Plasma-Enhanced Pulsed Laser Deposition of copper oxide and zinc oxide thin ﬁlms

Plasma-Enhanced Pulsed Laser Deposition (PE-PLD) is a technique for depositing metal oxide thin ﬁlms that combines traditional PLD of metals with a low-temperature oxygen background plasma. This proof-of-concept study shows that PE-PLD can deposit copper oxide and zinc oxide ﬁlms of similar properties to ones deposited using traditional PLD, without the need for substrate heating. Varying the pressure of the background plasma changed the stoichiometry and structure of the ﬁlms. Stoichiometric copper oxide and zinc oxide ﬁlms were deposited at a pressure of 13 and 7.5 Pa respectively. The deposition rate was approximately 5 nm/min and the ﬁlms were polycrystalline with a crystal size in the range of 3 - 15 nm. The dominant phase for ZnO was (110) and for CuO they were (020) and (111), where (020) is known as a high-density phase not commonly seen in PLD ﬁlms. The resistivity of the CuO ﬁlm was 0.76 ± 0.05 Ω cm, in line with ﬁlms produced using traditional PLD. Since PE-PLD does not use substrate heating or post-annealing, and the temperature of the oxygen background plasma is low, deposition of ﬁlms on heat-sensitive materials such as plastics is possible. Stoichiometric, amorphous zinc oxide and copper oxide ﬁlms were deposited on polyethylene (PE) and polytetraﬂuoroethylene (PFTE).


I. INTRODUCTION
Pulsed Laser Deposition (PLD) is a well-established and widely used deposition technique for e.g. dielectric, ferroelectric and magnetic oxide thin films. In particular, metal-oxide widebandgap group II-VI semiconductors are widely studied 1,2 . One of the main advantages of PLD as a deposition technique is the ability to achieve stoichiometric transfer of material from target to substrate. However, in practice, for metal oxides a background atmosphere of oxygen gas is often needed to avoid oxygen-deficient films that are formed under vacuum conditions 3,4 . In addition, the targets required are of the same complexity as the desired film, making the manufacturing of targets more demanding, compared to pure metal targets. Finally, like many other deposition techniques, PLD often requires elevated substrate temperatures, or post-annealing processes, to achieve high-quality films 3,5 , preventing direct deposition on heat-sensitive substrates like plastics.
In this paper, a proof-of-concept study for a modified version of PLD, Plasma-Enhanced PLD (PE-PLD) 6 , is presented, which aims to overcome some of the limitations of standard PLD, i.e. the need for multi-element targets and elevated substrate temperatures, for deposition metal-oxide films. The main idea is to combine a standard PLD set up using a metal target with an electricallyproduced low-temperature background oxygen plasma. In this way, the sources for metal and oxygen in the deposited film are separated, similar to the approach taken in reactive magnetron sputtering techniques 7 . Our previous modelling investigations indicate that metal atom densities in the order of 10 14 -10 15 cm −3 can be expected in the plasma plume in front of a substrate a few centimeters from the target 6 . In addition, a low-pressure rf-driven Inductively Coupled Plasma (ICP) used as the background plasma can provide reactive oxygen species, e.g. O and O * 2 , at densities of approximately 10 14 -10 15 cm −3 , depending on operating conditions 6 . Since these densities are similar, it seems feasible that significant interaction between the plasma plume and background plasma is possible, resulting in deposition of both metal and oxygen at comparable rates. Preliminary investigations of very thin films, 25-50 nm, deposited onto quartz and analysed with Medium Energy Ion Scattering showed broadly stoichiometric films for ZnO and Cu 2 O 8 , but no further characterisation of the film properties was done, nor any variation of deposition parameters. In addition, the oxygen plasma not only provides the oxygen atoms for the thin film, it also provides (chemical) energy to the substrate to assist the growth of the film without additional heating of the substrate. Importantly, the oxygen plasma is a non-equilibrium, pulsed plasma to ensure that its temperature remains low 9 , eliminating significant conductive heating of the substrate from the plasma, potentially allowing deposition of films onto sensitive substrates. The current paper aims to provide a more comprehensive proof-of-concept for PE-PLD of copper oxide and zinc oxide films, focusing not only on stoichiometry, but also on film structure, morphology and film resistance. In addition, a small range of substrate materials was investigated, in particular heat-sensitive materials.
The combination of a standard PLD plume with a secondary plasma has been reported in literature before. Notably, Dinescu et al., developed a plasma beam assisted PLD system in which a Zn target was ablated in standard PLD, with an additional oxygen plasma beam source also impinging on the substrate, creating ZnO films 10,11 . The oxygen plasma beam was generated in a separate chamber, with a beam of plasma flowing from this chamber onto the substrate, where this afterglow oxygen plasma interacted with the PLD plume and the growing film on the substrate.
With this system, in combination with substrate heating to 800 K, they achieved high-quality ZnO films. Basillais et al. followed a similar approach for the deposition of AlN films 12,13 . A pure Al target was ablated by a laser, while a nitrogen plasma was created in a separate chamber after which it flowed into the ablation chamber, interacted with the metal plasma plume and thin film growth was achieved on heated substrates.
In our work, a similar approach is followed for separating the source of metal and oxygen to deposit metal oxide films, however in our work no substrate heating is applied. The idea is that by using an active plasma that is in direct contact with the ablation plume, instead of an afterglow plasma beam, more reactive and energetic plasma particles impinge to the substrate. The higher energy of these particles, compared to an afterglow plasma or a neutral background gas, means that the diffusion length of these species is longer, resulting in more crystalline films 14 . In other words, the energy needed for good surface diffusion is provided by the background plasma rather than the heated substrate. In addition, the background plasma provides particles to the substrate for a much longer time that a laser-produced plasma plume. Tricot et al. showed that when using a pulsed-electron beam deposition system, plasma species are being delivered to the substrate for times much longer than in conventional PLD, resulting in polycrystalline films deposited at room temperature, since the probability for an incoming particle to find a good site on the surface for crystalline growth is increased 15 . In our case, it is only the oxygen species that are delivered over a much longer time scale, but nevertheless, it can be anticipated that this can promote (poly)crystalline growth.
Huang et al., report ZnO films deposited at room temperature with their RF-PEPLD system 16 and De Giacomo et al., investigated TiO 2 films produced by Plasma-Assisted PLD 17 . The layout of both these systems was very similar to ours, however, for both studies, the target was a metal oxide, not a pure metal as in our case. The reason is that the main focus of their work was to reduce droplets in the PLD plume by using a plasma background, not use the background plasma to supply the oxygen for the film. They both showed that droplet contamination is reduced by the plasma, but at the same time, the PLD plume was still capable of depositing high-quality films.
Investigations with the aim of room-temperature deposition of metal oxide films from a metal target have not been reported to our knowledge.
The materials chosen for our proof-of-concept study are copper oxide and zinc oxide. There are two common forms of copper oxide: cuprous oxide (Cu 2 O) and cupric oxide (CuO). Both are p-type semiconductors with a bandgap of 1.9 -2.1 and 2.1 -2.6 eV respectively 18 . Cu 2 O films are mainly investigated for applications in thin-film transistors (TFTs) 18 and solar cells 19 . CuO thin films find applications in gas sensors 20 and supercapacitors 21 . One of the main issues in this field is often the poor quality of the films and the high substrate temperatures that are needed for deposition. In addition, controlling the stoichiometry and obtaining single phase CuO or Cu 2 O has proven to be challenging 18 .
Zinc oxide, ZnO, is an n-type wide-bandgap semiconductor with a direct bandgap of approximately 3.3 eV that finds applications in electronic displays, thin film transistors and solar cells [22][23][24][25] . Background oxygen gas pressure is known to play a crucial role in determining the surface, optical and electrical properties of the resulting films, with the optimum pressure often being different for different types of properties 22 .
In this paper, a proof-of-concept study for PE-PLD of copper oxide and zinc oxide films is presented. The experimental system making the films is a combination of a standard PLD setup with a pulsed Inductively Coupled Plasma (ICP). The deposited films are characterised using Scanning Electron Microscopy (SEM), Tunneling Electron Microscopy (TEM), Selected Area Electron Diffraction (SAED), Energy Dispersive X-ray spectroscopy (EDX), X-ray Diffraction (XRD) and four-point probe resistivity measurements.

A. Stoichiometry
The stoichiometry of films deposited with the PE-PLD experimental system was investigated as a function of oxygen ICP pressure. Since all the oxygen in the resulting films comes from the ICP, it is likely that ICP pressure is the most sensitive control parameter for stoichiometry.
Copper oxide thin films were deposited on quartz substrates, from a copper target in an oxygen ICP background, for oxygen pressures ranging from 4 to 25 Pa. Zinc oxide thin films were deposited on quartz substrates, from a zinc target with ICP oxygen pressures between 7.5 and 13.5 Pa. The pressure range for copper oxide was chosen to be wider than zinc oxide since more stoichiometry The EDX results show that for both CuO and ZnO it is possible to deposit stoichiometric films from pure metal targets using the PE-PLD technique.

B. Structural characterisation of films
In order to determine the crystal structure of the copper oxide and zinc oxide films described in the previous section, XRD (Rigaku Smart Lab, λ = 1.54 Å) was performed. Fig. 4 shows the measured XRD spectra for the copper oxide films, as a function of ICP pressure, deposited on quartz. The signal intensity was not normalised to film thickness. The observed peaks were identified using the ICDD 00-045-0937 28 reference data for CuO and 00-005-0667 28 Fig. 2, there has to be additional Cu (for 4 Pa) and O (for 20 Pa) in the remainder of the film.
Therefore, the film deposited at an ICP pressure of 13 Pa seems to be the most promising in that both the stoichiometry as well as the structure are (polycrystalline) CuO. In addition, a few of the strong observed phases, e.g. the (020) phase, are known to be known high density phases, with high surface energies. These are not commonly observed with other deposition techniques, indicating that there might be significant energy transferred from the ICP to the substrate during the deposition process to allow these high-energy phases to be formed. The observed ZnO peaks were identified using the COD 2300112 30 reference data for ZnO.
The peaks in Fig. 5  performed, since this was outside the scope of this proof-of-concept study.
The crystal size of the deposited films can be estimated from the width of the measured XRD peaks using the Scherrer method 31 . A crystal size for the ZnO film of 3.1± 0.5 nm was found using the (002) peak. For the copper oxide films, the crystal size was determined from the dominant (020) peak. The results are shown in Fig. 6 and it can be seen that the crystal size increases with ICP pressure, from 5.  Fig. (a) shows the sample structure, consisting of the SiO 2 substrate, CuO thin film and the C, Pt-Pd and Pt protective layers. Fig. (b) shows a high-magnification image of the sample, focussing on the thin film and film-substrate interface. Fig. (c) is a dark-field image highlighting the grain structure in the film.
standard PLD-deposited films 32,33 . The increase of crystal size with background pressure has been reported before for metal-oxide thin films 34,35 and can be explained by the increased interaction of the metal plume with the background plasma for increasing pressure. This leads to on average larger clusters to be formed before deposition on the surface, eventually leading to larger crystal size in the deposited film 35 .
For further investigation of the PE-PLD-deposited CuO film at 13 Pa, TEM (JEOL 2010) analysis was undertaken. For this investigation, a thinner film of about 25 nm was deposited using the same operating conditions, but a reduced deposition time of 5 minutes, instead of 60 minutes.
An ICP pressure of 13 Pa was chosen since this seemed to have the best stoichiometry and crystal structure. In preparation for the TEM analysis, protective layers of C, Pt-Pd and Pt were deposited onto the CuO thin film. Using a Focused Ion Beam (FEI Nova Nanolab), a section of roughly 15 by 1 µm was milled out and removed from the film. This was then mounted and thinned to approximately 230 nm for TEM analysis. Fig. 7 presents images from TEM analysis of this sample. Fig   7(a) shows an overview of the sample, indicating the quartz substrate, CuO thin film and the three protective layers. It also gives an indication of the homogeneity of the film thickness. Fig. 7  was derived from the XRD analysis (Fig. 6). In addition, the thickness of the film measured with the TEM was 26 nm, which means the average deposition rate was 5.  Table I. local variations in the structure of the film which is not unlikely given the many different phases that are present in these polycrystalline films. Also, the composition is heavily dominated by CuO phases, but there is some evidence that there are also some Cu 2 O grains present in the film. C. Surface morphology and film resistance

D. Thin film deposition on plastic substrates
Since the substrate temperature is kept low, close to ambient temperature, deposition of thin films on heat-sensitive substrates such as plastics should be possible with PE-PLD. As a proof-of- suggests the copper oxide film is Cu 2 O as was the case for the samples deposited on quartz at this pressure (Fig. 2). XRD analysis showed no metal-oxide peaks for any of the films, suggesting that they are all amorphous. Fig. 10 shows an SEM image of the copper oxide film on a PE substrate. The film appears homogeneous without any micron-sized particulates. The vertical ridges that can be seen originate from the structure of the PE substrate and are not a result of the film deposition process. The adhesion of the films onto the substrates was not investigated in any detail. However, it was not possible to remove the films by bending and flexing the substrates after deposition, suggesting some reasonable adhesion properties. Of course, this is only a preliminary investigation and more detailed characterisation of the adhesion properties will need to be performed in the future to allow use of these films in applications and devices.

IV. CONCLUSIONS
PE-PLD with a pure metal target and a background low-temperature oxygen plasma (ICP) can successfully be used to deposit copper oxide and zinc oxide thin films on a quartz substrate without external substrate heating or post-annealing. The films were polycrystalline and the stoichiometry of the copper oxide films could be tuned by varying the ICP pressure. Many of the characteristics of the deposition technique and resulting films are similar to standard PLD using metal oxide targets and substrate heating, e.g. similar crystal size and deposition rate 29,37,38 . The resistivity of the film was also in line with what is reported for PLD-deposited films. Deposition of amorphous, stoichiometric zinc oxide and copper oxide films onto sheets of PE and PTFE was shown to be possible with PE-PLD. In conclusion, PE-PLD is capable of depositing films similar to PLD in stoichiometry and crystallinity but using pure metal targets instead of metal oxide ones and without substrate heating. Future investigations into performance of PE-PLD-deposited films in application-focused devices is needed to ensure PE-PLD can indeed produce the same, or better, quality films as conventional PLD.