Dimethylmalonato zinc complexes as molecular precursors for electronic grade zinc oxide: Towards a systematic ligand design by understanding molecular decomposition mechanisms

Article abstract of DOI:10.1002/ejic.201301525

Zinc complexes with dimethyl 2-hydroxyimino- and 2-nitromalonate (dmm-NOH and Hdmm-NO₂, respectively) were synthesized and examined as potential precursors for nanocrystalline zinc oxide. The functionalised 1,3-diketones are sufficiently acidic to allow direct synthesis from hydrozincite or zinc 2-propoxide as well as their full spectroscopic and analytical characterization. Their nominal elemental compositions of [Zn ₄O(dmm-NO)₆] and [Zn₃(OH)₄(dmm- NO₂)₂] suggest the presence of multinuclear cage structures with oxo- or hydroxo-bridging ligands. This assumption is in accordance with the obtained spectroscopic data. Thermogravimetry coupled with mass spectrometry and infrared spectroscopy (TG/MS and TG-IR) indicate the formation of dimethylcarbonate and methanol during the decomposition. Both precursors exhibit good film formation properties and the obtained zinc oxide was further characterized by XRD, SEM, TEM and PL. Finally, the semiconducting electronic behaviour of the obtained nanocrystalline ZnO was studied in a field-effect transistor (FET) device. Copyright

Full text of DOI:10.1002/ejic.201301525

FULL PAPER
DOI:10.1002/ejic.201301525
Dimethylmalonato Zinc Complexes as Molecular
Precursors for Electronic Grade Zinc Oxide: Towards a
Systematic Ligand Design by Understanding Molecular
Decomposition Mechanisms
Rudolf C. Hoffmann[a] and Jörg J. Schneider*[a]
Keywords: Semiconductors / Ceramics / Zinc oxide / Thin films / Transparent conducting oxide
Zinc complexes with dimethyl 2-hydroxyimino- and 2-nitro-
malonate (dmm-NOH and Hdmm-NO2, respectively) were
synthesized and examined as potential precursors for nano-
crystalline zinc oxide. The functionalised 1,3-diketones are
sufficiently acidic to allow direct synthesis from hydrozincite
or zinc 2-propoxide as well as their full spectroscopic and
cordance with the obtained spectroscopic data. Thermogra-
vimetry coupled with mass spectrometry and infrared spec-
troscopy (TG/MS and TG-IR) indicate the formation of di-
methylcarbonate and methanol during the decomposition.
Both precursors exhibit good film formation properties and
the obtained zinc oxide was further characterized by XRD,
analytical characterization. Their nominal elemental compo- SEM, TEM and PL. Finally, the semiconducting electronic
sitions of [Zn4O(dmm-NO)6] and [Zn3(OH)4(dmm-NO2)2]
suggest the presence of multinuclear cage structures with
oxo- or hydroxo-bridging ligands. This assumption is in ac-
behaviour of the obtained nanocrystalline ZnO was studied
in a field-effect transistor (FET) device.
based molecular precursors.[10] The introduction of such a
tailored ligand substitution could offer a way to facilitate
the decomposition process while avoiding unwanted subli-
mation of the metal oxide precursors. Moreover, metal
complexes of 1,3-diketones can be synthesized by a large
variety of methods.
The metathesis reaction using metal salts and the 1,3-
diketone in the presence of a suitable base is a commonly
used procedure. Apart from that, the direct reaction of
metal alkoxides, carbonates, carbonyls or acetates with suf-
ficiently acidic 1,3-diketones has been described as a valu-
able route to these compounds. These routes seem particu-
larly attractive because no contamination of the precursors
by inorganic salt elimination reactions or other by-products
is expected.[11]
Our present study involved the synthesis and characteri-
zation of malonatozinc(II) complexes with functionalized
dimethylmalonato molecules as ligands. To the best of our
knowledge, this class of compounds has been neither pre-
viously described nor investigated for use as metal oxide
precursors. The decomposition pathways into ZnO were in-
vestigated by in situ thermogravimetric analysis coupled
with MS and IR. With respect to the influence of various
residuals on the backbone of the malonic acid ester ligand
and its subsequent thermal decomposition and ceramis-
ation, we were keen to find a correlation between the sub-
stituent and decomposition process in order to allow a sys-
tematic molecular precursor design based on these ligand
frameworks. Finally, the obtained nanoscale zinc oxide was
Introduction
Molecular precursors for solution deposition of oxide
thin films have to meet a number of specific require-
ments.[1–3] Precursors which are typically employed in gas-
phase deposition routes, e.g. alkoxides, carboxylates or 1,3-
diketones[4,5] and which often have advantages during gas
phase deposition such as low conversion temperatures and
low carbon residues in the final ceramic are, however, sub-
ject to limitations when employed in solution deposition
routes. Precursors which are suited for solution deposition
routes should allow for an appropriate wetting of substrates
and should yield homogeneous layers after necessary drying
and processing steps.[6,7] Furthermore, partial sublimation
of such precursors is undesired in a strict sense because
ceramic materials with a defined multicomponent or
multinary composition would therefore be inaccessible.[8,9]
The 1,3-diketone ligand framework offers a range of pos-
sibilities for synthetic modifications especially in the 2-posi-
tion (β-position) but also at the substituents of the carbonyl
atoms (γ-position). This allows tuning their properties, thus
rendering these ligand structures interesting for solution
[a] Fachbereich Chemie, Eduard-Zintl-Institut für Anorganische
und Physikalische Chemie, Technische Universität Darmstadt,
Alarich-Weiss-Str. 12, 64287 Darmstadt, Germany
E-mail: joerg.schneider@ac.chemie.tu-darmstadt.de
http://www.chemie.tu-darmstadt.de/schneider/startseite_aks/
index.de.jsp
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejic.201301525.
Eur. J. Inorg. Chem. 2014, 2241–2247
2241
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
characterized by microscopic (SEM, TEM), diffraction residual mass of the thermogravimetric measurements (Fig-
(XRD) as well as optical techniques (photoluminescence, ure 1). Based on these analytical and spectroscopic data,
PL) and was studied for its functional behaviour as a semi- the structures of the new compounds were assigned to be
conductor material in a field-effect transistor (FET) device. like those of frequently found and well established tri- and
tetranuclear zinc complexes.[13–16]
In order to understand the decomposition pathway in the
formation of ZnO we conducted in situ TG-MS and TG-
IR studies. The course of the thermal decomposition is pre-
Results and Discussion
sented in Figure 2. The decay of both (1) and (2) proceeds
in a stepwise manner. The decomposition of (1) starts at
about 140 °C and is complete at about 400 °C. In the case
of (2), the decomposition begins at about 170 °C and ex-
tends over a temperature range up to 450 °C. In both cases
only the first step resulted in a sharp maximum of the
Gram–Schmidt signal, which was positioned at 180 °C and
220 °C, respectively. The corresponding IR spectra are
shown in Figure 3 and indicate a mixture of various gas-
eous decomposition products. The signals at 670 and
2370 cm–1 could be attributed to carbon dioxide[17] and ad-
ditional peaks around 2220 cm–1 were assigned to carbon
monoxide.[18] The presence of a carbonyl compound was
revealed by the signal at 1780 cm–1 (Figure 3, a–b). In com-
bination with signals in the region between 500–2000 cm–1
(fingerprint region), this signal pattern could be clearly at-
tributed to dimethyl carbonate (Figure 3c; Figure S2).
Moreover, the signal at about 1030 cm–1 was assigned to
methanol (Figure 3, d; Figure S2). The identification of de-
composition products resulting from the nitro and hy-
droxyimino functionalities was less clear, since only weaker
signals remained unassigned so far. In the case of (1), a
small peak at about 1600 cm–1 was attributed to nitrogen
dioxide, whereas broader signals at 1350 and 1700 cm–1
The reactions of the functionalised dimethyl malonates,
dimethyl 2-hydroxyimino- and 2-nitromalonate (dmm-
NOH and Hdmm-NO2, respectively), with various zinc
compounds were investigated (Figure 1). The addition of
hydrozincite to Hdmm-NO2 in water led to the formation
of [Zn3(OH)4(dmm-NO2)2] (1). The corresponding reaction
with dmm-NOH, however, only resulted in a sparingly solu-
ble compound with unsatisfactory yields. In a second series
of reactions under inert conditions, the addition of zinc 2-
propanolate to Hdmm-NO2 led to a moisture sensitive
product which rapidly decomposed after exposure to air.
The reaction of zinc 2-propanolate with dmm-NOH in
THF yielded [Zn4O(dmm-NO)6] (2) which, in contrast, was
stable under ambient conditions and could be characterized
accordingly. Both (1) and (2) are readily soluble in alcohols
(up to 5 wt.-%) and very soluble in aprotic polar solvents
such as N,NЈ-dimethylformamide or dimethyl sulfoxide
(Ͼ 10 wt.-%).
Figure 1. Schematic presentation of the synthesis and thermal de-
composition of complexes [Zn3(OH)4(dmm-NO2)2] (1) and
[Zn4O(dmm-NO)6] (2).
Bekermann et al. reported the synthesis of bis(dimethyl-
malonato)zinc (3) from diethylzinc and dimethyl malonate
in hexane by heating to reflux for several hours or stirring
for a prolonged period of time.[12] The electronegative resi-
dues in dmm-NOH and Hdmm-NO2 lead to higher acidity
of the 1,3-diketones and allowed fast complex formation
even at room-temperature. In contrast to the well known
zinc acetylacetonate, the homologous complex (3) is moist-
ure sensitive and rapidly hydrolyses under ambient condi-
tions. This supports the suggestion that (1) and (2) are mul-
tinuclear complexes which display an oxo- or hydroxo-
bridging situation between the metal centres. The latter can
be easily identified by means of IR (Figure S1) and NMR
spectroscopy (see Experimental Section). Nominal compo-
Figure 2. Thermogravimetric mass loss and corresponding Gram–
Schmidt signal of (a) [Zn3(OH)4(dmm-NO2)2] (1) and (b)
sitions could be derived from the elemental analysis and [Zn4O(dmm-NO)6] (2).
Eur. J. Inorg. Chem. 2014, 2241–2247
2242
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
might indicate nitric acid (Figure 3, a).[19,20] For (2), no sig- of mass spectroscopy.[12,27] In the investigations described
nals revealing the decay of the hydroxyimino group could herein, once again no specific decomposition products of
be clearly assigned (Figure 3, b).
uncoordinated dimethyl malonate (e.g. keto com-
pounds[28,29]) were identified. The introduction of strong
electron-withdrawing groups into the framework of di-
methyl malonate in (1) and (2) led to a decrease of the de-
composition temperature in comparison with (3), i.e. about
180 °C.[12] A direct comparison between (1), (2) and (3) re-
mains difficult, though, due to differences in structure and
the varying number of oxo- or hydroxo co-ligands.
The solid residues which were obtained by the decompo-
sition of (1) and (2) were analysed by means of gracing
incidence XRD (GI-XRD). The diffractograms of such thin
films fabricated by spincoating of the molecular precursors
on quartz and subsequent annealing at various tempera-
tures are shown in Figure 4. Films deposited at lower tem-
peratures (250 °C) were found to be amorphous, whereas at
higher temperatures (350 and 450 °C) crystalline films of
zincite (JCPDS 36–1451) were formed. The XRD measure-
ments indicated preferential growth on the substrate normal
to the (002) plane of the zincite structure. The formation of
this type of texture in zinc oxide thin films is well known
for sol-gel processes as well as solution deposition.[30] Thus,
the decomposition of zinc acetate to zinc oxide is frequently
used for the formation of seed layers for epitactic growth of
ZnO on different substrates.[31,32] For ZnO films obtained
by Chemical Bath Deposition (CBD), preferential orienta-
Figure 3. Gas phase IR spectra corresponding to the maximum of
the Gram–Schmidt signal in Figure 1 of (a) [Zn3(OH)4(dmm-
NO2)2] (1) and (b) [Zn4O(dmm-NO)6] (2) as well as reference spec-
tra of (c) dimethyl carbonate and (d) methanol. (Spectra showing
the range from 4000 to 500 cm–1can be found in Figure S2).
The gaseous decomposition products mentioned which tion of zincite might occur directly near the substrate but
were already identified by means of IR spectroscopy could can be lost with greater distance from the substrate surface,
be further confirmed by additional experiments using TG- i.e. with increasing film thickness.[33] The nanocrystallinity
MS (Figure S3). Thus, besides the masses of species less
specific than carbon dioxide (m/z+ = 44), carbon monoxide
(m/z+ = 28), water (m/z+ = 18) or methanol (m/z+ = 32), the
characteristic fragmentation pattern of dimethylcarbonate
(m/z = 90, 59, 45, 31, 29) could be identified. For the de-
composition of (1), the masses of nitrogen dioxide (m/z+ =
46) and nitrogen monoxide NO (m/z+ = 30) were found,
whereas there was no indication of nitric acid (m/z+ = 63)
or nitrous acid (m/z+ = 47). During the decomposition of
(2), a mass of m/z+ of 27 was detected, attributable to hy-
drogen cyanide. This finding is in accordance with a
Beckmann fragmentation of the ligand framework responsi-
ble for the decay of the (CNO) moiety and an N–O bond
cleavage.[21] Although uncoordinated α-hydroxyketoximes
can undergo several types of rearrangement and fragmenta-
tion reactions,[22,23] the presence of neighbouring quater-
nary carbonyl carbon atoms should favour the latter.[24,25]
Detailed studies on 2-hydroxyimino 1,3-diketones are cur-
rently not available for comparison.
The decomposition of (1) and (2) leads to the formation
of dimethyl carbonate and methanol (Figure 1). The course
of the reaction could proceed by means of a pericylic pro-
cess as well as by defined intermediates.[26] Interestingly, the
decomposition of the zinc[12] and zirconium complexes[27]
of dialkyl malonates follows
whereby the elimination of the alkyl or alkoxy side groups
occurs first. Thereby the scaffold of the 1,3-diketones ini-
a different mechanism,
Figure 4. GI-XRD of films obtained by spincoating solutions of
(a) [Zn3(OH)4(dmm-NO2)2] (1) and (b) [Zn4O(dmm-NO)6] (2) fol-
lowed by calcination at various temperatures. The broad signal at
tially remains intact and can indeed be identified by means 2θ of about 22° refers to the quartz substrate.
Eur. J. Inorg. Chem. 2014, 2241–2247
2243
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
of the ZnO formed by ceramisation of complexes (1) and
(2) was also confirmed by means of HRTEM. The ZnO
particles are highly crystalline with a mean diameter of
about 10–15 nm for those derived from 2 and slightly larger
(20–30 nm) for those from complex 1 (see Figure 5 and Fig-
ure S4).
Figure 6. AFM topographies of films obtained by spincoating of
solutions of (a) [Zn3(OH)4(dmm-NO2)2] (1) and (b) [Zn4O(dmm-
NO)6] (2) and subsequent annealing at 350 °C [film thickness about
10 nm, RMS values (a) 0.3 nm and (b) 0.2 nm].
Functional Properties of the Precursor Derived
Semiconductor ZnO Films
Figure 5. (a) and (b) HRTEM images at different magnifications
of nanocrystalline zinc oxide obtained from the precursor
[Zn4O(dmm-NO)6] (2) after processing at 350 °C. (c) Correspond-
ing SAED pattern.
A series of FETs with bottom gate/bottom contact set-
up (Figure 7) were fabricated by spincoating solutions of
[Zn4O(dmm-NO)6] in methoxyethanol on prefabricated
substrates (support/gate highly n-doped silicon, dielectric
silicon dioxide and interdigital gold electrodes) followed by
For further characterisation, photoluminescence (PL)
spectra of ZnO films on quartz slides obtained from the
decomposition of [Zn4O(dmm-NO)6] were recorded (Figure
S5). The amorphous film from the decomposition at 250 °C
showed intense fluorescence which seems, however, to be
unrelated to zinc oxide. The films obtained at higher tem-
peratures both exhibited a peak at about 380 nm and a fur-
ther very broad signal with maxima at 584 and 597 nm. The
first is commonly attributed to the e–/h exciton, whereas the
latter refers to structural point defects within the zincite
structure.[34–36] These findings also support the view that
significant amounts of organic residues are still present at
lower calcination temperatures. The topography of the zinc
oxide thin films from the decomposition of (1) and (2) was
further analysed by means of AFM (Figure 6). Solutions of
the precursors display excellent wetting of the substrate
which results in complete surface coverage. After annealing,
the ceramic films are uniform and exhibit low surface
roughness. The formation of nitrogen oxides from (1) or
hydrogen cyanide from (2) was not recognizable during the
thermal processing and posed no problems.
Figure 7. (a) Schematic setup of FET device. (b) SEM micrograph
showing a top-view of the FET device structure at the intersection
from the channel area (left) to the gold electrode (right).
Eur. J. Inorg. Chem. 2014, 2241–2247
2244
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
annealing on a hotplate at temperatures from 250 to 450 °C. the preparation of zinc oxide thin films. If employed in a
Prior to the FET measurement in a glove box, an additional solution deposition route they are certainly advantageous
annealing step was carried out under an argon atmosphere over their unfunctionalised counterparts. The multinuclear
at 150 °C. This treatment served to eliminate adsorbed oxy- complexes which were synthesised in this work show good
gen and/or water.
solubility in organic solvents and can be employed under
No performance was observed for a processing tempera- ambient conditions (air and light). Moreover, they can be
ture of the ZnO films of 250 °C. This was in accordance cleanly converted into zinc oxide by thermal annealing
with the before-mentioned findings from GI-XRD and PL- without any sign of sublimation. The functionalisation of
spectroscopy which indicated the presence of amorphous 1,3-diketones in the 2-position thus offers a valuable way to
ZnO films still containing some organic residues. However, introduce a predefined bonding breaking point into the li-
the FET characteristics obtained by annealing the films at gand framework of such zinc complexes (“molecular zip-
350 °C are shown in Figure 8 with characteristic values of per”). The variation of the side groups in the 1- and 3-
μ = 5.2ϫ10–2 cm2/Vs, IOn/off ≈ 245,000 and a threshold positions of the malonic acid ester allows further tailoring
voltage Vth = +6.3 V. Based on these values the transistor of the solubility and decomposition behaviour of the corre-
behaviour showed both a reasonable saturation and Ion/off sponding zinc complexes. This ligands of this class are
ratio. Moreover, no hysteresis was observed. Processing at therefore promising candidates for the synthesis of various
450 °C lead to even better charge carrier mobility above other metal oxide precursors.
0.3 cm2/Vs but resulted in a decrease in the Ion/off ratio by
one order of magnitude(μ = 0.307 cm2/Vs, IOn/off ≈ 10,000,
Vth +5.9 V). The observed values and behaviour are thus
comparable with earlier reports for solution derived nano-
crystalline zinc oxide.[6,37]
Experimental Section
Ligand Synthesis: Dimethyl 2-(hydroxyimino)malonate (dmm-
NOH) was obtained by the reaction of dimethyl malonate with
sodium nitrite in acetic acid.[38] The crude liquid product contained
some unreacted dimethyl malonate and was purified by fractional
microdistillation (bp. of product fraction 95 °C at 0.3 mbar). The
product solidifies after cooling to room temperature. C5H7NO5
(161.1): calcd. C 37.27, N 8.69, H 4.38; found C 37.51, N 8.68, H
1
4.29. H NMR (500 MHz, [D3]chloroform, 25 °C): δ = 3.81, 3.84
(ϪCH3), 10.68 (br., ϪOH) ppm. 13C{1H} NMR(500 MHz, [D3]-
chloroform, 25 °C):δ = 53.09, 53.37 (ϪCH3); 143.71 (ϪC=NOH);
160.55, 161.04 ppm (ϪCOO) as mixture of two E/Z-diastereomers.
Dimethyl 2-Nitromalonate (Hdmm-NO2) was synthesized by the
reaction of dimethyl malonate with fuming nitric acid.[39,40] No side
products or contamination were observed and a pure product was
obtained after distillation (bp. 82 °C at 0.4 mbar). C5H7NO6
(177.1): calcd. C 33.91, N 7.91, H 3.98; found C 34.46, N 7.85, H
3.97. 1H NMR (500 MHz, [D3]chloroform, 25 °C): δ = 3.85
(ϪCH3); 5.81 [ϪC(NO2)H]. 13C{1H} NMR(500 MHz, [D3]chloro-
form, 25 °C): δ = 54.35 (ϪCH3); 87.82 [ϪC(NO2)H]; 159.87
(ϪCOO).
Zinc Oxide Precursor Synthesis: Synthesis of dimethyl 2-nitro-
malonate-ZnII [Zn3(OH)4(dmm-NO2)2] (1): Hdmm-NO2 (10 g,
56.8 mmol) was added to distilled water (200 mL). After addition
of hydrozincite (10 g, 91.1 mmol Zn), a yellow suspension was ob-
tained which was stirred overnight. After removal of solid residues
by filtration, the solvent was completely removed with a rotary
evaporator (65 °C, 130 mbar). The remaining yellow oil was dis-
solved in dichloromethane (50 mL) and added to dropwise to
methyl-tert-butyl ether (200 mL). The product was collected by fil-
tration, dried in vacuo and obtained in the form of a yellowish
powder (6.4 g, 34.2%). Ceramic yield (CY) / elemental analysis
(CHN): found CY 39.87%, C 19.86, N 4.50, H 2.78%. calcd. for
Figure 8. Performance of an FET derived from [Zn4O(dmm-NO)6]
(1) by annealing at 350 °C in air and subsequently at 150 °C in
argon. (a) Output characteristics obtained from variation of the
drain-source voltage from 0–25 V for gate-source voltages from 0–
25 V in 5 V steps. Data were acquired for increasing as well as
decreasing drain-source voltages. (b) Transfer characteristics for
constant drain-source voltage of 25 V. (μ = 5.2ϫ10–2 cm2/Vs;
IOn/off ≈ 245,000; Vth = +6.3 V).
1
Zn3(OH)4(C5H6NO6)2 CY 39.63%, C 19.49, N 4.54, H 2.62%. H
NMR (500 MHz, [D6]dimethyl sulfoxide, 25 °C): δ = 3.57 (ϪCH3),
3.40 (ϪOH) ppm. 13C{1H} NMR (500 MHz, [D6]dimethyl sulfox-
ide, 25 °C): δ = 50.91 (ϪCH3); 109.07 (ϪCϪNO2); 163.72 (ϪCOO)
ppm. IR: 3460 (v.br., νO–H) cm–1.
Conclusions
Synthesis of [2-Hydroxyimino-dimethylmalonato-ZnII] [Zn4O(dmm-
NO)6] (2): The following operations were carried out under inert
Molecular zinc(II) complexes of 1,3-diketones which are
functionalised in the 2-position are suitable precursors for conditions. Two separate solutions of zinc 2-propanolate (1.33 g,
Eur. J. Inorg. Chem. 2014, 2241–2247
2245
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
7.2 mmol) and dmmNOH (3.50 g, 21.7 mmol) were prepared in
tetrahydrofuran (10 mL, each). After cooling the zinc compound
to 0 °C, the ligand was added dropwise, whereby a clear yellow
solution was formed. During warming to room temperature pre-
cipitation occurred. The product was filtered and washed twice
with methyl-tert-butyl ether (10 mL). After drying in high vacuum
over several hours, a yellowish powder (1.84 g, 82.3%) was ob-
tained. The product was stable under ambient conditions. Ceramic
yield (CY)/elemental analysis (CHN): Found CY 26.50%, C 29.52,
N 6.60, H 2.98. calcd. for Zn4O(C5H6NO5)6 CY: 26.29%, C 29.10,
N 6.79, H 2.92. 1H NMR (500 MHz, [D6]dimethyl sulfoxide,
25 °C): δ = 3.69, 3.72 (ϪCH3) ppm. 13C{1H} NMR (500 MHz,
[1] R. M. Pasquarelli, D. S. Ginley, R. O’Hayre, Chem. Soc. Rev.
2011, 40, 5406–5441.
[2] M. A. M. Leenen, V. Arning, H. Thiem, J. Steiger, R. Ansel-
mann, Phys. Status Solidi A 2009, 206, 588–597.
[3] R. Parashkov, E. Becker, T. Riedl, H. Johannes, W. Kowalsky,
Proc. IEEE 2005, 93, 1321–1329.
[4] A. C. Jones, M. L. Hitchman, Chemical Vapor Deposition: Pre-
cursors, Processes and Applications, RSC Publishing, Cam-
bridge, 2008.
[5] L. G. Bloor, C. J. Carmlat, D. Pugh, Coord. Chem. Rev. 2011,
255, 12931–1318.
[6] R. C. Hoffmann, J. Dilfer, A. Issanin, J. J. Schneider, Phys.
Status Solidi A 2011, 208, 1708–1713.
[7] Y. Aksu, M. Driess, A. Merkulov, D. V. Pham, S. Schutte, J.
Steiger, Patent No. DE102008058040-A1; WO2010057770-A2.
[8] M. Tsaroucha, Y. Aksu, E. Irran, M. Driess, Chem. Mater.
2011, 23, 2428–2438.
[9] Y. Zhao, G. Dong, L. Duan, J. Qiao, D. Zhang, L. Wang, Y.
Qiu, RSC Adv. 2012, 2, 5307–5313.
[10] A. V. Kelin, A. Maioli, Curr. Org. Chem. 2003, 7, 1855–1886.
[11] C. Pettinari, F. Marchetti, A. Drozdov, in: Comprehensive Co-
ordination Chemistry II (Eds.: J. A. McCleverty, T. J. Meyer),
Elsevier, Amsterdam, 2003, vol. 1, p. 97–111.
[D6]dimethyl sulfoxide, 25 °C):
δ = 51.62 (ϪCH3); 143.36
(ϪC=NO); 162.70, 164.69 (ϪCOO) ppm.
Zinc Oxide Film Preparation: Quartz slides (2ϫ2 cm) were cleaned
by ultrasonication in 2-propanol and distilled water. Substrates
(1.5ϫ1.5 cm) for FET devices (interdigital structure obtained from
Fraunhofer IWS-Dresden) consisted of n-doped silicon with a
90 nm layer of SiO2, on which gold electrodes were deposited with
an intermediate adhesive layer of indium tin oxide (ITO). The
channel width of source/drain electrodes was W = 10 mm with a
channel length L = 10 μm.[6,37,41] FET substrates were cleaned by
consecutive ultrasonication in acetone, 2-propanol and deionised
water for 5 min and dried by purging with argon. The deposited
solutions contained 3 wt.-% of the precursor (1) or (2) in methoxy-
ethanol. Films were produced by spincoating (6 s at 1000 rpm fol-
lowed by 20 s at 2000 rpm) and heating on a hotplate in air for
4 min. Thicker films can be obtained by iteration of this coating
procedure.
[12] D. Bekermann, D. Pilard, R. Fischer, A. Devi, ECS Trans.
2009, 25, 601–608.
[13] R. W. Saalfrank, U. Reinmann, M. Göritz, F. Hampel, A.
Scheurer, F. W. Heinemann, M. Büschel, J. Daub, V. Schüne-
mann, A. X. Trautwein, Chem. Eur. J. 2002, 8, 3614–3619.
[14] W. Clegg, D. R. Harbrob, C. D. Homan, P. A. Hunt, I. A.
Little, B. P. Straughan, Inorg. Chim. Acta 1991, 186, 51–60.
[15] R. M. Gordon, H. B. Silver, Can. J. Chem. 1983, 61, 1218–
1221.
Materials Characterisation: TG/IR/MS measurements: TG 209N1
(Netzsch) coupled with a Nicolet iS10 spectrometer (Thermo-
Scientfic) and Aelos QMS 403C (Netzsch). IR: Nicolet 6700 (Ther-
moScientifc). Samples were measured as liquids or powders with
an attenuated total reflection (ATR) unit. XRD: STOE Stadi-P,
Cu-Kα radiation. PL spectra: Fluorolog-3 (Horiba) with Xenon
lamp. AFM: CP-II (Bruker-Veeco), 0.5 Hz, silicon cantilevers.
SEM: JSM-7600F (Jeol). TEM: CM20 (Philips, FEI) operated at
200 kV. Samples were supported on a lacey-carbon copper grid
(300 mesh).
[16] B. Luo, B. E. Kucera, W. L. Gladfelter, Polyhedron 2010, 29,
2795–2801.
[17] A. Chedin, J. Teffo, J. Mol. Spectrosc. 1984, 107, 333–342.
[18] G. E. Ewing, J. Chem. Phys. 1962, 37, 2250–2256.
[19] J. Laane, J. R. Ohlsen, in: Progress in Inorganic Chemistry (Ed.:
S. J. Lippard), John Wiley & Sons, 1980, vol. 27, p. 465–509.
[20] R. E. Nightingale, A. R. Downie, D. L. Rotenberg, B. Craw-
ford, R. A. Ogg, J. Phys. Chem. 1954, 58, 1048–1050.
[21] A. J. L. Pombeiro, V. Y. Kukushkinin, Comprehensive Coordi-
nation Chemistry II (Eds.: J. A. McCleverty, T. J. Meyer), Elsev-
ier, Amsterdam, 2003, vol. 1, p. 631–637.
[22] M. A. Dahl, M. Manpadi, L. J. Williams, Angew. Chem. Int.
Ed. 2013, 52, 2–33.
FET characteristics were determined with an HP 4155A Semicon-
ductor Parameter Analyzer (Agilent) in a glove box in the dark
under an argon atmosphere. FET devices were stored in darkness
for about 24 h prior to measurement. Charge carrier mobility μSAT
and the threshold voltage Vth were derived from a linear fitting of
[23] H. P. Fischer, C. A. Grob, E. Renk, Helv. Chim. Acta 1962, 45,
2539–2553.
[24] G. W. Adams, J. H. Bowie, R. N. Hayes, J. Chem. Soc., Perkin
Trans. 2 1991, 1809–1818.
[25] A. F. Ferris, J. Org. Chem. 1960, 25, 12–18.
[26] L. M. Fell, P. J. A. Ruttink, P. C. Burgers, M. A. Trikoupis,
J. K. Terlouw, Int. J. Mass Spectrom. 2000, 195–6, 85–99.
[27] R. Pothiraja, A. Milanov, H. Parala, M. Winter, R. A. Fischer,
A. Devi, Dalton Trans. 2009, 654–663.
the square root of the source drain current (IDS0.5) as a function of
[6,37,41]
gate source voltage VGS
.
Supporting Information (see footnote on the first page of this arti-
cle): IR spectra of zinc oxide precursors, gas phase IR spectra as
well as mass spectra from TG coupling, additional TEM images
and photoluminescence spectra of zinc oxide films.
[28] S. Tobita, K. Ogino, S. Ino, S. Tajima, Int. J. Mass Spectrom.
1988, 85, 31–42.
[29] S. Tobita, S. Tajima, Int. J. Mass Spectrom. 1991, 105, 101–
118.
[30] B. A. Kang, K. S. Hwang, J. H. Jeong, J. Sol-Gel Sci. Technol.
2007, 43, 145–149.
Acknowledgments
[31] L. E. Greene, M. Law, D. H. Tan, M. Montano, J. Goldberger,
G. Somorjai, P. Yang, Nano Lett. 2005, 5, 1231–1236.
[32] S. Xu, Y. Wei, J. Liu, R. Yang, Z. L. Wang, Nano Lett. 2008,
8, 4027–4032.
[33] P. Lipowsky, S. Jia, R. C. Hoffmann, N. Y. Jin-Phillipp, J. Bill,
M. Rühle, Int. J. Mater. Res. 2006, 97, 607–613.
[34] D. Behera, B. S. Acharya, J. Lumin. 2008, 128, 1577–1586.
[35] A. Van Dijken, E. A. Meulenkamp, D. Vanmaekelbergh, A.
Meijerink, J. Lumin. 2000, 87–89, 454–456.
This work has been supported by the Deutsche Forschungsgemein-
schaft (DFG) through the DFG SPP 1596 program and by the
Merck Lab (a cooperative Research Lab Alliance between Merck
KGaA and TU Darmstadt). The authors thank M. Kaloumenos
(FET, MerckLab, Darmstadt), Dr. P. Atanasova (PL, University of
Stuttgart, Germany) and Dr. J. Broetz (GI-XRD, TU Darmstadt,
Germany) as well as T. Heinlein (TEM, TU Darmstadt, Germany)
for measurements.
Eur. J. Inorg. Chem. 2014, 2241–2247
2246
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
[36] A. Van Dijken, E. A. Meulenkamp, D. Vanmaekelbergh, A.
Meijerink, J. Lumin. 2000, 90, 123–128.
[37] J. J. Schneider, R. C. Hoffmann, J. Engstler, O. Soffke, W. Jae-
[40]
D. I. Weisblat, D. A. Lyttle, J. Am. Chem. Soc. 1949, 71, 3079–
3081.
[41] R. C. Hoffmann, M. Kaloumenos, S. Heinschke, E. Erdem, P.
Jakes, R. A. Eichel, J. J. Schneider, J. Mater. Chem. C 2013, 1,
2577–2584.
germann, A. Issanin, A. Klyszcz, Adv. Mater. 2008, 20, 3383–
3387.
[38] H. Adkins, E. W. Reeve, J. Am. Chem. Soc. 1938, 60, 1328–
Received: December 4, 2013
Published Online: March 25, 2014
1331.
[39] D. I. Weisblat, D. A. Lyttle, United States Patent No. 2, 644,
838.
Eur. J. Inorg. Chem. 2014, 2241–2247
2247
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim