Direct detection of i,i-diphenylgermene in solution and absolute rate constants for germene trapping reactions

Article abstract of DOI:10.1021/ja973756f

Direct irradiation of 1,1-diphenylgermetane in hexane solution affords 1,1,3,3-tetraphenyl-1,3 digermetane in high chemical yield. Photolysis in the presence of aliphatic alcohols leads instead to the formation of the corresponding alkoxymethyldiphenylgermane. These results are consistent with the formation of 1.1-diphenylgermene as a primary photochemical product from photolysis of the germetane. Nanosecond laser flash photolysis of the compound in hexane, acetonitrile, or tetrahydrofuran gives rise to the formation of a transient, assignable to the germene on the basis of its second order decay kinetics, UV spectrum (λ(max) = 325 nm), and the fact that it is quenched by addition of alcohols and acetic acid. Absolute rate constants for reaction of 1.1 diphenylgermene with methanol, ethanol, 2-propanol, tert-butyl alcohol, acetic acid, the O-deuterated isotopomers, and acetone were determined in the three solvents, using the germetane as the O-deuterated isotopomers, and acetone were determined in the three solvents, using the germetane as the precursor. The kinetics and mechanisms of these germene trapping reactions are discussed and compared to those of silenes.

Full text of DOI:10.1021/ja973756f

1172
J. Am. Chem. Soc. 1998, 120, 1172-1178
Direct Detection of 1,1-Diphenylgermene in Solution and Absolute
Rate Constants for Germene Trapping Reactions
Nicholas P. Toltl and William J. Leigh*
Contribution from the Department of Chemistry, McMaster UniVersity, 1280 Main Street West,
Hamilton, Ontario, Canada L8S 4M1
ReceiVed October 30, 1997
Abstract: Direct irradiation of 1,1-diphenylgermetane in hexane solution affords 1,1,3,3-tetraphenyl-1,3-
digermetane in high chemical yield. Photolysis in the presence of aliphatic alcohols leads instead to the formation
of the corresponding alkoxymethyldiphenylgermane. These results are consistent with the formation of 1,1-
diphenylgermene as a primary photochemical product from photolysis of the germetane. Nanosecond laser
flash photolysis of the compound in hexane, acetonitrile, or tetrahydrofuran gives rise to the formation of a
transient, assignable to the germene on the basis of its second-order decay kinetics, UV spectrum (λmax ) 325
nm), and the fact that it is quenched by addition of alcohols and acetic acid. Absolute rate constants for
reaction of 1,1-diphenylgermene with methanol, ethanol, 2-propanol, tert-butyl alcohol, acetic acid, the
O-deuterated isotopomers, and acetone were determined in the three solvents, using the germetane as the
precursor. The kinetics and mechanisms of these germene trapping reactions are discussed and compared to
those of silenes.
Introduction
diphenylgermene (2; eq 1) and ethylene, via [2 + 2]-cyclo-
The chemistry of stable organometallic compounds containing
a GedC double bond (germenes) has been extensively
investigated.1-4 Like its organosilicon homolog, the GedC
bond is of low intrinsic thermodynamic stability, so that stable
germene derivatives all contain sterically bulky substituents at
both germanium and carbon. Derivatives bearing less bulky
substituents are apparently highly reactive, and have been
suggested as reactive intermediates in a number of thermal and
photochemical reactions in organogermanium chemistry, on the
basis of the products observed in the presence of various
reagents included as germene traps (water, alcohols, dienes,
carbonyl compounds, etc.).3 Little direct evidence for their
involvement as reactive intermediates has been reported,
however, and the quantitative aspects of germene reactivity are
largely unknown. Our recent experiences with the photochemi-
cal generation and direct detection of reactive silenes in
solution5-7 suggested that similar techniques might be used for
the generation of reactive germenes in conditions under which
they might be detected directly and the kinetics of their
bimolecular reactions studied.
hν
+ C2H4
Ge CH2
(1)
Ge
2
1
reversion in the lowest excited singlet state. NLFP techniques
have allowed determination of the UV absorption spectrum of
the transient germene in polar and nonpolar solvents, and
absolute rate constants and kinetic deuterium isotope effects for
its reaction with a series of alcohols, acetic acid, and acetone.
The results allow direct comparisons to be made between the
kinetics and mechanisms of the reactions of germanium-carbon
double bonds with those of the corresponding silicon analogues.
Results
1,1-Diphenylgermetane was synthesized in 91% yield by
reaction of 1,3-dibromopropane with a ∼10-molar excess of
ground magnesium turnings, followed by addition of diphenyl-
germanium dichloride (see eq 2). The method is similar to that
In this paper, we report the results of a study of the
photochemistry of 1,1-diphenylgermetane (1) using steady-state
and nanosecond laser flash photolysis (NLFP) techniques. Our
results indicate that direct photolysis of this compound in
solution leads to the formation of the highly reactive 1,1-
Ph
1. Mg/Et2O
Ge
(2)
BrCH2CH2CH2Br
Ph
2. Ph2GeCl2
(1) Satge´, J. AdV. Organomet. Chem. 1982, 21, 241.
(2) Wiberg, N. J. Organomet. Chem. 1984, 273, 141.
(3) Barrau, J.; Escudie´, J.; Satge´, J. Chem. ReV. (Washington, D.C.) 1990,
90, 283.
(4) Escudie´, J.; Couret, C.; Ranaivonjatovo, H.; Satge´, J. Coord. Chem.
ReV. 1994, 130, 427.
(5) Leigh, W. J.; Bradaric, C. J.; Sluggett, G. W. J. Am. Chem. Soc.
1993, 115, 5332.
1
previously reported by Bickelhaupt and co-workers for the
synthesis of dialkylgermetanes,8 but for the substitution of a
large excess of (ordinary) magnesium turnings for the sublimed
magnesium used in the reported procedure for preparation of
(6) Leigh, W. J.; Bradaric, C. J.; Kerst, C.; Banisch, J. H. Organometallics
1996, 15, 2246.
(8) Seetz, J. W. F. L.; Van De Heisteeg, B. J. J.; Schat, G.; Akkerman,
(7) Bradaric, C. J.; Leigh, W. J. J. Am. Chem. Soc. 1996, 118, 8971.
O. S.; Bickelhaupt, F. J. Organomet. Chem. 1984, 277, 319.
S0002-7863(97)03756-6 CCC: $15.00 © 1998 American Chemical Society
Published on Web 02/03/1998

Direct Detection of 1,1-Diphenylgermene
J. Am. Chem. Soc., Vol. 120, No. 6, 1998 1173
the di-Grignard reagent. The compound was purified to >99%
purity by column chromatography, and identified on the basis
of the usual collection of spectroscopic and analytical data (see
Experimental Section).
Steady-state photolysis of deoxygenated, 0.01 M solutions
of 1 in cyclohexane or cyclohexane-d12 led to the gradual
disappearance of the germetane, and the formation of ethylene
1
and a single germanium-containing product. The H NMR
spectrum of the crude product mixture after 10-50% conversion
of 1 showed, in addition to a singlet at δ ) 5.30 ppm assignable
to ethylene, a singlet at δ ) 1.62 ppm, and increased complexity
in the aromatic region. Larger scale photolyses taken to higher
conversions allowed isolation of the compound as a crystalline
solid, which was identified as 1,1,3,3-tetraphenyl-1,3-diger-
metane (3) on the basis of spectroscopic and analytical data
(eq 3). Photolysis of less pure samples of 1 under similar
conditions led to the formation of ethylene, digermetane 3, and
significant amounts of a polymeric material which was not
identified.
Figure 1. Time-resolved UV absorption spectra of 1,1-diphenyl-
germene (2) in deoxygenated hexane solution at 23 °C. The spectrum
was recorded by laser flash photolysis of 4.4 × 10-3 M solutions of 1,
5-10 ms after excitation (248 nm). The insert shows a typical transient
decay trace, recorded at a monitoring wavelength of 325 nm.
decayed over several tens of microseconds with predominantly
second-order kinetics. Time-resolved absorption spectra, meas-
ured in point-by-point fashion 5-10 µs after the laser pulse,
showed an absorption band with a maximum at 325 nm in all
three solvents (Figure 1). The lifetime of the transient does
not appear to be sensitive to the presence of oxygen or 1,3-
octadiene (indicating that it is not a triplet species), but is
shortened upon addition of aliphatic alcohols or acetic acid to
the solution. On the basis of these observations, the transient
is assigned to 1,1-diphenylgermene (2).
A relatively weak transient with λmax ) 360 nm and τ )
150 ns in air-saturated solution could be detected in the presence
of enough methanol or acetic acid to shorten the lifetime of 2
to less than ∼100 ns. The lifetime of this transient did not
change upon further additions of nucleophile, but proved to be
sensitive to the presence of oxygen or aliphatic dienes. A value
of kq ) (6.1 ( 0.8) × 109 M-1 s-1 was measured for the rate
constant for quenching of the transient by 1,3-octadiene, in
deoxygenated solutions of 1 in hexane containing 0.5 M MeOH.
On the basis of these experiments, the 360-nm transient species
is assigned to a triplet, most likely that of the germetane.
However, it is difficult to rule out the possibility that it arises
from excitation of small amounts of impurities present in the
sample, because flash photolysis of less highly purified samples
of 1 gave correspondingly greater yields of a similarly absorbing
transient, superimposed on the absorptions due to 2. Addition
of 0.01 M diene to deoxygenated solutions of 1 (of any degree
of purity between 95 and 99+%) shortened the lifetime of the
triplet species to <20 ns but had no effect on the initial ∆-OD
of the transient assigned to 2, indicating that the triplet is not
responsible for the formation of 2.
Ph
Ge
Ph
Ge
Ph
Ph
hν
CH2
(3)
+ H2C
C6D12
Ge Ph
Ph
1
3
Photolysis of 1 in hexane or C6D12 containing 0.5 M
methanol, ethanol, or tert-butyl alcohol also led in each case to
the formation of ethylene and a single germanium-containing
product; digermetane 3 was not formed in detectable yields in
these experiments. These products were identified as the
corresponding alkoxymethyldiphenylgermanes 4a-c (eq 4),
Ph
a: R = Me
b: R = Et
PH2Ge
RO H(D)
hν
Ge
(4)
Ph
c: R = t-Bu
d: R = CH3CO
C6D12
ROH(D)
1
4
after isolation by semipreparative gas chromatography from
larger scale photolyses of deoxygenated solutions of 1 in the
neat alcohols, and/or by comparison to authentic samples
prepared by reaction of the appropriate alcohol with methyl-
diphenylgermanium chloride.9 Use of methanol-Od led to the
formation of 4a-d, in which deuterium was incorporated in the
methyl group of the alkoxygermane, as illustrated in eq 4.
Photolysis of a cyclohexane-d12 solution of 1 containing acetic
acid led to the formation of a product whose NMR spectrum
was consistent with acetoxygermane 4d (eq 4), but the
compound decomposed during attempts to isolate it by semi-
preparative GC. Photolysis of a 0.03 M solution of 1 in
cyclohexane-d12 containing 0.07 M acetone led to the formation
of polymeric material and ethylene; no other products were
detectable after ∼20% conversion of 1. Experiments in which
methanolic hexane solutions of 1 were photolyzed simulta-
neously with similar solutions of 1,1-diphenylsiletane (5)
indicated that the quantum yield for consumption of the
germetane is similar to that of the silicon analogue (Φ ) 0.24
( 0.046).
Above certain threshold concentrations of MeOH, EtOH,
2-PrOH, or t-BuOH (0.005-0.02 M), the decay of germene 2
proceeds with clean pseudo-first-order kinetics, and the decay
rate constant increases with increasing [ROH]. Plots of the
pseudo-first-order transient decay rate constant (kdecay) versus
alcohol concentration are approximately linear in hexane
solution, as shown in Figure 2. Linear least-squares analysis
of the data according to eq 5 gave acceptable fits, but with
NLFP of air-saturated 4.4 × 10-3 M solutions of 1 in hexane,
acetonitrile, or tetrahydrofuran (THF), with the pulses from a
KrF excimer laser (248 nm), gave rise to readily detectable
transient absorptions in the 300-360 nm spectral range, which
kdecay ) kdo + kq[Q]
(5)
negative intercepts; the apparent rate constants, given by the
slopes of these plots, are summarized in Table 1. Similar
experiments were carried out with MeOD and t-BuOD, and the
(9) Ranaivonjatovo, H.; Escudie´, J.; Couret, C.; Satge´, J.; Dra¨ger, M.
New J. Chem. 1989, 13, 389.

1174 J. Am. Chem. Soc., Vol. 120, No. 6, 1998
Toltl and Leigh
Table 1. Absolute Rate Constants for Reaction of 1,1-Diphenylgermene (2) with Alcohols, Acetic Acid, and Acetone in Air-Saturated Hexane
Solution at 23 °Ca
quencher
kq/108 M-1 s-1
kH/kD
MeOH
EtOH
i-PrOH
t-BuOH
HOAc
16 ( 2
0.9 ( 0.2
acetone
3.6 ( 0.4
1.3 ( 0.4
2.5 ( 0.2
1.6 ( 0.2
0.35 ( 0.03
1.9 ( 0.3
<0.004 ( 0.001
b
b
b
a From the slopes of plots of kdecay versus [Q] (eq 5). Errors are quoted as (2σ. b Not determined.
Figure 2. Plots of kdecay versus [ROH] for quenching of 1,1-
diphenylgermene (2) by MeOH (0) and t-BuOH (O) in air-saturated
hexane solution at 23 °C. The solid lines represent the least-squares
fits of the data to eq 4.
results are expressed in the table as kH/kD values. Plots of kdecay
versus [Q] for acetic acid and acetic acid-d exhibited excellent
linearity and positive intercepts. Reaction of 2 with acetone
was found to be too slow to be measured with our system, and
only a crude upper limit for the quenching rate constant could
be determined (from the lifetime of 2 in the presence of 0.23
M acetone). The rate constants obtained in these experiments
are included with those for alcohol quenching in Table 1.
Addition of methanol or t-butanol to acetonitrile solutions
of 1 also resulted in a shortening of the lifetime of germene 2,
but plots of kdecay versus [ROH] exhibited strong positive
curvature over the ranges of alcohol concentration studied. In
fact, the decay rate constants were found to fit well to the square
of the alcohol concentration (r2 > 0.99) according to eq 6.
Alternative fits to a polynomial expression containing both first
and second order terms in [ROH] (eq 7) gave first-order
coefficients identical to zero within two standard deviations.
Figure 3a shows representative plots of kdecay versus [ROH],
for quenching of germene 2 by MeOH and t-BuOH in MeCN
solution.
Figure 3. Plots of kdecay versus [ROH] for quenching of 2 at 23 °C;
(a) MeOH (0) and t-BuOH (O) in air-saturated MeCN; (b) MeOH (0)
and t-BuOH (O) in air-saturated THF. The solid lines represent the
least-squares fit of the data to eqs 6 (a) or 7 (b).
Table 2. Absolute Rate Constants for Reaction of
1,1-Diphenylgermene (2) with MeOH, t-BuOH, and HOAc in
Air-Saturated Acetonitrile and Tetrahydrofuran Solution at 23 °Ca
quencher
MeOH
t-BuOH
HOAcb
MeCN
THF
kq/107 M-1 s-1
k2q/107 M-2 s-1
kq/107 M-1 s-1
k2q/107 M-2 s-1
0.03 ( 0.2c
2.2 ( 0.2
0.5 ( 0.2
1.4 ( 0.4
0.04 ( 0.3c
0.14 ( 0.02
0.06 ( 0.03
0.13 ( 0.03
77 ( 4
-
kdecay ) kdo + k2q[Q]2
(6)
(7)
18 ( 1
-
kdecay ) kdo + kq[Q] + k2q[Q]2
a From polynomial least-squares fitting of kdecay versus [Q] data
b
according to eq 7. Errors are quoted as (2σ. kdecay varies linearly
with [HOAc] according to eq 5. c Indistinguishable from zero within
experimental error.
Plots of kdecay versus [MeOH] in THF solution were also
curved, but curve-fitting of the data to eq 7 afforded nonzero
values for both the first- and second-order terms in [ROH].
Figure 3b shows representative plots of this type for quenching
of 2 by MeOH and t-BuOH in THF at 23 °C. Plots of kdecay
versus [HOAc] for quenching of 2 by acetic acid in both
acetonitrile and THF solution were linear, with positive
intercepts. The second- and third-order rate constants (kq and
k2q, respectively) obtained from the experiments in MeCN and
THF solution are collected in Table 2.
Discussion
The products of direct irradiation of 1,1-diphenylgermetane
(1), either alone or in the presence of alcohols, are consistent
with the formation of 1,1-diphenylgermene (2) and ethylene as
the primary photochemical products. This behavior is exactly
analogous to that of 1,1-diphenylsiletane (5), which yields the

Direct Detection of 1,1-Diphenylgermene
J. Am. Chem. Soc., Vol. 120, No. 6, 1998 1175
head-to-tail [2+2]-dimer of 1,1-diphenylsilene (6; eq 8) in the
Scheme 1
Ph
Si
Ph
Si
Ph
Ph
hν
x 2
Ph
Ph
Si CH2
(8)
C6H14
(–C2H4)
Ph
Si
Ph
7
5
6
absence of silene traps,10 or alkoxysilanes in the presence of
alcohols.5-7 These similarities between germenes and silenes
are well known.3,4 Compounds 1 and 5 both undergo [2+2]-
photocycloreversion in solution with quantum yields on the order
of ∼0.25, and because the initial yields of 2 and 6 (as estimated
from the initial intensities of the transient absorption observed
by flash photolysis) are insensitive to the presence of oxygen
or diene, it can be concluded that the process is singlet-derived
in both cases.6 Thus, the photochemical behavior of 1,1-
diphenylgermetane (1) is identical in almost every respect to
that of its silicon analogue 5.
The kinetic behavior of the 325-nm transient observed in laser
flash photolysis experiments is also consistent with its assign-
ment as 2; it decays with predominant second-order kinetics in
hexane solution (indicative of dimerization) and with pseudo-
first-order kinetics in the presence of alcohols or acetic acid.
The UV spectrum is indistinguishable from that of 1,1-
diphenylsilene (6).5 Like that of 6, the position and breadth of
the absorption band are identical in hexane and acetonitrile, but
there is a slight broadening of the spectrum in THF solution.
This may result from weak complexation with the solvent, but
the effect is much smaller than that exhibited by the silicon
analogue.6
As expected, 1,1-diphenylgermene reacts rapidly with ali-
phatic alcohols and acetic acid, although it does so significantly
slower than its silicon homologue 6.6 The rate constants
measured for 2 in hexane are mechanistically intractable (vide
infra), but allow the qualitative conclusion that it exhibits the
same ordering of relative reactivity as 6 toward these reagents:
HOAc > MeOH > EtOH > 2-PrOH > t-BuOH. The 1H NMR
spectra of crude product mixtures from photolysis of 1 in the
presence of HOAc suggest that reaction of 2 with the carboxylic
acid proceeds with the same regiochemistry as the reaction with
alcohols, yielding methyldiphenyl-germyl acetate (3d). Unfor-
tunately however, the acetoxygermane proved to be too labile
for us to isolate.
The kinetic data for reaction of 2 with MeOH and t-BuOH
in acetonitrile suggest that the mechanism for the reaction of
the germene with alcohols differs significantly from that of
silenes. The latter is well understood; it is initiated by fast,
reversible nucleophilic attack at silicon to form a zwitterionic
complex, which collapses to alkoxysilane by two competing
proton-transfer pathwayssone intramolecular, and one which
involves a second molecule of alcohol.6,11,12 Although intra-
molecular proton-transfer is the slowest step in the sequence, it
proceeds with a lower enthalpy of activation than that of
reversion of the complex to starting materials. As a result, the
rate constants typically show significant deuterium isotope
effects, but exhibit negative overall activation energies.7,13,14 The
mechanism is shown in Scheme 1.
The intervention of the pathway involving reaction of the
zwitterionic complex with a second molecule of alcohol was
first proposed in order to explain a dependence of alkoxysilane
product distribution on bulk alcohol concentration, in the
reaction of a transient cyclic silene with aliphatic alcohols.11
Other examples of such behavior have been reported for other
transient silenes as well.6,12,15 In cases where the intracomplex
proton transfer is slow enough, this mechanism also manifests
itself in a quadratic dependence of the pseudo-first-order rate
constant for silene decay on alcohol concentration.12,16 For 1,1-
diphenylsilene however, proton transfer by the intramolecular
route is so fast that kinetic analysis reveals only the first-order
dependence of kdecay on alcohol concentration in both acetonitrile
and hexane solution.6,7
In fact, the kinetic data for reaction of germene 2 with
alcohols and acetic acid are consistent with a mechanism very
similar to that for addition of these reagents to the corresponding
silene, but with one crucial differencesthe proton transfer
leading to conversion of the intermediate complex to alkoxy-
germane proceeds entirely (within our limits of detection) by
the general base catalyzed route described by k3 in Scheme 1.
The base required for this step can either be the solvent (e.g.,
THF) or a second molecule of alcohol; in general, the two will
compete for deprotonation of the complex according to their
relative basicities. Deprotonation by the solvent should result
in a first-order dependence of the pseudo-first-order rate constant
for germene decay on [ROH], and deprotonation by ROH should
result in a second-order dependence on [ROH]. Application
of the equilibrium assumption for the germene-alcohol complex
leads to the expression shown in eq 9 for the pseudo-first-order
k1[ROH]
o
kdecay ) kd
+
(k3′[THF] + k3[ROH])
(9)
k-1
rate constant for decay of 2 in the presence of alcohol, where
k1, k-1, k3, and k3′ are defined in Scheme 1. THF is significantly
less basic than MeOH in nonaqueous solvents;17 thus, kdecay is
(10) Jutzi, P.; Langer, P. J. Organomet. Chem. 1980, 202, 401.
(11) Kira, M.; Maruyama, T.; Sakurai, H. J. Am. Chem. Soc. 1991, 113,
3986.
(12) Leigh, W. J.; Sluggett, G. W. J. Am. Chem. Soc. 1994, 116, 10 468.
(13) Kerst, C.; Byloos, M.; Leigh, W. J. Can. J. Chem. 1997, 75, 975.
(14) Bradaric, C. J.; Leigh, W. J. Can. J. Chem. 1997, 75, 1393.
(15) Steinmetz, M. G.; Udayakumar, B. S.; Gordon, M. S. Organo-
metallics 1989, 8, 530.
(16) Sluggett, G. W.; Leigh, W. J. J. Am. Chem. Soc. 1992, 114, 1195.

1176 J. Am. Chem. Soc., Vol. 120, No. 6, 1998
Toltl and Leigh
predicted to vary with [ROH] according to eq 7 in THF, with
kq ) (k1/k-1)k3′[THF] and k2q ) (k1/k-1)k3. In weakly basic
solvents such as MeCN or hexane (where k3′ , k3), this
expression collapses to eq 6.
in both cases, addition across the metal-carbon double bond
involves initial complexation, followed by proton transfer from
oxygen to carbon via four-membered transition states. The
calculated barriers to proton transfer within the complexes are
similar, suggesting that germene and silene should show similar
reactivities toward nucleophilic addition. Our experimental
results for alcohol additions to 2 and 6 are consistent with the
intermediacy of complexes in both cases, but indicate that
silene-alcohol complexes are substantially more reactive toward
intramolecular proton transfer than the germanium analogues.
We believe that the difference can most likely be ascribed to
complexation being significantly weaker with germenes than
with silenes, resulting in a lower degree of charge transfer from
oxygen to the R-carbon. This would be consistent with the
differences in the calculated charge distributions in germene
and silene,20 as well as with the differences in the UV absorption
spectra of 2 and 6 in tetrahydrofuran solution. The spectrum
of 6 is broadened considerably and slightly red-shifted in THF
compared to MeCN,6 while the spectrum of 2 in THF shows at
most a very slight broadening compared to that in MeCN. These
results indicate that the silene is solvated significantly more
strongly than the germene by the ether solvent.
Reaction with carbonyl compounds, which generally yields
1,2-germoxetane derivatives, is one of the best known trapping
reactions of stable and transient germenes.3,4 Transient silenes
are also known to react with acetone and other aliphatic ketones
to yield silyl enol ethers and/or siloxetane derivatives.22,23 The
rate constants, at least for enol ether formation, are typically
substantial; for example, 6 reacts with acetone with a rate
constant k ∼ 3 × 108 M-1 s-1 in hexane at 23 °C.5 We were
thus surprised to find that no reaction between germene 2 and
acetone could be detected by steady-state or flash photolysis
methods. Because recent results indicate that the reaction of
silenes with acetone is also initiated by complex formation,24 it
is possible that the much-reduced reactivity of 2 with acetone
is due to similar factors as those proposed above for alcohol
addition. The weaker complex formed in the germene case does
not provide enough negative charge density at the germenic
carbon to allow for rapid transfer of the weakly acidic R-proton.
The results obtained for quenching of 2 by MeOH and
t-BuOH in THF and MeCN are in excellent agreement with
this mechanism, as can be seen from the data in Figure 3. In
THF solution, the second-order rate coefficient kq yields (k1/
k-1)k3′ ) (4 ( 2) × 105 M-2 s-1, which corresponds to the
reaction pathway involving deprotonation of the complex by
the solvent. This is a factor of ∼35 lower than that involving
deprotonation by a second molecule of alcohol, which is in
reasonable agreement with the difference between the acid
+
dissociation constants of protonated THF (pKa 1.1) and MeOH2
(pKa 2.36) in acetonitrile solution.17 The third-order rate
constant for quenching of 2 by MeOH is slightly lower in THF
than in MeCN, which could result from either weak complex-
ation of the germene with the ether solvent or a difference in
the basicity of the alcohol in the two solvents.
Although the plots of kdecay versus [ROH] in hexane solution
are linear within experimental error, the fact that the least-
squares analyses yield negative intercepts indicates that the
quenching kinetics are more complex than the apparent linearity
would imply. Over the alcohol concentration range employed
in our kinetic experiments, both methanol and tert-butyl alcohol
are substantially oligomerized.18 In a plot of kdecay versus bulk
ROH concentration, this would have the effect of linearizing
the quadratic dependence on [ROH] that is predicted by the
mechanism of Scheme 1. A quantitative treatment of the hexane
data, taking alcohol oligomerization effects into account, is
difficult to carry out without considerably more data than we
have collected.
Quenching by acetic acid proceeds with strict overall second-
order kinetics in all three solvents, as evidenced by the strictly
linear dependences of the germene decay rate constant on
[HOAc]. If addition of HOAc also involves initial complex-
ation, as it very likely does with 1,1-diphenylsilene,14 then the
kinetics suggest that intracomplex proton transfer is much faster
than is the case in the reaction with alcohols. This would be
expected on the basis of the substantially greater acidity of the
carboxylic acid, as well as the fact that proton transfer can
proceed via a six-membered transition state if complexation
occurs at the carbonyl oxygen. In hexane solution, quenching
is rapid enough that a rate constant can be determined at acetic
acid concentrations below that at which dimerization of the acid
starts to become significant.19 It should be noted that a
mechanism for HOAc addition which involves initial protonation
of the GedC bond cannot be ruled out on the basis of the data
presented here, although we have not been able to detect
additional transient species in any of our time-resolved experi-
ments with the carboxylic acid. The substantially greater
reactivity of 2 toward HOAc compared to MeOH actually makes
this a reasonable possibility (6 exhibits similar reactivities toward
the two reagents6,14), and we are continuing to investigate it.
Nagase and co-workers have reported ab initio theoretical
calculations on the addition of water to silene (H2SidCH2)20
and germene (H2GedCH2).20,21 The calculations predict that
Summary and Conclusions
The transient 1,1-diphenylgermene (2) can be generated and
detected in solution by laser flash photolysis of 1,1-diphenyl-
germetane (1). The germene undergoes head-to-tail [2+2]-
dimerization in hexane solution, or can be trapped by aliphatic
alcohols as the corresponding alkoxymethyldiphenylgermanes.
Kinetic evidence suggests that this reaction proceeds by a
mechanism involving the initial formation of a germene-alcohol
complex, as is the case with alcohol additions to silenes.
Silene-alcohol complexes collapse to product by rapid in-
tramolecular proton transfer, as well as by an intermolecular
general base-catalyzed process which usually becomes important
only at relatively high alcohol concentrations. In contrast,
proton transfer in 1,1-diphenylgermene-alcohol complexes
appears to proceed only by the general base-catalyzed pathway.
This involves a second molecule of alcohol in hexane or
acetonitrile, in addition to the solvent in more basic media like
THF. If this mechanism is correct, then further analogy with
silene chemistry6,11,12,15 would predict that the addition of
alcohols and water to germenes should proceed nonstereo-
selectively or perhaps even with predominant anti-stereochem-
(17) Izutsu, K. Acid-base dissociation constants in dipolar aprotic
solVents. IUPAC Chemical Data Series #35; Blackwell Scientific Publica-
tions: Oxford, 1990; pp 17-35.
(18) Landeck, H.; Wolff, H.; Gotz, R. J. Phys. Chem. 1977, 81, 718.
(19) Fujii, Y.; Yamada, H.; Mizuta, M. J. Phys. Chem. 1988, 92, 6768.
(20) Nagase, S.; Kudo, T.; Ito, K. In Applied Quantum Chemistry; Smith,
V. H. J., Schaefer, H. F., Morokuma, K., Eds.; D. Reidel: Dordrecht, 1986;
pp 249-267.
(22) Raabe, G.; Michl, J. Chem. ReV. (Washington, D.C.) 1985, 85, 419.
(23) Brook, A. G.; Brook, M. A. AdV. Organomet. Chem. 1996, 39, 71.
(24) Bradaric, C. J.; Leigh, W. J. Organometallics 1998, in press.
(21) Nagase, S.; Kudo, T. Organometallics 1984, 3, 324.

Direct Detection of 1,1-Diphenylgermene
J. Am. Chem. Soc., Vol. 120, No. 6, 1998 1177
for C15H16Ge 270.0464, found 270.0470. Anal. Calcd for C15H16Ge:
C, 67.01, H, 6.00. Found: C, 67.07, H, 6.05.
istry. This and other aspects of the chemistry of germene
reactive intermediates are under continued investigation in our
laboratory.
Methyldiphenylgermanium chloride, from which authentic samples
of 4a and 4c were prepared, was synthesized as follows. Methyldi-
phenylgermane25 (1.75 g, 7.2 mmol) was dissolved in carbon tetra-
chloride (25 mL) in a 100 mL round-bottom flask, and then chlorine
gas was bubbled through the solution until it assumed a slightly yellow
tinge. The chlorine gas source was removed, and the solution was
irradiated with a sunlamp for 8 min. Removal of the solvent by vacuum
distillation afforded a slightly yellow oil, which was identified as
methyldiphenylgermanium chloride26 (1.97 g, 7.1 mmol, 99%) on the
basis of the following spectroscopic data: 1H NMR δ ) 0.99 (s, 3H),
7.26-7.30 (m, 6H), 7.50-7.55 (m, 4H); 13C NMR δ ) 1.91, 128.84,
130.41, 133.72, 137.76; IR (neat) 3070.8 (m), 2914.0 (w), 2861.2 (w),
1958.9 (w), 1883.9 (w), 1818.6 (w), 1484.9 (m), 1433.1 (s), 1334.9
(m), 1245.0 (m), 1094.1 (s), 1066.7 (m), 1026.6 (m), 802.5 (s), 735.0
(s), 696.9 (s); MS (EI) m/z (I) ) 278 (20), 263 (90), 243 (35), 201
(10), 151 (20), 117 (25), 91 (7), 77 (25); exact mass calcd for C13H13-
GeCl 277.9917, found 277.9914.
Methoxy- and tert-butoxymethyldiphenylgermane (4a,c) were syn-
thesized by reaction of methyldiphenylgermanium chloride with
methanol and tert-butanol, respectively, in dry benzene in the presence
of triethylamine.9 For example, methyldiphenylgermanium chloride
(0.25 g, 0.9 mmol) and dry methanol (0.1151 g, 3.6 mmol) were placed
in benzene (10 mL) and stirred at room temperature. Triethylamine
(0.3636 g, 3.6 mmol) was added rapidly to the stirred solution, causing
the immediate formation of a white precipitate. The reaction mixture
was stirred under reflux for 24 h (72 h for tert-butyl alcohol), cooled
and filtered, and then concentrated to ca. 1 mL by vacuum distillation.
GC analysis of the crude reaction mixture indicated the conversion of
the chloro- to alkoxygermane to be quantitative in both cases. The
alkoxygermanes were isolated from the crude reaction mixtures by
semipreparative gas chromatography.
Methoxymethyldiphenylgermane (4a). 1H NMR δ ) 0.71 (s, 3H),
3.44 (s, 3H), 7.10-7.35 (m, 6H), 7.42-7.61 (m, 4H); 13C NMR δ )
-3.67, 52.76, 128.57, 129.90, 134.35, 137.61; IR (neat) 3069.9 (m),
2955.3 (m), 2925.9 (s), 2816.8(m), 1962.9 (w), 1884.4 (w), 1829.2 (w),
1484.9 (m), 1431.8 (s), 1376.3 (m), 1185.8 (w), 1095.3 (s), 1052.7 (s),
998.6 (m), 847.7 (m), 790.9 (m); MS (EI) m/z (I) ) 274 (6), 259 (100),
229 (60), 167 (20), 151 (50), 89 (20), 77 (10). Anal. Calcd for C14H16-
GeO: C, 61.62, H, 5.91. Found: C, 61.95, H, 5.88.
tert-Butoxymethyldiphenylgermane (4c). 1H NMR δ ) 0.80 (s,
3H), 1.18 (s, 9H), 7.22-7.25 (m, 6H), 7.47-7.51 (m, 4H); 13C NMR
δ ) 0.55, 33.03, 73.17, 128.40, 129.55, 134.26, 140.40; IR (neat) 3071.5
(m), 2973.5 (s), 2886.3 (w), 2857.5 (w), 1950.3 (w), 1882.6 (w), 1817.3
(w), 1431.9 (s), 1361.5 (s), 1240.3 (m), 1192.4 (s), 1095.1 (s), 970.8
(s), 851.7 (w), 613.0 (m); MS (EI) m/z (I) ) 301 (30), 243 (100), 223
(10), 183 (7), 151 (25), 125 (3), 91 (15), 77 (15), 43 (40); exact mass
calcd for (M - 15) C16H19GeO 297.0678, found 297.0669.
Experimental 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 an HP-5971A mass-selective detector and a DB-5 fused
silica capillary column (30 m × 0.25 mm; Chromatographic Specialties,
Inc.). High-resolution desorption electron impact (DEI) and chemical
ionization (CI) 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). Elemental
analyses were performed by Galbraith Laboratories Inc.
Analytical gas chromatographic analyses were carried out using a
Hewlett-Packard 5890 gas chromatograph 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.).
Semipreparative GC separations employed a Varian 3300 gas chro-
matograph equipped with a thermal conductivity detector, and a 6 ft ×
0.25 in. stainless steel OV-101 packed column (Chromatographic
Specialties Inc.). Radial chromatographic separations employed a
Chromatotron (Harrison Research, Inc.), 2- or 4-mm silica gel 60
thick-layer plates, and hexane/ethyl acetate mixtures as eluant.
Acetonitrile (Caledon Reagent) was refluxed over calcium hydride
(Fisher) for several days and distilled under dry nitrogen. Hexane
(Caledon Reagent) was stirred for several days over concentrated
sulfuric acid, washed several times with water, washed once with
saturated aqueous sodium bicarbonate, dried over anhydrous sodium
sulfate, and distilled from sodium. Tetrahydrofuran (BDH Omnisolv)
was refluxed for several days over sodium under nitrogen and distilled.
Methanol, 2-propanol, tert-butanol (HPLC grade), 1,3-octadiene, and
1,3-dibromopropane were used as received from Aldrich Chemical Co.
Absolute ethanol was predried with calcium hydride, distilled from
magnesium under nitrogen, and stored over 3 Å molecular sieves.
Benzene (Caledon) was distilled from sodium. Deuterated materials
were used as received from Cambridge Isotope Labs. Diphenyl-
germanium dichloride was used as received from Gelest, Inc.
1,1-Diphenylgermetane (1) was prepared by a modification of the
method of Bickelhaupt and co-workers.8 Freshly ground magnesium
turnings (45 g, 1.85 g-atom) and anhydrous ether (300 mL) were placed
in a flame-dried, two-neck, 500 mL round-bottom flask fitted with a
condenser and addition funnel. A solution of 1,3-dibromopropane (26.0
g, 0.129 mol) in anhydrous ether (40 mL) was added dropwise with
vigorous stirring over a period of 4 h, and then left to stir overnight at
room temperature. Diphenylgermanium dichloride (5.50 g, 0.0185 mol)
was then added rapidly with a syringe and the mixture was left to stir
for a further 8 h. The reaction mixture was decanted into a large flask,
placed in an ice bath, and slowly quenched with saturated aqueous
ammonium chloride (200 mL). The organic layer was separated and
washed sequentially with water, 5% sodium bicarbonate, and water,
and then dried with anhydrous magnesium sulfate. The solvent was
removed on the rotary evaporator to yield a slightly yellow oil.
Purification was accomplished using silica gel column chromatography
with hexane as the eluting solvent, affording the product as a colorless
oil (4.58 g, 0.0168 mol, 91%) in >99% purity. It was identified as
1,1-diphenylgermetane (1) on the basis of the following spectroscopic
data: 1H NMR δ ) 1.96 (t, 4H), 2.40 (m, 2H), 7.37-7.41 (m, 6H),
7.56-7.61 (m, 4H); 13C NMR δ ) 20.45, 21.46, 128.29, 129.07, 134.01,
138.81; IR (neat), 3067.2 (s), 2926.3 (s), 1952.6 (w), 1817.1 (w), 1483.1
(m), 1430.5 (s), 1303.5 (m), 1262.0 (w), 1182.0 (w), 1091.1 (s), 843.2
(s), 788.7 (s); MS (EI) m/z (I) ) 270 (5), 242 (5), 227 (7), 151 (8), 84
(20); [NH3-CI], m/z (I) ) 288 (95), 269 (15), 244 (8), 210 (100), 181
(10), 162 (10), 137 (12), 113 (13), 78 (45), 52 (82); exact mass calcd
Preparative irradiations of 1 employed a Rayonet Photochemical
Reactor (Southern New England Ultraviolet Co.) equipped with 8-10
RPR-2537 lamps. For the synthesis of digermetane 3, a deoxygenated
solution of 1 (0.25 g, 0.9 mmol) in hexane (15 mL) was placed in a
quartz photolysis tube sealed with a rubber septum, and irradiated for
8 h. The solvent was removed on the rotary evaporator to yield a
slightly yellow oil, the NMR spectrum of which indicated that it
consisted of 1 and 3 in a 3:2 ratio. The germene dimer was isolated
by radial chromatography using hexane as eluant. After elution of
starting material (1, 0.15 g), the solvent was changed to 10%
dichloromethane in hexane. The product was collected as a colorless
oil (0.08 g, 0.16 mmol, 91% yield based on recovered 1) which
crystallized upon addition of a small amount of pentane and cooling
to -10 °C. A second recrystallization from pentane afforded the
compound as colorless needles (mp 120-121 °C) which were identified
as 1,1,3,3-tetraphenyl-1,3-digermetane (3) on the basis of the following
spectroscopic data: 1H NMR δ ) 1.62 (s, 4H), 7.31-7.36 (m, 12H),
7.50-7.56 (m, 8H); 13C NMR δ ) 7.71, 128.22, 128.96, 133.67,
(25) Castel, A.; Riviere, P.; Satge, J.; Ko, H. Y. Organometallics 1990,
9, 205.
(26) Mochida, K.; Wakasa, M.; Sakaguchi, Y.; Hayashi, H. J. Am. Chem.
Soc. 1987, 109, 7942.

1178 J. Am. Chem. Soc., Vol. 120, No. 6, 1998
Toltl and Leigh
139.53; IR (neat), 3070.2 (m), 2955.3 (m), 2926.2 (m), 1952.5 (w),
1879.7 (w), 1814.8 (w), 1484.6 (m), 1431.4 (s), 1373.7 (m), 1354.4
(m), 1091.8 (s), 954.6 (m), 908.9 (s); MS (EI) m/z (I) ) 482 (10), 406
(15), 391 (50), 305 (100), 271 (5), 243 (40), 227 (40), 151 (50); exact
mass calcd for C26H24Ge2 484.0261, found 484.0302. Anal. Calcd
for C26H24Ge2: C, 64.84, H, 5.02. Found: C, 64.64, H, 5.06.
Solutions of 1 (0.1 g, 3.66 × 10-4 mol) in hexane (15 mL) containing
1.0 M of MeOH, MeOD, EtOH, or t-BuOH were placed in quartz tubes,
sealed with a rubber septum, deoxygenated with a stream of dry argon,
and then photolyzed for 1 h. Periodic monitoring of the course of the
photolyses by GC revealed the formation of a single detectable product
in all cases. The solvent was removed on the rotary evaporator to yield
colorless oils containing product and small amounts of residual 1. The
products were isolated by semipreparative GC. Those from photolysis
with MeOH and t-BuOH were identical to the authentic samples of 4a
and 4c (vide supra). Those from photolysis with MeOD and EtOH
were identified as 4a-d and 4b, respectively, on the basis of the
following spectroscopic data.
These experiments gave comparable results to those described above;
in addition, the formation of ethylene was evident from the appearance
of a singlet at 5.30 ppm, which was verified by the addition of a small
amount of an authentic sample to one of the photolysis mixtures.
Attempted acetone trapping experiments were carried out using a 0.03
M solution of 1 containing 0.07 M dry acetone. Photolysis to ∼20%
conversion of 1 led only to the formation of a polymer characterized
by a singlet at 0.19 ppm and broad resonances in the aromatic region.
NLFP experiments employed the pulses (248 nm; ca. 16 ns; 70-
120 mJ) from a Lumonics 510 excimer laser filled with F2/Kr/He
mixtures, and a microcomputer-controlled detection system.27 The
system incorporates a brass sample holder, the temperature of which
is controlled to within 0.1 °C by a VWR 1166 constant temperature
circulating bath. 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. 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.
Methoxy(methyl-d)diphenylgermane (4a-d) 1H NMR δ ) 0.69 (t,
2H), 3.44 (s, 3H), 7.27-7.35 (m, 6H), 7.46-7.61 (m, 4H); 13C NMR
δ ) -4.30, -3.92, -3.54, 52.77, 128.77, 129.90, 134.34, 137.59; IR
(neat) 3069.1 (m), 2926.6 (m), 2854.1 (m), 1960.2 (w), 1889.6 (w),
1824.6 (w), 1486.1 (m), 1431.3 (s), 1157.6 (m), 1094.9 (s), 844.7 (s),
734.2 (s); MS (EI) m/z (I) ) 275 (5), 259 (100), 244 (20), 229 (70),
168 (20), 151 (60), 125 (30), 105 (15), 77 (10).
1
Ethoxymethyldiphenylgermane (4b) H NMR δ ) 0.71 (s, 3H),
1.12 (t, 3H), 3.62 (q, 2H), 7.24-7.31 (m, 6H), 7.47-7.51 (m, 4H);
13C NMR δ ) -3.12, 19.59, 60.74, 128.54, 129.84, 134.77, 138.07;
IR (neat) 3069.4 (m), 2966.4 (m), 2872.7 (m), 1960.8 (w), 1885.1 (w),
1824.9 (w), 1484.8 (m), 1431.5 (s), 1383.3 (m), 1241.5 (w), 1095.7
(s), 1061.1 (s), 1026.6 (w), 848.6 (m), 788.9 (m); MS (EI) m/z (I) )
288 (5), 273 (50), 243 (70), 229 (100), 165 (20), 151 (80), 123 (15),
77(20); exact mass calcd for C15H18GeO 288.0581, found 288.0569.
Anal. Calcd for C15H18GeO: C, 62.80, H, 6.32. Found: C, 62.50, H,
6.25.
Acknowledgment. We wish to thank D. Hughes for his
assistance with high-resolution NMR spectra, and R. Smith and
F. Ramelan of the McMaster Regional Centre for Mass
Spectrometry for high-resolution mass spectra and exact mass
determinations. The financial support of the Natural Sciences
and Engineering Research Council (NSERC) of Canada is also
gratefully acknowledged.
JA973756F
NMR scale photolyses were carried out in quartz NMR tubes, using
∼0.1 M solutions of 1 in cyclohexane-d12. For trapping experiments
the solutions also contained 1.0 M of MeOH, EtOH, t-BuOH, or HOAc.
(27) Leigh, W. J.; Workentin, M. S.; Andrew, D. J. Photochem.
Photobiol. A: 1991, 57, 97.