and others
experience with the disilane analogue 9a 7,16,19,26 and a
recently acquired knowledge of the quantitative aspects
of transient germene reactivity,28,29 it seemed possible
to us that these transients might better be assigned to
the (1-germa)hexatriene species 10c,d . This suspicion
provided further justification for the present study of
7b-d , since the corresponding transient 1,1-diphenyl-
(1-metalla)hexatrienes, if obtained, would be expected
to exhibit absorption maxima at significantly longer
wavelengths than those that have been reported by
various workers (under a variety of conditions) for either
Me2Ge:32,34-37 or Ph2Ge:.30,35,36,38
(22) West, R. In The chemistry of organic silicon compounds; Patai,
S., Rappoport, Z., Eds.; J ohn Wiley & Sons: New York, 1989; pp 1207-
1240.
(23) Michl, J . Synthet. Met. 1992, 50, 367.
(24) Mazieres, S.; Raymond, M. K.; Raabe, G.; Prodi, A.; Michl, J .
J . Am. Chem. Soc. 1997, 119, 6682.
(25) Kira, M.; Sakamoto, K.; Sakurai, H. J . Am. Chem. Soc. 1983,
105, 7469.
(26) Leigh, W. J .; Sluggett, G. W. J . Am. Chem. Soc. 1994, 116,
10468.
In this paper, we report the results of a study of the
photochemistry of the trimethyltriphenylgermasilane
(13) Ishikawa, M.; Fuchikami, T.; Kumada, M. J . Organomet. Chem.
1976, 118, 155.
(14) Ishikawa, M.; Fuchikami, T.; Kumada, M. J . Organomet. Chem.
1977, 133, 19.
(15) Sakurai, H. J . Organomet. Chem. 1980, 200, 261.
(16) Shizuka, H.; Okazaki, K.; Tanaka, M.; Ishikawa, M.; Sumitani,
M.; Yoshihara, K. Chem. Phys. Lett. 1985, 113, 89.
(17) Leigh, W. J .; Sluggett, G. W. J . Am. Chem. Soc. 1993, 115, 7531.
(18) Kira, M.; Miyazawa, T.; Sugiyama, H.; Yamaguchi, M.; Sakurai,
H. J . Am. Chem. Soc. 1993, 115, 3116.
(19) Leigh, W. J .; Sluggett, G. W. Organometallics 1994, 13, 269.
(20) Sluggett, G. W.; Leigh, W. J . Organometallics 1992, 11, 3731.
(21) Miller, R. D.; Michl, J . Chem. Rev. 1989, 89, 1359.
(27) Toltl, N. P.; Leigh, W. J . Organometallics 1996, 15, 2554.
(28) Toltl, N. P.; Leigh, W. J . J . Am. Chem. Soc. 1998, 120, 1172.
(29) Leigh, W. J . Pure Appl. Chem. 1999, 71, 453.
(30) Mochida, K.; Wakasa, M.; Sakaguchi, Y.; Hayashi, H. Bull.
Chem. Soc. J pn. 1991, 64, 1889.
(31) Mochida, K.; Kikkawa, H.; Nakadaira, Y. J . Organomet. Chem.
1991, 412, 9.
(32) Bobbitt, K. L.; Maloney, V. M.; Gaspar, P. P. Organometallics
1991, 10, 2772.
(33) Pannell, K. H.; Sharpe, L. A. Organometallics 1991, 10, 1655.
(34) Sakurai, H.; Sakamoto, K.; Kira, M. Chem. Lett. 1984, 1379.
(35) Ando, W.; Itoh, H.; Tsumuraya, T. Organometallics 1989, 8,
2759.

3234 Organometallics, Vol. 19, No. 16, 2000
Leigh et al.
determined to be on the order of 65-75% after ca. 50%
conversion of starting material in all three cases.
Compounds 11b and 11c, which were isolated, ex-
hibited spectroscopic data similar to those reported
previously for 11a 19 and the DMB adducts of 10b,c.31,32
A pure sample of 11d could not be isolated in sufficient
quantities for rigorous characterization and so was
identified on the basis of comparisons of its GC retention
time to those of 11a -c and its mass spectrum (obtained
by GC/MS) to that of 11c. The mass spectra of both 11c
and 11d exhibit parent ions at M-83 due to loss of C6H11
from the molecular ion, a fragmentation pathway that
is relatively minor in compounds 11a ,b. The product
mixtures also contained four isomeric compounds with
similar but slightly lower GC retention times than the
precursors (7); these are collectively identified in eqs
1-3 as 15Ge (from 7b,d ) and 15Si (from 7c). They were
tentatively identified on the basis of their mass spectra
as dimers of the carbon-centered free radicals (17)
derived from addition of Me3M• to the diene, as shown
in eq 4. The mass spectra of the four (C9H19Si)2 isomers
Resu lts a n d Discu ssion
Stea d y Sta te P h otolysis a n d Tr a p p in g of Rea c-
tive In ter m ed ia tes. Photolysis of deoxygenated cyclo-
hexane solutions of 7a (0.05 M) in the presence of 2,3-
dimethyl-1,3-butadiene (DMB) affords the silatriene-
DMB adduct 11a in ∼70% yield, along with minor
amounts of an unidentified isomeric adduct.19 In con-
trast, photolysis of 7b-d under similar conditions in
each case afforded complex mixtures of more than a
dozen products; those that we have been able to identify,
either by GC co-injection with authentic samples, after
isolation by semipreparative GC, or by GC/MS on the
basis of comparisons to literature data, are shown in
eqs 1-3. The product yields listed were determined by
(15Si) are identical and exhibit a weak molecular ion
(M+ ) 310), prominent (M/2)+ ions, and a (M+ ) 73)
parent ion due to Me3Si. The spectra of the germanium
analogues (15Ge) are also identical to one another and
contain the M+ ) 119 (Me3Ge) parent ion, but show only
the (M/2)+ ion in the higher mass range. The relative
yields of the four isomeric products were the same from
7b and 7d . The corresponding products from trapping
of Ph3M• radicals were not detected in the photolysates,
not unexpectedly considering their very high molecular
weights.
The mechanism for formation of Ph3M-H (14Si and
14Ge) is not clear. The mass spectra of these compounds
formed in the photolyses of 7b-d in C6H12 and C6D12
indicated that no deuterium was incorporated in either
of these products in the deuterated solvent, indicating
that they are not formed from reaction of Ph3M• radicals
with the solvent. Furthermore, no evidence could be
obtained for the presence of products that might be
consistent with disproportionation of Ph3M• and 17 in
any of the three cases. Since these products would have
the same molecular formula as those derived from
trapping of Me2MdCH2 by the diene (C9H18M), we can
similarly rule out the possibility that 14 is formed by
either Ph3M•/Me3M• disproportionation or concerted
dehydrometalation. At an early point in the study,
photolyses of 7b-d were carried out in hexane contain-
ing n-decanol (0.1 M), to probe for the formation of
Me2MdCH2 by trapping with the alcohol. The expected
alkoxymetallanes (18) are indeed formed under these
conditions, but are accompanied by increased yields of
14 in all three cases. Thus, the most likely mechanism
for formation of 18 in these experiments is direct
reaction of the alcohol with the excited singlet states of
7b-d , a reaction that is particularly well established
to occur with aryldisilanes.26,40 We have not explored
the point further.
GC/MS after 30-50% conversion of starting material,
as simple percentages of the total detectable products,
and are uncorrected for differences in detector response
factors.39 This method underestimates the yields of the
relatively low molecular weight products (13Me and 16)
because they elute much earlier than the starting
1
material and other products. This was verified by H
NMR spectroscopy on crude photolysis mixtures in
C6D12 solution, which showed the yields of 12 to be
similar to those of 13 in the photolyses of 7b,d ; the same
was found to be true of the relative yields of 13 and 16
in the photolyses of 7c,d . The three photolysates also
contained several other minor products, each of which
were formed in yields of less than ∼3%. The material
balances (starting material + identified products) were
(36) Mochida, K.; Yoneda, I.; Wakasa, M. J . Organomet. Chem. 1990,
399, 53.
(37) Mochida, K.; Kanno, N.; Kato, R.; Kotani, M.; Yamauchi, S.;
Wakasa, M.; Hayashi, H. J . Organomet. Chem. 1991, 415, 191.
(38) Konieczny, S.; J acobs, S. J .; Braddock Wilking, J . K.; Gaspar,
P. P. J . Organomet. Chem. 1988, 341, C17-C22.
(39) Product yields are listed as simple percentages of the total
photolyzed amount of 7, although the majority of them contained only
one of the Ph 3M or Me3M′ groups of the precursors. This was taken
into account for determination of material balances.
(40) Steinmetz, M. G.; Yu, C.; Li, L. J . Am. Chem. Soc. 1994, 116,
6, 932.

Group 14 1,1,1-Trimethyl-2,2,2-triphenyldimetallanes
Organometallics, Vol. 19, No. 16, 2000 3235
basis of comparisons to the published spectra41,42 and
the fact that they are quenched by chloroform with rate
constants on the order of kq ≈ (1.7 ( 0.2) × 108 M-1
s-1, which is also in excellent agreement with literature
data.41 The published spectra of these radicals also show
evidence for a weaker absorption band centered at ∼450
nm,41 which is obscured in the spectra of 7b,c by the
long wavelength bands due to the second transient
species. However, the lifetime of the latter can be
shortened to <20 ns upon addition of methanol to the
solution (vide infra), and under these conditions spectra
that are in excellent agreement with the literature
spectra of these radicals are obtained.
Photolysis of 7b-d (0.05 M) in deoxygenated cyclo-
hexane-d12 solution containing chloroform (0.07 M)
afforded the trimethylchlorometallanes 19M (30-40%),
the coproducts of germylene extrusion (12Si from 7b,
16Si from 7c, and 12Ge + 16Ge from 7d ; 10-15%),
1,1,2,2-tetrachloroethane (20), and dichloromethane
(21), according to 1H NMR analysis of the product
mixtures between 10 and 50% conversion. Qualitative
GC analysis of the product mixtures verified that the
corresponding triphenylchlorometallanes 22M were also
formed, along with a number of other, more minor
products which were not identified (see eq 5). The
Addition of trans-1,3-octadiene to the solutions of
7b-d caused reductions in the initial ∆OD’s at 330 nm,
indicating that free radical formation is suppressed
upon addition of the diene. This is consistent with the
process being derived from the triplet states of 7b-d ,
as suggested above. On the other hand, the initial ∆OD’s
at 480-490 nm were unaffected by the addition of diene
(beyond the slight reductions expected due to screening
of the excitation light by the diene), indicating these
products to be singlet-derived. Quenching of radical
formation by the diene was investigated quantitatively
in the case of 7c, by monitoring the ∆OD330 as a function
of diene concentration. To correct for screening of the
excitation light by the diene and minor variations in
laser intensity, the ∆OD330 value at each diene concen-
tration was normalized against the ∆OD at 490 nm; the
resulting ∆ OD330/∆ OD490 ratios are thus proportional
to the quantum yield for radical formation (relative to
that for formation of the 490 nm species) at each diene
concentration. A plot of the data according to the Stern-
Volmer expression of eq 6 is shown in Figure 2. The
slope is the product of the lifetime and rate constant
for quenching by diene of the triplet state of the
germasilane (kqτT ) 21.4 ( 1.3 M-1) and affords an
estimate of 2.7 ( 0.3 ns for the lifetime of the triplet
formation of 19M and 22M is ascribed to reaction of
Ph3M• and Me3M′• radicals with the halocarbon. The
resulting dichloromethyl radicals then dimerize to yield
20 or abstract hydrogen (most likely from the triphen-
ylmetallanes (14), which could not be detected in the
photolysates by either GC or NMR) to yield CH2Cl2 (21).
The much higher relative yields of radical-derived
products in these experiments, compared to those in the
presence of 0.5 M DMB, provide good evidence that they
are formed predominantly (if not completely; vide infra)
via the triplet states of 7b-d .
These results suggest that the lowest excited singlet
states of the three molecules react via two competing
pathways to form reactive intermediates which are
subsequently trapped by DMB: [1,3]-SiMe3 or [1,3]-
GeMe3 migration to yield the (1-metalla)hexatriene
species 8b-d (trapped by DMB as 11) and extrusion of
dimethyl- or diphenylgermylene (trapped by DMB as
13) with the concomitant formation of Ph3MCH3 (12M)
or PhM′(CH3)3 (16M). Germylene extrusion is the domi-
nant excited singlet state reaction pathway in each case
and is particularly prominent in the case of 7d . This
process does not occur in significant yields upon pho-
tolysis of the disilane analogue 7a .17,19,26 Free radical
formation occurs in much higher yields from 7b-d than
it does from 7a ,13 perhaps indicating that intersystem
crossing is enhanced in the germanium-containing
compounds compared to the disilane.
Dir ect Detection a n d Id en tifica tion of Rea ctive
In ter m ed ia tes by La ser F la sh P h otolysis. Laser
flash photolysis experiments employed continuously
flowing, deoxygenated hexane solutions of 7b-d (5 ×
10-4 M) and a KrF excimer laser as excitation source
(248 nm, ∼20 ns pulses, ∼50 mJ ). Transient absorption
spectra recorded 0.2-1.0 µs after the laser pulse are
shown in Figure 1 for 7b and 7c, along with typical
transient decay traces recorded at short (320-300 nm)
and long (480-490 nm) wavelengths. In both cases, the
spectra consist of absorptions due mainly to two tran-
sient species. The 330 nm wavelength transient absorp-
tions are readily identifiable as due to triphenylsilyl-
(from 7b) and triphenylgermyl (from 7c) radicals, on the
state of 7c, assuming a value of kq ) 8 × 109 M-1 s-1 43
.
This compares favorably with our previously reported
estimate of τT ) 6 ( 3 ns for the triplet lifetime of 7a ,
determined under similar conditions.17
(Φ0/Φ)Ph3Ge• ) (∆OD330/∆OD490)0/(∆OD330/∆OD490) )
1 + kqτT[diene] (6)
The value of the Stern-Volmer constant determined
above for 7c is large enough that under the conditions
of our steady state photolysis experiments in the pres-
ence of 0.5 M DMB (vide supra), the triplet state of this
compound should be quenched with more than 90%
efficiency. The fact that significant yields of radical-
derived products are still observed thus indicates that
radical formation is probably both singlet- and triplet-
derived in 7b-d . While photolysis of the disilyl deriva-
tive (7a ) leads to Si-Si bond homolysis only in polar
solvents via reaction of the lowest triplet state,17,19 we
really cannot rule out the possibility that some portion
of the reaction may be derived from the lowest excited
singlet state in this case as well.
(41) Chatgilialoglu, C.; Ingold, K. U.; Lusztyk, J .; Nazran, A. S.;
Scaiano, J . C. Organometallics 1983, 2, 1332.
(42) Chatgilialoglu, C. In CRC handbook of photochemistry, Vol II;
Scaiano, J . C., Ed.; CRC Press: Boca Raton, FL, 1989; pp 3-11.
(43) Murov, S. L.; Carmichael, I.; Hug, G. L. Handbook of photo-
chemistry; Dekker: New York, 1993.

3236 Organometallics, Vol. 19, No. 16, 2000
Leigh et al.
F igu r e 1. Time-resolved UV absorption spectra recorded with deoxygenated 5.0 × 10-4 M solutions of Ph3SiGeMe3 (7b,
a) and Ph3GeSiMe3 (7c, b) in hexane at 23 °C, 0.2-1.0 µs after 248 nm pulsed laser excitation. The insets show decay
traces recorded at monitoring wavelengths of 330 (- - -) and 490 nm (s).
DMB, and carbon tetrachloride affected the lifetime of
the triphenylsilyl radical absorption at 330 nm as well
as that of the longer wavelength transient. In contrast,
addition of chloroform to the solution had no effect on
the lifetime of the long wavelength transient absorption
within the concentration range examined ([CHCl3] <
0.02 M), but quenched that of the radical efficiently (vide
supra).
The lifetime of the 480 nm transient absorption from
photolysis of Ph3GeSiMe3 (7c) was also quenched upon
addition of DMB, HOAc, acetone, carbon tetrachloride,
or oxygen, but with the exception of the halocarbon,
quenching was much less efficient than with the ana-
logous transient from 7b. The decay rate constants
varied in direct proportion to concentration of added
reagent in each case, as is illustrated by the quenching
plots shown in Figure 4 for reaction of the transient with
acetic acid, DMB, and acetone. Again, addition of up to
∼0.06 M chloroform had no effect on the lifetime of the
long wavelength absorption, but quenched the radical
absorption at shorter wavelengths. Addition of up to 1.7
M MeOH had no discernible effect on the lifetimes at
either monitoring wavelength.
F igu r e 2. Stern-Volmer plot of the ratios of the ∆ODs
at 330 and 490 nm (Φ0/Φ; see eq 6), from photolysis of 7b
in the presence of 1,3-octadiene (0-0.02 M).
Addition of methanol (MeOH), DMB, acetone, acetic
acid, carbon tetrachloride, or oxygen to the solutions of
Ph3SiGeMe3 (7b) resulted in an increase in the pseudo-
first-order rate constant for decay of the long-wave-
length transient absorption. Plots of kdecay versus con-
centration of added reagent were curved for methanol
(Figure 3a), but fit acceptably to the second-order
The transient absorption spectrum recorded by flash
photolysis of a deoxygenated hexane solution of
Ph3GeGeMe3 (7d ) showed similarities to those from the
other two compounds (Figure 5a), but the long wave-
length absorption band was significantly less intense
than the short wavelength absorption at 330 nm. Still,
transient decays recorded at the two wavelengths
showed marked differences in lifetime, indicating that
two different transient species are being observed. As
with 7c, addition of 0.065 M chloroform to the solution
shortens the lifetime of the 330 nm transient to <25 ns
(consistent with its assignment to triphenylgermyl
radicals) but has little effect on that of the 480 nm
transient (τ ≈ 2.5 µs). This allows isolation of the
spectrum due to the long-wavelength-absorbing species,
as is shown in Figure 5b. Addition of DMB, carbon
tetrachloride, or oxygen to the chloroform/hexane solu-
tions of 7d resulted in shortening of the lifetime of this
transient, and the transient decay rate varied linearly
with concentration. Quenching by HOAc, acetone, and
MeOH was not investigated in this case.
0
polynomial expression shown in eq 7, where kd is the
0
kdecay ) kd + kMeOH[MeOH] + k2MeOH[MeOH]2 (7)
decay rate constant in the absence of methanol, and
kMeOH and k2MeOH are the overall second- and third-order
rate constants for reaction of the transient with the
alcohol. Similar plots for the other reagents were linear
in concentration and were thus analyzed according to
eq 8. Representative plots of kdecay vs concentration for
0
kdecay ) kd + kq[Q]
(8)
the quenching of the long wavelength transient from
7b by MeOH, acetic acid, DMB, and acetone are shown
in Figure 3. Of the six reagents examined, only oxygen,

Group 14 1,1,1-Trimethyl-2,2,2-triphenyldimetallanes
Organometallics, Vol. 19, No. 16, 2000 3237
F igu r e 3. Stern-Volmer plots for quenching of silene 8b by (a) methanol and (b) acetic acid (0), 2,3-dimethyl-1,3-butadiene
(O), and acetone (4) in deoxygenated hexane solution at 23 °C. The solid lines represent the best fits of the data to eqs 7
(a) and 8 (b).
assignment of this transient to Me2Ge:, which our
product studies show is formed in roughly twice the
yield of 8b (eq 1), can be ruled out on the basis of sev-
eral arguments. First, dimethylgermylene has been
reported to exhibit an absorption maximum in the 420-
450 nm range in hydrocarbon solvents at room temper-
ature,32,36,37,44 significantly lower than the absorption
maximum of the transient obtained from germasilane
7b. Second, the transient shows the same high reactivity
toward methanol (and the other five reagents) as 8a ;
dimethylgermylene is known to be relatively unreactive
toward alcohols (kROH , 104 M-1 s-1).37,44 Third, the
transient shows no detectable reactivity toward trieth-
ylsilane and chloroform, reagents which are known or
can be expected to react rapidly with Me2Ge:.37,45
Finally, the extinction coefficient of the long wavelength
(π,π*) absorption of the 1-silahexatriene (8b) can be
F igu r e 4. Stern-Volmer plots for quenching of germene
8c by acetic acid (0), 2,3-dimethyl-1,3-butadiene (O), and
acetone (4) in deoxygenated hexane solution at 23 °C. The
solid lines represent the best fits of the data to eq 8.
expected to be at least an order of magnitude larger than
that of the (n,p) absorption of dimethylgermylene; this
leads to the expectation that the absorption due to 8b
should be at least 5 times greater than for the ger-
mylene, given they are formed in relative yields of 1:2.46
Table 1 summarizes the absolute rate constants for
quenching of the 480-490 nm transient absorptions
from flash photolysis of 7b-d by the six quenchers
studied, along with the analogous data for quenching
of the transient 1-silahexatriene (8a ) derived from 7a
in the same solvent.19,26 Addition of up to 4 M trieth-
ylsilane (HSiEt3) to the solutions of 7b,c had no effect
on the lifetimes of either the radicals or the long-
wavelength-absorbing transients; upper limits for the
rate constants for quenching of the 480 nm transients
from these two compounds, calculated from the lifetimes
in the absence and presence of 4 M silane, are included
in Table 1 along with analogous data for 8a .
The 490 nm transient obtained from 7b exhibits an
absorption spectrum nearly identical to that of 8a and
strikingly similar rate constants for reaction with all
six of the quenchers examined, suggesting it should be
assigned to the 7-trimethylgermyl-substituted 1-sila-
hexatriene derivative 8b. The similarities in the UV
spectra and reactivity of 8a and 8b are quite reasonable,
given that the only structural difference between the
two molecules is the Me3M substituent at the 7-position
of the 1-metallahexatrienyl moiety. The alternative
Similarly, we assign the 480 nm transient products
observed by flash photolysis of 7c,d to the 1-germa-
hexatrienes 8c,d . These two compounds can be expected
to exhibit absorption maxima and extinction coefficients
similar to the silicon analogues 8a ,b, since 1,1-diphen-
ylsilene (23a ) and 1,1-diphenylgermene (23b) exhibit
nearly identical UV absorption spectra in hydrocarbon
solvents (λmax ) 325 nm,28,29 ꢀmax ) 7500-11 000 M-1
cm-147). The transient assigned as 8c also exhibits the
expected decrease in reactivity compared to 8a ,b toward
the nucleophilic quenchers MeOH, HOAc, and acetone,
judging from a comparison of the absolute reactivities
of 23a and 23b toward these same reagents.28,29 While
(44) Mochida, K.; Tokura, S. Bull. Chem. Soc. J pn. 1992, 65, 1642.
(45) Mochida, K.; Hasegawa, A. Chem. Lett. 1989, 1087.
(46) The extinction coefficient at the absorption maximum for
dimesitylgermylene is approximately 15 times less than that of
tetramesityldigermene, on the basis of experiments in which the two
species are generated together in a 1:1 ratio. See: Toltl, N. P.; Leigh,
W. J .; Kolleger, G. M.; Stibbs, W. G.; Baines, K. M. Organometallics
1996, 15, 3732. The extinction coefficient of tetramesityldigermene at
its absorption maximum is ca. 20 000 M-1 cm-1 (Baines, K. M. private
communication). Thus, this allows an estimate of 800 ( 400 M-1 cm-1
for the extinction coefficient of dimesitylgermylene.

3238 Organometallics, Vol. 19, No. 16, 2000
Leigh et al.
F igu r e 5. Time-resolved UV absorption spectra recorded with deoxygenated 5.0 × 10-4 M solutions of Ph3GeGeMe3 (7d )
at 23 °C in (a) hexane and (b) hexane containing 0.07 M chloroform. The spectra were recorded 0.2-1.0 µs after 248 nm
pulsed laser excitation.
Ta ble 1. Absolu te Ra te Con sta n ts (in u n its of 108
unsuccessful in both cases. Flash photolysis of 7b in
hexane containing 0.1 M MeOH, to shorten the life-
M-1 s-1) for Qu en ch in g of Tr a n sien t
1-Meta lla h exa tr ien es (8a -d ) by
time of the (1-sila)hexatriene 8b to <100 ns, afforded
2,3-Dim eth yl-1,3-bu ta d ien e (DMB), Aceton e, Acetic
a transient absorption spectrum assignable to only
Acid , Meth a n ol, Oxygen , Ca r bon Tetr a ch lor id e,
Ph3Si•. The long wavelength absorption band in the
a n d Tr ieth ylsila n e in Hexa n e Solu tion a t
spectrum of the radical8,41 is expected to overlap with
that due to dimethylgermylene, yet the decay kinetics
in this monitoring wavelength range were identical to
those measured at the maximum of the short wave-
length absorption band of the radical at ∼330 nm.
Furthermore, addition of triethylsilane or n-bromohex-
ane had no effect on the lifetime at either wavelength.
Since product studies indicate that the yield of Me2Ge:
from 7b is in fact fairly high (vide supra), our failure to
detect the germylene by flash photolysis is presumably
due to its low expected extinction coefficient compared
to that of the radical. Similar results were obtained in
flash photolysis experiments with 7c in methanolic
hexane solution.
23 °Ca
Steady state photolysis of a hexane solution of 7b
containing 0.5 M DMB and up to 4 M triethylsilane led
to the formation of no new products and no reduction
in the yield of 13Me relative to that of 12Si compared to
that obtained in the absence of silane. This is surprising,
since Me2Ge: has been reported to be only ∼5 times less
reactive toward hydridosilanes than DMB on the basis
of time-resolved quenching experiments using perm-
ethylcyclopolygermane precursors (kEt3SiH ≈ 4 × 106 M-1
s-1).36,37,44 We do not understand the reasons for this,
although it should be pointed out that the published
kinetic results were not actually accompanied by sup-
porting product studies.
a
From polynomial least-squares fitting of kdecay vs [Q] data
b
according to eqs 7 or 8. Errors are quoted as (2σ. Not deter-
mined. c Addition of up to 1.7 M MeOH had no effect on the lifetime
d
in deoxygenated solution. Addition of up to 4 M HSiEt3 had no
effect on the lifetime in deoxygenated solution.
nucleophilic quenching of 8d was not investigated, the
similarities in the rate constants for quenching by DMB,
oxygen, and CCl4 to those for 8c provides reasonably
compelling support for its assignment. Again, the al-
ternative assignment of these transients to diphenyl-
or dimethylgermylene can be ruled out by their lack of
reactivity toward triethylsilane and/or chloroform; fur-
thermore, the product studies indicate that their yields
relative to those of 8c,d are too low (when the differ-
ences in extinction coefficient are factored in) for them
to dominate the observed transient absorption spectra.
It is interesting to note that the trends in the absolute
reactivity of (1-germa)hexatriene 8c toward O2, CCl4,
DMB, and MeOH are very similar to those reported by
Mochida and co-workers for Me2Ge:, using permethyl-
cyclopolygermanes as germylene precursors.37,44 We
believe that these unanticipated similarities have led
to erroneous assignments in some of the time-resolved
studies of Me2Ge: and Ph2Ge: in solution which have
been reported, in particular, those that have employed
arylgermasilanes (e.g., 9c31,32) or aryl(disila)germanes
(e.g., 2438) as germylene precursors. While steady state
Attempts to obtain direct evidence for germylene
formation from 7b,c by laser flash photolysis have been

Group 14 1,1,1-Trimethyl-2,2,2-triphenyldimetallanes
Organometallics, Vol. 19, No. 16, 2000 3239
trapping experiments clearly demonstrate that in both
cases Me2Ge: and Ph2Ge: (respectively) are the major
products of photolysis of these compounds in solution,
they also show that the corresponding (1-germa)-
hexatriene derivatives are formed as well in significant
yields. Flash photolysis of 9c and 24 was reported to
lead to the formation of transients absorbing in the
420-460 nm region, which were assigned to the ger-
mylenes on the basis of their absorption maxima, the
fact that in the absence of trapping agents they decay
with second-order kinetics (consistent with dimerization
to the corresponding digermene, whose second-order
growth was also observed), and on the basis of absolute
rate constants for reaction with reagents such as O2,
DMB, CCl4, MeOH, and Et3SiH. However, several of
these rate constants (in particular, those for reaction
with Et3SiH) are in clear disagreement with those
determined using other Me2Ge:36,37,45 and Ph2Ge:36
precursors. On the basis of what we now know of the
effects of substituents on the UV spectra and reactivity
of transient silenes and germenes of homologous struc-
ture in solution,19,28,29,47-49 information that was not
available at the time of the original studies, the data
reported are more in line with those expected for the
(1-germa)hexatriene analogues 10c and 25, respectively.
ing the rate at which it collapses to product by intramo-
lecular proton transfer relative to reverting to starting
materials.
On the other hand, the reactivities of these com-
pounds toward the diene, oxygen, and carbon tetrachlo-
ride are much less dependent on the identity of the
group 14 heteroatom than is the case with nucleophilic
quenchers. These three reagents do not react with silene
23a at measurable rates in solution (k e 106 M-1 s-1),53
from which it can be concluded that their unusually high
reactivity toward 8a ,b is due to the unique structure of
the (1-sila)hexatriene system, which appears to ac-
centuate reactivity (compared to simpler metallaethene
systems) via biradical or biradicaloid mechanisms.19
The reaction of simple silenes and germenes with
aliphatic dienes proceeds nonconcertedly, generally to
yield a mixture of ene-, [2+2]-, and [4+2]-cycloaddition
products.54-58 In contrast, the reaction of 8a with DMB
proceeds largely via ene-addition, although there is at
least one other minor product formed in yields that are
too low to allow isolation.19 The same is true of 8b-d ,
according to GC/MS analysis of the reaction mixtures.
We thus conclude that the reaction of dienes with 8 is
also nonconcerted, most likely involving the biradical
intermediate (26) shown in eq 9. The enhanced reactiv-
ity of 8 compared to simpler silenes and germenes
presumably results from the facts that the initial step
involves formation of a biradical with significantly
greater resonance stabilization and that collapse of the
biradical to the ene-adduct 11 enjoys an additional
driving force due to aromatization. In this reaction, the
germanium analogues 8c,d are roughly 10 times less
reactive than the silicon analogues 8a ,b.
Reaction of oxygen with aryldisilane-derived 1-(sila)-
hexatrienes such as 10a and 8a proceeds with bimo-
lecular rate constants on the order of 109 M-1 s-1 and
yields phenyltrimethylsilane (16Si) and silanone oligo-
mers (eq 9; M ) M′ ) Si).19 The reaction presumably
must proceed via a triplet biradical intermediate (27)
such as that shown in eq 10, although the mechanism
Com p a r ison of th e Rea ctivities of th e SidC a n d
GedC Bon d s in 8a -d . As mentioned earlier, the (1-
sila)hexatriene derivatives 8a ,b are roughly 3 orders of
magnitude more reactive than the germanium ana-
logues 8c,d toward methanol, acetic acid, and acetone
(Table 1), a difference that is also mirrored in the
simpler MdC derivatives 23a ,b.29 Both silenes and
germenes react with alcohols (and probably acetic acid
as well50) via a mechanism involving initial complex-
ation of the nucleophile at the heteroatom, followed by
proton transfer.7,26,51,52 In the case of alcohols, the
proton-transfer step occurs by two competing mecha-
nisms, one involving unimolecular migration of the
proton from oxygen to carbon and one involving cataly-
sis by a second molecule of alcohol. Kinetically, the
result of this is a mixed first- and second-order depen-
dence of the silene/germene decay rate constant on
alcohol concentration.7,26 The much lower reactivity of
germenes compared to silenes toward nucleophiles can
be explained as due primarily to the higher electrone-
gativity of germanium relative to silicon, which leads
to a less polar MdC bond.28,29 This has the effect of
slowing down the initial complexation step of these
reactions and, because the complex is weaker, of reduc-
for collapse of the biradical to the observed products is
not well understood. It is interesting to note that the
products correspond formally to those of R2Si: extrusion
(47) Toltl, N. P.; Stradiotto, M. J .; Morkin, T. L.; Leigh, W. J .
Organometallics 1999, 18, 5643.
(48) Leigh, W. J .; Boukherroub, R.; Kerst, C. J . Am. Chem. Soc.
1998, 120, 9504.
(53) Leigh, W. J .; Bradaric, C. J .; Sluggett, G. W. J . Am. Chem. Soc.
1993, 115, 5332.
(54) Barton, T. J .; Kline, E. A.; Garvey, P. M. J . Am. Chem. Soc.
1973, 95, 3078.
(49) Leigh, W. J .; Kerst, C.; Boukherroub, R.; Morkin, T. L.; J enkins,
S.; Sung, K.; Tidwell, T. T. J . Am. Chem. Soc. 1999, 121, 4744.
(50) Bradaric, C. J .; Leigh, W. J . Can. J . Chem. 1997, 75, 1393.
(51) Kira, M.; Maruyama, T.; Sakurai, H. J . Am. Chem. Soc. 1991,
113, 3986.
(52) Leigh, W. J .; Bradaric, C. J .; Kerst, C.; Banisch, J . H. Orga-
nometallics 1996, 15, 2246.
(55) Ishikawa, M.; Fuchikami, T.; Kumada, M. J . Organomet. Chem.
1978, 149, 37.
(56) Wiberg, N.; Kim, C.-K. Chem. Ber. 1986, 119, 2980.
(57) Conlin, R. T.; Bobbitt, K. L. Organometallics 1987, 6, 1406.
(58) Leigh, W. J .; Bradaric, C. J .; Sluggett, G. W.; Venneri, P.;
Conlin, R. T.; Dhurjati, M. S. K.; Ezhova, M. B. J . Organomet. Chem.
1998, 561, 19.

3240 Organometallics, Vol. 19, No. 16, 2000
Leigh et al.
from the disilane. Similar products are known to be
formed upon reaction of silenes of this type with
dimethyl sulfoxide.19,59 To verify that quenching of the
480-490 nm transients from flash photolysis of 7b and
7c with oxygen results in the same chemistry (and to
provide further evidence consistent with their assign-
ment), the two germasilanes were photolyzed in oxygen-
saturated cyclohexane solution. As expected, photolysis
of 7b under these conditions led to the formation of
phenyltrimethylgermane (16Ge, 7%), a product that is
not formed in the absence of oxygen. Similarly, photoly-
sis of 7c under these conditions led to an enhancement
in the yield of 16Si, relative to that obtained in experi-
ments in nitrogen-saturated solution.
Presumably, reaction of 8 with CCl4 also proceeds via
a radical mechanism, which would explain the similar
reactivities of 8a ,b and 8c,d , as well as the fact that
simpler silenes and germenes such as 23a ,b show much
lower reactivity toward this reagent. It seems most
likely that the reaction is initiated by electron transfer,
on the basis that quenching of 8a -d by CCl4 is at least
100 times faster than by chloroform. The mechanistic
details are extremely difficult to ascertain, because CCl4
also reacts rapidly with the excited singlet states of
aryldisilanes (including 7a ) by electron transfer, which
ultimately results in Si-Si bond cleavage and the
formation of the corresponding chlorosilanes.19 Similar
behavior can be expected for 7b-d .
coefficient of the long wavelength n,p absorption band
of the germylene can be expected to be at least an order
of magnitude lower than that of the π,π* absorption
band of the silene.46
The long wavelength transient from flash photolysis
of Ph3GeSiMe3 (7c) absorbs with λmax ≈ 480 nm and
reacts with nucleophilic quenchers (HOAc, MeOH, and
acetone) with rate constants that are on the order of
103 lower than those for reaction of 8a ,b with the same
reagents. Much higher reactivity is observed toward O2,
CCl4, and dienes, with absolute rate constants that are
only 2-10 times lower than the corresponding values
for reaction of 8a ,b. The transient has been assigned to
the (1-germa)hexatriene 8c on the basis of these data.
Again, an alternative assignment to Ph2Ge:, which
product studies suggest is formed in roughly twice the
yield of 8c, can be ruled out on the basis of the extinction
coefficient arguments discussed earlier, its low reactivity
toward triethylsilane and chloroform, and the fact that
Ph2Ge: has been reported to absorb at somewhat
shorter wavelengths (λmax 445-460 nm30,35,36,38,60) than
the transient observed from 8c. A transient with the
same absorption maximum and identical reactivity
toward O2, CCl4, and DMB is observed in the case of
the digermane 7d and is identified as (1-germa)-
hexatriene 8d .
The 1000-fold lower reactivity of 8c,d compared to
8a ,b toward oxygen-containing nucleophilic trapping
agents such as alcohols, carboxylic acids, and ketones
parallels the differences reported previously for 1,1-
diphenylsilene (23a ) and 1,1-diphenylgermene (23b).28,29
Much smaller differences in reaction rate are observed
for DMB, O2, and CCl4, reagents that are normally
relatively unreactive toward either silenes or germenes
of simpler structure (e.g., 23a ,b). This unusually high,
relatively indiscriminate reactivity is presumably due
to the presence of the cyclohexadienyl substituent at the
metallaene carbon atoms in 8a -d , which would clearly
be expected to promote reactions that proceed through
radical or biradical intermediates.
Su m m a r y a n d Con clu sion s
The direct irradiation in solution of arylgermasilanes
and -digermanes of the general structure Ph3MM′Me3
(7b-d ) leads mainly to M-M′ bond homolysis, at least
partially via the lowest excited triplet state which is
produced by intersystem crossing from the excited
singlet state manifold in competion with the formation
of a mixture of germylene, silene, and germene reactive
intermediates. The latter are formed in similar yields
and can be trapped by aliphatic dienes. This same
general behavior has been reported previously for other
arylgermasilanes and -digermanes,30-32 as well as aryl-
disilagermanes38 and aryltrigermanes.36,45 As with the
disilane analogue 7a , flash photolysis of 7b-d results
in the formation of strong transient absorptions due to
arylsilyl or arylgermyl radicals (λmax ) 330-335 nm),
superimposed on absorptions due to a second transient
Exp er im en ta l Section
1H and 13C NMR spectra were recorded on Bruker AC200
or DRX500 NMR spectrometers in deuterated chloroform
solution and are referenced to tetramethylsilane. Ultraviolet
absorption spectra were recorded on Hewlett-Packard HP8451
or Perkin-Elmer Lambda 9 spectrometers. Low-resolution
mass spectra and GC/MS analyses were determined using a
Hewlett-Packard 5890 gas chromatograph equipped with a
HP-5971A mass selective detector and a DB-5 fused silica
capillary column (30m × 0.25 mm; Chromatographic Special-
ties, Inc.). High-resolution desorption electron impact (DEI)
mass spectra and exact masses were recorded on a VGH ZABE
mass spectrometer. Exact masses employed a mass of 12.000000
for carbon. Infrared spectra were recorded on a BioRad FTS-
40 FTIR spectrometer and are reported in wavenumbers
(cm-1).
Analytical gas chromatographic analyses were carried out
by GC/MS or using a Hewlett-Packard 5890 gas chromato-
graph equipped with a flame ionization detector, a Hewlett-
Packard 3396A recording integrator, conventional heated
splitless injector, and a DB-1 fused silica capillary column (15
m × 0.20 mm; Chromatographic Specialties, Inc.). Semi-
preparative GC separations employed a Varian 3300 gas
chromatograph equipped with a thermal conductivity detector
and a 6′ x 0.25′′ stainless steel OV-101 packed column
species which absorbs at longer wavelengths (λmax
480-490 nm).
)
In the case of Ph3SiGeMe3 (7b), this transient reacts
with MeOH, AcOH, acetone, DMB, O2, and CCl4 in
hydrocarbon solution with rate constants nearly identi-
cal to that of the (1-sila)hexatriene (8a ) observed under
similar conditions from Ph3SiSiMe3 (7a ) and can thus
be identified as the (1-sila)hexatriene derivative 8b.
Neither the absorption maximum nor the reactivity of
the transient is consistent with a possible alternative
assignment to Me2Ge:, on the basis of the various
studies of this transient germylene that have been
previously reported. This is not unexpected, since
product studies indicate that the germylene is formed
in only ca. twice the yield of 8b, and the extinction
(59) Soysa, H. S. D.; Weber, W. P. J . Organomet. Chem. 1979, 173,
269.

Group 14 1,1,1-Trimethyl-2,2,2-triphenyldimetallanes
Organometallics, Vol. 19, No. 16, 2000 3241
(Chromatographic Specialties). Radial chromatographic sepa-
rations employed a Chromatotron (Harrison Research, Inc.),
2- or 4-mm silica gel 60 thick-layer plates, and hexane/ethyl
acetate mixtures as eluant.
Hexane (BDH Omnisolv) was used as received without
additional purification or drying. Methanol, 2,3-dimethyl-1,3-
butadiene, acetone (HPLC grade), acetic acid, chloroform, and
carbon tetrachloride were used as received from Aldrich
Chemical Co. Deuterated materials were used as received from
Cambridge Isotope Labs. Compounds 7b-d were prepared and
purified as previously reported.61
For the synthesis of 18Si, a solution of trimethylchlorosilane
(4 g, 36.9 mmol) in anhydrous ether was added dropwise to a
stirred solution of triethylamine (7.5 g, 73.8 mmol) and
n-decanol (5.8 g, 36.9 mmol) in anhydrous ether (10 mL), which
were contained in a 250 mL round-bottom flask fitted with
magnetic stirrer, condenser, dropping funnel, nitrogen inlet,
and ice bath. A colorless precipitate of triethylamine hydro-
chloride was formed immediately. The mixture was stirred for
1 h and then filtered, and the solvent was removed under
vacuum to afford a slightly yellow liquid. Atmospheric distil-
lation of the material afforded 1-trimethylsiloxydecane (18Si
)
Solutions of 7b-d (0.1 g, 3.66 × 10-4 mol) in hexane (15
mL) containing 2,3-dimethyl-1,3-butadiene (0.5 M) were placed
in quartz tubes, which were then sealed with rubber septa and
deoxygenated with a stream of dry argon. The samples were
then photolyzed for 1 h in a Rayonet reactor equipped with a
merry-go-round and six RPR2537 lamps, with periodic moni-
toring of the course of the photolysis by GC. Methyltriphenyl-
silane (12Si), triphenylsilane (14Si), trimethylphenylsilane
(16Si), methyltriphenylgermane (12Ge), triphenylgermane (14Ge),
and trimethylphenylgermane (16Ge) were identified by GC and/
or GC/MS co-injection with authentic samples obtained from
Aldrich Inc. or synthesized by known routes. The germylene-
diene adducts 13Me and 13P h were isolated by semipreparative
GC, from the photolysis mixtures from 7b and 7c, respectively,
after evaporation of the solvent and dissolution of the resulting
colorless oils in a small amount of pentane. The two compounds
were identified by comparison of their 1H NMR and mass
spectra to previously reported data.32 The (1-metalla)hexatriene
adducts 11b,c were isolated by semipreparative GC and
identified on the basis of the spectroscopic data listed below.
Adduct 11d (from 7d ) was identified on the basis of GC
retention time and mass spectral comparisons to 11a -c.
1-(Tr im et h ylger m yl)-2-((2,3-d im et h yl-1-b u t en -4-yl)-
as a colorless liquid (bp 234-240 °C; 2.87 g, 12.5 mmol, 34%).
Compound 18Ge was synthesized in similar fashion from
trimethylgermanium chloride (4.1 g, 26.2 mmol), n-decanol (4.1
g, 26.2 mmol), and triethylamine (5.3 g, 52.4 mmol). Distilla-
tion under reduced pressure afforded 1-trimethylgermoxyde-
cane (18Ge) as a colorless liquid (bp 130 °C (20 mm); 1.24 g,
4.5 mmol, 19%). The two compounds exhibited the following
NMR and mass spectral data.
1-Tr im eth ylsiloxyd eca n e (18Si). 1H NMR: δ 0.11 (s, 9H),
0.88 (t, 3H, J ) 7.0 Hz), 1.27 (bd s, 14H), 1.53 (t, 2H, J ) 6.5
Hz), 3.57 (t, 2H, J ) 6.7 Hz). 13C NMR: δ -0.13, 14.5, 23.1,
26.2, 29.7, 29.8, 30.0, 30.0, 32.3, 33.1, 63.1. 29Si NMR: δ 17.1.
MS (EI): m/z (I) 230 (5), 215 (100), 199 (5), 187 (5), 157 (10),
143 (10), 129 (10), 115 (20), 103 (30), 75 (75), 73 (50).
1
1-Tr im eth ylger m oxyd eca n e (18Ge). H NMR: δ 0.39 (s,
9H), 0.89 (3H, t, J ) 6.9 Hz), 1.27 (bd s,14H), 1.53 (t, 2H, J )
6.9 Hz), 3.55 (2H, t, J ) 6.9 Hz). 13C NMR: δ 0.48, 14.5, 23.1,
26.4, 29.7, 29.8, 30.0, 30.1, 33.2, 34.0, 64.9. MS (EI): m/z (I)
276 (5), 261 (10), 231 (5), 203 (5), 189 (5), 173 (7), 149 (75),
119 (100), 105 (30), 89 (30), 75 (5).
Nanosecond laser flash photolysis experiments employed the
pulses (248 nm; ca. 16 ns; ca. 50 mJ ) from a Lambda-Physik
Compex 120 excimer laser filled with F2/Kr/He mixtures and
a microcomputer-controlled detection system.62 Solutions were
prepared at concentrations such that the absorbance at the
excitation wavelength (248 nm) was ca. 0.7 (4.4 × 10-3 M) and
were flowed continuously through a 3 × 7 mm Suprasil flow
cell connected to a calibrated 100 mL reservoir. Solution
temperatures were measured with a Teflon-coated copper/
constantan thermocouple, which was inserted directly into the
flow cell, and are accurate to within 0.1 °C. Quenchers were
added directly to the reservoir by microliter syringe as aliquots
of standard solutions. Rate constants were calculated by linear
least-squares analysis of decay rate-concentration data (6-
15 points, depending on whether the plots were linear or
curved), which spanned at least 1 order of magnitude in the
transient decay rate. Errors are quoted as twice the standard
deviation obtained from the least-squares analysis in each
case.
1
d ip h en ylsilyl)ben zen e (11b). H NMR (CD2Cl2): δ 0.06 (s,
9H), 0.84 (d, 3H), 1.46 (m, 2H), 1.63 (s, 3H), 2.48 (m, 1H), 4.54
(s, 1H), 4.58 (s, 1H), 7.21-7.66 (m, 14H). 13C NMR (CD2Cl2;
quaternary carbons not resolved): δ 2.12, 19.07, 22.23, 22.85,
37.90, 108.74, 127.47, 128.28, 129.09, 129.73, 134.47, 135.96,
136.41. MS (EI): m/z (I) 460 (1), 445 (4), 377 (20), 367 (15),
341 (15), 299 (15), 259 (50), 222 (100), 197 (50), 181 (50), 135
(25), 105 (25). Exact mass: calcd for (M - 83) C21H2370GeSi,
373.0811; found 373.0795.
1-(Tr im e t h ylsilyl)-2-((2,3-d im e t h yl-1-b u t e n -4-yl)d i-
p h en ylger m yl)ben zen e (11c). 1H NMR: δ 0.02 (s, 9H), 0.92
(d, 3H), 1.55 (m, 2H), 1.64 (s, 3H), 4.56 (m, 2H), 7.32-7.51
(m, 14H). MS: m/z (I) 460 (5), 445 (1), 377 (100), 362 (5), 347
(5), 299 (30), 243 (20), 227 (30), 195 (20), 181 (20), 151 (40),
135 (35), 105 (10), 73 (20). Exact mass: calcd for (M - 83)
C
21H2370GeSi, 373.0811; found 373.0793.
1-(Tr im et h ylger m yl)-2-((2,3-d im et h yl-1-b u t en -4-yl)-
d ip h en ylger m yl)ben zen e (11d ). MS (EI): m/z (I) 489 (1),
421 (100), 407 (10), 389 (10), 343 (25), 310 (20), 268 (20), 243,
(50), 227 (50), 181 (50), 151 (80), 119 (25), 91 (20).
Ack n ow led gm en t. We wish to thank D. Hughes for
his assistance with high-resolution NMR spectra, the
McMaster Regional Centre for Mass Spectrometry for
high-resolution mass spectra and exact mass determi-
nations, and Professors Kunio Mochida and Peter Gas-
par for helpful comments. The financial support of the
Natural Sciences and Engineering Research Council
(NSERC) of Canada (to W.J .L.) and the Robert Welch
Foundation (to K.H.P.) is also gratefully acknowledged.
Solutions of 7b-d (0.1 g, 3.66 × 10-4 mol) in hexane (15
mL) containing n-decanol (0.1 M) were photolyzed as described
above and monitored by GC, which revealed the formation of
n-decyloxytrimethylsilane (18Si) or n-decyloxytrimethylger-
mane (18Ge), along with other products. Compounds 18 were
identified by GC/MS and by GC co-injection with authentic
samples, which were synthesized under a nitrogen atmosphere
using dry oxygen-free solvents, as follows.
OM000174P
(60) Ando, W.; Tsumuraya, T.; Sekiguchi, A. Chem. Lett. 1987, 317.
(61) Pannell, K. H.; Kapoor, R. N.; Raptis, R.; Parkanyi, L.; Fulop,
V. J . Organomet. Chem. 1990, 384, 41.
(62) Leigh, W. J .; Workentin, M. S.; Andrew, D. J . Photochem.
Photobiol. A: Chem. 1991, 57, 97.