Addressing the chemical sorcery of "GaI": Benefits of solid-state analysis aiding in the synthesis of P→Ga coordination compounds

Article abstract of DOI:10.1021/ic501139w

The differing structures and reactivities of "GaI" samples prepared with different reaction times have been investigated in detail. Analysis by FT-Raman spectroscopy, powder X-ray diffraction, 71Ga solid-state NMR spectroscopy, and 127I nuclear quadrupole

Full text of DOI:10.1021/ic501139w

Article
pubs.acs.org/IC
Addressing the Chemical Sorcery of “GaI”: Benefits of Solid-State
Analysis Aiding in the Synthesis of P→Ga Coordination Compounds
Brian J. Malbrecht, Jonathan W. Dube, Mathew J. Willans, and Paul J. Ragogna*
The Department of Chemistry and The Centre for Advanced Materials and Biomaterials Research (CAMBR), The University of
Western Ontario, 1151 Richmond Street, London, Ontario N6A 5B7, Canada
S
* Supporting Information
ABSTRACT: The differing structures and reactivities of “GaI”
samples prepared with different reaction times have been
investigated in detail. Analysis by FT-Raman spectroscopy,
powder X-ray diffraction, 71Ga solid-state NMR spectroscopy,
and 127I nuclear quadrupole resonance (NQR) provides
concrete evidence for the structure of each “GaI” sample
prepared. These techniques are widely accessible and can be
implemented quickly and easily to identify the nature of the
“GaI” in hand. The “GaI” prepared from exhaustive reaction
times (100 min) is shown to possess Ga2I3 and an overall
formula of [Ga0]2[Ga+]2[Ga2I62−], while the “GaI” prepared
with the shortest reaction time (40 min) contains GaI2 and has
the overall formula [Ga0]2[Ga+][GaI4 ]. Intermediate “GaI”
−
samples were consistently shown to be fractionally composed of each of these two preceding formulations and no other
distinguishable phases. These “GaI” phases were then shown to give unique products upon reactions with the anionic
bis(phosphino)borate ligand class. The reaction of the early-phase “GaI” gives rise to a unique phosphine Ga(II) dimeric
coordination compound (3), which was isolated reproducibly in 48% yield and convincingly characterized. A base-stabilized
GaI→GaI3 fragment (4) was also isolated using the late-phase “GaI” and characterized by multinuclear NMR spectroscopy and
X-ray crystallography. These compounds can be considered unique examples of low-oxidation-state P→Ga coordination
compounds and possess relatively long Ga−P bond lengths in the solid-state structures. The anionic borate backbone therefore
results in interesting architectures about gallium that have not been observed with neutral phosphines.
In most cases, the synthesis of low-valent gallium compounds
begins with the Ga(I) synthon “GaI”.25 While the synthesis of
“GaI” has been well established,26,27 there has been some recent
debate regarding the true composition of this reagent. A report
by Coban has been commonly cited as assigning “GaI” to have
the dominant composition Ga2I3 (in the solid state:
[Ga+]2[Ga2I6]2−) with the possible presence of other gallium
subiodides.28 However, this report is inaccessible, which makes
data comparisons and analysis impossible. Contradicting the
Coban proposal is a solid-state NMR (ssNMR) investigation by
Bryce and co-workers which has recently suggested that “GaI”
is best assigned as [Ga0]2[Ga]+[GaI4]− (where [Ga]+[GaI4]− is
also described as GaI2).29 To attempt to resolve this
discrepancy and to provide the necessary data for stand-
ardization of the “GaI” synthesis across different laboratories,
we undertook a full characterization of “GaI” samples at several
different stages of its preparation. Analysis by numerous
techniquesincluding FT-Raman spectroscopy, 71Ga solid-
state NMR spectroscopy (ssNMR), 127I nuclear quadrupole
resonance (NQR), and powder X-ray diffraction (pXRD)
INTRODUCTION
■
Modern research in synthetic main-group chemistry frequently
focuses on the synthesis of compounds that possess multiple
bonds or are coordinatively unsaturated. Many of the resulting
coordination complexes have been shown to possess the
coordination space and accessible electronic structure that
allows them to perform roles traditionally associated with
transition metals.1,2 The synthesis of low-valent gallium
complexes is interesting for their unusual bonding environ-
ments (A; Figure 1) and for their potential in bond activation
and catalysis. To this end, N-heterocyclic Ga(I) compounds (B,
C) have been synthesized and shown to serve as suitable
ligands for both p- and d-block elements.3−14 The structure and
bonding of a unique series of base-stabilized cationic gallium(I)
compounds have also been recently reported by Krossing and
co-workers (D).15−20 Ga(II) complexes typically feature a Ga−
Ga bond, and one dimer in particular (E) has been shown to
catalyze the hydroamination of phenylacetylene.21,22 This
reactivity constitutes a rare example of an organic C−N bond
forming reaction catalyzed by a purely main group coordination
compound and highlights the potential of low-valent gallium
compounds to have an effect beyond the field of coordination
chemistry.23,24
Received: May 15, 2014
Published: September 3, 2014
© 2014 American Chemical Society
9644
dx.doi.org/10.1021/ic501139w | Inorg. Chem. 2014, 53, 9644−9656

Inorganic Chemistry
Article
Figure 1. Selection of neutral and charged low-valent gallium(I) compounds (A−D), a gallium(II) compound that catalyzes hydroamination
reactions (E), gallium−phosphine coordination compounds isolated from “GaI” (F−H), and compounds reported in this work (3−5).
Figure 2. Raman spectra of “GaI” as a function of reaction time: from top to bottom 40, 60, 80, and 100 min. Insets show toluene suspensions of the
“GaI” solids that correspond to the Raman spectra.
revealed that “GaI” can be controllably synthesized in two
different phases. An “early” gray phase is the first to be formed
during synthesis of “GaI”, followed by a “late” green phase that
represents the true end point of the “GaI” synthesis.
Intermediate phases containing a mixture of the early and
late phases were also isolated and identified.
Having isolated and characterized these different phases of
“GaI”, we additionally wished to demonstrate their variable
utility in the synthesis of low-valent gallium species. While the
preparation and reactivity of low-valent gallium complexes with
anionic nitrogen-based chelates have been well explored, the
coordination chemistry available from phosphorus chelates is
relatively unknown.30−32 Consequently, we chose to explore
the reactivity of the different “GaI” phases with the
bis(phosphino)borate ligands established by Peters et al., as
they have been successfully utilized in stabilizing unique low-
coordinate transition-metal and main-group centers.33−41 Using
these ligands and the two distinctly different phases of “GaI”,
we were able to isolate complexes of Ga in all its known formal
oxidation states (Ga(I), Ga(II), and Ga(III)). To the best of
our knowledge, this is the first report that significantly different
gallium products can be isolated from different batches of
“GaI”. These structures differ both from those typically formed
with the anionic nitrogen chelates as well as those formed with
simple neutral, monodentate phosphines, and, consequently,
this report constitutes a novel expansion of the chemistry of
low-valent gallium.
RESULTS AND DISCUSSION
■
Investigations into the Nature of “GaI”. The composi-
tion of “GaI” is thought to contain a variety of gallium
subiodides and also gallium metal. Information regarding
structure and Raman signatures of the gallium subiodides that
are of relevance to “GaI” is given below and is pertinent for the
following discussion.
(1) GaI2: alternatively written as Ga2I4, the bonding of GaI2
−
is best described by the formula [Ga+][GaI4 ]. The GaI4− anion
9645
dx.doi.org/10.1021/ic501139w | Inorg. Chem. 2014, 53, 9644−9656

Inorganic Chemistry
Article
Figure 3. Powder diffraction patterns of “GaI” as a function of reaction time: 40 min (orange) and 100 min (green). The uneven baseline is a result
of the Scotch tape used to prevent exposure of the sample to air.
is in a distorted-tetrahedral geometry, with the Ga+ cation being
weakly stabilized by eight iodide atoms in the unit cell.42 It
crystallizes in the R3c space group, and its Raman spectrum
features a prominent signal at 143 cm−1 and weaker signals at
214 and 235 cm−1.43
(2) Ga2I3: alternatively written as Ga4I6, though the formula
[Ga+]2[Ga2I62−] is a more descriptive representation of its
composition. The dianion [Ga2I62−], with gallium in the formal
Ga(II) oxidation state, possesses a Ga−Ga bond with all Ga−I
bonds being terminal.42 It crystallizes in the P21/c space group,
and its Raman spectrum features a very strong absorption at
124 cm−1 with weaker absorptions occurring at 292, 186, and
79 cm−1.43
(3) GaI3: structurally exists, and is sometimes written, as
Ga2I6 with two bridging and four terminal iodine atoms and no
Ga−Ga bond. The gallium atom is formally Ga(III) and is thus
distinct from [Ga2I62−]. We have obtained the Raman spectrum
of commercially available GaI3 and observed very strong
absorption at 142 cm−1 along with weaker signals at 267, 227,
194, 163, and 85 cm−1.
We prepared several batches of “GaI” using Green’s method
of sonicating from the elements, with the reaction being
stopped precisely after 40, 60, 80, and 100 min.26 It should be
noted that while the synthesis demonstrated a reasonably
reliable time course, care must be taken to ensure
reproducibility among reactions; changing the reaction vessel,
temperature, or amount of solvent can all have dramatic
influences on the rate of “GaI” conversion. The phases of “GaI”
synthesized here are stable for at least 1 year at −35 °C under
an inert atmosphere and showed no changes to the Raman
spectra or reactivity. However, early or gray phases will begin to
show a slight green color over the course of 1 week if left at
room temperature under N2. We are not certain how long this
transition would require to achieve complete conversion to the
exhaustively sonicated (green) phase, though the conversion
may be easily identified using the techniques described below.
Both phases are highly air sensitive, decomposing in minutes in
an open atmosphere. Sonication for longer reaction times
under the standard conditions gave a green powder with
characteristics identical with those of the 100 min sample.
As the gallium subiodides have been most thoroughly
characterized by Raman spectroscopy,43−47 this was a logical
entry point into the characterization of our different “GaI”
samples. The synthesis of “GaI” initially yields a product with a
strong absorption in the Raman spectrum at 141 cm−1
accompanied by weaker absorptions at 230 and 85 cm−1
(Figure 2). As the reaction is extended for longer times, the
absorptions at 230 and 141 cm−1 diminish and are replaced by a
strong absorption at 124 cm−1 and weaker absorptions at 292
and 188 cm−1. There was also a corresponding color change in
the produced powder from light gray to green (authentic “GaI”
is generally referred to as “green”). Comparison of the
vibrations observed in the Raman spectrum for the phases of
“GaI” to literature values for gallium subiodides suggests that
GaI2 ([Ga+][GaI4 ]) is the dominant gallium iodide present in
−
the early stage “GaI”, while in the late-stage “GaI” Ga2I3
([Ga+]2[Ga2I62−]) is the main gallium iodide species present.43
The resonances observed for the “GaI” samples also strongly
correlate with the Raman spectra of salts containing the
relevant GaI4 and Ga2I6 anions.48−50 While some peaks for
GaI3 do overlap with those of GaI2, the complete spectra are
quite distinct and do not match either phase prepared here.
To corroborate this proposal, the powder X-ray diffraction
(pXRD) patterns of our 40 and 100 min “GaI” samples were
obtained. The powders do not diffract strongly; however, the
observed patterns are quite distinct (Figure 3). Comparison of
the pXRD patterns obtained for the two pure phases prepared
by us to the literature patterns for pure GaI2 and Ga2I3 indicates
the expected trend; the early-phase “GaI” resembles GaI2, while
the late-phase “GaI” resembles Ga2I3 (Figures S-8 and S-9 in
the Supporting Information).42
−
2−
Therefore, on the basis of data obtained by Raman
spectroscopy, by powder diffraction, and by application of
mass and charge balance, we propose that the early-stage “GaI”
sample is largely composed of [Ga0]2[Ga+][GaI4 ] (simplified,
−
[Ga0]2[Ga2I4]), while the late-stage “GaI” sample has the
9646
dx.doi.org/10.1021/ic501139w | Inorg. Chem. 2014, 53, 9644−9656

Inorganic Chemistry
Article
composition [Ga0 ]2 [Ga+ ]2 [Ga2 I6
[Ga0]2[Ga4I6]).
]
(simplified,
frequencies and no signal was observed.53 For the late-stage
“GaI” sample, an exhaustive search led to the observation of
peaks at three distinct frequencies106.35, 107.83, and 123.54
MHzand no signal was obtained using the transmitter
frequencies where NQR peaks were observed for either GaI2 or
GaI3 (Figure 4). The observation of three distinct peaks is
consistent with the assignment of late-stage “GaI” as Ga2I3,
which has three crystallographically unique I atoms in its solid-
state structure. The 127I NQR data are therefore consistent with
the previously discussed methods and also confirm that the
early- and late-stage “GaI” phases contain only one gallium
subiodide component.
2 −
We further characterized the early and late phases of “GaI”
using 127I nuclear quadropole resonance (NQR) spectroscopy
to confirm our proposal and also identify if minor quantities of
other gallium subiodides were present.51 The frequencies of the
127I NQR peaks obtained for the early-stage (40 min) and late-
stage (100 min) “GaI” samples are summarized in Table 1 and
Table 1. 127I NQR Frequencies for the Early- and Late-Stage
a
“GaI” Samples
127I NQR freq (MHz)
The 71Ga ssNMR spectra of our “GaI” samples were
obtained to provide a direct comparison to the work of Bryce
and co-workers.29 The 71Ga ssNMR spectra of early-stage “GaI”
(40 min), which from the above characterization methods we
believe to be GaI2, acquired at two different magnetic fields are
presented in Figure 5, while the NMR spectroscopic parameters
and relative amounts are summarized in Table 2. The 71Ga
ssNMR spectra feature a sharp, strongly deshielded peak at
4484 ppm, which has been previously assigned to Ga metal
([Ga0]),29 and a broad powder pattern centered at about −400
ppm. The 71Ga powder pattern is actually a convolution of two
unique powder patterns arising from two unique Ga sites within
sample
site 1
site 2
site 3
site 4
ref
early-stage “GaI”
late-stage “GaI”
GaI2
113.69
106.35
113.65
133.69
132.04
107.83
131.94
173.65
134.39
123.54
134.27
174.59
163.71
this study
this study
52
163.71
GaI3
53
a
Frequency of the mI = 1/2 ↔ 3/2 transition at room temperature.
the spectra are shown in Figure 4. For the early stage “GaI”
sample, peaks were observed at four distinct frequencies:
113.69, 132.04, 134.39, and 163.71 MHz. These values are in
excellent agreement with those obtained from the previous 127I
NQR study of GaI2.52 To verify that no GaI3 was present, we
performed 127I NQR experiments at the known GaI3
−
the sample: one from the [GaI4 ] component and the other
from the [Ga+] component. As shown in Figure 5, each of these
sites can be independently simulated and the experimental
spectrum is effectively simulated by summing the two unique
Ga subspectra in a 1:1 ratio. The simulation reveals that one of
the Ga sites has a chemical shift of −511 ppm, a relatively small
CQ value of 1.81 MHz, and a CSA value of 85 ppm, thus giving
rise to a relatively narrow powder pattern. The other site has a
similar CSA value (80 ppm) but is significantly deshielded with
respect to site 2 (δ(71Ga) −335 ppm) and has a significantly
larger CQ value of 7.1 MHz and thus a much broader powder
pattern for this site. Past liquid NMR studies have shown that
−
the 71Ga chemical shift of [GaI4 ] varies from −505 to −450
ppm.55 Therefore, the early-stage site 1 peak, which has a shift
of −511 ppm and relatively small CQ value, can be assigned to
the [GaI4 ] site within GaI2 and the [Ga+] site assigned to the
−
broader NMR site that has a chemical shift of −335 ppm.
The 71Ga NMR spectrum of the late-stage “GaI” (100 min)
sample, which we believe to contain Ga2I3, is presented at the
bottom of Figure 6. Similar to the 71Ga ssNMR spectrum of the
early stage sample, there is a heavily deshielded peak at 4484
ppm, which corresponds to the [Ga0] site, and a powder
pattern at about −400 ppm. The shape of the powder pattern
of the late-stage sample is, however, very different from that of
the early-stage sample. Even though there should be two Ga
sites in Ga2I3the [Ga+] site and the single Ga site in the
[Ga2I62−] dimerthe powder pattern obtained for the late-
stage sample could be simulated using only one Ga site. The
parameters obtained for this site are summarized in Table 2 and
are similar to those previously determined for “GaI”.29
The 71Ga NMR spectra of the intermediate-stage (60 and 80
min) “GaI” samples are presented in Figure 6. Similar to the
case for the early-stage and late-stage samples, the intermediate-
stage samples have a strong, deshielded peak at 4484 ppm and a
powder pattern centered at about −400 ppm. The powder
pattern is a convolution of the GaI2 sites and the Ga2I3 site, and
as the reaction time increases, the relative amounts of the GaI2
sites decrease with respect to the Ga2I3 site. For example, at 60
min, the ratio of GaI2 sites 1 and 2 to Ga2I3 is 1:1:0.3, but at 80
Figure 4. Experimental 127I NQR spectra of (a) the four unique I sites
in the early-stage “GaI” (40 min) and (b) the three unique sites in the
late-stage “GaI” (100 min).
9647
dx.doi.org/10.1021/ic501139w | Inorg. Chem. 2014, 53, 9644−9656

Inorganic Chemistry
Article
Figure 5. Experimental and simulated stationary sample solid-state 71Ga NMR spectra of the early-stage “GaI” (40 min) sample, acquired at (a) B0 =
9.04T and (b) B0 = 14.1 T. The experimental spectrum is composed of two subspectra from the two distinct Ga sites within the sample. Each
subpectrum has been simulated and then summed together in a 1:1 ratio to form the complete simulation. Inset: a strongly deshielded 71Ga peak is
also observed at both magnetic field strengths.
Table 2. 71Ga Solid-State NMR Parameters for the Observed Ga Sites
site
a
b
deshielded peak,
Ga0
early stage sit−e 1, GaI2/
[GaI4
early stage site 2, GaI2/
[Ga+]
late-stage site 1, Ga2I3/
[Ga+]
late-stage site 2, Ga2I3/
2−
]
[Ga2I6
]
Parameters
δiso (ppm)
κ
4484.6(3)
−511(2)
−0.3(1)
85(5)
1.81(5)
1
−335(5)
+1
−425(3)
−0.07(7)
145(10)
3.1(1)
1
15(5)
Ω (ppm)
CQ (MHz)
ηQ
80(30)
7.1(3)
25(1)
0.38(5)
0.05(5)
α (deg)
β (deg)
γ (deg)
0
0
41(10)
132(5)
24(10)
70(5)
0
0
0
Ratio of Ga Sites
reaction time
40 min
present
present
present
present
1
1
1
0
1
1
1
0
0
60 min
0.3(1)
1.3(5)
1
80 min
100
min
present
a
Due to the large chemical shift difference between the Ga0 peak and the remaining sites, it was difficult to determine the relative amounts of these
sites with a high degree of accuracy. For example, for the late-stage (100 min) sample, depending on how the 71Ga NMR spectrum was acquired, the
b
ratio of the Ga0 site to the Ga2I3 site 1 was as low as 0.25 to as high as 1.2. Due to the large breadth of this site’s powder pattern, we were unable to
determine CSA parameters and also the amount of this site relative to those of the other sites in GaI2 and Ga2I3.
min, the ratio is now 1:1:1.3. Using these ratios, the 71Ga NMR
spectra of the 60 and 80 min samples can be simulated. This
transition from GaI2 to Ga2I3 was also observed in the FT-
Raman spectra of the four samples; however, the relative
amounts cannot be quantified using Raman spectroscopy, while
they can be using ssNMR spectroscopy.
As mentioned above, it was surprising that for the late-stage
“GaI” sample, which has been shown to possess Ga2I3, only one
Ga site was observed even though there are two distinct Ga
sites in Ga2I3.42 A closer examination of the 71Ga ssNMR
spectrum of Ga2I3 revealed several small bumps in the baseline,
which we originally assumed were simply artifacts. To be
certain, the 71Ga ssNMR spectrum of Ga2I3 was reacquired
using a quadrupolar Carr−Purcell−Meiboom−Gill
(QCPMG)56 experiment combined with WURST pulses.57
Using these approaches, we were able to obtain the 71Ga NMR
spectrum of the second Ga site within Ga2I3 (Figure 7). The
powder pattern for this site is extremely broad, over 700 kHz at
9648
dx.doi.org/10.1021/ic501139w | Inorg. Chem. 2014, 53, 9644−9656

Inorganic Chemistry
Article
Figure 6. Experimental and simulated stationary sample solid-state 71Ga NMR spectra of the “GaI” early-stage (40 min), intermediate-stage (60 and
80 min), and late-stage (100 min) samples acquired at (a) B0 = 9.04 T and (b) B0 = 14.1 T. The displayed region of the spectrum of the early-stage
−
sample was simulated using two distinct Ga sites that correspond to the [Ga+] and [GaI4 ] environments in GaI2. In the displayed region of the late-
stage sample, which we know to be Ga2I3, only one of its [Ga+] and [Ga2I62−] Ga sites give rise to an observable NMR signal. The simulated spectra
of the intermediate-stage samples are convolutions of the spectra of GaI2 and Ga2I3, where the relative amount of GaI2 decreases with respect to that
of Ga2I3 as the reaction time increases.
B0 = 9.4 T, in comparison to the other Ga site in Ga2I3 and
either Ga site in GaI2. The 71Ga CQ value of this second Ga2I3
site is 25 MHz, while for the first site the CQ value was only 3.1
MHz. The massive difference in the breadths of the two powder
patterns has two consequences. The first is that the intensity of
site 2 is very weak in comparison to that of the much narrower
site 1, and as a result the signal from site 2 was simply lost in
the baseline of the spin−echo NMR spectra. The second is that
it is not possible to accurately determine the relative amounts
of the two Ga sites present in the late-stage “GaI” sample.
In addition to having a significant difference in CQ values, the
chemical shifts also differ significantly between the two sites:
−425 ppm for the first site, in comparison to +15 ppm for the
second site. Thus, the first site in Ga2I3 has a CQ value and
chemical shift that are comparable to the values for the two
sites found in GaI2, whereas the second site in Ga2I3 has
significantly different CQ and chemical shift values. From this, a
tentative assignment of the NMR sites to the two Ga sites in
Ga2I3 can be made. In Ga2I3, there is a [Ga+] similar to that
found in GaI2, plus a [Ga2I62−] dimer that is unique to Ga2I3. It
would be expected that the NMR spectroscopic parameters for
the [Ga+] site in Ga2I3 would be similar to those observed for
the [Ga+] site in GaI2. From this, we assign the first NMR site
in Ga2I3 to [Ga+]. The second NMR site in Ga2I3 is thus the
[Ga2I62−] dimer, where each Ga in the dimer sits in the center
of a distorted I3−Ga−Ga tetrahedron. Of the two “GaI”
samples, this is the only Ga site where the nearest neighbor
atoms are not all I. When GaI4− is considered, it is well-known
that as the I atoms are replaced by other atoms the 71Ga
chemical shift becomes more deshielded.58 For example, the
−
71Ga chemical shift of [GaI4 ] in CH2Cl2 is −455 ppm and
−
increases to −310 ppm for [GaBrI3 ] and to −235 ppm for
59
−
2−
[GaClI3 ]. Therefore, it is not surprising that the [Ga2I6
]
dimer site in Ga2I3 has a unique chemical shift value in
comparison to the other Ga sites in Ga2I3 and GaI2.
There exists a curious disagreement between the 127I NQR
data and 71Ga solid-state NMR data acquired in our study
versus those acquired in the study by Bryce and co-workers.29
The 71Ga solid-state NMR parameters for our late-stage “GaI”,
which we have assigned to Ga2I3, match the 71Ga ssNMR
parameters determined by Bryce and co-workers for their “GaI”
sample. However, for their “GaI” sample, they reported 127I
NQR frequencies of 113.69, 132.03, and 134.375 MHz and
thus assigned their compound to be GaI2 (or more formally
[Ga0]2[Ga]+[GaI4]−).29 For our late-stage “GaI” sample, no 127
I
NQR signals were obtained at those frequencies. These known
127I NQR frequencies for GaI2 were present in our early-stage
“GaI”, although the solid-state 71Ga ssNMR parameters for that
“GaI” sample do not match those determined by Bryce and co-
9649
dx.doi.org/10.1021/ic501139w | Inorg. Chem. 2014, 53, 9644−9656

Inorganic Chemistry
Article
are controllably synthesized and can be stored for long periods
of time, and their identity is quickly confirmed using common
solid-state techniques.
Reactions of Bis(phosphino)borate Ligands with
“GaI”. Having obtained and characterized “GaI” phases of
differing structural composition, we wished to determine
whether or not each phase possessed a unique reactivity.
Consequently, we attempted the 1:2 stoichiometric reaction of
Tl[Ph2B(CH2PPh2)2] (1)33 with early “GaI” in both THF and
benzene solutions (Scheme 1). Upon mixing the rapid
precipitation of an orange powder was observed, consistent
with the elimination of TlI. Phosphorus-31 NMR spectroscopy
of the reaction mixture showed a dramatic transition from a
1
broad doublet that characterizes the free ligand (δP 52.6, JTl−P
= 4166 Hz) to several sharp resonances below δP 0, with the
most significant occurring as a singlet (δP −1.8 in THF).
Structural characterization of this product was achieved from X-
ray diffraction analysis of single crystals grown by vapor
diffusion of pentane into a THF solution of the purified
powder. Modeling of the X-ray data indicates that the product
is a Ga(II) dimer featuring a formally dicationic Ga2I2 fragment
(3).
The reaction was then repeated, varying only the batch of
“GaI” that was employed. Monitoring the resulting reaction
mixtures by 31P{1H} NMR spectroscopy (Figure 8, left)
demonstrated that each “GaI” unique combination of early and
late “GaI” phases produced different mixtures of products.
Aside from 3, these products proved difficult to isolate, and to
date we have not been able to isolate and characterize any other
species in these reaction mixtures. Therefore, it was decided to
employ
a related bis(phosphino)borate, [Li-
(THF)2(PiPr2CH2)2BPh2] (2),33 in the hope of obtaining
more tractable product mixtures (Scheme 1). The reaction of 2
with the different batches of “GaI” in THF gave a new set of
unique product mixtures (Figure 8, right). X-ray-quality crystals
grown from the reaction with early “GaI” yielded the solid-state
structure of 4 (δP 15 in THF), the structural analog of 3,
though we were unable to isolate this product cleanly in bulk.
An additional product, as determined by 31P{1H} NMR
spectroscopy (δP 14 in THF), was isolated as a white powder
from the reaction mixture involving the late “GaI” (100 min).
The 1H NMR data revealed an asymmetric ligand environment
and a substantial amount of THF, even after prolonged drying.
Single crystals of this product that were suitable for X-ray
diffraction were grown by vapor diffusion of pentane into a
THF solution of the purified powder. Analysis of the resulting
X-ray diffraction data revealed the product to be 5 (Scheme 1),
a lithium−THF salt involving a {Ga2I4} bis(phosphino)borate
Figure 7. Experimental and simulated stationary sample solid-state
WURST QCPMG 71Ga NMR spectra of late-stage (100 min) “GaI”
acquired at B0 = 9.04 T: (a) experimental spectrum, with the vertical
scale normalized to the height of the site 1 peaks; (b) spectrum with
the vertical scale increased to emphasize the second, broader Ga site;
(c) simulated spectrum of the second Ga site.
workers. We do not propose an explanation for this
discrepancy, although we do note that, when performing the
127I NQR, we performed 71Ga ssNMR both before the NQR
experiments and after the NQR experiments to verify the
composition of the sample.
In summary, the results obtained from Raman spectroscopy,
I127 NQR spectroscopy, 71Ga ssNMR spectroscopy, and pXRD
are self-consistent and indicate that the initial phase of “GaI”
−
obtained in Green’s synthesis is [Ga0]2[Ga+][GaI4 ], while the
late phase of “GaI” is [Ga0]2[Ga+]2[Ga2I62−]. These reagents
Scheme 1. Synthesis of Dimeric (3, 4) Ga2I2 Gallium−Phosphorus Coordination Compounds and the Unusual Base-Stabilized
Ga2I4 Fragment (5) Isolated from the Reaction of Bis(phosphino)borate Ligands (1, 2) with Different “GaI” Samples
9650
dx.doi.org/10.1021/ic501139w | Inorg. Chem. 2014, 53, 9644−9656

Inorganic Chemistry
Article
Figure 8. Effect of the type of “GaI” on the outcome of the reaction with the bis(phosphino)borate ligands 1 (left) and 2 (right). A stack plot of 31
P
NMR spectra is shown, highlighting the range of products observed. Total reaction times for the preparation of the “GaI” used in each reaction are
noted adjacent to the relevant spectrum.
Figure 9. Solid-state structures of the feature compounds 3−5 (from left to right, respectively). Ellipsoids are drawn at the 50% probability level, and
hydrogen atoms are removed for clarity. Selected metrical parameters are given in Table 4, while crystallographic details are given in Table 5
anion. In contrast to the reactions with the thallium
bis(phosphino)borate 1, salt elimination does not occur and
the charge of the anionic gallium complex is balanced by a
lithium cation that possesses four THF molecules in its
coordination sphere.
of peaks consistent with a related complex are also observed.
We hypothesize that this compound may be indicative of a
Ga(II) product, though we were unable to isolate and
characterize such a species (N-heterocyclic Ga(II) compounds
have precedent in the literature).5,60 This would be consistent
with our observation that Ga(II) products are produced from
the early-phase “GaI” as well as literature reports of the
formation of Ga(II) products from the reaction between GaI2
and simple donor ligands,61−63 despite the formal Ga(I) and
Ga(III) centers in the crystal structure.64 The increased ratio of
Ga(I) product for the late-phase “GaI” in comparison to the
early-phase “GaI” is potentially a result of the larger mole
percent of Ga+ ions in the overall formula: 8.9% Ga+ in the
To determine if the variable reactivity of the different “GaI”
phases would continue to hold across a range of reactions, we
chose to attempt the reaction of different “GaI” samples with
the β-diketiminate ligand Li{(NDippCMe)2CH (6; Dipp =
i
C6H3 Pr2-2,6), to yield N-heterocyclic gallium species pre-
viously reported by Power et al.6 By treating 6 with 1.5
stoichiometric equivalents of “GaI” in benzene and stirring for 1
1
h, crude mixtures whose contents could be analyzed by H
−
NMR spectroscopy were obtained. Signals that are diagnostic of
Ga(I) and Ga(III) products for Ga{(NDippCMe)2CH} (7)
and I2Ga{(NDippCMe)2CH} (8), respectively, in addition to
some protonated ligand were observed (Figures S-34−S-40,
Supporting Information). For the early-phase “GaI” a fourth set
early phase “GaI” ([Ga0]2[Ga+][GaI4 ]) and 11.8% Ga+ in the
late-phase “GaI” ([Ga0]2[Ga+]2[Ga2I62−]). Obviously the
structure−reactivity relationship is affected by complex redox
and disproportionation reactions that are still not understood.
However, it is hoped that the correct assignment of the
9651
dx.doi.org/10.1021/ic501139w | Inorg. Chem. 2014, 53, 9644−9656

Inorganic Chemistry
Article
structure of different phases of “GaI” will provide some
measure of control in the synthesis of new low-valent gallium
compounds.
X-ray Crystallography. The solid-state structures of 3−5
are shown in Figure 9, while the X-ray parameters and relevant
bond distances and angles are given in Tables 4 and 5,
lengths are slightly different at 2.401(2) and 2.448(2) Å and are
long in comparison to traditional Ga−P covalent bonds (cf.
2.31−2.37 Å)60,65,66 and comparable to those in compounds
that can be described as donor→acceptor complexes (cf. 2.40−
2.48 Å).30,32,67 The P−Ga−P bond angle is fairly small at
92.98(7)°, while the I−Ga−Ga−I torsion angle is 0° due to the
symmetry of the molecule. This structure may be compared to
a
a related Ga(II) dimer isolated by Schnockel et al. which
̈
Table 4. Significant Metrical Parameters for 3−5
consists of a {Ga2I4} fragment stabilized by two P(CH2CH3)3
molecules.31 The Ga−Ga, Ga−P, and Ga−I bond lengths are all
quite similar at 2.436(2), 2.414(3), and 2.58−2.59(1) Å,
respectively. Compound 4, [(2)Ga2I2(2)], has the same
phosphine-stabilized [Ga2I2]2+ core seen in 3; however, 4
adopts a different structural conformation than 3 with the I−
Ga−Ga−I torsion angle being 97.85(2)° instead of perfectly
linear. As a result, the bis(phosphino)borate ligands are twisted
relative to each other, highlighting the unique structural
changes that can be observed by varying the substituents on
phosphorus. The Ga−Ga bond length and two Ga−I bond
lengths are 2.4999(7), 2.6257(6), and 2.6367(6) Å, respec-
tively. The Ga−P bond lengths are again consistent with dative
bonds at 2.4239(14), 2.4439(13), 2.4246(13), and 2.4529(13)
Å. The P−Ga−P bond angles are somewhat different at
95.29(5) and 99.34(4)°, a likely result of the flexibility of the
ligand framework. For compound 5, [Li(THF)4]+[(2)-
GaIGaI3]−, the Ga−Ga and two Ga−P bond lengths are
2.4521(11), 2.3906(15), and 2.4027(16) Å, respectively. The
3
4
5
Ga−P
2.401(2)
2.448(2)
2.4239(14)
2.4439(13)
2.4246(13)
2.4529(13)
2.4999(7)
2.6257(6)
2.6367(6)
2.3906(15)
2.4027(16)
Ga−Ga
Ga−I
2.4666(17)
2.5755(14)
2.4521(11)
2.6167(14)
2.6055(14)
2.6082(11)
2.6181(11)
96.92(5)
P−Ga−P
92.98(7)
0
95.29(5)
99.34(4)
97.85(2)
I−Ga−Ga−I
a
Bond lengths are given in Å and bond angles in deg.
respectively. The monomeric form of compound 3 sits on an
inversion center. The Ga−Ga bond length is 2.4666(17) Å,
while the Ga−I bond length is 2.5755(14) Å. The Ga−P bond
Table 5. X-ray Details for 3−5
3
4
5
formula
C76H68B2Ga2I2P4
1520.1
C67H99B2Ga2I2P4
1443.27
0.034 × 0.47 × 0.121
colorless needle
monoclinic
P21/n
C42H74BGa2I4LiO4P2
1369.74
formula wt
crystal dimens, mm
0.065 × 0.080 × 0.110
colorless prism
triclinic
0.100 × 0.120 × 0.302
colorless needle
triclinic
crystal color and habit
crystal syst
space group
P1
P1
̅
̅
temp, K
150
150
150
a, Å
12.644(3)
13.368(3)
13.342(3)
106.16(3)
93.01(3)
110.49(3)
2014.1(7)
1
10.4865(9)
29.148(2)
22.9354(18)
90
11.109(2)
15.256(3)
17.476 (4)
105.18(3)
93.18(3)
104.20(3)
2748.2(10)
2
b, Å
c, Å
α, deg
β, deg
100.284(2)
90
γ, deg
V, Å3
6897.7(10)
4
Z
F(000)
762
2956
1340
ρ, g/cm
1.253
1.390
1.655
λ(Mo Kα), Å
μ, cm‑1
0.71073
1.551
0.71073
1.806
0.71073
3.319
max 2θ for data collection, deg
measd fraction of data
no. of rflns measd
no. of rflns measd
Rmerge
50.70
51.42
55.84
0.977
0.993
0.976
22320
60725
23164
7200
13067
12856
0.0835
0.1278
0.0445
no. of rflns included in refinement
no. of params in least squares
7200
13067
12856
388
710
635
a
R1, wR2
0.0626, 0.1401
0.1124, 0.1530
0.980
0.390, 0.0421
0.1175, 0.0505
0.706
0.0491, 0.1091
0.1107, 0.1313
1.038
a
R1, wR2 (all data)
a
GOF
min, max peak heights on final ΔF map, e/Å
1.796, −1.312
0.578, −0.600
1.620, −1.760
a
2
4
2
R1 = ∑(|Fo| − |Fc|)/∑Fo. wR2 = [∑(w(Fo − Fc2)2)/∑(wFo ) ]1/2. GOF = [∑(w(Fo − Fc2)2)/((no. of rflns) − (no. of params))]1/2
.
9652
dx.doi.org/10.1021/ic501139w | Inorg. Chem. 2014, 53, 9644−9656

Inorganic Chemistry
Article
from Caledon Laboratories and dried and deoxygenated in an MBraun
Atmosphere Controlled Solvent Purification System. Dried solvents
were collected under vacuum and stored as follows. Acetonitrile was
stored under a nitrogen atmosphere in Strauss flasks or in the glovebox
over 3 Å molecular sieves. All other solvents were stored under a
nitrogen atmosphere in Strauss flasks or in the glovebox over 4 Å
molecular sieves. All solvents used in NMR spectroscopy were
obtained from Cambridge Isotope Laboratories and dried with CaH2,
distilled under vacuum, and stored in the glovebox over 4 Å molecular
sieves. Solution 1H, 13C{1H}, 11B{1H}, and 31P{1H} NMR spectra
were recorded on a Varian INOVA 400 MHz or a Varian INOVA 600
MHz spectrometer as noted (for the 400 MHz spectrometer, 1H
400.09 MHz,11B{1H}128.23 MHz, 13C{1H} 100.52 MHz, 31P{1H}
161.82 MHz; for the 600 MHz spectrometer, 1H 599.5 MHz 13C{1H}
Ga−P bond lengths in 5 are nearly identical, and the average
distance is slightly less than that observed in 3. The Ga−I bond
distance for the {GaI} fragment is 2.6167(14) Å, while those
for the {GaI3} fragment are 2.6055(14), 2.6082(11), and
2.6181(11) Å. All four of these bond lengths are longer than
the Ga−I distance observed in 3. The P−Ga−P bond angle is
96.92(5)°, while the P−Ga−I bond angles are significantly
smaller (cf. 100−101°) than the P−Ga−Ga bond angles (cf.
121−126°). The Li−O bond lengths for the Li(THF)4 cation
are reasonably consistent, considering the inherent disorder of
the THF molecules and range from 1.90 to 1.94 Å. The gallium
atoms in both 3 and 4 are in a distorted-tetrahedral geometry,
consistent with being four-coordinate and electronically
satisfied.
1
150.78 MHz, 31P{1H} 242.89 MHz). All H and 13C NMR samples
were referenced to the solvent signal relative to Si(CH3)4 (CDCl3, 1H
δH 7.26, 13C δC 77.16; C6D6, 1H δH 7.16, 13C{1H} δC 128.06; CD3CN,
1H δH 1.95, 13C{1H} δC 1.32, 118.26). Chemical shifts for 31P{1H} and
11B{1H} NMR spectroscopy were referenced to an external standard
(85% H3PO4, δP 0 ppm; BF3(Et2O), δB 0 ppm). Fourier transform
infrared spectroscopy was performed on samples as KBr pellets using a
Bruker Tenser 27 FT-IR spectrometer with a resolution of 4 cm−1.
Fourier transform Raman spectroscopy was performed on samples that
were flame-sealed in glass capillaries using a Bruker RFS 100/S
spectrometer, with a resolution of 4 cm−1. Mass spectrometry was
recorded in house in positive and negative ion modes using an
electrospray ionization Micromass LCT spectrometer. Melting or
decomposition points were determined by flame-sealing the sample in
capillaries and heating using a Gallenkamp Variable Heater. The “GaI”
was prepared in a Elma E 60 H Elmasonic ultrasonic bath set to 30 °C.
Single-Crystal X-ray Crystallography. The single-crystal X-ray
diffraction studies were performed at the Western University X-ray
facility. Crystals were selected under Paratone(N) oil, mounted on a
Mitegen polyimide micromount, and immediately put under a cold
stream of nitrogen for data to be collected on a Nonius Kappa-CCD
area detector or Bruker Apex II detector using Mo Kα radiation (λ =
0.71073 Å). The Nonius and Bruker instruments operate SMART68
and COLLECT69 software, respectively. The unit cell dimensions were
determined from a symmetry -constrained fit on the full data set,
which was composed of φ and ω scans. The frame integration was
performed by SAINT,70 the resulting raw data were scaled, and
absorption was corrected using a multiscan averaging of symmetry-
equivalent data using SADABS.71 The SHELXTL/PC V6.14 for
Windows NT suite of programs was used to solve the structure by
direct methods.72 Subsequent difference Fourier syntheses allowed the
remaining atoms to be located, while hydrogen atoms were placed in
the calculated positions and allowed to ride on the parent atom. In all
cases the gallium bis(phosphino)borate components were well ordered
and refined with anisotropic thermal parameters. The Li(THF)4 cation
in 4 possesses three disordered solvate molecules that can each be
refined over two positions, while the fourth THF solvate is partially
disordered. The model refines satisfactorily, with all carbon atoms
being refined anisotropically. The thermal parameters were restrained
with SIMU and DELU commands. The related C−C and C−O bond
lengths in these disordered molecules were restrained to be identical
by using the SAME command. Certain C−C and C−O bond lengths
were restrained to sensible values with the aid of the DFIX command.
Two disordered solvents (THF and CH2Cl2) were present in the unit
cell of 3. We were unable to model the solvent molecules, even with
the use of restraints, and thus their electron density was treated as a
diffuse contribution to the overall scattering by Squeeze/Platon (total
129 electrons).
CONCLUSIONS
■
By examining the composition of different “GaI” samples,
prepared by varying the sonication time, convincing new
structural insights regarding the appropriate assignment of
“GaI” were obtained. It was demonstrated through compre-
hensive solid-state characterization methods that GaI2 is the
first gallium subiodide phase produced when using Green’s
method of sonication of the elements, followed by quantitative
conversion to Ga2I3 over the course of the reaction. Gallium
metal is present in both phases to give an overall structural
composition of [Ga0]2[Ga+][GaI4 ] (simplified, [Ga0]2[Ga2I4])
−
for the early-stage “GaI” and [Ga0]2[Ga+]2[Ga2I62−] (simplified,
[Ga0]2[Ga4I6]) for the late-stage “GaI”. The intermediate
phases contain a mixture of both extremes with no other
observable gallium iodine compounds (i.e. GaI3). These phases
are easily and reproducibly prepared by controlling the reaction
time, while the samples may be routinely analyzed by FT-
Raman spectroscopy and powder X-ray diffraction. In addition,
ssNMR and NQR spectroscopy may also be used to quickly
characterize, and identify, the “GaI” phase present after
synthesis. Gallium chemists can now use widely accessible
techniques to provide diagnostic information on the “GaI” they
have prepared and potentially gain a handle on the reactivity it
may exhibit. This is significant, as it was further demonstrated
that each unique composition possesses a unique reactivity
through the synthesis of gallium−phosphorus coordination
compounds. The zwitterionic Ga(II) dimers 3 and 4 feature a
formal “Ga2I22+” core stabilized by the anionic bis(phosphino)-
borate ligands, while the product 5 may be thought of as a
Lewis acid−base adduct between a bis(phosphino)borate-
stabilized {GaI} fragment and GaI3. The synthetic results
underscore the importance of identifying the nature of the
“GaI” prepared to obtain consistent reactivity.
EXPERIMENTAL SECTION
■
General Procedures. All inert atmosphere syntheses were
performed in a nitrogen-filled MBraun Labmaster 130 glovebox or
using standard Schlenk-line techniques unless otherwise stated. The
bis(phosphino)borates were prepared from the materials described
below by the literature procedures.33 Phenylmagnesium bromide (1 M
in THF), boron trichloride (1 M in heptane), methyllithium (1.6 M in
Et2O), chlorodiphenylphosphine, n-butyllithium (2 M in cyclohexane),
and diisopropylchlorophosphine were obtained from Sigma-Aldrich.
Dimethyltin dichloride, thallium(I) nitrate, and dimethyldiphenyltin
were obtained from Alfa Aesar. N,N,N′,N′-Tetramethylethylenedi-
amine (TMEDA, obtained from Alfa Aesar) was stirred over NaOH,
distilled under vacuum, and stored in a Strauss flask under N2. All
other reagents were used as received. Both elemental gallium and
iodine were obtained from Sigma-Aldrich. Solvents were obtained
Powder Diffraction. The powder diffraction studies were
performed on an Inel CPS Powder diffractometer using Cu Kα
radiation from an Inel XRG 3000 generator and a CPS 120 detector.
The samples were ground to a fine powder with a mortar and pestle
and sealed on an aluminum dish with Scotch tape (Scotch 3M Magic
Tape 810D). After 90 min the signals attributable to “GaI” and the
Scotch tape (broad signal between 5 and 20° 2θ) were clearly
observed. These data were processed using the ACQ software and
9653
dx.doi.org/10.1021/ic501139w | Inorg. Chem. 2014, 53, 9644−9656

Inorganic Chemistry
Article
compared to literature patterns using the Match software. The powder
patterns for GaI2 and Ga2I3 are accessible from the PDF-4+ database
with the numbers 04-007-1340 and 04-007-1339, respectively. It
should be noted that no suitable diffraction pattern is observed if the
samples are packed in a flame-sealed capillary.
“GaI” samples using a Varian Inova 600 NMR spectrometer equipped
with a Varian 4 mm triple-resonance HXY magic-angle spinning NMR
probe. The samples were packed tightly into 4 mm o.d. ZrO2 rotors
inside a nitrogen-filled glovebox and then sealed before being
transferred to the probe. The probe was placed roughly 3 m from
the edge of the NMR magnet and was purged continuously with
nitrogen gas. For the 40 min sample, the spectra were acquired using a
π/2−τ1−π−τ2−acq spin−echo pulse sequence, where τ1 was 30 μs and
τ2 was 15 μs. A total of 2048 scans were summed using a 1.05 μs π/2-
pulse width, 500 kHz spectral width, 0.5 s recycle delay, and 256 μs
acquisition time. The transmitter frequencies attempted included the
Solid-State NMR Spectroscopy. Solid-state 71Ga NMR experi-
ments were performed using a Varian Infinity Plus 400 NMR
spectrometer (νL(71Ga) = 121.78 MHz) equipped with a Varian 5 mm
quadrupole-resonance HFXY magic-angle spinning NMR probe and a
Varian Inova 600 NMR spectrometer (νL(71Ga) = 182.67 MHz)
equipped with a Varian 3.2 mm triple-resonance HXY magic-angle
spinning probe. The powder samples were stored inside a nitrogen-gas
glovebox filled with nitrogen gas and packed tightly into either 5 or 3.2
mm o.d. ZrO2 rotors and then sealed. At both magnetic field strengths,
the FIDs were acquired using either a π/2−τ1−π/2−τ2−acq or a π/
2−τ1−π−τ2−acq spin−echo pulse sequence, where τ2 < τ1, and the
spectra were referenced with respect to the 71Ga peak of a 1.0 M
aqueous Ga(NO3)3 solution (δ(71Ga) 0.0 ppm). On the 400 MHz
spectrometer, 4096 scans were summed using a selective 1.7 μs π/2-
pulse width, 800 kHz spectral width, 30 μs τ1, 1 s recycle delay, and
12.8 ms acquisition time. On the 600 MHz spectrometer, between
3104 and 16000 scans were summed using a selective 0.25 μs π/2-
pulse width, 500 kHz spectral width, 30 μs τ1, 1 s recycle delay, and 4.1
ms acquisition time. For processing, the FIDs were left-shifted to the
top of the half-echo; 1 zero-fill and 400 Hz of line broadening were
applied before Fourier transform.
58
known 127I NQR frequencies for GaI3 (ν(mI = 1/2 ↔ 3/2) =
133.69, 173.65, and 174.59 MHz) and for GaI2 (ν(mI = 1/2 ↔ 3/
2) = 113.65, 131.94, 134.27, and 163.71 MHz). For processing, the
FIDs were left-shifted to the top of the half-echo; 1 zero-fill and 500
Hz of line broadening were applied before Fourier transform. For the
100 min sample, experiments were performed and processed in the
same manner as for the 40 min sample, except the transmitter
frequency was varied from 176.6 to 104.0 MHz in 0.2 MHz increments
and 256 scans were summed. Once an NQR frequency was observed,
the transmitter was placed “on-resonance” and 2048 scans were
summed.
Literature Procedures. Tl[Ph2B(CH2PPh2)2] (1) and [Li-
(THF)2(PiPr2CH2)2BPh2] (2) were prepared according to the
procedure of Peters,33 with the exception that a solution of TlNO3
in 1/1 EtOH/H2O was often used as a substitute for the ethanolic
solution of TlPF6 with negligible changes in the purity or yield of the
product. Li{(NDippCMe)2CH (6) was prepared according to the
literature procedure.77
To observe the broad site in the 100 min sample, stationary-sample
71Ga quadrupolar Carr−Purcell−Meiboom−Gill (QCPMG)56 spectra
were acquired on both the Infinity Plus 400 and Inova 600 NMR
spectrometers. On the latter instrument, a total of 13 individual spectra
were acquired, where the transmitter frequency varied by 50 kHz
between each spectrum, and summed together to generate the entire
powder pattern. Each individual spectrum was acquired using 1024
scans, a 4.0 μs π/2-pulse width, 500 kHz spectral width, 1 s recycle
delay, 4.96 ms total acquisition time, and an echo train consisting of 48
π-pulses. The interpulse delays were set in order to achieve a 10 kHz
spacing between the individual spikelets. On the former instrument,
the WURST-QCPMG57 variant was utilized and thus the powder
pattern could be fully excited in one experiment. The spectrum was
acquired using 12600 scans, 10 μs WURST pulses, 700 kHz offset,
2500 kHz spectral width, 1 s recycle delay, 5.39 ms total acquisition
time, and an echo train consisting of 55 WURST pulses. The
interpulse delays were set in order to achieve a 10 kHz spacing
between the individual spikelets.
Synthesis of “GaI”. Note: we found it most reliable to prepare 500
mg of “GaI” at one time. The reaction can be scaled to prepare greater
than 10 g of “GaI”; however, the reaction time must be adjusted
accordingly.
Gallium metal (0.1863 g, 2.674 mmol, 1 equiv) was weighed into a
100 mL pressure tube in the glovebox. The gallium metal was heated
until it melted and spread about the bottom of the flask in an effort to
maximize surface area. Toluene (4.5 mL) was added, followed by
iodine (0.3393 g, 1.337 mmol, 0.5 equiv). Residual iodine was rinsed
with toluene (4.5 mL) and added to the reaction mixture, and the
vessel was sealed with an O-ring fitted screw-thread Teflon stopper.
The resulting purple solution was then sonicated at 30 °C in 20 min
intervals for 40−120 min, with vigorous physical agitation between
each interval. Toluene was removed in vacuo to yield a gray to green
powder depending on the time. Yield: 100%, 0.525 g, 2.67 mmol.
FT-Raman spectroscopy (cm−1 (intensity normalized to 2)): 40
min sample, 267 (0.03), 230 (0.11), 213 (0.03), 141 (2), 124 (0.04),
86 (0.26); 60 min sample, 292 (0.02), 232 (0.06), 213 (0.05), 141 (2),
124 (0.43), 84 (0.30); 80 min sample, 292 (0.12), 232 (0.05), 213
(0.04), 188 (0.04), 141 (1.66), 124 (2), 84 (0.70); 100 min sample,
292 (0.12), 188 (0.04), 141 (0.04), 124 (2), 84 (0.49).
Stationary-sample 71Ga NMR spectra are broadened by the
quadrupolar interaction between the nuclear quadrupole moment of
71Ga and the molecule’s electric field gradient (EFG), plus the
orientation dependence of the chemical shift (chemical shift
anisotropy, CSA). The EFG and CSA are both second-rank interaction
tensors that in their principal axis system can be described by three
principal components. The EFG tensor is represented by VXX, VYY, and
VZZ, where |VXX| ≤ |VYY| ≤ |VZZ|, and the CS tensor can be represented
by δ11, δ22, and δ33, where δ11 ≥ δ22 ≥ δ33. Simulations of the
experimental NMR spectra were performed using the WSolids1
software developed by Klaus Eichele73 and require parameters
describing the quadrupolar interaction, the CS tensor, and Euler
angles that describe the relative orientation of the EFG and CS
tensors.74−76 The quadrupolar interaction is described by two
parameters: the quadrupolar coupling constant CQ = eQV33h−1,
where e is the elementary charge, Q is the 71Ga nuclear quadrupole
moment, and h is Planck’s constant, and the asymmetry parameter ηQ
= (VXX − VYY)/VZZ. The chemical shift tensor is described by three
parameters: the isotropic chemical shift δiso = (δ11 ≥ δ22 ≥ δ33)/3, the
span Ω = δ11 − δ33, and the skew κ = 3(δ22 − δiso)Ω−1.79 The relative
orientations of the EFG and CS tensors are described by three Euler
angles: α, β, and γ. Different conventions for the Euler angles exist, and
we utilized the ZYZ convention as implemented in the WSolids1
software.
FT-Raman spectroscopy for other gallium iodides: GaI2, 235 (w),
214 (w), 143 (vs);43 Ga2I3, 292 (s), 186 (w), 124 (vs), 79 (m);43 GaI3,
267 (0.05), 227 (0.20), 194 (0.03), 163 (0.10), 142 (2.0), 85 (0.35).
Synthesis of the Gallium(II) Dimer [Ph2B(CH2PPh2)2(GaI)]2
(3). A suspension of Tl[Ph2B(CH2PPh2)2] (1; 0.7777 g, 1.013 mmol,
1 equiv) in THF (3 mL) was prepared. In a separate vial, further THF
(3 mL) was added to “GaI” (0.3877 g, 1.972 mmol, 2 equiv; 40 min
preparation time) to give a fluid gray-green slurry. This slurry was
immediately added to the suspension of 1 and rinsed with 3 mL of
THF, resulting in an immediate color change to bright orange. After
the mixture was stirred for 5 min, solids were removed by
centrifugation, yielding a colorless supernatant that was concentrated
in vacuo to give an off-white powder. Sequential washes with diethyl
ether (3 mL) and CH3CN (2 × 3 mL) and further drying in vacuo to
remove residual solvent yielded a white powder. Single crystals suitable
for X-ray diffraction were grown by vapor diffusion of pentane into a
THF solution. Yield: 48%, 0.3172 g, 0.2087 mmol. Mp: 176−177 °C
1
dec. H NMR (400 MHz, CDCl3, 25 °C, δ): 1.94 (br, 4H), 2.36 (br,
Nuclear Quadrupole Resonance. 127I nuclear quadrupole
resonance experiments were performed on the 40 and 100 min
4H), 6.74 (m, 16H), 6.93 (t, 8H, 3JH−H = 7.6 Hz), 6.99 (q, 12H, 3JH−H
3
= 8.0 Hz), 7.13 (m, 16H), 7.19 (t, 4H, JH−H = 7.6 Hz), 7.29 (t, 4H,
9654
dx.doi.org/10.1021/ic501139w | Inorg. Chem. 2014, 53, 9644−9656

Inorganic Chemistry
Article
3JH−H = 7.6 Hz). 31P{1H} NMR (161.82 MHz, CDCl3, 25 °C, δ):
−1.75. 11B{1H} NMR (128.23 MHz, CDCl3, 25 °C, δ): −13.2.
13C{1H} NMR (100.5 MHz, CDCl3, 25 °C, δ): 17.5−18.5 (br), 122.8,
ACKNOWLEDGMENTS
■
We thank the Natural Science and Engineering Research
Council of Canada (NSERC), the Ontario Government, the
Canadian Foundation of Innovation (CFI), and The University
of Western Ontario for their generous funding. Prof. Robert
Schurko from the University of Windsor is thanked for sending
us the code for the WURST-QCPMG pulse sequence.
1
123.3, 126.4, 126.6, 128.0, 128.8, 130.7, 131.1, 131.1 (d, JP−C = 56.5
1
Hz), 132.3, 132.8, 132.9 (d, JP−C = 54.3 Hz), 133.5, 134.3, 159.5−
162.0 (br). FT-IR (cm−1 (ranked intensity)): 476 (13), 492 (7), 507
(4), 691 (1), 736 (3), 867 (5), 932 (11), 1098 (6), 1136 (12), 1307
(15), 1435 (2), 1484 (8), 3005 (14), 3038 (10), 3057 (9). FT-Raman
(cm−1 (ranked intensity)): 86 (4), 101 (3), 143 (2), 204 (11), 220
(9), 234 (12), 262 (13), 1001 (1), 1032 (8), 1099 (10), 1155 (15),
1586 (6), 2884 (14), 3041 (7), 3057 (5). Anal. Calcd (found): C,
60.05 (59.25); H, 4.51 (4.23).
REFERENCES
■
(1) Power, P. P. Nature 2010, 463, 171.
Synthesis of Compound 4. Compound 4 was prepared by a
method similar to that for 3 using the 40 min “GaI” sample and
[Li(THF)2(PiPr2CH2)2BPh2] (2); however, this compound could not
be isolated in greater than 85% purity. 1H NMR (400 MHz, C6D6, 25
°C, δ): 0.67 (dd, 6H, 3JH−H = 6.4 Hz, 3JP−H = 13.2 Hz), 0.80−0.95 (m,
12H), 1.13 (dd, 6H, 3JH−H = 6.8 Hz, 3JP−H = 13.2 Hz), 1.57 (br t, 2H),
2.05 (br t, 2H), 2.68−2.78 (br, 4H), 7.05 (t, 1H, 3JH−H = 7.6 Hz), 7.20
(2) Power, P. P. Acc. Chem. Res. 2011, 44, 627.
(3) Schmidt, E. S.; Jockisch, A.; Schmidbaur, H. J. Am. Chem. Soc.
1999, 1, 9758.
(4) Asay, M.; Jones, C.; Driess, M. Chem. Rev. 2010, 111, 354.
(5) Baker, R. J.; Farley, R. D.; Jones, C.; Kloth, M.; Murphy, D. M. J.
Chem. Soc., Dalton Trans. 2002, 3844.
(6) Hardman, N. J.; Eichler, B. E.; Power, P. P. Chem. Commun.
2000, 53, 1991.
(7) Jones, C.; Junk, P. C.; Platts, J. A.; Stasch, A. J. Am. Chem. Soc.
2006, 128, 2206.
(8) Jones, C.; Mills, D. P.; Platts, J. A.; Rose, R. P. Inorg. Chem. 2006,
45, 3146.
(9) Green, S. P.; Jones, C.; Stasch, A. Inorg. Chem. 2007, 46, 11.
(10) Baker, R. J.; Jones, C.; Platts, J. A. J. Am. Chem. Soc. 2003, 125,
10534.
(11) Arnold, P. L.; Liddle, S. T.; McMaster, J.; Jones, C.; Mills, D. P.
J. Am. Chem. Soc. 2007, 129, 5360.
3
3
(t, 1H, JH−H = 7.2 Hz), 7.30 (t, 2H, JH−H = 7.2 Hz), 7.35 (t, 2H,
3JH−H = 8.0 Hz), 7.70 (d, 2H, 3JH−H = 7.6 Hz), 7.90 (d, 2H, 3JH−H = 7.6
Hz). 31P{1H} NMR (161.82 MHz, C6D6, 25 °C, δ): 21.5. 11B{1H}
NMR (128.23 MHz, C6D6, 25 °C, δ): −13.0.
Synthesis of [Li(THF)4][GaI3-GaI(PiPr2CH2)2BPh2] (5). A sol-
ution of [Li(THF)2(PiPr2CH2)2BPh2] (2; 57.9 mg, 0.10 mmol, 1.0
equiv) in THF (3 mL) was prepared. Separately, a vial was charged
with “GaI” (40.0 mg, 0.20 mmol, 2.0 equiv; 100 min preparation time)
followed by THF (3 mL) to give a suspension of green particles. The
solution of 2 was immediately added to this suspension in a rapid
dropwise fashion and the resulting mixture stirred for 5 min. Solids
were removed by centrifugation, and the supernatant was concentrated
to an off-white solid in vacuo. Washing this solid with Et2O (3 × 3
mL) and further drying in vacuo yielded a white powder. Yield: 28%,
0.0387 g, 0.282 mmol. 1H NMR (400 MHz, C6D6, 25 °C, δ): 0.85 (dd,
(12) Green, S. P.; Jones, C.; Mills, D. P.; Stasch, A. Organometallics
2007, 26, 3424.
(13) Bonello, O.; Jones, C.; Stasch, A.; Woodul, W. D. Organo-
metallics 2010, 29, 4914.
(14) Burford, N.; Ragogna, P. J.; Robertson, K. N.; Cameron, T. S.;
Hardman, N. J.; Power, P. P. J. Am. Chem. Soc. 2001, 124, 382.
(15) Lichtenthaler, M. R.; Higelin, A.; Kraft, A.; Hughes, S.; Steffani,
A.; Plattner, D. A.; Slattery, J. M.; Krossing, I. Organometallics 2013,
32, 6725.
(16) Higelin, A.; Sachs, U.; Keller, S.; Krossing, I. Chem. Eur. J. 2012,
18, 10029.
(17) Higelin, A.; Haber, C.; Meier, S.; Krossing, I. Dalton Trans.
2012, 41, 12011.
3
3
3
6H, JH−H = 6.8 Hz, JP−H = 14.6 Hz), 0.95 (dd, 6H, JH−H = 6.4 Hz,
3JP−H = 14.6 Hz), 1.15 (dd, 6H, JH−H = 7.0 Hz, JP−H = 16.0 Hz),
3
3
2
1.30−1.40 (overlapping signals, 22H), 1.60 (t, 2H, JP−H = 15.4 Hz),
3
1.95 (t, 2H, JP−H = 15.4 Hz), 2.40−2.60 (overlapping doublet of
3
3
septets, 4H), 3.50 (t, 16H, JH−H = 6.4 Hz), 7.05 (t, 1H, JH−H = 7.6
3
3
Hz), 7.20 (t, 1H, JH−H = 8.0 Hz), 7.30 (t, 2H, JH−H = 7.2 Hz), 7.35
3
3
(t, 2H, JH−H = 7.6 Hz), 7.70 (d, 2H, JH−H = 7.2 Hz), 7.90 (d, 2H,
3JH−H = 8.0 Hz). 31P{1H} NMR (161.82 MHz, C6D6, 25 °C, δ): 17.9.
11B{1H} NMR (128.23 MHz, C6D6, 25 °C, δ): −13.4.
(18) Higelin, A.; Keller, S.; Gohringer, C.; Jones, C.; Krossing, I.
Angew. Chem., Int. Ed. 2013, 52, 4941.
̈
Reactions of “GaI” with Li{(NDippCMe)2CH (6). A solution of 6
(49.5 mg, 0.118 mmol, 1 equiv) was prepared in 1.5 mL of benzene-d6.
Separately, a suspension of “GaI” (35.3 mg, 0.177 mmol, 1.5 equiv) in
1.5 mL of benzene-d6 was prepared. The solution of 6 was added to
the “GaI” suspension dropwise. After the mixture was stirred for 1 h,
the crude suspension was filtered and a 1H NMR spectrum obtained to
characterize the product mixture.
(19) Slattery, J. M.; Higelin, A.; Bayer, T.; Krossing, I. Angew. Chem.,
Int. Ed. 2010, 49, 3228.
(20) Wehmschulte, R. J. Angew. Chem., Int. Ed. 2010, 49, 4708.
(21) Fedushkin, I. L.; Nikipelov, A. S.; Lyssenko, K. A. J. Am. Chem.
Soc. 2010, 132, 7874.
(22) Fedushkin, I. L.; Nikipelov, A. S.; Morozov, A. G.; Skatova, A.
A.; Cherkasov, A. V.; Abakumov, G. A. Chem.Eur. J. 2012, 18, 255.
(23) Cooper, B. F. T.; Macdonald, C. L. B. Mixed Valent
Compounds of Al, Ga, In and Tl. In The Group 13 Metals Aluminium,
Gallium, Indium and Thallium. Chemical Patterns and Peculiarities,;
Aldridge, S., Downs, A. J., Eds.; Wiley: Chichester, U.K., 2011.
(24) Jones, C.; Stasch, A. The Chemistry of the Group 13 Metals in
the +1 Oxidation State. In The Group 13 Metals Aluminium, Gallium,
Indium and Thallium. Chemical Patterns and Peculiarities; Aldridge, S.,
Downs, A. J., Eds.; Wiley: Chichester, U.K., 2011.
ASSOCIATED CONTENT
■
S
* Supporting Information
Text, figures, tables, and CIF files giving complete NMR
spectral data, solid-state NMR spectral data, the solid-state
structure of the potassium reduction product, and crystallo-
graphic data for 3−5 and the the potassium reduction product.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(25) Baker, R. J.; Jones, C. Dalton Trans. 2005, 1341.
(26) Green, M. L. H.; Mountford, P.; Smout, G. J.; Speel, S. R.
Polyhedron 1990, 9, 2763.
(27) There have been considerable efforts in the synthesis of
gallium−iodine phases in varying stoichiometries, including a 1:1
phase that could be considered a type of “GaI”, prior to Green’s report.
For a brief description of this history and the appropriate citations, see
the Supporting Information.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail for P.J.R.: pragogna@uwo.ca.
Notes
(28) Coban, S. Diplomarbeit; Universitat Karlsruhe, Karlsruhe,
The authors declare no competing financial interest.
Germany, 1999.
9655
dx.doi.org/10.1021/ic501139w | Inorg. Chem. 2014, 53, 9644−9656

Inorganic Chemistry
Article
(29) Widdifield, C. M.; Jurca, T.; Richeson, D. S.; Bryce, D. L.
Polyhedron 2012, 35, 96.
(30) Doriat, C. U.; Friesen, M.; Baum, E.; Ecker, A.; Schnockel, H.
patterns can be obtained much more efficiently than when traditional
procedures are used: O’Dell, L. A.; Schurko, R. W. Chem. Phys. Lett.
2008, 464, 97−102.
(58) Multinuclear NMR; Mason, J., Ed.; Plenum Press: New York and
London, 1987.
Angew. Chem., Int. Ed. 1997, 1969.
(31) Schnepf, A.; Doriat, C.; Mollhausen, E.; Schnockel, H. Chem.
̈
(59) McGarvey, B. R.; Talyor, M. J.; Tuck, D. J. Inorg. Chem. 1981,
Commun. 1997, 0, 2111.
20, 2010.
(32) Cheng, F.; Hector, A. L.; Levason, W.; Reid, G.; Webster, M.;
(60) Antcliff, K. L.; Baker, R. J.; Jones, C.; Murphy, D. M.; Rose, R. P.
Inorg. Chem. 2005, 44, 2098.
(61) Beamish, J. C.; Boardman, A.; Worrall, I. J. Polyhedron 1991, 10,
95.
Zhang, W. Inorg. Chem. 2007, 46, 7215.
(33) Thomas, J. C.; Peters, J. C. Inorg. Chem. 2003, 42, 5055.
(34) Mankad, N. P.; Antholine, W. E.; Szilagyi, R. K.; Peters, J. C. J.
Am. Chem. Soc. 2009, 131, 3878.
(62) Ali, S. M.; Brewer, F. M.; Chadwick, J. R.; Garton, G. J. Inorg.
Nucl. Chem. 1959, 9, 124.
(63) Brewer, F. M.; Chadwick, J. R.; Garton, G. J. Inorg. Nucl. Chem.
1961, 23, 45.
(64) Tuck, D. G. Polyhedron 1990, 9, 377.
(35) Thomas, J. C.; Peters, J. C. J. Am. Chem. Soc. 2001, 123, 5100.
(36) Lu, C. C.; Peters, J. C. J. Am. Chem. Soc. 2002, 124, 5272.
(37) Thomas, J. C.; Peters, J. C. J. Am. Chem. Soc. 2003, 125, 8870.
(38) Dube, J. W.; Macdonald, C. L. B.; Ragogna, P. J. Angew. Chem.,
Int. Ed. 2012, 124, 13203.
(65) Arif, A. M.; Benac, B. L.; Cowley, A. H.; Geerts, R.; Jones, R. A.;
Kidd, K. B.; John, M.; Schwab, S. T. J. Chem. Soc., Chem. Commun.
1986, 1543, 1543.
(66) Petrie, M. A.; Power, P. P. Organometallics 1993, 12, 1592.
(67) Gans-Eischler, T.; Jones, C.; Aldridge, S.; Stasch, A. Anal. Sci.
2008, 24, 109.
(39) Weicker, S. A.; Dube, J. W.; Ragogna, P. J. Organometallics 2013,
32, 6681.
(40) Dube, J. W.; Ragogna, P. J. Chem.Eur. J. 2013, 19, 11768.
(41) Dube, J. W.; Macdonald, C. L. B.; Ellis, B. D.; Ragogna, P. J.
Inorg. Chem. 2013, 52, 11438.
(42) Gerlach, V. G.; Honle, W.; Simon, A. Z. Anorg. Allg. Chem. 1982,
(68) SMART; Bruker AXS Inc., Madison, WI, USA, 2001.
(69) COLLECT; Nonius BV, Delft, The Netherlands, 2001.
(70) SAINT; Bruker AXS Inc., Madison, WI, USA, 2007.
(71) SADABS; Bruker AXS Inc.: Madison, WI, USA, 2001.
(72) Sheldrick, G. M. Acta Crystallogr., Sect. A 2008, 64, 112.
(73) Eichele, K. WSolids, 2013; http://casgm3.anorg.chemie.uni-
tuebingen.de/klaus/soft/index.php?p=wsolids1/wsolids1.
(74) Cheng, J. T.; Edwards, J. C.; Ellis, P. D. J. Phys. Chem. 1990, 94,
553.
(75) Power, W. P.; Wasylishen, R. E.; Mooibroek, S.; Pettitt, B. A.;
Danchura, W. J. Phys. Chem. 1990, 94, 591.
(76) Mason, J. Solid State Nucl. Magn. Reson. 1993, 2, 285.
(77) Stender, M.; Wright, R. J.; Eichler, B. E.; Prust, J.; Olmstead, M.
M.; Roesky, H. W.; Power, P. P. J. Chem. Soc., Dalton Trans. 2001,
3465.
486, 7.
(43) Beamish, J. C.; Wilkinson, M.; Worrall, I. J. Inorg. Chem. 1978,
17, 2026.
(44) Lind, W.; Worrall, I. J. J. Organomet. Chem. 1972, 40, 35.
(45) Lind, W.; Waterworth, L.; Worrall, I. J. Inorg. Nucl. Chem. Lett.
1971, 7, 611.
(46) Waterworth, L. G.; Worrall, I. J. J. Inorg. Nucl. Chem. 1973, 35,
1535.
(47) Wilkinson, M.; Worrall, I. J. J. Organomet. Chem. 1975, 93, 39.
(48) Woodward, L. A.; Singer, G. H. J. Chem. Soc. 1958, 716.
(49) Tan, K. H.; Taylor, M. J. Inorg. Nucl. Chem. Lett. 1974, 10, 267.
(50) Evans, C. A.; Taylor, M. J. J. Chem. Soc. D Chem. Commun. 1969,
1201.
(51) 127I nuclear quadrupole resonance (NQR) spectroscopy is an
ideal technique to study “GaI”, because each crystallographically
unique I atom will give rise to a peak at a specific transmitter
frequency. Since there are four crystallographically unique I atoms in
−
the [GaI4 ] component of GaI2 and three distinct I types in the
[Ga2I62−] component of Ga2I3, 127I NQR is able to unambiguously
52
differentiate between the two species. Previous studies on GaI2 and
GaI3,53 show the range of frequencies to be from 113.6 to 174.6 MHz,
which conveniently fall within the standard tuning range of our NMR
probes and are thus easily accessible. Although it is possible to use
solid-state 127I NMR to study “GaI”, each distinct I site will give rise to
a broad powder pattern spanning many megahertz.54 The unique I
sites in “GaI” will thus overlap and make it difficult to unambiguously
identify the number of unique I sites present in a “GaI” sample. The
advantage of 127I NQR over 127I ssNMR is that, even though the NQR
peaks are relatively broad (line widths of a few hundred kilohertz) and
featureless, they are spread over a broad frequency range and typically
do not overlap.
(52) Okuda, T.; Hamamoto, H.; Ishihara, H.; Negita, H. Bull. Chem.
Soc. Jpn. 1985, 58, 2731.
(53) Segel, S. L.; Barnes, R. G. J. Chem. Phys. 1956, 25, 578.
(54) Widdifield, C. M.; Chapman, R. P.; Bryce, D. L. In Annual
Reports on NMR Spectroscopy; Webb, G. A., Ed.; Elsevier: San Diego,
CA, 2009; Vol. 66, pp 195−326.
(55) Buslaev, Y. A.; Tarasov, V. P.; Petrosyants, S. P.; Melnikov, N.
N. Zh. Strukt. Khim. 1974, 15, 617.
(56) QCPMG NMR spectra are comprised of spikelets that mimic
the overall shape of the spin-echo powder pattern but provide a
dramatic signal enhancement in comparison to typical spin-echo
experiments: Larsen, F. H.; Jakobsen, H. J.; Ellis, P. D.; Nielsen, N. C.
J. Phys. Chem. A 1997, 101, 8597.
(57) It has been shown that when QCPMG is combined with
adiabatic WURST pulses, dramatic increases in the excitation
bandwidth of the QCPMG experiment is achieved and broad powder
9656
dx.doi.org/10.1021/ic501139w | Inorg. Chem. 2014, 53, 9644−9656