Reactions of a gallium(ll)-diazabutadiene dimer, [(([(H)C(BU t)N]2)GaI)2], with [MESiMe3J 2] (M = Li or Na; e = N, P, or As): Structural, EPR, and ENDOR characterization of paramagnetic gallium(lll) pnictide complexes

Article abstract of DOI:10.1021/ic0486825

The reactions of the paramagnetic gallium(ll) complex [((Bu ^t-DAB)Gal)₂] (Bu^t-DAB = {(Bu ^tNC(H)}₂) with the alkali metal pnictides [ME(SiMe ₃)₂] (M = Li or Na; E = N, P, or As) have been carried out under a range of stoichiometries. The 1:2 reactions have led to a series of paramagnetic gallium(III)-pnictide complexes, [(Bu^t-DAB) Ga{E(SiMe₃)₂}I] (E = N, P, or As), while two of the 1:4 reactions afforded [(Bu^t-DAB)Ga{ E(SiMe₃) ₂}₂] (E = P or As). In contrast, treatment of [{(Bu ^t-DAB)Gal)₂] with 4 equiv of [NaN(SiMe₃) ₂} resulted in a novel gallium heterocycle coupling reaction and the formation of the diradical species [(Bu^t-DAB)Ga{N(SiMe ₃)₂){[CC(H)N₂(Bu^t) ₂]Ga[N(SiMe₃J₂]CH₃)]. The mechanism of this unusual reaction has been explored, and evidence suggests it involves an intramolecular transmethylation reaction. The X-ray crystal structures of all prepared complexes are reported, and all have been characterized by EPR and ENDOR spectroscopies. The observed spin Hamiltonian parameters provide a detailed picture of the distribution of the unpaired spin density over the molecular frameworks of the complexes.

Full text of DOI:10.1021/ic0486825

Inorg. Chem. 2005, 44, 2098−2105
Reactions of a Gallium(II)
−
Diazabutadiene Dimer,
t
[
{{[(H)C(Bu )N]2
}
GaI 2
}
], with [ME(SiMe ) ] (M
)
Li or Na; E ) N, P, or
3 2
As): Structural, EPR, and ENDOR Characterization of Paramagnetic
Gallium(III) Pnictide Complexes
Karen L. Antcliff, Robert J. Baker, Cameron Jones,* Damien M. Murphy,* and Richard P. Rose
Center for Fundamental and Applied Main Group Chemistry, School of Chemistry,
Cardiff UniVersity, P.O. Box 912, Park Place, Cardiff CF10 3TB, United Kingdom
Received September 20, 2004
t
t
t
The reactions of the paramagnetic gallium(II) complex [
{
(Bu -DAB)GaI
}
2
] (Bu -DAB ) {(Bu )NC(H)
}
2
) with the
alkali metal pnictides [ME(SiMe ] (M Li or Na; E
stoichiometries. The 1:2 reactions have led to a series of paramagnetic gallium(III)
3
)
2
)
)
N, P, or As) have been carried out under a range of
pnictide complexes,
−
t
t
[
(
(Bu -DAB)Ga
{
E(SiMe
3
)
2
}
I] (E
)
N, P, or As), while two of the 1:4 reactions afforded [(Bu -DAB)Ga
{
3 2
E(SiMe ) }
2
]
t
E
)
P or As). In contrast, treatment of [
{(Bu -DAB)GaI
}
2
] with 4 equiv of [NaN(SiMe
3
)
2
] resulted in a novel gallium
t
(But)
heterocycle coupling reaction and the formation of the diradical species [(Bu -DAB)Ga
{N(SiMe }{[CC(H)N ]-
)
3 2
2
2
3 2 3
Ga[N(SiMe ) ]CH }]. The mechanism of this unusual reaction has been explored, and evidence suggests it involves
an intramolecular transmethylation reaction. The X-ray crystal structures of all prepared complexes are reported,
and all have been characterized by EPR and ENDOR spectroscopies. The observed spin Hamiltonian parameters
provide a detailed picture of the distribution of the unpaired spin density over the molecular frameworks of the
complexes.
Introduction
halide and L ) ether, amine, or phosphine. These lead to an
array of remarkable suboxidation state “metalloid” cluster
The chemistry of gallium(III) pnictide complexes is very
4
-
3 2
compounds, e.g. [Ga84{N(SiMe ) }20], via controlled dis-
1
well developed. In this field the bis(silyl)pnictide ligands,
3
proportionation reactions. Related clusters derived from the
-
[
E(SiMe
)
3 2
] , E ) N, P, As, or Sb, are especially important,
t
-
dialkyl phosphide ligand, [PBu
2
] , have also been described
as their gallium complexes have been widely used as
precursors to the binary gallium pnictides, GaE, via thermal
decomposition pathways.1 Recently, this chemistry has been
in the last 2 years, e.g. [Ga16(PBut
Br ] . In gallium(II) chemistry, amide complexes are rare
6
10] and [Ga51(PBut
4
)
2 14
-
2
)
3- 5
6
,2
and, to the best of our knowledge, there is only one
extended to gallium(I) with the reactions of [LiN(SiMe
3 2
) ]
7
structurally characterized phosphide complex and no known
with “metastable” gallium(I) halides, [{GaX(L)} ], X )
n
arsenides. Of the known crystallographically authenticated
-
gallium(II) complexes, the [E(SiMe
3
)
2
] ligands have not
*
Authors to whom correspondence should be addressed. E-mail:
been featured. Considering their importance to Ga(III)
Jonesca6@cardiff.ac.uk (C.J.); murphydm@cardiff.ac.uk (D.M.M.).
(
1) E.g.: (a) Carmalt, C. J. Coord. Chem. ReV. 2001, 223, 217. (b) Schulz,
S. Coord. Chem. ReV. 2001, 215, 1. (c) Wells, R. L. Coord. Chem.
ReV. 1992, 112, 273. (d) Taylor, M. J.; Brothers, P. J. In Chemistry of
Aluminium, Gallium, Indium and Thallium; Downs, A. J., Ed.;
Chapman and Hall: London, 1993; pp 517-524. (e) Lappert, M. F.;
Power, P. P.; Senger, A. R.; Srivastava, R. C. Metal and Metalloid
Amides; Ellis-Horwood: Chichester, U.K., 1980; and references
therein.
(3) Schnepf, A.; Schn o¨ ckel, H.-G. Angew. Chem., Int. Ed. 2002, 41, 3532
and references therein.
(4) Steiner, J.; St o¨ sser, G.; Schn o¨ ckel, H.-G. Angew. Chem., Int. Ed. 2003,
42, 1971.
(5) Steiner, J.; St o¨ sser, G.; Schn o¨ ckel, H.-G. Angew. Chem., Int. Ed. 2004,
43, 302.
(6) (a) Linti, G.; Kotsler, W.; Rodig, A. Z. Anorg. Allg. Chem. 2002, 628,
1319. (b) Schmidt, E. S.; Schier, A.; Mitzel, N. W.; Schmidbaur, H.
Z. Naturforsch. B 2001, 56, 458. (c) Pott, T.; Jutzi, P.; Schoeller, W.
W.; Stammler, A.; Stammler, H. G. Organometallics 2001, 20, 5492.
(d) Linti, G.; Frey, R.; Schmidt, M. Z. Naturforsch. B 1994, 49, 958.
(7) Li, W. W.; Wei, P.; Beck, B. C.; Xie, Y.; Schaefer, H. F., III; Su, J.;
Robinson, G. H. Chem. Commun. 2000, 453.
(
2) E.g.: (a) Carmalt, C. J.; Mileham, J. D.; White, A. J. P.; Williams, D.
J. Dalton Trans. 2003, 4255. (b) Jouet, R. J.; Wells, R. L.; Rheingold,
A. L.; Incarvito, C. D. J. Organomet. Chem. 2000, 601, 191. (c) Janik,
J. F.; Baldwin, R. A.; Wells, R. L.; Pennington, W. T.; Schimek, G.
L.; Rheingold, A. L.; Liable-Sands, L. M. Organometallics 1996, 15,
5385. (d) Barry, S. T.; Richeson, D. S. Chem. Mater. 1994, 6, 2220.
2098 Inorganic Chemistry, Vol. 44, No. 6, 2005
10.1021/ic0486825 CCC: $30.25
© 2005 American Chemical Society
Published on Web 02/11/2005

Paramagnetic Gallium(III) Pnictide Complexes
Chart 1a
flow He Cryostat. The ENDOR spectra were obtained using 8 dB
rf power from a ENI A-300 RF amplifier with 75 or 250 kHz rf
modulation depth. Computer simulations were carried out using
Bruker’s Simfonia program.13 Mass spectra were recorded using a
VG Fisons Platform II instrument under APCI conditions. IR spectra
were recorded using a Nicolet 510 FT-IR spectrometer as Nujol
mulls between NaCl plates. Melting points were determined in
sealed glass capillaries under argon and are uncorrected. Mi-
t
10
croanalyses were obtained from Medac Ltd. [{(Bu -DAB)GaI}
2
],
14
15
15
[
NaN(SiMe ) ], [LiP(SiMe ) .DME], and [LiAs(SiMe ) .DME]
3 2 3 2 3 2
were synthesized by literature procedures.
t
Preparation of [(Bu -DAB)Ga{N(SiMe
3
)
2
}I] (5). To a solution
t
3
of [{(Bu -DAB)GaI}
2
] (0.30 g, 0.41 mmol) in Et
] (0.15 g, 0.83 mmol) in Et O (15 cm ) at -78
C over 5 min. The resultant solution was warmed to room
2
O (15 cm ) was
3
added [NaN(SiMe
)
3 2
2
a
t
i
R ) Bu or C6H3Pr 2-2,6.
°
temperature and stirred overnight to yield a yellow solution and
white precipitate. Volatiles were removed in vacuo, and the residue
was extracted with hexane (20 cm ). Filtration, concentration, and
chemistry and the amide’s ability to stabilize suboxidation
state gallium clusters, it was our intention to prepare gallium-
3
(
II)-bis(silyl)pnictide complexes and investigate their prop-
erties.
The ability of the diazabutadiene class of ligand to stabilize
cooling to -30 °C overnight yielded orange crystals of 5 (0.10 g,
-1
4
6%). Mp: 154-156 °C. IR (ν/cm ; Nujol): 1262 (s), 1202 (s),
9
19 (sh), 883 (sh), 829 (s), 775 (w), 760 (w), 721 (s), 669 (w). MS
diamagnetic and paramagnetic gallium complexes with the
+
+
t
(m/z; APCI): 524 [M , 100%], 397 [M - I, 55%], 169 [Bu -
+
metal in a variety of oxidation states, e.g. 1-4 (Chart 1), is
DABH , 23%]. Anal. Calcd for C H N GaSi I: C, 36.58; H, 7.29;
16
38
3
2
now well-known.6
c,8-11
In our group, and that of Schmid-
N, 8.00. Found: C, 36.11; H, 7.36; N, 8.31.
t
baurs, this ability has been most evidently exploited in the
formation of the valence isoelectronic N-heterocyclic carbene
analogues, 4, the coordination chemistry of which is currently
emerging.12 As a component of those studies, we have
Preparation of [(Bu -DAB)Ga{P(SiMe
of [{(Bu -DAB)GaI}
3
)
2
}I] (6). To a solution
t
3
2 2
] (0.30 g, 0.41 mmol) in Et O (15 cm ) was
added [LiP(SiMe .DME] (0.22 g, 0.82 mmol) in Et O (15 cm )
3
3
)
2
2
at -78 °C over 5 min. The resultant solution was warmed to room
temperature and stirred overnight to yield a red solution. Volatiles
developed a synthetic route to 3, which we have used as a
were removed in vacuo, and the residue was extracted with hexane
t 10
precursor to 4, R ) Bu . Additionally, we saw 3 as a
3
(
20 cm ). Filtration, concentration, and cooling to -30 °C overnight
potential precursor to gallium(II) pnictide complexes. To this
yielded red crystals of 6 (0.10 g, 45%). Mp: 124-126 °C. IR (ν/
t
t
end, the reactivity of [{(Bu -DAB)GaI}
2
], 3 (Bu -DAB )
-1
cm ; Nujol): 1261 (m), 1206 (w), 1096 (w), 1018 (w), 719 (m).
t
{
(Bu )NC(H)}
2
), toward [ME(SiMe
3
)
2
] (M ) Li or Na; E )
+
+
3 2
MS (m/z; APCI): 414 [M - I, 20%], 365 [M - P(SiMe ) , 31%],
t
+
N, P, or As) has been examined. The unexpected results of
this study are reported here.
169 [Bu -DABH , 100%].
Preparation of [(Bu -DAB)Ga{As(SiMe ) }I] (7). To a solution
t
3
2
t
3
of [{(Bu -DAB)GaI}
added [LiAs(SiMe
2
] (0.30 g, 0.41 mmol) in Et
.DME] (0.26 g, 0.82 mmol) in Et O (15 cm )
2
O (15 cm ) was
Experimental Section
3
3
)
2
2
at -78 °C over 5 min. The resultant solution was warmed to room
temperature and stirred overnight to yield a red solution. Volatiles
General Methods. All manipulations were carried out using
standard Schlenk and glovebox techniques under an atmosphere
of high-purity argon. Diethyl ether and hexane were distilled over
were removed in vacuo, and the residue was extracted with hexane
3
(
20 cm ). Filtration, concentration, and cooling to -30 °C overnight
Na/K alloy, toluene-d
was distilled over CaH
8
was distilled over potassium, and CD
and then freeze/thaw degassed prior to use.
2 2
Cl
yielded red crystals of 7 (0.08 g, 33%). Mp: 130-132 °C. IR (ν/
2
-
1
cm ; Nujol): 1457 (s), 1368 (s), 1361 (s), 1328 (sh), 1262 (s),
The continuous wave (CW) EPR/ENDOR spectra were recorded
on an X-band Bruker ESP300E series spectrometer equipped with
an ESP360 DICE ENDOR unit, operating at 12.5 kHz field
modulation in a Bruker EN801 cavity. The ENDOR spectra were
recorded at 10 K using an Oxford instruments ESR 900 continuous-
1
244 (s), 1213 (s), 836 (m), 776 (s), 747 (s), 691 (s), 620 (s). MS
+
+
(m/z; APCI): 291 [GaAs(SiMe
3
)
2
, 10%], 221 [As(SiMe
3
)
2
, 5%],
I: C,
t
+
1
3
69 [Bu -DABH , 100%]. Anal. Calcd for C16 GaAsSi
H
38
N
2
2
2.78; H, 6.53; N, 4.78. Found: C, 32.16; H, 6.59; N, 4.51.
t
Preparation of [(Bu -DAB)Ga{P(SiMe
3
)
2
}
2
] (8). To a solution
t
3
(
8) (a) Cloke, F. G. N.; Hanson, G. R.; Henderson, M. J.; Hitchcock, P.
B.; Raston, C. L. J. Chem. Soc, Dalton Trans. 1989, 1002. (b) Kaim,
W.; Matheis, W. J. Chem. Soc., Chem. Commun. 1991, 597. (c) Power,
P. Chem. ReV. 2003, 103, 789 and references therein.
of [{(Bu -DAB)GaI}
added LiP(SiMe ‚DME (0.45 g, 1.60 mmol) in Et
78 °C over 5 min. The resultant solution was warmed to room
2 2
] (0.30 g, 0.41 mmol) in Et O (15 cm ) was
3
3
)
2
2
O (15 cm ) at
-
temperature and stirred overnight. Volatiles were removed in vacuo,
(9) Brown, D. S.; Decken, A.; Cowley, A. H. J. Am. Chem. Soc. 1995,
3
1
17, 7, 5421.
and the residue was extracted with hexane (20 cm ). Filtration,
(
10) (a) Baker, R. J.; Farley, R. D.; Jones, C.; Kloth, M.; Murphy, D. M.
J. Chem. Soc., Dalton Trans. 2002, 3844, (b) Baker, R. J.; Farley, R.
D.; Jones, C.; Mills, D. P.; Kloth, M.; Murphy, D. M. Chem.sEur.
J., in press.
concentration, and cooling to -30 °C overnight yielded red crystals
-
1
of 8 (0.08 g, 33%). Mp: 160-162 °C. IR (ν/cm ; Nujol): 1369
(s), 1337 (s), 1262 (sh), 1243 (s), 1210 (s), 937 (s), 832 (m), 744
(
11) (a) Schmidt, E. S.; Jockisch, A.; Schmidbaur, H. J. Am. Chem. Soc.
+
+
(
s), 680 (s). MS (m/z; APCI): 593 [M , 65%], 415 [M
-
1999, 121, 9578. (b) Schmidt, E. S.; Schier, A.; Schmidbaur, H. J.
t
+
3 2
P(SiMe ) , 75%], 169 [Bu -DABH , 100%]. Anal. Calcd for
Chem. Soc., Dalton Trans. 2001, 505.
(
12) (a) Baker, R. J.; Jones, C.; Kloth, M.; Platts, J. A. Angew. Chem., Int.
Ed. 2003, 42, 2660. (b) Baker, R. J.; Jones, C.; Platts, J. A. J. Am.
Chem. Soc. 2003, 125, 10534. (c) Baker, R. J.; Jones, C.; Platts, J. A.
J. Chem. Soc., Dalton Trans. 2003, 3673. (d) Baker, R. J.; Jones, C.;
Kloth, M.; Platts, J. A. Organometallics 2004, 23, 4811.
(13) WINEPR SIMFONIA, version 1.25; Br u¨ ker Analytische: Messtechnick,
Germany, 1996.
(14) Clark, D. L.; Sattelberger, A. P. Inorg. Synth. 1997, 31, 307.
(15) Fritz, G.; Hoelderich, W. Z. Anorg. Allg. Chem. 1976, 422, 104.
Inorganic Chemistry, Vol. 44, No. 6, 2005 2099

Antcliff et al.
Table 1. Summary of Crystallographic Data for Compounds 5-10
5
6
7
8
9
10
empirical formula
fw
cryst system
space group
a (Å)
C16H38GaIN3Si2
525.29
orthorhombic
Pnma
18.408(4)
14.517(3)
8.9790(18)
90
2399.4(8)
4
C16H38GaIN2PSi2
542.25
monoclinic
P21/c
15.011(3)
18.299(4)
9.2890(19)
92.58(3)
2549.0(9)
4
C16H38AsGaIN2Si2
586.20
monoclinic
P21/c
15.009(3)
18.462(4)
9.3350(19)
92.77(3)
2583.7(9)
4
C22H56GaN2P2Si4
592.71
monoclinic
P2/c
31.505(6)
9.822(2)
16.731(3)
90.69(3)
5176.9(18)
6
C22H56As2GaN2Si4 C33H78Ga2N6Si4
680.61
monoclinic
C2/c
810.81
orthorhombic
Pna21
65.250(13)
9.859(2)
16.775(3)
103.72(3)
10483(4)
12
25.772(5)
15.892(3)
11.214(2)
90
4592.9(16)
4
b (Å)
c (Å)
â (deg)
3
V (Å )
Z
F(calcd) (g‚cm-3)
1.454
2.537
1068
1.413
2.449
1100
1.507
3.625
1172
1.141
1.042
1914
1.294
1.173
-
1
µ (mm
F(000)
)
2.817
1.305
4260
1744
cryst size (mm)
θ range (deg)
reflcns collcd
Rint
0.25 × 0.25 × 0.20 0.30 × 0.15 × 0.10 0.35 × 0.25 × 0.15 0.30 × 0.25 × 0.20 0.30 × 0.25 × 0.20 0.25 × 0.25 × 0.15
3.17-27.47
13 527
2.94-27.48
18 139
2.93-27.14
16 234
3.08-26.02
31 339
3.05-26.36
35 496
3.01-25.32
19 377
0.0533
0.0724
0.1266
0.0802
0.0795
0.0652
data/restraints/params 2847/0/123
5798/0/221
1.022
0.0424
5522/0/221
1.020
0.0766
10 138/24/479
1.050
0.0551
10 518/18/479
1.026
0.0456
7754/1/433
1.024
0.0543
2
goodness of fit on F
R1 indices [I > 2σ(I)] 0.0283
1.034
a
a
wR2 indices (all data) 0.0608
0.0868
0.2097
0.1391
0.1012
0.1274
largest peak and hole
0.487/-0.660
0.838/-0.534
2.814 (near As1)/
-2.045 (near I1)
1.135/-0.580
0.874/-0.628
1.830 (near Ga1)/
-0.435
-
3
(e‚Å )
a
R1(F) ) {Σ(Fo| - Fc|)/Σ|Fo|} for reflections with Fo > 4(σ(Fo)). wR2(F ) ) {Σw(Fo| - Fc| ) /Σw|Fo | }1/2, where w is the weight given each reflection.
2
2
2 2
2 2
Table 2. Selected Metrical Parameters for Complexes 5-9
av Ga-N
Ga-I
(Å)
Ga-E
(Å)
N-Ga-N
(deg)
Σ angles at E
av C-N
(in heterocycle) (Å)
C-C
(in heterocycle) (Å)
t
(
Bu -DAB) (Å)
(deg)
5
6
7
8
9
1.956
1.963
1.957
1.986
1.991
2.5906(5)
2.5893(8)
2.5944(12)
1.868(2)
85.86(11)
85.88(13)
85.6(3)
84.74(15)
84.62(14)
360.0
323.3
316.8
330.4 (av)
322.6 (av)
1.329
1.326
1.318
1.330
1.328
1.406(5)
1.395(6)
1.411(12)
1.394(6)
1.382(6)
2.2991(11)
2.3893(12)
2.343 (av)
2.437 (av)
C
22
H
56
N
2
GaP
2
Si
4
: C, 43.58; H, 9.52; N, 4.73. Found: C, 43.90;
diffractometer using a graphite monochromator with Mo KR
radiation (λ ) 0.710 73 Å). The data were collected at 150 K, and
H, 9.73; N, 4.84.
t
2
Preparation of [(Bu -DAB)Ga{As(SiMe
3 2
) }
2
] (9). To a solution
the structures were solved by direct methods and refined on F by
t
3
16
of [{(Bu -DAB)GaI}
added a solution of LiAs(SiMe
Et
warmed to room temperature and stirred overnight to yield a red
solution. Volatiles were removed in vacuo, and the residue was
2
] (0.30 g, 0.41 mmol) in Et
‚DME (0.52 g, 1.60 mmol) in
O (15 cm ) at -78 °C over 5 min. The resultant solution was
2
O (15 cm ) was
full-matrix least squares (SHELX97) using all unique data. All
3
)
2
non-hydrogen atoms are anisotropic with H atoms included in
calculated positions (riding model). Crystal data and details of data
collections and refinement are given in Table 1. Selected metrical
parameters for 5-9 are compiled in Table 2.
3
2
3
extracted with hexane (20 cm ). Filtration, concentration, and
cooling to -30 °C overnight yielded red crystals of 9 (0.15 g, 54%).
Mp: 158-160 °C. IR (ν/cm ; Nujol): 1458 (s), 1376 (s), 1260
Results and Discussion
-
1
Synthetic and Structural Studies. The reactions of 3 with
(
w), 1241 (w), 1208 (w), 834 (m), 722 (w). MS (m/z; APCI): 680
+
+
t
+
2 equiv of [ME(SiMe ) ] (M ) Li or Na; E ) N, P, or As)
[M , 18%], 624 [M - Bu , 18%], 459 [M - As(SiMe
3
)
2
, 32%],
Si : C,
3 2
t
+
afforded good yields of the mono(pnictido)gallium(III)
1
3
69 [Bu -DABH , 100%]. Anal. Calcd for C22
H
56
N
2
GaAs
2
4
8.82; H, 8.29; N, 4.12. Found: C, 38.33; H, 8.37; N, 4.13.
complexes, 5-7 (Scheme 1). The mechanism of formation
of these compounds is uncertain, but it must involve salt
elimination, Ga-Ga bond cleavage, and disproportionation
reactions. In this respect, it is noteworthy that related
t
(But)
Preparation of [(Bu -DAB)Ga{N(SiMe
3
)
2
}{[CC(H)N
2
2
]-
t
Ga[N(SiMe
(
(
3
)
2
]CH
0.30 g, 0.41 mmol) in Et
0.30 g, 1.60 mmol) in Et
3
}] (10). To a solution of [{(Bu -DAB)GaI}
2
]
]
3
2 3
O (15 cm ) was added [NaN(SiMe )
2
3
2
O (15 cm ) at -78 °C over 5 min. The
reactions of organodigallium(II) diiodides, [{GaI(R)}
R ) C(SiMe , with carboxylate salts do not lead to Ga-
Ga bond cleavage but to salt elimination and the formation
carboxylate-bridged gallium(II) complexes, e.g. [{Ga(R)}
2
],
resultant solution was warmed to room temperature and stirred
3 3
)
overnight to yield a brown suspension. Volatiles were removed in
3
vacuo, and the residue was extracted with hexane (20 cm ).
2
-
Filtration, concentration, and cooling to -30 °C overnight yielded
17
-
1
{µ-O
2 2
C(Ph)} ]. In the formation of 5-7, the only identified
olive crystals of 10 (0.10 g, 30%). Mp: 123-125 °C. IR (ν/cm
;
byproducts were gallium metal and small amounts of the
known Ga(III) complex [Ga(Bu -DAB) ], 1. It is interesting
Nujol): 1295 (w), 1244 (w), 1200 (w), 957 (s), 904 (w), 875 (s),
t
8
t
8
33 (w), 721 (w), 669 (m). MS (m/z; APCI): 413 [(Bu -DAB)Ga-
2
+
t
+
(Me){N(SiMe
)
3 2
} , 14%], 398 [(Bu -DAB)Ga{N(SiMe
3
)
2
} , 4%],
H , 100%].
X-ray Crystallography. Crystals of 5-10 suitable for X-ray
t
+
+
252 [(Bu -DAB)GaMe , 16%], 161 [N(SiMe
3
)
2
(16) Sheldrick, G. M. SHELX-97; University of G o¨ ttingen: G o¨ ttingen,
Germany, 1997.
(
17) (a) Uhl, W.; El-Hamdan, A. Eur. J. Inorg. Chem. 2004, 969. (b) Uhl,
W. Chem. Soc. ReV. 2000, 29, 259. (c) Uhl, W. AdV. Organomet.
Chem. 2004, 51, 53.
structural determination were mounted in silicone oil. Crystal-
lographic measurements were made using a Nonius Kappa CCD
2100 Inorganic Chemistry, Vol. 44, No. 6, 2005

Paramagnetic Gallium(III) Pnictide Complexes
Scheme 1
Figure 1. Thermal ellipsoid plot (30% probability surface) of the molecular
t
structure of [(Bu -DAB)Ga{N(SiMe3)2}I] (5). Hydrogen atoms are omitted
for clarity. Selected metrical parameters are listed in Table 2. Symmetry
transformation used to generate equivalent atoms: ′, x, -y + 1/2, -z.
that when the reactions were carried out in 1:1 stoichiom-
etries, 5-7 were formed in reduced yields and significant
amounts of 3 were recovered unreacted. Therefore, it appears
that the mechanism of formation of these compounds requires
2
equiv of the alkali-metal pnictide. Moreover, due to the
observed disproportionation processes, it is clear that each
dimeric molecule of 3 can give rise to only one molecule of
monomeric 5-7 in these reactions. It should also be
mentioned that attempts to prepare the antimonide analogue
of 5-7 by reaction of 3 with [LiSb(SiMe
3 2
) ] were not
successful and led to the formation of the known distibine
18
{Sb(SiMe
) }
3 2 2
via an oxidative coupling of the antimonide
fragment.
Due to their paramagnetic nature, no meaningful NMR
spectroscopic data could be obtained for 5-7. Consequently,
X-ray crystallographic studies were required to elucidate their
structures. The molecular structures of 5 and 7 are depicted
in Figures 1 and 2, while that for 6 was found to be
isomorphous to 7 and, therefore, has been deposited as
Supporting Information (see also Table 1). The geometries
of the diazabutadiene ligands in 5-7 (Table 2) are similar
to each other and are suggestive of significant delocalization,
as has been found in related paramagnetic complexes, e.g.
Figure 2. Thermal ellipsoid plot (30% probability surface) of the molecular
t
structure of [(Bu -DAB)Ga{As(SiMe ) }I] (7). Hydrogen atoms are omitted
3 2
for clarity. Selected metrical parameters are listed in Table 2.
arsenido-gallane fragments, respectively. The Ga-P bond
in 6 [2.2991(11) Å] is one of the shortest yet reported and
can be compared with the mean Ga-P(terminal phosphide)
20
distance for all previously reported structures (2.39 Å).
t
Moreover, it is very close to that in [(Bu )
2
Ga{P(Mes*)-
t
10
[GaI
2
(Bu -DAB)]. Likewise, the geometries about the
t
SiPh
3
6 2 3
}] (Mes* ) C H Bu -2,4,6) [2.295(3) Å] which has
gallium centers of the complexes are comparable. The only
obvious trend in the series involves the angles about the
pnictide centers in 5-7. As would be expected, the amido
N-center in 5 is trigonal planar, while the geometries of
the P- and As-centers in 6 and 7 tend toward pyramidal.
The Ga-N(amide) bond length of 1.868(2) Å in 5 is greater
been postulated as having a weak Ga-P π-contribution to
21
the bond. Clearly, in 6 this cannot be the case as the gallium
and phosphorus centers do not have trigonal planar geom-
(19) (a) Atwood, D. A.; Atwood, V. O.; Cowley, A. H.; Jones, R. A.;
Atwood, J. L.; Bott, S. G. Inorg. Chem. 1994, 33, 3251. (b) Brothers,
P. J.; Wehmschulte, R. J.; Olmstead, M. M.; Ruhlandt-Senge, K.;
Parkin, S. R.; Power, P. P. Organometallics 1994, 13, 2792. (c)
K u¨ hner, S.; Kuhnle, R.; Hausen, H.-D.; Weidlein, J. Z. Anorg. Allg.
Chem. 1997, 623, 25.
t
than its Ga-N(Bu -DAB) interactions but identical with
19
the Ga-N distances in [Ga{N(SiMe
3 2 3
) } ]. Complexes 6
and 7 contain rare examples of terminal phosphido- and
(
20) Determined from a survey of the Cambridge Crystallographic Database,
September, 2004.
(
18) Becker, G.; Freudenblum, H.; Witthauer, C. Z. Anorg. Allg. Chem.
(21) Petrie, M. A.; Ruhlandt-Senge, K.; Power, P. P. Inorg. Chem. 1992,
31, 4038.
1982, 492, 37.
Inorganic Chemistry, Vol. 44, No. 6, 2005 2101

Antcliff et al.
etries. Importantly, 7 possesses the shortest Ga-As single
bond yet reported [2.3893(12) Å], which is significantly
shorter than in related complexes, e.g. 2.421 Å average in
22
[Ga{As(SiMe
3 2 3
) } ]. The only shorter Ga-As interactions
are the resonance stabilized double bonds [2.318(1) Å] in
3 2 2 3 4
} Ga )} ].
{As(SiPri
23
[
{Li(THF)
Considering the formation of 5-7, it is perhaps not
surprising that the treatment of 3 with 4 equiv of
[
LiE(SiMe
3
)
2
] (E ) P, As) afforded compounds of the type
t
[(Bu -DAB)Ga{E(SiMe ], E ) P (8) and As (9), in
3 2 2
) }
moderate yields (Scheme 1). Similarly, treating 6 or 7 with
equiv of [LiE(SiMe ] led to these complexes. More
unexpected was the result of the related reaction of 3 with 4
equiv of [NaN(SiMe ]. This led, reproducibly, to a moder-
1
3 2
)
3 2
)
ate yield of the unusual coupled diradical product, 10, as
the only identifiable product.
Figure 3. Thermal ellipsoid plot (30% probability surface) of the molecular
structure of [(Bu -DAB)Ga{As(SiMe3)2}2] (9). Hydrogen atoms are omitted
for clarity. Selected metrical parameters are listed in Table 2.
The mechanism of formation of 10 has been investigated,
t
t
and it is believed that the initial reaction product is [(Bu -
DAB)Ga{N(SiMe
3 2 2
) } ] (cf. 8 and 9). This is then thought
to undergo an intramolecular transmethylation reaction to
t
give the intermediate [(Bu -DAB)Ga{N(SiMe
3
)
2
}Me], 11.
with
3 2 3 3
either [LiN(SiMe ) ] or N(SiMe ) , which both give rise
This proposal has precedent in the reactions of GaCl
3
24
25
to Si-C bond scissions and methyl migrations to the gallium
centers. It must be said that, in the reaction mixture that gave
10, we have not been able to isolate the expected elimination
product, {(Me Si)NSi(Me) . The final product, 10, could
3
2 2
}
be formed by deprotonation of the diazabutadiene backbone
of one molecule of the intermediate, 11, by the GaMe moiety
of another (i.e. CH elimination). This has been disproved
4
by intentionally preparing 11 from the reaction of 5 with
MeLi. The product was found to be stable toward the
formation of 10. Alternatively, 10 could be formed by
deprotonation of the backbone of the intermediate, 11, with
excess [NaN(SiMe
carbanion could then attack [(Bu -DAB)Ga{N(SiMe ) }I], 5,
3 2
) ] in the reaction mixture. The resulting
t
3 2
which must also be a reaction intermediate, to give 10 via
NaI elimination. To test this hypothesis, an equimolar mixture
of 5, 11, and [NaN(SiMe ) ] in diethyl ether was warmed
3 2
from -78 to 25 °C. Although the reaction was not clean,
compound 10 was isolated from it in a low yield (ca. 10%).
Presumably, similar coupling reactions were not observed
in the preparations of 8 and 9 as the pyramidal geometries
at their pnictogen centers (vide infra) circumvent close
interactions between their gallium centers and SiMe groups.
3
This would be a likely prerequisite for methyl migration
reactions to occur.
Figure 4. Thermal ellipsoid plot (30% probability surface) of the molecular
structure of [(Bu -DAB)Ga{N(SiMe3)2}{[CC(H)N2(Bu )2]Ga[N(SiMe3)2]-
CH3}] (10). Hydrogen atoms are omitted for clarity. Selected bond lengths
t
t
(
Å) and angles (deg): Ga(1)-N(1) 2.036(5), Ga(1)-N(2) 1.986(5), Ga-
(1)-N(3) 1.910(5), Ga(1)-C(17) 2.023(6), Ga(2)-N(4) 1.990(5), Ga(2)-
N(5) 1.972(5), Ga(2)-N(6) 1.909(4), Ga(2)-C(33) 1.981(6), N(1)-C(1)
1
.333(8), N(2)-C(2) 1.344(8), N(4)-C(17) 1.359(8), N(5)-C(18) 1.315-
Crystals of 8-10 suitable for X-ray diffraction were grown
from hexane solutions, and the molecular structures of 9 and
(
8), C(1)-C(2) 1.382(9), C(17)-C(18) 1.432(8), N(1)-Ga(1)-N(2) 84.5-
(2), N(1)-Ga(1)-N(3) 116.5(2), N(1)-Ga(1)-C(17) 105.1(2), N(2)-
Ga(1)-N(3) 107.2(2), N(2)-Ga(1)-C(17) 124.5(2), N(3)-Ga(1)-C(17)
10 are shown in Figures 3 and 4, respectively. Compound 8
1
15.5(2), N(4)-Ga(2)-N(5) 83.91(18), N(4)-Ga(2)-N(6) 118.4(2), N(4)-
was found to be isostructural to 9, and as a result, its
molecular structure has been deposited as Supporting Infor-
Ga(2)-C(33) 107.6(2), N(5)-Ga(2)-N(6) 115.5(2), N(5)-Ga(2)-C(33)
108.1(3), N(6)-Ga(2)-C(33) 118.1(2).
(
22) Wells, R. L.; Self, M. F.; Baldwin, R. A.; White, P. S. J. Coord. Chem.
mation (see also Table 1). The asymmetric units of both 8
and 9 contain 1.5 crystallographically independent molecules
which show no significant geometrical differences. Conse-
quently, the structure of only the full independent molecule
of each will be discussed here (Table 2). Both are monomeric
1994, 33, 279.
(
(
23) von H a¨ nisch, C.; Hampe, O. Angew. Chem., Int. Ed. 2002, 41, 2095.
24) Luo, B.; Young, V. G., Jr.; Gladfelter, W. L. J. Organomet. Chem.
2002, 649, 268.
(25) Carmalt, C. J.; Mileham, J. D.; White, A. J. P.; Williams, D. J.; Steed,
J. W. Inorg. Chem. 2001, 40, 6035.
2102 Inorganic Chemistry, Vol. 44, No. 6, 2005

Paramagnetic Gallium(III) Pnictide Complexes
Figure 6. X-band EPR spectra of (a) 8 and (b) 9 recorded in hexane at
98 K.
2
Figure 5. X-band EPR spectra of (a) 5, (b) 6, and (c) 7 recorded in hexane
at 298 K.
2
0.1, and 20.5 mT, respectively. Attempts to successfully
with distorted tetrahedral geometries about their gallium
centers. The geometries of the heterocyclic fragments are
similar to those in 5-7, while the average Ga-E bond
lengths of 8 and 9 are significantly greater than those in 6
and 7, presumably due to steric reasons.
The molecular structure of 10 confirms that a ligand
coupling reaction has occurred in its formation. Both the
heterocycles in this compound have similar geometries which
imply significant delocalization over their diazabutadiene
backbones (cf. 5-9). Moreover, the two Ga-N(amide) bond
lengths, 1.910(5) and 1.909(4) Å, are almost identical but
greater than that in 5. Finally, both Ga-C bonds [Ga(1)-
C(17) ) 2.023(6) Å, Ga(2)-C(33) ) 1.981(6) Å] are in the
normal range for such interactions.20
simulate the spectra using commercial simulation packages
1
3
(e.g. SIMFONIA ) proved very difficult due to slight
6
9,71
differences in
Ga hyperfine couplings and isotropic g
values. For example, while an excellent fit with the outer
lines could be achieved (i.e. essentially due to the wider
71
contribution of the Ga isotope), the shape of the inner lines
6
9
was distorted due to overlap with the smaller Ga compo-
nent. This resulted from slight differences in the isotropic g
values which we could not satisfactorily reproduce in the
simulation. Nevertheless, the computer simulations did reveal
an approximate hyperfine splitting of ca. 100 MHz to the
69,71
Ga nucleus of each compound, which represents ca. 0.7%
spin density on the gallium nucleus. This can be rationalized
in terms of a preferential polarization of the unpaired electron
EPR Spectroscopic Studies. The EPR spectra of the
paramagnetic complexes, 5-7, were recorded at X-band
frequency (9 GHz). The resulting room-temperature (298K)
X-band spectra, recorded under identical conditions, for the
2 2 2
away from the N C H fragment and toward the gallium
nuclei due to the influence of the N-, P-, and As-centers. As
a result, a small amount of the unpaired spin density can be
found at the N-, P-, and As-nuclei, as manifested by a
significantly increased number of lines in the EPR spectra.
Despite the increased spin density on the gallium nuclei, the
unpaired electron remains primarily localized on the diaza-
butadiene ligands of 5-7, as confirmed by the relatively large
complexes are shown in Figure 5a-c. A previous EPR study
t
on the related complex [(Bu -DAB)GaI
2
] revealed a relatively
small degree of spin delocalization on the gallium nucleus,
the observed hyperfine splittings being only 3.64 and 4.62
6
9
71
14
1
MHz for Ga and Ga, respectively (representing a 0.03%
N and smaller H EPR hyperfine splittings of ca. 25 MHz
isotropic unpaired spin density on 6 Ga). Hyperfine
splittings to two equivalent nitrogen nuclei (24.14 MHz) and
two R protons of the diazabutadiene ligand (3.92 MHz) and
very weak couplings to the remote 127I nuclei (3.64 MHz)
were also identified in the EPR spectrum of that compound.
As a result of these superimposed hyperfine patterns in the
isotropic spectrum, the overall width of the final spectrum
was 6.0 mT (168 MHz).
9,71
10
and ca. 3.59 MHz, respectively, which are similar to those
t
1
2
for [(Bu -DAB)GaI ]. The H hyperfine couplings were more
accurately determined by ENDOR spectroscopy, as discussed
in the next section.
The room-temperature EPR spectra of 8 and 9 were also
measured, and the resulting spectra are shown in Figure 6.
The widths of the EPR spectra have decreased to 14.8 and
18.1 mT for 8 and 9, respectively (cf. 20.1 and 20.5 mT for
6 and 7). This result indicates that as slightly more electron
By comparison, the EPR spectra for 5-7 are much more
complex and significantly wider, with spectral widths of 19.5,
3 2 3 2
spin is delocalized toward the two P(SiMe ) or As(SiMe )
Inorganic Chemistry, Vol. 44, No. 6, 2005 2103

Antcliff et al.
Figure 7. X-band EPR spectrum of 10 recorded in hexane at 298 K.
substituents, less spin remains on the gallium nucleus, and
therefore smaller 69,71Ga hyperfine splittings are observed,
thus producing a decreased spectral width.
Figure 8. X-band ENDOR spectrum of 7 at (a) 250 kHz modulation depth
and (b) 75 kHz modulation depth (recorded in CD2Cl2/C7D8 at 10 K).
Figure 7 depicts the EPR spectrum for 10 recorded at 298
K. The spectrum is substantially different compared to the
previous spectra, revealing a significantly altered structure
for this paramagnetic complex. Despite the presence of two
unpaired electrons in the two respective diazabutadiene
ligands (i.e. a diradical), the resulting EPR spectrum is not
typical of a system with an S ) 1 triplet ground state and
can best be interpreted as a composite spectrum with isotropic
contributions from two S ) 1/2 species. The frozen solution
spectrum of 10 (see Supporting Information) revealed an
easily identified quartet of ca. 3.0 mT (84 MHz) arising from
the predominantly isotropic hyperfine splitting to one gallium
nucleus, while the room temperature spectrum shows an
unmistakable quartet feature (most clearly seen in the
expanded outer wings of the spectrum) of 0.13 mT (3.64
MHz) separation which is due to a smaller hyperfine
interaction with a second gallium nucleus. The former
gallium splitting of ca. 84 MHz is approximately of the same
order of magnitude as those observed for 5-7 while the latter
spectrum is obtained.27 This is the typical case expected for
carbon-based organic free radicals, and narrow lines will only
be obtained if the hyperfine anisotropy pertaining to the
nucleus is small. For R protons, the anisotropy is about half
of the hyperfine coupling, a, so that the principal values of
the hyperfine tensor for R protons should occur near a/2, a,
and 3/2a. The ENDOR spectra are thus expected to extend
over a wider range (from a/2 to 3/2a) but with some build-
up of intensity at the three principal values of the hyperfine
tensor corresponding to those molecules with their respective
principal axes oriented along the magnetic field.
The X-band ENDOR spectrum of 7 is shown in Figure 8.
As the largest coupling (3/2a) is expected to produce a broad
and weak signal, the ENDOR spectrum was recorded using
a large rf modulation depth of 250 kHz (Figure 8a). The
three principal values of the hyperfine tensor expected for
an R-proton (a/2, a, and 3/2a) are clearly visible; the
measured values are 1.897, 3.795, and 5.692 MHz, respec-
tively (as shown by the stick diagram in Figure 8a). For
R-protons the sign of the isotropic hyperfine coupling is
expected to be negative, compared to â-protons where a
t
coupling of 3.64 MHz is analogous to that observed for [(Bu-
DAB)GaI ]. The EPR spectrum and in particular the dis-
2
crimination of hyperfine interactions to two independent
gallium nuclei therefore confirm the identity of 10 as a
dimeric gallium complex with two noninteracting S ) 1/2
spin systems.
ENDOR Spectroscopic Studies. To extract further in-
formation on the unpaired spin densities in these complexes,
the continuous wave (CW) ENDOR spectra were measured.
The ENDOR spectra of the complexes in disordered solids
26a,27c
positive sign is predicted.
The reason for the difference
in sign relates to the different mechanisms of spin transfer
from the π-center to such protons (spin polarization for R
protons and hyperconjugation for â protons). Knowing the
isotropic coupling is ca. 1.4 G (∼3.9 MHz) from the room-
temperature EPR spectrum, and that this can be assigned a
negative value, the three observed hyperfine tensor compo-
nents of these protons, obtained from the frozen solution
ENDOR spectrum, can therefore be assigned negative values
(Table 3).
A number of additional peaks can also be detected in the
CW ENDOR spectrum with pronounced smaller couplings.
These couplings undoubtedly arise from weaker interactions
to remote protons of the complex. To enhance the resolution
of these additional lines, the spectrum was recorded using a
(
i.e. frozen solutions) are expected to be complicated by the
2
6
broadened nature of the ENDOR response observed. Due
to absorptions from a range of orientations of the radical
with respect to the direction of the magnetic field, the
ENDOR lines arising from weak superhyperfine interactions
to I * 0 nuclei will broaden and may become very weak.
Unless the shape of the EPR spectrum is dominated by a
particular dipolar interaction, there is little or no orientational
selection in the ENDOR experiment and a powder-type
(
27) (a) Hurreck, H.; Kirste, B.; Lubitz, W. Electron Nuclear Double
Resonance Spectroscopy of Radicals in Solution; Applications to
Organic and Biological Chemistry; VCH Publishers: Weinheim,
Germany, 1988. (b) Gerson, F. Acc. Chem. Res. 1994, 27, 63. (c) Hyde,
J. S.; Rist, G. H.; Ericksson, L. E. G. J. Phys. Chem. 1968, 72, 4269.
(
26) (a) O’Malley, P. J.; Babcock, G. T. J. Am. Chem. Soc. 1986, 108,
3
995. (b) Kevan, L.; Narayana, P. A. In Multiple Electron Resonance
Spectroscopy; Dorio, M. M., Freed, J. H., Eds.; Plenum Press: New
York, 1979; Chapter 6, p 229.
2104 Inorganic Chemistry, Vol. 44, No. 6, 2005

Paramagnetic Gallium(III) Pnictide Complexes
Table 3. Relative Sign and Magnitude (MHz) of the Hyperfine
Couplings to the R-Protons and the Remote Protons of the tert-Butyl
Groups of 7 As Determined by ENDOR Spectroscopy
protons
A1
A2
A3
aiso
R-H
tert-butyl
-1.897
+2.859
-3.795
-1.234
-0.863
-5.692
-1.234
-0.863
-3.795
+0.13
+0.19
+
2.301
lower rf modulation depth of 75 kHz (Figure 8b). The
unresolved broad line at the nuclear Lamour frequency for
1
H (ν ) 14.41 MHz at 3.385 mT) is due to a matrix
N
ENDOR line. This line arises from almost purely dipolar
couplings of the unpaired electron with surrounding (remote
matrix) magnetic nuclei.2 The remaining lines can then be
assigned to weak couplings with the remote protons of the
tert-butyl groups (shown by the stick diagram in Figure 8b).
In the case of â-protons, considerably less anisotropy is
generally observed compared to R-protons. As a result, these
interactions exhibit much stronger ENDOR lines in disor-
dered solids. This is the situation for the remote tert-butyl
protons in complex 10 which give rise to small hyperfine
splittings (Figure 8b) which are slightly different for the two
tert-butyl groups, thus implying a small inequivalency in the
unpaired spin distribution on these substituents.
6b
Figure 9. X-band ENDOR spectra of (a) 5, (b) 6, and (c) 7 recorded in
CD2Cl2/C7D8 at 10 K.
Conclusions
In summary, the reactions of a gallium(II) dimeric
t
complex, [{(Bu -DAB)GaI}
2
], with the alkali-metal pnictides,
3 2
[ME(SiMe ) ] (M ) Li or Na; E ) N, P, or As), have been
As discussed earlier in the EPR analysis, the presence of
the electronegative elements in the E(SiMe ) substituents
3 2
produced a noticeable redistribution of electron spin density
carried out under a range of stoichiometries. The reactions
have led to a series of paramagnetic gallium(III)-pnictide
t
complexes, [(Bu -DAB)Ga{E(SiMe
3
)
2
}I] (E ) N, P, As) and
onto the 6 Ga, N, P, and As nuclei. However, it must
be clearly noted that the theoretical isotropic hyperfine
couplings of these elements are very large (435.68 and 553.58
9,71
14
31
75
t
[
(Bu -DAB)Ga{E(SiMe
3
)
2
}
2
] (E ) P, As). In addition, a
novel gallium heterocycle coupling reaction has been ob-
served and its mechanism explored. All prepared complexes
have been characterized by X-ray crystallography, which in
the case of one compound, 7, has highlighted the shortest
Ga-As single bond yet reported. Moreover, each of the
paramagnetic compounds have been characterized in solution
6
9,71
14
31
mT for
Ga; 64.6 mT for N; 474.8 mT for P; 523.11
75
mT for As), so even a very small spin density on the nuclei
will produce an appreciable hyperfine coupling. We were
unable to clearly detect any of these couplings in our CW
ENDOR experiment. Despite the apparently large splittings
to these nuclei observed in the EPR experiments, the unpaired
electron distributions around the diazabutadiene ligand and
the tert-butyl groups remain very similar for complexes 5-9.
1
by EPR spectroscopy and the frozen-solution H ENDOR
spectra of several complexes have been acquired and
analyzed. These spin resonance studies have enabled the
estimation of the spin density over the molecular frameworks
of the compounds. This has shown that although the EPR
spectra of the various complexes appear very different, the
spin densities on the peripheral atoms (e.g. the tert-butyl and
E(SiMe
) substituents) do not significantly differ between
3 2
the complexes, while most of the electron spin remains on
their diazabutadiene backbones.
1
This is confirmed by analysis of the H ENDOR spectra for
5-7 shown in Figure 9 (analogous spectra were also recorded
for 8 and 9ssee Supporting Information). The spectra (and
therefore the associated hyperfine couplings responsible for
the lines) for all complexes are virtually identical, revealing
that the very weak couplings to the protons of the tert-butyl
groups and the larger couplings to the R-protons remain
predominantly unchanged over the series.
The CW ENDOR spectrum of 10 (see Supporting Infor-
mation) is clearly different from those of 5-9 and can be
interpreted in terms of two superimposed patterns arising,
first, from the paramagnetic heterocycle containing the
gallium center that bridges to the other heterocycle (produc-
ing a spectrum analogous to those observed for 5-9) and,
second, from the singly deprotonated heterocycle which
displays a larger coupling to the remaining R-proton. The
ENDOR spectrum supports the earlier claim that the EPR
spectrum of 10 (Figure 7) reveals a substantially different
structure for 10 compared to those of 5-9.
Acknowledgment. We thank the EPSRC (partial stu-
dentship for R.P.R. and postdoctoral fellowship for R.J.B.)
for financial support and for funding the National ENDOR
Service (Grant GR/R17980/01).
Supporting Information Available: Crystallographic data as
CIF files for 5-10, molecular structures of 6 and 8, an EPR
spectrum of 10 in hexane (100 K), and ENDOR spectra for 8-10
(10 K). This material is available free of charge via the Internet at
http://pubs.acs.org.
IC0486825
Inorganic Chemistry, Vol. 44, No. 6, 2005 2105