betaFeedback

WorksPhysics and Astronomy

Ultra-low-resistivity nitrogen-doped p-type Cu2O thin films fabricated by reactive HiPIMS

Abstract

We have successfully fabricated the nitrogen-doped cuprous oxide thin films on the amorphous standard soda-lime glass by reactive high-power impulse magnetron sputtering. The energy of film-forming particles was controlled by the value of pulse-averaged target power density, which has a significant impact on the elemental composition, structure and optoelectrical properties of the films. We have shown that the high-energy regime is more suitable for preserving Cu2O structure and leads to continuous substitution of oxygen by nitrogen compared with the low-energy regime. Moreover, in the high-energy regime, it is possible, to some extent, to independently control the electrical resistivity and optical properties. The electrical resistivity decreases down to 5 x10-2 ohm.cm at the optical band gap 2.0-2.3 eV. Special attention is paid to the formation of nitrogen molecules and their ability to form shallow acceptor states. Experimental results supported by our DFT calculations indicate that N2 replacing Cu in the Cu2O lattice is one possible (but not the only possible) acceptor. We have also found that the formation of nitrogen molecules is preferred in a high-energy regime.

Introduction

One of the challenging scientific topics of today is finding a suitable p-type TCO that would at least approach the optoelectronic properties of the n-type counterpart [1]. Finding such p-type material is a necessary condition for the further sustainable growth of civilization. Realizing p-n junctions using transparent conductive materials will enable the development of a new generation of invisible electronics, contribute to reducing the energy requirements of various optoelectronic devices or lead to the production of more efficient solar cells.

Transparent conductive materials based on Cu2O\mathrm{Cu_2O} appear to be among the most promising. This is mainly due to the abundance of elements used, their non-toxicity and interesting optoelectronic properties. One of the many applications of Cu2O\mathrm{Cu_2O} is as an absorption layer for solar cells. It is because of a proper optical band gap of 2.0-2.6 eV [2, 3, 4] and the possibility of reaching relatively high hole mobilities [5, 6, 7]. It turns out that Cu2O\mathrm{Cu_2O}-based solar cells have one of the lowest cost-per-kWh ratios of any known material [8]. However, as with most other p-type TCOs, the main problem with Cu2O\mathrm{Cu_2O} is its relatively high resistivity. Several strategies have been used and investigated to decrease this. Namely, deposition at elevated temperatures [9], post-deposition annealing [10, 11], or post-deposition laser treatment [3], which have been proven to increase the mobility of holes. Even though the concentration of holes mostly decreases at the same time (healing of defects in the layer that serve as acceptor states), the overall conductivity of the layer increases. Another strategy is to dope Cu2O\mathrm{Cu_2O} with other elements to create additional acceptor states near the valence band maximum (VBM). Successful doping of Cu2O\mathrm{Cu_2O} with elements such as B, Li or Na was reported [12, 13, 14], which led to a significant increase in hole concentration and electrical conductivity. Special attention is then paid to doping Cu2O\mathrm{Cu_2O} with nitrogen, an abundant and non-toxic element. The role of nitrogen in the Cu2O\mathrm{Cu_2O} layer was described, e.g. by Lai et al.{}, where nitrogen was added during rf-magnetron sputter deposition, resulting in a decrease in the electrical resistivity down to ≈1×10−1 Ωcm\approx1\times10^{-1}\ \Omega\mathrm{cm} for optimal nitrogen partial pressure [15]. Other authors also reported similar results [16, 17, 18]. It was generally believed that the increase in hole concentration observed in all the above-mentioned works is caused by a nitrogen molecule substituting a copper atom in the Cu2O\mathrm{Cu}_2\mathrm{O} lattice ([2, 4]). However, in a recent in-depth study of Soltanmohammad and Nilius, it seems that the main mechanism is rather the substitutional replacement of oxygen by atomic nitrogen in the Cu2O\mathrm{Cu}_2\mathrm{O} matrix [19].

Although reactive high-power impulse magnetron sputtering (r-HiPIMS) is an advanced magnetron deposition method and has been successfully applied to the deposition of various TCOs [20, 21] and even to the deposition of Cu2O\mathrm{Cu}_2\mathrm{O} [22], the potential of this method to prepare nitrogen-doped Cu2O\mathrm{Cu}_2\mathrm{O} has not been explored. The advantage of this technique is the possibility of influencing, to some extent, the energy and composition of the particle flux forming the growing film. This was demonstrated, for example, during HiPIMS deposition of IGZO [23] or AZO [24] layers, when changing the elemental composition and optoelectronic properties, was possible by adjusting the pulse target power density. In this publication, we have demonstrated that r-HiPIMS is a very suitable method for effective nitrogen doping of Cu2O\mathrm{Cu}_2\mathrm{O} films and, to some extent, for independent tuning of electrical/optical properties. The prepared p-type Cu2O\mathrm{Cu}_2\mathrm{O}:N films exhibited an extraordinarily low resistivity below 5×10−2 Ωcm5\times10^{-2}\ \Omega\mathrm{cm} while maintaining p-type character. Moreover, we found the role of the nitrogen molecule as one of the possible acceptors is probably still not ruled out.

Experimental setup

Film preparation

Cu-O-N thin films were deposited using reactive r-HiPIMS with a copper target (99.99% purity), with a diameter of 100 mm and a thickness of 6 mm (area, AtA_\mathrm{t}, of 78.54 cm278.54\ \mathrm{cm}^2), in an argon–oxygen–nitrogen gas mixture. A scroll and turbomolecular pump were used to evacuate the chamber to a base pressure of approximately 6×10−4 Pa6\times10^{-4}\ \mathrm{Pa}. The schematic picture of our experimental apparatus with a more detailed description could be found in our recent work [3].

The working pressure during the deposition was controlled in two stages. In the first stage, a constant pressure of 0.5 Pa (Ar + N2_2) was maintained by adjusting the pumping speed. Keeping the argon and nitrogen flow constant, the required nitrogen fraction in Ar + N2_2 mass flow, fN2f_{\mathrm{N}_2}, was set. In the second stage, the oxygen was added, and the oxygen partial pressure, poxp_{\mathrm{ox}}, varied from 100 mPa to 340 mPa. The magnetron was driven with a pulsed power supply (Melec GmbH) with the rectangular-shaped voltage pulse on-time tont_{\mathrm{on}} was 100 μ\mus. The value of the off-time (tofft_{\mathrm{off}}) was set to obtain the required pulse-averaged target power density Sda=Pdc×AtS_{\mathrm{da}} = \frac{P}{d_{\mathrm{c}} \times A_{\mathrm{t}}}, where PP is the average power (500 W for all cases) and dc=ton/(ton+toff)d_{\mathrm{c}} = t_{\mathrm{on}}/(t_{\mathrm{on}} + t_{\mathrm{off}}) is the duty cycle. The soda-lime glass (38 ×\times 26 ×\times 1 mm3^3) and Si (100) wafers (26 ×\times 15 mm2\,\mathrm{mm}^2) were used as substrates and were cleaned in isopropyl alcohol and distilled water before deposition (10 min for each process). The target-to-substrate distance was 100 mm, and the substrates were heated to 190 °C. The deposition time was between 110 s and 150 s, resulting in a film thickness of 160-250 nm.

Film characterization

The elemental composition of the selected Cu-O-N films was measured by wave dispersive spectroscopy (WDS) carried out in a scanning electron microscope (Hitachi, SU-70). Measurements were carried out at an accelerating voltage of 7 kV. Cuprite (Cu2_2O) was used as the standard for copper and oxygen, while ZrN (sputtered, 50:50 ratio) served as the standard for nitrogen. Calibration and quantification were conducted using the PROZA method, neglecting the influence of the underlying silicon substrate. The standards and the analyzed samples were coated with a 14 nm carbon layer to minimize charging effects and improve surface conductivity during analysis. To characterize the phase composition of thin films, we used a Raman spectroscopy (Horiba Jobin Yvon LABRAM HR Evolution Raman spectroscope), with a laser wavelength of 532 nm. The structure of the films was characterized by X-ray diffraction (XRD) using a PANalytical X'Pert PRO MPD diffractometer working in the Bragg-Brentano geometry using a CuKα\alpha (40 kV, 40 mA) radiation. The electrical resistivity, ρ\rho, was measured by the standard four-probe method. The optical band gap, EgE_g, was estimated using the Tauc plot method, which involves analyzing the optical absorption data. The transmittance TT and reflectance RR of the thin films were measured. The absorption coefficient α\alpha was calculated as

α=−[ln⁡T1−R]1t\alpha= -\left[\ln\frac{T}{1-R}\right]^{\frac{1}{t}}

where tt is the thickness of the thin film layer. For direct band gap materials, the relationship (αhν)2=(hν−Eg)(\alpha h\nu)^2=(h\nu-E_g) was applied, where hνh\nu is the photon energy. A plot of (αhν)2(\alpha h\nu)^2 versus hνh\nu was used to determine EgE_g. Photoelectrochemical (PEC) measurements were conducted in a custom cell made from chemically inert polytetrafluoroethylene (PTFE), with internal dimensions of 20 ×\times 17 ×\times 30 mm3^3. Optical illumination and electrochemical contact were defined by a 6 mm circular window sealed by the sample. The working electrode consisted of a Cu-O-N layer on the Si substrate. A spiral platinum wire, positioned 15 mm from the working electrode, served as the counter electrode, while a standard Ag/AgCl reference electrode was placed ∼\sim 6 mm away (corresponding to +0.20 V vs SHE and +0.53 V vs RHE). The supporting electrolyte was 0.1 M Na2_2SO4_4 (pH 5.66). Illumination was provided by a xenon arc lamp with an AM1.5G filter, calibrated to 100 mW⋅\cdotcm−2^{-2} at the sample surface, and delivered through a quartz window. Electrochemical measurements were carried out using a Squidstat Plus potentiostat (Admiral Instruments, USA). Open-circuit potential (OCP) measurements under dark and illuminated conditions were used to evaluate the semiconductor properties. The refractive index (nn) and extinction coefficient (kk) have been measured by spectroscopic ellipsometry using the J. A. Woollam Co. Inc. VASE instrument. The measurements were performed in reflection using angles of incidence of 55°, 60° and 65° in the wavelength (λ\lambda) range from 300 to 2000 nm. The optical model included the glass substrate, Cu-O-N layer and a surface roughness layer. The imaginary part of the permittivity (ε2\varepsilon_2) of Cu-O-N has been expressed by a sum of Cody-Lorentz formula (oscillator) with Lorentz oscillators (see the detailed discussion in Results). The formula for the former reads

ε2=L(E)G(E)for E≥Eg+Et(1)\varepsilon_2 = L(E)G(E) \quad\text{for } E \ge E_g + E_t \tag*{(1)}

and

ε2=L(Eg+Et)G(Eg+Et)Eg+EtEexp⁡(E−Eg−EtEu)for E<Eg+Et(2)\varepsilon_2 = L(E_g + E_t)G(E_g + E_t)\frac{E_g+E_t}{E}\exp\left(\frac{E-E_g-E_t}{E_u}\right) \quad\text{for } E < E_g + E_t \tag*{(2)}

where

L(E)=AB2EnE(E2−En2)+B2E2(3)L(E) = \frac{AB^2E_nE}{(E^2-E_n^2)+B^2E^2} \tag*{(3)}

(Lorentz oscillator in one of its possible forms)

and

G(E)=(E−Eg)2(E−Eg)2+Ep2(4)G(E) = \frac{(E-E_g)^2}{(E-E_g)^2+E_p^2} \tag*{(4)}

(Cody form of the variable band edge function) where AA is the amplitude, BB is the broadening (damping), EnE_n is the energy which the Lorentz oscillator is centered at, EpE_p is the transition energy between the onset behavior and the Lorentz oscillator behavior, EgE_g is optical gap in the narrow sense of the word, Eg+EtE_g+E_t is the optical gap in a broader sense of the word (including non-zero density of electronic states close to its edges) representing the transition energy between band-to-band transitions and the exponential Urbach absorption tail and EuE_u represents the shape of this tail.

DFT calculations

DFT calculations were performed using VASP with the PBE-GGA functional (Perdew et al.) [25] and the PAW method (Kresse and Joubert) [26]. A $2 \times 2 \times 2$ Cu$_2$O supercell was used, with a 550 eV cutoff energy and a $6 \times 6 \times 6$ Monkhorst-Pack k-point mesh. Structures were relaxed until forces were below 0.005 eV/Å and energy convergence was within $1\mathrm{e}{-7}$ eV. DFT-D3 dispersion corrections were included {citep}(Hassani et al.){citep}[27]`. Nitrogen atoms and nitrogen molecules were doped at interstitial, Cu, and O sites. Formation energies were calculated as:

Interstitial:Ef=ECu2O+N−ECu2O−EN2tot2\text{Interstitial:}\quad E_f = E_{\mathrm{Cu_2O+N}} - E_{\mathrm{Cu_2O}} - \frac{E_{\mathrm{N_2}}^{\mathrm{tot}}}{2}

Substitutional N (example for Cu-site): Ef=ECu2O+N−ECu2O−EN2tot2+Etot(Cu bulk)NCuE_f = E_{\mathrm{Cu_2O+N}} - E_{\mathrm{Cu_2O}} - \frac{E_{N_2}^{\mathrm{tot}}}{2} + \frac{E_{\mathrm{tot}}\left(\mathrm{Cu\ bulk}\right)}{N_{\mathrm{Cu}}} and

Substitutional N2_2 (example for Cu-site): Ef=ECu2O+N2−ECu2O−EN2tot+Etot(Cu bulk)NCuE_f = E_{\mathrm{Cu_2O+N_2}} - E_{\mathrm{Cu_2O}} - E_{N_2}^{\mathrm{tot}} + \frac{E_{\mathrm{tot}}\left(\mathrm{Cu\ bulk}\right)}{N_{\mathrm{Cu}}}

Where ECu2O+NE_{\mathrm{Cu_2O+N}} is the total energy of supercell with one nitrogen atom, ECu2OE_{\mathrm{Cu_2O}} is the total energy of supercell without defects, EN2totE_{N_2}^{\mathrm{tot}} is the total energy of nitrogen molecule, Etot(Cu bulk)E_{\mathrm{tot}}\left(\mathrm{Cu\ bulk}\right) is the total energy of the Cu supercell and NCuN_{\mathrm{Cu}} is the number of Cu atoms in that supercell.

Results and discussion

Elemental composition

The elemental composition of the Cu-O-N layers prepared in the high-energy regime (at an oxygen partial pressure pox=260p_{\mathrm{ox}} = 260 mPa) as a function of the nitrogen fraction in the Ar + N2_2 mixture, fN2f_{\mathrm{N_2}}, is shown in Figure 1a. We can see that the copper content remains practically constant (approx. 65 at.%) independent of the fN2f_{\mathrm{N_2}} value. Furthermore, a gradual substitution of oxygen by nitrogen is evident. Here, the previously reported advantage of the high pulse-averaged target power density, SdaS_{\mathrm{da}}, is utilized [28]. During this regime, the high electron temperature in the plasma discharge effectively dissociates nitrogen molecules to atoms with a higher sticking coefficient and nitrogen is more easily incorporated into the layer. This is also clear from Figure 1b: for Sda≤700 W cm−2S_{\mathrm{da}} \le700\ \mathrm{W\,cm}^{-2} the nitrogen content in the films is around ≈2\approx2 - 3 at. %, when the increase of SdaS_{\mathrm{da}} to 1000 and 1300 W cm−2\mathrm{W\,cm}^{-2} the nitrogen content increases to ≈4\approx4 and 6 at. %, respectively.

Image not extracted

Figure 1. (a) The elemental composition of Cu-O-N films fabricated at constant pulse-averaged target power density Sda=1300 Wcm−2S_{\mathrm{da}} = 1300\ \mathrm{Wcm}^{-2}, pox=260 mPap_{\mathrm{ox}} = 260\ \mathrm{mPa} and as a function of nitrogen fraction in Ar + N2_2 mass flow, fN2f_{\mathrm{N}_2}. (b) The elemental composition of Cu-O-N films fabricated at pox=260 mPap_{\mathrm{ox}} = 260\ \mathrm{mPa}, fN2=90%f_{\mathrm{N}_2} = 90\% as a function of SdaS_{\mathrm{da}}.

To demonstrate the fundamental differences in the growth conditions of Cu-O-N layers in the low-energy (Sda=100 Wcm−2S_{\mathrm{da}} = 100\ \mathrm{Wcm}^{-2}) and high-energy (Sda=1300 Wcm−2S_{\mathrm{da}} = 1300\ \mathrm{Wcm}^{-2}) regimes, a series of layers were prepared at different oxygen partial pressures, poxp_{\mathrm{ox}}. To highlight the observed effects, the layers were prepared at the highest fN2=90%f_{\mathrm{N}_2} = 90\%. At first glance, a clear difference between the two regimes is visible. In the low-energy regime (Figure 2a), there is a noticeable decrease in the copper content in the layer from about 63 at.% to 50 at.% with an increase in poxp_{\mathrm{ox}} from 160 mPa to 260 mPa. In contrast, in the high-energy regime (Figure 2b), the copper content is practically constant in a very wide interval of 160 - 340 mPa. In the low-energy mode, there is also a sharper increase in the oxygen content in the layer (with increasing poxp_{\mathrm{ox}}), which leads to an undesirable gradual transition to a material with a higher oxidation state of copper (Cu4O3\mathrm{Cu}_4\mathrm{O}_3, CuO or their combinations, see also below). Therefore, the high-energy mode is clearly more suitable for fine-tuning the nitrogen content in Cu2_2O:N films.

Image not extracted

Figure 2. (a) The elemental composition of Cu-O-N films fabricated in low-energy mode (Sda=100S_{\mathrm{da}}=100 Wcm−2^{-2}) at constant fN2=90%f_{\mathrm{N}_2}=90\% as a function of oxygen partial pressure, poxp_{\mathrm{ox}}. (b) The elemental composition of Cu-O-N films fabricated in high-energy regime (Sda=1300S_{\mathrm{da}}=1300 Wcm−2^{-2}) at fN2=90%f_{\mathrm{N}_2}=90\% as a function of SdaS_{\mathrm{da}}. The intervals when only Cu2_2O phase is detected are marked.

Structural properties

Evolution of Raman spectra for the films prepared at different fN2f_{\mathrm{N}_2} being in the range of 0 - 90% (at constant pox=260p_{\mathrm{ox}}=260 mPa and Sda=1300S_{\mathrm{da}}=1300 Wcm−2^{-2}) are depicted in Figure 3. We can observe several Raman peaks. The peak at about 145 cm−1^{-1} corresponds to the T1u_{1u} cuprite mode, and oxygen vacancies activate it. The intensity of this peak decreases with increasing fN2f_{\mathrm{N}_2}, indicating a decrease in the number of oxygen vacancies in the films [29]. Other significant peaks (around 208, 531 and 608 cm−1^{-1}) prove that only the Cu2_2O phase is present in all films containing up to ≈ 6 at.% of nitrogen. XRD results also double-checked this, where only two prominent peaks corresponding to the cubic Cu2_2O phase were detected for the films prepared at fN2=0f_{\mathrm{N}_2}=0 and 90% (see Figure 4a). The positions of the most significant peaks, 531 and 608 cm−1^{-1} for the film without nitrogen, are slightly shifted (10 - 15 cm−1^{-1}) to the lower wavenumber values when the nitrogen is introduced. Also, we can see the broadening of those two peaks with increasing nitrogen content, probably due to the higher density of crystal defects. Finally, there is a continuous increase of the peak located around 2250 cm−1^{-1} for the films with fN2≥10%f_{\mathrm{N}_2} \ge10\%. This peak corresponds to the nitrogen molecule [16].

Image not extracted

Figure 3. A Raman spectrum of the Cu-O-N films fabricated in the high-energy regime at constant pox=260p_{\mathrm{ox}}=260 mPa as a function of fN2f_{\mathrm{N}_2}. The elemental composition of individual films is plotted for better clarity.

Image not extracted

Figure 4. (a) XRD patterns of the Cu-O-(N) films fabricated in high-energy regime at constant pox=260p_{\mathrm{ox}}=260 mPa and different fN2f_{\mathrm{N_2}} (b) XRD patterns of selected films prepared in low-energy regime (bottom) and high-energy regime (top) at constant fN2=90%f_{\mathrm{N_2}}=90\% and for different values of poxp_{\mathrm{ox}}. The position of Cu2O\mathrm{Cu_2O} peaks (stress-free standard) and elemental composition of individual films are also marked for better clarity.

Raman spectra of the films prepared at different values of SdaS_{\mathrm{da}}, i.e. at varying energies of film-forming particles are shown in Figure 5. The film prepared at the lowest Sda=100S_{\mathrm{da}}=100 Wcm−2^{-2} exhibits CuO structure (which corresponds to elemental composition of Cu50_{50}O47_{47}N3_3), other films exhibit standard Raman patterns of Cu2_2O discussed above. A slight broadening and shifting of the peaks to the lower values of wavenumbers is caused by an increasing nitrogen concentration.

Image not extracted

Figure 5. A Raman spectrum of the Cu-O-N films fabricated at constant pox=260p_{\mathrm{ox}}=260 mPa and fN2=90%f_{\mathrm{N_2}}=90\% as a function of SdaS_{\mathrm{da}}. The elemental composition of individual films is plotted for better clarity.

The intensity of the peak around 2250 cm−1^{-1} (N2_2 molecule) also increases with the SdaS_{\mathrm{da}} value. Moreover, it is continuously shifting to the lower wavenumbers, probably caused by different bonding mechanisms of the nitrogen molecule under various growing conditions [16].

Interestingly, although films prepared at Sda=100S_{\mathrm{da}}=100-700 Wcm−2^{-2} have practically the same nitrogen concentration (≈3\approx3 at.%) and the films prepared at Sda=400S_{\mathrm{da}}=400 and 700 Wcm−2^{-2} even the very similar Cu and O content (≈64\approx64 at. % and 33 at. %, respectively), there is a visible increase in N2_2 molecule Raman peak (not detected for the Sda=100S_{\mathrm{da}}=100 Wcm−2^{-2}). This indicates that the higher energy of film-forming particles favours the formation of N2_2 molecules.

The next paragraph will discuss significant and qualitative differences in the structure of the films prepared at low- and high-energy conditions. Figure 6 shows Raman spectra of films prepared at both regimes and at different oxygen partial pressures. In the low-energy regime, the single-Cu2_2O structure is presented in the range of poxp_{\mathrm{ox}} from 100 to 160 mPa. Above that threshold, there is a mixture of Cu4_4O3_3+CuO phases (poxp_{\mathrm{ox}}=200 mPa) or a single-CuO phase film (poxp_{\mathrm{ox}}=260 mPa). For the high-energy regime, the situation is quite different. All films in the studied poxp_{\mathrm{ox}} interval (160 - 340 mPa) remain single-phase Cu2_2O. This was also checked by XRD (Figure 4b). From this point of view, the high-energy regime is more suitable for the preparation of Cu2_2O films without undesired admixture or other phases like Cu4_4O3_3 or CuO. The main reason is probably that, due to high-energy and high-density plasma, the target is prevented from being covered by copper (sub)oxides, and therefore, there is no significant drop in sputtering yield, as probably is the case in the low-energy regime. This means that there are still enough copper atoms ready to form Cu2_2O film, and the transition to Cu4_4O3_3 or CuO will occur only at very high oxygen partial pressures (beyond our experimental values). This behaviour was also observed in other work dealing with reactive sputtering at high target power densities [30]. Another interesting finding is the different trend in the N2_2 molecule Raman peak intensity. While in the low-energy regime we observe a significant decrease in the intensity of the N2_2 molecule signal with increasing poxp_{\mathrm{ox}} (i.e. with more oxygen in the chamber), practically down to zero for pox≥200p_{\mathrm{ox}} \ge200 mPa, in the high-energy regime, the opposite trend is observed. We see a gradual increase in the peak intensity from 160 mPa to 300 mPa. This is despite the fact that the total nitrogen concentration in the layer with poxp_{\mathrm{ox}} decreases (from approximately 10 at.% to 2 at.% with an increase in poxp_{\mathrm{ox}} from 160 to 300 mPa). The mechanism of this behaviour is not yet completely clear, but it is certain that it cannot be explained only by the elemental composition, which would possibly favour the formation of the N2_2 molecule. An example is the comparison of a layer formed in the low-energy regime at poxp_{\mathrm{ox}}=160 mPa and a layer formed in the high-energy regime at poxp_{\mathrm{ox}}=300 mPa: both layers show practically identical elemental composition (Cu63_{63}O33_{33}N4_4 vs Cu64_{64}O32_{32}N4_4), but while the peak of the N2_2 molecule in the “low-energy” layer is almost negligible (Figure 6a), in the “high-energy” layer it is very intense (Figure 6b).

Image not extracted

Figure 6. (a) A Raman spectrum of the Cu-O-N films fabricated in the low-energy regime at constant fN2=90%f_{\mathrm{N}_2}=90\% as a function of poxp_{\mathrm{ox}}, (b) A Raman spectrum of the Cu-O-N films fabricated in the high-energy regime at constant fN2=90%f_{\mathrm{N}_2}=90\% as a function of poxp_{\mathrm{ox}}. The elemental composition of individual films is plotted for better clarity.

Electrical and optical properties

The electrical resistivity and optical band gap determined from Tauc’s plots as a function of fN2f_{\mathrm{N}_2} in the high-energy regime (at pox=260p_{\mathrm{ox}}=260 mPa) are plotted in Figure 7a. Both the resistivity and the optical band gap decrease with fN2f_{\mathrm{N}_2}. In the case of electrical resistivity, we observe a very significant decrease from the value of 3×1033 \times10^{3} Ω\Omegacm for the nitrogen-free layer to about 5×10−25 \times10^{-2} Ω\Omegacm for the highest value of fN2=90%f_{\mathrm{N}_2}=90\%. This value of electrical resistivity surpasses, to our knowledge, all previously published values of electrical resistivity of nitrogen-doped Cu2_2O thin films prepared on amorphous substrates. Probably similar resistivity was achieved only in the work of Benz et. Al [16] (although the exact value is not given numerically in the article), however, it was a relatively slow RF deposition on a highly crystalline sapphire substrate. There is an observed decrease in the optical band gap with increasing fN2f_{\mathrm{N}_2} from approximately 2.50 eV for the nitrogen-free layer to 2.30 eV for the Cu64O30N6\mathrm{Cu}_{64}\mathrm{O}_{30}\mathrm{N}_{6} layer prepared at fN2=90%f_{\mathrm{N}_2}=90\%.

Image not extracted

Figure 7. (a) Electrical resistivity and optical band gap of the Cu-O-N films prepared in high-energy regime at constant pox=260p_{\mathrm{ox}} = 260 mPa as a function of fN2f_{\mathrm{N}_2}. (a) Electrical resistivity and optical band gap of the Cu-O-N films prepared at constant pox=260p_{\mathrm{ox}}=260 mPa and fN2=90%f_{\mathrm{N}_2}=90\% as a function of SdaS_{\mathrm{da}}. The intervals when only the Cu2_2O phase is detected are marked.

The gradual decrease in electrical resistivity and optical band gap is also observed if the films prepared at increasing SdaS_{\mathrm{da}} (at constant poxp_{\mathrm{ox}} and fN2f_{\mathrm{N}_2} of 260 mPa and 90%, respectively), see Figure 7b. More precisely, there is an increase in EgE_{\mathrm{g}} from 2.20 eV to 2.50 eV when SdaS_{\mathrm{da}} is increased from 100 Wcm−2^{-2} to 400 Wcm−2^{-2}. After that, EgE_{\mathrm{g}} decreases monotonically with SdaS_{\mathrm{da}}. The initial increase is due to the fact that the layer prepared for the lowest Sda=100S_{\mathrm{da}}=100 Wcm−2^{-2} exhibits CuO structure (which has lower EgE_{\mathrm{g}} than Cu2O\mathrm{Cu}_{2}\mathrm{O}).

The benefit of the high-energy regime for tuning of optoelectric properties is clearly seen in Figure 8. In the low-energy regime, the resistivity is very sensitive to the value of poxp_{\mathrm{ox}} and, mainly due to various phase transitions, changes between 100 mPa and 260 mPa by three orders of magnitude. In the high-energy regime, the resistivity is almost constant and at the same time very low (from ≈5×10−2 Ωcm\approx5 \times10^{-2}\ \Omega\mathrm{cm} to ≈2×10−1 Ωcm\approx2 \times10^{-1}\ \Omega\mathrm{cm}) in a wide range of conditions from pox=160p_{\mathrm{ox}}=160 to 300 mPa. At the same time, the optical band gap changes very significantly, from 2.0 eV for 160 mPa to 2.5 eV for 300 mPa. This independent tuning of the optical band gap can be advantageous, for example, for photovoltaic absorber applications, where it is necessary to precisely tune the band gap edge while maintaining a low electrical resistivity. Another interesting observation is that the Cu2O\mathrm{Cu}_{2}\mathrm{O}:N film prepared under the high-energy regime (pox=260p_{\mathrm{ox}}=260 mPa, see Figure 6b) containing 5.8 at.% nitrogen exhibits very similar optoelectrical properties (ρ≈5×10−2 Ωcm\rho\approx5 \times10^{-2}\ \Omega\mathrm{cm}, Eg≈2.35E_{\mathrm{g}} \approx2.35 eV) as the film prepared under low-energy regime containing 10.6 at.% of nitrogen (pox=100p_{\mathrm{ox}}=100 mPa, see Figure 6b). Although the nitrogen concentration is roughly half for the “high-energy” layer, the intensity of the Raman peak corresponding to the N2\mathrm{N}_{2} molecule is higher compared with the low-energy layer. This indicates that the fraction of nitrogen in molecular form is higher in the case of “high-energy” film and, in our opinion, still keeps the nitrogen molecule in play as a serious candidate responsible for increasing the hole concentration, see also later in the DFT calculations paragraph. However, further research is still needed.

Image not extracted

Figure 8. (a) Electrical resistivity and optical band gap of the Cu-O-N films prepared at constant fN2=90%f_{\mathrm{N}_2}=90\% as a function of poxp_{\mathrm{ox}} in the low-energy regime, and (b) in the high-energy regime. The intervals when only the Cu2_2O phase is detected are marked.

Spectroscopic ellipsometry

Optical properties of selected Cu-O-N materials studied are shown in Figure 9. As illustrated in Figure 9a for one of the samples, the materials have been represented by a sum of (i) the Cody-Lorentz dispersion formula (Eq. 1), having the optical gap Eg+EtE_g+E_t (2.2±\pm0.4 eV for samples dominated by Cu2O\mathrm{Cu}_2\mathrm{O}) as one of its parameters, with (ii) Lorentz oscillator (Eq. (3)) in the infrared, centered at En≈0.8±0.2E_n \approx0.8\pm0.2 eV (all samples prepared at a presence of N2\mathrm{N}_2, not necessary for pure Cu2O\mathrm{Cu}_2\mathrm{O}), (iii) Lorentz oscillator in the ultraviolet, centered at En≈3.7±0.2E_n \approx3.7\pm0.2 eV (all samples) and (iv) another Lorentz oscillator in the ultraviolet, centered at En≈4.7±0.5E_n \approx4.7\pm0.5 eV (all samples dominated by Cu2O\mathrm{Cu}_2\mathrm{O}, not necessary after the transition from Cu2O\mathrm{Cu}_2\mathrm{O} to Cu4O3\mathrm{Cu}_4\mathrm{O}_3 and CuO\mathrm{CuO}).

Image not extracted

Figure 9. Panel (a) shows an example of the optical model of Cu-O-N: imaginary part of relative permittivity as a sum of four oscillators discussed in the text. Panels b-d show spectral refractive index (left axes, dashed lines) and extinction coefficient (right axes, solid lines) at pox=260p_{\mathrm{ox}} = 260 mPa, Sda=1300S_{\mathrm{da}} = 1300 W cm−2^{-2} and varied fN2f_{\mathrm{N}_2} (b), fN2=90%f_{\mathrm{N}_2} = 90\%, low-energy regime and varied poxp_{\mathrm{ox}} (c) and fN2=90%f_{\mathrm{N}_2} = 90\%, high-energy regime and varied poxp_{\mathrm{ox}} (d). The insets show parameters of the Lorentz oscillator centred in the IR. Panel e shows the relationship of electrical resistivity and amplitude of the oscillator centred in IR (AIRA_{\mathrm{IR}}) for N-poor samples (where both quantities monotonically depend on [N]) as well as N-rich samples (where they almost saturate). Panel f shows the relationship of electrical resistivity and extinction coefficient at λ=550\lambda= 550 nm.

Note that although the aforementioned Tauc gaps obtained by spectrophotometry (2.0 - 2.5 eV) are not equal to the Eg+EtE_g+E_t values (1.8-2.6 eV) of Cu2O\mathrm{Cu}_2\mathrm{O} obtained by ellipsometry, there is an almost monotonic dependence between these quantities (not shown). The only significant outliers are O-rich samples containing Cu4O3\mathrm{Cu}_4\mathrm{O}_3 and CuO\mathrm{CuO} rather than Cu2O\mathrm{Cu}_2\mathrm{O}, characterized by the least sharp band gap edges and significantly different dispersion of optical constants (high pox=260p_{\mathrm{ox}} = 260 mPa in the low-energy regime; see two of the dependencies in Figure 9c).

The characteristics of the Lorentz oscillator centred in the infrared are arguably most revealing.

The inset in Figure 9b shows that increasing N2\mathrm{N}_2 partial pressure and, in turn, increasing N incorporation into the films (Figure 1a) leads to a steeply increasing amplitude of this oscillator (AIRA_{\mathrm{IR}}) from 0 to ≈4.4\approx4.4, at approximately constant position and broadening. Thus, a strong case can be made that this oscillator represents the electronic states inside the band which are localized on nitrogen. The position of this oscillator, En≈0.8±0.2E_n \approx0.8\pm0.2 eV, is a measure (upper bound of) the distance of these states from the band gap edge. Similar evolution has also been observed for the oscillator centered in the ultraviolet region at En≈4.70±5E_n \approx4.70\pm5 eV (not shown).

Increasing the N incorporation indirectly, by decreasing the partial pressure of oxygen, has an effect on the electronic structure of Cu-O-N as well. In the low-energy regime (Figure 9c), at mostly low [N] values (only one sample with >5 at.% N; Figure 2a), the phenomenon is similar to that observed in Figure 9b: AIRA_{\mathrm{IR}} increases with decreasing poxp_{\mathrm{ox}} (increasing [N]), once again up to ≈ 4.4. However, in the high-energy regime (Figure 9d), leading to only high [N] values (all samples with >5 at.% N; Figure 2b), AIRA_{\mathrm{IR}} does not increase with [N] and even slightly decreases.

This indicates that while the incorporation of up to ≈ 5 at.% N preserves the main features of the electronic structure of Cu2_2O except the addition of new electronic states inside the band gap, further N incorporation (here up to 12 at.% N) leads to further qualitative changes.

The presented phenomena are examined in Figure 9e, which shows a very good correlation (even including the two samples containing Cu4_4O3_3 and CuO) between two quantities which are both related to the concentration of free charge carriers: ρ\rho and concentration of electronic states inside the band gap in terms of AIRA_{\mathrm{IR}}. While 0–5% at.% N leads to a practically monotonic dependence of both these quantities on [N], at higher N contents both these quantities saturate close to ρ≈5×10−2\rho\approx5 \times10^{-2} Ωcm (see also the saturation in Figure 8b) and AIR≈4.4A_{\mathrm{IR}} \approx4.4.

In addition to the spectral dependencies of n(λ)n(\lambda) and k(λ)k(\lambda) shown in Figure 9b–Figure 9d, let us point out a representative and frequently used quantity: kk at λ=550\lambda= 550 nm, k550k_{550}. The relationship between key optoelectronic properties, ρ\rho and k550k_{550}, regardless the N content, is shown in Figure 9f. This time the samples containing Cu4_4O3_3 and CuO constitute outliers. However, the Cu2_2O-like samples yield a smooth dependence: k550k_{550} increases with decreasing ρ\rho from 0.06 to 0.26 until the latter reaches the aforementioned saturation value of 5×10−25 \times10^{-2} Ωcm. Then, k550k_{550} can still be varied in a wide range from 0.26 to 0.85 (which of these values is better depends on the application) at approximately constant ρ\rho.

Charge transport properties

Because it is very complicated or even impossible to measure charge transport properties (i.e. Hall mobility and concentration of major charge carriers) in Cu2O\mathrm{Cu}_{2}\mathrm{O} films heavily doped by nitrogen, we have performed illuminated open-circuit potential (OCP) measurements of selected films. Illuminated Open Circuit Potential (OCP) is the most straightforward photoelectrochemical method for determining the conductivity type of a semiconductor. Illumination above the bandgap generates electron–hole pairs, which are separated by the internal electric field within the depletion region. The minority charge carriers migrate toward the semiconductor–electrolyte interface to facilitate chemical reactions, while the majority charge carriers generate an opposing electric field. Consequently, if the OCP shifts positively (toward more anodic potentials) upon above-bandgap illumination, the material is identified as p-type [31].

Image not extracted

Figure 10. The open circuit potential for the selected films as a function of time. The elemental composition of individual films are marked for the better clarity.

We measured four films, all prepared with the highest nitrogen content (fN2=90%f_{\mathrm{N}_2} = 90\%). The first two were synthesised under a low-energy regime with oxygen partial pressures of pox=100p_{\mathrm{ox}} = 100 and 200 mPa, while the second pair was prepared under a high-energy regime (Sda=1300S_{\mathrm{da}} = 1300 Wcm−2^{-2}) with oxygen partial pressures of pox=160p_{\mathrm{ox}} = 160 and 260 mPa. These films were chosen to cover a broad range of conditions and corresponding material properties. All films exhibited moderate photoactivity. Although the OCP shifts were only on the order of millivolts, the consistently positive shifts indicate that all measured films possessed p-type conductivity.

Despite the low-energy regime film at pox=100p_{\mathrm{ox}} = 100 mPa differing significantly from the other three, including approx. 200 times higher electrical resistivity, distinct elemental and crystalline phase composition, the most pronounced differences in both dark OCP values and illuminated OCP responses were attributable to the regime of deposition (low- or high-energy). Films prepared under high-energy conditions exhibited lower dark OCP values, indicating a greater susceptibility to electron loss and corrosion due to lower electronegativity and/or a reduced ability to form protective oxide layers (passivation). More importantly, both high-energy regime films also demonstrated stronger and faster illuminated OCP responses. This indicates a more pronounced p-type semiconductor character, where hole conduction dominates. Such behaviour suggests reduced charge carrier recombination due to a lower density of defect states (recombination centres), thereby enabling higher charge carrier mobility.

DFT calculations

Table 1 confirms that all nitrogen-doped configurations in Cu2O\mathrm{Cu}_2\mathrm{O} have positive formation energies, indicating endothermic incorporation and underlining the advantage of the nonequilibrium technique used for the preparation of these configurations. Among them, the most favorable configurations correspond to nitrogen atoms substituting oxygen atoms and nitrogen molecules (N2\mathrm{N}_2) substituting copper atoms. These were selected for further analysis. The calculated band gap of pristine Cu2O\mathrm{Cu}_2\mathrm{O} is 0.51 eV (Figure 11a), consistent with previous DFT-GGA reports (0.55–0.70 eV, A. Larabi [32]), though significantly lower than the experimental value of 2.1-2.6 eV [2, 3] due to the known band gap underestimation (albeit at preserved trends) by GGA functionals. Upon nitrogen doping, the band gap increases to 0.89 eV for N2\mathrm{N}_2 doping at Cu sites and 0.72 eV for N doping at O sites (Figure 11b and Figure 11c, respectively). This band gap widening is accompanied by the key result of the DFT calculations: shift of the Fermi level toward the valence band in both cases, indicating increased hole carrier concentration (see the lowered electrical resistivity in Figure 7 and Figure 8) and further supporting the desired p-type character of the material. Additionally, although the new electronic states arising from nitrogen doping (projected density of states in Figure 11e and Figure 11f) are not exclusively localized near the Fermi level, Figure 11f shows disproportionately strong role of N at the Fermi level and just below it at the top of the valence band (but not at the bottom of the conduction band). This is consistent with the spectroscopic ellipsometry. While the results of ellipsometry (Figure 9) could be in part explained by new electronic states close to the edge of some band after N incorporation (a new Lorentz oscillator centred in the infrared), DFT allows us to specify that they are close to the edge of the valence band.

SitesN-Cu2_2O (O-site)N-Cu2_2O (Cu-site)Interstitial position (N atom)N2_2-Cu2_2O (O-site)N2_2-Cu2_2O (Cu-site)Interstitial position (N2_2)
EfE_f(eV)2.264.873.004.581.672.86

Table 1. Formation energies of N atom and N2_2 molecule at various sites in the Cu2_2O lattice.

Image not extracted

Figure 11. (a) Total DOS of pure Cu2_2O, (b) N2_2-doped Cu2_2O (Cu substitution, the most energetically preferred N2_2 site), (c) N-doped Cu2_2O (O substitution, the most energetically preferred N site), (d) optimized 2×2×22 \times2 \times2 Cu2_2O supercell structure (Cu: blue, O: red, N: green), (e) projected density of states of N2_2-doped Cu2_2O (Cu substitution), and (f) projected density of states of N-doped Cu2_2O (O substitution). The Fermi level is set to 0 eV.

Conclusions

The nitrogen-doped p-type cuprous oxide thin films with very low electrical resistivity down to ≈5×10−2 Ωcm\approx5 \times10^{-2}\ \Omega\mathrm{cm} were successfully fabricated. The main conclusion of this work could be summarized as follows:

  • High-energy regime (i.e. discharge at high pulse-averaged target power density of 1300 Wcm−21300\ \mathrm{Wcm}^{-2}) is more suitable for the preparation of Cu2_2O-like films and enables continuous replacement of oxygen by nitrogen (Cu65_{65}O32_{32}N4_4 – Cu65_{65}O23_{23}N12_{12}) just by changing the oxygen partial pressure during the deposition without undesirable transition to Cu4_4O3_3 or CuO. This results in the independent tuning of optical properties

(for instance, EgE_g from 2.0 to 2.5 eV) while maintaining a very low electrical resistivity of 5×10−25 \times10^{-2} Ωcm - 2×10−12 \times10^{-1} Ωcm.

  • The results from spectroscopic ellipsometry, supported by DFT calculations, clearly show that nitrogen is responsible for the new acceptor states in the form of substitution of atomic N for O or substitution of N2_2 molecule for Cu

  • The nitrogen molecule seems to be more important in the films grown under a high-energy regime and still is at least partially responsible for the increase in hole concentration in Cu2_2O:N films

CRediT authorship contribution statement

Jiří Rezek: Conceptualization, Methodology, Investigation, Visualization, Writing – original draft, Writing – review & editing. Jan Koloros: Investigation, Experiment Jiří Houška: Methodology, Investigation, Visualization, Writing – original draft. Radomír Čerstvý: Experiment. Stanislav Haviar Investigation. David Kolenatý: Experiment, Writing – original draft. Jemal Yimer Damte: Investigation, Visualization, Writing – original draft. Pavel Baroch: Supervising, Conceptualization.

Acknowledgement

This work was supported by the project Quantum materials for applications in sustainable technologies (QM4ST), funded as project No. CZ.02.01.01/00/22_008/0004572 by Programme Johannes Amos Comenius, call Excellent Research.

References

  1. [1]J. Singh, P. Bhardwaj, R. Kumar, V. Verma, Progress in Developing Highly Efficient p-type TCOs for Transparent Electronics: A Comprehensive Review, J Electron Mater (2024). https://doi.org/10.1007/s11664-024-11445-7.
  2. [2]Y. Wang, J.F. Pierson, Binary copper oxides as photovoltaic absorbers: Recent progress in materials and applications, J Phys D Appl Phys 54 (2021). https://doi.org/10.1088/1361-6463/abf165.
  3. [3]J. Rezek, M. Kučera, T. Kozák, R. Čerstvý, A. Franc, P. Baroch, Enhancement of hole mobility in high-rate reactively sputtered Cu2O thin films induced by laser thermal annealing, Appl Surf Sci 664 (2024) 160255. https://doi.org/10.1016/j.apsusc.2024.160255.
  4. [4]J. T-Thienprasert, S. Limpijumnong, Identification of nitrogen acceptor in Cu2O: First-principles study, Appl Phys Lett 107 (2015). https://doi.org/10.1063/1.4936760/29117.
  5. [5]M.A. Islam, Y.A. Wahab, M.U. Khandaker, A. Alsubaie, A.S.A. Almalki, D.A. Bradley, N. Amin, High mobility reactive sputtered CuXO thin film for highly efficient and stable perovskite solar cells, Crystals (Basel) 11 (2021) 1–13. https://doi.org/10.3390/cryst11040389.
  6. [6]A.S. Reddy, S. Uthanna, P.S. Reddy, Properties of dc magnetron sputtered Cu2O films prepared at different sputtering pressures, Appl Surf Sci 253 (2007) 5287–5292. https://doi.org/10.1016/J.APSUSC.2006.11.051.
  7. [7]B.S. Li, K. Akimoto, A. Shen, Growth of Cu2O thin films with high hole mobility by introducing a low-temperature buffer layer, J Cryst Growth 311 (2009) 1102–1105. https://doi.org/10.1016/J.JCRYSGRO.2008.11.038.
  8. [8]S. Chen, L. Wang, C. Zhou, J. Yang, A review of Cu2O solar cell, Journal of Renewable and Sustainable Energy 15 (2023). https://doi.org/10.1063/5.0167383/2920882.
  9. [9]A. Lakshmanan, Z.C. Alex, S.R. Meher, Cu2O thin films grown by magnetron sputtering as solar cell absorber layers, Mater Sci Semicond Process 148 (2022) 106818. https://doi.org/10.1016/J.MSSP.2022.106818.
  10. [10]Y. Wang, P. Miska, D. Pilloud, D. Horwat, F. Mücklich, J.F. Pierson, Transmittance enhancement and optical band gap widening of Cu2O thin films after air annealing, J Appl Phys 115 (2014). https://doi.org/10.1063/1.4865957/138680.
  11. [11]A.M. Koshy, A. Sudha, P. Gollapalli, S.K. Yadav, P. Swaminathan, Annealing-induced changes in optoelectronic properties of sputtered copper oxide films, Journal of Materials Science: Materials in Electronics 33 (2022) 13539–13546. https://doi.org/10.1007/S10854-022-08288-5/TABLES/3.
  12. [12]K. K. Markose, M. Shaji, S. Bhatia, P.R. Nair, K.J. Saji, A. Antony, M.K. Jayaraj, Novel Boron-Doped p-Type Cu2O Thin Films as a Hole-Selective Contact in c-Si Solar Cells, ACS Appl Mater Interfaces 12 (2020) 12972–12981. https://doi.org/10.1021/acsami.9b22581.
  13. [13]M. Nyborg, A. Azarov, K. Bergum, E. Monakhov, Deposition and characterization of lithium doped direct current magnetron sputtered Cu2O films, Thin Solid Films 722 (2021). https://doi.org/10.1016/j.tsf.2021.138573.
  14. [14]T. Minami, Y. Nishi, T. Miyata, Impact of incorporating sodium into polycrystalline p-type Cu2O for heterojunction solar cell applications, Appl Phys Lett 105 (2014). https://doi.org/10.1063/1.4902879/812762.
  15. [15]G. Lai, Y. Wu, L. Lin, Y. Qu, F. Lai, Low resistivity of N-doped Cu2O thin films deposited by rf-magnetron sputtering, Appl Surf Sci 285 (2013) 755–758. https://doi.org/10.1016/J.APSUSC.2013.08.122.
  16. [16]J. Benz, K.P. Hering, B. Kramm, A. Polity, P.J. Klar, S.C. Siah, T. Buonassisi, The influence of nitrogen doping on the electrical and vibrational properties of Cu2O, Physica Status Solidi (b) 254 (2017) 1600421. https://doi.org/10.1002/PSSB.201600421.
  17. [17]Y. Wang, J. Ghanbaja, D. Horwat, L. Yu, J.F. Pierson, Nitrogen chemical state in N-doped Cu2O thin films, Appl Phys Lett 110 (2017). https://doi.org/10.1063/1.4979140/33252.
  18. [18]F. Ye, J.-J. Zeng, Y.-B. Qiu, X.-M. Cai, B. Wang, H. Wang, D.-P. Zhang, P. Fan, Y.-Z. Xie, X.-F. Ma, F. Wang, The optical and electrical properties of nitrogen-doped cuprous oxide annealed at different temperatures, (2018). https://doi.org/10.1016/j.surfcoat.2018.12.109.
  19. [19]M. Soltanmohammadi, N. Nilius, Nitrogen doping of cuprous oxide films: A surface science perspective, Appl Surf Sci 681 (2025) 161507. https://doi.org/10.1016/J.APSUSC.2024.161507.
  20. [20]F.C. Carreri, A. Sabelfeld, H. Gerdes, R. Bandorf, M. Vergöhl, G. Bräuer, HIPIMS ITO films from a rotating cylindrical cathode, Surf Coat Technol 290 (2016) 65–72. https://doi.org/10.1016/j.surfcoat.2015.10.069.
  21. [21]C.-H. Wu, F.-C. Yang, W.-C. Chen, C.-L. Chang, Influence of oxygen/argon reaction gas ratio on optical and electrical characteristics of amorphous IGZO thin films coated by HiPIMS process, Surf Coat Technol 303 (2016) 209–214. https://doi.org/10.1016/j.surfcoat.2016.03.089.
  22. [22]M.J. Zhao, J. Huang, J.F. Zhang, C.H. Hsu, W.Y. Wu, P.H. Huang, S.F. Wei, S.Y. Lien, W.Z. Zhu, Effect of oxygen flow rate ratio on crystalline phase and properties of copper oxide films prepared by room-temperature high-power impulse magnetron sputtering, Surf Coat Technol 434 (2022) 128215. https://doi.org/10.1016/J.SURFCOAT.2022.128215.
  23. [23]J. Rezek, J. Houska, M. Prochazka, S. Haviar, T. Kozák, P. Baroch, In-Ga-Zn-O thin films with tunable optical and electrical properties prepared by high-power impulse magnetron sputtering, Thin Solid Films 658 (2018) 27–32. https://doi.org/10.1016/j.tsf.2018.05.029.
  24. [24]J. Rezek, P. Novák, J. Houška, A.D. Pajdarová, T. Kozák, High-rate reactive high-power impulse magnetron sputtering of transparent conductive Al-doped ZnO thin films prepared at ambient temperature, Thin Solid Films 679 (2019) 35–41. https://doi.org/10.1016/j.tsf.2019.04.009.
  25. [25]J.P. Perdew, K. Burke, M. Ernzerhof, Generalized Gradient Approximation Made Simple, Phys Rev Lett 77 (1996) 3865. https://doi.org/10.1103/PhysRevLett.77.3865.
  26. [26]G. Kresse, D. Joubert, From ultrasoft pseudopotentials to the projector augmented-wave method, Phys Rev B 59 (1999) 1758. https://doi.org/10.1103/PhysRevB.59.1758.
  27. [27]A. Hassani, M.T. Hamed Mosavian, A. Ahmadpour, N. Farhadian, A comparative theoretical study of methane adsorption on the nitrogen, boron and lithium doped graphene sheets including density functional dispersion correction, Comput Theor Chem 1084 (2016). http://profdoc.um.ac.ir/paper-abstract-1056525.html (accessed May 29, 2025).
  28. [28]J. Rezek, J. Vlček, J. Houška, R. Čerstvý, High-rate reactive high-power impulse magnetron sputtering of Ta-O-N films with tunable composition and properties, Thin Solid Films 566 (2014) 70–77.DOI
  29. [29]A.S. Zhakypov, R.R. Nemkayeva, Y. Yerlanuly, M.A. Tulegenova, B.Y. Kurbanov, M.B. Aitzhanov, A.A. Markhabayeva, M.T. Gabdullin, Synthesis and in situ oxidation of copper micro- and nanoparticles by arc discharge plasma in liquid, Sci Rep 13 (2023). https://doi.org/10.1038/s41598-023-41631-2.
  30. [30]J. Rezek, T. Kozak, N. Kumar, S. Haviar, Synergy of experiment and model for reactive HiPIMS: Effect of discharge parameters on WOx composition and deposition rate, J Phys D Appl Phys 54 (2021). https://doi.org/10.1088/1361-6463/abd1a3.
  31. [31]Z. Chen, H.N. Dinh, E. Miller, Photoelectrochemical water splitting: standards, experimental methods, and protocols, (2013) 126. https://doi.org/10.1007/978-1-4614-8298-7.
  32. [32]A. Larabi, A. Mahmoudi, M. Mebarki, M. Dergal, First principle investigation of hydrogen behavior in M doped CuS$_2$2$O (M $=$ Na, Li and Ti), Condens Matter Phys 22 (2019) 1–11. https://doi.org/10.5488/CMP.22.23702.

Paper details

vlnc:W_efhb6mpb
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 9

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents