**Cold Plasma Produced Catalytic Materials**

### Jacek Tyczkowski

Additional information is available at the end of the chapter

http://dx.doi.org/10.5772/61832

#### **Abstract**

The cold plasma techniques are widely applied to create new materials possessing unique properties, which cannot be prepared by any other methods. Among the many in‐ teresting substances produced with the participation of cold plasma, a special place is oc‐ cupied by materials with catalytic properties. The chapter gives a brief review of various cold plasma methods used for the preparation of catalytic materials – from the plasma modification of conventional catalysts via plasma-enhanced classical synthesis of cata‐ lysts to the advanced thin catalytic films fabricated by plasma sputtering processes but primarily by plasma deposition from metalorganic precursors (PECVD). Recently, the catalytic films have attracted considerable attention due to the possibility of depositing them as very thin coatings on virtually all supports without any change in their geome‐ try. Such coatings open the way for new reactor designs, so-called structured reactors, designated for various chemical processes. They can also be used as catalytic deposit on the surface of electrodes for fuel cells and photoelectrodes for water splitting processes. Recent developments in this field and further prospects for thin catalytic films are dis‐ cussed, all the more so because it is one of the main areas of research in our department.

**Keywords:** Catalysts, plasma treatment, PECVD, structured reactors, fuel cells, water splitting

### **1. Introduction**

For a long time, plasma techniques have been used in the creation of new materials possessing unique properties, which cannot be fabricated by other methods. A particularly important technique, giving plenty of possibilities, is the plasma deposition of entirely new, advanced materials in the form of solid coatings having unusual molecular structure and sophisticated nanomorphology. The plasma processes can also be used to modify conventional materials through treating them during or after "classical" synthesis, which generally leads to new

© 2016 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

substances with disparate properties, often more suitable than those of the unmodified materials.

Among the many interesting products fabricated by plasma techniques, materials with catalytic properties occupy a special place. There exist a lot of catalytic reactions, for example, combustion of organic volatile compounds, CO2 methanation, Fischer–Tropsch synthesis, water photo-splitting, fuel cell electrode processes, and many others, which challenge chemists all over the world to seek better catalysts or more effective catalyst preparation methods. It seems that the plasma techniques pave the way for novel solutions in this field.

In general, plasma techniques used in the preparation of catalysts can be categorized into one of two groups depending on the plasma type: thermal (equilibrium) plasma and cold (none‐ quilibrium) plasma [1—3]. The thermal plasma is employed in the preparation process mainly through plasma spraying of catalytically active compounds [4—6] as well as washcoats [7] on various carriers and usually in the form of thicker coatings (>> 1 µm). However, it can also be utilized for the synthesis of ultrafine catalysts whose particle diameters are in the range from a few to a few tens of nanometers and specific area is highly developed [2, 8]. Lately, graphene nanoflakes for catalytic applications have been produced by thermal plasma [9, 10]. Also, this plasma technique is used to recover and regenerate the spent conventional catalysts [11, 12]. Very recently, a more sophisticated method of the catalyst preparation employing thermal plasma has been discovered. For example, very small Ni particles (<100 nm) attached to MgO nano-rods of 10–20 nm in diameter were fabricated by an RF thermal plasma flame, into which a solid precursor consisting of Ni and MgO powders (~5 µm and ~100 nm, respectively) was injected [13]. In turn, Nehe et al. [14] proposed a new promising technique of depositing nanostructured films, namely the solution precursor plasma spray (SPPS) technique. This method was employed to the deposition of CuO−ZnO−Al2O3 layers that were later successfully used to catalyze the methanol reforming for the production of hydrogen (H2) gas.

On the other hand, the cold plasma, in which all processes proceed at much lower temperatures (up to a few hundred degrees centigrade) than those in the thermal plasma (much higher than 1273 K), is a very promising tool for the preparation of catalytic materials [1-3, 15]. Recently, these materials have attracted considerable attention due to the possibility of depositing them as very thin coatings (<< 1 µm) on virtually all supports without any change in their geometry. Such coatings open the way for new reactor designs, so-called structured reactors, for various chemical catalytic processes [16]. They can also be used as catalytic deposits on the surface of fuel cell electrodes without practically any changes in their electrical conductivity and gas permeability. Further advancement is also expected with regard to new sophisticated thin films for photocatalytic splitting of water molecules and efficient production of hydrogen [17]. Apart from the thin catalytic coating deposition, the cold plasma is also utilized in preparations of conventional catalysts, through both plasma enhanced "classical" synthesis and plasma modification of as-prepared "conventional" catalysts [2, 15].

The cold plasma preparation of catalytic films, carried out by plasma polymerization processes from organometallic complexes as precursors, and study of their properties are the main research fields in our laboratory. However, for the sake of completeness, cold plasma involve‐ ments in the preparation of "conventional" catalysts as well as sputtering processes are also discussed in this chapter.

### **2. Cold plasma for "conventional" catalysts**

There are two ways of application of plasma to the conventional production of catalysts: plasma treatment during formation of catalyst and plasma modification of catalysts that have been already conventionally formed for commercial application. The first way can be used to replace the thermal calcination or reduction of a catalyst, to modify its structure, for example, its dispersion and particle sizes, as well as to immobilize catalytic species on supports. Due to chemical treatment with active plasma, catalysts prepared in this way are usually very different from those synthesized only conventionally. Evident changes in the properties of a conventional catalyst can also be achieved by plasma treatment of its final form. Very often, the activity and selectivity of such a catalyst can be significantly improved this way [3, 15 and references therein].

#### **2.1. Plasma-enhanced preparation of "conventional" catalysts**

The results published by Chen et al. in 2004 can serve as the classic example of a comparison between a conventional catalyst and such a catalyst prepared by applying the plasma techni‐ que [18]. The catalyst was composed of Pd and α-Al2O3 and was tested in the selective hydrogenation of acetylene to ethylene. Both procedures of preparation are shown in Figure 1. In the plasma procedure, conventional thermal calcination and reduction processes are replaced with plasma treatment. Apart from the fact that plasma processing is much quicker, cleaner, and easier to control, it was found that the catalyst prepared by the plasma procedure exhibited a significant increase in activity and selectivity, compared with the conventional samples, especially at the lower reaction temperature. Typical results are given in Figure 2, where a drastic difference between the plasma and the conventionally fabricated Pd/α-Al2O3 catalysts is evident.

**Figure 1.** Two ways of preparing the Pd/α-Al2O3 catalyst (on the basis of Ref. [18]).

**Figure 2.** Conversion of C2H2 to C2H4 versus Pd content for the conventional and plasma-enhanced Pd/α-Al2O3 cata‐ lysts (on the basis of Ref. [18]).

The investigations simultaneously performed by Liu et al. [19] confirmed that plasma treat‐ ment during the fabrication of conventional catalysts is critical to the formation of their structure. They have shown that Ar plasma treatment between the preparation steps of drying after impregnation (PdCl2 solution) and the thermal calcination evidently modifies the structure of palladium catalysts supported by a zeolite (Pd/HZSM-5). An enhanced dispersion of PdO and increase in Brönsted and Lewis acidities, which leads to a remarkable improvement in the catalyst stability, have been observed. It has also been found that combustion of methane (to carbon dioxide and water) over the plasma-treated catalyst is close to 100% efficient at 723 K, but it is only approximately 50% efficient at the same temperature over the catalyst not subjected to plasma treatment. In turn, the investigations performed by Zhu et al. [20] on a Mo-Fe/ HZSM-5 catalyst, tested for the non-oxidative aromatization of methane, have shown that Ar plasma treatment benefits the formation of carbonaceous species associated with the active species of MoCX and disfavors the coke formation, which in turn leads to deactivation of the catalyst. The crucial role of plasma treatment in the catalyst structure creation was also shown in 2004 by Legard et al. [21]. They used remote hydrogen microwave plasma (2.45 GHz) for the preparation of gold-based metallic catalysts. Unfortunately, gold tends to sinter easily and the conventional reduction of gold performed in hydrogen at high temperatures very often leads to large particles that are not catalytically active. The particles obtained by using plasma are less than 5 nm in size, and are stable during thermal treatment.

In the following years, we could observe a significant growth of interest in employing cold plasma in the conventional catalyst preparation at various stages of this procedure [22–34]. These methods involve plasma initial decomposition of precursors, plasma replacement of calcination and reduction processes, plasma pretreatment of supports, as well as plasma breaking of thin films to form specific nano-sized catalysts. In general, all these plasma processes, in comparison with the entirely conventional method, lead to unusual catalyst structures, built of smaller particles, better dispersion, stronger interaction between the catalyst and the support, and consequently, to much higher activity, enhanced selectivity, and better stability during a given catalyzed reaction. The anti-carbon deposition and anti-sintering performance are also improved by the plasma treatment. Various types of cold plasma have been tested for this purpose, such as low-pressure glow discharges, atmospheric cold plasma jets, dielectric-barrier discharges, and corona discharges. The plasmas are generated by microwaves (MW), radio frequency (RF, mainly 13.56 MHz), audio frequency (AF, in the range of kHz), alternating current (AC, in the range of Hz), direct current (DC), and even a hot filament, using non-polymerized gases, e.g., Ar, H2, O2, N2, CO2, NH3.

Recently, much attention has been paid to the use of dielectric-barrier discharge (DBD) for plasma decomposition of inorganic precursors, such as carbonate mixtures of CuCO3 + ZnCO3 (for CuO–ZnO catalyst) [35] and NiCO3 + MgCO3 (for Ni/MgO catalyst) [36], as well as Ni(NO3)2 (for Ni catalyst) [37,38] and Co(NO3)2 (for Co3O4 catalyst). This interest stems from the fact that the DBD plasma has turned out to be more reactive than a typical glow discharge. For example, in contrast with the DBD plasma [37], the glow discharge does not lead to the full decomposition of the Ni(NO3)2 precursor. A specific hydrate is formed during the glow discharge treatment and thermal decomposition has to be carried out before further reactions [39]. The DBD plasma technique also enables simultaneous preparation, in the same process, a support and catalytically active component, as it was done by Hua et al. [36], who treated a powdered mixture of NiCO3 and MgCO3 with the DBD hydrogen plasma to prepare a Ni/MgO catalyst. For comparison, the Ni/MgO catalyst was also prepared conventionally (MgO powder was impregnated with an aqueous solution of Ni(NO3)2, then calcined at 973 K in air, and finally reduced at 1123 K in hydrogen). An evident difference in the Ni particle sizes between the plasma-fabricated and the conventional catalysts is clearly visible (Figure 3). It has also been found that the plasma-prepared Ni/MgO catalyst exhibits much higher activity and stability in CO2 reforming of methane.

**Figure 3.** Ni particle size distributions in plasma-treated (A) and conventional (B) Ni/MgO catalysts [36].

It should be emphasized that plasma treatment can generate specific changes in the molecular structure formed during the preparation of a catalyst, which cannot be other‐ wise obtained. Molecular mechanisms of reactions running with the participation of cold plasma, constituting a complex state containing highly active species, such as ions, electrons,

radicals, and excited molecules, are generally entirely different from "classic" chemical processes that occur during the conventional preparation. For example, Chen et al. [40] showed that plasma treatment (N2 plasma, 13.56 MHz) favored significant enrichment of Co3O4 (cobalt spinel structure) on the surface of a cobalt catalyst prepared by the sol–gel method for combustion of methane, in comparison with only thermal calcination. It is suggested that the bombardment of highly active species on the surface during plasma treatment, inter alia, leads to the breakage of –Si–O–Co– bonds formed in the sol–gel process and creates the Co3O4 structure. Such a structure considerably enhances the catalytic performance. Figure 4 presents a comparison between catalysts produced without and with the plasma step. The CH4 conversion is approximately two times higher in the case of plasma treatment.

The plasma treatment can also be used in more sophisticated processes, where the electronic structure of a material, which is crucial for its photocatalytic activity, will be modified accordingly. Recent studies performed on the so-called black TiO2 demonstrate that an enhanced solar absorption and excellent photocatalytic activity have been achieved through the introduction of disorder surface layers on the crystalline TiO2 nanoparticles by plasma hydrogenation. Hydrogen ion bombardment of the particle surface produces large amounts of oxygen vacancies and Ti–H bonds, thereby forming the disorder layer and, consequently, a different electronic structure of the whole nanoparticle. In the surface, tails of localized states occur, which narrows the band gap (to about 2.8 eV from 3.3 eV), but on the other hand, a smaller crystalline core causes widening of its band gap (to about 3.5 eV) [41].

**Figure 4.** Influence of plasma treatment on catalytic performance [40].

In plasma treatment, a particularly important role is played by electrons that are captured by the treated surface and thus create negative potential with reference to the plasma bulk. Under these conditions, positively charged species possessing high energy bombard the surface causing a considerable transformation in its molecular structure practically at room tempera‐ ture. Besides, the trapped electrons, by their mutual repulsion, elongate or distort bonds, which facilitate dispersion processes [42].

As an example of the plasma decomposition of an inorganic precursor, the process of NiO/ Al2O3 preparation from nickel nitrate impregnated to Al2O3 grains can be given [43]:

$$\text{Ni} \left( \text{NO}\_3 \right)\_2 + \text{M}^\* \rightarrow \text{NiO} + 2 \text{ NO} + ^3 / \_2 \text{O}\_2 + \text{M}\_\prime \tag{1}$$

where M\* is an energetic species.

Another example is the plasma formation of a Pd/HZSM-5 catalyst from palladium chloride [42]:

$$\text{PdCl}\_2 + \text{M}^\* \rightarrow \text{Pd}^{2+} + 2\text{ Cl}^\cdot + \text{M} \tag{2}$$

The catalyst can be further reduced by plasma to form pure metal Pd:

$$\text{Pd}^{2+} + 2\text{ e} \rightarrow \text{Pd}^{0}\tag{3}$$

As can be seen, an electron mechanism (involving free electrons) is proposed to explain the plasma reduction. Such a mechanism gives much broader capabilities than typical chemical reduction processes, which usually require high temperatures, resulting in disadvantageous aggregation of metal particles, and are not environmentally friendly. The plasma reduction can be performed at low temperatures in various types of plasma, even in oxygen plasma [44]. However, since the electron mechanism of plasma reduction is governed by electrochemistry and the standard potential of Mn+/M0 pairs determines whether the metal salt can be reduced or not, there are some cases where the plasma reduction process cannot be performed. It has been suggested that the reduction can be observed only when the standard potential is positive. Thus, PdCl2 (Pd2+/Pd0 , E = +0.92 V) can be reduced by plasma treatment (Eq. (3)). On the other hand, Ni(NO3)2 decomposed into metal oxide (Eq. (1)) cannot be reduced this way (Ni2+/Ni0 , E = −0.25 V) [42].

The plasma treatment is also engaged in appropriate adaptation of catalytic supports through developing their surfaces and increasing the catalyst dispersion. For example, the study focused on Al2O3-supported Ni catalysts suggests that the catalysts with Al2O3 subjected to plasma treatment before impregnation are relatively easier to reduce and exhibit higher activities under mild reduction conditions [45].

Often, the plasma treatment of supports is utilized to functionalize their surfaces, which enables efficient attachment of the catalysts. Médard et al. [46] used CO2 plasma to form carboxylic groups on the polyethylene surface, which were able to form pure covalent bonds with metallocene catalysts (Indene2MCl2, where M = Zr, Ti), then tested them successfully in styrene polymerization. Lopez et al. [47,48] applied similar treatment to poly(vinylidene fluoride) membranes. The membranes were treated with Ar and then with NH3 plasma in order to obtain a surface rich in amino groups, which are suitable anchor sites for the immo‐ bilization of tungsten-based catalysts. In particular, tungstate ions (WO2 4−), catalyzing the oxidation of secondary amines to nitrones, as well as decatungstate (W10O324−) and phospho‐ tungstate (W12PO403−) ions, which can both be used as catalysts in the degradation of organic pollutants such as phenol, were investigated. The plasma technique was also used to modify the interface between the catalyst nanoparticles and the support. Gold nanoparticles (for hydrogenation of acetylene) were anchored on a SiO2 support through (3-aminopropyl)trie‐ thoxysilane (APTES) molecules that should subsequently be removed. It turned out that O2 plasma destroyed these molecules giving much higher dispersion of Au nanoparticles (~3 nm) than thermal treatment at 773 K [49].

Among the materials that can be used as catalyst supports in fuel cells, carbon nanotubes (CNTs) and nanofibers (CNFs) have captured increasing attention owing to their high electrical conductivity, large surface-to-volume ratio, and corrosion resistance. However, the nonreactive and hydrophobic nature of the nanocarbons make deposition of catalytically active nanoparticles technically difficult. To overcome this problem, their modification by changing chemical composition of the surface has proven to be efficient. The plasma treatment allows the introduction of appropriate functional groups capable of chemical anchoring metal nanoparticles onto the surface, without affecting the bulk structure and morphology of nanocarbon supports. In this manner, carboxylic and phenolic groups, with control of their ratio, were formed by air or O2 MW plasma on CNFs [50] and CNTs [51] to attach Pd and Ru nanoparticles, respectively. Very recently, oxygen-plasma-functionalized CNTs have been presented as supports for Pt–Ru catalysts applied in direct methanol fuel cells for electro‐ chemical methanol oxidation. It has been shown that O2 plasma treatment leads to the formation of carbonyl (–CO) and carboxylic (–COO) groups on the CNT surface. Pt–Ru nanoparticles dispersed for an optimum plasma treatment time exhibit high catalytic activity toward oxidation of methanol [52]. Iron(II) phthalocyanine, used as an electrocatalyst for the oxygen reduction reaction (ORR) in fuel cells, was also deposited on carbon nanoparticles (Vulcan XC-72) with the aid of plasma pretreatment of the support (N2, Ar, Ar+O2, and NH3 RF plasmas were tested). In some cases, a fivefold increase in the electrocatalytic activity was observed [53].

Considering the issue of carbon nanotubes, we should also mention the use of cold plasma in the preparation of catalysts for the synthesis of CNTs. Gao et al. [54, 55] employed N2 microwave plasma for this purpose. A very thin Fe film was deposited on a Si substrate by the pulse laser technique and then it was broken by plasma treatment. The nano-sized Fe islands having the density of 1.9 × 1015 m–2 and diameters of about 15 nm were obtained. CNTs synthesized on this catalyst were well-aligned, vertically arranged, and had almost the same diameters and density as the nano-islands. Similarly, the nickel catalyst was obtained by converting a Ni thin film into nanoparticles by treatment with microwave H2 or H2/N2 plasmas [56, 57]. Recently, the direct current (DC) plasma as a source of effective ion bombardment, which is crucial for the creation of nano-islands, has been used in the CNT synthesis. It has turned out that gold nanoparticles produced in this way from a 0.5 nm thick Au film act as efficient catalysts promoting the growth of single-walled CNTs [58]. A small addition of C2H2 to H2 during the DC plasma treatment of Fe thin films causes the simultaneous formation of Fe nanoparticles and coating them with a very thin carbon layer, which reduces sintering of the nanoparticles during the growth of CNTs. This results in a very dense CNT forest, reaching 2.2–2.4 × 1016 m–2 [59].

#### **2.2. Plasma modification of as-prepared "conventional" catalysts**

Many catalysts that are produced for practical use or are already used commercially still need further improvement of their efficiency and selectivity. There have been a lot of research works on the subject and it appears that also in this case, good results can be achieved by plasma treatment. For example, Ar plasma treatment (a corona discharge) of the Ni catalyst used in partial oxidation of methane to syngas causes an increase in the catalytic activity and Ni dispersion. Moreover, the plasma treatment improves the catalytic stability by preventing the deposition of carbon on the Ni catalyst [60]. Similar changes in catalytic properties have been observed for Ni and Pt catalysts treated with He plasma in the dielectric-barrier discharge (DBD) under atmospheric pressure [61]. The DBD technique was also used in order to improve the activity of MnOx catalysts in the catalytic oxidation of nitrogen oxide (NO). The effect of plasma treatment is shown in Figure 5, where increased conversion of NO to NO2 by O2 at low temperature (323−523 K) is recorded [62, 63].

**Figure 5.** Oxidation activity of MnOX with and without plasma treatment [63].

An evident improvement in the catalytic properties after the plasma modification of the asprepared catalysts has recently been reported more and more often. Such catalysts as Pd/ TiO2 for the selective hydrogenation of acetylene [64, 65], Fe–Cu on active carbon for the hydrolysis of carbon disulfide [66], CuO nanowires for oxidation of carbon monoxide [67], or CuO/TiO2 employed in the reduction of nitrogen oxides [68] have shown higher efficiencies, only when the plasma treatment process was added after the catalyst preparation. Of course, in any case, it is necessary to choose an appropriate type of plasma and select optimized parameters of the treatment process. An example is given in Figure 6, where one can observe the effect of plasma exposure time (Ar MW plasma) on the conversion and selectivity for the above-mentioned CuO/TiO2 catalyst tested in the reaction [68]:

$$\text{NO} + \text{CO} \rightarrow \forall \text{N}\_2 \text{+ CO}\_2 \tag{4}$$

This result indicates that a remarkable increase in activity and selectivity is achieved depend‐ ing on the plasma treatment time. However, this clear improvement is attributed not only to a greater dispersion of the catalyst but also to changes in its chemical structure. It has been suggested that highly active oxygen species (O\*) are formed on the surface during plasma treatment, which leads to positive changes in the reaction path.

**Figure 6.** Catalytic performances for reaction (4) over CuO/TiO2 catalysts following different microwave plasma treat‐ ment times [68].

The plasma treatment has also turned out to be very useful for the regeneration of spent catalysts. It has been found that the rate of reduction is several times higher for the plasmatreated deactivated catalysts than for the untreated ones. Furthermore, activity of the regen‐ erated catalyst is usually higher than that of the fresh catalyst [1]. Application of the plasma method is especially important in the case of nanocatalysts, whose catalytic activity depends strongly on their size, so only low-temperature regeneration methods are acceptable. The plasma regeneration (O2 DBD plasma) of TiO2-supported gold nanoparticle catalysts (Au/ TiO2) has been successfully used to a great enhancement of catalytic activity for CO oxidation over the completely deactivated Au/TiO2 [69]. Recently, a more sophisticated plasma treatment has been used to regenerate tungsten carbide (WC), which is considered a promising replace‐ ment for precious metal-based electrocatalysts for fuel cells. Mild treatment of WC foil with atomic oxygen generated in a plasma source operating in the atom mode (with an ion trap) allows a controlled removal of graphitic carbon from the WC surfaces without causing oxidation of WC [70].

### **3. Plasma sputtering of catalytic nanoparticles**

Ultrafine particles are a particularly interesting form of catalysts due to their large specific surface and less-perfect crystal lattice with a large number of vacancies, which induces high catalytic activity. One of the most efficient methods of producing such particles of sizes from several to tens of nanometers is the cold plasma sputtering [71]. In this technique, positive ions that are produced in plasma generated in an inert gas, for example Ar, bombard the target surface and cause sputtering of its material. The sputtered material condenses on the substrate that is located outside of the plasma region. If some reactive gases are used (e.g., O2, N2, CH4), the target material takes part in chemical processes during sputtering and finally a new converted material is deposited. This is the so-called reactive plasma sputtering. Simultane‐ ously, two or more different targets can be specified, and the process can also be enhanced by magnetic field (magnetron plasma sputtering). So, as one can see, the plasma sputtering has great technological potential for the production of nanocatalysts. By choosing the appropriate process parameters, the structure, size, and quantity of nanoparticles can be controlled. As an example, the dependence of diameter of Pd nanoparticles on the sputtering time (Ar plasma, 100 MHz) is shown in Figure 7.

**Figure 7.** Pd nanoparticles diameter on the time of plasma sputtering (on the basis of Ref. [72]).

The metal and metal oxide nanoparticles are used primarily as catalytic materials for fuel cell electrodes, and mainly in this respect, the plasma sputtering has been intensively investigated over the past 15 years [17]. It has been shown that Pt, Pd, Rh, PtXRhY, PtXRuY, Pt-RuOX, NiXZrY, PtNiZr, NiZrPtRu, CoOX, NbOX, NbOXNY, PdXAuY, and others can be plasma-sputtered and deposited in the electrocatalytic active form [73–85]. In many cases, they exhibit higher activity than conventionally prepared electrodes, but the possibility to significantly reduce the amount of catalytic material used is the most important. It has been shown that the catalytic electrodes could be prepared with a platinum loading down to 0.005 mgPt/cm2 , which is drastically lower than that for conventional Pt electrodes (0.5–1.0 mgPt/cm2 ), with no detrimental effect on the fuel cell performance. Figure 8 presents the specific power on the current density for a PEM fuel cell with a Pt-sputtered anode, for various loadings of the catalyst. The advantageous effect of reducing the amount of catalyst is clearly visible [86].

**Figure 8.** Platinum utilization efficiency versus current density for plasma-prepared cathodes with different Pt loading (mgPt/cm2 ). For comparison, the results for "classic" Pt cathode with Pt loading of 0.5 mgPt/cm2 are presented [86].

The method of plasma spraying has proven to be very useful in the preparation of 3D-catalysts [17]. Already in the early 2000s, Brault et al. [87, 88] applied this method to introduce Pt clusters into a porous carbon material forming 3D-electrodes for PEM fuel cells. They obtained Pt nanoclusters of 3.5 nm, which penetrated the porous carbon electrode up to 2 µm deep. It should also be emphasized that although the electrode work was comparable to commercially available electrodes, it had the platinum density 4.5 times lower and hence was significantly more effective. Recently, a new pathway in the design of 3D-electrodes has been proposed, namely a combination of the methods of plasma polymerization (PECVD) and plasma sputtering. This dual-plasma process, i.e., the synthesis of catalytic thin films made of Pt nanoclusters (3–7 nm) embedded in a porous hydrocarbon matrix, was carried out by simul‐ taneous plasma-polymerization of ethylene and sputtering of a platinum target. The metal content in the films could be controlled over a wide range of atomic percentages (5– 80%) [89]. Great possibilities of this method have encouraged attempts to produce non-noble metal catalysts for PEM fuel cells. Metals such as copper, cobalt, and iron have been sputtered and embedded into the matrix formed via the PECVD of pyrrole [90–93]. Plasma polymerized pyrrole (pp-pyrrole) exhibits high electrical conductivity and gas permeability as well as good chemical stability, so that together with the sputtered metal clusters forms an excellent nanocomposite, which can act as a 3D-catalyst.

### **4. Plasma-deposited (PECVD) thin films with catalytic properties**

One of the most promising methods of producing new catalytic structures is the plasma deposition of thin films from organic and metalorganic precursors most frequently supplied to the plasma reactor in the gas phase. This method of thin-film deposition, well known as the plasma polymerization or plasma-enhanced chemical vapor deposition (PECVD), has already been used to fabricate a lot of thin-film materials for various practical applications [94]. In the late 1980s, a possibility of using PECVD to produce thin films having potential catalytic properties, such as Pd, Rh, Pt, was already mentioned [95–98]; however, only recently the involvement of this method in the field of catalysis has become a reality [3, 15, 17].

### **4.1. Background**

Metalorganic compounds (their molecules are composed of a metal atom surrounded by organic or organic-like (e.g., CO) ligands) are key precursors for the PECVD of thin films displaying catalytic properties, in which the catalytic activity is related to the presence of metal or metal oxide clusters in amorphous or nanocrystalline forms. However, thin films possessing potential catalytic properties plasma-polymerized (pp-) from typical organic (without metal) precursors, wherein organic functional groups act as active centers, have also attracted attention. It was found quite a long time ago that the films containing organic moieties of specific electronic structure, for instance, the quinone-type groups formed in pp-4-vinylpyri‐ dene [99] or the protonated amine groups created in pp-allylamine [100], exhibit electrocata‐ lytic activity. Recently, this idea has returned – thin films have been deposited onto silica– alumina powders by plasma polymerization of 1,2-diaminocyclohexane to prepare amineimmobilized solid base catalysts. The existence of amine moieties has been confirmed by solidstate 13C NMR, FTIR, and XPS analyses. The silica–alumina powders with these films have shown strong base catalytic activity, e.g., in Knoevenagel condensation reaction between benzaldehyde and ethyl cyanoacetate [101].

As already mentioned, the metalorganic precursors are introduced into plasma reactors mainly in the gaseous form. Sometimes, however, this poses a problem because of the low vapor pressure of such compounds. To get around this difficulty, an innovative method of production of catalytic films by the PECVD has been proposed lately. In this method, compounds are injected into a plasma reactor by means of, for example, an atomizer, in the form of aerosols created from liquid solutions of precursors. Cobalt oxide thin films for catalytic applications were deposited this way using the aerosol from a solution of cobalt carbonyl (Co2(CO)8) and hexane [102, 103]. Likewise, a series of hybrid silica-based catalysts containing various metals (Ti, V, Zr, Sn, Mn, Fe, Co) were produced utilizing nano-sized droplets (15–50 nm) sprayed from a solution of a selected metalorganic precursor, tetraethoxysilane (Si(OC2H5)4), and an organic solvent, and then injected into a DBD plasma reactor [104]. Precursors can also be introduced directly to the plasma chamber as a solid phase mixture with a powdered support. During plasma operation, the precursor is transferred into the gas phase and then involved in the deposition processes [105–107].

The plasma deposition (PECVD) of thin films from metalorganic precursors should also include a fairly sophisticated method called the plasma-assisted atomic layer deposition (ALD). This is true nanotechnology, allowing ultra-thin films of a few nanometers to be deposited in a precisely controlled way via subsequent cycles that generate layer-by-layer growth [108]. Very recently, the ALD has emerged as an interesting tool for the atomically precise design and synthesis of catalytic materials with a controlled distribution of size, composition, and active site [109].

Occasionally, an incipient wet impregnation of the support by a solution of metalorganic precursor, followed by plasma treatment (e.g., with O2 plasma) is applied to preparation of potential catalysts, such as CoOX on zirconia [110], or CrOX on silica [111]. However, this procedure is not a typical PECVD process in the gas phase.

Examples of metalorganic precursors used for the plasma deposition of thin films having potential catalytic activity are summarized in Table 1. Apparently, the PECVD method can be successfully used to produce a wide variety of films, comprising both metal and metal oxide catalytically active phases.



**Table 1.** Examples of metalorganic precursors for PECVD of catalytic films.

Generally, there are two possible ways to start the decomposition of metalorganic molecules in the PECVD process. The ligands of the precursor can be decomposed in the gas phase, and the products of the process then participate in the formation of the film on the substrate surface, or the precursor molecules are first adsorbed on the surface without decomposition. Then, their decomposition occurs during the interaction of plasma with the substrate. Investigations of the preparation process of CrOX deposited from chromium(III) acetyloacetonate (Cr(acac)3) in microwave plasma on a Zr-based support have evidently shown that during the PECVD process adsorption of the precursor molecules on the support probably takes place by cleavage of one ligand. Further plasma operation causes the gradual destruction of ligands and formation of a CrOX structure. This chemical process is schematically presented in Figure 9 [106]. On the other hand, the study on plasma-polymerized ZrO2 films from zirconium(IV) tetra-tert-butoxide in a mixture with oxygen has shown that the gas-phase reactions have a direct impact on the deposition process. Mass spectra of the plasma have revealed a number of species, including CH3 + , C2H4 + , CHO+ , C2H5 + , CH3CO+ , C4H6O+ , Zr+ , ZrO+ , ZrO2H+ , ZrO3H3 + , ZrO4H5 + , ZrO3C3H7 + , which initiate processes of the film formation [135].

**Figure 9.** Schematic representation of possibly adsorption of Cr(acac)3 on Zr-based support and stepwise plasma de‐ struction of the precursor molecule, which leads to CrOX with Cr3+ and Cr6+ species [106].

By carefully controlling the process parameters, the ligands can be completely removed, leaving only "pure" metal or its oxides. Very often, however, some amount of carbon is present in deposits. To eliminate the carbon contamination, treatment with oxygen plasma, heated substrates during the deposition, and heat treatment after the deposition have been practiced [3, 15].

An appropriate choice of the PECVD process conditions also permits control of the film structure. Amorphous films such as TiO2 films of 7–120 nm thickness (deposited from titanium(IV) butoxide) can be obtained and after thermal treatment at 713 K transformed into nanocrystalline anatase films having a grain size of about 15 nm [132]. Similar results con‐ cerning the conversion of TiO2 from amorphous in crystalline phase of anatase structure as a result of thermal treatment were shown by Cho et al. [129] for films deposited from titanium tetraisopropoxide. However, nanocrystalline structure can also be produced directly in the PECVD process, as is the case with TiO2 films deposited on the heated substrate, where the crystalline domains of 40–90 nm have been obtained [130]. The structure of the films can be controlled by the type of support as well. When TiO2 films were produced form Ti(O-i-Pr)2(thd)2, the use of quartz substrates resulted in the formation of films showing predomi‐ nantly the anatase crystallographic structure. On the other hand, only the rutile phase was found in the case of silicon wafers and nearly amorphous phase for metallic Ni [134].

Finally, the molecular structure of deposited films can be controlled using mixtures of metalorganic precursors. For example, interesting films of CoOX doped with Cu were prepared by the PECVD from a gas mixture of CpCo(CO)2 and Cu(acac)2. It was found that pure cobalt oxide films were mainly composed of Co3O4 in the form of nanoclusters, whereas the Cu doped films were much more complex: CoOX (also Co3O4), mixed Co–Cu oxides, and CuOX nano‐ clusters were detected. Preliminary catalytic tests showed that the films of CoOX doped with Cu initiated the catalytic combustion of n-hexane at a lower temperature, compared with the pure cobalt oxide films [124].

Among the biggest advantages of films produced by the PECVD method, beyond their structural diversity that cannot be achieved by any other method, is their very thin form (in the order of nm). This is of particular importance, on the one hand, to save materials, and on the other hand, such thin films are ideal for new constructions of catalytic structured reactors and electrodes for fuel cells. Currently, the most extensively studied plasma-deposited materials with possible applications as catalysts include some noble metal-based films (e.g., Pt, Pd, Ru) and non-noble transition metal-based films, which exist mainly in the form of oxides (e.g., TiO2, CoOX, FeOX).

#### **4.2. Noble metal–based films**

Platinum-based materials (Pt or Pt alloys) are the best catalysts for many reactions, especially for hydrogen and methanol oxidation as well as oxygen reduction, which are at the core of fuel cell technologies. Unfortunately, Pt is a precious and very expensive metal. Besides, the stability of Pt and Pt alloys becomes a serious problem. Hence, extensive research has been underway to overcome these difficulties and new methods to ensure consumption of smaller amounts of platinum and at the same time provide more stable and effective catalysts are being sought.

In addition to plasma sputtering, which is a very promising technique (see: Sec. 3), the plasma polymerization method (PECVD) has also been tested as a tool for the preparation of platinum thin-film catalysts. A platinum metalorganic complex, (Pt(acac)2), was used by Dhar et al. [112] as a precursor for the PECVD carried out in an RF discharge. The plasma-polymerized film was then calcined to drive off organic material, leaving behind a catalyst-loaded substrate. The same procedure was used to prepare a composite consisting of a ZrO2 support and Pt catalyst. The support and catalyst were deposited on a metallic substrate by the PECVD as alternate layers from Zr(acac)4 and Pt(acac)2, respectively. It was found that Pt agglomerates were embedded in the zirconia support [113]. Recently, the mechanism of formation of Pt nanopar‐ ticles on carbon black powder used as a support, during plasma deposition from Pt(acac)2 has been more closely investigated. It has been proposed that some oxygenated surface groups or structural defects are formed at the surface of carbon under the influence of plasma. These surface defects act as anchoring sites for the nucleation of Pt nanoparticles. Thus, by controlling the processes of precursor decomposition and formation of surface defects, through an appropriate choice of plasma parameters, the size and concentration of nanoparticles can be controlled [107].

Another important noble metal, which has been widely used as a catalyst in various syntheses, is palladium. In case of palladium, some attention has been paid to the plasma assisted ALD method. Pd thin films were deposited on substrates at low temperatures (≤ 373 K) by means of a remote inductively coupled hydrogen plasma and frequent use of Pd(hfac)2 as a precursor. This way, Ten Eyck et al. [115] obtained films of the thickness from 1.66 to 3.87 nm after 150 cycles of pulsed ALD, depending on the type of substrate. An additional O2 plasma step in each cycle yielded virtually 100% pure palladium thin films composed of nanometric crystal‐ line grains, the size of which could be controlled by the number of deposition cycles [116]. Investigations performed on Pd films prepared by the typical PECVD process (from Pd(acac)2) showed that the as-deposited films were amorphous and only after thermal treatment at 623 K, Pd nanocrystalline clusters of 5–10 nm size were formed [114].

The group of important noble metals used in catalysis includes ruthenium as well. As in the case of Pd plasma-deposited films, particular attention has been paid to the plasma-assisted ALD method. By using a Ru(EtCp)2 precursor and N2/H2/Ar or NH3/Ar plasmas as reactants for the appropriate steps in the deposition cycles, very thin films of ruthenium (with the growth rate of about 0.17 nm/cycle on SiO2 substrates), containing very low amounts of oxygen and carbon, were produced [117,118]. The same precursor (Ru(EtCp)2) was used in our laboratory to perform typical PECVD process for preparing ruthenium films. XPS analyses indicated that the as-deposited films contained only metallic ruthenium (Ru0 ); however, after calcination at 773 K in air, a fraction of RuO2 (Ru4+) appeared. Spectra of Ru 3d5/2 obtained for the asdeposited and calcined films are shown in Figure 10. The films were tested in the catalytic reactions, such as methanation of CO2 [141] and water splitting [142].

**Figure 10.** XPS spectra of Ru 3d5/2 for as-deposited (A) and calcined (B) films obtained by PECVD from Ru(EtCp)2.

#### **4.3. Non-noble transition metal–based films**

Transition metals, mainly in the form of oxides, are well known as catalysts in many chemical and photochemical reactions. Among these oxides, for a long time TiO2 has belonged to the most intensively studied compounds, mainly because of its unique photocatalytic activity [143]. The plasma deposition of TiO2 thin films is prominent among the methods of producing this material and many results have already been published on this subject; some of them were mentioned above [129, 130, 132, 134], others can be found in the reviews (for example, see [15, 144]). In most reported works, either TiCl4 or Ti alkoxides (mainly titanium tetraisopropoxide) are used as precursors of the PECVD process, resulting in amorphous or crystalline films having the nonstoichiometric (TiOX) or stoichiometric (TiO2) structure. Physicochemical and catalytic properties of all of these films are strongly dependent on their structure, which can be effectively controlled by the deposition conditions [125–128, 131, 133].

Other metal oxides important from the standpoint of catalysis are cobalt oxides (CoOX) and iron oxides (FeOX). Their thin films produced through the PECVD have recently been of particular interest.

#### *4.3.1. CoOX-based films*

The first attempts to produce films from cobalt metalorganic precursors by the PECVD method were undertaken in the early 1980s [145,146], but they have been more closely studied only recently [102, 103, 119–122, 124, 136]. This is related to wide possibilities of application of CoOx films in new structural reactors for catalytic synthesis as well as in electro- and photo-catalyses carried out in fuel cells and water splitting processes. Various precursors have been used (see, Table 1), but one of the most convenient is CpCo(CO)2, because it is liquid at room temperature and highly stable, which facilitates control of the plasma process.

The nanostructure of plasma-deposited CoOX films, which is crucial for catalytic properties, can be controlled by the type of precursor, plasma process conditions, and nature of supports. For example, small particles of CoOx 2–10 nm in diameter were found in the plasma-deposited films on TiO2 supports [119]. When CoOx films were produced on a glass substrate at elevated temperatures (423–673 K), columnar grains of the average diameter between 35 and 60 nm were formed on the film surface [147]. Highly pure and strongly oriented Co3O4 thin films, with features dependent on the used substrate and adopted growth temperature, were obtained by the PECVD on single crystals, such as MgO(100) and MgAl2O4(100) [120]. In turn, the plasma deposition from liquid solutions of Co compounds sprayed into the plasma reactor gave thin films containing nanocrystals 15 or 40 nm in size, depending on the type of precursor and the PECVD process conditions [102]. If such films are post-treated with Ar and O2 plasma, the crystal nano-catkins, which can increase the contact area of the catalyst, are formed containing particles of Co3O4 3–12 nm in size [103].

The research conducted in our laboratory on CoOX films produced by the PECVD from CpCo(CO)2 has confirmed that the structure of these films can be controlled by the conditions of the deposition [3, 122, 123]. As deposited, thin films (25–750 nm) composed of a hydrocarbon matrix and amorphous CoOX were obtained. However, only moderate thermal treatment was enough to transform the amorphous films into films possessing the nanocrystalline structure of cobalt spinel (Co3O4). The formation of such nanocrystals was confirmed not only by electron diffraction analysis but also by Raman spectroscopy measurements, an example of which is shown in Figure 11.

**Figure 11.** Raman spectra for: (A) thin film of hydrocarbon matrix with amorphous CoOX, (B) thin film of nanocrystal‐ line Co3O4 [123].

It has also been found that the size of nanocrystallites is controllable by the thermal treatment. To investigate this effect precisely, an energetic laser beam was used as the heat source. The dependence of the crystallite size on the time of laser treatment is illustrated in Figure 12. As can be seen, the average size of the Co3O4 nanocrystallites increases with the treatment time reaching a constant value. On the other hand, the concentration of nanocrystallites is almost constant throughout the treatment. The maximum size and concentration of the nanocrystal‐ lites can in turn be controlled by the parameters of the plasma process. It has already been shown [123] that with an increase in the precursor flow rate, the maximum size of nanocrys‐ tallites, after the thermal treatment, increases.

**Figure 12.** The size (A) and concentration (B) of Co3O4 nanocrystallites in a film plasma deposited from CpCo(CO)2, in dependence on the laser treatment time.

The above results represent a further step toward the molecular engineering of catalysts, where the desired nanostructure of material and thus its catalytic activity can be designed.

### *4.3.2. FeOX -based films*

The recent interest in the plasma-deposited FeOX thin films has been prompted primarily by the possibility of using these films as photocatalysts in the water splitting process [137, 139, 140, 148, 149]. Various types of FeOX films can be prepared, depending on the precursor used (see, Table 1) and deposition process conditions. Already twenty years ago, Fujii et al. [136] showed that using the same precursor and constant RF plasma power, a wide range of FeOX structures was obtained by controlling the O2 flow rate and substrate temperature (*TS*) during the deposition process. These results are summarized in Figure 13. As can be seen, at low values of the O2 flow rate, amorphous films are formed independent of *TS*. When the flow rate increases, thin films of spinel-type iron oxide (Fe3O4), which are composed of both Fe2+ and Fe3+, are created. Usually, however, it is not a pure fraction, but rather Fe3O4–γ-F2O3 inter‐ mediate phase. At the O2 flow rate higher than 30 cm3 /min, thin films composed of Fe3+ are formed, namely α-Fe2O3 (at *TS* = 600–750 K) and β-Fe2O3 (at *TS* = 450–650 K), although the films of single phase β-Fe2O3 cannot be obtained and this structure occurs only in coexistence with the spinel-type iron oxide. However, by appropriate selection of the PECVD conditions, it is possible to obtain films containing pure β-Fe2O3. Such films with columnar arrays of β-Fe2O3, characterized by a preferred (100) growth direction and highly porous structure, have been successfully plasma deposited from a fluorinated precursor (Fe(hfa)2TMEDA) in a mixture with Ar and O2 [137].

**Figure 13.** Formation diagram of iron oxide films deposited from Fe(acac)3 at a constant RF plasma power (400 W) (on the basis of Ref. [136]).

The structure of deposited iron oxide films can also be modified by post-treatment. In Figure 14, examples of the XPS spectra of Fe 3p, concerning films deposited from Fe(CO)5 in our laboratory, are shown. A clear change in the structure as a result of thermal treatment (723 K, in air) is observed. The as-deposited film contains Fe2+ and Fe3+, which in combination with the results of X-ray diffraction and Raman spectroscopy indicates a spinel-like structure. After thermal treatment, only the completely oxidized form Fe3+ is present on the film surface. It has been found that in these films Fe3+ is associated with γ-Fe2O3 nanoparticles [150], which are already known as an effective catalyst [151]. An opposite process, i.e. the change of the film structure containing only Fe3+ to that containing a mixture of Fe3+ and Fe2+, can be carried out using hydrogen plasma treatment. This way, hematite (α-Fe2O3) films were transformed to Fe3O4:α-Fe2O3, with precise control of the Fe3+ to Fe2+ ratio. This treatment substantially improved optical absorption, enhanced electrical conductivity, and improved transport properties due to a valence dynamic in the film among different iron oxidation states [140].

**Figure 14.** XPS spectra of Fe 3p for as-deposited (A) and calcined (B) films obtained by PECVD from Fe(CO)5.

A wide range of possibilities provided by the PECVD method allows us to design and produce more sophisticated structures at the molecular level. It is also true for iron oxide films, where, for example, Barreca et al. [137] replaced a part of oxygen atoms with fluorine atoms in the film structure. They showed that a noticeable switch from *n*- to *p*-type conductivity occurred upon increasing the fluorine amount in the films. This transformation in the electronic structure resulted in the extended charge carrier lifetime, making F-doped β-Fe2O3 films an efficient water oxidation catalyst. Recently, more complex nanosystems based on iron oxide films have also been prepared. Such nanosystems, for instance, including Fe2O3–CuO [148] and Fe2O3–Co3O4 [149], are synthesized by using a two-step plasma-assisted strategy. First, the iron oxide nanostructured thin film (as a host) is deposited by the PECVD and then CuO or Co3O4 nanoparticles (as guests) are over-deposited on the host matrix by means of plasma sputtering of Cu or Co and annealing in the air. The obtained structures based on the combi‐ nation of *p*-type (CuO and Co3O4) and *n-*type (Fe2O3) semiconducting oxides, which create the *p*-*n* junctions responsible for the improved separation of electron–hole pairs, are highly promising as catalysts for the photochemical splitting of water.

In addition, the advanced structures based on plasma-deposited iron oxides include Fe2O3 thin films that very recently have been produced by plasma-enhanced atomic layer deposition (ALD) from tertiary butyl ferrocene and O2 plasma. The films deposited below 523 K are amorphous but can be converted into hematite (α-Fe2O3) with (104) preferential orientation and average crystallite size of 35 nm by annealing in He [138].

#### **4.4. Potential applications of the plasma-deposited (PECVD) catalytic films**

The three main features of the PECVD method are the source of its attractiveness and broad prospects in the production of new catalytic species, namely – the ability of preparing very thin (<< 1 µm) films, the possibility of precise control over molecular structure and nanostruc‐ ture of the films by selecting appropriate precursors and conditions of the manufacturing process, and finally, the possibility of obtaining 3D structures through the use of copolymer‐ ization of two or more precursors in the PECVD process or combining the PECVD and plasma sputtering processes together.

Current research on the potential possibilities of the application of catalysts produced by the PECVD is focused on the structured reactors, where the deposition of very thin catalytic coatings is the key problem, and on the PEM fuel cells and photoelectrochemical cells for water splitting, which need cheap and efficient electro- and photocatalysts.

#### *4.4.1. Structured reactors*

Development of the new and improvement of the existing heterogeneous catalytic processes are actually tasks of great practical importance in such areas as chemical and petrochemical industries, as well as in environment protection, particularly in catalytic combustion of volatile organic compounds (VOCs) and CO2 conversion into value-added products. Conventionally used reactors for catalytic combustion and CO2 conversion are packed bed (filled with a bed of catalytic grains) and monolithic reactors. The first type can be regarded as a slightly oldfashioned solution showing high flow resistance and low effectiveness of the catalyst grains, which in turn leads to high spending on catalyst material. Monolithic reactors are currently the standard equipment used for catalytic combustion since they provide better performance compared with packed bed reactors. However, monolithic reactors often fail in this process. This is because the so-called fully developed laminar flow occurs throughout most of the monolithic channels. Therefore, mass transfer at the developed laminar flow is rather low and weakly affected by the fluid velocity. Taking into account the high dilution of VOCs in large gas streams, the process is mainly controlled by the mass transfer, which imposes the use of long reactors to secure high conversion of VOCs. This in turn gives rise to an undesired high pressure drop and poor catalyst exploitation. An ideal reactor for combustion of VOCs should thus allow both high mass transfer coefficients and reasonably low flow resistance. The same requirements apply to many other heterogeneous catalytic reactions [16,152].

One of the possible paths for progress in this area is the development and application of new geometrical forms of catalysts. Structural reactors equipped with metallic fillers made of wire gauzes displaying highly enhanced transport properties proved to be an interesting option for all types of catalytic processes. Although such a solution based on noble metal catalysts has been known for a long time, for example, reactors filled with stacked platinum woven wire gauzes used for ammonia oxidation, the extension of this concept to other non-noble metalbased catalysts faces serious problems. The crucial difficulty is the deposition of appropriate thin, uniform, well-adhered, and catalytically effective films on wire supports that form fine microstructural meshes, such as the one shown in Figure 15. The most important is the deposition process should not change the elaborate geometry of the meshes.

**Figure 15.** SEM image of a wire gauze made of kanthal with thin CoOX film, used in our experiments with structured reactors.

The PECVD is one of the most promising methods of producing thin catalytic films on metal gauzes. Research on this subject was started in our laboratory in 2007. Catalytic combustion of *n*-hexane, as a representative of VOCs, and CoOX as well as CoOX–CuOX based thin films produced by the PECVD, as catalysts of the combustion process, have been intensively investigated providing excellent results [122, 153–156]. It has been found that the plasmadeposited cobalt oxide catalyst showing a dispersed spinel (Co3O4) structure (see: Sec. 4.3.1) proves active in the *n*-hexane combustion as compared with commercial Pt catalysts. Even better results have been obtained for Cu-doped CoOX films that have the lowest reaction initiation temperature (493 K). It has also been concluded that the gauze carrier enhances the mass transport in the reactors preventing the diffusional limitation of the reaction rate. Compared with the standard monolithic converter, it allows a reduction in the reactor length by around 50 times with a 20% increase in the pressure drop. These results encouraged undertaking the research on a larger scale with a prospect of possible future industrial applications. Experiments with *n*-hexane combustion on the plasma-deposited Co3O4 catalysts were successfully carried out in a large structured reactor (gas stream up to 10 m3 /h STP) [16].

More recently, the research on CO2 methanation process using thin films of catalysts plasma deposited on wire gauze carriers has been initiated by our team. The CoOX (see: Sec. 4.3.1) and Ru-based (see: Sec. 4.2) films have been tested. Sample results are shown in Figure 16. Although these are only preliminary results, they confirm the catalytic activity of the films in the process of CO2 methanation and encourage further research in this area.

**Figure 16.** Conversion of CO2 to CH4 as a function of the process temperature for two different plasma-deposited cata‐ lytic films [141].

#### *4.4.2. Fuel cells*

Current studies on thin films produced by the PECVD, which can be used as catalytic systems in proton exchange membrane fuel cells (PEMFC), are primarily focused on finding new solutions that could compete with platinum electrodes – would be cheaper and exhibit at least comparable catalytic properties. Taking into account the promising electrocatalytic activity in the oxygen reduction reaction (ORR) demonstrated by nanoparticles of cobalt oxides [157], an attempt to produce such a material for the PEMFC electrodes by the PECVD method has been undertaken in our laboratory. The CoOX-based films obtained from the CpCo(CO)2 precursor (see: Sec. 4.3.1) were deposited on carbon paper substrate, and they were tested as the cathode in a hydrogen PEMFC (the anode was made of platinum). The preliminary results admittedly indicated the electrocatalytic activity of the films in the oxygen reduction reaction; however, it was far from the activity of Pt cathodes [17, 123]. Optimization of the conditions of plasma deposition and an enhanced procedure for the preparation of membrane electrode assemblies resulted in a significant improvement of the CoOX electrocatalytic activity. Figure 17 presents current–voltage characteristics of the PEMFC with the cathode made of CoOX deposited at various times (with various amounts of CoOX) and Pt anode. The characteristic of the cell with both Pt electrodes is also shown. As can be seen, the cell with the Pt anode and CoOX cathode is characterized by the open circuit voltage of 635 mV, which is already close to 963 mV obtained for the "pure" Pt cell [158]. Further progress in this area is expected, for example, by using doped CoOX films and increasing the dispersion of Co3O4 nanoclusters.

**Figure 17.** Current–voltage characteristics of PEMFC with the cathode of varying amounts of CoOX deposit. The char‐ acteristic for the reference cell with both Pt electrodes is shown for comparison [158].

Very recently, we have obtained interesting results concerning the PEMFC, in which plasmadeposited FeOX-based films were used as the cathode. The films have been deposited from Fe (CO)5 (see: Sec. 4.3.2). An example of the preliminary results recorded for the fuel cell with such a cathode (and Pt anode) is shown in Figure 18. Although the results are far from expected, they point the way to further research in this area [150].

**Figure 18.** Power and potential versus current density for PEMFC with FeOX cathode [150].

#### *4.4.3. Water splitting systems*

The most desirable method of hydrogen production, which represents a sustainable fuel of the future, is photoelectrochemical (PEC) splitting of water by visible light. A very important issue in the PEC hydrogen generation is the development of a high-performance photoelectrode that exhibits high efficiency in the conversion of solar energy into chemical energy, resistance to corrosion in aqueous environments, and low processing costs. However, after four decades of intensive research, since the first report on water photo-splitting [159], no material has been found to simultaneously satisfy all the criteria required for widespread PEC application. No wonder that a quest for new materials for photoelectrodes is still ongoing. The PECVD technique is also involved in this search [17].

Particular attention has recently been paid to FeOX-based films (see: Sec. 4.3.2). Among iron oxides, hematite (α-Fe2O3) is considered as an attractive photoanode material, largely due to its abundance, chemical stability in aqueous environments, and light absorption in the visible range of the solar spectrum (Eg = 2.1–2.2 eV). However, the reported water-splitting efficiency of α-Fe2O3 is much lower (< 3.0 %) than the maximum theoretical value (~13 %), mainly due to the low absorption coefficient and slow reaction kinetics, resulting in a high carrier recom‐ bination rate and short hole diffusion length. Singh et al. [139] achieved the photocurrent density of about 1.1 mA/cm2 at 0.9 V/SCE (with onset potential 0.6 V/SCE) for α-Fe2O3 films applied as a photoanode in 1M NaOH electrolyte (deposited from Fe(CO)5). Naturally, this value depends on the used light intensity and development of the substrate surface, thus it is difficult to compare this particular result with those obtained in other laboratories. Neverthe‐ less, it can be concluded that the films deposited in our laboratory from the same precursor under conditions that allow formation of γ-Fe2O3 nanoparticles reveal similar, if not better, photoactivity. For example, the onset potential is about 0.45 V/SCE.

A considerable improvement was observed in the photoactivity of α-Fe2O3 films treated by hydrogen plasma, which led to the formation of a complex structure of Fe3O4:α-Fe2O3 type. In this case, enhanced photocurrent densities (3.5 mA/cm2 at 1.8 V/RHE) and reduced photocur‐ rent onset potentials (from 1.68V/RHEfor non-treated films to 1.28 V/RHE) were measured in 1M NaOH electrolyte [140].

In search for further opportunities to improve photocatalytic properties of FeOX thin films, tuning of the system nano-organization and controlled tailoring of the chemical composition have been proposed. In particular, oriented columnar nanostructures offer the possibility of absorbing a significant fraction of light while providing short carrier transport distances to the electrolyte, thus minimizing recombination losses. Recently, columnar arrays of β-Fe2O3 showing a preferred (100) growth direction and highly porous structure have been successfully obtained in Ar/O2 plasma by Barreca et al. [137]. The decoration of such columnar arrays with other transition metal oxides, such as cobalt oxide [149] and copper oxide [148], in the form of nanoparticles, might profitably suppress charge carrier recombination and improve Fe2O3 catalytic activity in water splitting. For example, the Fe2O3–Co3O4 heterostructure nanosystem reveals a photocurrent increase of 40% with respect to bare Fe2O3, which clearly proves that nanosystems produced with the involvement of PECVD are very promising [149].

Some interesting results have been obtained very recently in our laboratory for Ru0 /RuO2 films produced by the PECVD (see: Sec. 4.2). In Figure 19, the photocurrent density (with subtracted dark current) measured under illumination from a 150 W Xe arc lamp equipped with an AM 1.5 filter, as a function of the applied voltage, is presented. As shown in the figure, the achieved value of the photocurrent is 9.1 mA/cm2 at 0.6 V/SCE, which, compared with the results obtained in the same experimental system for Fe2O3 films (0.65 mA/cm2 at 0.6 V/SCE) and Co3O4 films (0.48 mA/cm2 at 0.6 V/SCE), is a very promising finding [142].

**Figure 19.** Photocurrent–voltage characteristic for Ru0 /RuO2 photoanode prepared by PECVD (measurements in 1M NaOH electrolyte) [142].

### **5. Conclusions and outlook**

I hope that after reading this chapter, the reader will no longer have any doubt that the enormous potential is hidden in the cold plasma technology. This technology gives us almost unlimited possibilities of modifying various materials and producing entirely new structures. This is particularly true with respect to the field of catalysis.

On one hand, the high-temperature processes such as decomposition of precursors, pyrolysis, calcination, reduction, and regeneration can be replaced by low-temperature plasma treat‐ ment, which minimizes problems caused by high temperature, such as aggregation, crystallite size growth, and sublimation, often leading to the creation of a completely different structure of the resulting catalyst. Generally, the conventional catalysts prepared with the involvement of cold plasma methods exhibit much higher activity, enhanced selectivity, and better stability. On the other hand, the cold plasma techniques, such as plasma sputtering and especially plasma polymerization (PECVD), have paved the way for the design at the molecular level and production of new advanced catalysts. Very thin films of precisely controlled molecular structure, 3D nanostructured systems, nanocomposites, all these put the cold plasma at the forefront of the twenty-first century methods of catalyst synthesis.

It is hard to imagine further progress in the construction of structured reactors without very thin catalytic films easily and cheaply deposited on precisely designed, sophisticated fillers designated for these reactors. The current development of hydrogen technologies also pins great hopes on new catalyst systems produced using the cold plasma. Although we are only at the beginning of the road, the results already obtained in the field of plasma-prepared catalytic electrodes for fuel cells and photocatalytic electrodes for PEC cells are very promising.

The ability to design new plasma nanomaterials possessing the desired catalytic properties and their preparation through controlled processes carried out in the cold plasma still pose a serious scientific challenge. As of now, the relation between cold plasma processes and the resulting structure and catalytic properties of deposited films is not entirely clear either. Further research in this field will undoubtedly lead to significant progress in science and provide attractive catalytic systems for technology.

### **Acknowledgements**

I would like to thank the members of my team and my PhD students who are engaged in research on plasma-deposited catalytic films: Prof. P. Kazimierski, Dr. J. Balcerzak, Dr. J. Sielski, R. Kapica, Ł. Jóźwiak, M. Marchwicka, and W. Redzynia, for their help and assistance. I thank Prof. A. Kołodziej from the Institute of Chemical Engineering of Polish Academy of Science in Gliwice and Prof. J. Łojewska from Jagiellonian University in Kraków for their excellent scientific collaboration. I am also grateful to Dr. A. Kubiczek for his assistance in the preparation of this chapter.

This work was supported by the Polish National Science Center, on the basis of decision DEC-2012/07/B/ST8/03670. The financial support is gratefully acknowledged.

### **Author details**

Jacek Tyczkowski\*

Address all correspondence to: jacek.tyczkowski@p.lodz.pl

Department of Molecular Engineering, Faculty of Process and Environmental Engineering, Lodz University of Technology, Lodz, Poland

### **References**

