Synthesis of an organogallium(I) compound [Ga(CH2CMe2Ph)]n with EPR spectral evidence for gallium clusters

Article abstract of DOI:10.1016/S0022-328X(98)01182-6

The organogallium(I) compound [Ga(CH₂CMe₂Ph)]_n has been prepared by the reduction of Ga(CH₂CMe₂Ph)₂Cl by using either sodium or lithium with naphthalene in THF. The yellow dihydronaphthalene gallium(III) intermediate M₂{C₁₀H₈[Ga(CH₂CMe ₂Ph)₂Cl]₂} initially formed at -78°C but then decomposed at higher temperatures to form [Ga(CH₂CMe₂Ph)]_n, Ga(CH₂CMe₂Ph)₃ and MCl. EPR spectra, which were recorded as the two yellow intermediates Na₂{C₁₀H₈[Ga(CH₂CMe ₂R)₂Cl]₂} (R = Ph, Me) decomposed, indicated the presence of radicals. The first and second derivatives of the EPR signals, line-widths, g-values and hyperfine coupling constants are consistent with the radicals being clusters of organogallium species. The experimental spectra were simulated by the superposition of two spectra, a single Gaussian shaped line with a peak-to-peak width of 14 mT (92%) with the spectrum of five equivalent gallium nuclei, A = 2.1 mT (8%). The even number of lines observed in the experimental spectra indicate an odd number of gallium nuclei with at least five being required to give the number of observed lines.

Full text of DOI:10.1016/S0022-328X(98)01182-6

Journal of Organometallic Chemistry 582 (1999) 32–39
Synthesis of an organogallium(I) compound [Ga(CH2CMe2Ph)]n with
EPR spectral evidence for gallium clusters
O.T. Beachley Jr. *, Matthew J. Noble, Robert D. Allendoerfer 1
State Uni6ersity of New York at Buffalo, Natural Science and Mathematics Complex, Department of Chemistry, Box 603000, Buffalo,
NY 14260-3000, USA
Received 1 September 1998
Abstract
The organogallium(I) compound [Ga(CH2CMe2Ph)]n has been prepared by the reduction of Ga(CH2CMe2Ph)2Cl by using
either sodium or lithium with naphthalene in THF. The yellow dihydronaphthalene gallium(III) intermediate
M2{C10H8[Ga(CH2CMe2Ph)2Cl]2} initially formed at −78°C but then decomposed at higher temperatures to form
[Ga(CH2CMe2Ph)]n, Ga(CH2CMe2Ph)3 and MCl. EPR spectra, which were recorded as the two yellow intermediates
Na2{C10H8[Ga(CH2CMe2R)2Cl]2} (R=Ph, Me) decomposed, indicated the presence of radicals. The first and second derivatives
of the EPR signals, line-widths, g-values and hyperfine coupling constants are consistent with the radicals being clusters of
organogallium species. The experimental spectra were simulated by the superposition of two spectra, a single Gaussian shaped line
with a peak-to-peak width of 14 mT (92%) with the spectrum of five equivalent gallium nuclei, A=2.1 mT (8%). The even number
of lines observed in the experimental spectra indicate an odd number of gallium nuclei with at least five being required to give
the number of observed lines. © 1999 Elsevier Science S.A. All rights reserved.
Keywords: Organogallium compounds; Gallium clusters; EPR spectra; Gallium radicals
The syntheses of organogallium compounds in oxida-
tion states other than the typical +3 present interesting
challenges. The most common preparative route has
been the reaction of a low oxidation state gallium
halogen compound with an organo lithium or sodium
reagent [1–4]. An alternative route to low oxidation
state compounds involves the reduction of an
organogallium(III) halide with a metal. Sodium/potas-
sium alloy [5] reduced Ga[C6H2(CF3)3]2Cl at 25°C in
hexane to form Ga2[C6H2(CF3)3]4. When Ga[C6H3-
(C6H2Me3)2]Cl2 was reacted with sodium [6], or potas-
sium [7] metal, M2{Ga[C6H3(C6H2Me3)2]}3 (M=Na,
K), compounds with triangular arrays of gallium atoms
were formed whereas reduction with sodium of a
derivative with more bulky substituents [8]
Ga{C6H3[C6H2(2,4,6 i-Pr)3]2}Cl2 produced Na2[Ga{C6-
H3[C6H2(2,4,6 i-Pr)3]2}]2. The gallium(I) compound
Ga(C5Me5) was formed by a reductive dehalogenation
of Ga(C5Me5)I2 with potassium [9], whereas the reduc-
tion of LiGaCl3[C(SiMe2R)3] (R=Me, Et) with Rieke
magnesium produced [GaC(SiMe2R)3]4 [10]. When
sodium naphthalenide [11] was the potential reducing
agent and was reacted with Ga(CH2CMe3)2Cl at
−78°C, the initial product was a bright yellow dihy-
dronaphthalene gallium(III) derivative Na2{C10H8-
[Ga(CH2CMe3)2Cl]2} rather than a low oxidation state
compound. However, this intermediate decomposed at
room temperature (r.t.) to form the gallium(I) com-
pound [Ga(CH2CMe3)]n (n=6–12), Ga(CH2CMe3)3
and NaCl. The product [Ga(CH2CMe3)]n was con-
firmed as gallium(I) by oxidation with aqueous DCl to
form D2 and with HgCl2 to form Ga(CH2CMe3)Cl2 and
mercury metal and was believed to exist as clusters with
* Corresponding author.
1 Also corresponding author.
0022-328X/99/$ - see front matter © 1999 Elsevier Science S.A. All rights reserved.
PII: S0022-328X(98)01182-6

O.T. Beachley et al. / Journal of Organometallic Chemistry 582 (1999) 32–39
33
different numbers of Ga(CH2CMe3) moieties. However,
the only experimental evidence for clusters was the
cryoscopic molecular weight data which indicated an
average association of nine Ga(CH2CMe3) moieties.
Thus, rings and chains could not be ruled out.
In this paper, we report the synthesis of
[Ga(CH2CMe2Ph)]n and the results of our investiga-
tions of the EPR spectra which were recorded as the
yellow neophyl and neopentyl dihydronaphthalene gal-
lium(III) intermediates M2{C10H8[Ga(CH2CMe2R)2-
Cl]2} (R=Ph, Me; M=Li, Na) were transformed into
the final reddish–brown gallium(I) products. It is of
interest that Na2{C10H8[Ga(CH2CMe2R)2Cl]2} formed
the gallium(I) compound faster than Li2{C10H8[Ga-
(CH2CMe2R)2Cl]2}.
Fig. 2. EPR spectrum (second derivative) for the decomposition of
the yellow intermediate formed from the reaction of
Ga(CH2CMe2Ph)2Cl with Na(C10H8).
2.002–2.006 [12,13], and are typically 2–3 mT wide. A
second small EPR signal is observed at g=2.003 with a
peak-to-peak linewidth of 0.24 mT. Its width is dis-
torted by the large field modulation used to observe the
main signal in Fig. 1. This small signal, which never
amounts to more than 0.1% of the total signal, disap-
pears from view as the NaCl formed as a product in the
reaction settles to the bottom of the EPR tube. Based
on its g-value and linewidth, we believe this radical is
an insoluble Na+ or Cl− salt of an organic radical ion.
The EPR line in Fig. 1 reveals a series of ripples
which can be resolved into a distinct hyperfine splitting
pattern by taking the second derivative of the EPR
signal. The second derivative of the EPR signal for the
neophyl system reveals an unsymmetrical hyperfine
splitting pattern with nine or ten easily identifiable lines
(Fig. 2), whereas the spectrum for the neopentyl system
exhibits an unsymmetrical pattern with 12 or 13 lines
(Fig. 3). The unsymmetrical shape of the main EPR
signal, Figs. 2 and 3, could arise from a superposition
of signals from more than one radical with different
g-values or it could be caused by the lack of rotational
averaging of the anisotropic component of the hy-
(1)
The strong Lewis base tetramethylethylenediamine also
retarded the rate at which the intermediate was con-
verted to the gallium(I) product. The EPR spectra are
consistent with the presence of radicals as gallium
clusters. The EPR signals grew in intensity, decreased,
and then finally disappeared after 6–10 h. A typical
EPR spectrum for the neophyl system is shown in Fig.
1. All observations suggest that the main signal arises
from the presence of gallium based radicals rather than
from hydrocarbon radicals. The g-value is approxi-
mately 2.07 and the peak-to-peak first derivative
linewidth is 14 mT, whereas g-values for the hydrocar-
bon radicals which might be possible for these systems,
−
’
’
and PhMe2CCH2 , occur in the range of
C10H8
Fig. 3. EPR spectrum (second derivative) for the decomposition of
the yellow intermediate formed from reaction of Ga(CH2CMe3)2Cl
with Na(C10H8).
Fig. 1. EPR signal as Na2{C10H8[Ga(CH2CMe2Ph)2Cl]2} decom-
posed at room temperature.

34
O.T. Beachley et al. / Journal of Organometallic Chemistry 582 (1999) 32–39
complex to be resolved. The observed coupling con-
’
stants for [Ga(CH2CMe2R)2]5 (R=Me, Ph) agree well
’
with that found for the Ga4 radical {Ga[C(SiMe3)3]}4
and is about half of that found for the Ga2 radicals,
Ga2[Si(t-Bu)3]3 [15] and [Li(12-crown-4)2]{Ga2[(2,4,6-
’
’
(i-Pr)3C6H2]4} , [16] as expected.
1. Experimental
1.1. General
Compounds described in this investigation were ex-
tremely sensitive to oxygen and moisture and were
handled with a standard vacuum line or under a
purified argon atmosphere in a Vacuum Atmospheres
drybox. The starting compounds Ga(CH2CMe2Ph)2Cl
[17], Ga(CH2CMe2Ph)Cl2 [17], Ga(CH2CMe3)2Cl [11],
and Ga(CH2CMe3)Cl2 [11] were prepared and purified
by literature methods. Elemental analyses were per-
formed by either Schwarzkopf Microanalytical Labora-
tory, Woodside, NY or E&R Microanalytical
Laboratory, Corona, NY. IR spectra of Nujol mulls
between CsI plates were recorded by means of a
Fig. 4. Computer simulated EPR spectrum of overlapping signals for
[GaR]5 (8%) and (unresolved) (92%), one unpaired electron, A=2.1
mT.
’
’
perfine splittings. Low-temperature EPR experiments
tested the hypothesis based on the lack of rotational
averaging but failed to show any evidence of hyperfine
anisotropy. Thus, the asymmetry in the signal is be-
lieved to arise from a mixture of radicals. If the experi-
mental spectrum arises from a superposition of only
two spectra, it may be simulated by combining the
spectrum of a single Gaussian shaped line with a peak-
to-peak width of 14 mT (92%) with the spectrum of five
equivalent gallium nuclei, A=2.1 mT (8%) [14] as
shown in Fig. 4. The even number of lines in the
hyperfine pattern indicates that an odd number of
gallium nuclei are involved in the cluster and at least
five nuclei are required to give the number of observed
lines. Five equivalent gallium nuclei (I=3/2) should
give 16 lines while 9–13 lines are readily observed.
Since the species responsible for the EPR signal are
transient intermediates in low concentration, their iden-
tity cannot yet be confirmed chemically. Thus, we sug-
gest one plausible hypothesis to explain the above
observations. The decomposition of Na2{C10H8-
[Ga(CH2CMe2R)2Cl]2} might involve the formation of
1
Perkin–Elmer 683 spectrometer. The H-NMR spectra
were recorded at either 300 or 400 MHz by using either
a Varian Gemini 300 or a Varian VXR-400 spectrome-
ter, respectively. Proton chemical shifts (l) were refer-
enced to C6H6 at 7.15 ppm and SiMe4 at 0.00 ppm. All
samples for NMR spectra were contained in sealed
NMR tubes. Melting points were observed for sealed
capillaries under purified argon.
1.2. Reaction of Ga(CH2CMe2Ph)2Cl with Na[C10H8]
In a typical experiment, a side-arm dumper with
Ga(CH2CMe2Ph)2Cl (1.87 g, 5.03 mmol) was attached
to a reaction flask charged with finely cut sodium metal
(0.116 g, 5.03 mmol) and naphthalene (0.645 g, 5.03
mmol). After THF (50 ml) was vacuum distilled into
the reaction flask, the reaction mixture was stirred for
18 h to form the dark green sodium naphthalenide
solution. Then, a small amount of THF (10–20 ml) was
transferred by vacuum distillation from the flask to the
side-arm dumper. The solutions in the reaction flask
and the side-arm dumper were simultaneously cooled to
−78°C and the Ga(CH2CMe2Ph)2Cl/THF solution was
added to the sodium naphthalenide solution over 30
min. The resulting solution was stirred for 18 h at
−78°C. The color of the solution changed from brown/
green to bright yellow. No precipitate was observed. As
the yellow solution was slowly warmed to ambient
temperature, a large amount of a colorless precipitate
formed between −10 and 0°C. Then, the solution
slowly changed from bright yellow to dark brown over
2–5 h at ambient temperature. The solution was stirred
’
neutral Ga(CH2CMe2R)2 radicals by homolytic cleav-
age of the bond between gallium and the dihydronaph-
thalene moiety. The majority of these neutral radicals
could undergo ligand redistribution and/or dispropor-
tionation reactions to form the observed products
Ga(CH2CMe2R) and Ga(CH2CMe2R)3 whereas a small
concentration of these radicals could form clusters.
Those clusters with an odd number of units
’
[Ga(CH2CMe2R)2]n (n=1, 3, 5, 7,…) would be para-
magnetic and would be expected to be observed by
EPR spectroscopy. If the clusters with n=1 and 3 are
too reactive to be observed, the cluster with n=5 could
be the 8% of the observed radical with the remaining
92% being accounted for by clusters with n\5 (n=7,
9, 11,…) whose spectra would be expected to be too

O.T. Beachley et al. / Journal of Organometallic Chemistry 582 (1999) 32–39
35
for an additional 12 h followed by the removal of THF
by vacuum distillation. The material remaining after all
visible THF had been removed was subjected to dy-
namic vacuum for 24 h. These volatile materials were
weighed in a small trap (1.41 g) and identified by
1H-NMR spectroscopy as C10H8 (0.482 g, 3.76 mmol,
74.8% recovered based on the initial amount of C10H8),
and THF. The nonvolatile material was separated into
benzene soluble and insoluble fractions after 3–4 ex-
tractions. The dark brown, soluble product was iden-
tified as an equimolar mixture of Ga(CH2CMe2Ph) and
agent. (3) The yellow solutions with lithium decom-
posed more slowly (4–6 h) to the dark brown solution
than those with sodium (2–5 h).
1.4. ‘In flask’ acid hydrolysis of the products from
reactions of Ga(CH2CMe2Ph)2Cl with n(Na/C10H8),
n(Li/C10H8), or n(Li2/(C10H8)) (n=1, 2)
Reduction reactions of Ga(CH2CMe2Ph)2Cl with
Na[C10H8] and Ga(CH2CMe2Ph)2Cl with Li[C10H8] and
(Li2[C10H8]) were performed as previously described,
except that the solutions were stirred for 2–3 days to
insure complete decomposition of the yellow intermedi-
ate before hydrolysis. After removal of the THF and
other materials volatile at r.t. by dynamic vacuum
distillation for 2–3 days, the dark brown product was
reacted with dilute aqueous HCl for 2–4 days in a
60–120°C oil bath. Then the reaction flask was cooled
to −196°C and the H2 was measured with a Toepler
pump/gas buret assembly.
1
Ga(CH2CMe2Ph)3 by H-NMR spectroscopy. The ben-
zene insoluble, light brown solid was identified as im-
pure NaCl (0.276 g, 4.72 mmol, 93.9% yield based on
Na). The reaction of Ga(CH2CMe2Ph)2Cl with
Na[C10H8] at −78°C was repeated multiple times and
identical observations and results were obtained.
The dark brown soluble product mixture of
Ga(CH2CMe2Ph) and Ga(CH2CMe2Ph)3 was dissolved
in 3–5 ml of C6H6, chromatographed twice with a
Bio-Beads® column and then recrystallized from pen-
tane/benzene ( ca. 50 ml/ca. 10 ml, respectively) at
−78°C. The recrystallized product was a dark brown,
vitreous solid and identified as neophylgallium(I),
[Ga(CH2CMe2Ph)]n (0.313 g, 1.54 mm, 60.7% yield
based on Ga(CH2CMe2Ph)2Cl, Eq. 1).
Ga(CH2CMe2Ph)2Cl (1.86 g, 5.01 mmol), Na (0.116
g, 5.05 mol), C10H8 (0.645 g, 5.03 mmol) yielded 2.38
mmol H2 (94.8%). Ga(CH2CMe2Ph)2Cl (1.34 g, 3.61
mmol), Li (0.0254 g, 3.66 mmol), C10H8 (0.470 g, 3.67
mmol)
yielded
1.42
mmol
H2
(78.7%).
Ga(CH2CMe2Ph)2Cl (1.49 g, 4.01 mmol), Li (0.0278 g,
4.01 mmol), C10H8 (0.257 g, 2.01 mmol) yielded 1.54
mmol H2 (76.8%).
1.2.1. [Ga(CH2CMe2Ph)]n
M.p. 69.3–82.4°C (dec.). H-NMR (C6D6, l): 7.42,
1
7.25, 7.15, 7.10, 7.03 (all of the lines are part of br–m,
–C6H5)], 1.89 (s), 1.80 (s), 1.73 (s), 1.68 (s), 1.54 (s),
1.40 (s), 1.38 (s), 1.21 (s) (all of the lines are part of
br–m, –CMe2 and –CH2– of [Ga(CH2CMe2Ph)]n)],
0.87 (br, small). IR (Nujol, cm−1): 3100 (s), 3080 (s),
3050 (vs), 3010 (vs), 1938 (m), 1861 (m), 1792 (m), 1739
(m), 1595 (s), 1578 (m), 1490 (vs), 1300 (m), 1272 (vs),
1268 (vs), 1191 (s), 1185 (s), 1152 (m), 1140 (m), 1129
(m), 1082 (s), 1070 (s), 1025 (vs), 1010 (sh, s), 955 (m),
845 (s), 795 (vs), 781 (sh, s), 759 (vs), 730 (m), 718 (s),
690 (vs), 675 (sh, m), 622 (m), 601 (m), 550 (m), 500
(m), 470 (m), 442 (w). Anal. Calc.: C, 59.19; H, 6.45;
Ga, 34.36. Found: C, 58.76; H, 6.93; Ga, 34.91. Solubil-
ity: soluble in THF and benzene; slightly soluble in
pentane.
1.5. Identification of yellow intermediate from reaction
of Ga(CH2CMe2Ph)2Cl with Na[C10H8] by reaction
with anhydrous HCl
Sodium naphthalenide (0.0358 g, 1.56 mmol Na;
0.200 g, 1.56 mmol C10H8) was reacted with
Ga(CH2CMe2Ph)2Cl (0.574 g, 1.55 mmol) in 50 ml of
THF at −78°C as previously described to form a clear,
bright yellow solution (no precipitate) after 18 h. The
yellow solution was cooled to −196°C and then anhy-
drous HCl (0.066 g, 1.8 mmol) was vacuum distilled
into the reaction flask. The solution was slowly warmed
to ambient temperature and the solution changed from
bright yellow to colorless in 30 min. After the solution
was stirred for 12 h, the THF was removed by vacuum
distillation. Then, dynamic vacuum for 24 h at r.t. was
used to transfer the remaining volatile materials into a
small, weighable trap. The contents (0.320 g) were
identified by 1H-NMR spectroscopy as C10H10 (1,2- and
1,4-dihydronaphthalene), C10H8 (naphthalene) and
THF. The nonvolatile material remaining in the flask
was separated into pentane soluble and insoluble frac-
tions by extraction through a fine glass frit. The insolu-
ble colorless product was NaCl (0.058 g, 0.99 mmol,
64% yield based on Na). The soluble colorless solid
was identified as Ga(CH2CMe2Ph)2Cl (0.538 g, 1.45
mmol, 93.5% yield based on the initial amount of
1.3. Reaction of Ga(CH2CMe2Ph)2Cl with Li[C10H8] or
Li2[C10H8]
Lithium naphthalenide Li[C10H8] or dilithium naph-
thalenide
Li2[C10H8]
were
reacted
with
Ga(CH2CMe2Ph)2Cl as described above. The following
observations differentiated the reduction reactions with
lithium. (1) The yellow intermediates in the lithium
reactions formed in 10–20 min after mixing reagents
whereas 10–20 h was needed for sodium. (2) No precip-
itate formed when lithium was used as the reducing

36
O.T. Beachley et al. / Journal of Organometallic Chemistry 582 (1999) 32–39
Ga(CH2CMe2Ph)2Cl) by comparison of its melting
point and H-NMR spectrum with that of an authentic
1.6.2. Spectrum after 4 h
1
Solution color: light reddish/brown. 7.78 (m, C10H8),
7.40 (m, C10H8), 7.32 (d, o-Ar (–C6H5, Li2{C10H8-
[Ga(CH2CMe2Ph)2Cl]2}), 7.12 (m, –C6H5, Ga(CH2-
CMe2Ph)3+[Ga(CH2CMe2Ph)]n), 7.07 (m, m-Ar
(–C6H5, Li2{C10H8[Ga(CH2CMe2Ph)2Cl]2}), 6.90 (m, p-
Ar (–C6H5, Li2{C10H8[Ga(CH2CMe2Ph)2Cl]2}), 6.48
(br, 1,4- and 1,2-C10H10), 3.58 (m, THF), 1.74 (m,
THF), 1.34 (s, –CMe2, Li2{C10H8[Ga(CH2CMe2Ph)2-
Cl]2}), 1.18 (s, –CMe2, Ga(CH2CMe2Ph)3), 0.88 (S,
–CH2–, Li2{C10H8[Ga(CH2CMe2Ph)2Cl]2}), 0.63 (s, –
CH2–, Ga(CH2CMe2Ph)3).
sample [17].
1.5.1. Colorless solid—Ga(CH2CMe2Ph)2Cl
M.p. 83.0–85.3°C (lit. [17] m.p. 86.4–87.8°C). 1H-
NMR (C6H6, l): 1.40 (s, –CH2– (Ga(CH2CMe2-
Ph)2Cl)), 1.33 (s, –CMe3 (Ga(CH2CMe2Ph)2Cl)).
Solubility: soluble in THF, pentane, and benzene.
1.5.2. Volatile materials
1H-NMR (CH2Cl2, l): 7.93 (m, C10H8), 7.57 (m,
C10H8), 7.17 (m, C10H10), 6.74 (m, C10H10), 6.38 (m, s,
C10H10), 6.10 (m, C10H10), 3.77 (m, THF), 3.46 (s,
C10H10), 2.43 (m, C10H10), 1.93 (m, THF).
1.6.3. Spectra after 5, 15 and 25 days
Solution color: reddish/brown. 7.78 (m, C10H8), 7.40
(m, C10H8), 7.11 (m, –C6H5, (Ga(CH2CMe2Ph)3+
[Ga(CH2CMe2Ph)]n), 6.97 (m, –C6H5, Ga(CH2CMe2-
Ph)3+[Ga(CH2CMe2Ph)]n), 3.58 (m, THF), 1.74 (m,
THF), 1.19 (s, –CMe2, Ga(CH2CMe2Ph)3), 0.64 (s,
–CH2–, Ga(CH2CMe2Ph)3).
1.6. Identification of the yellow intermediate from
reaction of Ga(CH2CMe2Ph)2Cl with Li2[C10H8] by
1H-NMR spectroscopy
Lithium metal (0.0348 g, 5.01 mmol), naphthalene
(0.321 g, 2.51 mmol) and 20 ml of THF were reacted to
form the dark purple Li2[C10H8] solution (0.25 M).
Then, 2 ml (0.501 mmol) of Li2[C10H8]/THF was added
to 2–3 ml of a solution of Ga(CH2CMe2Ph)2Cl (0.186
g, 0.501 mmol) in d8-THF at −78°C. The bright yellow
solution was stirred for ca. 30 min and then a sample of
the solution was poured into the NMR tube and cooled
1.6.4. EPR spectroscopic studies of the decomposition
of the yellow intermediates,
M2{C10H8[Ga(CH2CMe2R)2Cl]2}), (M=Na, Li;
R=Ph, Me) in THF solution
EPR spectra were recorded with an IBM/Bruker
X-band EPR spectrometer model ER200D-SRC by us-
ing a TE102 mode rectangular cavity model 4102ST.
The sample temperature was controlled with an IBM/
Bruker ER 4111VT temperature controller. The spec-
trometer was interfaced to a Nicolet 2090 digital
oscilloscope for digital recording of spectra. The g-val-
ues were measured with a Magnion G-502 Gaussmeter
and a Hewlett Packard 5245L frequency counter by
using the procedure described previously [18]. The pre-
cision of the g-value measurements is limited to90.002
by the broad lines encountered with gallium radicals.
The concentration of an observed radical was estimated
by comparison of the signal intensity with the signal
intensity for a known concentration of the stable free
radical 2,2-diphenyl-1-picryl hydrazyl (DPPH) in the
same size sample tube. The first-derivative EPR spec-
trum of a 2.7×10−4 M DPPH solution was obtained
by using 100 kHz Zeeman field modulation of 0.08 mT
and a microwave power of 5 mW. First-derivative
spectra of the gallium radicals were obtained with field
modulation of 0.8 mT and microwave power of 20 mW.
The area under the EPR signal was determined by
double integration by using the program STICKS [19] on
an IBM-PC/XT. The signal to mol ratio obtained for
DPPH was recalculated for the gallium radical condi-
tions before making the comparison.
1
to −78°C. The H-NMR spectrum of the yellow solu-
tion was recorded immediately after warming to r.t.
and at 45 min, 4 h, 5, 15 and 25 days thereafter. The
spectra at 5, 15 and 25 days were identical. No precipi-
tate formed in the NMR tube.
1.6.1. Spectrum immediately after warming to ambient
temperature
Solution color: bright yellow. 1H-NMR (d8-THF/
THF, l): 7.78 (m, C10H8), 7.42 (m, C10H8), 7.32 (d,
o-Ar (–C6H5, Li2{C10H8[Ga(CH2CMe2Ph)2Cl]2}), 7.08
(m, m-Ar (–C6H5, Li2{C10H8[Ga(CH2–CMe2Ph)2-
Cl]2}), 6.91 (m, p-Ar(–C6H5, Li2{C10H8[Ga(CH2-
CMe2Ph)2Cl]2}), 6.66 (br, (1,4- and 1,2-C10H10), 3.58
(m, THF), 1.73 (m, THF), 1.31 (s, –CMe2,
Li2{C10H8[Ga(CH2CMe2Ph)2Cl]2}), 0.88 (s, –CH2–,
Li2{C10H8[Ga(CH2CMe2Ph)2Cl]2}), solution color: dark
yellow. 7.78 (m, C10H8), 7.40 (m, C10H8), 7.32 (d, o-Ar
(–C6H5, Li2{C10H8[Ga(CH2CMe2Ph)2Cl]2}), 7.12 (m, –
C6H5, Ga(CH2CMe2Ph)3+[Ga(CH2CMe2Ph)]n), 7.07
(m, m-Ar (–C6H5, Li2{C10H8[Ga(CH2CMe2Ph)2Cl]2}),
6.90 (m, p-Ar(–C6H5, Li2{C10H8[Ga(CH2–CMe2Ph)2-
Cl]2}), 6.68 (br, 1,4- and 1,2-C10H8), 3.58 (m, THF),
1.74 (m, THF), 1.31 (s, –CMe2, Li2{C10H8[Ga(CH2-
CMe2Ph)2Cl]2}), 1.18 (s, –CMe2, Ga(CH2CMe2Ph)3),
0.88 (s, –CH2–, Li2{C10H8[Ga(CH2CMe2Ph)2Cl]2}),
0.63 (s, –CH2–, Ga(CH2CMe2Ph)3).
In
a
typical experiment, the gallium reagent
Ga(CH2CMe2R)2Cl (R=Ph, Me) was added to a THF
solution of the alkali metal naphthalenide at −78°C to

O.T. Beachley et al. / Journal of Organometallic Chemistry 582 (1999) 32–39
37
form the bright yellow intermediate. Details are given
below. Then, a portion of the yellow solution at −
78°C was poured into a quartz EPR tube which had
been precooled to −78°C and then flame sealed. After
the sample was rapidly warmed to r.t., EPR spectra
(first and second derivative spectra, as appropriate)
were recorded. In all cases the EPR signal was observed
to grow in intensity with time, decrease until a broad
unresolved signal was apparent and then finally disap-
pear. The solution changed color from bright yellow to
brown. When sodium was used, a precipitate appeared
during the decomposition. However, when lithium was
the reducing agent, no precipitate was observed. The
following give quantities of reagents, specific experi-
mental observations and details of the EPR spectra for
each reagent combination.
small, broad, unresolved signal, whereas no second
derivative spectrum was detected; all other experimental
observations were identical to those described for the
previous experiment.
1.6.8. Ga(CH2CMe2Ph)2Cl—sodium (second-deri6ati6e
spectra at low temperature)
Ga(CH2CMe2Ph)2Cl 2.29 g, 6.18 mmol), Na (0.142 g,
6.19 mmol), C10H8 (0.795 g, 6.20 mmol), THF (23 ml),
initial gallium concentration (0.27 M); the second-
derivative spectra of the yellow solution were recorded
at ambient temperature until the signal reached a max-
imum intensity at which time the tube was cooled to
−50°C; the spectra were recorded at −50, −25, 0 and
20°C; the unresolved signal, g=2.06790.002, re-
mained after 3–6 h at 20°C; no gallium radicals were
observed after an additional 6–10 h. The second-
derivative spectrum at 20°C before cooling was a well
resolved signal with an unsymmetrical splitting pattern
with the center of the hyperfine pattern at g=2.0389
0.002, Aave(Ga)=2.2 mT. The second-derivative spec-
trum at −50 and −25°C depicted only an unresolved
spectrum. At 0°C, a slightly resolved signal was ob-
served. At 20°C, the spectrum depicted the same well
resolved signal as observed previously. At all tempera-
tures, the overall lineshape of the signal is the same and
the relative ratios of the individual hyperfine lines, once
resolved, appear to be independent of temperature.
1.6.5. Ga(CH2CMe2Ph)2Cl—sodium
Ga(CH2CMe2Ph)2Cl (1.88 g, 5.07 mmol), Na (0.117
g, 5.07 mmol), C10H8 (0.650 g, 5.07 mmol), THF (18
ml), the initial gallium concentration (0.28 M); a yellow
solution formed after 18 h at −78°C; the time for the
EPR signal to reach maximum intensity was 2–3 h; the
first derivative spectrum was a broad unresolved signal
with a peak-to-peak distance of 13.8 mT and g=
2.06790.002; the second derivative spectrum was a
broad but well resolved signal with an unsymmetrical
hyperfine splitting pattern. The center of the hyperfine
pattern falls at g=2.03890.002, Aave(Ga)=2.2 mT;
maximum observed concentration of gallium radicals
was 2.5×10−5 M; broad unresolved signal remained
after 3–6 h; no gallium radicals were observed after an
additional 6–10 h.
1.6.9. Ga(CH2CMe2Ph)2Cl—sodium with added
TMEDA
Ga(CH2CMe2Ph)2Cl (0.613 g, 1.65 mmol), Na
(0.0381 g, 1.66 mmol), C10H8 (0.213 g, 1.66 mmol),
THF (30 ml), initial gallium concentration (0.06 M); a
yellow solution formed after 18 h at −78°C. The
bright yellow solution was cooled to −196°C and
TMEDA (2.45 g, 21.1 mmol) was added to the flask.
The resulting solution was warmed to −78°C and a
heavy, colorless precipitate formed. After stirring for 12
h, a sample was poured into the quartz EPR tube and
the tube was flame-sealed. The tube was warmed to r.t.
and inserted into the EPR spectrometer. However, no
gallium radicals were observed. The solution remained
yellow with a colorless precipitate for more than 5 days
before slowly turning brown. A signal for an unknown
organic radical at a g-value of 2.00390.001 was ob-
served in the precipitate (NaCl).
1.6.6. Ga(CH2CMe2Ph)2Cl—lithium
Ga(CH2CMe2Ph)2Cl (1.86 g, 5.01 mmol), Li (0.0349
g, 5.03 mmol), C10H8 (0.645 g, 5.03 mmol), THF (18
ml), initial gallium concentration (0.28 M); a yellow
solution formed in less than 1 min but the solution was
stirred at −78°C for 12 h before an EPR spectal study;
the time for the EPR signal to reach maximum intensity
was 2–3 h; the concentration of gallium radicals was so
low that only the g-value from the first derivative
spectrum could be accurately measured; the first deriva-
tive spectrum was a broad unresolved signal with a
peak-to-peak distance of 15.4 mT and g=2.0689
0.002; the second derivative spectrum was a small,
broad and unresolved signal; the broad unresolved
signal remained after 3–6 h; no gallium radicals were
observed after an additional 6–12 h.
1.6.10. Ga(CH2CMe3)2Cl—sodium
Ga(CH2CMe3)2Cl (1.24 g, 5.01 mmol), Na (0.115 g,
5.02 mmol), C10H8 (0.644 g, 5.02 mmol), THF (18 ml),
initial gallium concentration (0.28 M); a yellow solution
with slight precipitate formed after 18 h at −78°C; the
time for EPR signal to reach maximum intensity was
2–3 h; the first-derivative spectrum was broad unre-
solved signal with a peak to peak distance of 14.7 mT
1.6.7. Ga(CH2CMe2Ph)2Cl—lithium
Ga(CH2CMe2Ph)2Cl (1.57 g, 4.22 mmol), Li (0.0293
g, 4.22 mmol), C10H8 (0.272 g, 2.12 mmol), THF (14
ml), initial gallium concentration (0.30 M); the first
derivative spectrum of the yellow solution exhibited a

38
O.T. Beachley et al. / Journal of Organometallic Chemistry 582 (1999) 32–39
and g=2.06390.002; the second-derivative spectrum
was a broad but well-resolved signal with an unsym-
metrical hyperfine splitting pattern; the center of the
hyperfine pattern falls at g=2.03690.002, Aave(Ga)=
2.0 mT; maximum observed concentration of gallium
radicals was 1.2×10−4 M; a broad unresolved signal
remained after 3–6 h; no gallium radicals were ob-
served after an additional 6–10 h.
vations were made when 2-methyl-tetrahydrofuran was
used instead of THF.
1.6.14. Ga(CH2CMe3)2Cl—sodium with added
TMEDA
Ga(CH2CMe3)2Cl (0.312 g, 1.50 mmol), Na (0.0346
g, 1.51 mmol), C10H8 (0.194 g, 1.51 mmol), THF (30
ml), initial gallium concentration (0.05 M); a yellow
solution formed after 18 h at −78°C. The bright
yellow solution was cooled to −196°C and TMEDA
(3.46 g, 29.7 mmol) was added to the flask. The result-
ing solution was warmed to −78°C and a heavy,
colorless precipitate formed. After stirring for 12 h, a
sample was poured into the quartz EPR tube and the
tube was flame-sealed. The tube was warmed to r.t. and
inserted into the EPR spectrometer. No gallium radi-
cals were observed. The solution remained yellow with
a colorless precipitate for more than 5 days before
slowly turning brown. A signal for an unknown organic
radical at a g-value of 2.00390.001 was observed in
the precipitate (NaCl).
1.6.11. Ga(CH2CMe3)2Cl—lithium
Ga(CH2CMe3)2Cl (1.55 g, 6.28 mmol), Li (0.0437 g,
6.30 mmol), C10H8 (0.808 g, 6.30 mmol), THF (23 ml),
initial gallium concentration (0.27 M); a yellow solution
formed after 15 min at −78°C but the solution was
stirred for 12 h before an EPR spectral study; the time
for EPR signal to reach maximum intensity was 1–2 h;
the concentration of gallium radicals was so low that
only the g-value from the first-derivative spectrum
could be accurately measured; the first-derivative spec-
trum was a broad unresolved signal with a peak to peak
distance of 12.3 mT and g=2.06290.002; the second-
derivative spectrum was a small, broad and unresolved
signal; no gallium radicals were observed after an addi-
tional 6–10 h.
References
[1] W. Uhl, Angew. Chem. Int. Ed. Engl. 32 (1993) 1386.
[2] C. Dohmeier, D. Loos, H. Schno¨ckel, Angew. Chem. Int. Ed.
Engl. 35 (1996) 129.
1.6.12. Ga(CH2CMe3)2Cl—lithium
Ga(CH2CMe3)2Cl (1.57 g, 6.33 mmol), Li (0.044 g,
6.3 mmol), C10H8 (0.406 g, 3.17 mmol), THF (23 ml),
initial gallium concentration (0.28 M); the first-deriva-
tive spectrum of the yellow solution exhibited a broad,
unresolved signal with a peak to peak distance of 11.7
mT and g=2.06690.002; the second-derivative spec-
trum was a small, broad and unresolved signal; all
other experimental observations were identical to those
described for the previous experiment.
[3] G. Linti, J. Organomet. Chem. 520 (1996) 107.
[4] W. Ko¨stler, G. Linti, Angew. Chem. Int. Ed. Engl. 36 (1997)
2644.
[5] R.D. Schuter, A.H. Cowley, D.A. Atwood, R.A. Jones, M.R.
Bond, C.J. Carrano, J. Am. Chem. Soc. 115 (1993) 2070.
[6] X.-W. Li, W.T. Pennington, G.H. Robinson, J. Am. Chem. Soc.
117 (1995) 7578.
[7] X.-W. Li, Y. Xie, P.R. Schreiner, K.D. Gripper, R.C. Critten-
don, C.F. Campana, H.F. Schaefer, G.H. Robinson,
Organometallics 15 (1996) 3798.
[8] J. Su, X.-W. Li, R.C. Crittendon, G.H. Robinson, J. Am. Chem.
Soc. 119 (1997) 5471.
[9] P. Jutzi, B. Neumann, G. Reumann, H.-G. Stammler,
Organometallics 17 (1998) 1305.
[10] W. Uhl, A. Jantschak, J. Organomet. Chem. 555 (1998) 263.
[11] O.T. Beachley Jr., P.C. Pazik, M.J. Noble, Organometallics 13
(1994) 2885.
1.6.13. Ga(CH2CMe3)2Cl—sodium (second-deri6ati6e
spectra at low temperature)
Ga(CH2CMe3)2Cl (1.98 g, 8.00 mmol), Na (0.184 g,
8.02 mmol), C10H8 (1.03 g, 8.02 mmol), THF (30 ml),
initial gallium concentration (0.27 M); the second-
derivative spectra of the yellow solution were recorded
at ambient temperature until signal reached maximum
intensity at which time the tube was cooled to −50°C;
spectra were recorded at −50, −35, −20, −5, 15
and 20°C; a broad unresolved signal remained after
1–3 h at 20°C; no gallium radicals were observed after
an additional 6–10 h. The second-derivative spectrum
at 15°C before cooling was a well-resolved signal with
an unsymmetrical splitting pattern. At −50 and −
35°C, the signal for the gallium radicals had an unre-
solved spectrum whereas at −5 and 15°C the signal
was well resolved with an unsymmetrical hyperfine
splitting pattern; after 1 h at 15°C the hyperfine split-
ting pattern could no longer be observed. Similar obser-
[12] A.J. Gordon, R.A. Ford, The Chemist’s Companion, Wiley,
New York, 1972, pp. 336–339.
[13] E.J. Hamilton, H. Fisher, Helv. Chim. Acta 2 (1973) 795.
[14] Gallium has two important isotopes 69Ga (60.4% abundant) and
71Ga (39.6% abundant) with nuclear spins of I=3/2. 71Ga is
expected to have a 27% larger hyperfine coupling constant than
69Ga for the same electron density. Thus, the hyperfine splitting
pattern for a radical with five or seven equivalent Ga nuclei is
expected to be very complex. In the present case, the component
linewidth is considerably greater than the difference in the 69Ga
and 71Ga hyperfine splittings so that only an average splitting is
observed. We have simulated spectra both with a single weighted
average coupling constant and with two separate coupling con-
stants for the gallium isotopes. The simulations are indistinguish-
able. The values reported are for the best match of the single
weighted average coupling simulations. The separate isotopic
hyperfine values can be calculated as 69A=Aave0.9033 and
71A=Aave1.148.

O.T. Beachley et al. / Journal of Organometallic Chemistry 582 (1999) 32–39
39
[15] X. He, R.A. Bartlett, M.M. Olmstead, K. Ruhlandt-Senge, B.E.
Sturgeon, P.P. Power, Angew. Chem. Int. Ed. 32 (1993) 717.
[16] N. Wiberg, K. Amelunxen, H. No¨th, H. Schwenk, W. Kaim, A.
Klein, T. Scheiring, Angew. Chem. Int. Ed. Engl. 36 (1997) 1213.
[17] O.T. Beachley Jr., M.J. Noble, M.R. Churchill, C.H. Lake,
Organometallics 11 (1992) 10.
[18] R.D. Allendoerfer, J. Chem. Phys. 55 (1971) 3615.
[19] R.D. Allendoerfer, R.J. Male, J. Phys. Chem. 92 (1988) 6237.
.