An investigation of 1:1 adducts of gallium trihalides with triarylphosphines by solid-state 69/71Ga and 31P NMR spectroscopy

Article abstract of DOI:10.1002/chem.201202954

Several 1:1 adducts of gallium trihalides with triarylphosphines, X ₃Ga(PR₃) (X=Cl, Br, and I; PR₃=triarylphosphine ligand), were investigated by using solid-state ^69/71Ga and ³¹P NMR spectroscopy at different magnetic-field strengths. The ^69/71Ga nuclear quadrupolar coupling parameters, as well as the gallium and phosphorus magnetic shielding tensors, were determined. The magnitude of the ⁷¹Ga quadrupolar coupling constants (C _Q(⁷¹Ga)) range from approximately 0.9 to 11.0 MHz. The spans of the gallium magnetic shielding tensors for these complexes, δ₁₁-δ₃₃, range from approximately 30 to 380 ppm; those determined for phosphorus range from 10 to 40 ppm. For any given phosphine ligand, the gallium nuclei are most shielded for X=I and least shielded for X=Cl, a trend previously observed for In^III-phosphine complexes. This experimental trend, attributed to spin-orbit effects of the halogen ligands, is reproduced by DFT calculations. The signs of C _Q(^69/71Ga) for some of the adducts were determined from the analysis of the ³¹P NMR spectra acquired with magic angle spinning (MAS). The ¹J(^69/71Ga,³¹P) and ΔJ(^69/71Ga, ³¹P) values, as well as their signs, were also determined; values of ¹J(⁷¹Ga,³¹P) range from approximately 380 to 1590 Hz. Values of ¹J( ^69/71Ga,³¹P) and ΔJ(^69/71Ga, ³¹P) calculated by using DFT have comparable magnitudes and generally reproduce experimental trends. Both the Fermi-contact and spin-dipolar Fermi-contact mechanisms make important contributions to the ¹J( ^69/71Ga,³¹P) tensors. The ³¹P NMR spectra of several adducts in solution, obtained as a function of temperature, are contrasted with those obtained in the solid state. Finally, to complement the analysis of NMR spectra for these adducts, single-crystal X-ray diffraction data for Br₃Ga[P(p-Anis)₃] and I₃Ga[P(p-Anis) ₃] were obtained.

Full text of DOI:10.1002/chem.201202954

DOI: 10.1002/chem.201202954
An Investigation of 1:1 Adducts of Gallium Trihalides with Triarylphosphines
31
by Solid-State 69/71Ga and P NMR Spectroscopy
Fu Chen, Guibin Ma, Guy M. Bernard, Roderick E. Wasylishen,* Ronald G. Cavell,
Robert McDonald, and Michael J. Ferguson[a]
1
Abstract: Several 1:1 adducts of galli-
um trihalides with triarylphosphines,
X3GaACHTUNGTRENNUNG(PR3) (X=Cl, Br, and I; PR3 =
phine ligand, the gallium nuclei are
most shielded for X=I and least
shielded for X=Cl, a trend previously
observed for InIII–phosphine com-
plexes. This experimental trend, attrib-
uted to spin-orbit effects of the halogen
ligands, is reproduced by DFT calcula-
also determined; values of J
G
range from approximately 380 to
1590 Hz. Values of 1J(69/71Ga,31P) and
DJ(69/71Ga,31P) calculated by using DFT
have comparable magnitudes and
triarylphosphine ligand), were investi-
gated by using solid-state 69/71Ga and
31P NMR spectroscopy at different
magnetic-field strengths. The 69/71Ga
nuclear quadrupolar coupling parame-
ters, as well as the gallium and phos-
phorus magnetic shielding tensors,
were determined. The magnitude of
the 71Ga quadrupolar coupling con-
generally
reproduce
experimental
trends. Both the Fermi-contact and
spin-dipolar Fermi-contact mechanisms
make important contributions to the
1J(69/71Ga,31P) tensors. The 31P NMR
spectra of several adducts in solution,
obtained as a function of temperature,
are contrasted with those obtained in
the solid state. Finally, to complement
the analysis of NMR spectra for these
adducts, single-crystal X-ray diffrac-
tions. The signs of CQACHUTNGRNENUG(
69/71Ga) for some
of the adducts were determined from
the analysis of the 31P NMR spectra ac-
quired with magic angle spinning
(MAS). The 1J(69/71Ga,31P) and DJ(69/71Ga,
31P) values, as well as their signs, were
stants (CQACHTUNGTRENNUNG
(71Ga)) range from approxi-
mately 0.9 to 11.0 MHz. The spans of
the gallium magnetic shielding tensors
for these complexes, d11ꢀd33, range
from approximately 30 to 380 ppm;
those determined for phosphorus range
from 10 to 40 ppm. For any given phos-
Keywords: chemical shifts
· cou-
pling constants · density functional
calculations · gallium · NMR spec-
troscopy · phosphorus
tion data for Br3Ga[P
ACHTUNGTRNEUN(NG p-Anis)3] and
I3Ga[P(p-Anis)3] were obtained.
AHCTUNGTRENNUNG
Introduction
has drawn considerable recent attention[2] because they have
many important applications, particularly in materials chem-
istry.[3,4,5] For example, they are single-source precursors for
Interest in the study of adducts formed by Group 13 Lewis
acids with Group 15 Lewis bases dates back to the early
1800s with the first synthesis of F3B·NH3 by Gay-Lussac and
Thꢀnard.[1] The preparation of adducts formed by the
Group 13 (B, Al, Ga, In, Tl) trihalides or trialkyls with
Lewis bases containing Group 15 elements (N, P, As, Sb, Bi)
preparing
a wide range of semiconductors based on
Group 13 and 15 elements.[3] In the case of gallium, GaN
and GaP are used extensively in light-emitting diodes
(LEDs),[5] GaSb is used in thermal-imagining devices[5] and
GaAs is widely used in solar cells.[5] The gallium trihalide
phosphine adducts were first prepared and characterized by
Carty and co-workers,[6] and have remained of interest be-
cause of their importance in the preparation of GaP-based
semiconductors.[7]
[a] Dr. F. Chen, Dr. G. Ma, Dr. G. M. Bernard, Prof. R. E. Wasylishen,
Prof. R. G. Cavell, Dr. R. McDonald, Dr. M. J. Ferguson
Department of Chemistry, University of Alberta
Edmonton, Alberta, T6G 2G2 (Canada)
Gallium trihalide phosphine adducts have often been
characterized by using single-crystal X-ray diffraction or vi-
brational spectroscopy.[6,8–10] Herein, we show that solid-
state 69/71Ga and 31P NMR spectroscopy are excellent com-
plementary techniques for characterizing 1:1 adducts of gal-
lium trihalides with triarylphosphines. The information
available from NMR spectroscopy includes the chemical
shift (CS) tensors, indirect spin–spin coupling constants be-
tween 69/71Ga and 31P, and the electric field gradient (EFG)
tensors for the quadrupolar nuclei, 69Ga and 71Ga. Previous-
ly, solid-state 69/71Ga NMR spectroscopy has been used to
Fax : (+1)780-492-8231
E-mail: Roderick.Wasylishen@ualberta.ca
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201202954. It contains back-
ground information on nuclear quadrupole coupling, magnetic shield-
ing, and indirect spin–spin coupling; sample preparation; quantum
chemical calculations; single-crystal X-ray diffraction data for
Br3Ga[PACHTUNGTRENNUNG(p-Anis)3] and I3Ga[PCAHTUNGTRNEN(UGN p-Anis)3]; a summary of 2d values;
computational gallium magnetic shielding parameters; additional
solid-state 31P and 69/71Ga NMR spectra for all adducts at different
magnetic-field strengths; and solution-phase 31P NMR spectra for all
adducts at different temperatures.
determine the nuclear quadrupolar coupling constants (CQ-
[11]
ACTHNGUTRNENUG(
69/71Ga)) for Ga2O3 and various semiconductors, including
2826
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 2826 – 2838

FULL PAPER
GaN,[12] GaP,[13] and GaAs;[14] nevertheless, the study of gal-
lium complexes by using solid-state 69/71Ga NMR spectrosco-
py is limited.[15] Gallium isotropic chemical shifts (diso(Ga))
range over approximately 1400 ppm;[16] therefore, one might
expect gallium chemical shift anisotropy (CSA) to be large.
However, there have only been a few reports of gallium
CSAs.[11a,17]
Previously, we undertook a detailed solid-state NMR in-
vestigation of some related indium trihalide phosphine ad-
ducts.[18,19] In those studies, we showed that a careful analysis
of 115In and 31P NMR spectra could provide considerable in-
formation about these adducts. Hence, a goal of this study is
to determine whether the similarity of the gallium trihalide
triarylphosphines with the 1:1 indium–phosphine adducts in-
vestigated in the earlier studies will be reflected in their
NMR properties. Specifically, 1:1 adducts of gallium triha-
pling interactions between 69/71Ga and coupled 31P nuclei
(I=1= , NA=100%).
2
NMR spectroscopists typically measure chemical shifts, d,
the nuclear magnetic shielding of a given nucleus relative to
that for a reference compound: dꢁsrefꢀssample, in which the
latter two terms refer to the magnetic shielding of the refer-
ence compound and of the sample, respectively.[21] When
discussing NMR spectra, it is convenient to use three chemi-
cal shift parameters: the isotropic chemical shift, diso =(d11 +
d22 +d33)/3, in which d11 ꢂd22 ꢂd33, the span, W=d11ꢀd33, de-
scribing the maximum orientation dependence of the mag-
netic shielding interaction, and the skew, k=3
G
which is unitless with ꢀ1ꢃkꢃ +1;[21] note that for an axially
symmetric chemical shift tensor, k= ꢄ1.
The effects of the nuclear quadrupolar coupling interac-
tion on solid-state NMR spectra of quadrupolar and spin-1=
lide with a triarylphosphine, X3Ga
G
nuclei have been discussed extensively;[22] a summary is also
presented in the Supporting Information. Likewise, readers
are encouraged to consult the original literature[22d,23,24] or
Supporting Information for a detailed discussion of the ef-
fects of direct and indirect spin–spin interactions on NMR
spectra.
PR3 =triarylphosphine ligand), are characterized by using
solid-state 69/71Ga and 31P NMR spectroscopy; these NMR
results are corroborated by the results of density functional
theory (DFT) calculations. Possible causes for the observed
sensitivity of the gallium CS tensors to the nature of the hal-
ogen ligands are considered. In addition, analyses of the
31P NMR spectra allowed the determination of J(69/71Ga,31P)
31P NMR spectroscopy of the 69/71Ga–31P spin pair: A
31P NMR spectrum spin–spin coupled to either 69Ga or 71Ga
should consist of four peaks (Figure 1).[25] If CQ is significant,
the direct dipolar interaction for the 69/71Ga–31P spin pair is
and DJ(69/71Ga,31P), that is, the anisotropy in J(69/71Ga,31P).
1
The signs for DJ(69/71Ga,31P), J(69/71Ga,31P), and C
were also determined for these adducts. In addition, since
knowledge of the structures of the adducts under study is in-
valuable for a proper analysis of the 69/71Ga and 31P NMR
not averaged to zero by magic angle spinning (MAS),[25]
a
consequence of the fact that the gallium nuclei are not
quantized exactly along B0.[26] In this case, one observes an
uneven spacing between adjacent peaks (Figure 1); analysis
of these spectra yield the residual dipolar coupling (d),
which is directly related to CQ and Reff. In favorable cases,
such analyses will also yield the sign of CQ; see the Support-
ing Information for a presentation of the underlying theory.
spectra, single-crystal X-ray diffraction data for Br3Ga[PACHTGNUTRENNUNG
Anis)3] and I3Ga[P
are also presented.
ACHTUNGTRENNUNG(p-Anis)3] (Anis=anisole, C6H4ACHTUNGTRENNUNG
Gallium NMR spectroscopy: Gallium has two naturally oc-
curring isotopes: 69Ga and 71Ga. The nuclear-spin properties
and natural abundance (NA) of the two isotopes, 69Ga (S=
3/2, X=24.00%, Q=17.1 fm2, NA=60.1%) and 71Ga (S=3/2,
X=30.50%, Q=10.7 fm2, NA=
39.9%),[20] make them potential
candidates for NMR spectro-
scopic study. In the solid state,
71Ga is the preferred gallium
NMR isotope because of its
larger Larmor frequency and
receptivity, as well as its smaller
nuclear quadrupolar moment.
Nevertheless, in many studies,
including this work, both 69Ga
and 71Ga are investigated. In an
applied magnetic field, B0, the
69/
interactions involving the
71Ga nuclei of X3Ga
ACTHNUTRGENNU(G PR3) in-
clude the Zeeman, nuclear
quadrupolar, and nuclear mag-
netic shielding interactions, as
well as the direct dipolar and
Figure 1. Calculated splitting pattern in a 31P NMR spectrum of a MAS sample containing a 69/71Ga–31P spin
1
pair with d=0, d>0, and d<0. The spin states as indicated are those for a positive sign for J(69/71Ga,31P), but
indirect nuclear spin–spin cou- the splitting pattern is invariant with this sign.
Chem. Eur. J. 2013, 19, 2826 – 2838
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2827

R. E. Wasylishen et al.
Results and Discussion
Because of the importance of molecular structure in the
analysis of the NMR results, those for the triarylphosphine
gallium trihalide adducts are presented first, followed by a
discussion of the solid-state 69/71Ga NMR spectra of each
1
adduct. Next, the J(69/71Ga,31P) and DJ(69/71Ga,31P) values, as
well as their signs, and the CQACHUTNGRNENUG(
69/71Ga) values for all com-
plexes, determined through 31P NMR spectroscopy, are re-
ported. Finally, results of relativistic DFT calculations of the
NMR parameters for these compounds are compared with
experimental values and the causes of the observed sensitivi-
ty of the gallium CS tensors to the nature of the directly
bonded halogen ligands are explored.
Structures of the triarylphosphine gallium trihalide adducts:
Structures for Cl3Ga
(PPh3),[9] obtained from X-ray crystallography, have been re-
ported; those for Br3Ga[P(p-Anis)3] and I3Ga[P(p-Anis)3]
(PPh3),[8] Br3Ga(PPh3),[8] and I3Ga-
ACHUTGTNRENNUG CAHTUNGTRENNUGN
ACHTUNGTRENNUNG
A
ACHTUNGTRENNUNG
were determined as part of this study and are shown in
Figure 2. See Tables S1 and S2 in the Supporting Informa-
tion for more detailed X-ray diffraction data and for select-
ed structural information. All five adducts studied by X-ray
crystallography have approximately tetrahedral coordination
Figure 2. Molecular structures of Br3Ga[P(p-Anis)3] (top) and I3Ga[P(p-
Anis)3] (bottom). For clarity, the hydrogen atoms are not shown.
ꢀ
nent of each tensor along the Ga P bond; hence hQ =0 and
about the Ga and P atoms; there is a C3 axis along the Ga
P bond of I3Ga(PPh3) and Br3Ga[P(p-Anis)3] and an ap-
k= ꢄ1. The central transition peaks in the 71Ga NMR spec-
ꢀ
N
U
tra of MAS and stationary samples of Br3Ga[PACTHNGUTRE(NNUG p-Anis)3],
proximate C3 axis along this bond for the other adducts.
We were unable to obtain crystals suitable for X-ray dif-
shown in Figure 3a and b, respectively, are broad and fea-
tureless; the broadening, attributed to the 71Ga–79/81Br resid-
fraction studies for the adduct of GaCl3 with [P
N
ual dipolar interactions (79Br: I=3= , X=25.05%, NA=
2
and for the three adducts containing the tris(2,4,6-trimethox-
yphenyl)phosphine (TMP) ligand. However, elemental anal-
50.69%; 81Br: I=3= , X=27.01%, NA=49.31%),[20] leads to
2
difficulties in determining the CQACTHNUTRGNEUNG
(71Ga) value. The value for
ysis indicated that both moieties were present in a 1:1 ratio
CQACHTUNTGRENNUNG
(71Ga), (+1.00ꢄ0.10) MHz, was obtained by analyzing
31
the full 71Ga NMR spectrum of a MAS sample, including
ꢀ
and solid-state P NMR spectroscopy confirmed the Ga P
connectivity.
Table 1. Experimental 69/71Ga NMR parameters for X3Ga(PR3) (X=Cl, Br, and I) adducts.
69/71Ga NMR of solid triaryl-
phosphine gallium adducts:
69/71Ga NMR parameters deter-
mined from analyses of the
spectra for the triarylphosphine
gallium trihalide adducts inves-
tigated herein are listed in
Table 1. See Figures S1–S4 in
the Supporting Information, as
well as those shown below, for
examples of Ga NMR spectra.
Gallium chemical shift tensor
diso
d11
d22
d33
W
k
[ppm]
[ppm]
[ppm]
[ppm]
[ppm]
Cl3Ga(PPh3)
Br3Ga(PPh3)
I3Ga(PPh3)
Cl3Ga[P(p-Anis)3]
Br3Ga[P(p-Anis)3]
I3Ga[P(p-Anis)3]
Cl3Ga(TMP)
Br3Ga(TMP)
I3Ga(TMP)
260.0ꢄ10.0
137.0ꢄ10.0
357.5ꢄ10.0
178.2ꢄ10.0
215.0ꢄ10.0
119.7ꢄ10.0
207.5ꢄ10.0 150.0ꢄ15.0 ꢀ0.90ꢄ0.10
113.2ꢄ10.0
65.0ꢄ15.0 ꢀ0.80ꢄ0.10
ꢀ150.0ꢄ10.0 ꢀ30.0ꢄ10.0 ꢀ30.0ꢄ10.0 ꢀ390.0ꢄ10.0 360.0ꢄ15.0
1.00
256.0ꢄ3.0
138.0ꢄ7.0
342.0ꢄ3.0
164.7ꢄ7.0
ꢀ10.0ꢄ7.0
295.0ꢄ10.0
140.0ꢄ10.0
220.0ꢄ3.0
124.7ꢄ7.0
ꢀ31.0ꢄ7.0
196.0ꢄ10.0
111.0ꢄ10.0
207.0ꢄ3.0
124.7ꢄ7.0
135.0ꢄ4.0
ꢀ0.80ꢄ0.10
40.0ꢄ10.0 ꢀ1.00
ꢀ135.0ꢄ7.0
225.0ꢄ10.0
120.0ꢄ10.0
ꢀ140.0ꢄ10.0
ꢀ365.0ꢄ7.0
355.0ꢄ10.0
0.88ꢄ0.10
185.0ꢄ10.0 110.0ꢄ15.0 ꢀ0.80ꢄ0.10
110.0ꢄ10.0
30.0ꢄ15.0 ꢀ0.90ꢄ0.10
ꢀ7.0ꢄ10.0 ꢀ26.0ꢄ10.0 ꢀ387.0ꢄ10.0 380.0ꢄ15.0
0.90ꢄ0.10
The signs for CQACHTUNGTRENNUNG(
69/71Ga) report-
Gallium quadrupolar coupling
CQ(69Ga)
CQ(71Ga)
hQ
a
b
g
ed in this section were deter-
mined from an analysis of the
31P NMR spectra of MAS sam-
ples, or from the results of DFT
calculations (see below).
[MHz]
[MHz]
[8]
[8]
[8]
Cl3Ga(PPh3)
Br3Ga(PPh3)
I3Ga(PPh3)
Cl3Ga[P(p-Anis)3]
Br3Ga[P(p-Anis)3]
I3Ga[P(p-Anis)3]
Cl3Ga(TMP)
Br3Ga(TMP)
I3Ga(TMP)
+3.00ꢄ0.20
ꢄ1.45ꢄ0.20
+5.20ꢄ0.20
+2.40ꢄ0.10
ꢄ1.60ꢄ0.10
ꢀ6.80ꢄ0.20
ꢀ14.30ꢄ0.40
ꢀ17.40ꢄ0.40
ꢀ17.10ꢄ0.40
+1.90ꢄ0.20
ꢄ0.90ꢄ0.20
+3.30ꢄ0.20
+1.50ꢄ0.10
ꢄ1.00ꢄ0.10
ꢀ4.25ꢄ0.20
ꢀ9.00ꢄ0.40
ꢀ11.00ꢄ0.40
ꢀ10.70ꢄ0.40
0.24ꢄ0.10
0.10ꢄ0.10
0
65ꢄ10
80ꢄ10
0
75ꢄ10
50ꢄ10
0
32ꢄ5
30ꢄ10
0
0.16ꢄ0.10
80ꢄ10
85ꢄ5
5ꢄ10
Because Br3Ga[PACHTUNRGTNEUNG(p-Anis)3]
has a C3 symmetry axis along
0
–
90
0
0.15ꢄ0.10
0.20ꢄ0.10
0.24ꢄ0.10
0.35ꢄ0.10
75ꢄ15
80ꢄ10
30ꢄ10
60ꢄ10
15ꢄ10
85ꢄ5
85ꢄ5
10ꢄ10
45ꢄ10
10ꢄ10
10ꢄ10
50ꢄ10
ꢀ
the Ga P bond, the Ga CS and
EFG tensors are axially sym-
metric with the unique compo-
2828
www.chemeurj.org
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 2826 – 2838

69/71Ga and 31P NMR Spectroscopy of X3Ga
ACTHNUTRGENUN(G PR3)
FULL PAPER
arising from quadrupolar coupling cannot be discerned in
these spectra, but the 71Ga NMR spectrum of a stationary
sample obtained at 21.14 T (Figure 3b) is asymmetric; this is
attributed to anisotropic gallium magnetic shielding. This al-
lowed the determination of the span of the gallium CS
tensor, (40.0ꢄ10.0) ppm, and of k, ꢀ1.00. Thus, the unique
components of the Ga CS and EFG tensors, d11 and VZZ, re-
spectively, are along the C3 axis, in agreement with results of
the DFT calculations.
The central transition peaks in the 71Ga NMR spectra of
both MAS and stationary samples (Figure S3 in the Support-
ing Information) of Br3Ga
ACHTUNGRTEN(NUNG PPh3) are similar to those of
Br3Ga[P(p-Anis)3], and thus were analyzed in the same
AHCTUNGTRENNUNG
manner. The broad and featureless peaks are also attributed
to the 71Ga–79/81Br residual dipolar interactions. This adduct
ꢀ
has an approximate C3 symmetry axis about the Ga P
bond.[8] Analysis of the 71Ga NMR spectra yielded similar
values as for Br3Ga[P
G
C
(71Ga)=(ꢄ0.90ꢄ
Figure 3. Experimental (lower traces) and calculated (upper traces) cen-
tral transition in the 71Ga NMR spectra of a) MAS and b) stationary
powder samples of Br3Ga[P(p-Anis)3] acquired at 11.75 and 21.14 T.
c) The NMR spectrum of the central and satellite transitions for a MAS
sample (20.0 kHz) acquired at 21.14 T is shown in the lower trace with
the simulated spectrum of a stationary sample shown in the upper trace;
the peak arising from the central transition has been truncated in these
images to emphasize the line shape of the spectrum arising from the sat-
ellite transitions.
0.20) MHz, hQ =(0.10ꢄ0.10), W=(65.0ꢄ15.0) ppm, and k=
(ꢀ0.80ꢄ0.10). Hence, the EFG and CS tensors are almost
axially symmetric with VZZ coincident with d11, both aligned
approximately along the Ga P bond.
Ga NMR spectra of the central transition of stationary
samples of I3GaACTHNUTRGNENUG(PPh3) obtained at different magnetic-field
strengths are shown in Figure 4. From the simulations, it is
apparent that for 71Ga at 7.05 T, the broadening arising from
the magnetic shielding anisotropy is somewhat greater than
that arising from the second-order nuclear quadrupolar in-
teraction, the opposite of what is observed for the 69Ga
the satellite transitions, as shown in Figure 3c. The small
value for C
that is, a negligible d value (see below). Powder line shapes
QACHTUNGTRENNUNG
(71Ga) is consistent with the 31P NMR spectra,
Figure 4. Experimental (lower traces) and calculated (upper traces) central transition in the a) 71Ga and b) 69Ga NMR spectra of stationary powder sam-
ples of I3Ga(PPh3) acquired at 7.05, 11.75, and 21.14 T.
Chem. Eur. J. 2013, 19, 2826 – 2838
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2829

R. E. Wasylishen et al.
nuclei of this sample. For 69Ga NMR spectra acquired at
11.75 T, the broadening arising from these two interactions
is comparable. The contribution from the second-order
quadrupolar interaction is negligible for 71Ga NMR spectra
acquired at 21.14 T. The differences discussed above arise
because 69Ga has a larger nuclear quadrupole moment and a
smaller magnetic moment, combined with the fact that, in
frequency units, the effects of the second-order quadrupolar
interaction are inversely proportional to B0, whereas those
of the magnetic shielding anisotropy are proportional to B0.
The value of CQACHTUNGTRENNUNG
(71Ga) is (+3.30ꢄ0.20) MHz and W=
(360.0ꢄ15.0) ppm. Both the EFG and CS tensors at Ga in
I3GaACHTUNGTRENNUNG(PPh3) are axially symmetric, as expected from the
presence of a C3 symmetry axis, with hQ =0 and k=1.00.
ꢀ
The skew indicates that d33 is oriented along the Ga P
bond, in contrast to Br3Ga
A
G
which d11 is parallel or approximately parallel to the this
bond. DFT calculations reproduce these orientations (see
below).
Figure 6. Experimental (lower traces) and calculated (upper traces) cen-
tral transition in the 71Ga NMR spectra of a) MAS and b) stationary
powder samples of Cl3Ga[P(p-Anis)3] acquired at 7.05 and 11.75 T. The
corresponding 69Ga spectra of this compound are shown in c) (MAS) and
d) (stationary).
The Ga NMR spectra of the central transition of station-
ary samples of I3Ga[P
magnetic-field strengths (Figure 5), are similar to those of
I3Ga(PPh3). The CQA
(71Ga) value is (ꢀ4.25ꢄ0.20) MHz and
W=(355.0ꢄ10.0) ppm; both the EFG and CS tensors at Ga
in I3Ga[P(p-Anis)3] are close to axially symmetric (hQ =
ACHTUNGTNER(NUNG p-Anis)3], obtained at three different
A
CHTUNGTRENNUNG
CQACHTUGNTRENNNUG
(71Ga) value, (+1.50ꢄ0.10) MHz, was determined from
T
an analysis of the 71Ga spectra of a MAS sample. The
breadth of the peaks in the 71Ga NMR spectra of a station-
ary sample, acquired at 7.05 and 11.75 T, plotted on the ppm
scale in Figure 6b, are almost equal; this observation is a
clear indication that magnetic shielding anisotropy domi-
(0.15ꢄ0.10) and k=(0.88ꢄ0.10)), as expected from the ap-
proximate C3 symmetry axis that includes the gallium atom.
The values of the asymmetry parameters indicate that d33
ꢀ
and VZZ are oriented approximately along the Ga P bond.
69/71Ga NMR spectra of Cl3Ga[P
(p-Anis)3] are shown in
nates the 71Ga NMR spectra of Cl3Ga[P
ACTHNGUTER(NNUG p-Anis)3]. The span
Figure 6. The splitting in the 71Ga NMR spectrum of a MAS
of the Ga CS tensor is (135.0ꢄ4.0) ppm. The gallium hQ
1
sample obtained at 11.75 T is attributed to J
(71Ga,31P). The
and k values, (0.16ꢄ0.10) and (ꢀ0.80ꢄ0.10), respectively,
Figure 5. Experimental (lower traces) and calculated (upper traces) central transition in the a) 71Ga and b) 69Ga NMR spectra of stationary powder sam-
ples of I3Ga[P(p-Anis)3] acquired at 7.05, 11.75, and 21.14 T.
2830
www.chemeurj.org
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 2826 – 2838

69/71Ga and 31P NMR Spectroscopy of X3Ga
ACTHNUTRGENUN(G PR3)
FULL PAPER
indicate that the EFG and CS
tensors at Ga are close to axial-
ly symmetric with d11 and VZZ
both approximately in the di-
ꢀ
rection of the Ga P bond.
71Ga NMR spectra of Cl3Ga-
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
C CHTUNGTRENNUNG
0.2) MHz, is slightly larger in
magnitude than that for the
latter, as is the span of the Ga
CS tensor, (150ꢄ15) ppm. The
gallium hQ and k values, (0.24ꢄ
0.10) and (ꢀ0.9ꢄ0.1), respec-
tively, indicate that the EFG
and CS tensors at Ga have a
similar orientation to those for
Cl3Ga[P
above.
G
discussed
The Ga NMR spectra of the
central transition of stationary
samples of X3GaACHTUNGTRENNUNG(TMP) (X=
Cl, Br and I), obtained at
11.75 T, are shown in Figure 7;
the results of the analyses are
summarized in Table 1. For a
given adduct, the Ga NMR
spectra have similar line shapes
apart from the positions of the
discontinuities in each spec-
trum, suggesting that they have
similar Ga EFG tensors, but
different Ga CS tensors.
Figure 7. Experimental (lower traces) and calculated (upper traces) central transition in the 71Ga (left) and
69Ga NMR (right) spectra of stationary powder samples of a) Cl3Ga(TMP), b) Br3Ga(TMP), and
c) I3Ga(TMP), acquired at 11.75 T.
The jC
(71Ga)j values for the
gallium complexes investigated
herein range from 0.9 to
11.0 MHz (Table 1). These are
greatest for the adducts con-
taining the bulky TMP ligand,
but otherwise there is no clear
relationship between the galli-
um CQ value and the structures
of the adducts. A comparison of
adducts with the same triaryl-
phosphine ligand indicates that
Figure 8. Spans of the CS tensors for X3Ga(TMP), X3Ga[P(p-Anis)3], and X3Ga(PPh3).
the gallium nuclei are most and least shielded for X=I and
Cl, respectively, as shown in Figure 8 and summarized in
Table 1; this has been attributed to the spin-orbit effect of
the halogen ligand.[27] Reported Ga chemical shift values for
largest for X=I (355 to 380 ppm) and smallest for X=Br
(30 to 65 ppm).
Solid-state 31P NMR spectroscopy: With a couple of excep-
tions, the 1J(69/71Ga,31P) and Reff(69/71Ga,31P) values for the ad-
ducts considered herein are too small to be extracted from
the Ga NMR spectra of stationary samples, but are available
from 31P NMR spectra acquired with MAS. Figures 9–11
provide examples of 31P NMR spectra for several X3Ga-
X3GaACHTUNGTRENNUNG(PPh3) (X=Cl, Br, and I) in CD2Cl2 are 264, 152, and
ꢀ151 ppm, respectively,[8] similar to the values and the trend
obtained in this study. The same trend in indium chemical
shifts has also been observed for the indiumACTHUNRTGNEUNG(III) trihalide
triarylphosphine adducts.[18] The spans of the gallium mag-
netic shielding tensors for the adducts considered herein are
AHCTUNTGREGN(NUN PR3) adducts. From the relative magnetogyric ratios for
Chem. Eur. J. 2013, 19, 2826 – 2838
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2831

R. E. Wasylishen et al.
d=(3CQReff)/ACTHNUTRGNEUNG
(10vS)[18] (see also
the Supporting Information).
Table S3 in the Supporting In-
formation summarizes the 2d
values obtained in this study.
With CQ determined from the
Ga NMR spectra and
d de-
termined from the 31P NMR
spectra of MAS samples,
Reff(69/71Ga,31P) values can be
determined from analyses of
the 31P NMR spectra. In our
earlier work,[18] we also demon-
strated that a 2.5% correction
for librational effects[28] on
measured values of Reff or RDD
was sufficient for heavier atoms
such as those in the Ga–P spin
pairs considered herein. Hence,
experimental dipolar couplings
Figure 9. Experimental (lower traces) and calculated (upper traces) 31P NMR spectra of MAS powder samples
of a) Cl3Ga[P(p-Anis)3], b) Cl3Ga(TMP), c) I3Ga[P(p-Anis)3], and d) Br3Ga(TMP) acquired at 7.05 T.
69Ga and 71Ga, 31P NMR spectra of MAS samples are ex-
pected to consist of two 1:1:1:1 quartets, arising from indi-
rect spin–spin coupling to 69Ga
obtained from 31P NMR spectra have been reduced by
2.5%.
and 71Ga, with an intensity ratio
Table 2. Experimental 31P NMR parameters for X3Ga(PR3) (X=Cl, Br, and I) adducts.
of 3:2, respectively, based on
the natural abundances of the
two gallium isotopes. Thus, one
may expect to see eight peaks
in the isotropic region of a
31P NMR spectrum, as shown in
Figure 9a. However, in most
cases, not all the peaks are re-
solved as a consequence of the
similar magnetogyric ratios for
69Ga and 71Ga, combined with
relatively small 1J(69/71Ga,31P)
values and in some cases, broad
peaks. Another important spin–
spin parameter for the present
work, Reff(69/71Ga,31P), may be
estimated from 31P NMR spec-
tra of either stationary or MAS
samples. The parameters de-
rived from analyses of the
31P NMR spectra are summar-
ized in Table 2.
Phosphorus chemical shift tensor
diso
d11
d22
d33
W
k
[ppm]
[ppm]
[ppm]
[ppm]
[ppm]
Cl3Ga(PPh3)
Br3Ga(PPh3)
I3Ga(PPh3)
Cl3Ga[P(p-Anis)3]
Br3Ga[P(p-Anis)3]
I3Ga[P(p-Anis)3]
Cl3Ga(TMP)
Br3Ga(TMP)
I3Ga(TMP)
ꢀ8.1ꢄ1.0
ꢀ13.5ꢄ1.0
ꢀ28.5ꢄ1.0
ꢀ18.9ꢄ1.0
ꢀ26.6ꢄ1.0
ꢀ32.4ꢄ1.0
ꢀ41.4ꢄ1.0
ꢀ41.7ꢄ1.0
ꢀ54.1ꢄ1.0
5.6ꢄ1.0
ꢀ14.4ꢄ1.0
ꢀ19.8ꢄ1.0
ꢀ35.2ꢄ1.0
ꢀ30.3ꢄ1.0
ꢀ39.9ꢄ1.0
ꢀ39.4ꢄ1.0
ꢀ47.3ꢄ1.0
ꢀ45.7ꢄ1.0
ꢀ56.8ꢄ1.0
ꢀ15.4ꢄ1.0
ꢀ20.8ꢄ1.0
ꢀ35.2ꢄ1.0
ꢀ32.2ꢄ1.0
ꢀ39.9ꢄ1.0
ꢀ43.9ꢄ1.0
ꢀ49.5ꢄ1.0
ꢀ47.2ꢄ1.0
ꢀ57.8ꢄ1.0
21.0ꢄ2.0
21.0ꢄ2.0
20.0ꢄ2.0
38.0ꢄ2.0
40.0ꢄ2.0
30.0ꢄ2.0
22.0ꢄ2.0
15.0ꢄ2.0
10.0ꢄ2.0
ꢀ0.90ꢄ0.10
ꢀ0.90ꢄ0.10
ꢀ1.00
0.2ꢄ1.0
ꢀ15.2ꢄ1.0
5.8ꢄ1.0
ꢀ0.90ꢄ0.10
ꢀ1.00
0.1ꢄ1.0
ꢀ13.9ꢄ1.0
ꢀ27.5ꢄ1.0
ꢀ32.2ꢄ1.0
ꢀ47.8ꢄ1.0
ꢀ0.70ꢄ0.20
ꢀ0.80ꢄ0.10
ꢀ0.80ꢄ0.10
ꢀ0.80ꢄ0.10
69Ga–31P spin–spin coupling
1J(69Ga,31P)
Reff(69Ga,31P)
Rdd(69Ga,31P)
[Hz]
DJ(69Ga,31P)
[Hz]
[Hz]
[Hz][a]
Cl3Ga(PPh3)
Br3Ga(PPh3)
I3Ga(PPh3)
Cl3Ga[P(p-Anis)3]
Br3Ga[P(p-Anis)3]
I3Ga[P(p-Anis)3]
Cl3Ga(TMP)
Br3Ga(TMP)
I3Ga(TMP)
780ꢄ20
620ꢄ20
300ꢄ20
770ꢄ20
640ꢄ20
480ꢄ20
1250ꢄ20
1120ꢄ20
850ꢄ20
650ꢄ50
550ꢄ50
550ꢄ50
750ꢄ50
580ꢄ50
700ꢄ50
520ꢄ50
520ꢄ50
500ꢄ50
860ꢄ20
840ꢄ20
810ꢄ20
860ꢄ20[b]
730ꢄ20
830ꢄ20
860ꢄ20[b]
730ꢄ20[b]
810ꢄ20[b]
630ꢄ150
870ꢄ150
780ꢄ150
330ꢄ150
450ꢄ150
390ꢄ150
1020ꢄ150
630ꢄ150
930ꢄ150
71Ga–31P spin–spin coupling
From the values for Reff-
1J(71Ga,31P)
Reff(71Ga,31P)
Rdd(71Ga,31P)
[Hz]
DJ(71Ga,31P)
[Hz]
69/71Ga,31P) and Rdd(69/71Ga,31P),
[Hz]
[Hz][a]
(
the latter of which can be calcu-
Cl3Ga(PPh3)
Br3Ga(PPh3)
I3Ga(PPh3)
Cl3Ga[P(p-Anis)3]
Br3Ga[P(p-Anis)3]
I3Ga[P(p-Anis)3]
Cl3Ga(TMP)
Br3Ga(TMP)
I3Ga(TMP)
990ꢄ20
780ꢄ20
380ꢄ20
980ꢄ20
810ꢄ20
610ꢄ20
1590ꢄ20
1420ꢄ20
1080ꢄ20
830ꢄ50
700ꢄ50
700ꢄ50
950ꢄ50
740ꢄ50
890ꢄ50
660ꢄ50
660ꢄ50
630ꢄ50
1090ꢄ20
1060ꢄ20
1030ꢄ20
1090ꢄ20[b]
930ꢄ20
780ꢄ150
1080ꢄ150
990ꢄ150
420ꢄ150
570ꢄ150
480ꢄ150
1290ꢄ150
810ꢄ150
1200ꢄ150
ꢀ
lated from the Ga P bond
length, DJ(69/71Ga,31P) may in
some cases be estimated. In an
earlier study, we showed that if
the J and dipolar tensors are
coincident with an exact or ap-
proximate C3 symmetry axis in-
corporating the spin pair, then [a] Reff
1050ꢄ20
1090ꢄ20[b]
930ꢄ20[b]
1030ꢄ20[b]
(
69/71Ga,31P) values were obtained by correcting the observed values by ꢀ2.5%. [b] Estimated (see text).
2832
www.chemeurj.org
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 2826 – 2838

69/71Ga and 31P NMR Spectroscopy of X3Ga
ACTHNUTRGENUN(G PR3)
FULL PAPER
Figure 11. Experimental (lower traces) and calculated (upper traces)
31P NMR spectra of a) MAS and b) stationary powder samples of
I3Ga(PPh3) acquired at 4.70 and 7.05 T.
Figure 10. Experimental (lower traces) and calculated (upper traces)
31P NMR spectra of a) MAS and b) stationary powder samples of
Br3Ga[P(p-Anis)3] acquired at 4.70, 7.05, and 11.75 T.
significantly affect the values obtained. As discussed above,
these analyses assumed that the signs of 1J(69/71Ga,31P) for
these adducts are positive; DFT calculations (see below)
support this assumption.
A detailed discussion of the analysis of the data for the
Br3Ga[PACHTUNGTRENNUNG(p-Anis)3] adduct (Figure 10) is presented as an ex-
ample. The equally spaced peaks in this spectrum indicate
that d is negligible and therefore that the gallium CQ values
are very small, precluding an experimental assignment of a
sign for this interaction; these were thus assigned from the
ꢀ
Although Ga P bond lengths are unavailable for several
adducts considered herein, examination of the data for simi-
lar adducts in References [8] and [9] and in this study
(Table S2 in the Supporting Information) indicate that there
DFT calculations (see below). Values of (810ꢄ20) and
1
ꢀ
(740ꢄ50) Hz for J
G
G
is little variation in the Ga P bond lengths for a given
were obtained from analyses of the spectra of MAS and sta-
halide. Since Reff values, determined for all adducts, are of
comparable magnitudes (Table 2), DJ(69/71Ga,31P) is expected
to be significant in all cases considered herein. Hence,
values for DJ(69/71Ga,31P) have been estimated based on the
ꢀ
tionary samples (Figure 10). From the known Ga P bond
length, 2.500 ꢂ, an uncorrected value of (950ꢄ20) Hz was
obtained for RDD; this is reduced to (930ꢄ20) Hz after ap-
plying the correction for librational motion discussed above.
Furthermore, since this adduct contains a C3 symmetry axis
smallest corrected value of RDD (and thus the smallest value
31
for DJ(69/71Ga, P)) predicted from the Ga P bond lengths
ꢀ
ꢀ
along its Ga P bond, DJ can be calculated by using DJ=
(RDDꢀReff), allowing an estimate of DJ
for a given halide; these are summarized in Table 2.
Previously, the nature of the bond between a spin–spin
coupled pair has been explained based on the assumption
that the coupling is dominated by the Fermi-contact mecha-
nism.[33] Although a similar conclusion appears to be consis-
tent with the results of DFT calculations discussed below, it
is important to consider that the Fermi contact term does
not contribute to DJ;[23,24] hence the observation of DJ(69/
71Ga,31P) for these adducts indicates that mechanisms other
than the Fermi-contact term also make significant contribu-
tions to 1J(69/71Ga,31P).[24]
3
G
(71Ga,31P), which de-
ACHTUNGTRENNUNG
pends on the sign of Reff. Values for DJ of (+570ꢄ150) or
(+5010ꢄ150) Hz are thus obtained. An analysis of
31P NMR spectra of stationary samples showed that 1J-
ACHTUNGTRENNUNG CHTUNGTRENNUNG
(71Ga,31P) and R (71Ga,31P) have the same sign. Since one-
effA
bond spin–spin coupling constants in analogous compounds
are positive,[18,29,30] 1J(71Ga,31P) and Reff are expected to be
positive[31,32] and thus, the smaller of the two possible values
for DJ is more probable. Hence, DJ
(71Ga,31P) and 1J-
(71Ga,31P) are both positive and have similar magnitudes.
This conclusion is supported by calculations (see below).
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
ACHTUNGTRENNUNG
1
To summarize, J(69/71Ga,31P) and DJ(69/71Ga,31P) values, as
ꢀ
The I3Ga
(PPh3) adduct also has a C3 axis along the Ga P
well as their signs, have been determined for all X3Ga
adducts investigated herein. These are of comparable magni-
tude and are positive. Signs for CQA
69/71Ga) have also been
ACHTUNGTRENNUNG(PR3)
CTHUNGTRENNUNG(
A
determined for those adducts that have C3 symmetry and for
which the 69/71Ga–31P residual dipolar coupling was observa-
ble.
CTHUNGTRENNUNG
C
CHTUNGTRENNUNG
These results agree with those obtained from DFT calcula-
tions (see below). A comparable analysis was performed for
Solution-phase 31P NMR spectroscopy: 31P{1H} NMR spectra
of Br3Ga[PACHTNUGTRENUNG(p-Anis)3] dissolved in deuterated chloroform
Cl3Ga
(PPh3) and I3Ga[P
E
and acquired at various temperatures are shown in Figure S9
of the Supporting Information. At 323 K, two overlapping
The analysis outlined above assumes that the J and dipo-
lar tensors are coincident and axially symmetric. Significant
differences between Reff and RDD must be a consequence of
DJ(69/71Ga,31P), and since the analysis yields comparable
values of DJ(69/71Ga,31P) for closely related adducts, the devi-
ation from axial symmetry is not thought to be sufficient to
1
1
multiplets due to J
G
E
1
The value of J
N
error 10%) in agreement with the value obtained in the
solid state, (810ꢄ20) Hz (see Table 2). At temperatures less
than 300 K, the 69/71Ga–31P spin–spin coupling splitting pat-
Chem. Eur. J. 2013, 19, 2826 – 2838
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2833

R. E. Wasylishen et al.
Table 4. Calculated 69/71Ga NMR parameters for X3Ga(PR3) and model X3Ga(PMe3)
terns collapse as a result of rapid 69/71Ga spin-lattice
relaxation.[34] At 213 K, the 69/71Ga nuclei are essen-
tially self-decoupled from the 31P nuclei (i.e.,
(X=Cl, Br, and I) adducts.[a]
Gallium chemical shift tensors
diso
{2p[J(69/71Ga,31P)][T
results were obtained for I3Ga[P
ure S10 in the Supporting Information) and for
I3Ga(PPh3).
(
69/71Ga)]}2 !0.1).[34,35] Similar
d11
d22
d33
W
k
[ppm]
[ppm]
[ppm]
[ppm]
[ppm]
AHCTUNGTRENNUNG
Cl3Ga(PPh3)
Br3Ga(PPh3)
I3Ga(PPh3)
Cl3Ga[P(p-Anis)3]
Br3Ga[P(p-Anis)3]
I3Ga[P(p-Anis)3]
Cl3Ga(TMP)
Br3Ga(TMP)
I3Ga(TMP)
344
171
473
205
ꢀ30
468
152
ꢀ8
291
156
ꢀ30
295
119
ꢀ34
278
86
ꢀ87
304
138
ꢀ15
270
153
203
52
545
200
33
517
156
62
547
180
22
ꢀ0.79
ꢀ0.89
1.00
ꢀ0.73
ꢀ1.00
0.90
ꢀ0.69
ꢀ0.75
0.57
ACHTUNGTRENNUNG
ꢀ212
ꢀ575
In some cases, we observed broadening of the
overlapping 1:1:1:1 multiplets with increasing tem-
perature, which we attribute to ligand exchange at
higher temperatures. Likewise, Cheng et al. found
that it was necessary to cool samples below room
344
268
130
119
ꢀ189
ꢀ525
313
409
140
31
253
101
78
ꢀ190
ꢀ515
temperature to decrease the rate of ligand exchange Cl3Ga(PMe3)
Br3Ga(PMe3)
354
469
144
37
288
ꢀ0.83
sufficiently to observe 1J(69/71Ga,31P) for X3Ga
ACTHNUTRGEN(UNG PPh3)
135
ꢀ178
122
ꢀ555
0.44
I3Ga(PMe3)
591
0.82
(X=Cl and Br) dissolved in CH2Cl2.[8] Clearly, an
interplay of the magnitude of the 69/71Ga nuclear
quadrupolar coupling constants, rotational correla-
tion times, rate of ligand exchange, and the magni-
Gallium quadrupolar coupling
CQ(69Ga)
CQ(71Ga)
hQ
a
b
g
[MHz]
[MHz]
[8]
[8]
[8]
Cl3Ga(PPh3)
Br3Ga(PPh3)
I3Ga(PPh3)
Cl3Ga[P(p-Anis)3]
Br3Ga[P(p-Anis)3]
I3Ga[P(p-Anis)3]
Cl3Ga(TMP)
Br3Ga(TMP)
I3Ga(TMP)
+8.3
+4.6
+5.6
+5.9
+11.0
ꢀ3.3
+5.2
+2.8
+3.5
+3.7
+6.9
ꢀ2.1
ꢀ4.3
ꢀ7.1
ꢀ2.0
+4.1
+8.0
ꢀ2.6
0.17
0.39
0.00
0.47
0.00
0.52
0.36
0.04
0.32
0.11
0.01
0.17
73
80
0
88
84
0
17
50
0
10
0
47
11
0
30
6
1
tude of J(69/71Ga, 31P) will dictate exactly what line
shape one will observe in solution as a function of
temperature. To obtain accurate values of 1J(69/
71Ga,31P) from 31P NMR spectra acquired in solu-
tion, it would be necessary to carry out detailed
line-shape analyses as outlined in the review by
Mlynꢃrik,[34] but we were able to obtain reasonable
83
90
85
79
27
72
84
17
86
88
90
37
77
87
46
73
89
30
ꢀ6.8
ꢀ11.4
ꢀ3.2
Cl3Ga(PMe3)
Br3Ga(PMe3)
I3Ga(PMe3)
+6.6
+12.7
ꢀ4.2
estimates of these values for the X3GaACTHNUTRGNEUN(G PR3) ad-
18
30
ducts (see Table 3). As shown in Tables 2 and 3, the
same trend is observed for solid samples and for
samples in solution. 1J(69/71Ga,31P) values decrease
slightly as X changes from Cl to Br and more dra-
matically as X changes from Cl or Br to I.
[a] Chemical shifts converted from shielding values according to diso(Ga)=
1814ꢀsiso(Ga) (see text).
ison of experimental and calculated
CQAHCTUNTGERNNUNG(
69/71Ga) values
(Tables 1 and 4) shows that the DFT calculations do not re-
produce experimental trends, but they do correctly predict
that these values are small (e.g., for the diatomic gallium(I)
Table 3. Solution-phase 31P NMR parameters for X3Ga(PR3) (X=Cl, Br,
and I) adducts.
diso(31P)
1J(71Ga,31P)
T
halides,
CQACHTUNTGRENNUNG
(69Ga) ranges from ꢀ106.6 for GaF to
[ppm]
[Hz]
[K]
ꢀ81.1 MHz for GaI, an order of magnitude larger than the
Cl3Ga(PPh3)
Br3Ga(PPh3)
I3Ga(PPh3)
ꢀ5.5ꢄ0.2
ꢀ5.6ꢄ0.2
721ꢄ10
793ꢄ10
693ꢄ10
696ꢄ10
466ꢄ10
470ꢄ10
970ꢄ10
810ꢄ10
183[a]
193
values for the gallium
ACHTUNGTRENNUNG
ꢀ10.7ꢄ0.2
ꢀ10.3ꢄ0.2
ꢀ29.7ꢄ0.2
ꢀ28.7ꢄ0.2
ꢀ18.8ꢄ0.2
ꢀ24.8ꢄ0.2
ꢀ31.4ꢄ0.2
ꢀ41.7ꢄ0.2
ꢀ45.8ꢄ0.2
ꢀ52.8ꢄ0.2
273[a]
193
addition, the calculated sign for C CTHNUGTRENNUNG(
for the seven adducts for which it is known from experi-
ments. Furthermore, the calculated relative orientations of
the gallium EFG and shielding tensors, given by the Euler
angles a, b, and g, are in qualitative agreement with experi-
mental values. For example, consider the two complexes
ꢀ
273[a]
253
RT
323
333
193
233
193
Cl3Ga[P(p-Anis)3]
Br3Ga[P(p-Anis)3]
I3Ga[P(p-Anis)3]
Cl3Ga(TMP)
Br3Ga(TMP)
I3Ga(TMP)
535ꢄ10
[b]
that have an exact C3 axis along the Ga P bond, Br3Ga[PACHTUNGTRENNUNG(p-
[b]
[b]
Anis)3] and I3GaACTHNUTRGNE(UNG PPh3). Despite the small gallium and phos-
phorus spans, calculations correctly predict that the unique
components of the EFG and CS tensors are VZZ and d11 for
the former and VZZ and d33 for the latter.
[a] Reference [8]. [b] Only one broad peak was observed and the
1J(71Ga,31P) values could not be resolved.
Converting calculated shielding into chemical shifts re-
quires the establishment of an absolute shielding scale for
the nucleus of interest.[37] Since no such scale has been es-
tablished for gallium, the gallium shielding constant for [Ga-
These results demonstrate that, because of the competing
mechanisms occurring for samples in solution, it generally is
easier to characterize indirect spin–spin coupling constants
involving quadrupolar nuclei in the solid state.
AHCTUNGTRENNUNG
(H2O)6]3+ was calculated to approximate that of the accept-
ed gallium chemical shift reference, GaACTHNUTRGNEUNG
(NO3)3 (1.0m).[20]
DFT calculations—Comparison with experiment: Calculated
This value, s=1814 ppm, was set to diso(Ga)=0 ppm and
Ga NMR parameters are summarized in Table 4. A compar-
thus the calculated chemical shift values in Table 4 were ob-
2834
www.chemeurj.org
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 2826 – 2838

69/71Ga and 31P NMR Spectroscopy of X3Ga
ACTHNUTRGENUN(G PR3)
FULL PAPER
tained from calculated magnetic shielding values by using
the relationship d(Ga)calcd =1814ꢀs(Ga)calcd
.
Plots of experimental versus calculated diso(Ga), k, and W
values are shown in Figure 12; these indicate that the calcu-
lated chemical shift parameters are qualitatively in agree-
ment with the corresponding experimental values. For the
three series of complexes studied, X3Ga
Anis)3], and X3Ga(TMP), diso is greatest for X=Cl and least
for X=I, in agreement with experiments; this is illustrated
in Figure 13. Likewise, calculated and experimental
ACHUTNGTRNENUG(PPh3), X3Ga[PACHTUGNTREN(NUGN p-
ACHTUNGTRENNUNG
W
values are in agreement (compare Figure 13 with Figure 8).
To investigate the origin of the increased gallium shielding
as the atomic number of the halogen atom increases, the so-
called normal halogen effect (NHE), the diamagnetic (sdia),
paramagnetic (spara), and spin-orbital (sspin–orbit) contributions
to the gallium isotropic magnetic shielding constants were
calculated (see Table S4 in the Supporting Information).
The value of sdia is virtually independent of the halogen, and
spara tends to decrease slightly as X changes from Cl to I;
however, it is clearly sspin–orbit that is responsible for the ob-
served NHE, with contributions from sspin–orbit to the isotrop-
ic shielding of gallium of approximately 160, 370, and
745 ppm for X=Cl, Br, and I, respectively. Significantly, al-
though both spara and sspin–orbit are significantly anisotropic,
there is little variation in the anisotropy of the former, re-
gardless of the halogen atom. In contrast, the anisotropy in
sspin–orbit varies from less than 30 ppm for the trichloro ad-
ducts to more than 700 ppm for the triiodo adducts
(Table S4 in the Supporting Information). Hence, sspin–orbit is
also responsible for the large spans in the gallium chemical
shift tensors for the triiodo adducts in comparison with
those for the tribromides and trichlorides.
Figure 12. Experimental versus calculated gallium CS tensor values
a) diso, b) k, and c) W for the gallium trihalide triarylphosphine adducts
X3Ga(PR3). The solid line indicates perfect agreement between experi-
mental and calculated values. The standard deviations (s=
[S(calcdꢀexptl)2/(nꢀ1)]1/2 in which n is the number of data points) are
64 ppm, 0.14, and 111 ppm for diso, k, and W, respectively.
In contrast to the GaIII adducts, GaI monohalides exhibit
an inverse halogen dependence (IHD).[38] Likewise, we re-
cently reported an NHE for a
series of InIII trihalides in con-
trast to the IHD found for
some InI monohalides.[18] In
both the gallium and indium
series of compounds, a large
paramagnetic term (largest for
X=I and smallest for X=Cl) is
mainly responsible for the IHD
of the isotropic chemical shifts
for the monohalides (see
Table S5 in the Supporting In-
formation).
Calculated 1J(69/71Ga,31P) and
DJ(69/71Ga,31P) values for the
X3GaACHTUNGTRENNUNG(PR3) adducts are listed in
Table 5. The former are gener-
ally smaller than the observed
values; nevertheless the experi-
1
mental trend that J(69/71Ga,31P)
is greatest for X=Cl and small-
est for X=I is reproduced by
Figure 13. Calculated spans of the CS tensors for X3Ga(TMP), X3Ga[P(p-Anis)3], X3Ga(PPh3), and
the calculations. Calculated X3Ga(PMe3).
Chem. Eur. J. 2013, 19, 2826 – 2838
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2835

R. E. Wasylishen et al.
Table 5. Calculated 1J(69/71Ga,31P) and DJ(69/71Ga,31P) for X3Ga(PR3) and
model X3Ga(PMe3) (X=Cl, Br, and I) adducts.
21.14 T, the contribution from this interaction to the NMR
line shape of the central transition is of the same magnitude
1J(69Ga,31P) 1J(71Ga,31P) DJ(69Ga,31P) DJ(71Ga,31P)
or greater than that from CQACHUTNGRNENUG(
69/71Ga). Values for 1J(69/
71Ga,31P) were determined for all adducts investigated
herein. In addition, DJ(69/71Ga,31P) values and their signs are
reported; these were derived from known or estimated
values for RDD. As for the 115In–31P J tensors,[18] the 1J(69/
71Ga,31P) and DJ(69/71Ga,31P) values are comparable in mag-
nitude and positive.
[Hz]
[Hz]
[Hz]
[Hz]
Cl3Ga(PPh3)
Br3Ga(PPh3)
I3Ga(PPh3)
Cl3Ga[P(p-Anis)3] +404
Br3Ga[P(p-Anis)3] +218
I3Ga[P(p-Anis)3]
Cl3Ga(TMP)
Br3Ga(TMP)
Cl3Ga(PMe3)
Br3Ga(PMe3)
I3Ga(PMe3)
+344
+211
+6
+437
+268
+8
+511
+484
+391
+536
+480
+448
+555
+471
+549
+469
+466
+650
+615
+496
+682
+609
+569
+705
+598
+697
+596
+592
+513
+277
+114
+596
+322
+492
+229
+66
+89
+469
+254
+388
+181
+52
Solution-phase 31P NMR spectroscopic results indicate
that the gallium T1 relaxation times for I3Ga
Br3Ga[P(p-Anis)3], and I3Ga[P(p-Anis)3] are very short at
lower temperatures, whereas those for Cl3Ga(PPh3), Br3Ga-
(PPh3), and Cl3Ga[P(p-Anis)3] do not change significantly in
ACHTUNGTRENNUNG(PPh3),
A
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
A
ACHTUNGTRENNUNG
comparison to changes in the Ga–phosphine exchange rates.
The differences between the solution-phase 31P NMR spec-
tra for these adducts at different temperatures are attributed
to a competition between variations in the 69/71Ga T1 values
values of DJACHTUNGTRENNUNG
(71Ga,31P) are positive and range from approxi-
mately 500 to 700 Hz, whereas the experimental values
range for (420ꢄ150) to (1290ꢄ150) Hz. Although the
errors in the experimental values are large, it is clear that
the anisotropies in the indirect Ga–P spin–spin coupling ten-
sors are on the same order of magnitude as the isotropic
values. The calculated contribution of the Fermi-contact
and in the Ga-phosphine exchange rates. The 31P NMR pa-
1
rameters, J
(71Ga,31P) and diso
N
generally in agreement with those obtained from solid-state
NMR spectroscopy.
1
mechanism to J(69/71Ga,31P)iso (not shown) for the model ad-
The experimental results have been complemented by rel-
ativistic DFT calculations. These calculations qualitatively
reproduce the available experimental 69/71Ga EFG and CS
tensors, including their relative orientations, and they also
reproduce the normal halogen effect observed between the
isotropic gallium magnetic shielding and the halogen li-
gands; calculations confirm that this effect is due to the
spin-orbital effect of the halogen ligand. Although calculat-
ducts is approximately 99%. The calculations indicate that
the nonzero DJ(69/71Ga,31P) values result from the spin-dipo-
lar Fermi-contact (SDꢄFC) cross term. Thus, both the
Fermi-contact and spin-dipolar Fermi-contact mechanisms
make important contributions to the J(69/71Ga,31P) tensors in
these adducts. A similar effect was noted for the J(113/
115In,31P) tensors of indium
ducts.[18]
ACHTUNGTNER(NUNG III) trihalide phosphine ad-
1
ed values of J(69/71Ga,31P) and DJ(69/71Ga,31P) are generally
lower than the experimental values, calculations undertaken
1
for the X3Ga
N
Conclusion
DJ(69/71Ga,31P) are comparable in magnitude with positive
signs. Calculations indicate that both the Fermi-contact and
spin-dipolar Fermi-contact mechanisms are important fac-
tors when considering the J tensor between coupled gallium
and phosphorus nuclei.
The results of an investigation of several triarylphosphine
gallium trihalide adducts, X3GaACTHNUTRGNE(UNG PR3) (X=Cl, Br, or I;
PR3 =triarylphosphine ligand) by using solid-state 69/71Ga
and 31P NMR spectroscopy and single-crystal X-ray diffrac-
We have provided herein an interpretation of solid-state
69/71Ga NMR spectra for some gallium–phosphine complexes.
The conclusions reached in this study are comparable to
those reached in an earlier study of indium–phosphine com-
plexes:[18] one may not ignore the effects of gallium magnet-
ic shielding anisotropy, particularly for spectra acquired at
high magnetic-field strengths; relativistic DFT calculations
are helpful in the analysis of 69/71Ga NMR spectra; and
tion has been presented. The experimental C
for these adducts range from (+3.3ꢄ0.2) to (ꢀ11.0ꢄ
0.4) MHz; the signs of CQA
69/71Ga) for some adducts with an
(71Ga) values
QACHTUNGTRENNUNG
CTHUNGTRENNUNG(
exact or approximate C3 axis were determined from analyses
of the 31P NMR spectra. The 69/71Ga nuclear quadrupolar
coupling constants reported herein are small compared with
others that we have previously measured in our lab. The rel-
atively small EFGs at the gallium nuclei in these compounds
are in agreement with the predictions of Bancroft and co-
workers.[39]
NMR investigations of spin-1= nuclei, such as 31P, coupled to
2
gallium complement the 69/71Ga NMR studies.
Trends observed in this study are similar to those ob-
served for the related In–phosphine adducts.[18] In particular,
the gallium nuclei are most shielded for X=I and least
shielded for X=Cl, that is, the normal halogen effect. The
anisotropic gallium magnetic shielding is measurable for all
adducts, with spans as large as (380ꢄ15) ppm. Similar to ob-
servations for the In–phosphine adducts, the spans are great-
est for the adducts with X=I; for 69/71Ga spectra acquired at
Experimental and Computational Details
Sample preparation: Gallium
(GaBr3), gallium(III) iodide (GaI3), triphenylphosphine (PPh3), tris-para-
methoxyphenylphosphine (P(p-Anis)3), and tris(2,4,6-trimethoxyphenyl)-
ACHUTGTNREN(NUG III) chloride (GaCl3), galliumACHTUNGTNER(NUGN III) bromide
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
phosphine (TMP) were purchased from either Aldrich or Strem and used
as received. Owing to the ease of hydrolysis of the anhydrous halides and
oxidation of the phosphine ligands, all operations were carried out in a
2836
www.chemeurj.org
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 2826 – 2838

69/71Ga and 31P NMR Spectroscopy of X3Ga
ACTHNUTRGENUN(G PR3)
FULL PAPER
dry glovebox filled with argon. See the Supporting Information for more
details of these syntheses and of the syntheses of the X3Ga(TMP) com-
pounds, X=Cl, Br, and I, which have not previously been reported.
Single-crystal X-ray diffraction: Single crystals of Br3Ga[P(p-Anis)3] and
I3Ga[P(p-Anis)3] suitable for X-ray diffraction were grown by very slow
of the halides on the gallium magnetic shielding, CS tensors for model
X3Ga(PMe3) (X=Cl, Br, or I) adducts were also calculated. See the Sup-
porting Information for more details.
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
ACHTUNGTRENNUNG
evaporation of solutions of these complexes in a 1:1 mixture of ethyl ace-
tate and dichloromethane. Suitable crystals were mounted on glass fibers
by means of paratone-N oil, and data were collected at 193 K by using
graphite-monochromated MoKa radiation (0.71073 ꢂ) on a Bruker PLAT-
FORM/SMART 1000 CCD diffractometer. The structures were solved
by direct methods using SHELXS-97[40] and refined using full-matrix
least-squares on F2 (SHELXL-97).[40] All nonhydrogen atoms in the com-
pounds were refined with anisotropic displacement parameters. Selected
Acknowledgements
We thank Victor Terskikh and Shane Pawsey for helpful comments and
for acquiring some NMR spectra at 21.14 T. We are very grateful to
Jason Clyburne, Gang Wu, Neil Burford, Klaus Eichele, Chris Kirby,
Mike Lumsden, and Kenneth Wright for early contributions to this proj-
ect. Access to the 900 MHz NMR spectrometer was provided by the Na-
tional Ultrahigh-Field NMR Facility for Solids (Ottawa, Canada), funded
by the Canada Foundation for Innovation, the Ontario Innovation Trust,
Recherche Quꢀbec, the National Research Council of Canada, and
Bruker BioSpin and managed by the University of Ottawa
(www.nmr900.ca). R.G.C. and R.E.W. thank the Natural Sciences and
Engineering Research Council of Canada (NSERC) for financial support
through the Discovery Grant Program and for a Major Resources Sup-
port grant. R.E.W. also thanks the Canada Research Chairs program for
research support.
crystal data and structure refinement details for Br3Ga[P
I3Ga[P(p-Anis)3] are listed in Tables S1 and S2 in the Supporting Infor-
mation. CCDC-896761 (Br3Ga[P(p-Anis)3]) and CCDC-896762 (I3Ga[P-
(p-Anis)3]) contain the supplementary crystallographic data for this
ACHTUNGTRNEUN(NG p-Anis)3] and
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
ACHTUNGTRENNUNG
paper. These data can be obtained free of charge from The Cambridge
Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
NMR spectroscopy: Solution-phase 31P{1H} NMR spectra were acquired
at 161.902 MHz on a three-channel Varian Inova 400 MHz (1H) spec-
trometer by using a pulsed field gradient direct detection broadband
switchable probe. Reported temperatures are based on a calibration
curve for that probe. The low-temperature portion of the calibration
curve was constructed by using a standard methanol sample from Varian.
[1] J. L. Gay-Lussac, J. L. Thꢀnard, Mem. Phys. Chim. Soc. dꢀArcueil
1809, 2, 210, as cited in: V. Jonas, G. Frenking, J. Chem. Soc. Chem.
Commun. 1994, 1489–1490.
The solvent used for the X3Ga
N
ACHTUNGTREN(NGNU TMP) adducts was
CD2Cl2 whereas CDCl3 was used for the X3Ga[PACHTUNGTRENNNUG
Solid-state 31P NMR spectra of MAS and stationary samples were ac-
quired on a Chemagnetics CMX Infinity 200 (B0 =4.70 T), as well as
Bruker Avance 300 and 500 NMR spectrometers by using the combina-
tion of standard cross-polarization (CP) with proton TPPM decoupling.[41]
Samples were packed in 4 mm o.d. rotors. Proton 908 pulse widths of
4.0 ms, contact times of 5 to 10 ms, and pulse delays of 4 to 60 s were used
to acquire most 31P NMR spectra. 31P chemical shifts were referenced
with respect to 85% aqueous phosphoric acid by setting the isotropic
peak of an external solid ammonium dihydrogen phosphate sample to
0.81 ppm.[42] Spectra of MAS samples were acquired at ambient tempera-
ture with a spinning frequency of 8.0 to 15.0 kHz.
69/71Ga NMR spectra of MAS and stationary samples were acquired on
Bruker Avance 300 (B0 =7.05 T), 500 (B0 =11.75 T) and Avance II
900 MHz (B0 =21.14 T) NMR spectrometers by using Bruker 4 mm
MAS probes. Spectra of MAS samples were acquired at ambient temper-
ature with a spinning frequency of 12.5 to 20.0 kHz. A p/2–t1–p/2–t2–
ACQ echo sequence was used to acquire all Ga NMR spectra, with pulse
[2] See, for example: a) S. J. Geier, T. M. Gilbert, D. W. Stephan, Inorg.
Chem. 2011, 50, 336–344; b) E. I. Davydova, T. N. Sevastianova,
A. V. Suvorov, A. Y. Timoshkin, Coord. Chem. Rev. 2010, 254,
2031–2077; c) F. Dornhaus, S. Scholz, I. Sꢅnger, M. Bolte, M.
Wagner, H.-W. Lerner, Z. Anorg. Allg. Chem. 2009, 635, 2263–2272;
d) T. J. Clark, C. A. Jaska, A. Turak, A. J. Lough, Z.-H. Lu, I. Man-
ners, Inorg. Chem. 2007, 46, 7394–7402.
[3] a) A. Staubitz, A. P. M. Robertson, M. E. Sloan, I. Manners, Chem.
Rev. 2010, 110, 4023–4078; b) C. W. Hamilton, R. T. Baker, A. Stau-
bitz, I. Manners, Chem. Soc. Rev. 2009, 38, 279–293; c) J. M. Brunel,
B. Faure, M. Maffei, Coord. Chem. Rev. 1998, 178–180, 665–698.
[4] See, for example: a) A. Kuczkowski, S. Schulz, M. Nieger, Eur. J.
Inorg. Chem. 2001, 2605–2611; b) A. Kuczkowski, S. Schulz, M.
Nieger, P. R. Schreiner, Organometallics 2002, 21, 1408–1419;
c) R. L. Wells, R. A. Baldwin, P. S. White, W. T. Pennington, A. L.
Rheingold, G. P. A. Yap, Organometallics 1996, 15, 91–97; d) R. L.
Wells, S. R. Aubuchon, S. S. Kher, M. S. Lube, P. S. White, Chem.
Mater. 1995, 7, 793–800.
[5] a) P. OꢆBrien, N. L. Pickett in Comprehensive Coordination Chemis-
try II, Vol. 9 (Eds.: J. A. McCleverty, T. J. Meyer), Elsevier, Oxford,
2003, pp. 1005–1063; b) I. R. Grant in Chemistry of Aluminium,
Gallium, Indium and Thallium (Ed.: A. J. Downs), Blackie, Glas-
gow, 1993, Chapter 5, pp. 292–321.
[6] a) A. J. Carty, Can. J. Chem. 1967, 45, 345–351; b) A. J. Carty, Can.
J. Chem. 1967, 45, 3187–3192; c) A. J. Carty, T. Hinsperger, P. M.
Boorman, Can. J. Chem. 1970, 48, 1959–1970.
[7] For recent reviews, see a) S. Arai, N. Nishiyama, T. Maruyama, T.
Okumura, IEEE J. Sel. Top. Quantum Electron. 2011, 17, 1381–
1389; b) S. L. Pyshkin, J. M. Ballato in Advances and Applications in
Electroceramics, Vol. 226 (Eds.: K. M. Nair, Q. Jia, S. Priya), Wiley,
Hoboken, 2011, Chapter 18, pp. 77–90; c) D. J. Friedman, J. M.
Olson, S. Kurtz in Handbook of Photovoltaic Science and Engineer-
ing, 2nd ed. (Eds.: A. Luque, S. Hegedus), Wiley, Chichester, 2011,
pp. 314–337.
[8] F. Cheng, H. L. Codgbrook, A. L. Hector, W. Levason, G. Reid, M.
Webster, W. Zhang, Polyhedron 2007, 26, 4147–4155.
[9] M. A. Brown, J. A. Castro, D. G. Tuck, Can. J. Chem. 1997, 75, 333–
341.
lengths (tp(sel)) that selectively excited the central transition, t(sel) =t(non-sel)
/
(S+1= )=t(non-sel)/2 for 69/71Ga.[43] Proton TPPM decoupling[41] was used to
2
acquire all spectra. Each step of the Ga NMR spectra acquired at 7.05 or
11.75 T is the sum of 4096 to 72000 scans. Ga NMR chemical shifts were
referenced with respect to an external solution of Ga(NO3)3 (1.0m).[20]
The relaxation delay was 0.5 s.
311P and 69/71Ga NMR parameters were determined by visual comparison
of experimental NMR spectra with those simulated by using the WSolids
software package.[44] This software includes the quadrupolar interaction
to second-order perturbation theory for simulations of 69/71Ga NMR spec-
tra, and includes spin–spin interactions with these nuclei for simulations
of 31P NMR spectra of MAS samples.
Quantum chemical calculations: DFT calculations of gallium EFG[45] and
CS[46] tensors, as well as 1J(69/71Ga,31P) and DJ(69/71Ga,31P) values,[47] were
performed by using the Amsterdam Density Functional (ADF) pro-
gram.[48] Geometries used for calculations for the Cl3Ga(PPh3),[8]
Br3Ga(PPh3),[8] I3Ga(PPh3),[9] Br3Ga[P(p-Anis)3], and I3Ga[P(p-Anis)3]
ꢀ
adducts were those obtained from X-ray diffraction, except for the C H
bond lengths, which were fixed at 1.08 ꢂ. Because single-crystal structure
data for Cl3Ga[P(p-Anis)3], Cl3Ga(TMP), Br3Ga(TMP), and I3Ga(TMP)
were unavailable, the geometries for Cl3Ga(PPh3), Br3Ga[P(p-Anis)3],
and I3Ga[P(p-Anis)3] were used, with the H atoms replaced by a methoxy
group where needed (e.g., the para-H atom of PPh3 was replaced with a
methoxy group to model the P(p-Anis) ligand). To investigate the effects
[10] a) F. Cheng, A. L. Hector, W. Levason, G. Reid, M. Webster, W.
Zhang, Inorg. Chem. 2007, 46, 7215–7223; b) J. F. Janik, R. A. Bald-
win, R. L. Wells, W. T. Pennington, G. L. Schimek, A. L. Rheingold,
L. M. Liable-Sands, Organometallics 1996, 15, 5385–5390; c) L.-J.
Chem. Eur. J. 2013, 19, 2826 – 2838
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2837

R. E. Wasylishen et al.
Baker, L. A. Kloo, C. E. F. Rickard, M. J. Taylor, J. Organomet.
Chem. 1997, 545–546, 249–255; d) L.-J. Baker, C. E. F. Rickard,
M. J. Taylor, J. Organomet. Chem. 1994, 464, C4–C6; e) M. Sigl, A.
Schier, H. Schmidbaur, Eur. J. Inorg. Chem. 1998, 203–210; f) J. C.
Carter, G. Jugie, R. Enjalbert, J. Galy, Inorg. Chem. 1978, 17, 1248–
1254.
[24] a) R. E. Wasylishen in Encyclopedia of Nuclear Magnetic Resonance,
Vol. 3 (Eds.: D. M. Grant, R. K. Harris), Wiley, Chichester, 1996,
pp. 1685–1695; b) R. E. Wasylishen in Encyclopedia of Nuclear
Magnetic Resonance, Vol. 9 (Eds.: D. M. Grant, R. K. Harris), Wiley,
Chichester, 2002, pp. 274–282, and references therein.
[25] a) R. K. Harris, A. C. Olivieri, Prog. Nucl. Magn. Reson. Spectrosc.
1992, 24, 435–456, and references cited therein; b) R. K. Harris,
A. C. Olivieri in Encyclopedia of Nuclear Magnetic Resonance,
Vol. 9 (Eds.: D. M. Grant, R. K. Harris), Wiley, Chichester, 2002,
pp. 141–150; c) P. Grondona, A. C. Olivieri, Concepts Magn. Reson.
1993, 5, 319–339.
[26] D. L. VanderHart, H. S. Gutowsky, T. C. Farrar, J. Am. Chem. Soc.
1967, 89, 5056–5057.
[27] a) M. Sugimoto, M. Kanayama, H. Nakatsuji, J. Phys. Chem. 1993,
97, 5868–5874; b) H. Takashima, M. Hada, H. Nakatsuji, Chem.
Phys. Lett. 1995, 235, 13–16.
[28] J. Manz, J. Am. Chem. Soc. 1980, 102, 1801–1806.
[29] a) H. Nçth, B. Wrackmeyer in NMR Basic Principles and Progress,
Vol 14, Nuclear Magnetic Resonance Spectroscopy of Boron Com-
pounds (Eds.: P. Diehl, E. Fluck, R. Kosfeld), 1978, Springer, Berlin,
pp. 104–105; b) H. C. E. McFarlane, W. McFarlane, D. S. Rycroft, J.
Chem. Soc. Faraday Trans. 2 1972, 68, 1300–1305.
[30] a) R. W. Rudolph, C. W. Schultz, J. Am. Chem. Soc. 1971, 93, 6821–
6822; b) A. H. Cowley, M. C. Damasco, J. Am. Chem. Soc. 1971, 93,
6815–6821.
[31] a) C. J. Jameson in Multinuclear NMR (Ed.: J. Mason), Plenum
Press, New York, 1987, Chapter 4, pp. 89–131; b) C. J. Jameson in
Phosphorus-31 NMR Spectroscopy in Stereochemical Analysis, Or-
ganic Compounds and Metal Complexes (Eds.: J. G. Verkade, L. D.
Quin) VCH, Deerfield Beach, 1987, Chapter 6, pp. 205–230.
[32] R. E. Wasylishen, K. C. Wright, K. Eichele, T. S. Cameron, Inorg.
Chem. 1994, 33, 407–408.
[11] a) T. Vosegaard, I. P. Byriel, L. Binet, D. Massiot, H. J. Jakobsen, J.
Am. Chem. Soc. 1998, 120, 8184–8188; b) T. Vosegaard, D. Massiot,
Chem. Phys. Lett. 2007, 437, 120–125; c) L. A. OꢆDell, S. L. P. Savin,
A. V. Chadwick, M. E. Smith, Appl. Magn. Reson. 2007, 32, 527–
546; d) J. T. Ash, P. J. Grandinetti, Magn. Reson. Chem. 2006, 44,
823–831; e) D. Massiot, I. Farnan, N. Gautier, D. Trumeau, A.
Trokiner, J. P. Coutures, Solid State Nucl. Magn. Reson. 1995, 4,
241–248.
[12] a) J. P. Yesinowski, Phys. Status Solidi C 2005, 2, 2399–2402; b) W.-
S. Jung, O. H. Han, S.-A. Chae, Mater. Chem. Phys. 2006, 100, 199–
202; c) A. P. Purdy, J. P. Yesinowski, A. T. Hanbicki, Phys. Status
Solidi C 2005, 2, 2437–2440; d) J. P. Yesinowski, J. Magn. Reson.
2006, 180, 147–161; e) J. P. Yesinowski, A. P. Purdy, J. Am. Chem.
Soc. 2004, 126, 9166–9167.
[13] a) D. Franke, H. Eckert, R. B. Kaner, R. E. Treece, Anal. Chim.
Acta 1993, 283, 987–995; b) D. P. Tunstall, A. Rogerson, J. Phys. C:
Solid State Phys. 1979, 12, 3105–3120.
[14] a) J. M. Kikkawa, D. D. Awschalom, Science 2000, 287, 473–476;
b) T. Ota, N. Kumada, G. Yusa, S. Miyashita, Y. Hirayama, Phys.
Status Solidi C 2007, 4, 1759–1762; c) J. Takeuchi, H. Nakamura, H.
Yamada, E. Kita, A. Tasaki, T. Erata, Solid State Nucl. Magn.
Reson. 1997, 8, 123–128; d) L. D. Potter, Y. Wu, J. Magn. Reson.
Ser. A 1995, 116, 107–112; e) O. H. Han, H. K. C. Timken, E. Old-
field, J. Chem. Phys. 1988, 89, 6046–6052.
[15] For recent examples, see a) A. T. Durant, K. J. D. MacKenzie, H.
Maekawa, Dalton Trans. 2011, 40, 4865–4870; b) F. Blanc, D. S.
Middlemiss, Z. Gan, C. P. Grey, J. Am. Chem. Soc. 2011, 133,
17662–17672; c) H. B. Yahia, L. van Wꢇllen, S. Balamurugan,
U. Ch. Rodewald, H. Eckert, R. Pçttgen, Z. Naturforsch. B 2011, 66,
14–20.
[16] J. F. Hinton, R. W. Briggs in NMR and the Periodic Table (Eds.:
R. K. Harris, B. E. Mann), Academic Press, London, 1978, Chap-
ter 9, p. 285–286.
[33] F. Chen, S.-W. Oh, R. E. Wasylishen, Can. J. Chem. 2009, 87, 1090–
1101.
[34] V. Mlynꢃrik, Prog. Nucl. Magn. Reson. Spectrosc. 1986, 18, 277–305.
[35] J. Bacon, R. J. Gillespie, J. W. Quail, Can. J. Chem. 1963, 41, 3063–
3069.
[36] V. B. Singh, J. Phys. Chem. Ref. Data 2005, 34, 23–37.
[37] C. J. Jameson in Encyclopedia of Nuclear Magnetic Resonance, Vol 2
(Eds.: D. M. Grant, R. K. Harris), Wiley, Chichester, 1996, pp. 1273–
1281.
[17] T. Vosegaard, D. Massiot, N. Gautier, H. J. Jakobsen, Inorg. Chem.
1997, 36, 2446–2450.
[18] F. Chen, G. Ma, G. M. Bernard, R. G. Cavell, R. McDonald, M. J.
Ferguson, R. E. Wasylishen, J. Am. Chem. Soc. 2010, 132, 5479–
5493.
[19] F. Chen, G. Ma, R. G. Cavell, V. V. Terskikh, R. E. Wasylishen,
Chem. Commun. 2008, 5933–5935.
[38] M. Gee, R. E. Wasylishen in ACS Symposium Series, No. 732, Mod-
eling NMR Chemical Shifts, Gaining Insights into Structure and En-
vironment (Eds.: J. C. Facelli, A. C. de Dios), American Chemical
Society, Washington, 1999, Chapter 19, pp. 259–276.
[39] a) G. M. Bancroft, R. H. Piatt, Adv. Inorg. Chem. Radiochem. 1972,
15, 59–258; b) G. M. Bancroft, T. K. Sham, J. Magn. Reson. 1977,
25, 83–90.
[20] R. K. Harris, E. D. Becker, S. M. Cabral de Menezes, R. Goodfellow,
P. Granger, Pure Appl. Chem. 2001, 73, 1795–1818.
[21] J. Mason, Solid State Nucl. Magn. Reson. 1993, 2, 285–288.
[22] For several current articles, see a) P. P. Man in NMR of Quadrupolar
Nuclei in Solid Materials (Eds.: R. E. Wasylishen, S. E. Ashbrook, S.
Wimperis), Wiley, Chichester, 2012, Chapter 1, pp. 3–16; b) A. J.
Vega in NMR of Quadrupolar Nuclei in Solid Materials (Eds.: R. E.
Wasylishen, S. E. Ashbrook, S. Wimperis), Wiley, Chichester, 2012,
Chapter 2, pp. 17–44; c) S. E. Ashbrook, S. Wimperis in NMR of
Quadrupolar Nuclei in Solid Materials (Eds.: R. E. Wasylishen, S. E.
Ashbrook, S. Wimperis), Wiley, Chichester, 2012, Chapter 3, pp. 45–
61; d) D. L. Bryce, R. E. Wasylishen in NMR of Quadrupolar Nuclei
in Solid Materials (Eds.: R. E. Wasylishen, S. E. Ashbrook, S. Wim-
peris), Wiley, Chichester, 2012, Chapter 4, pp. 63–74; e) D. C. Ap-
perley, R. K. Harris, P. Hodgkinson, Solid State NMR: Basic Princi-
ples and Practice, Momentum Press, New York, 2012, Chapter 6,
pp. 141–175; see also, f) J. P. Amoureux, C. Fernandez, P. Granger
in Multinuclear Magnetic Resonance in Liquids and Solids—Chemi-
cal Applications (Eds.: P. Granger, R. K. Harris), Kluwer Academic,
Dordrecht, 1990, pp. 409–424; g) A. Abragam, Principles of Nuclear
Magnetism, Clarendon Press, Oxford, 1961, Chapter VII, pp. 216–
263.
[40] G. M. Sheldrick, Acta Crystallogr. Sect. A 2008, 64, 112–122.
[41] A. E. Bennett, C. M. Rienstra, M. Auger, K. V. Lakshmi, R. G. Grif-
fin, J. Chem. Phys. 1995, 103, 6951–6958.
[42] K. Eichele, R. E. Wasylishen, J. Phys. Chem. 1994, 98, 3108–3113.
[43] P. R. Bodart, J.-P. Amoureux, Y. Dumazy, R. Lefort, Mol. Phys.
2000, 98, 1545–1551.
[44] K. Eichele, R. E. Wasylishen, WSOLIDS, Version 2.0.18, University
of Alberta, Edmonton, 2000.
[45] E. van Lenthe, E. J. Baerends, J. Chem. Phys. 2000, 112, 8279–8292.
[46] G. Schreckenbach, T. Ziegler, J. Phys. Chem. 1995, 99, 606–611.
[47] a) J. Autschbach, T. Ziegler, J. Chem. Phys. 2000, 113, 9410–9418;
b) J. Autschbach, T. Ziegler, J. Chem. Phys. 2000, 113, 936–947.
[48] a) ADF, 2006.01, Theoretical Chemistry, Vrije Universiteit, Amster-
dam, 2006, http://www.scm.com; b) G. te Velde, F. M. Bickelhaupt,
E. J. Baerends, C. Fonseca Guerra, S. J. A. van Gisbergen, J. G.
Snijders, T. Ziegler, J. Comput. Chem. 2001, 22, 931–967; c) C. Fon-
seca Guerra, J. G. Snijders, G. te Velde, E. J. Baerends, Theor. Chem.
Acc. 1998, 99, 391–403.
Received: August 19, 2012
[23] A. D. Buckingham, I. Love, J. Magn. Reson. 1970, 2, 338–351.
Published online: January 10, 2013
2838
www.chemeurj.org
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 2826 – 2838