The reactivity of diazabutadienes toward low oxidation state Group 13 iodides and the synthesis of a new gallium(i) carbene analogue

Article abstract of DOI:10.1039/b206605j

The reactions of two diazabutadiene ligands, Ar-DAB [(ArN=CH)₂, Ar = 2,6-Prⁱ₂C₆H₃] and Bu^t-DAB [(Bu^tN=CH)₂] with either GaI or an AlI₃/Al mixture have afforded the paramagnetic compounds, [{IGa^II(Bu^t-DAB)}₂], [I₂Ga^III- (Bu^t-DAB)] and [I₂Al^III(Ar-DAB)] which have been characterised by X-ray crystallography and EPR spectroscopy. In addition, the diamagnetic, ionic complex, [I₂Al^III(Ar-DAB)]I, has been prepared and structurally characterised. The reduction of [I₂Ga^III(Ar-DAB)] with potassium metal in the presence of various Lewis bases has led to (three structurally characterised compounds, [{(Et₂O)KGa^I(Ar-DAB)}₂], [{(TMEDA)KGa^I(Ar-DAB)}₂] and [:Ga^I(Ar-DAB)]₂[{K(18-crown-6)}₂(μ-18- crown-6)], which contain the second example of an anionic gallium(I) carbene analogue. In addition, the former two compounds display Ga · · · Ga interactions in the solid state which are unprecedented for this complex type.

Full text of DOI:10.1039/b206605j

View Article Online / Journal Homepage / Table of Contents for this issue
The reactivity of diazabutadienes toward low oxidation state
Group 13 iodides and the synthesis of a new gallium(I) carbene
analogue†
Robert J. Baker, Robert D. Farley, Cameron Jones,* Marc Kloth and Damien M. Murphy
Department of Chemistry, University of Wales, Cardiff, P.O. Box 912, Park Place, Cardiff,
UK CF10 3TB
Received 8th July 2002, Accepted 2nd September 2002
First published as an Advance Article on the web 18th September 2002
The reactions of two diazabutadiene ligands, Ar-DAB [(ArN᎐CH) , Ar = 2,6-Pri C H ] and But-DAB [(ButN᎐CH) ]
᎐
᎐
2
2
6
3
2
with either GaI or an AlI3/Al mixture have afforded the paramagnetic compounds, [{IGaII(But-DAB)}2], [I2GaIII-
(But-DAB)] and [I2AlIII(Ar-DAB)] which have been characterised by X-ray crystallography and EPR spectroscopy.
In addition, the diamagnetic, ionic complex, [I2AlIII(Ar-DAB)]I, has been prepared and structurally characterised.
The reduction of [I2GaIII(Ar-DAB)] with potassium metal in the presence of various Lewis bases has led to
three structurally characterised compounds, [{(Et2O)KGaI(Ar-DAB)}2], [{(TMEDA)KGaI(Ar-DAB)}2] and
[:GaI(Ar-DAB)]2[{K(18-crown-6)}2(µ-18-crown-6)], which contain the second example of an anionic
gallium() carbene analogue. In addition, the former two compounds display Ga ؒ ؒ ؒ Ga interactions
in the solid state which are unprecedented for this complex type.
Introduction
Much interest has been paid to diazabutadiene Group 13 com-
plexes over the last 15 years and many complex types have
arisen from this work. Some of the more interesting complexes
that have come forward are the paramagnetic gallium()
complex, 1,1 the diamagnetic gallium() complex, 2,2 and the
remarkable anionic gallium() heterocycle, 3,3 which is a valence
isoelectronic N-heterocyclic carbene (NHC) analogue. This
anion has been obtained in an uncoordinated state (i.e. there is
no Ga: K coordination) as its potassium salt, [:Ga{N(But)-
C2H2N(But)}][K(18-crown-6)(THF)2],3 and as a dimeric potas-
sium complex, 4.4 Through our work with NHC-stabilised
indium hydride complexes5 we have become interested in
forming indium analogues of 3 which are unknown to date. To
this end we recently reported that the reaction of a sterically
encumbered diazabutadiene ligand, (ArN᎐CH) , Ar-DAB, Ar
᎐
2
= 2,6-Pri2C6H3, with In()Cl yields the unusual paramagnetic
indium() complex, 5, via a 1-electron reduction of the Ar-
DAB ligand by the indium centre.6 We have now fully extended
this study by (i) attempting the reduction of 5 to an indium()
NHC analogue, (ii) investigating the reactivity of diazabuta-
dienes toward Ga()I and a AlI3/Al mixture and (iii) preparing a
new gallium() NHC analogue that displays an unprecedented
Ga ؒ ؒ ؒ Ga interaction in the solid state. The results of these
investigations are reported herein.
low temperature gallium metal deposition occured and a deep
red solution resulted which upon work-up afforded a high yield
(> 90%) of the paramagnetic gallium() complex, 6 (Scheme
1). Since carrying out this work Jutzi et al. have reported8
a
similar preparation of this compound, the spectroscopic and
crystallographic data for which were found to be identical to
our own and thus they will not be discussed here. It is worth
noting, however, that this reaction presumably proceeds via a
gallium() intermediate, [GaI(Ar-DAB)], which undergoes an
intramolecular ligand reduction and subsequent dispropor-
tionation reaction to give 6 and gallium metal. The formation
of 6 contrasts with the analogous reaction that gave 5. How-
ever, the yield of 5 was low (15%) and considerable indium
deposition was seen in this reaction which we believed was due
to the formation of an indium analogue of 6 as the major
product. In a parallel study9 we have since confirmed this,
though the paramagnetic indium() product formed is 5-
Results and discussion
Considering the unexpected formation of 5 from the reaction
of Ar-DAB with In()Cl it was decided to investigate the
analogous reactions of diazabutadienes with Ga()I. This
reagent is readily prepared by ultrasonicating gallium metal
with 0.5 equiv. of I2 in toluene.7 When the reaction of GaI with
Ar-DAB was carried out in a 2 : 1 stoichiometry in toluene at
ؒ
coordinate, [InCl2(THF)(Ar-DAB )], with the indium centre
being additionally coordinated by a molecule of THF.
In contrast to the reactions that gave 5 and 6, the 1 : 1 or 1 : 2
† Electronic supplementary information (ESI) available: experimental
EPR spectra for compounds 9 and 7. See http://www.rsc.org/suppdata/
dt/b2/b206605j/
reactions of But-DAB, (ButN᎐CH) , with Ga()I in toluene
᎐
2
yielded predominantly the dimeric complex, 7, but also small
3844
J. Chem. Soc., Dalton Trans., 2002, 3844–3850
This journal is © The Royal Society of Chemistry 2002
DOI: 10.1039/b206605j

View Article Online
Table 1 Isotropic g and hyperfine coupling values for complexes 7, 8, 9
and 5, obtained via simulation of their room temperature EPR spectra
Complex
giso
Ma
1Hb
14Na
127Ic
7
8
2.0030
2.0038
69Ga = 0.13
71Ga = 0.16
69Ga = 0.13
71Ga = 0.165
27Al = 0.285
115In = 2.62
113In = 2.614
0.14
0.14
0.842
0.862
0.12
0.13
0.04
9
2.0038
2.0012
0.59
0.50
0.67
0.50
56
All isotropic hyperfine couplings in mT. a M refers to gallium, alu-
minium or indium. b 1H refers to two equivalent imine protons and 14
N
refers to the two equivalent nitrogen nuclei. c Two equivalent iodine
nuclei simulated in 8 and 9, but only one in 7.
the Ar-DAB ligand is acting as a localised Lewis base because
the protons on the diimine backbone of the ligand resonate
near those of the free ligand and the neutral indium com-
pound, [InBr3(Ar-DAB)], but significantly downfield from
the imine protons in related systems incorporating reduced
diazabutadiene ligands, e.g. 2–4.
The EPR spectra of 7–9 were recorded at X-band frequen-
cies. The spectrum and associated computer simulation for 8 is
shown in Fig. 1.13 This spectrum was particularly well resolved,
Scheme 1 Reagents and conditions: i, GaI, toluene, R = Ar; ii, GaI,
toluene, R = But; iii, AlI3/Al, toluene, R = Ar.
amounts of the monomeric species, 8. It is interesting that the
dimeric gallium() species is the major product in this reaction
whereas in the analogous Ar-DAB reaction the monomeric
gallium() compound, 6, is in the majority. We do not
know why this is but presumably it stems from the steric and
electronic differences between the two diazabutadiene ligands.
An attempt was made to prepare the aluminium analogue
of 6 by reacting Ar-DAB with a 1 : 2 mixture of AlI3 and Al
powder in toluene. This was successful in that over 12 hours
a proportion of the aluminium powder was consumed and a
deep red solution was formed. Volatiles were removed from this
solution and the residue recrystallised from diethyl ether to give
a good yield of the paramagnetic aluminium() compound, 9.
Interestingly, a trace by-product, 10, was also isolated from the
reaction mixture. This was subsequently intentionally prepared
in good yield (66%) by reacting AlI3 with one equivalent of
Ar-DAB in toluene. Presumably the mechanism of formation
of 9 involves the initial formation of 10 which is then reduced
by the aluminium powder, followed by an intramolecular
ligand reduction reaction to give 9. It is also of interest that the
reaction of Ar-DAB with AlI3 yields an ionic product, 10,
which is comparable to the product of the 2 : 1 reaction of
GaCl3 with But-DAB, viz. [GaCl2(But-DAB)][GaCl4].10 How-
ever, these products contrast with that from the analogous
reaction of Ar-DAB with InBr3 which we have recently shown
to be a neutral 5-coordinate adduct, [InBr3(Ar-DAB)].11 The
results of these reactions are also different to that from the
reaction of BCl3 with Ar-DAB which leads to chloroboration
of the diimine and formation of [BCl(ArNC(H)(Cl)C(H)(Cl)-
NAr)].12
Fig. 1 X-Band EPR spectrum and computer simulation of 8 in
CD2Cl2/C7D8 (50 : 50) at 298 K.
allowing accurate spin Hamiltonian parameters to be obtained
(Table 1). Similar spin Hamiltonian parameters were used to
simulate the spectrum of 7 (with only one interacting 127I
nucleus) indicating the similarity of the radical fragment in
both cases. In both gallium complexes, 7 and 8, the spin density
on the 69,71Ga nuclei was very small (0.03%) and negligible on
the 127I nuclei (0.008%). Nevertheless, without this contribution
from the 127I nuclei an accurate simulation could not be
obtained. The hyperfine couplings to the imine protons (0.14
mT) and nitrogen nuclei (ca. 0.85 mT) of 7 and 8 were found to
be different compared to the aluminium complex, 9, and the
previously reported indium complex, 5, where aiso values close
to 0.5 mT were observed (see Table 1). It is noteworthy that
Jutzi et al. reported similar couplings of 0.41 mT and 0.7 mT
for the protons and nitrogen nuclei respectively in 6. Therefore,
it seems that the presence of the tert-butyl groups in 7 and 8
have a significantly different influence on the spin density
around the imine protons and nitrogen nuclei, compared to the
No meaningful NMR data could be obtained on compounds
7–9 due to their paramagnetic nature but all other spectroscopic
data pointed towards their formulations. The ionic compound,
1
10, is not paramagnetic but its H NMR spectrum showed
significantly broadened signals. It is thought that this is due to
an exchange of IϪ ligands at the aluminium centre on the NMR
time scale which could proceed via a 5-coordinate neutral
intermediate, [AlI3(Ar-DAB)]. This proposal seems feasible in
light of the fact that the indium analogue of this compound,
[InBr3(Ar-DAB)], is neutral and 5-coordinate in the solid
1
state.11 The H NMR spectrum of 10 could not be resolved
when solutions of the compound were heated to 70 ЊC or
cooled to 0 ЊC. Further cooling of solutions of 10 led to signifi-
cant precipitation of the compound. Despite the broadening of
the signals in the 1H NMR spectrum of 10 the data suggest that
J. Chem. Soc., Dalton Trans., 2002, 3844–3850
3845

View Article Online
influence of the aryl groups in 5, 6 and 9. This influence by the
tert-butyl groups on the electron spin density is also manifested
in the hyperfine couplings to the metal centres, which was found
to be 0.03% in 7 and 8, but ca. 0.3% in 5 and 9. In all complexes,
7–9, the g values are close to free spin, indicating the essentially
organic nature of the radical.
The X-ray crystal structures of all compounds 7–10 were
obtained and are depicted in Fig. 2–5 (see Table 2). The dimeric
Fig. 4 Molecular structure of compound 9. Selected bond lengths (Å)
and angles (Њ): Al(1)–I(1) 2.4957(11), Al(1)–I(2) 2.5318(11), Al(1)–N(1)
1.889(2), N(1)–C(1) 1.337(3), C(1)–C(1Ј) 1.409(5); N(1)–Al(1)–N(1Ј)
87.42(13), N(1)–Al(1)–I(1) 116.50(7), N(1)–Al(1)–I(2) 111.93(7), I(1)–
Al(1)–I(2) 110.78(4), C(1)–N(1)–Al(1) 108.93(17), N(1)–C(1)–C(1Ј)
116.71(14). Symmetry operation: Ј x, Ϫy ϩ 1/2, z.
Fig. 2 Molecular structure of compound 7. Selected bond lengths (Å)
and angles (Њ): Ga(1)–Ga(1Ј) 2.4232(7), Ga(1)–I(1) 2.6169(5), Ga(1)–
N(2) 1.958(3), Ga(1)–N(1) 1.966(3), N(1)–C(1) 1.322(4), N(2)–C(2)
1.333(4), C(1)–C(2) 1.395(5); N(2)–Ga(1)–N(1) 85.23(11), N(2)–Ga(1)–
Ga(1Ј) 124.87(8), N(1)–Ga(1)–Ga(1Ј) 116.26(9), N(2)–Ga(1)–I(1)
103.47(8), N(1)–Ga(1)–I(1) 107.33(9), I(1)–Ga(1)–Ga(1Ј) 115.153(14),
C(1)–N(1)–Ga(1) 108.8(2), C(2)–N(2)–Ga(1) 109.0(2), N(1)–C(1)–C(2)
118.7(3), N(2)–C(2)–C(1) 118.0(3). Symmetry operation: Ј y, x, Ϫz.
Fig. 5 Molecular structure of the cation of compound 10. Selected
bond lengths (Å) and angles (Њ): Al(1)–I(1) 2.491(3), Al(1)–I(2)
2.487(3), Al(1)–N(2) 1.903(8), Al(1)–N(1) 1.933(8), N(1)–C(1)
1.289(11), N(2)–C(2) 1.325(12), C(1)–C(2) 1.442(13); N(2)–Al(1)–N(1)
85.7(3), N(2)–Al(1)–I(2) 116.4(3), N(1)–Al(1)–I(2) 110.7(2), N(2)–
Al(1)–I(1) 112.8(3), N(1)–Al(1)–I(1) 116.8(3), I(2)–Al(1)–I(1)
112.12(11), C(1)–N(1)–Al(1) 109.9(6), C(2)–N(2)–Al(1) 110.7(6), N(1)–
C(1)–C(2) 117.7(8), N(2)–C(2)–C(1) 115.4(9).
and its gallium centre has a distorted tetrahedral geometry. The
Ga–N bond lengths [1.930 Å avge.] are close to those in 6
[1.9449(10)], as are the backbone C–N and C–C distances
[8: 1.336 Å avge, 1.450(11) Å; 6: 1.3386(15), 1.406(2) Å]. Both
sets of values are comparable to those observed in the singly
reduced But-DAB ligand in the related Ga() compound,
[Ga(But-DAB)2] 1, and are strongly suggestive of delocalised
diazabutadiene ligand systems.
The paramagnetic aluminium() compound, 9, is isomorph-
ous to 6 and both compounds have similar geometries and
bond lengths about their metal centres, which is not surprising
given the almost equivalent covalent radii for the two metals
involved.14 In addition, the bond lengths within the diazabuta-
diene ligand in 9 are indicative of a similar degree of delocalis-
ation as possessed by 6. There appears to be less delocalisation
over the Ar-DAB ligand in the cation of 10 compared to 9,
despite their structural similarities. This is evidenced by the
shorter N–C and longer C–C and Al–N distances in the former
which is consistent with its formulation as a diamagnetic
Ar-DAB adduct of the AlI2ϩ cation.
Fig. 3 Molecular structure of compound 8. Selected bond lengths (Å)
and angles (Њ): I(1)–Ga(1) 2.5312(7), Ga(1)–N(2) 1.896(6), Ga(1)–N(1)
1.965(6), N(1)–C(1) 1.316(10), N(2)–C(2) 1.357(10), C(1)–C(2)
1.450(11); N(1)–Ga(1)–N(2) 87.1(3), N(2)–Ga(1)–I(1) 115.89(8), N(1)–
Ga(1)–I(1) 113.66(9), I(1)–Ga(1)–I(1Ј) 109.33(4), C(1)–N(1)–Ga(1)
108.3(5), C(2)–N(2)–Ga(1) 110.8(5), N(1)–C(1)–C(2) 118.6(7), N(2)–
C(2)–C(1) 115.2(7). Symmetery operation: Ј x, Ϫy ϩ 3/2, z.
gallium compound, 7, is closely related to 5 and as in that com-
pound the bond lengths within the diazabutadiene framework
show the ligand to be delocalised. The gallium centres possess
distorted tetrahedral geometries and the N–Ga bond lengths
and the N–Ga–N angle are close to those in 6. Similarly, the
Ga–Ga distance of 2.4232(7) Å compares well with the metal–
metal interaction in related diazabutadiene complexes, e.g. 2
[2.333(1) Å]. Compound 8 is similar to 6 in that it is monomeric
Compounds 5, 6 and 9 seemed to be ideally suited as pre-
cursors to the Group 13 valence isoelectronic NHC analogues,
[:M(Ar-DAB)]Ϫ, M = In, Ga or Al, via reduction reactions. To
3846
J. Chem. Soc., Dalton Trans., 2002, 3844–3850

View Article Online
date the only known example of such an anion is 3 which is
prepared in a multi-step synthesis. This involves the reaction
of [Li2(But-DAB)] with GaCl3 to give [GaIIICl2{µ-GaIII(But-
DAB)2}] which is reduced over 10 days to give the gallium()
complex, 2. This is further reduced with potassium metal over
10 days to give 3 or 4 depending on the presence of either 18-
crown-6 or TMEDA. It is worth mentioning that two closely
related neutral aluminium and gallium() carbene analogues,
[:M[{N(Ar)CMe}2CH]], M = Al15 or Ga,16 have been recently
reported and their chemistry has begun to be explored.17
Unfortunately the reduction of either 5 or 9 with an excess of
potassium metal in THF did not lead to the M() carbene ana-
logues but instead to decomposition products which included
the free Ar-DAB ligand and elemental Group 13 metal. The
analogous reduction of 6 was, however, more successful. After
8 hours reaction time at room temperature and subsequent
recrystallisation from diethyl ether the new gallium() carbene
analogue, 11, was formed in good yield (Scheme 2). In a similar
Scheme 2 Reagents and conditions: i, K, THF; ii, Et2O or TMEDA/
Et2O; iii, 18-crown-6.
fashion the reduction of the dimeric gallium() compound, 7,
with an excess of potassium led to a high yield of the known
complex, 4, after recrystallisation from a diethyl ether/TMEDA
mixture.
The spectroscopic data for 11 are consistent with its proposed
structure and are indicative of a 1 : 1 ratio of coordinated ether
to gallium heterocycle. As in the case of 3 the 1H and 13C reson-
ances for the olefinic fragment of the Ar-DAB backbone are
significantly shifted to higher field relative to those for the free
ligand, as would be expected in a reduced anionic system.
The molecular structure of 11 is depicted in Fig. 6 (Table 2).
This shows the molecule to be dimeric and to sit on a centre of
inversion. It can be considered as consisting of monomeric
units which comprise a gallium “carbene” heterocycle η5-
coordinated to a K(Et2O) fragment. Dimerisation of these
fragments occurs via intermolecular interactions of gallium
lone pairs with two potassium centres. At first glance this seems
to be a very similar arrangement to that seen for 4. The
geometries of the essentially planar gallium heterocycles in 4
and 11 are similar and close to that predicted by theoretical
studies for the model heterocycle, [:GaN(H)C(H)C(H)-
N(H)]Ϫ.18 In addition, the Ga K distances in 4 [3.438(1) Å]
and 11 [3.4223(10) Å], and the η5-interactions from the hetero-
cycles to the gallium centres are comparable [Ga–K 3.4681(5) Å
4, 3.3784(13) Å 11; K–N avge. 2.889 Å 4, 2.962 Å 11; K–C avge.
2.999 Å 4, 2.997 Å 11]. Despite these similarities there are also
significant differences between 4 and 11. For example, in 4 the
Ga K bond forms an angle of 20.8Њ with the C2N2Ga ring
plane whereas in 11 this angle is only 3.4Њ. This in turn means
J. Chem. Soc., Dalton Trans., 2002, 3844–3850
3847

View Article Online
Fig. 7 Molecular structure of compound 12. Selected bond lengths
(Å) and angles (Њ): Ga(1)–Ga(1Ј) 2.8746(15), Ga(1)–N(1) 2.014(4),
Ga(1)–N(2) 2.011(4), Ga(1)–K(1Ј) 3.4620(16), Ga(1)–K(1) 3.5318(18),
K(1)–N(3) 2.996(6), K(1)–N(4) 3.055(7), K(1)–N(1Ј) 3.027(4), K(1)–
C(1Ј) 3.009(5), K(1)–N(2Ј) 3.034(4), K(1)–C(2Ј) 3.016(5), K(1)–Ga(1Ј)
3.4620(16), C(1)–C(2) 1.350(6); N(1)–Ga(1)–N(2) 82.05(15), N(1)–
Ga(1)–Ga(1Ј) 109.79(12), N(2)–Ga(1)–Ga(1Ј) 109.79(12), N(1)–Ga(1)–
K(1) 138.84(10), N(2)–Ga(1)–K(1) 139.09(11), Ga(1)–N(1)–C(1)
110.4(3), C(2)–N(2)–Ga(1) 110.4(3), N(1)–C(1)–C(2) 117.2(4), N(2)–
C(2)–C(1) 118.0(4). Symmetery operation: Ј Ϫx, Ϫy ϩ 1, Ϫz.
Fig. 6 Molecular structure of compound 11. Selected bond lengths
(Å) and angles (Њ): Ga(1)–Ga(1Ј) 2.8640(13), Ga(1)–N(1) 2.005(2),
Ga(1)–N(2) 2.009(2), Ga(1)–K(1Ј) 3.3784(13), Ga(1)–K(1) 3.4223(10),
K(1)–O(1) 2.718(2), K(1)–N(1Ј) 2.936(3), K(1)–C(1Ј) 2.977(3), K(1)–
N(2Ј) 2.987(2), K(1)–C(2Ј) 2.997(3), K(1)–Ga(1Ј) 3.3784(13), C(1)–C(2)
1.356(4); N(1)–Ga(1)–N(2) 81.88(9), N(1)–Ga(1)–Ga(1Ј) 107.95(7),
N(2)–Ga(1)–Ga(1Ј) 109.68(7), N(1)–Ga(1)–K(1) 138.96(6), N(2)–
Ga(1)–K(1) 139.15(6), Ga(1)–N(1)–C(1) 111.05(17), C(2)–N(2)–Ga(1)
110.65(16), N(1)–C(1)–C(2) 116.8(2), N(2)–C(2)–C(1) 117.3(2).
Symmetry operation: Ј Ϫx, Ϫy, Ϫz ϩ 1.
In order to test whether a Ga ؒ ؒ ؒ Ga interaction would
remain between two of the gallium heterocycles in the absence
of coordinated potassium centres, compound 11 was treated
with an excess of 18-crown-6 in THF and the crude product
recrystallised from diethyl ether to give a good yield of 13
(Scheme 2). Again, the spectroscopic data for the gallium
heterocycle are very similar to those in 11 and 12 but it could
not be ascertained whether there was any heterocycle associ-
ation from this data alone. As a result an X-ray crystal structure
analysis of the compound was carried out and its molecular
structure is shown in Fig. 8 (Table 2). The asymmetric unit
contains two crystallographically independent [:Ga(Ar-DAB)]Ϫ
anions and two K(18-crown-6)1.5 cations, the latter of which are
generated into dicationic units, [(18-crown-6)K(µ-18-crown-
6)K(18-crown-6)]2ϩ, by inversion centres. One of these is shown
in Fig. 8(b) and it is clear that the terminal 18-crown-6 ligands
are η6-coordinated to a potassium centre whilst the bridging
crown is η2-coordinated to both K centres. There are no con-
tacts between [:Ga(Ar-DAB)]Ϫ anions which strongly suggests
that the Ga ؒ ؒ ؒ Ga interactions in 11 and 12 are weak and will
not persist in the absence of the coulombic assistance provided
by the partially solvated potassium ions. The GaN2C2 five
membered rings of both [:Ga(Ar-DAB)]Ϫ anions are position-
ally disordered over two sites in an 80 : 20 ratio. This disorder
was successfully modelled and the geometries of the two major
occupancy sets are similar so only one is shown in Fig. 8(a).
Although, the [:Ga(Ar-DAB)]Ϫ units do not have an interaction
with each other there appears to be a weak contact between one
methyl group on each heterocycle and a potassium centre,
e.g. C(11)–K(1) 3.432 Å. The protons on this carbon will very
likely have a closer interaction but they could only be included
in calculated positions so comment on such interactions is not
valid. Although geometrical constraints were used to model
the disorder in each [:Ga(Ar-DAB)]Ϫ anion the bond lengths
and angles within the heterocycle are similar to those observed
in the only other uncoordinated gallium carbene analogue,
3, and as in that anion the Ga centres are undoubtedly 2-
coordinate.
that the distance between the Ga centres in 4 is 4.21 Å and the
heterocycle centroid–Ga–Ga angle is 106.0Њ whereas in 11 the
same angle is 119.2Њ and the Ga–Ga distance is only 2.8640(13)
Å. Although the difference in these Ga–Ga distances is almost
1.4 Å the shorter interaction is still longer than normally seen
for Ga–Ga single bonds, e.g. 2.541(1) Å in [(tmp)2Ga–Ga-
(tmp)2], tmp = 2,2,6,6-tetramethylpiperidine,19 but not markedly
so, especially considering the expected larger covalent radius of
the Ga() centre in 11 relative to the Ga() centres in [(tmp)2-
Ga–Ga(tmp)2] and other related dimers. This could mean that
in 11 the two anionic gallium centres are held together, not only
by an electrostatic attraction to the potassium cations but per-
haps also by a partial interaction of electron density from the
lone pair on each gallium centre with the p-orbital on the other.
If this were the case it is only a weak interaction and it is not
known why it does not occur in 4, which should be more open to
an approach of the two gallium centres, at least on steric grounds.
It was thought of sufficient worth to attempt the replacement
of the coordinated Et2O ligand of 11 with the bidentate
TMEDA ligand to form the direct analogue of 4, i.e. 12, for
purposes of comparison. This was readily achieved and 12
was obtained in good yield after recrystallisation from diethyl
ether (Scheme 2). The spectroscopic data for 12 are very similar
to those for 11 with the exception that they suggest a 1 : 1
ratio of TMEDA to each potassium centre, which would
indicate chelation. The X-ray crystal structure of 12 (Fig. 7,
Table 2) confirmed this and shows that the geometry of the
dimeric core of the molecule is very similar to that of 11
[Ga–Ga 2.8746(15) Å; Ga(1)–K(1) 3.5318(18) Å; Ga(1)–K(1Ј)
3.4620(16) Å; K(1Ј)–C avge. 3.013 Å; K(1Ј)–N avge. 3.030 Å
avge.]. This observation confirms that the shortness of the
Ga ؒ ؒ ؒ Ga interaction in 11 compared to 4 is not due to the
potassium centre in the former being less electronically satisfied
by the monodentate ether ligand than the bidentate TMEDA
ligand in the latter.
3848
J. Chem. Soc., Dalton Trans., 2002, 3844–3850

View Article Online
for 7–9 could not be obtained due to their paramagnetic nature.
A useful 13C NMR spectrum of 10 could not be obtained due
to the fluxional nature of the compound in solution. The EPR
spectra were recorded at room temperature on an X-band
Bruker ESP 300e series spectrometer operating at 12.5 kHz
field modulation in a Bruker EN801 cavity. The spectra were
obtained with a 2.5 mW power. Solutions of the complexes
were prepared in CD2Cl2/C7D8. The g values were obtained
using a Bruker ER035 gaussmeter calibrated using the perylene
radical cation in concentrated H2SO4 (g = 2.002569). Computer
simulations were performed using the SIMFONIA Bruker
software.20 Mass spectra were recorded using a VG Fisons
Platform II instrument under APCI or EI conditions. Melting
points were determined in sealed glass capillaries under argon
and are uncorrected. Microanalyses were obtained from the
Warwick Microanalytical Service. Where reproducible micro-
analyses could not be obtained the compound was either highly
air sensitive, contained solvent of crystallisation or could not be
separated from trace by-products. In all cases, however, all data
pointed to the bulk materials having a purity of > 95%. The
starting materials Ar-DAB21 and But-DAB22 were prepared by
literature procedures and 6 was prepared by a variation of the
published method.8 All other reagents were used as received.
[{IGa(But-DAB)}2] 7 and [I2Ga(But-DAB)] 8
To a suspension of Ga metal (0.24 g, 3.4 mmol) in toluene
(15 ml) was added I2 (0.43 g, 1.7 mmol) at 25 ЊC. The mixture
was sonicated until the solution had become colourless and a
pale green precipitate was produced. To this suspension was
added a solution of But-DAB (0.57 g, 3.4 mmol) in toluene
(15 ml) and the resulting suspension stirred for 4 hours to yield
a red solution. This was filtered, concentrated to ca. 10ml and
placed at Ϫ30 ЊC overnight to yield a mixture of 7 as dark green
crystals (0.37 g, 32%) and 8 as red crystals (0.06 g, 4% based on
But-DAB) which were manually separated.
Fig. 8 Molecular structure of (a) the anion and (b) the dication of
compound 13. Selected bond lengths (Å) and angles (Њ): Ga(1)–N(1)
1.983(3), Ga(1)–N(2) 1.956(3), N(1)–C(1) 1.423(6), N(2)–C(2) 1.395(5),
C(1)–C(2) 1.374(6), K(1)–O(1) 2.875(3), K(1)–O(2) 2.871(3), K(1)–O(3)
2.807(3), K(1)–O(4) 2.957(3), K(1)–O(5) 2.859(3), K(1)–O(6) 2.850(3),
K(1)–O(13) 2.932(3), K(1)–O(14) 2.816(3), N(1)–Ga(1)-N(2) 83.02(11);
Ga(1)–N(1)–C(1) 110.2(4), Ga(1)–N(2)–C(2) 114.7(3), N(1)–C(1)–C(2)
118.5(6), N(2)–C(2)–C(1) 113.6(6).
Data for 7: mp 126–131 ЊC; MS APCI: m/z (%) 492 (Ar-
ϩ
DABGaI2 , 27), 169 (Ar-DABHϩ, 100); IR (Nujol) ν/cmϪ1 1634
(m), 1527(s), 1260(s), 1204(m), 779(m); C20H40Ga2I2N4 requires
C 32.97, H 5.52, N 7.68%; found: C 33.43, H 5.64, N 7.74%.
Data for 8: mp 85–92 ЊC; MS APCI: m/z (%) 492 (Mϩ, 5),
169 (Ar-DABHϩ, 100); IR (Nujol) ν/cmϪ1 1634(m), 1265(s),
1199(m), 1015(m), 799(m); C10H20N2GaI2 requires C 24.42,
H 4.10, N 5.70%; found: C 24.66, H 4.14, N 5.48%.
Conclusions
In conclusion we have developed a high yielding synthetic route
to the monomeric, paramagnetic Group 13 diazabutadiene
complexes, [I2MIII(Ar-DAB)], M = Ga 6, Al 9, the former of
which has very recently been independently reported by
another group.8 In addition, the syntheses and characterisations
of the new Ga(), Ga() and Al() complexes, [{IGaII-
(But-DAB)}2] 7, [I2GaIII(But-DAB)] 8 and [I2AlIII(Ar-DAB)]I 10
have been described. The reduction of 6 comprises a new and
facile synthetic route to Ga() NHC analogues, only the second
example of which is reported as three potassium salts in which
the alkali metal is coordinated by Et2O 11, TMEDA 12 or 18-
crown-6 13. Compounds 11 and 12 exhibit Ga ؒ ؒ ؒ Ga inter-
actions which were not observed in the only other reported
Ga() NHC analogue. We are currently developing a range of
other mono- and multi-dentate Group 13 NHC analogues and
investigating their coordination chemistry. The results of these
investigations will be reported in forthcoming publications.
[I2Al(Ar-DAB)] 9
To a suspension of Al powder (0.073 g, 2.7 mmol) in toluene
(15 ml) was added I2 (0.342 g, 0.135 mmol) at 25 ЊC. The
resulting suspension was sonicated for 4 hours to yield an AlI3/
Al mixture. A solution of Ar-DAB (1.0 g, 2.7 mmol) in toluene
(15 ml) was added to this over 5 min at 25 ЊC to give an immedi-
ate deep red colour to the suspension. This was stirred over-
night during which time a proportion of the Al powder was
seen to dissolve. Volatiles were removed in vacuo and the residue
extracted into Et2O (20 ml). Cooling to Ϫ30 ЊC yielded red
crystals of 9 overnight (0.70 g, 40% based upon Ar-DAB),
mp 224–227 ЊC; MS APCI: m/z (%) 377 (Ar-DABHϩ, 100),
531 (Mϩ Ϫ I, 61); IR (Nujol) ν/cmϪ1 1516(s), 1219(s), 1055(s),
886(m), 789(m); C26H36N2AlI2 requires C 47.50, H 5.52, N
4.26%; found: C 47.16, H 5.57, N 4.10%.
Experimental
General remarks
[I2Al(Ar-DAB)]I 10
All manipulations were carried out using standard Schlenk and
glove box techniques under an atmosphere of high purity argon
or dinitrogen. The solvents diethyl ether, toluene and THF
were distilled over either potassium or Na/K alloy then freeze/
To a solution of AlI3 (0.54 g, 1.33 mmol) in toluene (15 ml) was
added a solution of Ar-DAB (0.50 g, 1.33 mmol) in toluene
(15 ml) over 5 min at 25 ЊC. The solution immediately became a
deep red colour and was stirred for 5 hours whereupon volatiles
were removed in vacuo and the residue extracted into Et2O
(20 ml) and the resulting solution placed at Ϫ30 ЊC to yield red
crystals of 10 overnight (0.69 g, 66%), mp 163–176 ЊC
(decomp.); 1H NMR (400 MHz, C6D6, 300 K) δ 0.95 (br, 24H,
1
thaw degassed prior to use. H and 13C NMR spectra were
recorded on Bruker DPX400 or Jeol Eclipse 300 spectrometers
in deuterated solvents and were referenced to the residual 1H or
13C resonances of the solvent used. Meaningful NMR spectra
J. Chem. Soc., Dalton Trans., 2002, 3844–3850
3849

View Article Online
CH3), 2.98 (br, 4H, CH), 7.02 (br, 6H, ArH), 8.05 (br, 2H,
NCH); MS EI: m/z (%) 332 (Ar-DAB Ϫ Priϩ, 100), 531 (Mϩ Ϫ
2I, 10); IR (Nujol) ν/cmϪ1 1532(s), 1276(s), 1255(m), 1107(s),
799(m).
Structure determinations
Crystals of 7–13 suitable for X-ray structure determination
were mounted in silicone oil. Crystallographic measurements
were made using a Nonius Kappa CCD diffractometer. The
structures were solved by direct methods and refined on F 2 by
full matrix least squares (SHELX97)23 using all unique data.
Crystal data, details of data collections and refinements are
given in Table 1. The molecular structures of the complexes are
depicted in Fig. 2–8 and show ellipsoids at the 30% probability
level.
[{(Et2O)KGa(Ar-DAB)}2] 11
A solution of 6 (0.75 g, 1.07 mmol) in THF (15 ml) was stirred
over a potassium mirror (0.80 g, 20 mmol) at 25 ЊC for 8 hours
after which the yellow-orange solution was filtered and volatiles
removed in vacuo. The residue was extracted into Et2O (20 ml)
and the resulting solution was concentrated to ca. 10 ml and
cooled to Ϫ30 ЊC overnight to yield 11 as red crystals (0.45 g,
CCDC reference numbers 189504–189510.
See http://www.rsc.org/suppdata/dt/b2/b206605j/ for crystal-
lographic data in CIF or other electronic format.
1
75%), mp 180–187 ЊC (decomp.); H NMR (400 MHz, C6D6,
300 K) δ 1.28 (t, 12H, 3JHH = 6.4 Hz, CH3), 1.46 (d, 48H, 3JHH
=
6.8 Hz, CH3), 3.44 (q, 8H, 3JHH = 6.4 Hz, CH2), 3.83 (sept., 8H,
3JHH = 6.8 Hz, CH), 6.48 (s, 4H, NCH), 7.23 (t, 4H, 3JHH = 7.4
Hz, p-Ar), 7.37 (d, 8H, 3JHH = 7.4 Hz, m-Ar); 13C NMR (100.6
MHz, C6D6, 300 K) δ 15.3 (CH2CH3), 25.3 (CHCH3), 28.0
(CH), 65.6 (CH2), 122.5 (CN), 122.8 (m-ArC), 123.7 (p-ArC),
145.8 (o-ArC), 150.2 (ipso-ArC); MS EI: m/z (%) 377 [DABϩ,
5], 189 [ArNCHϩ, 100]; IR (Nujol) ν/cmϪ1 1583(s), 1557(m),
1255(s), 1096(s), 927(m).
Acknowledgements
We gratefully acknowledge financial support from the EPSRC
for personnel (studentship for M. K. and postdoctoral fellow-
ship for R. J. B.) and for the National Service for Electron
Paramagnetic Resonance (GR/R17980). Thanks also go to
Dr Peter Junk of Monash University, Melbourne for helpful
discussions regarding the crystallographic disorder in 13.
References and notes
[{(TMEDA)KGa(Ar-DAB)}2] 12
1 (a) F. G. N. Cloke, G. R. Hanson, M. J. Henderson, P. B. Hitchcock
and C. L. Raston, J. Chem. Soc., Chem. Commun., 1989,
1002; (b) T. Pott, P. Jutzi, B. Neumann and H. G. Stammler,
Organometallics, 2001, 20, 1965.
2 D. S. Brown, A. Decken and A. H. Cowley, J. Am. Chem. Soc., 1995,
117, 5421.
3 E. S. Schmidt, A. Jockisch and H. Schmidbaur, J. Am. Chem. Soc.,
1999, 121, 9578.
4 E. S. Schmidt, A. Schier and H. Schmidbaur, J. Chem. Soc., Dalton
Trans., 2001, 505.
5 C. Jones, Chem. Commun., 2001, 2293.
6 R. J. Baker, R. D. Farley, C. Jones, M. Kloth and D. M. Murphy,
Chem. Commun., 2002, 1196.
7 M. L. H. Green, P. Mountford, G. J. Smout and S. R. Peel,
Polyhedron, 1990, 9, 2763.
8 T. Pott, P. Jutzi, W. Kaim, W. W. Schoeller, B. Neumann,
A. Stammler, H. G. Stammler and M. Wanner, Organometallics,
2002, 21, 3169.
9 R. J. Baker, C. Jones and M. Kloth, unpublished results.
10 J. A. C. Clyburne, R. D. Culp, S. Kamepalli, A. H. Cowley and
A. Decken, Inorg. Chem., 1996, 35, 6651.
11 R. J. Baker, A. J. Davies, C. Jones and M. Kloth, J. Organomet.
Chem., 2002, 656, 203.
12 F. S. Mair, R. Manning, R. G. Pritchard and J. E. Warren, Chem.
Commun., 2001, 1136.
13 The experimental EPR spectra of 7 and 9 are included as ESI†.
14 J. Emsley, The Elements, Clarendon Press, Oxford, 2nd edn., 1991.
15 C. Cui, H. W. Roesky, H. G. Schmidt, M. Noltemeyer, H. Hao and
F. Cimpoesu, Angew. Chem., Int. Ed., 2000, 39, 4274.
16 N. J. Hardman, B. E. Eichler and P. P. Power, Chem. Commun., 2000,
1991.
17 See for example: N. J. Hardman, P. P. Power, J. D. Gorden,
C. L. B. Macdonald and A. H. Cowley, Chem. Commun., 2001, 1866.
18 A. Sundermann, M. Reiher and W. W. Scoeller, Eur. J. Inorg. Chem.,
1998, 305.
To a solution of 11 (0.40 g, 0.76 mmol) in Et2O (25 ml) at
Ϫ78 ЊC was added TMEDA (0.40 g, 3.5 mmol) and the result-
ing solution warmed to room temperature and stirred for
4 hours after which time volatiles were removed in vacuo and
the residue extracted into Et2O (10 ml). The resulting solution
was cooled to Ϫ30 ЊC overnight to yield 12 as red crystals
(0.31 g, 68%), mp 163–166 ЊC (decomp.); 1H NMR (400 MHz,
C6D6, 300 K) δ 1.41 (d, 24H, 3JHH = 6.9 Hz, CH3), 1.43 (d, 24H,
3JHH = 6.9 Hz, CH3), 2.05 (2, 24H, NCH3), 2.20 (s, 8H, NCH2),
3
3.82 (sept., 8H, JHH = 6.9 Hz, CH), 6.42 (s, 4H, NCH), 7.23
(t, 4H, 3JHH = 7.5 Hz, p-Ar), 7.32 (d, 8H, 3JHH = 7.5 Hz, m-Ar);
13C NMR (100.6 MHz, C6D6, 300 K) δ 26.0 (CHCH3), 28.7
(CH), 46.2 (NCH3), 58.2 (NCH2), 123.1 (CN), 123.4 (m-ArC),
124.3 (p-ArC), 146.3 (o-ArC), 151.0 (ipso-ArC); MS EI: m/z (%)
115 [TMEDAHϩ, 100], 189 [ArCNHϩ, 3]; IR (Nujol) ν/cmϪ1
1588(s), 1562(m), 1245(s), 1096(s) 1081(m).
[:Ga(Ar-DAB)]2[{K(18-crown-6)}2(ꢀ-18-crown-6)] 13
To a solution of 11 (0.29 g, 0.26 mmol) in THF (15 ml) at
Ϫ50 ЊC was added a solution of 18-crown-6 (0.30 g, 1.14 mmol)
in THF (15 ml). The resulting orange solution was allowed
to warm to 25 ЊC and was stirred for 2 hours whereupon
volatiles were removed in vacuo. The residue was extracted into
Et2O (10 ml) and the resulting solution cooled to Ϫ30 ЊC to
1
yield 13 as orange needles (0.27 g, 59%), mp 134–139 ЊC; H
3
NMR (400 MHz, C6D6, 300 K) δ 1.58 (d, 24H, JHH = 6.9 Hz,
CH3), 1.64 (d, 24H, 3JHH = 6.9 Hz, CH3), 3.27 (s, 72H, OCH2),
3
4.45 (sept., 8H, JHH = 6.9 Hz, CH), 6.82 (s, 4H, NCH), 7.28
(t, 4H, 3JHH = 7.5 Hz, p-Ar), 7.43 (d, 8H, 3JHH = 7.5 Hz, m-Ar);
13C NMR (100.6 MHz, C6D6, 300 K) δ 26.0 (CH3), 26.7 (CH3),
28.6 (CH), 70.7 (CH2), 122.2 (CN), 122.4 (m-ArC), 122.7 (p-
ArC), 146.5 (o-ArC), 152.6 (ipso-ArC); MS EI: m/z (%) 443
[Ga(Ar-DAB)ϩ 17], 377 [DABϩ, 31], 265 [18-crown-6ϩ, 100]; IR
(Nujol) ν/cmϪ1 1573(s), 1552 (m), 1255 (s), 1112(s) 794(s);
C88H144O18N4Ga2K2 requires C 59.93, H 8.23, N 3.18%; found:
C 58.92, H 8.25, N 3.10%.
19 G. Linti, R. Frey and M. Schmidt, Z. Naturforsch., Teil B, 1994, 49,
958.
20 WINEPR SIMFONIA Version 1.25, Bruker Analytische
Messtechnick GmbH, 1996.
21 L. Jafarpour, E. D. Stevens and S. P. Nolan, J. Organomet. Chem.,
2000, 606, 49.
22 J. M. Kleigman and R. K. Barnes, Tetrahedron, 1970, 26, 2555.
23 G. M. Sheldrick, SHELXL-97, University of Göttingen, 1997.
3850
J. Chem. Soc., Dalton Trans., 2002, 3844–3850