Synthesis and structural characterization of β-ketoiminate-stabilized gallium hydrides for chemical vapor deposition applications

Article abstract of DOI:10.1002/chem.201402998

Bis-β-ketoimine ligands of the form [(CH₂) _n{N(H)C(Me)-CHC(Me)=O}₂] (LⁿH₂, n=2, 3 and 4) were employed in the formation of a range of gallium complexes [Ga(Lⁿ)X] (X=Cl, Me, H), which were characterised by NMR spectroscopy, mass spectrometry and single-crystal X-ray diffraction analysis. The β-ketoimine ligands have also been used for the stabilisation of rare gallium hydride species [Ga(Lⁿ)H] (n=2 (7); n=3 (8)), which have been structurally characterised for the first time, confirming the formation of five-coordinate, monomeric species. The stability of these hydrides has been probed through thermal analysis, revealing stability at temperatures in excess of 200C. The efficacy of all the gallium β-ketoiminate complexes as molecular precursors for the deposition of gallium oxide thin films by chemical vapour deposition (CVD) has been investigated through thermogravimetric analysis and deposition studies, with the best results being found for a bimetallic gallium methyl complex [L³{GaMe₂}₂] (5) and the hydride [Ga(L³)H] (8). The resulting films (F5 and F8, respectively) were amorphous as-deposited and thus were characterised primarily by XPS, EDXA and SEM techniques, which showed the formation of stoichiometric (F5) and oxygen-deficient (F8) Ga₂O₃ thin films.

Full text of DOI:10.1002/chem.201402998

DOI: 10.1002/chem.201402998
Full Paper
&
Chemical Vapor Deposition
Synthesis and Structural Characterization of b-Ketoiminate-
Stabilized Gallium Hydrides for Chemical Vapor Deposition
Applications
[a]
[b]
[a]
[a]
Peter Marchand, David Pugh, Ivan P. Parkin, and Claire J. Carmalt*
Abstract:
Bis-b-ketoimine
ligands
of
the
form
of all the gallium b-ketoiminate complexes as molecular pre-
cursors for the deposition of gallium oxide thin films by
chemical vapour deposition (CVD) has been investigated
through thermogravimetric analysis and deposition studies,
with the best results being found for a bimetallic gallium
n
[
(CH ) {N(H)C(Me)ÀCHC(Me)=O} ] (L H , n=2, 3 and 4) were
2
n
2
2
employed in the formation of a range of gallium complexes
n
[Ga(L )X] (X=Cl, Me, H), which were characterised by NMR
spectroscopy, mass spectrometry and single-crystal X-ray dif-
3
3
fraction analysis. The b-ketoimine ligands have also been
methyl complex [L {GaMe } ] (5) and the hydride [Ga(L )H]
2 2
used for the stabilisation of rare gallium hydride species
(8). The resulting films (F5 and F8, respectively) were amor-
phous as-deposited and thus were characterised primarily
by XPS, EDXA and SEM techniques, which showed the for-
mation of stoichiometric (F5) and oxygen-deficient (F8)
Ga O thin films.
n
[Ga(L )H] (n=2 (7); n=3 (8)), which have been structurally
characterised for the first time, confirming the formation of
five-coordinate, monomeric species. The stability of these
hydrides has been probed through thermal analysis, reveal-
ing stability at temperatures in excess of 2008C. The efficacy
2
3
Introduction
terials highly desirable for use as solid electrolytes in solid
[6]
oxide fuel cells (SOFCs). In addition, gallium oxide thin films
have been used as white-light emitting phosphors for use in
Gallium oxide has found use in a diverse range of industrially
and technologically important applications. There are five
known crystallographic phases of Ga O , with the most ther-
[7]
light-emitting devices (LEDs), as well as a potential deep-UV
[8]
transparent conducting oxide. Further application includes
use within catalytic zeolite systems to enhance the conversion
2
3
modynamically stable being monoclinic b-Ga O on account of
2
3
[
1]
[9]
its low density. b-Ga O is a wide band-gap material (4.2–
of lower alkanes into aromatic molecules.
2
3
[
2]
4
.9 eV), which is electrically insulating at room temperature,
Given the range of applications of thin films of gallium
oxide, it is no surprise that a number of techniques have been
explored for the deposition of the material. Common fabrica-
however displays n-type semiconductivity at elevated tempera-
[
3]
tures.
At temperatures above 5008C, Ga O acts as a sensor for re-
[
10]
[11]
tion techniques such as sputtering,
sol-gel processes,
2
4]
3
[
[12]
[13]
[14]
ducing gases such as CO and H2, whereas at temperatures
spray pyrolysis, screen printing, atomic layer deposition,
above 9008C it can be used as a catalytically inactive oxygen
as well as chemical vapour deposition (CVD) processes includ-
[
3]
[15]
[16]
sensor. More recently, Ga O nanowires have been found to
ing low-pressure (LP), atmospheric-pressure (AP) and aero-
2
3
[17]
show a reversible response to O and CO at much lower work-
sol-assisted (AA) CVD, have all been reported for the produc-
tion of thin films of Ga O . Amongst the many available tech-
2
[
5]
ing temperatures in the range of 100–5008C. Lanthanum gal-
2
3
late (LaGaO ) doped with various metals was observed to have
niques for the deposition of Group 13 metal oxides, CVD is fa-
vourable due to the production of high quality, uniform films
that are largely free from contamination.
3
superior oxygen-ion-conductivities in comparison to other ma-
terials such as yttria-stabilised zirconia (YSZ), making these ma-
AACVD relies on the solubility of the chosen precursor
rather than depending on its volatility, thus broadening the
range of potential precursors that can be employed. Further-
more, the solution-based process offers an additional means
with which to control film morphology and hence its proper-
ties, and is therefore of great interest for the deposition of ma-
[
a] Dr. P. Marchand, Prof. I. P. Parkin, Prof. C. J. Carmalt
Materials Chemistry Centre, Department of Chemistry
University College London
2
0 Gordon Street, London, WC1H 0AJ (UK)
Fax: (+44)020-7679-7463
E-mail: c.j.carmalt@ucl.ac.uk
[18]
terials.
[b] Dr. D. Pugh
Interest in the use of single-source precursors for the CVD of
main group materials has been driven by their potential for
the repeatable deposition of thin films with the desired stoi-
Department of Chemistry, University of Southampton
Highfield, Southampton, SO17 1BJ (UK)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201402998.
[19]
chiometries and good homogeneity. This has prompted a sig-
Chem. Eur. J. 2014, 20, 10503 – 10513
10503
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
nificant amount of research into the development of novel
precursor systems, in particular those containing pre-existing
MÀO bonds, for the deposition of metal oxides. To date, the
field of single-source precursors to gallium oxide thin films is
largely dominated by mono- and donor-functionalised alkox-
Results and Discussion
Synthesis
The synthesis of chloro-gallium b-ketoiminate complexes
[
20]
ides and b-diketonates. More recently, b-ketoimines have re-
ceived some interest in this regard, mainly due to the potential
for additional functionalisation of the imino residue allowing
for the formation of monomeric
1 and 2 has been previously reported, and was achieved
through deprotonation of the pro-ligands L H (n=2 and 3;
n
2
Scheme 1) with NaH followed by salt metathesis with
species with enhanced thermal
stabilities and solubilities.
Furthermore, gallium hydrides
are an attractive class of com-
pound for application as molec-
ular precursors, with advantages
including
i) low
molecular
masses, ii) clean decomposition
pathways due to the weakness
of the GaÀH bond, iii) ready MÀ
H bond cleavage with respect to
MÀC and MÀCl bonds and
iv) low levels of carbon contami-
nation within deposited films as
a result of the absence of direct
metalÀcarbon bonds within the
[
21]
precursor. Although their ther-
mal instability has limited the
use of gallium hydrides as mo-
lecular precursors, progress in
the last 25 years has led to the
isolation of some systems that
exhibit considerable robustness.
In particular, gallane adducts sta-
bilised by N-heterocyclic car-
benes have been shown to have
thermal stabilities as high as
Scheme 1. The synthesis of gallium b-ketoiminate derivatives 1–6.
[
22]
[24–25]
2148C. Examples of mixed gallium hydride species contain-
GaCl .
The tetramethylene derivative of the pro-ligand,
3
4
ing MÀO bonds also exist in the form of mono- and donor-
L H , was synthesised by condensation of acetylacetone with
2
functionalised alkoxides, with the latter being successfully uti-
1,4-diaminobutane, which was subsequently converted into
the corresponding gallium complex 3 through the same salt
metathesis pathway (Scheme 1). Colourless, crystalline blocks
that were suitable for analysis by single-crystal X-ray diffraction
were formed by slow solvent diffusion of hexane into a concen-
[
23]
lised in the deposition of Ga O thin films. Recent studies
2
3
have demonstrated the use of bis-b-ketoimines as suitable
ligand systems for the formation and isolation of monomeric
gallium species, with the notable formation of the b-ketoimi-
2
[24]
nate-stabilised gallium hydride complex [Ga(L )H] (7).
Al-
trated CH Cl2 solution of the crude product. The molecular
2
though the application of the hydride as a molecular precursor
to Ga O by AACVD was reported, structural characterisation
structure of 3 is shown in Figure 1 and selected bond lengths
and angles are given in Table 1. The monomeric complex 3
crystallised as a five-coordinate species in the triclinic P-
1 space group, with the gallium cation in a slightly distorted
2
3
was not obtained and no systematic study of the potential of
the bis-b-ketoiminate systems as precursors was carried out.
Herein, we report the synthesis and structural characterisation
of a range of gallium bis-b-ketoiminates with chloride, methyl
and hydride co-ligands. Thermal analysis and deposition stud-
ies of the gallium b-ketoiminate precursors has been used to
investigate their efficacy towards the formation of Ga O by
[26]
trigonal bipyramidal geometry (t=0.86).
Chloride, as the
largest ligand, predictably occupies one of the equatorial sites,
with the other two sites being occupied by one oxygen and
one nitrogen of the chelating b-ketoiminate ligand. The axial
positions are occupied by the remaining donor atoms of the
b-ketoiminate ligand, with the O(1)-Ga-N(2) angle bending
away from linearity to 175.34(6)8. Bond lengths within the
complex are consistent with those found within the propylene
2
3
AACVD.
[24]
analogue 2. As expected, bond lengths within the ketoimine
Chem. Eur. J. 2014, 20, 10503 – 10513
www.chemeurj.org
10504 ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Whereas salt metathesis provides a clean route to the forma-
tion of a range of gallium b-ketoiminate complexes, an explo-
ration of a range of alternative co-ligands was of interest with
a view to their application as molecular precursors to Ga O
2
3
through CVD. Compounds 4–6 were synthesised by the meth-
n
ane elimination reaction of GaMe with pro-ligands L H (n=2,
3
2
3
and 4), according to Scheme 1. In all cases, the evolution of
gas was observed at ca. 608C. Needle-like crystals of 4 were
isolated by slow cooling of a hot hexane solution of the crude
compound. Crystallographic analysis revealed the formation of
a five-coordinate, monomeric complex shown in Figure 2. In
contrast to 3, the gallium centre in 4 lies in a distorted square-
based pyramidal geometry (t=0.19). The basal plane is
formed of the N O unit of the ketoiminate ligand, with the
2
2
substituent methyl group occupying the axial position. This is
in contrast to the geometries seen for the chloride complexes
1
–3, which all tended towards trigonal bipyramidal geome-
tries.
Figure 1. ORTEP diagram of compound 3. Thermal ellipsoids at 50% proba-
bility, hydrogen atoms omitted for clarity.
Table 1. Selected bond lengths [ꢁ] for 3, 4, 7 and 8.
4
2
2
3
Bond
[Ga(L )Cl] (3)
Ga(L )Me (4)
[Ga(L )H] (7)
[Ga(L )H] (8)
GaÀO
1.881(1)
1.942(4)
1.947(4)
2.030(5)
2.034(5)
1.971(5)
1.295(6)
1.302(6)
1.303(7)
1.312(7)
1.937(2)
1.957(2)
2.024(2)
2.043(2)
1.56(3)
1.288(3)
1.304(3)
1.299(3)
1.312(3)
1.915(2)
2.023(2)
1.994(2)
2.082(2)
1.75(3)
1.290(3)
1.303(4)
1.314(3)
1.313(4)
1
.951(1)
1.960(2)
.031(2)
GaÀN
2
GaÀX
CÀO
2.2632(5)
1.295(2)
1
.309(2)
1.309(2)
.325(2)
CÀN
1
moieties are intermediate between those of single and double
[27]
bonds, suggesting a delocalisation of electron density. GaÀO
bond lengths show good agreement with similar gallium b-di-
[
28]
Figure 2. ORTEP diagram of compound 4. Thermal ellipsoids at 50% proba-
bility, hydrogen atoms omitted for clarity.
ketonate complexes,
and are longer than GaÀO bond
lengths within gallium donor-functionalised alkoxide com-
[
29]
plexes. This is a result of the bidentate, monoanionic nature
1
13
1
of the bonding, making each individual GaÀO bond weaker.
Despite washing with hexane, H and C{ H} NMR spectro-
scopic analysis of the crystals showed the presence of a small
amount of the free ligand. However, the resonances corre-
sponding to the product were in agreement with the findings
of the structural characterisation and could be easily distin-
guished from those corresponding to the free ligand. The for-
mation of a monomethyl complex was confirmed by the pres-
ence of a three-proton singlet appearing at À0.46 ppm, corre-
1
13
1
The H and C{ H} NMR spectra of 3 were in agreement with
the observed solid-state structure. Of note was the absence of
1
the broad NH resonance in the H NMR spectrum, characteris-
tic of protonated b-ketoimines, which acted as a good indica-
tion that deprotonation in situ was complete and that no pro-
ligand remained within the reaction mixture. The methylene
groups of the tetramethylene bridging moiety were observed
as two broad singlets at 3.52 and 1.78 ppm. This is likely to be
a result of Berry pseudo-rotation of the complex in the solu-
tion phase, leading to movement in the tetramethylene bridg-
ing group and broad, “averaged” resonances being observed
within the spectrum. Mass spectrometry showed the presence
of the parent ion at m/z 354, in addition to various ions result-
ing from the fragmentation of the complex, in particular the
[30]
sponding to GaÀCH . Mass spectrometry showed the pro-
3
tonated parent ion at m/z 307, as well as the loss of CH indi-
3
cated by a fragment at m/z 291. Elemental analysis of the
sample gave values close to those expected, although a slight
discrepancy was found, likely due to the presence of residual
free ligand within the sample.
In an attempt to form the propylene and tetramethylene an-
alogues of the mono-metallic methyl gallium complex, the re-
loss of Cl (m/z 319) and CH (m/z 339) groups.
3
Chem. Eur. J. 2014, 20, 10503 – 10513
www.chemeurj.org
10505
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
3
2
actions between trimethyl gallium and the free ligands L H
servations made in the attempted synthesis of [In(L )H], which
2
4
[24]
and L H were carried out by applying the same method used
has been previously reported. Given the successful synthesis
2
1
for the synthesis of 4. H NMR spectra of both reaction prod-
and isolation of 7 and 8 without decomposition of the gallane
starting material, it is likely that some reaction occurred with
the tetramethylene pro-ligand, however, any product formed
was insufficiently stable to allow its isolation, leading to the
formation of gallium metal as a grey precipitate.
ucts showed the presence of two sets of resonances, the first
corresponding to the free ligand (including the presence of
the characteristic NH proton) and the second presumably cor-
responding to the methane elimination product. In both cases,
the ratio between the two sets of resonances was 1:1. Further-
more, the peak corresponding to the organometallic protons
1
Initial characterisation of 8 was carried out by H NMR spec-
troscopy, with the formation of the hydride being indicated by
the absence of any NH proton of the pro-ligand and by the
presence of a very broad singlet at 5.83 ppm. This peak broad-
(
GaÀMe) showed an integration of 12H, as opposed to the ex-
pected 3H for the formation of the desired product, indicating
that mono-substitution had taken place, with each equivalent
6
9
ening is due to the two quadrupolar gallium nuclei, Ga and
7
1
3
of the free ligand reacting with two equivalents of GaMe , re-
Ga, with I= / and is characteristic of hydrogen atoms form-
2
3
[23c,32]
sulting in the formation of the bimetallic species 5 and 6, as
shown in Scheme 1.
ing terminal bonds to gallium.
IR spectroscopic analysis
À1
showed a broad band at 1867 cm , characteristic of a GaÀH
[23a]
To confirm this, the reactions were repeated with a ligand-
to-metal stoichiometry of 1:2, resulting in the complete forma-
tion of the bimetallic species. The absence of any resonances
corresponding to the presence of the pro-ligands (in particular
stretch.
Figure 3 shows the molecular structures of hydride com-
plexes 7 and 8, which both crystallised into the monoclinic
P2 /c space group. Analysis of the geometries around the galli-
1
1
that of the NH protons) within the H NMR spectrum con-
firmed that the reaction had gone to completion. Peaks at
À0.38 ppm correspond to the organometallic methyl protons,
which show a slight upfield shift with respect to the unreacted
[
31]
starting material GaMe (À0.18 ppm). Mass spectrometry of
3
5
confirmed the formation of the bimetallic product by the de-
tection of a fragment at m/z 419, corresponding to the molec-
ular ion with loss of a CH group. A fragment at m/z 321 corre-
3
3
sponds to the loss of GaMe (perhaps forming [Ga(L )Me] in
3
the fragmentation process) and loss of a further CH resulted
3
in a fragment detected at m/z 305. The mass spectrum of 6
showed an identical fragmentation pattern, again confirming
the formation of the bimetallic species. Attempts to form crys-
tals of 5 and 6 by numerous methods were unsuccessful, in
contrast to the relatively facile recrystallisation of 4, owing to
the oily nature of the bimetallic species compared with the
solid mono-metallic complex.
Figure 3. ORTEP diagram of a) 7 and b) 8. Thermal ellipsoids at 50% proba-
bility, hydrogen atoms (except H(1)) are omitted for clarity.
The synthesis of gallium hydride derivative 7 has been re-
um centres by calculation of their respective t values revealed
that 7 exhibits a distorted square-based pyramidal geometry
(t=0.35), whereas the cation of 8 lies in a distorted trigonal bi-
pyramidal geometry (t=0.61). In both cases, the degree of dis-
tortion is much greater than observed in chloride and methyl
complexes, which typically showed geometries closer to the
idealised extremes. Applying the same geometrical analysis to
the corresponding ethylene-bridged chloride complex 1 report-
[
24]
ported and discussed in a previous publication,
however,
neither its structural characterisation nor the synthesis of its
propylene analogue were reported. Moreover, apart from ap-
plication in the AACVD of Ga O thin films, no investigation
2
3
into its thermal stability was carried out. Although initial at-
tempts at crystallisation afforded only microcrystalline material,
pale-yellow crystals of 7 were grown by vapour diffusion of
hexane into a CH Cl solution of the crude product. The pro-
[25]
ed by Vohs et al. gave a t value of 0.03, showing a near-per-
2
2
pylene analogue 8 was also synthesised by the same method
fect square-based pyramidal geometry. All three gallium com-
3
2
as 7, by direct reaction of L H with a freshly prepared ethereal
plexes of the form [Ga(L )X] (X=Cl (1), Me (4) and H (7)) were,
2
solution of [GaH (NMe )] (Scheme 1). In this case, the crude
therefore, found to exhibit square-based pyramidal geometry
(albeit to different degrees of distortion), whereas those of the
3
3
product was dissolved in a minimal volume of toluene and
3
cooled to À208C, yielding colourless platelet crystals after four
form [Ga(L )X] (X=Cl (2), Me (8)) were both found to have
4
days. The synthesis of the tetramethylene analogue [Ga(L )H]
trigonal bipyramidal geometries. This observation is attributed
2
3
was investigated by the same synthetic pathway outlined for
both 7 and 8, however numerous attempts led only to the for-
mation of a small amount of grey precipitate within the reac-
tion medium. Isolation of the solution, removal of the solvent
to the lack of flexibility offered by L compared with L and is
further demonstrated by NÀGaÀN bite angles found within the
complexes (Table 2). The constraint of the ethylene-bridged
2
ligand L prohibits the bite angle approaching 908, resulting in
and NMR spectroscopic analysis of the resulting off-white solid
a square-based pyramidal geometry. In contrast, the additional
4
3
product showed only the presence of L H , similar to the ob-
flexibility offered by L allows bite angles closer to 908, thus
2
Chem. Eur. J. 2014, 20, 10503 – 10513
www.chemeurj.org
10506
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Table 2. Comparison of N-Ga-N bite angles and complex geometries
represented by t values) for five-coordinate gallium b-ketoiminate com-
(
plexes.
Complex
t
Bite angle
8]
Complex
t
Bite angle
[8]
[
2
3
[
[
[
Ga(L )Cl] (1)
0.03
0.19
0.35
81.65(5)
80.4(2)
79.58(8)
[Ga(L )Cl] (2)
0.82
0.86
0.61
89.64(5)
93.63(6)
87.77(8)
2
4
Ga(L )Me] (4)
[Ga(L )Cl] (3)
2
3
Ga(L )H] (7)
[Ga(L )H] (8)
square-based pyramidal
trigonal bipyramidal
complexes 2 and 8 display trigonal bipyramidal geometries. In-
creasing the carbon chain length of the bridging group further,
as in chloride complex 3, results in a further increase in the
bite angle (93.63(6)8) and a trigonal bipyramidal geometry is
again observed. As would be expected for main-group com-
plexes, it has been found that the preferred geometry within
these systems is trigonal bipyramidal, however, when geomet-
ric constraints are enforced by the ligand system, square-based
pyramidal geometries are observed.
Selected bond lengths for complexes 7 and 8 are given in
Table 1. The GaÀH bond length of 1.56(3) ꢁ observed in 7 was
[33]
in keeping with literature values for gallium hydride species.
However, within 8, the GaÀH bond was found to be signifi-
cantly longer (1.75(3) ꢁ). Although this bond is also longer
than those found in the majority of structurally characterised
gallium hydride complexes, it is comparable to that of the
[
34]
amido-bridged dimer [H GaNEt ] (1.72 ꢁ) and is likely to be
2
2 2
a reflection of the uncertainty in accurately locating hydrogen
compared with heavier atoms. The remaining bond lengths
within the structures of 7 and 8 are consistent with those in
the b-ketoiminate complexes previously discussed. In particu-
lar, CÀC, CÀO and CÀN bond lengths within the six-membered
ring formed by the gallium centre and b-ketoiminate moieties
show a delocalisation of electron density, as has been dis-
cussed for chloride complex 3.
Figure 4. Thermogravimetric analysis of gallium b-ketoiminate complexes.
a) Complexes 1, 4 and 7 (L ); b) 2, 5 and 8 (L ).
2
3
The hydride complexes showed good stability at room tem-
perature over a period of a few weeks under an inert atmos-
phere. However, the complexes were susceptible to hydrolysis
upon exposure to air and moisture. Thermogravimetric analysis
was carried out to establish the thermal properties and decom-
position characteristics, however, it is pertinent to note that
the onset of decomposition of 8 was not observed until ap-
proximately 2358C, representing considerable thermal stability
for such species and thus presenting hydrides as feasible pre-
cursors for the deposition of gallium oxide by CVD.
tion chamber in the gas phase. All compounds showed a sizea-
ble temperature window between melting and decomposition,
suggesting their suitability for CVD (Table 3).
Bimetallic species 5 and 6 showed melting points that were
significantly lower than those for the monomeric species (65
and 738C, respectively), whereas their decomposition onset
temperatures were comparable (215–2308C), representing
Table 3. Melting points and decomposition onset temperatures [8C] of
compounds 1–8 determined by DSC and TGA, respectively.
Thermal analysis
Compound
Melting point
8C]
Decomp. onset temp.
[8C]
[
To gain an understanding of the general efficacy of b-ketoimi-
nates towards the deposition of Ga O , thermogravimetric
1
2
3
4
5
6
7
8
200
161
191
120
65
73
126
127
260
260
245
210
215
230
215
235
2
3
analysis and differential scanning calorimetry studies were car-
ried out on complexes 1–8 between 25 and 6008C (Figure 4).
One important characteristic of a molecular precursor for ther-
mal CVD processes is a sufficient temperature window be-
tween the melting point and the onset of decomposition to
allow for the successful transport of the precursor to the reac-
Chem. Eur. J. 2014, 20, 10503 – 10513
www.chemeurj.org
10507
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
a very large temperature window of approximately 1508C. As
discussed, the decomposition onset temperatures found for
the hydride complexes 7 (2158C) and 8 (2358C) represent sig-
nificant stability for such species and indicate that such com-
pounds are feasible candidates for CVD application, particularly
due to the low mass and clean decomposition offered by the
hydride co-ligand. Figure 4 shows representative TGA mass
loss profiles for the complexes based on ethylene (Figure 4a)
and propylene-bridged (Figure 4b) ligand derivatives. Chloride
complexes 1–3 showed higher decomposition temperatures
volatilisation took place to allow transfer of the precursor to
[24]
the reaction chamber. To overcome this problem, aerosol-as-
sisted delivery of the precursors was used. Generation of an
aerosol from a precursor solution facilitates the transport of
the precursor to the reaction chamber by use of a carrier gas,
circumventing the high temperatures necessary for thermal
CVD and preventing decomposition of the precursor prior to it
entering the reaction chamber.
Despite the suggestion from TGA that decomposition of the
chloride precursors occurs between 300 and 5008C, which is
within the desired temperature window for the deposition of
Ga O on glass, no deposition was observed for any of the
(up to 1008C higher for complete decomposition) and path-
ways involving multiple steps. Furthermore, residual masses
were higher than would be necessary for full decomposition to
Ga O , suggesting that unwanted ligand substituents were re-
2
3
chloride complexes under any experimental conditions em-
ployed. The substrate temperature was varied between 400
2
3
À1
tained. The TGA profile for hydride complex 7 shows an initial
mass loss at approximately 1008C, which can be attributed to
the loss of residual solvent (toluene) within the sample. No fur-
ther mass loss was subsequently observed until the onset of
decomposition at 2158C, with decomposition occurring by
a relatively clean pathway thereafter. The residual mass (43%)
was again higher than necessary for complete decomposition
to Ga O (32%). The propylene derivative 8, however, showed
and 6008C, flow rates between 0.2 and 1 Lmin and quanti-
ties of precursor between 0.2 and 0.6 g. It was initially thought
that the lack of film growth was due to insufficient solubility of
the precursor in toluene, owing to the presence of a residue of
the precursor within the AACVD flask after the nebulisation
and transport of the solvent to the CVD reactor. However, use
of a number of alternative solvents (CH Cl , tetrahydrofuran
2
2
(THF) and MeOH), in which the solubility of the precursors was
sufficient for no residue to be left, did not lead to any observa-
ble deposition.
2
3
a very clean decomposition profile and the total mass loss ob-
served was 71%, compared to a theoretical mass loss of 69%
for decomposition to the oxide. In our earlier attempts to form
similar hydride species based on b-diketonate ligand moieties,
the only gallium-containing products isolated were tris-diketo-
Depositions from precursors 5 and 8 were successful in the
formation of thin, transparent films, observable by the appear-
ance of interference fringes on the silica-coated glass sub-
strates. Depositions were optimised on the basis of substrate
coverage, with optimum conditions being given in Table 4. All
films were determined to be adherent by the Scotchꢂ tape
test, however they were easily scratched by a brass or stainless
steel stylus.
[
35]
nates of the form [Ga(bdk) ] (bdk=b-diketonate). However,
3
of note was the isolation of a doubly-reduced b-diol from the
reaction mixture owing to the hydride reduction of the b-di-
ketonate ligand. A grey precipitate was also observed within
the reaction mixture, attributed to gallium metal. These obser-
vations provide evidence that such systems are prone to the
formation of metallic gallium, with the hydride acting as a re-
ducing agent, and it is therefore likely that decomposition of 7
and 8 yields metallic gallium in addition to Ga O , given the re-
Table 4. Optimised AACVD conditions for the deposition of Ga
single-source precursors 5 and 8 and XPS determined elemental ratios.
2 3
O from
[
a]
2
3
ducing potential of the hydride species. Indeed, this consider-
ation could account for the small discrepancy between the ex-
perimental and calculated mass losses for 8, although this
greater than anticipated mass loss could also be attributed to
sublimation of the precursor.
Film No.
Precursor
Mass
[g]
Solvent
volume [mL]
Deposition
time [h]
Substrate
temp. [8C]
F5
F8
5
8
0.2
0.6
30
40
2
2.5
550
500
[
a] All depositions were carried out using precursor solutions in anhy-
The best decomposition profiles were observed for the
methyl derivatives 4–6, with clean, single-step decompositions
occurring between 200–4008C and major decomposition steps
reaching completion at ca. 3508C. Total mass loss was greater
than necessary for decomposition to Ga O , suggesting either
À1
2
drous toluene and N as a carrier gas at a flow rate of 0.5 Lmin .
Glancing angle powder X-ray diffraction was carried out on
films F5 and F8 as an initial tool for characterisation. However,
in both cases, the as-deposited films were found to be amor-
phous, as expected for the growth of Ga O at substrate tem-
2
3
sublimation of the precursor prior to decomposition, or de-
composition to gallium metal. Whereas the TGA profiles pres-
ent the methyl complexes as suitable precursors, the bimetallic
species 5 and 6 are likely to deposit oxygen-deficient films
owing to the gallium-to-oxygen ratio of 1:1 within the precur-
sor molecule (cf. Ga O =1:1.5).
2
3
[
29,36]
peratures below approximately 7008C.
Energy dispersive
X-ray analysis (EDXA) confirmed the presence of gallium within
the films, however the gallium-to-oxygen ratios could not be
accurately measured due to breakthrough to the underlying
glass substrate. From SEM imaging (Figure 5), both films were
found to be made up of spherical particles, indicating that film
growth had occurred through a Volmer–Weber-type island
growth mechanism.
2
3
Aerosol-assisted chemical vapour deposition
We had previously attempted to utilise 7 in an LPCVD process,
resulting in decomposition of the precursor before sufficient
Chem. Eur. J. 2014, 20, 10503 – 10513
www.chemeurj.org
10508
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 5. SEM images of Ga
2 3
O films of a) F5 (ꢃ30000 magnification) and
b) F8 (ꢃ40000 magnification).
XPS analysis further confirmed the presence of both gallium
and oxygen within the films by the presence of Ga 3d and O
1
s peaks, respectively (Figure 6, Table 5). For reference, a Ga O3
2
standard was also analysed by XPS (Figure 6a). Peaks at 19.87,
9.54 and 19.41 eV for F5, F8 and standard, respectively, were
1
[37]
consistent with literature values for Ga 3d5/2 within Ga O . In
2
3
all samples, including the standard, the presence of another
gallium species was indicated by the appearance of a doublet
at 18.21 and 18.66 eV, likely to be metallic gallium by compari-
son to the Ga 3d binding energy of 18.4 eV reported for ele-
5/2
[
38]
mental gallium. The observation of this impurity was also re-
ported by Ghosh et al. in the XPS analysis of a Ga O stan-
2
3
[
39]
dard. The oxygen 1s peak of Ga O is reported to appear at
2
3
[
40]
a binding energy of approximately 531 eV, therefore peaks
observed at 530.44, 530.35 and 530.52 eV were attributed to O
1s of the native oxide. A small additional O 1s peak at approxi-
mately 532 eV within the XPS spectra for the films and stan-
dard were attributed to the presence of trace amounts of
[40b]
other oxygen-containing impurities within the samples.
Gal-
lium-to-oxygen ratios (Table 5) were calculated from the XPS
data, showing the standard sample and film F5 to be near stoi-
chiometric Ga O (Ga/O ratio 0.67 and 0.62, respectively), with
2
3
slight discrepancies being attributed to trace impurities at the
surface level. Film F8, however, was found to be oxygen-defi-
cient (Ga/O ratio 0.95).
Oxygen-deficient film F8 showed low to moderate transpar-
ency in the visible region, as determined from the UV/Vis/IR
transmission spectrum shown in Figure 7. The sample showed
approximately 20–60% transmission in the visible region,
which decreased towards the UV wavelengths. The spectrum
of the near-stoichiometric film F5, however, showed much im-
proved transparency (60–80% in the visible). Whereas our pre-
vious report of Ga O thin films deposited from 7 showed com-
2
3
Figure 6. XPS peaks for Ga 3d. a) Ga O standard; b) film F5; c) film F8.
2
3
[24]
paratively higher transparency (>80% in the visible), these
films had been post-treated by annealing in air at 10008C. It is
therefore likely that the reduced transmission in F5 and F8
arises from carbon contamination. Indeed, XPS analysis re-
vealed a carbon content at the surface of the as-deposited
films, though accurate quantification was unreliable due to the
presence of surface contaminants.
displayed by the complexes range between the idealised ex-
tremes of trigonal bipyramidal and square-based pyramidal ge-
ometries and are heavily influenced by the nature of the
ligand system, in particular by the length of the bridging
carbon chain between the two b-ketoimine moieties. In partic-
ular, the stabilising ability of these ligands towards rare gallium
hydride species has been demonstrated through the synthesis
of monomeric hydride species 7 and 8, the crystal structures of
which are also reported. Moreover, the complexes showed
considerable thermal stability, with the onset of decomposition
occurring above 2008C. The general instability of such gallium
Conclusion
A range of gallium species utilising bis-b-ketoimine ligand sys-
tems have been synthesised and the crystal structures of
a number of such species have been reported. The geometries
Chem. Eur. J. 2014, 20, 10503 – 10513
www.chemeurj.org
10509
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[
42]
[
GaH (NMe )], were synthesised by literature procedures, as were
3 3
Table 5. Ga 3d5/2 and O 1s XPS peaks and gallium-to-oxygen ratios for
samples F5, F8 and a Ga O standard. Literature values included for refer-
2 3
ence.
n
ligands [(CH ) {N(H)C(Me)ÀCHC(Me)=O} ] (L H , n=2, 3 and 4) with
2
n
43]
2
2
[
adaptations. Details of the ligand syntheses can be found in the
Supporting Information.
Sample
Binding energy [eV]
Ga 3d5/2
Ga/O
All manipulations were carried out under an inert atmosphere
using standard Schlenk techniques. For hydride syntheses, glass-
ware was flame-dried prior to use. Nitrogen (99.99%) was obtained
from BOC and passed through activated molecular sieves to
remove moisture. All solvents were dried over activated alumina
by the Grubbs method using anhydrous engineering equipment,
O 1s
F5
F8
19.87
19.54
19.41
530.44
530.35
530.52
530.70
0.62
0.95
0.67
–
Ga
Ga
2
2
O
O
3
(standard)
(lit.)
[
36]
[41]
3
19.6
[44]
such that the water concentration was 5–10 ppm.
4
Complex [Ga(L )Cl] (3)
NaH (1.0 g, 42.0 mmol) was suspended in THF (50 mL) and added
4
to a stirring solution of L H (5.0 g, 19.8 mmol) in THF (50 mL) and
2
the mixture was stirred overnight. Solvent was removed from the
resulting suspension in vacuo to afford a pale-yellow solid product,
4
4
L Na . Subsequently, L Na (0.59 g, 2.0 mmol) was suspended in
2
2
THF (20 mL) and added slowly to a stirring solution of GaCl3
(
0.35 g, 2.0 mmol) in THF (20 mL) at À788C. The mixture was
stirred for approximately 30 min before being warmed slowly to RT
and subsequently heated to reflux for 16 h. The resulting yellow
supernatant was separated from the white solid precipitate (NaCl)
by filtration and the solvent was removed in vacuo to afford
a solid yellow crude product. The product was then dissolved in
minimal CH Cl (ca. 3 mL), filtered into a narrow Schlenk flask and
2
2
layered with hexane (ca. 10 mL). The product was left overnight,
during which time colourless crystalline blocks formed. Yield:
1
0
.71 g (67%); H NMR (600 MHz, CDCl , 258C, TMS): d=5.10 (s, 2H;
3
Figure 7. UV/Vis/IR transmission spectrum for films F5 and F8.
CH), 3.52 (br s, 4H; CH CH CH CH ), 2.06 (s, 6H; CH CO), 2.03 (s,
2
2
2
2
3
1
3
1
6
H; CH CN), 1.78 ppm (br s, 4H; CH CH CH CH ); C{ H} NMR
3 2 2 2 2
(
4
600 MHz, CDCl , 258C, TMS): d=182.9 (CO), 174.9 (CN), 99.5 (CH),
8.7 (CH CH CH CH ), 28.6 (CH CH CH CH ), 26.6 (CH CN), 22.5 ppm
2 2 2 2 2 2 2 2 3
3
hydrides has previously limited their feasible use as single-
source precursors in thin-film deposition, thus the isolation of
thermally stable hydrides is of great significance for their appli-
cation in this area.
(
CH CO); MS: m/z: 354 [M], 339 [MÀCH ], 319 [MÀCl], 311
3
3
[MÀCH CO]; elemental analysis calcd (%) for C H N O Ga: C
3
14 22
2
2
47.30, H 6.24, N 7.88; found: C 46.72, H 6.20, N 6.17.
The general efficacy of b-ketoiminate precursors has been
investigated through the deposition of Ga O thin films from
2
2
3
Complex [Ga(L )Me] (4)
methyl (5) and hydride (8) precursor derivatives. In both cases,
the deposition of gallium- and oxygen-containing thin films
was confirmed through EDX and XPS analysis. The formation
of oxygen-deficient films from the hydride precursor 8 is
a likely result of hydride acting as a reducing agent, resulting
in the co-deposition of gallium metal and gallium oxide within
the resultant films. In the absence of hydride, films deposited
from methyl derivative 5 showed near-stoichiometric Ga O ,
despite the gallium-to-oxygen ratio of 1:1 within the precursor.
The work demonstrates the potential of b-ketoiminate based
precursor systems for the deposition of metal oxide materials
by CVD and further investigation into alternative ligand deriva-
tives is ongoing.
2
L H (7.8 g, 34.8 mmol) was dissolved in anhydrous toluene (30 mL)
2
and added slowly to a stirring solution of GaMe (4.0 g, 34.8 mmol)
3
in toluene (30 mL) at À788C. The mixture was stirred at À788C for
approximately 30 min before being warmed slowly to RT. The reac-
tion mixture was then heated to 808C in an oil bath and stirred at
this temperature for 16 h. The resulting dark-yellow/orange solu-
tion was filtered and toluene was removed under reduced pressure
to give a dark-yellow waxy product. The product was dissolved in
a minimum volume of hexane with gentle heating and allowed to
cool to RT, during which time needle-like, yellow crystals formed.
Yield: 8.16 g (77%); H NMR (600 MHz, CDCl , 258C, TMS): d=4.91
s, 2H; CH), 3.37–3.46 (m, 4H; CH CH ), 1.99 (s, 6H; CH CO), 1.94 (s,
2
3
1
3
(
2
2
3
13
1
6
H; CH CN), À0.45 ppm (s, 3H; GaCH ); C{ H} NMR (600 MHz,
3
3
CDCl , 258C, TMS): d=182.1 (CH CO), 171.0 (CH CN), 97.5 (CH), 45.9
3
3
3
(
3
CH CH ), 26.9 (CH CN), 21.6 (CH CO), À7.1 ppm (GaCH ); MS: m/z:
2
2
3
3
3
+
+
07 [M+H] , 291 [MÀCH ] ; elemental analysis calcd (%) for
3
C H GaN O : C 50.85, H 6.89, N 9.12; found: C 49.23, H 7.29, N
13
21
2
2
Experimental Section
9
.24.
General procedures
3
Complex [L (GaMe ) ] (5)
2
2
GaCl (10 mesh beads, 99.99%), LiH, NaH and amines were pur-
3
3
chased from Sigma–Aldrich and b-diketones from Alfa Aesar: all
L H (0.6 g, 2.5 mmol) was dissolved in anhydrous toluene (20 mL)
2
were used without further purification. GaMe3 was supplied by
and added slowly to a stirring solution of GaMe (0.6 g, 5.0 mmol)
in toluene (20 mL) at À788C. The mixture was stirred at À788C for
3
[24]
SAFC Hitech Ltd. Gallium complexes 1, 2 and 7,
along with
Chem. Eur. J. 2014, 20, 10503 – 10513
www.chemeurj.org
10510
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
approximately 30 min before being warmed slowly to RT. The reac-
tion mixture was then heated to approximately 808C and stirred at
this temperature for 16 h. The yellow solution was cooled to RT, fil-
tered and the solvent was removed from the filtrate in vacuo. Re-
crystallisation of the resulting yellow solid powder was unsuccess-
tion, two substrates were placed within the reactor with the lower
surface of the ‘top-plate’ sitting at a distance of approximately
8 mm above the upper surface of the ‘bottom-plate’. The reactor
was heated to the relevant temperature by a graphite block con-
taining a Whatman cartridge heater and the temperature of the
block was monitored by using a Pt-Rh thermocouple. All deposi-
tion temperatures quoted refer to this temperature. The tempera-
ture of the top-plate was generally found to be approximately 50–
708C lower than that of the bottom-plate. After reaching the de-
sired temperature, the temperature within the reactor was allowed
to equilibrate for 30 min prior to commencing the deposition.
ful from a range of solvent and temperature combinations. Yield:
1
0
.63 g (58%); H NMR (600 MHz, CDCl , 258C, TMS): d=4.85 (s, 2H;
3
CH), 3.23–3.27 (m, 4H; CH CH CH ), 1.96 (s, 6H; CH CO), 1.91 (s,
2
2
2
3
6
H; CH CN), 1.70–1.80 (m, 2H; CH CH CH ), À0.38 ppm (s, 12H;
3
2
2
2
13 1
Ga(CH ) ); C{ H} NMR (600 MHz, CDCl , 258C, TMS): d=180.4 (CO),
3
2
3
1
71.5 (CN), 98.2 (CH), 46.4 (CH CH CH ), 30.5 (CH CH CH ), 26.4
2 2 2 2 2 2
(
3
CH CN), 20.9 (CH CO), À7.3 ppm (Ga(CH ) ); MS: m/z 421 [MÀCH ],
3
3
3 2
3
The precursors were dissolved in anhydrous solvent using standard
Schlenk techniques, from which an aerosol was generated at RT by
use of an ultrasonic humidifier. The aerosol was carried into the re-
actor by passing nitrogen through a brass baffle to obtain laminar
flow. The total time for each deposition was dependent upon the
solvents, volumes and flow rates being used. Depositions were car-
ried out at substrate temperatures ranging between 400–6008C.
21 [MÀGaMe ], 305 [MÀGaMe ÀCH ]; elemental analysis calcd
3
3
3
(
%) for C H N O Ga : C 46.84, H 7.40, N 6.43; found: C 45.30, H
17 32 2 2 2
8
.01, N 6.28.
Complex [L4(GaMe2)2] (6)
4
L H (0.6 g, 2.5 mmol) was dissolved in anhydrous toluene (20 mL)
2
and added slowly to a stirring solution of GaMe (0.6 g, 5.0 mmol)
3
Analysis techniques
in toluene (20 mL) at À788C. The mixture was stirred at À788C for
approximately 30 min before being warmed slowly to RT. The reac-
tion mixture was then heated to approximately 808C and stirred at
this temperature for 16 h. The yellow solution was cooled to RT, fil-
tered and the solvent was removed from the filtrate in vacuo. Re-
crystallisation of the resulting yellow solid powder was unsuccess-
Thermogravimetric analysis was carried out with a Netzsch system
from 20 to 6008C. Samples of approximately 10–20 mg were ana-
lysed under an inert atmosphere of helium in sealed aluminium
pans. Scanning electron microscopy (SEM) was performed to deter-
mine surface morphology and film thickness with a JEOL JSM-
6301F field emission SEM at an accelerating voltage of 5 keV.
Images were captured with SEMAfore software. Samples were cut
to approximately 5ꢃ5 mm coupons and coated with a fine layer of
gold to avoid charging.
ful from a range of solvent and temperature combinations. Yield:
1
0
.89 g (79%); H NMR (600 MHz, CDCl , 258C, TMS): d=4.85 (s, 2H;
3
CH), 3.28 (br s, 4H; CH CH CH ), 1.96 (s, 6H; CH CO), 1.92 (s, 6H;
2
2
2
3
CH CN), 1.53 (br s, 2H; CH CH CH ), À0.38 ppm (s, 12H; Ga(CH ) );
3
1
2
2
2
3 2
1
3
C{ H} NMR (600 MHz, CDCl , 258C, TMS): d=180.0 (CO), 171.3
3
X-ray photoemission spectroscopy (XPS) was performed with
a Thermo Scientific K-alpha photoelectron spectrometer using
monochromatic AlKa radiation. Survey scans were collected in the
range 0–1100 eV (binding energy) at a pass energy of 160 eV.
Higher resolution scans were recorded for the principal peaks of
Ga (3d), O (1s) and C (1s) at a pass energy of 50 eV. Peak positions
were calibrated to carbon and plotted by using the CasaXPS soft-
ware. Energy dispersive (EDX) analysis was carried out with a JEOL
JSM-6301F field emission instrument with an acceleration voltage
of 20 kV. UV/Vis transmittance and reflectance spectra were ob-
tained with a PerkinElmer Lambda 950 spectrometer using an air
background and were recorded between 250 and 2000 nm.
(
CN), 98.3 (CH), 48.6 (CH CH CH ), 27.9 (CH CH CH ), 26.4 (CH CN),
2 2 2 2 2 2 3
2
1.2 (CH CO), À7.4 ppm (Ga(CH ) ); MS: m/z: 435 [MÀCH ], 335
3
3 2
3
[MÀGaMe ], 319 [MÀGaMe ÀCH ]; elemental analysis calcd (%) for
3
3
3
C H N O Ga : C 48.05, H 7.62, N 6.23; found: C 47.15, H 7.81, N
1
8
34
2
2
2
6
.15.
3
Complex [Ga(L )H] (8)
A freshly prepared solution of [GaH (NMe )] (5.0 mmol) in Et O
3
3
2
3
(
(
50 mL) was cooled to À788C and added to a suspension of L H
2
1.2 g, 5.0 mmol) in Et O (20 mL) at À788C. The suspension was
2
stirred for 15 min, then warmed to RT and stirred for 16 h. During
this time, a pale-yellow solution with a white suspension formed.
The reaction was filtered, volatiles were removed in vacuo and the
crude product was dissolved in minimal toluene. The toluene solu-
Crystallography
Crystals were obtained as described above. Details of the crystallo-
graphic data collection and refinement are given in Table 6. Diffrac-
tometers: for compounds 3 and 7, Rigaku R-Axis Spider including
curved Fujifilm image plate and a graphite monochromated sealed
tion was cooled to À208C, from which colourless crystals were
1
formed overnight. Yield: 0.91 g (59%; based on GaCl ); H NMR
3
(
600 MHz, C D , 258C, TMS): d=5.83 (v br s, 1H; GaH), 4.85 (s, 2H;
6 6
CH), 2.80–2.89 and 2.79–2.85 (m; CH CH CH , each 2H), 2.15–2.19
tube Mo generator (l =0.71073 ꢁ) was used; for compound 4,
2
2
2
1
(
1
m, 1H; CH CH CH ), 1.98 (s, 6H; COCH ), 1.36 (s, 6H; CNCH ), 1.12–
a Bruker SMART APEX CCD diffractometer using graphite mono-
2
2
2
3
3
13
1
.16 ppm (m, 1H; CH CH CH ); C{ H} NMR (600 MHz, C D , 258C,
chromated MoKa radiation (l =0.71073 ꢁ) was used; for com-
2
2
2
6
6
1
TMS): d=182.95 (CO), 171.1 (CN), 97.9 (CH), 49.0 (CH CH CH ), 26.8
pound 8, a Rigaku AFC12 goniometer equipped with an enhanced
sensitivity (HG) Saturn724+ detector mounted at the window of
an FR-E+ SuperBright molybdenum rotating anode generator
2
2
2
(
CNCH ), 26.5 (CH CH CH ), 21.2 ppm (COCH ); MS: m/z: 305
3
+
2
2
2
3
À1
[MÀH ]; IR (KBr): n˜ =1867 cm (br, Ga-H); elemental analysis calcd
(
%) for C H N O Ga: C 50.85, H 6.89, N 9.12; found: C 49.56, H
(l =0.71073 ꢁ) with VHF Varimax optics (100 mm focus) was used.
1
1
3
21
2
2
7
.31, N 8.48.
For compounds 3, 7 and 8: cell determination, data collection,
data reduction, cell refinement and absorption correction were car-
[45]
ried out by using CrystalClear-SM Expert 3.1 b18. For compound
, cell determination, data reduction and integration were carried
out with SAINT+ and absorption corrections were applied by
Aerosol-assisted chemical vapour deposition
4
Films were grown on SiO -coated float-glass substrates supplied by
2
[46]
Pilkington NSG, of dimensions approximately 90ꢃ45ꢃ4 mm. Prior
to use, the substrates were cleaned by using water and detergent,
propan-2-ol and acetone and were dried in air. For each deposi-
using SADABS. Structure solution and refinement were carried
[47]
out by using WinGX and software packages. H atoms attached
to C atoms were placed in geometrically assigned positions, with
Chem. Eur. J. 2014, 20, 10503 – 10513
www.chemeurj.org
10511
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[
[
7] S. C. Vanithakumari, K. K. Nanda, Adv. Mater. 2009, 21, 3581–3584.
8] M. Orita, H. Hiramatsu, H. Ohta, M. Hirano, H. Hosono, Thin Solid Films
Table 6. Crystallographic data for the compounds reported in this paper.
2
002, 411, 134–139.
Compound
3
4
7
8
[
9] P. Meriaudeau, C. Naccache, J. Mol. Catal. 1990, 59, L31–L36.
formula
M [gmol
T [K]
C
14
H
22ClGaN
2
O
2
C
13
H
21GaN
2
O
2
C
12
H
19GaN
2
O
2
C
13
H
21GaN
2
O
2
[10] F. Shi, S. Zhang, C. Xue, J. Alloys Compd. 2010, 498, 77–80.
[11] S. Mishra, S. Daniele, S. Petit, E. Jeanneau, M. Rolland, Dalton Trans.
2009, 2569–2577.
À1
]
355.51
120(2)
307.04
150(2)
293.01
120(2)
307.04
100(2)
crystal system triclinic
monoclinic
P2 /c (14)
monoclinic
P2 /c (14)
monoclinic
P21/c (14)
[12] J. Hao, M. Cocivera, J. Phys. D 2002, 35, 433.
[13] R. Binions, C. J. Carmalt, I. P. Parkin, Meas. Sci. Technol. 2007, 18, 190–
200.
space group
No.)
P-1 (2)
1
1
(
a [ꢁ]
b [ꢁ]
c [ꢁ]
a [8]
b [8]
g [8]
8.646(1)
9.218(1)
11.043(2)
79.726(3)
77.402(3)
65.573(3)
778.2(2)
2
8.148(2)
13.014(3)
14.885(3)
90
110.30(1)
90
1480.4(6)
4
1.855
640
7.9220(2)
12.0209(3)
13.5746(5)
90
93.045(2)
90
1290.88(7)
4
2.124
608
13496
2956
0.041
1.121
7.4450(5)
15.1760(9)
12.4920(8)
90
97.112(1)
90
1400.6(2)
4
1.961
640
[14] D. J. Comstock, J. W. Elam, Chem. Mater. 2012, 24, 4011–4018.
[15] L. Mꢅinea, S. Suh, S. G. Bott, J.-R. Liu, W.-K. Chu, D. M. Hoffman, J. Mater.
Chem. 1999, 9, 929–935.
[16] R. Binions, C. J. Carmalt, I. P. Parkin, K. F. E. Pratt, G. A. Shaw, Chem.
Mater. 2004, 16, 2489–2493.
[17] S. Basharat, C. J. Carmalt, R. Palgrave, S. A. Barnett, D. A. Tocher, H. O.
Davies, J. Organomet. Chem. 2008, 693, 1787–1796.
[18] P. Marchand, I. A. Hassan, I. P. Parkin, C. J. Carmalt, Dalton Trans. 2013,
42, 9406–9422.
[19] A. H. Cowley, R. A. Jones, Polyhedron 1994, 13, 1149–1157.
[20] a) L. G. Bloor, C. J. Carmalt, D. Pugh, Coord. Chem. Rev. 2011, 255, 1293–
1318; b) D. Pugh, L. Bloor, S. Moniz, C. J. Carmalt in Comprehensive Inor-
ganic Chemistry II, Vol. 1 (Eds.: J. Reedijk, K. Poeppelmeier), Elsevier,
Oxford, 2013, pp. 1021–1052.
[21] a) A. T. S. Wee, A. J. Murrell, N. K. Singh, D. O’Hare, J. S. Foord, J. Chem.
Soc. Chem. Commun. 1990, 11–13; b) C. Jones, G. A. Koutsantonis, C. L.
Raston, Polyhedron 1993, 12, 1829–1848.
3
U [ꢁ ]
Z
m(MoKa) [mm ] 1.942
F(000)
total reflns
unique reflns
À1
368
9706
3531
0.034
1.084
12073
3432
9263
2895
R
int
0.094
0.892
0.060
0.117
0.138
0.160
0.032
1.116
0.041
0.044
0.113
2
GooF on F
b
R
R
1
[I
o
>2s(I
o
)] 0.026
0.028
)] 0.063
0.065
0.037
0.039
0.095
0.096
1
(all data)
wR
wR
b
2
[I
o
>2s(I
o
(all data)
0.116
[22] C. D. Abernethy, M. L. Cole, C. Jones, Organometallics 2000, 19, 4852–
2
4
857.
23] a) M. Veith, S. Faber, H. Wolfanger, V. Huch, Chem. Ber. 1996, 129, 381–
84; b) G. A. Koutsantonis, F. C. Lee, C. L. Raston, Main Group Chem.
995, 1, 21–28; c) D. Pugh, L. G. Bloor, I. P. Parkin, C. J. Carmalt, Chem.
Eur. J. 2012, 18, 6079–6087.
[
3
1
CÀH distances of 0.95 (CH), 0.98 (CH ) or 0.99 ꢁ (CH ) and refined
3
2
by using a riding model, with U (H)=1.2U (C) (CH, CH ) or
iso
eq
2
1
.5U (C) (CH ). The GaÀH protons in 7 and 8 were located in the
[24] D. Pugh, P. Marchand, I. P. Parkin, C. J. Carmalt, Inorg. Chem. 2012, 51,
6385–6395.
[25] J. K. Vohs, D. O. Miller, D. R. Denomme, J. W. Ziller, B. D. Fahlman, Acta
Crystallogr. Sect. C 2005, 61, m287–m288.
eq
3
Fourier difference map and allowed to refine freely with U (H)=
eq
[48]
1
.2U (Ga). enCIFer was used to prepare CIFs for publication.
eq
CCDC-994235 (3), -994236 (4), -994237 (7) and -994238 (8) contain
the supplementary crystallographic data for this paper. These data
can be obtained free of charge from The Cambridge Crystallo-
graphic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
[
[
[
[
26] A. W. Addison, T. N. Rao, J. Reedijk, J. Vanrijn, G. C. Verschoor, J. Chem.
Soc. Dalton Trans. 1984, 1349–1356.
27] F. H. Allen, O. Kennard, D. G. Watson, L. Brammer, A. G. Orpen, R. Taylor,
J. Chem. Soc. Perkin Trans. 2 1987, S1–S19.
28] B. Ballarin, G. A. Battiston, F. Benetollo, R. Gerbasi, M. Porchia, D. Favret-
to, P. Traldi, Inorg. Chim. Acta 1994, 217, 71–78.
29] S. Basharat, C. J. Carmalt, S. J. King, E. S. Peters, D. A. Tocher, Dalton
Trans. 2004, 3475–3480.
Acknowledgements
We would like to thank the EPSRC (grant nos. EP/F035675 and
EP/H00064X) and UCL (Impact Studentship Scheme) for finan-
cial support. We would also like to thank SAFC Hitech Ltd. for
supplying GaMe and InMe , Pilkington NSG for supplying the
[30] C. E. Knapp, L. Pemberton, C. J. Carmalt, D. Pugh, P. F. McMillan, S. A.
Barnett, D. A. Tocher, Main Group Chem. 2010, 9, 31–40.
[
[
31] H. Tang, R. Fabicon, H. G. Richey, Organometallics 1998, 17, 139–145.
32] a) C. R. Pulham, A. J. Downs, M. J. Goode, D. W. H. Rankin, H. E. Robert-
son, J. Am. Chem. Soc. 1991, 113, 5149–5162; b) J. L. Atwood, S. G. Bott,
F. M. Elms, C. Jones, C. L. Raston, Inorg. Chem. 1991, 30, 3792–3793;
c) N. N. Greenwood, E. J. F. Ross, A. Storr, J. Chem. Soc. A 1966, 706–711;
d) E. Johnsen, A. J. Downs, T. M. Greene, P. F. Souter, K. Aarset, E. M.
Page, D. A. Rice, A. N. Richardson, P. T. Brain, D. W. H. Rankin, C. R.
Pulham, Inorg. Chem. 2000, 39, 719–727; e) A. J. Downs, L. A. Harman,
P. D. P. Thomas, C. R. Pulham, Polyhedron 1995, 14, 935–945; f) O. T.
Beachley, R. G. Simmons, Inorg. Chem. 1980, 19, 783–785.
3
3
float glass, and Drs. Peter Horton, Mateusz Pitak and Graham
Tizzard at the EPSRC National Crystallography Service for data
set collection of compounds 3, 7 and 8.
Keywords: chelates · chemical vapour deposition · gallium
oxides · hydrides · thin films
[
[
33] S. Aldridge, A. J. Downs, Chem. Rev. 2001, 101, 3305–3366.
34] J. Lorberth, R. Dorn, W. Massa, S. Wocadlo, Z. Naturforsch., B: J. Chem.
Sci. 1993, 48, 224–226.
[
[
1] a) L. M. Foster, H. C. Stumpf, J. Am. Chem. Soc. 1951, 73, 1590–1595;
b) M. Marezio, J. P. Remeika, J. Chem. Phys. 1967, 46, 1862–1865.
2] a) T. Harwig, G. J. Wubs, G. J. Dirksen, Solid State Commun. 1976, 18,
[35] D. Pugh, L. G. Bloor, S. Sathasivam, I. P. Parkin, C. J. Carmalt, Eur. J. Inorg.
Chem. 2011, 1953–1960.
1
223–1225; b) Z. Hajnal, J. Miro, G. Kiss, F. Reti, P. Deak, R. C. Herndon,
[36] a) P. Marchand, C. J. Carmalt, Coord. Chem. Rev. 2013, 257, 3202–3221;
b) S. Basharat, W. Betchley, C. J. Carmalt, S. Barnett, D. A. Tocher, H. O.
Davies, Organometallics 2007, 26, 403–407.
[37] F. J. Grunthaner, P. J. Grunthaner, R. P. Vasquez, B. F. Lewis, J. Maserjian,
A. Madhukar, J. Vac. Sci. Technol. 1979, 16, 1443–1453.
[38] H. A. Budz, M. C. Biesinger, R. R. LaPierre, J. Vac. Sci. Technol. B 2009, 27,
637–648.
J. M. Kuperberg, J. Appl. Phys. 1999, 86, 3792–3796.
[
[
3] M. Fleischer, H. Meixner, Sens. Actuators B 1991, 4, 437–441.
4] a) M. Fleischer, H. Meixner, Sens. Actuators B 1992, 6, 257–261; b) M.
Fleischer, H. Meixner, Sens. Actuators B 1995, 26, 81–84.
5] Z. F. Liu, T. Yamazaki, Y. Shen, T. Kikuta, N. Nakatani, Y. X. Li, Sens. Actua-
tors B 2008, 129, 666–670.
[
[
6] A. Cꢄneyt Ta s¸ , P. J. Majewski, F. Aldinger, J. Am. Ceram. Soc. 2000, 83,
[39] S. C. Ghosh, M. C. Biesinger, R. R. LaPierre, P. Kruse, J. Appl. Phys. 2007,
101, 114322–114328.
2954–2960.
Chem. Eur. J. 2014, 20, 10503 – 10513
www.chemeurj.org
10512
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[
40] a) A. Trinchi, S. Kaciulis, L. Pandolfi, M. K. Ghantasala, Y. X. Li, W. Wlodar-
[44] A. B. Pangborn, M. A. Giardello, R. H. Grubbs, R. K. Rosen, F. J. Timmers,
Organometallics 1996, 15, 1518–1520.
ski, S. Viticoli, E. Comini, G. Sberveglieri, Sens. Actuators, B 2004, 103,
1
2
29–135; b) W. Mi, J. Ma, Z. Zhu, C. Luan, Y. Lv, H. Xiao, J. Cryst. Growth
012, 354, 93–97.
41] G. Hollinger, R. Skheyta-Kabbani, M. Gendry, Phys. Rev. B 1994, 49,
1159–11167.
[45] CrystalClear-SM Expert 3.1 b18, Rigaku Corporation, Tokyo, Japan, 2013.
[46] SMART, SAINT, SADABS, Bruker AXS Inc., Madison, WI, 1998.
[47] L. J. Farrugia, J. Appl. Crystallogr. 2012, 45, 849–854.
[48] F. H. Allen, O. Johnson, G. P. Shields, B. R. Smith, M. Towler, J. Appl. Crys-
tallogr. 2004, 37, 335–338.
[
[
[
1
42] P. C. Andrews, M. G. Gardiner, C. L. Raston, V.-A. Tolhurst, Inorg. Chim.
Acta 1997, 259, 249–255.
43] a) A. E. Martell, R. Linn Belford, M. Calvin, J. Inorg. Nucl. Chem. 1958, 5,
170–181; b) Y. Elerman, M. Kabak, H. Kara, K. Guven, C. Nakipoglu, Acta
Received: April 8, 2014
Crystallogr. Sect. C 1999, 55, 1508–1511.
Published online on July 17, 2014
Chem. Eur. J. 2014, 20, 10503 – 10513
www.chemeurj.org
10513
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim