Plasma Jet Sputtering as an Efficient Method for the Deposition of Nickel and Cobalt Mixed Oxides on Stainless-Steel Meshes: Application to VOC Oxidation

Abstract

Hollow cathode plasma sputtering is an advantageous method of preparing catalysts in the form of thin oxide films on supports. Such catalysts are particularly suitable for processes such as catalytic total oxidation of volatile organic compounds (VOCs), representing an economically feasible and environmentally friendly method of VOC abatement. Catalysts with Ni:Co molar ratios of 1:4, 1:1, and 4:1 were prepared on stainless-steel meshes and compared with single-component Ni and Co oxide catalysts. The properties of the catalysts were characterized by EDX, SEM, powder XRD, temperature-programmed reduction (H₂-TPR), Raman spectroscopy, and XPS. Powder XRD revealed the formation of various crystalline phases that were dependent on molar the Ni:Co ratio. NiO and Co₃O₄ were identified in the single-component Ni and Co oxide catalysts, whereas Ni-Co mixed oxides with a spinel structure, together with NiO, were found in the catalysts containing both Ni and Co. Raman spectra of the catalysts prepared at high working pressures showed a slightly lower intensity of bands, indicating the presence of smaller oxide particles. The TPR profiles confirmed the improved reducibility of the Ni-Co oxide catalysts compared to the single-component Ni and Co catalysts. Catalytic activity was investigated in the deep oxidation of ethanol and toluene, which were used as model volatile organic compounds. In ethanol oxidation, the activity of sputtered catalysts was up to 16 times higher than that of the commercial Cu-Mn oxide catalyst EnviCat^® VOC-1544. The main benefits of the sputtered catalysts are the much lower content of Ni and Co oxides and a negligible effect of internal diffusion. Moreover, the process of plasma jet sputtering can be easily implemented on a large scale.

1. Introduction

Mixed oxides of transition metals are used as catalysts in many chemical reactions, such as hydrogenation, oxidation, and water gas shift, or as gas purification adsorbents. They also make an important contribution to improving the quality of the environment because they are used in the oxidation of volatile organic compounds (VOCs). In particular, oxides of Mn, Co, Cu, Ni, V, Cr, Fe, and their various mixtures are active catalysts in this process [1]. We previously showed that cobalt oxide [2], Co-Mn mixed oxides [3], and Co-Cu oxides [4] prepared by precipitation and subsequent calcination or by magnetron sputtering are active in the complete oxidation of ethanol in air, arising, e.g., from pharmaceutical plants. In these processes, pelletized catalysts are most commonly used, but a significant part of the active mass in the pellets is unused in the process due to the large effect of internal diffusion limitations.

The nickel–cobalt oxide system is most commonly used in the process of dry reforming of methane (DRM) to produce synthesis gas [5,6,7], selective hydrogenation of pyrolysis gasoline [8], or ethanol decomposition reactions [9]. Owing to their similar electronic configuration, Co and Ni can easily form alloyed bimetallic particles that exhibit superior performance in DRM reactions. Such catalysts show outstanding performance, which is ascribed to the strong metal-support interaction and synergy between Ni and Co active metals. Moreover, the presence of Co in Ni-Co catalysts eliminates carbon deposition due to the enhancement of CO₂ adsorption/reaction in DRM. It has been proven that Ni-Co catalysts contain an active phase with small particle size and excellent dispersion [10]. Recently, Jianrui Niu et al. [11] demonstrated that NiCo₂O₄ spinel was very efficient for toluene oxidation.

Low-temperature plasma sputtering is one of the most important methods used in industry to deposit thin films on various substrates. Using a planar magnetron, thin films of metals or dielectrics can be reliably prepared over large areas [12]. This deposition system is used to deposit, e.g., tribological, catalytic, optical, or decorative coatings [13,14,15]. An alternative deposition method based on sputtering the use of a metal nozzle with a flowing gas has attracted attention in the past. Inside the nozzle, a very intense discharge is ignited, and the working gas (typically argon) flows into a vacuum chamber, where a so-called plasma jet [16] entrains the plasma. It should be noted that the pressure inside the hollow cathode is up to two orders of magnitude higher than that in the vacuum chamber, leading to the formation of nanoclusters of sputtered particles that are deposited on the substrate, forming nanostructured thin films [17]. Extensive diagnostics of plasma jets have been performed in the past, showing, for example, a different EEDF compared to a magnetron [18]. Plasma jet sputtering is a technique that has also been widely investigated for thin film deposition; for example, textured ZnO thin films [19], high-quality polycrystalline Si thin films [20], nanostructured Co₃O₄ thin films with a large surface area [21], piezoelectric PZT ceramics on a polymer substrate [22], and nanostructured Fe₂O₃ thin films as photoanodes for solar water splitting [23] have been prepared. However, plasma jets have not been applied for the preparation of heterogeneous catalysts.

In the view of the activity of Ni-Co catalysts in the abovementioned reactions, we focused on the study of supported Ni-Co oxide catalysts deposited as thin films on stainless steel. Stainless-steel meshes were chosen as a support, as they have a lower pressure drop of the catalytic bed than the bed of shaped pellets. Catalysts in the form of meshes are used especially when large linear velocities of reactants are applied in catalytic reactors.

In this study, we aimed to investigate the effect of the Ni:Co molar ratio on the properties of Ni-Co oxide catalysts prepared by plasma jet sputtering. The selected active components, i.e., nickel and cobalt, are readily available metals from which the targets required for sputtering can be well prepared; therefore, the process can thus be easily implemented on a large scale. The influence of the working conditions of catalyst preparation, especially the working gas pressure, on the oxide deposition process was also investigated in detail. Catalysts with different Ni:Co molar ratios were designated as NiCo14, NiCo11, and NiCo41 (the numbers indicate the intended Ni:Co molar ratio). Single-component Ni and Co oxide catalysts (labeled as Ni and Co, respectively) were also prepared. To reveal the effect of the gas pressure applied during thin film deposition on the properties of the catalysts, the catalysts were prepared at different pressures. The pressure (in Pa) at which the sample was prepared was indicated by a suffix in the sample name, e.g., NiCo11₁₅. In addition to the standard deposition time of 120 min, the NiCo11₁₅ catalyst was also prepared with an extended deposition time of 370 min to achieve an increased amount of oxides on the support. This sample was designated as NiCo11₁₅*. For the same purpose, the NiCo11₁₆₀ catalyst was prepared using a higher working pressure of 160 Pa and a deposition time of 200 min. Details of the catalyst preparation process are given in the Experimental section.

2. Results

2.1. Plasma Jet Sputtering of Ni-Co Oxides on Stainless-Steel Meshes

The key parameters affecting the amount of metal oxide thin film deposited on the support are composition of the hollow cathode, deposition pressure, power absorbed in the discharge, and deposition time. Deposition experiments revealed that the material composition of the hollow cathode has a significant impact on the discharge behavior. For example, in the case of a pure Ni hollow cathode, we observed that parasitic discharge is often ignited, which causes plasma instability and aggravates the sputtering of the hollow cathode. Therefore, we used a lower discharge current of 1.05 A to suppress undesired parasitic discharge and ensure more stable deposition conditions.

To determine the effect of the material composition of the hollow cathode on the deposition rate of the thin metal oxide film, we calculated the amount of deposited material normalized to kWh of energy delivered to the discharge of the hollow cathode. The mass deposited on the support per kWh as a function of the hollow cathode alloy is shown in Figure 1. For a deposition pressure of 80 Pa, the mass deposited on the mesh increased linearly from 0.6 mg/kWh to 1.15 mg/kWh with increasing amount of nickel in the hollow cathode alloy increased, followed by a sudden drop to 0.45 mg/kWh for the pure nickel cathode. A lower deposition pressure of 15 Pa resulted in a slightly different deposition rate depending on the material of the hollow cathode. The pure Co hollow cathode and Ni20Co80 alloy exhibited a roughly 20% higher deposition rate (mg/kWh) compared to the 80 Pa deposition pressure, whereas the deposition rates for pure Ni and Ni80Co20 alloy were virtually comparable to those for the 80 Pa deposition pressure. On the other hand, we observed an approximately 50% decrease in deposition rate for Ni50Co50 alloy at a deposition pressure of 15 Pa, possibly due to the plasma instability. The same is true for the pure Ni hollow cathode, which had the lowest deposition rate.

2.2. Composition of the Catalysts

A commonly used method for the investigation of catalyst composition, i.e., the determination of Ni and Co concentrations by atomic absorption spectroscopy (AAS) after the dissolution of the samples could not be used because the stainless-steel support contained nickel as one of the major components. Moreover, the amount of oxides deposited on the mesh was very low compared to the total weight of the catalyst. Therefore, the sputtered catalysts were analyzed by EDX (Table S1). Analysis of the bare stainless-steel support showed that it contained 6.50 at.% Ni and; therefore, the EDX analysis of catalytically active oxide layers had to be corrected accordingly. The Ni/Fe intensity ratio (Kα line) for the mesh was equal to 0.07. Because the Fe in all samples originated from the support, the Ni intensity in the samples was reduced by the Fe intensity multiplied by the Ni/Fe intensity ratio (from the support itself). This procedure allowed us to determine an approximate correction to the Ni concentration as the ratio between the reduced and measured Ni intensity (Kα). This correction was performed for an approximate and rapid determination of the Ni content in the deposited oxide layers, although it may have led to a reduced accuracy of the determination. The Ni:Co molar ratios in the deposited oxides determined by the EDX method were close to the nominal values in the used cathodes (

Table 1. Ni and Co concentrations in the oxide layers deposited on meshes in at% (determined by EDX).

Sample	Nia at. %	Co at. %	Ni/(Ni+Co)EDX mol mol−1	Ni:CoEDX mol mol−1
Co15	0.0	31.5	0	1:0
NiCo1415	7.5	28.5	0.21	1:3.80
NiCo1115	15.6	14.7	0.52	1:0.92
NiCo4115	27.2	6.8	0.80	4:1
Ni15	29.8	0.2	1	0:1
NiCo1115 *	20.2	19.5	0.50	1:0.97
Co80	0.0	28.1	0	1:0
NiCo1480	6.3	24.0	0.21	1:3.80
NiCo1180	18.5	17.6	0.51	1:0.96
NiCo4180	28.1	7.2	0.80	4:1
Ni80	27.2	0.3	1	0:1
NiCo11160	19.0	19.2	0.50	1:1.01

^a Ni average atomic concentration in the respective catalysts (correction factor applied) (at%); NiCo11₁₅ * deposition time: 370 min.

).

2.3. Powder X-ray Diffraction

Powder XRD patterns of catalysts prepared using hollow cathode plasma jet sputtering in the oxidation Ar+O₂ atmosphere and calcined at 500 °C in air are shown in Figure 2. The most intensive diffraction lines observed in powder XRD patterns of all catalysts at 51.0° and 59.7° 2θ corresponded to iron (PDF 04–014–0264) or various iron-containing alloys such as Fe_0.65Cr_0.1Ni_0.25 (PDF 04–019–2390) and were ascribed to the stainless-steel support. The Ni₁₅ and Ni₈₀ catalysts showed diffraction peaks characteristic of NiO (bunsenite) at approximately 43.6 and 74.7° 2θ (PDF 04–005–4472); the most intensive NiO diffraction line at approximately 50.6° 2θ coincided with that of the stainless-steel support. Powder XRD patterns of the Co₁₅ and Co₈₀ catalysts contained diffraction peaks at 22.1, 36.4, 43.1, 52.4, 65.6, 70.2, and 77.5° 2θ, corresponding to Co₃O₄ with a spinel structure (PDF 04–014–7752).

In the catalysts combining Ni and Co active components, both NiO-like and spinel-type mixed oxides (e.g., Co₃O₄, Ni_0.82Co_2.18O₄, and NiCo₂O₄) were revealed. Their relative content and lattice parameters were dependent on the Ni:Co molar ratio in the deposited oxides (

Table 2. Crystallite sizes (D) and lattice parameters (a) of the NiO-like and Ni-Co spinel-type oxides evaluated according to the powder XRD patterns of the catalysts. The NiCo11₁₅ * catalyst was prepared with an extended deposition time of 370 min.

Sample	NiO		Ni-Co Spinel
	D (nm)	a (Å)	D (nm)	a (Å)
NiCo1415	-	-	20.3	8.119
NiCo1115	-	-	19.6	8.104
NiCo4115	17.5	4.180	-	-
NiCo1115 *	15.4	4.175	17.1	8.120
NiCo1480	-	-	24.2	8.122
NiCo1180	18.7	4.171	15.8	8.109
NiCo4180	14.2	4.179	11.4	8.104
NiCo11160	15.3	4.181	17.3	8.129

NiCo11₁₅ * catalyst was prepared with an extended deposition time of 370 min.

). The catalysts with low Ni content prepared at 15 Pa (NiCo14₁₅ and NiCo11₁₅) contained only spinel-like NiCo₂O₄ (PDF 04–013–0797), whereas the NiCo41₁₅ with high Ni content showed only relatively low and broad diffraction peaks corresponding to NiO. The NiCo11₁₅* catalyst prepared using a longer deposition time, which led to an increased amount of deposited oxides, contained NiO and Ni-Co spinel. The catalysts prepared at a higher working pressure (80 Pa) contained analogous crystalline phases to those of samples prepared at 15 Pa; a small amount of Ni-Co spinel was also identified in the NiCo41₈₀ sample. The catalysts prepared under a deposition pressure of 160 Pa (NiCo11₁₆₀ sample) with a longer deposition time (NiCo11₁₅* sample) showed more intensive diffraction peaks compared to other samples due to higher content of deposited oxides. Both these samples contained Ni-Co spinel, together with NiO.

Lattice parameters of the oxide phases present in the catalysts were evaluated according to their powder XRD patterns (Table 2). The NiO and spinel lattice parameters were calculated as the average value of all identified diffraction peaks. The NiO-like phase in the Ni-Co samples showed lattice parameters ranging from 4.171 to 4.181 Å. Fedorov et al. [24] reported a lattice parameter of 4.173 Å for nickel oxide on alumina. The slightly larger values found in the NiCo41₁₅ and NiCo41₈₀ samples could be explained by the partial incorporation of larger cobalt cations into the NiO lattice (Co⁺ and Ni²⁺ ionic radii of 0.72 and 0.69 Å, respectively). Analogously, the changes in the spinel lattice parameters from 8.104 to 8.129 Å were likely connected with various contents of Ni and Co cations in the Ni-Co mixed oxides.

The mean coherence lengths evaluated according to powder XRD data reflected the ordering of the structure of the crystalline oxide phases and can be considered as approximate crystallite sizes. The values were calculated based on the integral width (B_obs) of the diffraction lines determined by HighScore Plus software once the powder XRD patterns were subjected to standard profile fitting. The next step involved calculating the crystallite sizes using the Scherrer equation, taking into account the integral width values of each diffraction peak. The largest spinel crystallite size (24.2 nm) was found for the NiCo14₈₀ sample, and the smallest crystallite size (11.4 nm) was associated with the NiCo41₈₀ catalyst (Table 2). Structure ordering of the spinel-type Ni-Co mixed oxides, reflecting their calculated crystallite sizes, could be affected by the interaction between NiO and Co₃O₄ arising during plasma jet sputtering and following calcination, resulting in mixed oxide formation. This interaction could also explain the observed slight change in the spinel lattice parameters. On the contrary, differences in the crystallite size of the NiO-like phase were not so significant.

2.4. Surface Morphology and Textural Properties

SEM images of the meshes with deposited NiCo11 oxide layers prepared using the gas pressure of 15 and 80 Pa are shown in Figure 3. These two samples were compared with NiCo11₁₅* also prepared under 15 Pa but with three times longer deposition time than the NiCo11₁₅ catalyst. The appearance of oxide particles in the samples was very similar; in the case of the NiCo11₁₅ catalyst, the small clusters of oxides were connected with even smaller particles, forming a homogeneous layer covering the mesh. Using a higher working pressure (80 Pa) to prepare the NiCo11₈₀ catalyst of identical composition, the larger particles formed a thicker layer, which indicated a tendency to crack. Such a process is more distinguishable at in the image of the NiCo11₁₅* catalyst. A detailed SEM image of this sample shows a cracked oxide layer due to the increased thickness. There were small, almost spherical particles on the oxide layer covering the surface of the mesh, which seem to have led to the formation of a new oxide layer.

Figure 3 (bottom) shows images of the two NiCo14 catalysts prepared at 15 and 80 Pa and of the NiCo11₁₆₀ catalyst, which was prepared at a working pressure of 160 Pa. On the surface of the NiCo14₁₅ catalyst, significantly larger particles can be clearly distinguished compared to the NiCo11₈₀ catalyst. Finally, the surface of the NiCo11₁₆₀ catalyst is similar to that of the NiCo11₁₅* catalyst.

The difference in the size of the deposited particles shown in Figure 3 was reflected in their textural properties (

Table 3. Surface area, initial reduction temperatures (T_onset), reduction temperature of maxima (T_max), and H₂ consumption determined during TPR measurements for the Ni-Co catalysts.

Sample	SBET m2 goxide−1	Tonset °C	Tmax °C	mmol H2 goxide−1
Co15	2.6	242	359	16.4
NiCo1415	8.1	163	323	17.6
NiCo1115	7.9	147	315	19.8
NiCo4115	8.8	190	339	14.2
Ni15	2.2	314	371	15.7
NiCo1115 *	2.9	240	314	15.1
Co80	2.8	162	376	20.0
NiCo1480	9.4	195	328; 346; (512)	18.4
NiCo1180	10.0	137	301; 316	22.8
NiCo4180	9.4	199	328; 356	12.4
Ni80	4.1	342	364	13.1
NiCo11160	2.8	245	320	12.5

* Deposition time: 370 min.

). The surface area of oxides deposited on the meshes varied from 2.2 to 10.0 m² g_oxide⁻¹. The lowest surface area was associated with the single-component Co and Ni catalysts prepared at both deposition pressures, whereas the catalysts prepared with the Ni-Co alloys as hollow cathode material showed the highest values. An approximately linear dependence of the absolute value of the surface area on the amount of oxides deposited on the meshes was found.

2.5. Temperature-Programmed Reduction

The reduction properties of the catalysts were investigated using H₂ temperature-programmed reduction. The maximum in the profiles corresponds to the maximum of the reduction rate, and the shift of the position of temperature maximum indicates a change in the metal–oxygen bond strength [25]. The TPR profiles (Figure 4) reflect the differing contents of Ni and Co in the catalysts. The single-component Co catalyst prepared at 15 Pa showed one main reduction peak with T_max 362 °C, similarly to our previously published results [26]. This peak was attributed to the unresolved reduction of Co³⁺ to Co²⁺ and Co²⁺ to Co⁰ [25,27]. The reduction profile of the single-component Ni catalyst also showed one unresolved peak with maximum at 372 °C. According to Richardson et al. [28], the nickel oxide reduction mechanism includes direct Ni⁰ and H₂O formation. Thus, the higher reduction temperature maximum of the Ni oxide-based catalyst shows poorer reducibility than the Co-based catalyst. Garces et al. [29] observed the same behavior.

Numerous studies on Ni-Co mixed oxide reduction by the H₂-TPR method have been reported. For example, Goncalves et al. [30] found that the reduction of Ni-Co oxides supported on alumina depends on the molar ratio of Ni:Co and that the reduction of Ni-Co oxide particles was easier than the reduction of Ni and Co oxides alone. As shown in Figure 4, the TPR profiles of the examined Ni-Co catalysts sputtered at 15 Pa changed depending on the phase composition. The Ni-Co mixed oxides were reduced at lower temperatures of between 315 and 338 °C. Their reduction peaks were sharp; only the profile of NiCo41₁₅ showed enlargement towards higher temperatures, indicating possible reduction of another phase—probably NiO—which was identified in the catalyst by XRD (Figure 2).

The catalysts prepared at 80 Pa exhibited a very similar course of reduction as those prepared at 15 Pa, with a slight shift to higher temperatures with an increasing amount of cobalt in the catalysts. The reduction of the binary Ni-Co oxide catalysts showed an indication of peak splitting into two parts, which may involve either the reduction of several different components present in the catalysts or the reduction of particles of different sizes.

2.6. Raman Spectra

The Raman spectra of all catalysts (Figure 5) exhibited an intense band at 674 cm⁻¹, which corresponds to the presence of Co₃O₄ [31,32], and a band at 484 cm⁻¹ that can be ascribed to the first-order transverse optical phonon mode of NiO [33,34]. The spectra of NiCo41₁₅ and NiCo41₈₀ catalysts featured an intense Raman band at 530 cm⁻¹, corresponding to the first-order longitudinal optical vibrational mode of NiO [33,34]. This vibration was also present as a broad, weak band in the spectra of the NiCo11₁₅ and NiCo11₈₀ catalysts but less pronounced in the spectra of the NiCo14₁₅ and NiCo14₈₀ catalysts. The broad band at 1090 cm⁻¹ in the spectra of the NiCo41₁₅ and NiCo41₈₀ catalysts, together with the shoulder at 706 cm⁻¹ in the spectrum of the NiCo41₁₅ catalyst, can be attributed to the combination of transverse and longitudinal second-order optical vibrational modes of NiO [35]. Finally, the band at 193 cm⁻¹, which is particularly pronounced in the spectra of the NiCo41₁₅ and NiCo41₈₀ catalysts, can be attributed to the tetrahedral sites of Co₃O₄ [4].

There is a significant difference in the width and intensity of the bands. The broadening of the peaks and shifts in the positions of the peak maxima compared to the Co₃O₄ spectrum indicate the formation of Ni-Co spinel, which is consistent with the results of XRD measurements.

As the intensity is also connected with the size of particles [35], we can conclude that the binary Ni-Co oxide catalysts contain smaller particles than the single-component Ni and Co catalysts. According to the Raman spectra, the smallest particles were revealed in the NiCo14₈₀ catalyst.

2.7. Surface Composition of the Catalysts

XPS measurements were used to determine the surface composition of the catalysts (Figure S1). XPS results corresponding to the binding energies (BE) of several core levels (C 1s, Ni 2p, Co 2p, and O 1s), surface molar fractions of each element, and atomic ratios obtained for the examined catalysts are summarized in

Table 4. Surface concentrations of elements (in at. %) and comparison of the surface (XPS), bulk (EDX), and nominal bulk concentrations expressed as Ni/(Ni+Co) molar ratios.

Sample	C 1s	O 1s	Co 2p	Ni 2p	Ni/(Ni + Co)nom	Ni/(Ni + Co)EDX	Ni/(Ni + Co)XPS
	At.%	At.%	At.%	At.%	mol/mol	mol/mol	mol/mol
Co15	50.79	38.02	11.19	0	0	0	-
NiCo1415	54.62	32.94	8.07	4.38	0.2	0.21	0.35
NiCo1115	50.47	34.42	6.63	8.48	0.5	0.52	0.56
NiCo4115	45.61	33.28	2.34	18.78	0.8	0.8	0.89
Ni15	47.94	34.71	0	17.35	1	1	-
NiCo1115 *	50.66	32.75	5.34	11.24	0.5	0.50	0.68
Co80	52.27	36.74	10.99	0	0	0	-
NiCo1480	53.57	34.36	8.32	3.75	0.2	0.21	0.31
NiCo1180	42.64	36.00	5.53	15.83	0.5	0.51	0.74
NiCo4180	49.22	31.92	2.14	16.71	0.8	0.80	0.89
Ni80	52.95	30.97	0	16.08	1	1	-
NiCo11160	49.43	36.05	5.32	9.2	0.5	0.50	0.63

The NiCo11₁₅* catalyst was prepared with an extended deposition time of 370 min.

. No iron signal was detected in any of the spectra of the examined samples.

XPS measurements were complicated by the fact that the samples were in the form of meshes. The signal of the measured carbon largely originated from the underlying carbon tape, which was easily accessible through the mesh. This large excess of carbon originating from the carbon tape led to increased uncertainty upon carbon calibration (setting the main peak position to 284.8 eV), as the carbon on the surface of the mesh may have a spectrum slightly differently shifted than the spectrum from the carbon tape. Binding energies of approximately 287 and 288 eV were observed in the carbon spectrum, which are usually assigned to C−O and C=O bonds.

Binding energies of Ni and Co in the catalysts were observed at 853.7 and 855.2 eV (main lines for Ni 2p_3/2) and at 779.5 and 781.0 eV (main lines for Co 2p_3/2), respectively, which are typical for the oxidation states of Ni²⁺ and Co³⁺/Co²⁺. For some samples, the binding energy of Ni 2p_3/2 was shifted towards higher energies than those reported in the literature [36] for NiO. This may indicate the presence of Ni hydroxide, as its peak could be found at higher energy levels than the oxide peak (approximately 854.6 eV). Because the spectra have a similar shape to that of NiO, the shift was probably due to the formation of OH groups on the surface after exposure to air. The Ni and Co concentrations did not correspond to the expected composition of NiCo11₁₅ and NiCo11₈₀ catalysts. Remeasurement of these samples showed the same spectra for Co and Ni. The ratios of the Ni and Co surface and bulk concentrations are presented in Table 4. A comparison of these quantities with EDX analysis showed that the surface of the catalysts was enriched with Ni. For both NiCo14₁₅ and NiCo14₈₀ catalysts, the surface concentration of nickel was found to be 1.5 times higher than the bulk concentration. This ratio decreased to 1.1 with increasing nickel content in the catalysts.

The oxidation state of Co and Ni in the catalysts was determined by deconvolution of the XPS spectra. Deconvolution of the Ni 2p spectra of all catalysts (Figure S1) showed no difference between the spectra (except for the intensity of the peaks). The results (

Table 5. Relative atomic concentration of deconvoluted Co 2p and O 1s peaks (at. %), molar ratio of Co³⁺/Co²⁺, and oxygen vacancies O_v related to surface concentration of (Ni+Co) ions.

Sample	Co3+	Co2+	Co2+	Co3+/Co2+	Ol	Ov	Ow	Ov/(Ni+Co)
BE, eV	779.5 ± 0.1	781.0	783.1	Peak 1/(2 + 3)	529.0–529.9	530.5–531.0	531.7–533.0
Co15	34.44	25.99	2.83	1.2	34.5	0.0	65.5	0.0
NiCo1415	33.39	23.44	5.42	1.16	41.3	24.2	34.5	1.9
NiCo1115	30.79	25.24	6.04	0.98	42.5	22.3	35.2	1.5
NiCo4115	25.81	21.76	6.53	0.91	50.5	25.5	24.0	1.2
Ni15	-	-	-	-	44.5	23.3	32.3	1.3
NiCo1115 *	32.07	23.87	3.03	1.2	46.3	19.3	34.4	1.2
Co80	33.95	25.15	2.3	1.2	34.0	0.0	66.0	0.0
NiCo1480	35.70	24.01	2.4	1.35	42.1	25.6	32.3	2.1
NiCo1180	31.29	26.4	4.4	1.02	52.0	29.3	18.8	1.4
NiCo4180	22.87	21.62	5.36	0.85	45.3	24.4	30.3	1.3
Ni80	-	-	-	-	38.9	18.5	42.6	1.2
NiCo11160	28.33	25.56	6.2	0.9	32.4	3.8	63.8	0.3

) led to the conclusion that Ni was present only in the form of NiO, which was confirmed by the structure of the Ni 2p_3/2 line fitted by two nearby peaks and a satellite contribution. The determined NiO position (853.7 eV) is consistent with the reported literature data [36].

However, it is not possible to distinguish between NiO and Ni-containing oxide with a cubic spinel structure (e.g., NiFe₂O₄), as demonstrated by Biesinger et al. [37]. The similarity of spectra was also demonstrated between Ni(OH)₂ and spinels with a tetragonal structure, such as NiCr₂O₄. According to our data, the Ni 2p spectrum resembles that of NiO and/or cubic spinel for the NiCo41₁₅ and NiCo41₈₀ samples. On the other hand, the spectra of NiCo14 catalysts more closely resemble those of Ni(OH)₂ and/or tetragonal spinel. XRD measurements confirm the difference in the composition of crystalline phases between the two samples (Table 2). XRD measurements revealed the presence of Ni-Co spinel in NiCo14₈₀ and NiCo11₈₀ catalysts, with the closest match from the PDF database with NiCo₂O₄.

Cobalt can occur as Co²⁺ or Co³⁺ in catalysts, as Co²⁺ species can also be partially oxidized to Co³⁺ species. Co2p spectral deconvolution revealed the presence of both Co²⁺ and Co³⁺; the Co³⁺/Co²⁺ ratio varied from 0.85 to 1.35 (Table 5). The observed position of the most intense Co 2p_3/2 line corresponded to the position of Co₃O₄. The structure of the satellite peaks observed at 940 and 943 eV also confirmed the presence of Co₃O₄. The average values of binding energies of the Co 2p_3/2 published in the literature are 780.17, 780.39, and 779.90 eV for Co₃O₄, CoO, and Co₂O₃, respectively [36]. The oxidation state of cobalt changed only slightly with catalyst composition (Figure S1). In some catalysts, the Co³⁺ oxidation state was slightly predominant (Table 5)—more frequently in the catalysts prepared at 80 Pa.

To confirm the abovementioned findings, LMM spectra (Figure S2) were measured, which are more sensitive to changes in Ni and Co oxidation states and bonding. The spectra of NiCo11 and NiCo41 samples were equally interleaved so that the samples had the same oxidation state of Ni corresponding to NiO. NiCo14₁₅ and NiCo14₈₀ samples had a slightly different spectral shape corresponding to Ni(OH)₂, which appears to occur on the catalyst surface after exposure to air.

The deconvolution of oxygen spectra of the samples is shown in Figure S3 and summarized in Table 5. There are three peaks in the spectra, with binding energies of 529.1–529.3, 530.6–530.9, and 532.0–532.7eV. Oxygen at 529.2 eV corresponds to metal oxides. Atomically, the concentration of this oxygen is equal to the sum of the concentrations of Co and Ni. Biesinger et al. [37] identified oxygen at 531eV as defect sites, adsorbed oxygen, and hydroxides, with most emphasis on the defect sites. According to [37], the binding energy of oxygen in NiO is 529.3 eV and that in CoO is 529.8 eV. Oxygen at 531 eV in the deconvoluted spectra corresponded to C=O bonds or to oxygen bound to metal as O⁻ or OH⁻. The other oxygen appeared at position 533 eV, corresponding to C-O bonds. About one-third to one-half of oxygen may originate from the carbon tape and other carbon impurities. The 532.8eV peak was determined as adsorbed H₂O or O₂. The first peak at 529.2eV is not sufficient to satisfy the oxygen demand for Co and Ni; therefore, some of the oxygen from the second peak must be bound to the metal. However, the sum of the first and second peaks would probably result in excessive oxygen. Therefore, it follows that the second peak contributes only partially to the metal-bound oxygen, with the remainder probably occurring in the form of a hydroxyl group. However, some defects may also shift the oxygen binding energy.

2.8. Catalytic Activity

The objective of this study was to determine the effect of the preparation conditions of plasma jet sputtering and Ni:Co molar ratio in the oxide catalysts on their catalytic activity in the total oxidation of ethanol and toluene. The light-off curves of ethanol and toluene oxidation are shown in Figure S4. It can be seen that for ethanol oxidation, the conversion curves of the catalysts prepared at 15 and 80 Pa are very similar; the NiCo11₁₅* and NiCo11₁₆₀ catalysts are the most active, and single-component Ni₁₅ and Ni₈₀ catalysts are the least active. A similar trend was observed for the oxidation of toluene.

At first, the activity of the catalysts was evaluated according to the T₅₀ temperature (the temperature at which the conversion of the reactant reaches 50%), which is a frequently used parameter in this field of catalysis. The T₅₀ values obtained in the oxidation of ethanol and toluene as a function of the Ni/(Ni+Co) molar ratio are summarized in

Table 6. T₅₀ and T₉₀CO₂ temperatures in the oxidation of ethanol and toluene. Reaction conditions: ca. 3.8 g of sputtered catalysts; 760 ppm of ethanol or 780 ppm of toluene; GHSV 21 L g_cat^−l h^−l; temperature ramp: 2 °C min^−l.

Catalyst	Ni/(Ni + Co)	Ethanol		Toluene
	mol mol−1	T50, °C	T90CO2, °C	T50, °C	T90CO2, °C
Co15	0	260	324	345	402
NiCo1415	0.2	265	335	334	359
NiCo1115	0.5	274	329	364	355
NiCo4115	0.8	269	332	363	396
Ni15	1	296	402	455	>480
NiCo1115 *	0.5	249	301	343	342
Co80	0	258	324	343	379
NiCo1480	0.2	252	329	328	356
NiCo1180	0.5	274	324	338	382
NiCo4180	0.8	270	329	356	389
Ni80	1	296	402	430	>450
NiCo11160	0.5	247	294	314	346

NiCo11₁₅ *—deposition time: 370 min.

. According to the T₅₀ values, the composition of catalysts did not considerably affect their activity in either reaction, which could indicate that the T₅₀ parameter is not very sensitive to changes in the catalyst composition. The oxidation of toluene on Ni, Co, and Ni-Co oxides was more difficult than the oxidation of ethanol; the temperature required to achieve 50% conversion of toluene was approximately 80–100 °C higher. However, both single-component Ni catalysts (Ni₁₅ and Ni₈₀) needed the highest temperature to achieve 50% conversion of both reactants (ethanol and toluene).

The T₉₀CO₂ temperature, reflecting the temperature required to achieve 90% conversion of reactants to CO₂, seems to be a more important parameter when evaluating the performance of catalysts designed for VOC total oxidation. The T₉₀CO₂ temperatures for the catalysts with different Ni:Co molar ratios are presented in Table 6. The values show that the catalysts with Ni/(Ni + Co) = 0.5 prepared at both 15 and 80 Pa have the lowest T₉₀CO₂ temperature. In the case of toluene oxidation, the minimum temperature was reached for the catalysts with the a Ni/(Ni + Co) ratio of 0.2–0.5.

The activity of the catalyst components in the oxidation of ethanol and toluene can be more accurately compared when specific reaction rates (moles of reactant converted per gram of active components per hour) are used, as the amount of oxides varied from catalyst to catalyst. The specific reaction rate (R) was calculated for two reaction temperatures because oxidation of toluene proceeded at higher temperatures than the oxidation of ethanol. Therefore, R₂₀₀ was chosen for ethanol oxidation, and R₃₂₅ was used for the oxidation of toluene. Figure 6 shows that in the series of catalysts prepared under identical conditions differing only in the value of the working pressure, both single-component Co catalysts were the most active, and both single-component Ni catalysts were the least active in the oxidation of ethanol. The activity of binary Ni-Co catalysts decreased with an increasing amount of Ni in the mixed oxides.

In the oxidation of toluene, a similar dependence of catalyst activity on molar Ni/(Ni + Co) ratio was found for the catalysts prepared at 15 Pa (Figure 6); the specific reaction rates decreased proportionally with an increasing amount of Ni in the catalysts. However, the catalysts prepared at 80 Pa showed quite different behavior. The NiCo14₈₀ catalyst and the NiCo11₁₅* and NiCo11₁₆₀ catalysts prepared under slightly different conditions—either nearly three times longer deposition time (370 min) or substantially higher working pressure (160 Pa)—were significantly more active than the analogous catalysts prepared at 15 Pa. The value of R₃₂₅ for toluene oxidation reached maximum for the NiCo14₈₀ catalyst. It is worth mentioning that the dependence of the T₉₀CO₂ values on catalyst composition roughly corresponded to that revealed for the specific reaction rates.

3. Discussion

The plasma jet sputtering experiments showed that the amount of oxides deposited on the meshes from the hollow cathodes depends essentially on the deposition rate (Figure 1), defined as mg/kWh. Beyond the scope of this work, further research on the sputtering process from hollow cathodes is required. Plasma sputtering of metallic materials, especially alloys, is a very complex process involving the interaction of high-energy noble gas ions with ordered material composed of different atoms. For example, the hollow cathode alloy Ni50Co50 was used to deposit a thin film of oxides on the support with a higher deposition pressure of 160 Pa. The deposition rate was reduced by 20% compared to a deposition pressure of 80 Pa. Thus, it can be assumed that an increase in the deposition pressure leads to a lower deposition rate of thin films (normalized to energy consumed in the discharge). As shown in

Table 7. Conditions of catalyst preparation.

Sample	Hollow Cathode Material	Ni/(Ni+Co) Molar Ratio	Cathode Voltage (V)	Discharge Current (A)	Absorbed Power in Plasma (W)	Deposition Pressure (Pa)	Deposition Time (min)	Average Thin Film Weight (mg)
Co15	Co	0	−312	1.6	499	15	120	1.47
NiCo1415	Ni20Co80	0.2	−312	1.6	499	15	120	1.99
NiCo1115	Ni50Co50	0.5	−306	1.6	490	15	120	1.41
NiCo4115	Ni80Co20	0.8	−314	1.6	502	15	120	2.22
Ni15	Ni	1	−318	1.05	334	15	240	1.10
NiCo1115 *	Ni50Co50	0.5	−336	1.5	504	15	370	3.38
Co80	Co	0	−356	1.6	567	80	120	1.34
NiCo1480	Ni20Co80	0.2	−336	1.6	538	80	120	1.69
NiCo1180	Ni50Co50	0.5	−340	1.6	544	80	120	2.18
NiCo4180	Ni80Co20	0.8	−354	1.6	566	80	120	2.59
Ni80	Ni	1	−345	1.05	362	80	240	1.25
NiCo11160	Ni50Co50	0.5	−320	1.5	480	160	200	2.60

The NiCo11₁₅ * catalyst was prepared with an extended deposition time of 370 min.

and described in detail in the Experimental section, the cathode voltage at the hollow cathode is approximately -300 V for a pressure of 15 Pa, −350 V for a pressure of 80 Pa, and −320 V for a pressure of 160 Pa. Then, the sputtering yields (Y) of argon ions bombarding the surface of the hollow cathode with energies of 300 eV and 350 eV for pure Ni and Co metals are Y^Ni₃₀₀ = 0.74, Y^Co₃₀₀ = 0.7, Y^Ni₃₅₀ = 0.84, and Y^Co₃₅₀ = 0.79, respectively [38]. However, our experiments show that a higher sputtering rate (sputtering yield times of the ion flux at the hollow cathode surface given by the discharge current) in the case of higher deposition pressure does not yield the expected higher deposition rate (see Figure 1). This phenomenon can be explained by the fact that higher deposition pressure probably leads to more diffusive dispersion of the sputtered particles, which affects the number of sputtered particles that reach the mesh.

The physicochemical properties of the active components in sputtered catalysts prepared at two working pressures (15 and 80 Pa) differ in some cases. Increasing the amount of nickel in the deposited Ni-Co oxides led to the formation of various oxide phases (Ni-Co mixed oxides with a spinel structure and NiO-like oxides) in the calcined samples. The formation of Ni-Co mixed oxides was also confirmed by Raman spectroscopy. Their content in the catalysts varied depending on the molar ratio of Ni:Co.

In the H₂-TPR experiments, all catalysts were reduced before reaching 500 °C (Figure 2). The Ni-Co catalysts exhibited lower T_max temperatures than the single-component Ni and Co catalysts. The difference in T_max values was greater for the catalysts containing more nickel. On the other hand, for the catalysts with higher Co content, a new reduction peak around 340 °C appeared, in addition to the main reduction peak at approximately 300 °C. The preparation of the sputtered catalysts with a higher amount of cobalt possibly resulted in the formation of smaller nickel–cobalt oxide particles, which were reduced at lower temperatures. This assumption is supported by the surface area values, which are the highest for the NiCo14₈₀ and NiCo11₈₀ catalysts. Smaller particles in these sputtered catalysts can be also seen in the SEM images (Figure 3), and the low intensity of the bands in the in Raman spectra of these catalysts also indicates the presence of small particles [35].

The XPS measurements proved the presence of Ni²⁺, Co²⁺, and Co³⁺ species on the catalyst surfaces. Surface and bulk Ni and Co concentrations slightly varied with Co content. Surprisingly, higher surface concentrations of Ni were found for the catalysts with higher Co content.

Because the activity of Co oxide in oxidation reactions is higher than that of Ni oxide [39], it can be expected that catalysts with higher Ni content in Ni-Co catalysts will exhibit lower values. This assumption was confirmed by the R₂₀₀ values for ethanol oxidation over the catalysts prepared at 15 Pa and 80 Pa, which decrease with a decreasing amount of Co in the catalysts. In the case of toluene oxidation, similar dependence was found for the R₃₂₅ values obtained over the catalysts prepared at 15 Pa. However, the activity of catalysts prepared at 80 Pa showed a completely different dependence on the molar ratio of Ni:Co, as it reached a maximum for the NiCo14₈₀ catalyst (Figure 6). The explanation for this finding may lie in the different physicochemical properties of the samples. The XRD measurements revealed the presence of spinel-like Ni-Co components in the catalysts with a low Ni/(Ni+Co) molar ratio (≤ 0.5) prepared at a working pressure of 80 Pa (Table 2), and deconvolution of the Co 2p line of XPS spectra demonstrated the presence of Co³⁺ and Co²⁺ species in the catalysts (Figure S1 in, Table 5). Therefore, the high catalytic activity of NiCo14₈₀ in toluene oxidation could be attributed to the higher content of Co³⁺ ions in the catalyst. A roughly proportional dependence of the R₃₂₅ on the Co³⁺/Co²⁺ ratio was found for all sputtered catalysts prepared either at 80 or 15 Pa working pressure (Figure 7). The NiCo14₈₀ catalyst with the highest Co³⁺/Co²⁺ ratio showed the highest activity in the oxidation of toluene. Therefore, it can be concluded that in the oxidation of toluene, the oxidation/reduction properties of the catalysts are important.

In the case of ethanol, such dependence of R₂₀₀ on Co³⁺/Co²⁺ was not revealed, possibly because this reaction proceeds via another mechanism in which acid–base properties of catalysts are important. Therefore, electronegativity may be an important parameter for assessment of ethanol oxidation activity because acidity is related to the electronegativity of metal ions present in the mixed oxides of transition metals. There is a direct relationship between the acidity of binary metal oxides and their electronegativity. The electronegativity (X) of a binary oxide [40,41] can be calculated as an average of the two metal cations using Pauling’s scale of electronegativity of metals:

X_ion = (1 + 2z) X_metal

(1)

where z is the valence of the metal cation.

The electronegativity of Ni and Co mixed oxides was calculated according to the electronegativity of the corresponding metal ions (X_ion) under the assumption of additive behavior of this property and using molar fractions of metal species as weights:

X= x_NiX_Ni2+ + x_CoX_Co3+

(2)

In our calculations, we assumed that the valence of Ni is 2 and the valence of Co is 3 (although Co²⁺ is also present in the samples, as shown by XPS). The calculated values of electronegativity and their relation to catalytic activity in ethanol oxidation (R₂₀₀) can be seen in Figure 8. Based on the dependence presented in Figure 8, it can be concluded that the activity of Ni-Co catalysts in the oxidation of ethanol is higher when their acidity is higher. It is well accepted that oxidation of ethanol starts with its adsorption on the oxide surface [42], and a higher concentration of adsorbed ethanol then leads to a higher oxidation rate.

The deconvolution of the O 1s line of the XPS spectra of the catalysts showed three oxygen forms: the neutral singlet O₂ and the ionic O₂⁻ and O ⁻ species, which are strongly electrophilic reactants that attack the organic molecule in the region of its highest electron density. Electrophilic oxygen O₂⁻ and O⁻ species present at the surface lead to the total oxidation of hydrocarbons [39]. Unfortunately, the differentiation between individual oxygen species determined by XPS could not reveal a clear dependence of the activity on the concentration of oxygen forms. This may be due to problems both in the actual XPS measurement (contamination of the samples by adsorbed carbon and carbon originating from the underlying tape) or in the deconvolution of the O 1s spectra.

With respect to the stability of sputtered catalyst activity, it was monitored for the NiCo11₁₅ catalyst. As shown in Figure 8, in repeated measurements, the specific reaction rate of ethanol oxidation gradually increased from 6.8 to 17.2 mmol_EtOH g⁻¹ h⁻¹. For toluene oxidation, the specific reaction rate increased slightly in the second experiment, but the catalyst activity began to decrease in subsequent experiments. The reason for the differing behavior of the two reactants is probably the reaction temperature, which is sufficiently low for ethanol (up to 300 °C) and much higher for toluene (up to 450 °C). At a low reaction temperature, the properties of the active components seem to be optimized, but at high temperatures, a decrease in the surface area of the active components can proceed.

The efficiency of the studied catalysts in the oxidation of ethanol was compared with that of the EnviCat^® VOC-1544 catalyst [43]. This commercial Cu-Mn/Al₂O₃ catalyst contains 5.4 wt.% Mn and 3.3 wt.% Cu. In ethanol oxidation (1000 ppm at the GHSV of 71 L g_cat ⁻¹ h⁻¹), the T₅₀ temperature was reached at 142 °C. This result may seem to be better than that obtained with the Ni-Co oxide catalysts on meshes (the T₅₀ varied between 240–270 °C). However, the amount of metal oxides in the catalytic bed was much higher for the EnviCat^® VOC-1544 catalyst (1.39 g compared to 0.3–0.5 g for the sputtered catalysts). The conversion of ethanol achieved with the EnviCat^® VOC-1544 catalyst at 200 °C was 87.5 %; therefore, the catalytic activity (R₂₀₀) was 3.5 mmol g_oxides⁻¹ h⁻¹. The specific activity (R₂₀₀) of the Ni-Co oxide catalysts varied from 1 to 16 mmol g_oxides⁻¹ h⁻¹ during ethanol oxidation, with the highest value observed for the Co₈₀ catalyst.

4. Conclusions

Ni-Co oxide catalysts were prepared by plasma jet sputtering on stainless-steel meshes using a hollow cathode at two working pressures (15 ad 80 Pa) and thoroughly characterized and tested in the oxidation of model volatile organic compounds (ethanol and toluene).

Powder XRD revealed the formation of Ni-Co mixed oxides with a spinel structure, together with NiO-like oxides; their contents and cation composition varied with the molar ratio of Ni:Co in the prepared samples. The formation of Ni-Co mixed oxides was confirmed by the Raman spectroscopy. Moreover, the catalysts prepared at a higher working pressure showed slightly lower intensity of the Raman peaks, indicating smaller particles. The TPR profiles confirmed better reducibility of the Ni and Co components in the Ni-Co catalysts compared to that in the single-component Ni and Co oxides.

The sputtered catalysts containing very low concentrations of active components (approximately 0.3–0.5 wt.%) were active in the total oxidation of ethanol and toluene. Ethanol was 90% converted at temperatures between 250 and 300 °C, whereas the temperatures required for the conversions of toluene were 100 °C higher. Regardless of the selected working pressure, the single-component nickel oxide catalysts showed the lowest activity of all catalysts in both ethanol and toluene oxidation. The catalysts containing only Co (i.e., the Co₁₅ and Co₈₀ samples) showed the highest activity in ethanol oxidation. The catalytic activity of the Ni-Co catalysts decreased with increasing nickel content in the catalysts, probably due to strong interactions between the active components. In the oxidation of toluene, the NiCo14₈₀ catalyst showed the highest catalytic activity, in contrast to ethanol oxidation. The catalytic activity of all the investigated catalysts was roughly proportional to the surface Co³⁺/Co²⁺ ratio during toluene oxidation, whereas during ethanol oxidation, the activity was proportional to the acidity of the catalysts.

In ethanol oxidation, the activity of the sputtered catalysts was higher than that of the commercial Cu-Mn oxide catalyst EnviCat^® VOC-1544— up to 16 times higher for the sputtered single-component Co catalyst. The main benefit of sputtered catalysts is considerably lower content of active Ni and Co components and a negligible effect of internal diffusion.

5. Experimental

5.1. Preparation of Catalysts

A plasma-aided deposition method based on sputtering of a hollow cathode, which was partially described in our recent report [23], was used to prepare the catalysts. Circular stainless-steel meshes (71 wt. % Fe, 16 wt. % Cr, 11 wt. % Ni, 2 wt. % Mn; mesh size: 0.40 mm; wire diameter: 0.22 mm) with an outer diameter of 25 mm were used as supports. A standard washing procedure (initial washing in a detergent and thorough rinsing in distilled water, followed by degreasing in acetone and ethanol for 10 minutes in an ultrasonic bath and final rising in distilled water) was applied on the meshes prior to the deposition of the thin oxide layer.

A high-vacuum sputtering system equipped with one hollow cathode plasma source was used to coat the stainless-steel meshes with Ni, Co, and Ni-Co oxide thin films. The hollow cathode plasma jet sputtering gun was horizontally mounted in a cylindrical vacuum chamber evacuated by a turbomolecular pump to a base pressure lower than 10⁻³ Pa. Hollow cathodes differing in composition were used as sources for thin oxide films with various Ni:Co molar ratios. Specifically, we used hollow cathodes with an outer diameter of 12 mm, an inner diameter of 5 mm, and a length of 60 mm made of Ni, Co, and Ni-Co alloys (Ni:Co molar ratios of 80:20, 50:50, and 20:80) with 99.95% purity (Kurt J. Lesker). The hollow cathodes were fixed in a water-cooled copper block connected to a direct-current (DC) power supply. The argon mass flow rate through the hollow cathode was held at 100 mL min^-1 under standard conditions, and oxygen was supplied to the reactor by a separated port with a mass flow rate of 100 mL min^-1 under standard conditions to ensure the same Ar+O₂ oxidation atmosphere for all the experiments. A working pressure in the range of 15 Pa to 160 Pa was set by a gate valve, which throttled the turbomolecular pump. Discharge in the hollow cathode was generated by a TDK-Lambda GEN 600-2.6 DC power supply (TDK Co., Japan) connected to the cathode via a 100 Ω resistor. The DC power supply was regulated in constant current mode (CC), holding constant discharge current at values of 1.05, 1.5, or 1.6 A. The voltage on the cathode was measured using a digital oscilloscope with a voltage probe (Tektronix P5100A).

Then, 10 meshes were placed in a vacuum chamber on an electrically grounded circular rotating substrate holder (rotation speed: 5 rpm), allowing all 10 samples to be deposited under the same process conditions. The direct distance between the hollow cathode outlet and the mesh was 55 mm. This distance ensured a homogeneity of the film thickness better than 10% for all the meshes placed on the holder. The meshes were not intentionally heated, and their temperature was lower than 60 °C during the deposition process. The deposition process by plasma jet sputtering was carried out for 120 min and repeated on both sides of the meshes to ensure that both sides were evenly covered. Details of deposition conditions can be seen in Table 7.

After the deposition, the coated meshes were calcined in air at 500 °C for 4 h. The prepared samples with various Ni:Co molar ratios were labeled as NiCo14, NiCo11, and NiCo41 (numbers denote the intended Ni:Co molar ratio). Single-component Ni and Co oxide catalysts (labeled as Ni and Co, respectively) were also prepared. To reveal the effect of the gas pressure applied during the thin film deposition on the properties of the catalysts, the catalysts were prepared at different pressures. The pressure (in Pa) at which the samples were prepared are indicated by a suffix in the sample name, e.g., NiCo11₁₅.

In addition to the standard deposition time of 120 min, the NiCo11₁₅ catalyst was also prepared with an extended deposition time of 370 min to achieve a higher amount of oxides on the support. This sample was designated as NiCo11₁₅*. For the same purpose, the NiCo11₁₆₀ catalyst was prepared using a higher working pressure of 160 Pa and a deposition time of 200 min. Because our deposition experiments proved that the deposition rate of nickel oxide was significantly lower than that of other investigated cathode compositions, the deposition time for nickel oxide catalysts (samples Ni₁₅ and Ni₈₀) was doubled in comparison to the standard deposition time to obtain a comparable amount of nickel oxide on the meshes.

5.2. Characterization of Catalysts

The Ni and Cu contents in the deposited oxides were determined by energy-dispersive X-ray spectroscopy (EDX, Quantax 200, Bruker, Billerica, MA, USA).

Powder X-ray diffraction (XRD) patterns were recorded using a Bruker AXS D8 diffractometer (Bruker, Berlin, Germany) with Co Kα radiation (λ = 0.179 nm) in the 2θ range of 20–80°, with a step size of 0.02°. Qualitative analysis was performed with the HighScore Plus 4.8 software package (PANanalytical, Malvern, Almelo, The Netherlands, 2018). Lattice parameters were evaluated using cell refinement and profile fitting. The mean coherence lengths (considered as approximate crystallite sizes) were evaluated using the Scherrer equation.

The surface morphology of the oxide particles deposited on the stainless-steel supports was investigated using a Tescan FERA 3 scanning electron microscope (Tescan Indusem, Tescan Brno s.r.o., Brno, Czech Republic).

The surface area of the sputtered catalysts was determined according to the adsorption isotherms of physisorbed krypton at T = 77 K using the BET equation relative to the mass of oxides present on the meshes. An ASAP device (Norcross, Atlanta, GA, Micromeritics, USA) and a specially designed stainless-steel vessel were used for the characterization of bulky catalysts [44], enabling measurements of surface area of large samples (<30 mm). The samples were outgassed in a vacuum at 150 °C prior to data collection.

H₂-TPR was performed using 10 mol. % H₂ in nitrogen with a 20 °C min⁻¹ temperature ramp. Changes in H₂ concentration were detected with a katharometer. The grained CuO (0.160–0.315 mm) was reduced to calculate absolute values of the hydrogen consumed during reduction of the samples. Prior to the measurements, the meshes were cut into small pieces with dimensions of approximately 1 × 1 mm.

Raman spectra were measured using a dispersive Raman spectrometer (DXR Raman microscope, Thermo Scientific, Brno, Czech Republic) equipped with an Olympus confocal microscope and a thermoelectrically cooled CCD detector. A solid-state Nd:YAG laser (wavelength, 532 nm; maximum power, 10 mW) was used as an excitation source. The following parameters were used to collect the spectra: 900 lines/mm grating, 50× magnification objective, laser power of 0.3 mW, acquisition time of 30 s per scan, and 40 repetitions.

Surface elemental analyses were performed using a Kratos ESCA 3400 X-ray photo-808 electron spectrometer (X-ray photoelectron spectroscopy ESCA 3400, Kratos, Manchester, UK). Ni, Co, O, and C elements were detected. A piece of mesh (approximately 8 × 8 mm) was cut from the original sample and placed on carbon tape. All spectra were corrected by shifting the main carbon C 1s peak to 284.8 eV. The Shirley background was subtracted, and elemental compositions of the layers were calculated from corresponding areas.

5.3. Catalytic Experiments

The Ni-Co oxides deposited on stainless-steel meshes were tested in a fixed-bed reactor with an inner diameter of 25 mm. Eight pieces of circular stainless-steel meshes with an outer diameter of 25 mm were used for the measurements. The catalytic tests in total oxidation of ethanol and toluene were performed with 760 ppm of ethanol and 780 ppm of toluene in air at GHSV of 21 L g_cat⁻¹ h⁻¹ using a temperature ramp of 2 °C min⁻¹. An Agilent 8890 gas chromatograph coupled with a mass spectrometer was used for analysis of the composition of reaction products. The T₅₀ temperature at which 50% conversion of ethanol or toluene was achieved was used to compare the performance of the catalysts. Catalytic activity was evaluated as the amount of ethanol or toluene converted per gram of oxides and hour at selected reaction temperatures (200 and 325 °C in oxidation of ethanol and toluene, respectively). The T₉₀(CO₂) temperature at which 90% conversion to CO₂ was achieved was employed to compare the effectiveness of the catalysts in the total oxidation of both model volatile organic compounds.