Description The present invention is in the field of gas phase processes for the production of conductive metal oxide films. The present invention further relates to conductive metal oxide films and electronic devices comprising such conductive metal oxide films.

The performance of electronic devices depends on the conductivity of its active parts. Inorganic thin layers are attractive due to their high conductivity and good stability. As electronic devices become smaller and smaller it is challenging to produce very thin inorganic films with sufficient conductivity with reasonable effort.

Kovalenko et al. (Science, 324 (2009), 1417-1420) have shown that conductive inorganic films can be made out of colloidal CdSe or Au nanocrystals from solution.

The conductivity of PbSe quantum dots deposited on a surface from solution could be increased by infilling the quantum dot films with amorphous alumina using atomic layer deposition as disclosed by Liu et al. (Nano Letters, 13 (2013), 1578-1587).

It was an objective of the present invention to provide a process which makes available conductive metal oxide films without any deleterious ligands. It was a further objective to provide a process for producing conductive metal oxide films with high conductivity, i.e. low resistivity. It was further aimed at a process which comprises few steps and is able to provide conductive metal oxide films of easily adjustable thickness. A further objective was to provide a process for producing conductive metal oxide films at high velocity on large scale.

These objectives were achieved by a process for producing a conductive metal oxide film comprising

(a) preparing a mixture comprising a metal precursor in the gaseous state and an oxygen- containing compound in the gaseous state,

(b) converting the mixture into particles which are in the aerosol state,

(c) depositing the particles from the aerosol state onto a substrate,

(d) coating the particles deposited on the substrate with a metal oxide. The present invention further relates to a conductive metal oxide film comprising particles which comprise a metal oxide and have a weight average particle size of 1 to 100 nm and a coating which is on the particles and comprises the same or a different metal oxide.

The present invention further relates to the use of the conductive metal oxide film according to the invention in electronic devices.

The present invention further relates to an electronic device comprising the conductive metal oxide film according to the present invention.

Preferred embodiments of the present invention can be found in the description and the claims. Combinations of different embodiments fall within the scope of the current invention.

According to the present invention a mixture comprising a metal precursor and an oxygen- containing compound in the gaseous state is prepared. The mixture can contain one or more than one metal precursor such as two or three. Independently, the mixture can contain one or more than one oxygen-containing compound such as two or three. The metal precursor contains one or more metals including earth alkaline metals such as Be, Mg, Ca, Sr, Ba; main group metals such as Al, Si, Ga, Ge, In, Sn, Tl, Pb, Bi; transition metals such as Sc, Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, Y, Zr, Nb, Mo, Tc, Ru, Rh, Pd, Ag, Cd, Hf, Ta, W, Re, Os, Ir, Pt, Au, Hg, Tl, Pb or Bi; lanthanoids such as La, Ce, Pr, Nd, Pm, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu. Preferably, the metal precursor contains Al, Si, Ca, Ti, V, Zn, Ga, Ge, Sr, Y, Zr, Nb, In, Sn, Ba, Ta, W or Pb. More preferably, the metal precursor contains Zn, In or Sn. The present invention has the advantage that metals can be chosen which are inexpensive and abundant to obtain metal oxide films of high conductivity. Any metal precursor which can be brought into the gaseous state is suitable. Preferably, the metal precursor is a metal-organic compound. Metal organic compounds include alkyl metals such as dimethyl zinc or dibutyl tin; metal alkoxylates such as diethoxy zinc or tetra-isopropoxy zirconium; cyclopentadiene adducts like ferrocene or titanocene; metal carbenes such as tanta- lum-pentaneopentylat or bisimidazolidinylenrutheniumchloride; metal halogenides such as tita- nium tetrachloride; carbon monoxide complexes like chromium hexacarbonyl or nickel tetracar- bonyl. More preferably, the metal precursor is an alkyl metal.

It is possible to use more than one metal precursor. Preferably, a second metal precursor is present at 1 to 30 mol-% with respect to the total molar amount of metal precursors, more pref- erably at 2 to 15 mol-%. In this case, metal-doped metal oxides are accessible, for example aluminum-doped zinc oxide, tin-doped indium oxide, or antimony-doped tin oxide. Alternatively, in order to obtain halogen-doped metal oxides it is possible to use a halogen-containing metal precursor or a halogen-containing compound in addition to the metal precursor preferably in an amount of 1 to 30 mol-% with respect to the total molar amount of metal precursor and halogen- containing compound, more preferably of 2 to 15 mol-%. Examples for such halogen-containing compounds are chlorine gas, ammonium fluoride or tin tetrachloride. In this case, halogen- doped metal oxides are accessible. According to the present invention the oxygen-containing compound can be chosen from a wide variety of compounds. It is possible to use a compound which only contains oxygen like elemental oxygen, an oxygen plasma, or ozone. It is further possible to use water or hydrogen peroxide. Elemental oxygen and/or water are preferred, in particular oxygen. It is further possible that the metal precursor contains oxygen. In this case the metal precursor also acts as oxygen- containing compound such that the mixture can contain only one compound. Furthermore, it is possible to use more than one oxygen-containing compound or a combination of oxygen- containing compound and a metal precursor which contains oxygen. In order to prepare the mixture the metal precursor and the oxygen-containing compound are brought into the gaseous state. This can be effected by heating and/or decreasing the pressure. Usually the metal precursor is brought into the gaseous state separated from the oxygen- containing compound and both are mixed in the gaseous state. Preferably, the pressure at the place where the metal precursor or the oxygen-containing compound are brought into the gase- ous state is from 10 ^"6 to 1500 mbar, more preferably from 10 ^"2 to 100 mbar, in particular from 0.1 to 10 mbar. It is advantageous if the vapor pressure of the metal precursor and the oxygen- containing compound is high. Preferably, the vapor pressure is at least 0.1 mbar at 25 °C, more preferably at least 1 mbar at 25 °C. Furthermore, it is advantageous if the metal precursor is stable upon heating. Preferably, the pristine metal precursor can be heated under inert atmos- phere without significant decomposition to at least 250 °C. Significant decomposition in the present context means that more than 20 wt.-% of the metal precursor is decomposed when kept at a certain temperature for one hour.

According to the present invention the mixture of the metal precursor and the oxygen-containing compound in the gaseous state typically comprises the metal precursor and the oxygen- containing compound in a molar ratio of 1 : 99 to 90 : 10, preferably of 2 : 98 to 60 : 40, more preferably of 5 : 95 to 30 : 70, in particular of 10 : 90 to 20 : 80.

According to the present invention the mixing of the metal precursor and the oxygen-containing compound in the gaseous state usually does not need any action due to the high diffusion of gases. It can be advantageous if this mixture additionally comprises an inert gas such as argon, nitrogen or carbon dioxide. Preferably, the mixture contains 20 to 95 mol-% of inert gas with regard to the total mixture, preferably 50 to 90 mol-%, in particular 70 to 85 mol-%. According to the present invention the mixture is converted into particles which are in the aerosol state. If the metal precursor and the oxygen-containing compound are sufficiently reactive, no action is required for this conversion. This is for example the case for most alkyl metals, such as diethyl zinc, if mixed with water. It is also possible to heat the mixture in order to effect the conversion, for example to a temperature of 25 to 500 °C, preferably of 100 to 300 °C. Normally it is enough to heat the mixture for a couple of seconds such as 1 to 10 s. Preferably, the conversion is effected by the application of a radiofrequency electric field as described for example by Mangolini et al. (Nano Letters 5 (2005) 655-659). The frequency of the electric field is gener- ally 1 to 100 MHz, preferably 10 to 50 MHz. The energy density exerted by the generator of the radiofrequency electric field is usually 1 to 2000 W/cm ³ , preferably 100 to 1500 W/cm ³ , in particular 300 to 1000 W/cm ³ . Often, only parts of the energy density actually reaches the mixture, for example 10 to 30 %. The residence time of the mixture in the zone of the radiofrequency electric field is generally 100 and 1 s, preferably 1 ms to 200 ms, in particular 10 ms to

50 ms. This time is calculated by the length of the electric field along the mixture flow axis divided by the mixture flow velocity. Usually, the longer the residence time in the zone of the radiofrequency electric field is the larger the particles become. The pressure at which the mixture is converted into particles is usually from 10 ^-6 to 1500 mbar, preferably from 10 ^"2 to 100 mbar, in particular from 0.1 to 10 mbar. Preferably, the pressure at which the mixture is converted into particles equals the pressure at which the metal precursor and/or the oxygen-containing compound is brought into the gaseous state.

According to the present invention the thus obtained particles are deposited onto a substrate. The deposition can be effected by various methods. Examples are diffusion of the particles, thermophoresis, electrophoresis, or inertial impaction. The choice of the method may depend on the choice of method with which the mixture is converted into particles in the aerosol state. Thermophoresis is preferred in the case where the conversion of the mixture to particles in the aerosol state is effected by heating. The substrate is then at a lower temperature, such as room temperature. The preferred temperature difference is 10 to 300 °C, more preferably 50 to 150 °C. In the case of electrophoresis, a voltage between the substrate and the place of particle generation is applied such as 10 to 5000 V, preferably 50 to 200 V. In the case of inertial impaction, the substrate is surrounded by a lower pressure than the aerosol comprising the generated particles. The pressure difference causes the aerosol to be accelerated towards the substrate onto which the particles are deposited. The pressure of the aerosol is preferably 2 to 10 ⁴ times the pressure surrounding the substrate, more preferably 10 to 1000 times the pressure surrounding the substrate, in particular 100 to 500 times the pressure surrounding the substrate. Inertial impaction is preferred.

The particles deposited on the substrate usually form a film. A film in the context of the present invention is an object with a length in one dimension which is at least ten times smaller than in the other two dimensions. The surface of a film can be rough or smooth. Smooth typically means that the thickness of the film at the place of minimum thickness is at least half the thickness of the film at the place of maximum thickness. Normally, a film resembles the structure of the substrate, e.g. if the substrate is bent, the film is also bent. A film can contain defects like holes. These defects, however, generally constitute less than half of the surface area covered by the film.

Upon deposition of the particles forming the film according to the present invention, the deposi- tion rate is preferably 1 to 200 nm/s, more preferably 10 to 100 nm/s. The deposition rate in nm/s refers to the thickness growth of the film upon deposition of the particles. It can be measured by measuring the thickness of the film by ellipsometry as described in PAS 1022 EN (Reference procedure for the determination of optical and dielectric material properties and the layer thickness of thin films by ellipsometry; February 2004) after depositing the particles for a defined time and then divide the measured thickness by the time. The dimensions of the film parallel to the substrate surface can be controlled by the dimension of the particle flux. If the particles are focused, e.g. by an orifice separating the volume in which the particles are generated and the volume in which the substrate is located. It is also possible to move the substrate such that larger areas of the substrate are hit by particles. Preferably, a large substrate can be moved relative to the particle flux such that the film covers the substrate completely, in particular in a roll-to-roll process.

According to the present invention the substrate can be any solid material. These include for example metals, semimetals, oxides, nitrides, and polymers. It is also possible that the substrate is a mixture of different materials. Examples for metals are aluminum, steel, zinc, and copper. Examples for semimetals are silicon, germanium, and gallium arsenide. Examples for oxides are silicon dioxide, titanium dioxide, and zinc oxide. Examples for nitrides are silicon nitride, aluminum nitride, titanium nitride, and gallium nitride. Examples for polymers are polyethylene terephthalate (PET), polyethylene naphthalene-dicarboxylic acid (PEN), polyolefines, pol- ycycloolefines and polyamides. Polymers are preferred. The substrate can have any shape. These include sheet plates, films, fibers, particles of various sizes, and substrates with trenches or other indentations. The solid substrate can be of any size. If the solid substrate has a particle shape, the size of particles can range from below 100 nm to several centimeters, preferably from 1 μηη to 1 mm. Figure 1 shows the schematic of a reactor which is suitable for producing particles in the aerosol state and depositing them on a substrate according to the present invention. The metal precursor and the oxygen-containing compound each with optionally an inert gas can be injected by the ports (a) and/or (b). Typcially the oxygen-containing compound is injected via port (a) and the metal precursor via port (b). Preferably, mass flow controllers are attached to the ports (a) and (b) to adjust the amounts of metal precursor and oxygen-containing compound. The pressure is controlled by a pressure gage (f). Any pressure gauge able to measure the pressure in the range given for the reaction above is suitable, such as a McLeod gauge, aneroid gauges equipped with a Bourdon tube, a diaphragm, a capsule, or a set of bellows, thermal conductivity gauges, e.g. a Pirani gauge, ionization gauges such as a Bayard-Alpert gauge.

The container in which the particles are formed (c) should be of an inert material such as glass, ceramics, or metal coated with noble metals such as gold. The container can have any form, preferably a tube with a length (I) chosen such that the reaction time at a given flow rate of metal precursor and oxygen-containing compound matches the values described above, such as 5 cm to 1 m, for example 15 cm to 50 cm. The diameter depends on the amount of material to be processed and the size of the substrate on which the film is deposited, e.g. the diameter can be 0.5 cm to 25 cm. If the conversion of the mixture comprising the metal precursor and the oxygen-containing compound is performed by a radiofrequency electric field, a radiofrequency electrode (e), e.g. a ring electrode, can be used. This electrode can be placed before, at or after the mixing zone. Preferably, the electrode is placed before the mixing zone. In this case it is preferable to place it around the port from which the less reactive reagent, usually the oxygen- containing compound, is injected. The electrode (e) is connected to a radiofrequency generator.

Preferably, the part of the container (c) furthest away from the port (a), can be capped with an orifice (m) to focus the particles and control the pressure inside and around the container (c). The orifice can have any shape, for example a round hole, a square hole or a rectangular slit. The orifice can also be a nozzle. A rectangular slit is preferred. The size of the slit depends on the pressure difference across the orifice and the desired particle flow rate. For example, a slit can be 1 mm to 1 m long and 0.1 mm to 3 mm wide, for example 10 to 100 mm long and 0.5 mm to 1.5 mm wide. The substrate (o) is placed in the distance (p) from the optionally capped end of the container (c), such as 1 mm to 10 cm, preferably 0.5 cm to 5 cm. The substrate (o) can preferably be moved. For large-area films the movement speed can be 1 m/s. The substrate (o) and the end of container are surrounded by a container (h). The container (h) can be of any material which is form stable and has a low vapor transmission rate, such as metals, glass or ceramics. Preferably, the container (h) is electrically conductive to avoid static charge. Therefore, the container (h) is preferably out of a metal or a material which has a conductive coating, e.g. a metal or a conductive polymer. It is even more preferably if the container (h) is conductive and electrically grounded. If larger areas of substrate are coated, the container (h) can contain several containers (c) in parallel.

The container (h) can be connected to a pressure gauge (f) as described above and a pump (r) in case the pressure around the substrate is below atmospheric pressure. Suitable pumps include membrane pumps, piston pumps, turbo pumps, cryo pumps and getter pumps. Turbo pumps are preferred.

Preferably, the particles deposited on the substrate are heated to a temperature of 150 °C to 600 °C, more preferably of 250 to 400 °C. Preferably, the temperature is applied for 10 minutes to 5 hours, more preferably from 30 minutes to 2 hours. It is possible to heat the particles on the substrate in an inert atmosphere, such as nitrogen or argon; in a reducing atmosphere, such as hydrogen or carbon monoxide; or in an oxidizing atmosphere, such as oxygen or air. It is preferable to mix an inert gas with a reducing or oxidizing gas, more preferably with a reducing gas, in particular hydrogen. Preferably, the inert gas contains 1 to 50 mol-% of the oxidizing or reducing gas with respect to the total gas mixture, preferably 5 to 40 mol-%, in particular 10 to 30 mol-%.

According to the present invention the particles deposited on the substrate are coated with a metal oxide. Various coating techniques can be used. Preferably gas phase coating techniques like chemical vapor deposition (CVD) or atomic layer deposition (ALD) are used. The chemical vapor deposition of metal oxides including reaction parameters and precursor selection is de- scribed by Kuzminykh et al. (Surface & Coatings Technology, 230 (2013) 13-21 ). ALD is a technique in which a series of self-limiting surface reactions are conducted which builds up confor- mal coatings of precise thickness depending on the number of self-limiting reactions performed. The process is described in detail by George (Chemical Reviews 1 10 (2010), 1 1 1 -131 ). ALD is preferred. Preferably, the coating material is a metal oxide different to the metal oxide of the particles deposited on the substrate.

In an ALD process a metal precursor as described for the particle generation can be used. The ALD metal precursor is deposited on the surface of the particle film by exposing the metal oxide film to the ALD metal precursor in the gaseous state. The ALD metal precursor reacts with the functional groups on the accessible surface of the particles. In this way typically a monolayer or a submonolayer of ALD metal precursor on the particles is formed. A subsequent treatment with a reactant converts the ALD metal precursor to a metal oxide, for example oxygen, ozone, plasma like oxygen plasma, water or hydrogen peroxide. The sequence of ALD metal precursor deposition and conversion to metal oxide, also referred to as ALD cycle, is normally performed for 10 to 1000 times, preferably for 25 to 500 times, in particular for 50 to 200 times. This usually generates coatings of a thickness of 1 to 100 nm, preferably of 3 to 50 nm, in particular of 5 to 20 nm which can be measured by high-resolution electron microscopy of a fractured sample. Typical pressures at which the ALD process is performed range from 100 to 10 ^"5 mbar, prefera- bly from 10 to 10 ^"3 mbar, more preferably from 1 to 0.1 mbar. The temperature for the ALD process is in the range of -20 to 500 °C, preferably 0 to 300 °C, in particular 50 to 200 °C. Typically, the reaction of a precursor with the surface in one ALD cycle takes 1 ms to 30 s, preferably 10 ms to 5 s, in particular 50 ms to 1 s. It is preferable to purge the substrate with an inert gas in between treating the surface with different precursors, normally for 0.1 s to 10 min, preferably for 1 s to 3 min, in particular for 10 s to 1 min.

It is possible to use the same ALD metal precursor in all ALD cycles or different ones in different cycles. If different ALD metal precursors are used doped metal oxide layers can be generated. Preferably, 99 to 70 % of the ALD cycles are performed with one ALD metal precursor and 1 to 30 % of the ALD cycles are performed with a different ALD metal precursor, more preferably 98 to 90 % of the ALD cycles are performed with one ALD metal precursor and 2 to 10 % of the ALD cycles are performed with a different ALD metal precursor. Examples for combinations of different ALD metal precursors are Al and Zn precursors, In and Sn precursors, Sn and Sb precursors.

Alternatively, different ALD metal precursors can be used in each ALD cycle simultaneously in order to produce doped metal oxide coatings. In this case a ALD metal precursor mixture is prepared preferably containing 99 to 70 mol-% with regard to the total ALD metal precursor mixture one ALD metal precursor and 1 to 30 mol-% of a different ALD metal precursor, more pref- erably 98 to 90 mol-% of one ALD metal precursor and 2 to 10 mol-% of a different ALD metal precursor. The same ALD metal precursor combinations as described above can be used. Preferably, in each ALD cycle one ALD metal precursor is used. The process according to the present invention yields conductive metal oxide films. Therefore the present invention further relates to conductive metal oxide films. The conductive metal oxide films have a resistivity of 10 ^-6 to 100 Ω-cm, preferably of 10 ^"5 to 1 Ω-cm, in particular 10 ^"4 to 10 ² Ω-cm. The resistivity is preferably measured according to the Van der Pauw method (Philips Research Reports 13 (1958) 1-9).

The conductive metal oxide film according to the present invention is preferably transparent. Transparent in the present context means that the conductive metal oxide film transmits at least 50 % of the intensity of light at a wavelength of 550 nm shined on the film parallel to the surface normal, more preferably at least 70 %, in particular at least 90 %.

The conductive metal oxide film according to the present invention normally has a thickness of 0.01 to 1000 μηη, preferably 0.1 to 100 μηη, more preferably 0.2 to 20 μηη, in particular 0.5 to 5 μηη.

The conductive metal oxide film according to the present invention comprises metal oxide particles having a weight average particle size of 1 to 100 nm, preferably 3 to 60 nm, in particular 5 to 30 nm. The film thickness and the particle size can be measured by cross-sectional SEM image analysis using the static image analysis method according to ISO 13322-1 (Static image analysis methods, 2004).

The particles comprise metal oxide including earth alkaline metal oxides such as BeO, MgO, CaO, SrO, BaO; main group metal oxides such as AI2O3, S1O2, Ga203, Ge02, Ιη2θ3, Sn02, TI2O, PbO, Pb0 ₂ , Bi ₂ 0 ₃ ; transition metal oxides such as Sc ₂ 0 ₃ , T1O2, V ₂ 0 ₅ , Cr0 ₂ , MnO, Mn ₂ 0 ₃ , FeO, Fe ₃ 0 ₄ , Fe ₂ 0 ₃ , CoO, Co ₂ 0 ₃ , NiO, Ni ₂ 0 ₃ , Cu ₂ 0, CuO, ZnO, Y2O3, Zr0 ₂ , Nb ₂ 0 ₅ , MoO, M0O2, Tc, Ru0 ₂ , Rh ₂ 0, PdO, Ag ₂ 0, CdO, Hf0 ₂ , Ta ₂ 0 ₅ , W0 ₃ , Re0 ₃ , Os0 ₄ , Ir0 ₂ , Pt0 ₂ , AuO, Hg ₂ 0; lantha- noid oxides such as La203, Ce203, Ce02, Pr203, Nd203, Prri203, Srri203, EU2O3, Gd203, Tb203, DV2O3, H02O3, Er203, Trri203, Yb203, LU2O3. Preferably, the particles comprise AI2O3, S1O2, CaO, T1O2, V2O5, ZnO, Ga ₂ 0 ₃ , Ge0 ₂ , SrO, Y2O3, Zr0 ₂ , Nb ₂ 0 ₅ , ln ₂ 0 ₃ , Sn0 ₂ , BaO, Ta ₂ 0 ₅ , W0 ₃ or Pb02. More preferably, the particles comprise ZnO, Ιη2θ3 or Sn02.

In order to increase the conductivity, the particles preferably comprise a metal oxide which is doped. Suitable dopants are other metals or halogens. If the particles comprise ZnO, it is more preferred if the dopant is Al, Ga, or In. If the particles comprise Sn02, it is more preferred if the dopant is F, CI, Br, I, As, Sb or Bi. If the particles comprise Ιη2θ3, it is more preferred if the dopant is Ge, Sn, Pb, As, Sb or Bi.

According to the present invention the film of particles comprising a metal oxide has a coating of the same or a different metal oxide on top of the particles. The definitions, examples and pre- ferred examples as described for the metal oxide comprised in the particles applies for the metal oxide comprised in the coating. Preferably, the coating is conformal. More preferably, the conformal coating comprises a different metal oxide than the particles. The coating preferably has a thickness of 1 to 100 nm, more preferably of 3 to 50 nm, in particular of 5 to 20 nm. The thickness can be measured by ellipsometry.

The conductive metal oxide film is particularly suitable for the use in electronic devices. There- fore, the present invention further relates to electronic devices comprising the conductive metal oxide film according to the present invention. Examples for electronic devices are power generators such as solar cells, illumination devices such as light emitting diodes, transistors such as thin film transistors or displays (television, computer monitor, smart-phone touch display). The conductive metal oxide film can serve different purposes in electronic devices, e.g. as electrode in a solar cell or as active material in a transistor. It is also possible to use the conductive metal oxide film on glass to change its infrared absorption and reflection characteristics for application in architectural glass or as a transparent contact in electrochromic smart windows.

The conductive metal oxide films according to the present invention are free of any potentially disturbing materials from precursors, show high conductivity and are very stable upon operation in an electronic device, and can be made easily in a large scale.

Description of the Figures Figure 1 schematically depicts a reactor suitable for the process according to the present invention. The references stand for: (a) top port; (b) side port; (c) discharge tube; (d) aerosol comprising particles; (e) radiofrequency ring electrode; (f) pressure gage; (g) plasma; (h) grounded metal chamber; (i) aerosol comprising particles; (k) dimension of plasma; (I) distance between reaction zone and orifice; (m) orifice; (n) particle beam; (o) substrate; (p) distance between ori- fice and substrate; (q) reciprocal substrate motion; (r) to pump.

Figure 2 depicts the measured voltage drop AV as a function of sourced current l _a for (a) example 3 and (b) example 4; as well as the hall voltage divided by sourced current RH as a function of magnetic field B for (b) example 3 and (d) example 4.

Figure 3 depicts cross-sectional scanning electron microscopy images of (a) example 1 , (b) example 3, (c) example 2, (d) example 4; as well as (f) a high angle annular dark field transmission electron microscopy image of example 4 and (e) an energy-dispersive x-ray spectroscopy map from a transmission electron microscope of the area indicated by the black box in panel (f) wherein the dark regions indicate aluminum and the bright regions zinc.

Figure 4 depicts the Fourier transform infrared absorption spectra for [1] example 3 and [2] example 1 on a NaCI substrate, v is the wavenumber and I the measured intensity in an arbitrary unit. The dashed line [3] is the modeled free carrier absorption of small ZnO nanocrystals em- bedded in AI2O3 with electron concentration of 5.5 ^■ 10 ¹⁹ cm ^"3 and mobility of 21 cm ² V- ¹ S ^"1 . The peak labeled with an ^* is CO2 in the spectrometer ambient. The large feature between 1000 and 500 cm- ¹ in the curve for [1 ] originates from aluminum oxide. Figure 5 depicts X-ray diffraction patterns of [1] example 1 , [2] example 2, [3] example 3, [4] example 4, 2Θ is the beam angle to the surface normal and Inorm is the normalized intensitiy. The arrows indicate the peaks of hexagonal zincite wherein the corresponding crystal planes are indicated by the numbers in round brackets.

Figure 6 depicts ultraviolet-visible-near infrared absorption spectra for [1 ] a 165 nm film of aluminum-doped zinc oxide, [2] a 400 nm film produced according to example 6 and [3] a 190 nm film of indium-doped tin oxide, λ is the wavelength and T the transmission. Examples

General Procedure

The reaction is performed in a reactor for which a schematic is presented in figure 1 . Diethyl zinc (DEZ, Sigma-Aldrich, Saint Louis MO) and molecular oxygen (O2) were fed into the reactor controlled with mass flow controllers (MKS 1 179 series and Edwards model 825 series B). DEZ, which is a volatile liquid at room temperature, was fed into the plasma using a bubbler with Ar as the carrier gas. DEZ and O2 were fed into the reactor by two different gas streams. The DEZ/Ar mixture was injected through the side port. The oxygen was diluted with argon and fed into the plasma through the top port. The feed rates into the reactor were: FDEZ = 4 SCCM, F _Ar ,DEz = 20 SCCM, F02 = 30 SCCM, F _Ar ,o2 = 250 SCCM. SCCM stands for standard cubic centimeter per minute (cm ³ min- ¹ ) at 273 K and atmospheric pressure. The feed rate of DEZ was calculated using the feed rate of argon into the DEZ bubbler, DEZ vapor pressure at the bubbler temperature (298 K) and the measured total bubbler pressure (Baratron model 727A, MKS, Andover MA) using the ideal gas law and conservation of mass.

The plasma inside the reactor was generated in a 2.54 cm glass tube (1 .90 cm inner diameter) with a top port with an outer diameter of 0.95 cm and a side port with an outer diameter of 0.63 cm. The plasma was generated by applying a radiofrequency (RF) field at 13.56 MHz, through a custom impedance matching network, to a copper ring wrapped around the top port just above the mixing zone. The pressure in the plasma (P _u ) was 4.8 mbar to 7.5 mbar. The residence time of the mixture in the plasma was approximately 27 to 42 ms. The quick disconnect vacuum fitting forming the seal with the chamber of the deposition stage served as a ground electrode. The plasma extended approximately 10 cm downstream of the mixing zone where the top port and side port meet.

Approximately 30 cm downstream of the mixing zone was a rectangular slit shaped orifice with a cross-sectional area of 0.8 mm x 15 mm. The slit was 2 mm long for the plasma pressure of 4.8 mbar, and 10 mm long for the plasma pressure of 7.5 mbar. The deposition stage down- stream was a chamber evacuated by a Leybold Turbovac 1000C turbo pump. The pressure in this chamber was Pdown = 1 .9 - 10 ^-3 mbar, thus the pressure ratio across the orifice was 257 to 400 depending on the upstream pressure. In the deposition stage chamber, the substrates were mounted to a push-rod (CRPP-1 , MDC vacuum, Hayward CA) moved back and forth under the particle beam in a reciprocating motion. The frequency of the reciprocating motion was approximately 1 Hz, and the stroke length approximately 4 cm. Two substrates were deposited on simultaneously, one polished silicon and the other Corning Eagle XG borosilicate glass (MTI Corporation, Richmond CA). Each substrate was approximately 1 x 1 cm. The glass substrates were used for electrical and ultraviolet- visible-near infrared transmission spectroscopy. The silicon substrates were used for x-ray diffraction (XRD), spectroscopic ellipsometry, and electron microscopy. Fourier-transform infrared transmission spectroscopy was performed on polished NaCI substrates.

Example 1 (Comparative)

A film was produced as described in the general procedure. Example 2 (Comparative)

A film as obtained in example 1 was annealed at 340 °C for 60 min in a custom tube furnace under a gas flow of Ar containing 20 mol-% hb with regard to the total gas phase. The tube had an outer diameter of 2.5 cm, and the total gas flow rate was 0.5 liters per minute. The sample was inserted into the furnace when it was at room temperature. Approximately 30 minutes were allowed to pass at room temperature to purge out air that may have diffused into the furnace during sample insertion. Power was then turned on, and the furnace was heated up to 340 °C over a period of approximately 25 minutes. After 60 minutes at 340 °C under flowing gas the power was turned off. The furnace was allowed to cool at the natural cooling rate of the system, which took approximately 3 hours. The sample was removed once the temperature was below 40 °C.

Example 3 (Inventive) A film as obtained in example 1 was coated with AI2O3 by atomic layer deposition carried out in a commercial reactor (Savannah S200, Cambridge Nanotech/Ultratech, Waltham MA). All depositions were carried out at a temperature of 180 °C. Before precursor pulsing, the film sample was allowed to outgas for 2000 seconds by sitting in the chamber at the deposition pressure of approximately 0.7 mbar under 20 SCCM flowing nitrogen (also used as carrier gas during depo- sition). The sample was exposed to trimethyl aluminum vapor for 0.1 s, the reactor was than purged for 30 s with nitrogen, whereupon the sample was exposed for 0.1 s with water vapor followed by purging with nitrogen for 30 s. This sequence was successively performed for 100 times. The samples were removed from the reactor after coating and allowed to cool in air. Example 4 (Inventive)

A film as obtained in example 2 was coated with AI2O3 by atomic layer deposition following the same procedure as for example 3.

Example 5 (Comparative)

The film obtained in example 2 was exposed to the conditions under which atomic layer deposition was performed in example 3 with the difference that no precursors were injected into the reactors.

Example 6 (Inventive)

A film as obtained in example 2 was coated with AhZnO by atomic layer deposition analogous to example 3 with the difference that instead of trimethyl aluminum diethyl zinc is used. In every 20th sequence, however, trimethyl aluminum is used and not dimethyl zinc.

Characterization Film thickness and ZnO volume fraction were measured by spectroscopic ellipsometry using a J.A. Woollam M44. The effective medium approximation (Bruggeman type) was used with the bulk optical constants of ZnO to model the data to extract the ZnO volume fraction and film thickness. The spectra were fit in the transparent wavelength region between 450 nm and 750 nm where no size effects on the optical constants of ZnO are expected.

The crystallite size was calculated by measuring the full-wide-at-half maximum (FWHM) of X-ray diffraction peaks, subtracting the instrument broadening, and then using the Scherrer equation (A. L. Patterson, Physical Reviews 56 (1939) 972-977). The diffraction patterns for examples 1 to 4 are depicted in figure 5.

Electrical resistivity measurements were carried out in a probe station (Janis Research, Wilmington MA) connected to an Agilent 4155C semiconductor parameter analyzer. All reported resistivity values were measured at room temperature and atmospheric pressure. Silver electrodes, 500 nm in thickness, were deposited on top of the films by thermal evaporation at a rate of 2 nm s ^_1 . For low resistivity samples, the Van der Pauw electrode configuration was used to measure the sheet resistance. For high resistivity samples, the electrode configuration was parallel stripes, 1 .0 x 0.1 cm positioned at the edges of the 1 x 1 cm glass substrate. The results are summarized in the following table. Example dfiim / μηη Xs/ nm Φζηθ p / Ω-cm

1 0.48 6.5 ± 1 34 % >10 ⁹

(comparative)

2 0.77 27 ± 5 33 % 2.9 ^■ 10 ⁶

(comparative)

3 0.48 6.5 ± 1 34 % 8.1 ^■ 10°

(inventive)

4 0.63 24 ± 3 31 % 1 .3 ^■ 10- ¹

(inventive)

5 0.99 5.1 ± 1 26 % >10 ⁹

(comparative)

6 0.40 4 ^■ 10- ³

(inventive)

In this table, dfiim is the film thickness, x _s is the particle size measured by XRD, φζηο is the ZnO volume fraction and p is the electrical resistivity. The current-voltage curves and the hall voltage curves for examples 3 and 4 are depicted in figure 2.

Scanning electron microscopy was performed in a JEOL 6700 FEG operating at 5 keV accelerating voltage. The samples were prepared by cleaving the silicon substrate on which the nano- crystal thin film was deposited, and then mounting the sample edge-on so the surface was parallel to the electron beam. Ultraviolet-visible-near infrared spectroscopy was carried out in a Cary 3000 (Agilent Technologies, Santa Clara CA) in transmission mode using a blank glass substrate for the baseline. Figure 3 depicts corresponding images of the examples 1 to 4.

Fourier transform infrared spectroscopy (FTIR) was carried out in a Nicolet Magna IR 550 (Thermo Scientific, Madison Wl) in transmission using a blank NaCI substrate as the back- ground. The sample compartment was allowed to purge for 2.5 minutes with the treated air before acquiring data. Figure 4 depicts the FTIR spectra of example 1 and 3 including a comparison to a calculated AI2O3 film.

Ultraviolet-visible-near infrared spectroscopy was carried out in a Cary 3000 (Agilent Technolo- gies, Santa Clara CA) in transmission mode using a blank glass substrate for the baseline. The spectra of example 6, a 165 nm aluminum-doped zinc oxide film and a 190 nm indium-doped tin oxide film are depicted in figure 6. It can be seen that the film according to the invention reach the transparency of conventional transparent conductive films.