Properties, Industrial Applications and Future Perspectives of Catalytic Materials Based on Nickel and Alumina: A Critical Review

Abstract

The bulk and surface properties of materials based on nickel and aluminum oxides and hydroxides, as such or after reduction processes, are reviewed and discussed critically. The actual and potential industrial applications of these materials, both in reducing conditions and in oxidizing conditions, are summarized. Mechanisms for reactant molecule activation are also discussed.

1. Introduction

Metals represent a main family of catalytic materials, with particular relevance in the fields of hydrogenation and dehydrogenation reactions, including steam reforming, as well as total and partial oxidation [1]. As is well known, platinum group metals (PGMs), mainly platinum itself and palladium, have extraordinarily high catalytic and electrocatalytic activities, much more than typical transition metals such as nickel, cobalt, iron and copper. For this reason, they are applied in many chemical and electrochemical technologies, some of which are strictly needed for the next energetic transition to obtain decarbonization or defossilization for our civilization [2]. However, PGMs are relatively rare elements, and their costs are extremely high. For these reasons, they are becoming critical metals [3] and may be unsustainable for use in some technologies.

Among non-PGM transition metals, nickel is frequently the most active one. It is by far cheaper than PGMs and may represent, for this reason, the best or at least second-best choice after PGMs for many catalytic [4] and electrocatalytic technologies [5].

As is well known, while bulk metal powders can be used sometimes in catalysis, metals are supported by different carriers in most cases. This allows one to reduce the amount of metal used, but it also frequently induces new and different properties to the metal itself. Among the carriers largely used in heterogeneous catalysis, aluminas are relatively cheap, strong and reactive materials that allow for the strong dispersion of supported species and the production of metal nanoparticles whose catalytic activities are frequently significantly superior to those of micrometric or millimetric unsupported powders and whose thermal stability is also significantly superior to that of unsupported nanoparticles. For these reasons, alumina-supported nickel catalysts have, for many years, represented versatile and very active heterogeneous catalytic systems [6]. On the other hand, materials based on nickel and alumina may have very different behaviors depending on the type of alumina used and the relative amount of nickel and alumina, as well as the oxidation state, pretreatments, dispersion, etc. In this critical review, we will summarize the literature data, in part arising from our own experience, concerning materials based on nickel and alumina, both at the laboratory level and concerning the industrial applications of these materials. We will take into consideration catalytic systems where Ni species and alumina or metal aluminates play a major role as the active phase and as the “support”. In

Table 1. Characteristics of reactions where Ni/Al₂O₃ catalysts might be applied.

Reaction Name	Stoichiometry	ΔH°298 (Gas)	Typical Reaction Conditions
		(kJ/molR)	Phase	T (°C)	P (Bar)
Acetylene partial hydrogenation in C2 cuts	HC≡CH + H2 → H2C=CH2	−172	Gas Liq	100–250 25–100	5 20–35
C3–C5 steam cracking cut purification	Diene and acetylenic hydrogenation	~−240	Liq	10–60	<100
Pygas hydrotreatment 1 step	Diene and acetylenic hydrogenation	~−150	Liq	70–200	20–50
Phenylacetylene partial hydrogenation	C6H5-C≡CH + H2 → C6H5-HC=CH2	~−120	Liq	15–50	<50
Linear diene hydrogenation for LAB synthesis	R-CH=CH-CH=CH2 → R-CH2CH2CH2CH3	−240	Liq	170–230	10–20
Benzene saturation	C6H6 + 3 H2 ⇆ C6H12	−208	Gas Liq	500 200–350	30–50
CO methanation	CO + 3 H2 → CH4 + H2O	−206	Gas	250–350	20–50
CO2 methanation	CO2 + 4 H2 → CH4 + 2 H2O	−165	Gas	300–400	20–50
Reverse water–gas shift	CO2 + H2 → CO + H2O	+41	Gas	>850	1–30
Acetonitrile hydrogenation	CH3C ≡ N + H2 → CH3CH2NH2	−121	Gas	100–200	1–20
n-octanal hydrogenation	C7H15-CH=O + H2 → C7H15CH2OH	−57	Gas	100–200	20–50
Methane steam reforming	CH4 + H2O → CO + 3H2	+206	Gas	500–800	1–40
Methane dry reforming	CH4 + CO2 → 2CO + 2H2	+247	Gas	500–700	1–20
Methane partial oxidation	CH4 + ½ O2 → CO + 2 H2	−36	Gas	500–1000	1–40
Methane combustion to CO	CH4 + O2 → CO + 2 H2O	−519	Gas	>200	1–40
Methane decomposition	CH4 → C + 2 H2	+76	Gas	>400	1–10
Ammonia decomposition	NH3 → ½ N2+ ³/₂ H2	+46	Gas	>400	1–10
Cyclohexane dehydrogenation	C6H12 ⇆ C6H6 + 3 H2	+208	Gas	400	1–3
Ammonia oxidation to NO	NH3 + 5/4 O2 → NO + 3/2 H2O	−226	Gas	>600	1–20
Selective oxidation of ammonia to N2	NH3 + ¾ O2 → ½ N2 + 3/2 H2O	−317	Gas	>600	1–20
CO oxidation	CO+ ½ O2 → CO2	−283	Gas	>0	1–20
Hydrogen combustion	H2 + ½ O2 → H2O	−242	Gas	>100	1–20
Ethane oxidative dehydrogenation	H3C-CH3 + ½ O2 → H2=CH2 + H2O	−105	Gas	500–700	1–10
Ozone decomposition	O3 → 3/2 O2	−143	Gas	−50–0	<1

, the ΔH°₂₉₈ (gas) values (calculated per mol of reactant molecule from the enthalpy of formation data [7]) and the practical conditions of the main reactions considered here are summarized.

2. Solid-State Chemistry of Nickel and Aluminum-Based Catalytic Materials

2.1. Oxide and Hydroxide Phases

2.1.1. Ni Oxides and Hydroxides

Precipitation procedures from Ni salts may produce hydroxides [8], which are relevant materials in electrochemical technologies in particular. Two phases exist of the Ni(OH)₂ compound: β-Ni(OH)₂, the mineral theophrastite, is isostructural with brucite, Mg(OH)₂, with Ni²⁺ in octahedral coordination; the second phase is actually a hydrated form of β-Ni(OH)₂, α-Ni(OH)₂·xH₂O, with 0.41 ≤ x ≤ 0.7, consisting of brucite-type layers intercalated by water molecules. By oxidation, two polymorphs of the oxy-hydroxide of trivalent Ni form, which have a main role in electrochemistry as cathode materials of Ni–Cd and Ni–MH (nickel–metal hydride) batteries. They are denoted as β- and γ-NiO(OH). However, their structure is still incompletely known, taking also into account that they could be impure compounds [9].

A single nickel oxide phase exists in nature and is easily produced, NiO, whose mineralogical name is bunsenite. NiO crystallizes in the rock salt or periclase (MgO) structure with all octahedral Ni²⁺ ions and melts at 1900 °C [10]. It has a green-colored, p-type, wide bandgap (3.6–4.0 eV) semiconductor and shows antiferromagnetic behavior underneath the Neel temperature of 523 K. The antiferromagnetic behavior of NiO becomes superparamagnetic when the particle size is small, i.e., around 100 nm [10].

Higher valence binary nickel oxides also exist, although their real structure is not fully clear. In particular, “black” nickel oxide, denoted as Ni₂O₃ [11] or NiO_1+x, can only be formed in a water-free environment at temperatures below 250 °C [12]. Both these phases would formally contain the trivalent Ni ion Ni³⁺ and would exist only in very strong oxidizing conditions. On the other hand, a number of trivalent nickel mixed oxides exist such as, e.g., the LaNiO₃ perovskite [13] and NaNiO₂, which both contain Ni³⁺ in octahedral coordination [14]. Among the high oxidation Ni oxide compounds, LiNiO₂ and Li_xNiO₂ oxides have been the object of investigation because of their role as electrodes in Li-ion batteries. Formally, in the parent layered LiNiO₂ compound, which is nearly isostructural with LiCoO₂ and LiMnO₂, nickel is trivalent, becoming partially tetravalent by the removal of a part of the Li ions, which stay mostly in the interlayer region between the oxide layers, although cation mixing can also occur [15,16]. Indeed, although high-valence nickel compounds (Ni³⁺ and Ni⁴⁺ based) are mostly quite instable structures, they can exist in highly oxidizing environments and, in particular, in structures.

2.1.2. Aluminum Oxides and Hydroxides

The chemistry of aluminum oxides and hydroxides is a very complex one and has been the object of several reviews [17,18]. Techniques for the preparation of alumina nanoparticles, as well as their characterization, have also been the object of a more recent review [19].

In conditions where the solubility is exceeded in Al³⁺ water solutions, usually “gelatinous” XRD amorphous precipitates form first. Depending primarily on temperature and pH, as well as on the aging time, the nature of anions present and the copresence of organic components [20,21], different crystalline hydroxides or oxyhydroxides form [22]. At a low temperature and in excess water, the following hydroxides are preferentially formed: bayerite α-Al(OH)₃ if 5.8 < pH < 9 or gibbsite γ-Al(OH)₃ for pH < 5.8 and for pH > 9. At temperatures higher than about 80 °C, the oxyhydroxides become thermodynamically more stable than the trihydroxides and thus tend to form. Boehmite γ-AlOOH, or the low crystallinity form pseudoboehmite, is most easily formed at atmospheric pressure while the production of diaspore α-AlOOH usually needs higher pressures. The additional Al(OH)₃ polymorphs norstrandite and doyleite, as well as the oxyhydroxide akdalaite, or tohdite, whose formula is Al₅O₇(OH), can be prepared only in particular conditions.

The thermal decomposition of hydroxides and oxyhydroxides gives rise to the many different alumina phases. The thermodynamically stable phase α-Al₂O₃ (corundum) can be directly synthesized by the decomposition of diaspore α-AlOOH at about 500 °C as a medium-surface area powder (65 m²/g [23]) or by the treatment of any Al oxide or hydroxide above 1000 °C. In the latter case, it is usually prepared as a stable low-surface-area ceramic abrasive and refractory material (T_melt 2040 °C). In catalysis, low-surface-area α-Al₂O₃ powders are used as inert and thermally stable supports of catalysts [24]. Higher-surface-area corundum powders can also be prepared by a resistive hotspot-induced phase transformation of γ-Al₂O₃ [25].

The most commonly used material in catalysis is cubic γ-Al₂O₃, which is usually prepared by the calcination of gelatinous pseudoboehmite or by crystalline boehmite at around 450 °C. When treated at a slightly higher temperature, the crystal structure can become slightly distorted and is denoted as δ-Al₂O₃. Much less used is the similar polymorph denoted as η-Al₂O₃, which is usually prepared by the decomposition of bayerite at 250–300 °C. These three slightly different materials, both constituted by cation defective spinel structures [26], can have a surface area of 150–300 m²/g, are stable until near 500 °C, and represent the most common supports for metal and sulfide catalysts working at temperatures < 500 °C. They are also active acid or acid–base catalysts as well as very active adsorbents at both vapor/solid and liquid/solid interfaces. The three cation defective spinel structures, when treated at 700–900 °C, give rise to θ-Al₂O₃, which has a typical surface area of 70–110 m²/g and is stable until ~900 °C. It represents the most common support for metal catalysts for moderately high-temperature reactions. Its monoclinic structure, the same of β-Ga₂O₃, is also derived from the structure of spinel. A further heat treatment of θ-Al₂O₃ at about 1000 °C gives rise to the progressive formation of low-surface-area corundum powders or mixed θ,α-Al₂O₃ composite powders.

It can be taken into consideration that many other carrier materials are currently available for supported catalyst manufacturing, such as other metal oxides (silica, titania, zirconia, ceria, magnesia, etc.) and carbon materials. However, for vapor-phase reactions in particular, alumina and silica are usually the preferred catalyst supports by catalyst producers because of their moderate costs, their well-known chemistry and their well-proven stabilities. In particular, aluminas are characterized by their dispersion ability, which allows them to tune supported metal particle sizes with loading and support surface area optimization.

2.1.3. Mixed Ni-Al Hydroxide and Oxide Phases

The basification of Ni²⁺ and Al³⁺-containing water solutions produces the coprecipitation of mixed hydroxides in a wide range of Ni:Al ratios. In the absence of carbonate ions, Al can enter the structures of both β- and α-Ni(OH)₂ [27,28]. With Ni:Al ratio 3 and in the presence of carbonate ions, a crystalline precipitate with the formula Ni₆Al₂CO₃(OH)₁₆·4H₂O, analogous to the mineral takovite, can be obtained. This is the Ni analogue of hydrotalcite, Mg₆Al₂CO₃(OH)₁₆·4H₂O, the best known among layered-double-hydroxide anionic clays [29,30]. With lower Ni:Al ratios, the precipitates look essentially amorphous, maybe with the presence of takovite too. This methodology for the preparation of Ni/Al₂O₃ catalysts, described in the scientific literature by J. Ross [31], allows one to manufacture, by further calcination, a homogeneous series of catalytic materials that are characterized in a large compositional range of spinel/rock-salt NiO/NiAl₂O₄/Al₂O₃ solid solutions [32,33].

By calcining the precipitate with Ni:Al ratio = 0.5, the spinel-phase NiAl₂O₄ can be prepared [33,34,35], which is the only thermodynamically stable intermediate phase in the NiO-Al₂O₃ phase diagram [36]. In contrast with what happens with spinel MgAl₂O₄, the Jahn Teller effect, due to the electronic structure of Ni²⁺, pushes it in octahedral coordination. Thus, Ni aluminate cation distribution is that of an “inverted” spinel, with nearly half of the Al³⁺ ions in tetrahedral coordination and both the second half of Al³⁺ ions and the large predominance of Ni²⁺ ions in octahedral coordination [37].

The partial solubility of alumina into a Ni aluminate spinel is known to produce Al-rich non-stoichiometric spinels that are thermodynamically stable at a high temperature [36]. These spinels are essentially solid solutions of the stoichiometric inverse spinel NiAl₂O₄ and the defective spinel γ-Al₂O₃, which have partial solubility. In metastable phases, a continuous series of Ni-deficient spinels can be prepared [33,38], including samples with composition NiAl₄O₇ [39] and the NiAl₁₀O₁₆ phase [40]. Interestingly, the presence of even small amounts (0.1 mol%) of NiO in spinel alumina hinders the phase transformation to α-Al₂O₃ [41], thus stabilizing spinel-type aluminas.

On the other hand, Al-deficient or Ni-rich spinels can also be produced, such as Ni₂Al₂O₅ [39], thanks to the reciprocal solubility of the two phases NiAl₂O₄ and NiO, which are both based on the cubic close packing array of oxide ions, forming cation excess non-stoichiometric spinel phases. The metastability of these Ni-rich spinels is confirmed by the segregation of NiO from NiAl₂O₄ upon high-temperature treatment [39].

Interestingly, some solubility of alumina into NiO exists; this is very limited in terms of thermodynamics and occurs only at a high temperature [36], but it is more abundant in metastable phases, producing, e.g., a bunsenite-type (rock-salt structure) crystalline material containing significant amounts of Al ions, e.g., with composition NiAl_0.35O_1.53 [33].

Similar Ni-Al mixed oxide materials can also be prepared with other techniques, such as liquid-feed flame spray pyrolysis [42] and the sol–gel technique followed by calcination [43,44,45]. It can also be taken into consideration that other coprecipitation techniques can produce less-homogeneous materials, producing rock salt and spinel phases with a Ni:Al ratio less than 2 [46].

The solid state reaction of NiO with Al₂O₃ has also been investigated to produce bulk NiAl₂O₄. It has been found that the spinel phase forms only at high temperatures, i.e., above 800 °C [47] or 1000 °C [48], when alumina spinel phases are no longer stable.

2.1.4. Deposition of NiO onto Transitional Aluminas

NiO-Al₂O₃ catalysts, as catalytic materials or as precursors of metallic Ni/Al₂O₃ catalysts, are frequently prepared by impregnation techniques (dry impregnation, wet impregnation, deposition precipitation, etc.). Common dry impregnation techniques allow for the large dispersion of Ni²⁺ ions (and the balancing O²⁻ anions) over the alumina surface until the amount corresponds to a full monolayer. For typical γ-Al₂O₃ with a surface area of about 200 m²/g, this limit corresponds roughly to 15–20 wt% NiO on alumina for well-dispersed samples. The techniques agree, showing no NiO formed until this loading is obtained (see Figure 1 for X-Ray Diffraction (XRD)). These highly dispersed species are essentially constituted by isolated Ni²⁺ complexes strongly interacting with the support. They cover the strongest acid–base sites of alumina located in the edges and defects and, in a second step, the less-active sites located on the main crystal faces [49,50]. This agrees with the reduction in the number of both of the strong acid sites [51] and basic sites [52] of γ-Al₂O₃ by the deposition of the Ni oxide species. When the monolayer coverage is full, the formation of a NiAl₂O₄-like surface layer is observed [53]. The XRD pattern of the samples undergoes slight changes from that of γ-Al₂O₃ (Figure 1) and becomes very similar to that of bulk NiAl₂O₄, although the positions of the peaks, the 400 and the 440 ones in particular (near 2θ = 46° and 67°, respectively), are shifted at lower 2θ values with repect to pure γ-Al₂O₃ but are still at significantly higher values than those observed in the pattern of NiAl₂O₄. Correspondigly, the calculated unit cell parameter is < 8.0 Å, which is only slightly higher than that of γ-Al₂O₃ (7.887 Å) but still definitely smaller than that of NiAl₂O₄ (a = 8.052 Å) [54]. On the other hand, spectroscopies, scheletal infrared in particular (Figure 2) and, even more, visible and near infrared spectroscopy, allow one to distinguish such a surface layer from bulk NiAl₂O₄ and NiO, as shown in Figure 3. The surface NiAl₂O₄-like layer protects the alumina from surface area loss and sintering, and it inhibits bulk phase transformation [55]. Treatment at 600 °C and above results in a fast drop in the surface area with alumina phase transformation occurring [56].

When the “monolayer” loading limit is overcome, further loading produces NiO particles to form, which is well evident by FESEM, XRD, FTIR, UV–visible spectroscopy and XPS techniques [57] (see Figure 1 for XRD). These NiO nanoparticles, although still interacting with the alumina support, are by far more easily reducible to Ni, as is evident by the Temperature Programmed Reduction (TPR) resembling the behavior of the unsupported Ni nanoparticles. As stated, while a NiAl₂O₄ surface layer forms, the formation of bulk nickel aluminate does not occur until at least 600 °C. The reaction between NiO and γ-Al₂O₃ is still slow and incomplete, even at 1000 °C [48], when γ-Al₂O₃ is already converted, usually into θ,α-Al₂O₃ composite powders.

Impregnated materials are frequently prepared to be later reduced in hydrogen for the final manufacturing of reduced–supported Ni catalysts. It can be taken into consideration that, even though coprecipitated Ni/Al₂O₃-based catalysts might be a little more active than impregnated ones [58], the less complete use of nickel in coprecipitated samples (where a large part of nickel remains catalytically inactive in the bulk) and the likely simpler preparation of catalysts by the co-impregnation of commercial alumina with nickel and dopants make impregnation techniques advantageous (also from the environmental point of view) with respect to more complex coprecipitation procedures at the industrial level.

2.1.5. Deposition of NiO over α-Al₂O₃ and Alkali Earth Aluminates

For reactions occurring at very high temperatures, i.e., >700 °C, transitional aluminas lack the stability to be used as supports. Thus, more stable ceramic supports must be used. Among them, corundum powders can be applied [59]. Also, in this case, the deposition of Ni results in the formation of a spinel NiAl₂O₄ layer at even high temperatures up to 1100 °C, protecting α-Al₂O₃ by sintering [60,61].

Alternatively, Mg²⁺ can be added to the catalyst recipe. This can be done using MgAl₂O₄ (spinel, T_melt 2300 °C) powder as the support [59] or preparing NiO-MgO-Al₂O₃ mixed oxide materials by coprecipitation techniques or co-impregnation techniques. The addition of magnesium not only allows one to increase thermal stability [62] but may also allow for a better utilization of nickel, which will remain limited at the surface of the catalyst [53]. The MgO-Al₂O₃ system has strong similarities with the NiO-Al₂O₃ system [63], except for the fact that MgAl₂O₄ is a normal spinel while NiAl₂O₄ is, as mentioned, an inverse spinel. In general, large solid-state solubility occurs in the system NiO-MgO-Al₂O₃ [64].

Additionally, calcium aluminates are also largely used as supports for nickel catalysts for high-temperature applications, e.g., for steam reforming and autothermal reforming (see below). A number of different structures exist in the CaO-Al₂O₃ system [65]. The calcium-poorer structure CaAl₁₂O₁₉ (hibonite) is declared by Clariant as the support of some of their catalysts [59]; it has the magnetolumbite structure, similar to the β-alumina-type layered structures with spinel-type Al₂O₃ layers separated by CaO-type layers. Other hexa-aluminates, such as Ca,K- [59], Ba-, Sr- and La- hexa-aluminates, may be used as catalyst supports for Ni-based catalysts [66] in high-temperature applications. Some Ni-based Johnson Matthey catalysts are instead supported on mixed calcium aluminate phases (calcium aluminate cement [67]).

The surface properties of the materials based on alkali-earth aluminates, such as MgAl₂O₄ [68] and BaAl₁₂O₁₉ [69], are similar to those of alkali-earth-doped spinel aluminas, with reduced Lewis acidity, increased basicity with respect to transitional aluminas and a slightly reduced nickel dispersion ability [70].

2.2. Metallic Phases in the Ni-Al System

Nickel crystallizes in the face-centered cubic (fcc) structure with a melting point of 1453 °C. Cubic nickel is ferromagnetic until the Curie temperature of 354 °C. However, it has been reported that the melting point is strongly reduced for nanoparticles down to 700 °C for 2 nm size particles produced by the NaBH₄ reduction of Ni nitrate [71]. It must, however, be considered that nanoparticles produced with this technique may be amorphous [72] and can contain relevant amounts of boron impurities [73]. The Curie temperature for nanoparticles decreases so that they are superparamagnetic at room temperature [72,74].

Interestingly, hexagonal (hcp) nickel, instead of cubic nickel, is sometimes found in spent catalysts. This metastable structure is likely stabilized by the presence of carbide species [75] and can also be prepared by introducing small amounts of Pt in Ni [76] or be produced as a surface layer of cubic nickel particles [77]. Such a metastable phase is assumed to be more active in the field of electrocatalysis for the hydrogen evolution reaction. It is also supposed to be more active than cubic phase for steam reforming reactions [72].

Aluminum crystallizes too in the fcc structure. It is paramagnetic and has a low melting temperature of 660 °C. Six intermetallic compounds can exist in the Al–Ni system, namely, Al₃Ni, Al₃Ni₂, Al₄Ni₃, AlNi, Al₃Ni₅ and AlNi₃, which are all conducting materials and show a very strong chemical interaction between Al and Ni [78,79,80]. On the other hand, it must be taken into consideration that the reducibility of Ni is much easier than that of aluminum; this is illustrated by the standard reduction potentials π°= −1.676 V for Al³⁺, with respect to π°= −0.257 V for Ni²⁺, as well as by the results of the Temperature Programmed Reduction experiments. In fact, TPR data show that NiO is reduced to Ni by hydrogen at 300–450 °C depending on particle size [81], although reduction can be not fully complete in such low-temperature conditions [82]. Instead, Al₂O₃ is unreducible by hydrogen, based on thermodynamics, until potentially very high temperatures [83,84]. The TPR experiments confirm that pure aluminas do not undergo any reduction until 1000 °C. This means that, starting from Ni-Al mixed oxides, surface Ni species can be reduced selectively. It must, however, be taken into consideration that Al is soluble in fcc Ni to a considerable extent, according to phase diagrams, up to about 8% mol at 400 °C and near 20% at 1360 °C [79]. Thus, the formation of Al-doped Ni metal particles after the reduction of Ni-alumina materials is unlikely but cannot be fully excluded. To our knowledge, never has this possibility been taken into consideration.

2.3. On the Reactivity of the Metallic Ni-Based Phases

Real metal catalysis occurs when the catalyst contains an active metallic phase in real reaction conditions. It must, however, be considered that, in many cases, the elements and compounds present in the reaction medium can react with metal phases, converting them at least partially to different compound phases.

2.3.1. Oxidation of Metallic Nickel

The oxidation of metallic nickel with oxygen-containing flows to NiO is particle-size dependent. While nickel nanoparticles with a size distribution between 10 and 100 nm start to be oxidized around 200 °C and may be fully oxidized at 700 °C [85], reaction rates are slower for polycrystalline nickel, and, consequently, higher temperatures are needed [86]. The surface oxidation of nickel polycrystals in O₂ is already found at 300 °C [87,88], while the NiO layer formed at T > 600 °C results in a significant passivation of its corrosion behavior [89], i.e., in a slowing of its bulk oxidation. These data indicate that in oxidizing conditions and temperatures above 300 °C or more, nickel catalysts are expected to be largely or fully oxidized.

2.3.2. Sulfidation of Metallic Nickel

As is well known, hydrocarbons coming from oil refineries but also flows coming from biomass treatments such as biogases and biomass gasification and pyrolysis, contain, or may contain, sulfur in the form of H₂S and alkyl sulfides. In the presence of these compounds, metallic nickel can become sulfurized [90] and its catalytic activity reduced [91]. In fact, sulfur deposition from H₂S decomposition occurs relatively fast on the surfaces of nickel-based materials due to the relatively stronger bonding of sulfur compounds and the small activation energies required for the H₂S decomposition [92] producing surface sulfide species that hinder catalytic activity [91]. However, prolonged sulfidation can produce bulk nickel sulfide species. Indeed, various nickel sulfide phases exist [93] such as NiS (two polymorphic phases), Ni_3+xS₂, Ni₃S₂, Ni₇S₆, Ni₉S₈, Ni₃S₄ and NiS₂ (two polymorphs). Studies on Ni/Al₂O₃ sulfidation suggest that NiS₂ is formed as an intermediate phase and Ni₃S₂ as a final sulfide species [94].

However, it must be taken into consideration that even supported nickel catalysts may have significant metal loadings (20–60 wt%), and, in the presence of small amounts of sulfur in the feed, their sulfurization can be only partial. In fact, heavily loaded nickel catalysts are frequently preferred to low-loading noble metal catalysts just because they may have a better resistance to deactivation by sulfur due to their much higher metal surface area.

2.3.3. Reactivity with Carbon: Carburization and Formation of Carbon Nanotubes and Ni(CO)₄

Nickel metal surface reactivity is also relevant with respect to carbon when in contact with carbon monoxide and hydrocarbons. Several different forms of carbon can be produced at the surface of metallic nickel: surface carbon species, amorphous layers, filamentous carbon, bulk carbide and graphitic carbon [91]. A quite fast deactivation of unsupported nickel was reported upon CO₂ methanation at atmospheric pressure, coproducing CO, mainly associated with the production of encapsulating carbon [95] (see Figure 4). In conditions of the Fischer–Tropsch synthesis, the carburization of nickel is less extensive than that of other metals such as iron and cobalt, resulting in a less-efficient hydrocarbon-chain-growth mechanism [96]. This is the main reason why nickel-based catalysts are very active for methanation and less useful for Fischer–Tropsch synthesis. In fact, only a partial carburization of nickel to Ni₃C was found with pure CO at 265 °C [97]. This reactivity is determined mainly by an electronic effect on the CO dissociation barrier, i.e., the increase in the dissociation barrier from left to right in the periodic table, which has been correlated with the metal d band center [98,99,100] and results in the strong CO dissociation activity of iron, the medium dissociation activity of nickel and cobalt and the minimum dissociation activity of copper.

However, in the presence of hydrocarbons or alcohols, as well as at higher temperatures (>450 °C), the reactivity of nickel with carbon species results in the formation of carbon “whiskers” or nanotubes [43,101,102,103] (see Figure 5). As discussed by researchers from Topsoe [101,102], they form specifically on the external surface of “spearhead-like” and “drill –like” Ni particles with the (111) face largely exposed. These materials do not directly cause the deactivation of the catalyst but increase the density of the bed occupying the empty spaces, thus increasing pressure losses. This chemistry causes difficulties in steam reforming processes (mainly pressure drops). However, this activity is useful for the preparation of nanotube materials, although the metal-catalyzed synthesis of carbon nanotubes produces materials that still contain metal particles, i.e., impure metal nanotube materials, that are not always easy to purify from the catalyst [104,105]. The formation of carbon species from hydrocarbons in the absence of steam, i.e., the decomposition of methane and other hydrocarbons, allows for the coproduction of hydrogen [106] (see below), supposing that the fate of the resulting catalyst-containing carbon material is determined.

Another relevant aspect of Ni catalysts when working in the presence of CO is the possibility of the formation of very volatile carbonyl Ni(CO)₄, with the abstraction of Ni from the catalyst. This may thermodynamically occur in the range of 120–200 °C [107]. The formation of nickel tetracarbonyl causes the corrosion of the catalyst and the transport of nickel out of the reactor, with catalytic activity loss and relevant process problems.

2.3.4. Nickel Reactivity with Hydrogen

Bulk nickel hydride can be formed from nickel and hydrogen only at extremely high pressures (>300 MPa) [108,109] with a stoichiometry near NiH. The smaller the nickel particles, the more hydrogen they absorb up to stoichiometries about NiH_1.8 above 100 atm [109]. However, at reduced pressures, hydrogen is essentially adsorbed as surface and subsurface atomic species (see below).

2.3.5. Nickel Reactivity with HCl

HCl can be present in gases coming from biomass or waste gasification or pyrolysis, as well as in the product of hydro-de-chlorination processes. The deactivation of nickel-based catalysts by HCl in the presence of hydrogen has been reported [110,111], attributed to the agglomeration of nickel crystallites. It has been concluded that NiCl₂ is formed in the presence of HCl and is reduced back to larger-crystal-size metallic nickel [112,113].

3. Surface Properties of Materials Based on Nickel and Alumina

3.1. Characterization of Unreduced Systems

3.1.1. Surface Properties of Aluminas

The surface properties of aluminas have been the object of many investigations and reviews [17,18,114]. Transition aluminas, i.e., essentially γ-, η- and θ-Al₂O₃, whose typical surface areas are, as mentioned, 150–250 m²/g stable until 400 °C for γ- and η-Al₂O₃ and 60–120 m²/g stable until 700 °C for θ-Al₂O₃, have quite similar chemical characteristics, with a strong ionicity and acid basicity of the surface coordinatively unsaturated Al³⁺ ions—oxide anions couple, dominated by the strong Lewis acidity of low coordination Al³⁺, most probably on trigonal coordination over the high-temperature activated surfaces [115]. Their surfaces are also usually highly hydroxylated, with a medium–low Brønsted acidity of the surface hydroxy groups.

In contrast, typical and α-Al₂O₃ powders have typical surface areas of <30 m²/g, are stable up to 1000 °C and exhibit weaker surface reactivity and acido-basicity.

3.1.2. Surface Properties of Nickel Oxide NiOx (x ≥ 1)

Nickel oxide has been the object of few surface characterization studies. The IR spectroscopy of adsorbed molecules [33,116,117] shows the presence of hydroxy groups (νOH 3680, 3600 cm⁻¹) and of weakly acidic Ni²⁺ sites that are able to adsorb carbon monoxide (νCO 2150–2160 cm⁻¹) and pyridine (ν8a 1595 cm⁻¹) with mode vibrational perturbation. The dissociative adsorption of water on NiO surfaces producing hydroxy groups has been the object of computational studies [118,119]. The surface hydroxy groups have poor Brønsted acidity, being unable to protonate pyridine [33] and ammonia [120]. IR reflection studies on NiO/Ni monocrystals show that NiO adsorbs carbon dioxide at −150 °C in the form of monodentate carbonate species [121]. A more recent calorimetric and IR spectroscopic study shows that the surface reactivity of NiO towards CO₂-producing surface carbonate species is not very strong, suggesting a moderate basicity of NiO [122]. On the other hand, using the optical basicity scale discussed by Bordes and Courtine [123], based on the spectroscopic data, NiO is quite as strong of a base as the other rock-salt-type bivalent metal oxides. The surface and the bulk basicity are, according to Busca [124], mainly associated with the size of the cation, which is relatively large, or to its polarizing power, which is relatively small. According to this approach, the NiO surface is expected to be quite basic. The cationic radius of the octahedrally coordinated Ni²⁺ is reported to be 0.69 Å, which is smaller than that of Mg²⁺ (0.72 Å), thus suggesting that NiO should be a little less basic than the isostructural MgO. It seems possible to conclude that stoichiometric NiO has a prevalent, moderately basic nature.

As mentioned, in strongly oxidizing conditions, NiO becomes oxygen-rich and non-stoichiometric. XPS studies show that bulk rock-salt NiO particles may contain up to near 40% Ni³⁺ at their surface if calcined at 400–800 °C [125].

3.1.3. Surface Properties of NiAl₂O₄ and NiAl₂O₄-Al₂O₃ Solid Solutions

The surface properties of stoichiometric bulk NiAl₂O₄ and NiAl₂O₄-Al₂O₃ solid solutions have been the object of investigations by the FTIR studies of the adsorption of probe molecules, such as carbon monoxide and pyridine [34,126]. The surface of stoichiometric NiAl₂O₄ is largely hydroxylated in normal conditions, whose hydroxy group has low Brønsted acidity similar to those of the transitional aluminas. Strong Lewis acidic Al³⁺ ions are present at the surface, together with medium Lewis acidity Ni²⁺ sites and active oxide ions too. Interestingly, the CO stretching frequency of Ni²⁺ carbonyls observed over Ni aluminates after the adsorption of CO is higher (2180–2185 cm⁻¹) than that observed on pure NiO (2150–2160 cm⁻¹), showing that the Ni²⁺ centers on these catalysts are more electron-withdrawing (i.e., stronger Lewis acids) than those at the surface of NiO. By increasing Al content on the surface of the NiAl₂O₄-Al₂O₃, the solid solutions become more and more similar to those of alumina, with a decreasing amount of Ni²⁺ ions at the surface, as expected indeed.

3.1.4. Surface Properties of Impregnated NiO/Al₂O₃ Materials

Several studies report on the surface properties of unreduced NiO-Al₂O₃ powders produced by the impregnation techniques. At loadings lower than that of the “monolayer coverage” (see above), i.e., with NiO < 25 wt%, the surface probe molecules provide evidence of surface hydroxy groups with poor Brønsted acidity and small amounts of strong Lewis acidic Al³⁺ (at least two types) as well as weak Lewis acidic Ni²⁺ ions at the surface [127], whose amount increases with the nickel loading. The surface of such materials appears to be similar to those of the NiAl₂O₄-Al₂O₃ solid solutions discussed above. The XPS of alumina-supported nickel oxide also revealed the presence of Ni³⁺ after calcination at 350 °C, even in the absence of NiO [128]. Experiments with samples with a higher nickel loading show the additional presence of NiO particles, as mentioned above, which also contain Ni³⁺ ions after calcination at 350 °C [128].

3.2. Characterization of Reduced Ni/Al₂O₃ Materials

Temperature Programmed Reduction (TPR) in H₂-containing feeds is a practical technique for testing the reducibility of catalysts [129,130]. However, it must be taken into account that, while the general behavior is similar in all cases for similar samples, the exact temperature for reduction acts may significantly differ as an effect of the actual conditions of the experiment (feed composition, flowrate and resulting contact time, heating rate, etc.). It must also be taken into consideration that, at the end of the TPR experiments, temperatures such as >1000 °C are usually raised. Thus, in the experiment, the NiO/γ-Al₂O₃ catalyst usually undergoes alumina sintering, loss of surface area and phase transition, with NiO segregation, agglomeration and reduction. Consequently, at the end of the TPR analysis, the catalyst is strongly modified, and further experiments (such as TPR-TPO cycles) have no sense. The indicative, typical H₂-TPR curves are reported in Figure 6.

As already stated, the TPR experiments show that, while the NiO particles are reduced in hydrogen starting at about 300 °C to metallic nickel particles, the reduction of the surface Ni²⁺ on NiAl₂O₄ starts at definitely higher temperatures, >600 °C [35,39]. Meanwhile, at about 800 °C, bulk reduction starts to be observed with the formation of metallic Ni and γ-Al₂O₃ [35]. Surface Ni²⁺ on Al-rich spinels starts to reduce at higher temperatures than for stoichiometric NiAl₂O₄, i.e., above 700 °C [39]. The reduction temperature of near-monolayer amounts of nickel oxide impregnated over the different transitional aluminas γ-, δ-, η- and θ-Al₂O₃ (precalcined at 500 °C) is similar, i.e., above 500 °C [131]. Increasing calcination temperature results in only a slight decrease in the surface area but a significant shift of the TPR peak at higher temperatures [132], which is likely due to the stabilization of the NiAl₂O₄ subsurface layer. For impregnated samples calcined at high temperature (>700 °C), the TPR peak of nickel reduction is observed above 800 °C with a threshold at about 650–700 °C but shifts slightly to lower temperatures the higher the loading [133,134] and shows an additional medium-temperature reduction peak (550 °C) when the monolayer coverage is approached [49,135]. This reasonably concludes that, for samples pretreated at moderately high temperatures, NiO reduction occurs first (usually 300–400 °C), and highly concentrated dispersed Ni²⁺ species reduce later (500–600 °C) while highly dispersed Ni²⁺ reduces even later (>700 °C).

Surface characterization studies performed by the IR spectroscopy of adsorbed CO (Figure 7) allow one to distinguish three types of surface Ni metal sites depending on Ni loading on reduced Ni/γ-Al₂O₃ [127,136]. For low Ni loading, nearly isolated metal sites that are able to adsorb CO in the form of polycarbonyls are observed, which are produced by the partial reduction of fully isolated Ni²⁺ surface ions. At medium loadings, together with such polycarbonyl species, on-top monocarbonyl species are observed, with a shifting behavior by increasing/decreasing CO loading typical of CO bonded on metal faces (coupling effect), suggesting that more concentrated Ni²⁺ ions at the surface of alumina can produce small metallic nanoparticles, which are also observed using Field Emission Scanning Electron Microscopy (FE-SEM) and XRD. When nickel loading is even higher, larger metal particles adsorbing CO as both terminal and bridging species (νCO 2000–1800 cm⁻¹) are observed. Thus, over catalysts with up to near the formal “NiO monolayer” loading, Ni metal particles formed by reduction have a very small particle size, few or few teens of nm at most. When loading is higher, larger and more metal particles are produced by reduction (Figure 8).

4. Mechanistic Aspects in Catalysis by Ni and Alumina Materials

4.1. The Activation of Hydrogen on Ni/Al₂O₃ Catalysts

The adsorption and reactivity of hydrogen on aluminas have been the object of early studies, see Refs. [137,138,139], showing the formation of different adsorbed species at low temperatures that tend to disappear with increasing temperature [140]. Adsorbed species that are stable at high temperatures are also observed. In particular, Kramer and Andre [139] showed that atomic hydrogen can be spilt over from metallic centers to alumina and remain stable until 480 °C. More recent theoretical and experimental studies suggest that spilt-over hydrogen can be adsorbed over γ-Al₂O₃-forming hydride species that are likely bonded on trigonal Al³⁺ centers [141,142,143]. It must, however, be taken into account that these studies fully neglect the possible role of transition metal impurities that are unavoidably present on and in γ-Al₂O₃, such as, in particular, the non-negligible amount of Fe³⁺ and Cr³⁺ [144] that can either be reduced by hydrogen or be involved in H₂ dissociation. On the other hand, hydrogen spillover is far more evident on reducible oxides than on alumina [145]. Indeed, quantitatively, the adsorption of hydrogen in non-reducible oxides including γ-Al₂O₃ is essentially negligible at temperatures up to 600 °C [146].

On the other hand, hydrogen activation over oxide-supported metal catalysts is mostly considered to occur on metallic centers [147], as well as when supports are active in adsorbing it. According to the DFT studies [148], the dissociation of the hydrogen molecule occurs over the topmost Ni atom, and the resulting H atoms may adsorb either on two free hollow sites (but the adjacent bridge sites must be free) or two bridge sites (the adjacent hollow sites must be free). On nickel powder, the TPR studies [149] evidenced three distinct forms of hydrogen permanently chemisorbed at −173 °C: γ-form, located in the subsurface layer and desorbed at about −87 °C, β-form adsorbed in the so-called second layer and evolved at about 54 °C and α-form, fixed directly on the nickel surface and desorbed at the 80–400 °C range. Alumina supports insignificantly affect hydrogen that is strongly adsorbed on nickel but significantly affect weakly adsorbed hydrogen. On the other hand, the nature of the supports also affects the metal particle size and stability of nanoparticles, as well as their electronic properties [150].

In conclusion, hydrogen activation occurs on nickel metal particles, while alumina support influences the size and nature of the supported nickel particles, producing very small nickel nanoparticles when the loading is low. These supported nanoparticles may be more reactive than the larger ones. A possible role of the alumina-uncovered surface for spilling-over hydrogen atoms cannot be excluded.

4.2. The Activation of Carbon Monoxide

CO adsorbs molecularly even at low temperatures on both metals and oxides. The molecular adsorption of CO on aluminas is quite a weak phenomenon, and it is definitely weaker than on zerovalent metal centers [151]. It adsorbs at a low temperature on electron-withdrawing centers on oxide surfaces, in particular, on the Al³⁺ Lewis acid sites of aluminas through the lone pair at its C atom [152,153], resulting in the strengthening of the CO bond and a shift of the stretching frequency to higher wavenumbers (up to 2220 cm⁻¹ with respect to 2140 cm⁻¹ for unbonded CO). It desorbs easily at room or slightly higher temperatures. At moderately high temperatures, CO adsorption on hydroxylated alumina may give rise to formate ions by reaction with hydroxy groups with an “associative” adsorption mechanism [154].

CO adsorption on metallic surfaces is generally far stronger than on oxides due to the possibility of back-bonding from the partially filled d orbitals of the metal atom to the empty antibonding 2π* orbitals of CO, which produces a quite strong metal–C bond and a weakening of the C-O bond, whose stretching frequency shifts to lower wavenumbers [155]. The adsorption of CO on Ni monocrystal faces has been the object of several studies, showing the possible formation of three forms: (i) terminal (on top), whose CO stretching frequency is in the range of 2120–2000 cm⁻¹; (ii) bridging between two Ni atoms, with CO stretching at 1950–1800 cm⁻¹; and (iii) triply (or multiply) bridging over three or more Ni atoms, with a further significant weakening of the CO bond down to the CO stretching range of 1900–1780 cm⁻¹ [155,156,157,158]. The surface carbonyl species are usually stable upon outgassing at room and slightly higher temperatures, with the stability trend triply bridging > bridging > on top.

As already stated, the formation of a strong Ni-CO bond easily results in the formation of Ni(CO)₄, a stable gaseous compound at temperatures below 200 °C, which can result in nickel corrosion and being transported away. At higher temperatures, as a result of this strong interaction, CO dissociation on the nickel surface can also occur, producing carbon and oxygen species [159]. CO dissociation occurs on Ni/η-Al₂O₃ with an increasing rate at 250–400 °C and with a deposition of increasing amounts of carbon [160], occurring faster on ensembles of several Ni atoms on sufficiently highly loaded samples [161]. The CO dissociation rate is fundamentally influenced by two main factors [162]: (i) an electronic effect on the dissociation barrier, i.e., the increase in the dissociation barrier from left to right in the periodic table (with the trend Fe > Ni > Co >> Cu), which has been correlated with the metal d band center [98,99,100], and (ii) an effect of particle and surface morphology, with lower dissociation barrier on steps, kinks and defects than on terraces. As a common feature in catalysis, activation energy for CO dissociation is correlated inversely with the adsorption strength. The produced carbon species from CO dissociation can finally give rise to encapsulating carbon, which is the main cause of catalyst deactivation [91], or to the formation of bulk Ni₃C.

Both carbon species produced by CO dissociation and surface carbonyl species can be the active surface intermediates in carbon oxide hydrogenation reactions, such as methanation and Fischer–Tropsch synthesis. Most of the data strongly converge for a predominant role of CO dissociation over the metal surface and carbide mechanisms for both CO and CO₂ hydrogenation to methane over nickel catalyst reactions when carried out with syngas as a reactant [163].

4.3. The Activation of Unsaturated Hydrocarbons

The activation of unsaturated hydrocarbons can occur over oxide surfaces thanks to three different mechanisms: (i) protonation of the C=C bond by Brønsted acidic hydroxy groups; (ii) π-bonding of the C=C bond over d⁰ and dⁿ cationic sites; (iii) activation at allylic bonds by C-H dissociation over Lewis acid–base pair sites. Previous studies show that, on alumina, the last interaction actually occurs at room temperature over well-activated and dehydroxylated surfaces [164], which likely represents the key interaction involved in the double-bond isomerization activity of alumina. However, it has also been shown that the protonic acidity of alumina’s surface OHs is not negligible, and the protonation of butenes can occur at high temperatures (>400 °C) as the main step in the olefin skeletal isomerization activity of alumina [164].

As in the case of CO activation, the adsorption and activation of unsaturated hydrocarbons (acetylenics, dienes, olefins and aromatics) on metal centers also primarily occurs thanks to the electron donation from the full π-orbitals of the organic molecule to the empty d-type orbitals of the transition metal and the simultaneous electron back-donation from the full d-type orbitals of the transition metal to the empty π*-antibonding orbitals of the unsaturated organic molecule. According to the DFT calculations and STM experiments, this so-called π-bonding is the stable form of ethylene adsorbed on the flat and stepped Ni(111) crystal faces, [165] as well as on the nickel clusters [166]. However, according to calculations, this primary interaction could evolve towards more strongly adsorbed species producing multiple σ-bonding. As an example, the most stable acetylene adsorbed form on nickel should be the so-called µ-bridge, where the acetylene molecule interacts with four carbon atoms and retains a single C-C bond [165,167]. Similarly, benzene and aromatics adsorb on metallic nickel first through π-bonding, finally producing multiple σ-bonds with more nickel metal atoms [168]. At moderately high temperatures, e.g., >400 °C, all hydrocarbons will decompose on nickel surfaces producing carbon species [169].

In the presence of hydrogen, these surface hydrocarbon intermediates should later undergo the attack of atomic hydrogen species adsorbed on the nearest sites, finally giving raise to the hydrogenation reactions.

It is, however, interesting to remark that Ni²⁺ ion, being a d⁸ cation, is able to produce π-bonding, e.g., with acetylene and ethylene. This is demonstrated by the activity of NiCl₂/Al₂O₃ powders as selective adsorbents of acetylene [170] and by the high catalytic activity of Ni²⁺ supported on acidic supports, such as silica-rich amorphous alumina [171] for the ethylene dimerization process to butene where the interaction of Ni²⁺ with the olefin π bonds is a key step [172].

4.4. The Activation of CO₂

The activation of CO₂ is relevant with respect to two types of catalysts based on Ni/Al₂O₃:CO₂ and CO_x methanation and methane dry reforming. The two reactions have different thermodynamic properties (the first being exothermic, the second endothermic) and are realized in very different conditions: 300–400 °C and 20–30 atm in H₂ for methanations and >600 °C at an atmospheric pressure for dry reforming.

CO₂ adsorption occurs quite weakly at room temperature on hydroxylated alumina, mainly in the form of bicarbonates and of linear-coordinated CO₂, while easy desorption is observed at moderate temperatures. However, it can become stronger in the form of carbonate species for high-temperature dehydroxylated alumina [114] or when alkali and alkali earth ions [173] or rare earth ions [174] are present on doped aluminas or on metal aluminates [175]. On the other hand, as seen above, carbon dioxide may adsorb in the form of carbonate species over nickel oxide [121,122], which is expected to show a higher basicity than alumina. Indeed TPR data indicate that CO₂ adsorption on NiO/Al₂O₃ is weak, with a desorption temperature at about 100 °C [176], which is in agreement with the lower-surface basicity of NiO/Al₂O₃, at least in terms of the number of basic sites with respect to alumina [52].

CO₂ adsorption on metal surfaces has been the object of early spectroscopic studies on monocrystal surfaces [177,178,179]. These studies report strong spectroscopic evidence for the formation, in addition to the physisorbed CO₂ that only forms at very low temperatures, of negatively charged bent CO₂^δ−, which, depending on the nature of the metal, may dissociate into CO and O or transform into CO₃²⁻ + CO. Theoretical studies of CO₂ adsorption on transition metals [180,181,182] indicate that the strength of the CO₂ exothermic adsorption forming bent CO₂^δ−, the charge of this species, and the exothermic dissociation energy to produce CO-O, all follow the trend Cu < Ni < Co < Fe, i.e., the trend of the d-band position. A recent ambient-pressure X-ray Photoelectron Spectroscopy study [183] indicated that carbonate species, adsorbed CO and graphitic carbon form on both Ni(111) and Ni(100) surfaces in the presence of 0.2 Torr CO₂. However, more than 90% of the adsorption species on the Ni(111) surface is the carbonate, whereas the Ni(100) surface is mainly covered by adsorbed CO and graphitic carbon [183].

The adsorption of CO₂ has been studied over reduced Ni/Al₂O₃ by DRIFT spectroscopy. In addition to the bicarbonate species bonded on the alumina support, carbonate and carboxylate species have been observed, that are supposed to be coordinated at the Ni–alumina interface and over metallic nickel, respectively, and the TPR data indicate that these species are adsorbed stronger than bicarbonates on bare alumina [184]. Interestingly, no CO coordinated to the Ni, which might be originated from the CO₂ dissociation. Consequently, it seems that the activation of CO₂ at low/moderate temperatures for its further hydrogenation on Ni/Al₂O₃ would occur mostly on Ni metal centers or at the support/metal interface, although, in the case of the presence of basic dopants, the support might have a role too [185]. On the other hand, data indicate that metallic nickel may be oxidized by CO₂ above 700 °C, producing mainly CO and partially oxidized nickel [186]. This reactivity might be involved in the dry reforming of the hydrocarbons that only occur near that temperature.

4.5. The Activation of Water

The adsorption of water on alumina results in the formation of hydroxy groups, by dissociation. A part of such hydroxides appears to have moderate acidity, being unable to protonate pyridine, but being able to protonate stronger bases such as amines and piperidine [17,18]. On the other hand, a part of the hydroxy groups shows moderate basicity/nucleophilicity, being able to produce hydrogencarbonate species reacting with CO₂ at low temperatures [114] and formate species reacting with CO at moderately high temperatures [154]. Thus, water adsorption on alumina can result in its activation as a base or as an acid. The un-reducibility of alumina does not allow for activation in terms of redox chemistry. A similar acid–base activation of water is observed on NiO and on NiO/Al₂O₃ powders, where surface hydroxy groups are observed too (see above) and are thought to have a predominantly basic character.

Water adsorbs also on nickel metal at very low temperatures. No nickel oxidation was found by reaction with steam at 25 °C [88]. However, XPS data on Ni polycrystalline surfaces show that, at 300 °C, water vapor produces a thin NiOx surface layer (<2 nm thickness) that also contains Ni³⁺ ions, similar to that produced by oxygen [88]. Thus, an oxidizing layer can be produced by the reaction of metallic nickel with water at relatively low temperatures. Experiments of the steam reforming of ethanol, a molecule whose activation is easy at a low temperature, over unsupported nickel allowed us to conclude that water activation is needed for such reactions and that the temperature threshold for water activation on unsupported nickel is less than 400 °C [72].

4.6. The Activation of Oxygen

Alumina as such is essentially unable to activate oxygen [140], except maybe in the case of impurity centers [144]. As reported above, calcined NiO, as well as NiO-Al₂O₃-based materials, do present XPS features attributed to Ni³⁺ (see above). In the case of NiO, this produces p-type semiconducting NiO_1+x materials where the excess oxide ion species are supposed to be on reticular positions and are balanced by nickel vacancies [187], which are likely mainly occurring near the surface. At the surface of NiO/Al₂O_3, a similar mechanism should occur, with the formation of additional surface oxide ions and the oxidation of Ni²⁺ to Ni³⁺. This is expected to be a mechanism of oxygen activation upon oxidation catalysis over nickel-oxide-based materials.

On the other hand, data show that polycrystalline Ni already undergoes surface oxidation by O₂ at 25 °C, producing a NiO_1+x layer containing both Ni²⁺ and Ni³⁺ [87]. Thus, metallic nickel is also able to activate oxygen at low temperatures. However, obviously nickel metal would no longer exist in an oxygen-containing atmosphere at moderately high temperatures, as discussed above.

4.7. The Activation of Saturated Hydrocarbons on Reduced and Oxidized Nickel

The activation of methane over heterogeneous catalyst surfaces has been the object of several studies [188]. Among the simple transition metal oxides, NiO has the highest catalytic activity for methane combustion, with an ignition temperature in the range of 450–500 °C. This is attributed by a reaction with surface oxide ions interacting with Ni³⁺ ions [189] and, thus, with an Eley−Rideal mechanism. Thus, oxygen activation is needed more than methane activation for its combustion. Similar activation mechanisms can be proposed for the catalytic combustion of higher paraffins, whose reactivity is expected to be higher than for methane.

In the absence of water and oxygen, the formation of carbon species from several light hydrocarbons on nickel foils occurs over the temperature range of 400–600 °C [169]. Carbon formation from paraffins is comparatively slower than from olefins and acetylene. In fact, methane activation on metal surfaces, including nickel, has been reported to start with a single C–H bond breaking as the molecule collides with the surface, leaving chemisorbed H and CH₃ fragments [190], which occur at about 100 °C [191]. This step is quite a slow one, is very temperature-sensitive and is supposed to be followed by the sequenced dehydrogenation of carbon species. On the other hand, the production of hydrogen by methane decomposition on nickel catalysts occurs at relatively high temperatures, i.e., in the 400–600 °C range [192,193].

When higher paraffins are reacted, a similar dissociative activation step can be presumed to occur on nickel metal particles as a first step for dehydrogenation reactions. On the other hand, nickel catalysts do mostly show low selectivity in paraffin dehydrogenation due to the too high activity in C-C bond cracking finally producing decomposition and carbon deposition. The selectivity in isobutane dehydrogenation on Ni/Al₂O₃ is, however, increased by partial sulfidation [194] that likely poisons the too-active sites. As is reported below, however, nickel catalysts—in particular, Ni/Al₂O₃—are reported to act as excellent catalysts for cycloparaffin aromatization. Early studies suggest that carbonized Ni metal surfaces are active in cyclohexane aromatization [195]. On the other hand, authors suggest that unreduced dispersed Ni²⁺ on alumina may act as the active site for propane dehydrogenation on Ni/Al₂O₃ [196]. Thus, the real active sites for paraffin dehydrogenation on Ni-Al₂O₃ catalysts are still the object of debate and can be different in different conditions.

In conditions where both paraffins (most commonly methane) with water vapor (steam reforming), or with CO₂ (dry reforming), or with oxygen (partial oxidation), or all together (tri-reforming), are present together, it is not evident whether alkane activation is needed and whether it occurs on nickel metal surface or over the oxide layer produced by water, CO₂ or oxygen activation. In fact, the temperature ranges of methane, water and CO₂ activation nearly superpose. Most authors believe that methane activation occurs on the nickel surface that is reacting further with adsorbed oxygen species in all cases [188], and the usual deactivation of such catalysts by carbonaceous materials or, at least, the evident formation of carbonaceous species agrees with this conclusion. Among the supported nickel catalysts, Ni/Al₂O₃ is reported to have the lowest onset methane conversion temperature [197], thus making alumina the best support for producing active nickel species.

5. Deactivation and Reactivation of Ni/Al₂O₃ Metal Catalysts

Different types of deactivation mechanisms may affect catalyst behavior [91]. Some of them are due to solid-state reactions slowly occurring at moderately high temperature, such as the loss of support surface area (intraparticle sintering), support phase transition (γ- to θ- to α-Al₂O₃) and the formation of bulk NiAl₂O₄. The sintering of supported metal particles usually occurs in parallel. These phenomena are essentially irreversible and do not allow for catalyst regeneration. In correct industrial conditions, they usually occur very slowly over years. When deactivation is excessive, the catalyst bed must be substituted by a fresh one.

Ni metal particle sintering can also occur without significant support sintering [198]. This phenomenon is also a thermal one but may also be favored by the adsorption of molecules, such as HCl (see above). In this case, at least partial regeneration can be attempted either by steaming or by the oxidation/reduction of the catalysts, allowing for nickel redispersion [199,200].

As already cited, Ni metal catalysts can react with species in the reactant medium and become deactivated by poisoning. To reduce this drawback, feed pretreatment can sometimes be realized: as an example, natural gas is virtually fully desulfurized before steam-reforming reactors to avoid nickel sulfidation. Thus, in most cases, the trace of contaminants still unavoidably present can cause deactivation very slowly (years on steam), and catalyst regeneration is consequently not needed.

When, however, the presence of poisons in the feed cannot be avoided, deactivation can be faster, and catalyst regeneration must be carried out. The surface or bulk sulfidation of nickel metal particles results in partial or total deactivation in most cases [198]. However, the adsorption and desorption of H₂S are reversible phenomena, and this may allow for the equilibration of adsorbed and gas–phase species, resulting in limited deactivation. On the other hand, treatment in sulfur-free feeds can allow for surface cleaning by H₂S desorption with the recovery of catalytic activity. Alternatively, oxidation may allow for the removal of sulfur as SO₂ [201]. Steaming can also allow for the removal of sulfur [202,203], although, in these cases, residual sulfate species may remain on the surface. In oxidizing conditions, sulfur can be present in the feed, mainly as SO₂, and this can also cause catalyst poisoning by sulfites and sulfates [204]. Oxidation and reduction treatments can remove sulfur poisons also in this case.

As said above, in the presence of carbon reactants (carbon oxides, hydrocarbons, alcohols, etc.), the formation of surface carbon deposits (surface carbide species, encapsulating carbon and carbon whiskers) also can occur at high temperature [205]. While carbon whiskers do not cause evident catalyst deactivation, at least initially, they causereactor overpressures with time [101,102]. Surface carbides and encapsulating carbon, instead, usually reduces catalytic activity down to total deactivation [95,198]. In this case, heating in air or oxygen results in the burning of carbon residues. Alternatively, steaming results in the gasification of surface carbon species. Finally, hydrogenation can remove carbon deposits also [91]. These procedures, followed by re-reduction treatments, can allow for the at-least-partial regeneration of the catalysts.

As is quite obvious, regeneration pretreatments must be carefully realized, avoiding excessive heating phenomena to not cause alumina sintering and phase transitions during regeneration.

6. Applications of Catalysts Based on Nickel and Alumina or Aluminates

6.1. Commercial Ni-Al₂O₃ Catalysts for Hydrogenation Reactions

In most hydrogenation reactions, noble metals such as Pt, Pd and Ru are far more active at the same loadings than nickel catalysts. However, thanks to its much lower cost, nearly 1000 times less than ruthenium and more than 2000 times less than platinum and palladium, nickel loading in commercial catalysts can be far higher. This allows for higher catalytic activity and the better management of sulfur poisoning. In fact, the same sulfur content in the reactant flow is more detrimental for <1 wt%-loaded noble metal catalysts than for 20–50 wt%-loaded supported Ni catalysts due to the much higher metal surface area. For these reasons, highly loaded Ni on alumina may definitely be competitive with extremely low-loading, supported, noble metal catalysts in several practical cases and is still much cheaper.

In this chapter, we will not take into consideration the typical hydrodesulfurization and hydrotreating catalysts that are usually based on alumina-supported MoS₂ and WS₂ where nickel may act as an activator (the so-called NiMo and NiW systems). This is because the main role in catalysis in this instance is due to the sulfides of molybdenum and tungsten.

6.1.1. Hydrogenation of Acetylenics and Dienes

Ni/Al₂O₃-based catalysts may be used instead of the more active Pd or Pt catalysts for hydrogenations of unsaturated hydrocarbons, such as to reduce acetylenics and aromatics in hydrocarbon mixtures coming from steam cracking or fluid catalytic cracking. In fact, Ni/Al₂O₃ [206] and Ni/NiAl₂O₄ [46] catalysts have interesting activities in acetylene-selective hydrogenation to ethylene. Ni/Al₂O₃, possibly containing copper, has excellent activity in the hydrogenation of acetylenics in C2, C3 and C4 cuts in the presence of butadiene [207] and can be preferential with respect to noble metal catalysts if the feed contains significant amounts of sulfur compounds. Clariant (previously Süd Chemie) offers multi-promoted NiO catalysts supported on silica alumina (0.6–2.6% NiO), denoted as the OleMax 100 series [208] for the treatment of FCC offgases. The hydrogenation of α-acetylenic compounds (1-butyne and vinylacetylene) may be performed with 100% yield to purify the steam cracking C4 cut in order to simplify the further extraction of 1,3-butadiene (e.g., by N-methyl pyrrolidone). This occurs [209] in the UOP-KLP process, using a single fixed bed of the KLP-60^TM catalyst, Cu-Ni/Al₂O₃, in the liquid phase, as well as with the Hüls Selective Hydrogenation Process (SHP) using the Ni-based H-15 SHP catalyst from the UOP. Catalysts of 1–25 wt%Ni on θ- or γ-alumina have also been patented for the selective hydrogenation of phenylacetylene impurities in styrene/ethylbenzene mixtures [210] where they were found to resist deactivation better than the Pd-based catalysts.

Small amounts of straight-chain C10 to C13 di-olefins are selectively hydrogenated to the mono-olefins in the feed to the alkylation reactor in the process of the production of Linear Alkyl Benzene (LAB). The reaction is carried out in a liquid phase at 170–230 °C and 10–20 bar on nickel–alumina (Ni/Al₂O₃) catalysts [211,212].

Pyrolysis gasoline is purified from acetylenics and dienes by hydrogenation before the hydrodesulfurization occurs. Johnson Matthey offers the HTC NI 200 (12 wt% Ni on alumina [213]), which is available in three forms—oxidic (OX), pre-reduced and air passivated (RP) and pre-reduced, air passivated and pre-sulfide (RPS) [214]—and has stronger heavy metal and sulfur resistance with respect to Pd catalysts. Also, the Topsoe SH-501 catalyst [215], as well as the different Axen catalysts [216] offered for the first step of pyrolysis gasoline hydrogenation, are based on the Ni/alumina systems.

6.1.2. Dearomatization Reactions

Supported nickel catalysts may be used for the dearomatization of hydrocarbon flows in competition with Pt catalysts. Due to the wide pore structure and small nickel crystals providing high strength, activity, selectivity and poison resistance, together with lower gum/polymer formation and side reactions, nickel catalysts can offer comparable performance to noble metal catalysts. Clariant (previously Süd Chemie) offers Ni-based catalysts (the NiSAT^® family, with NiO content from 43% to 77%) for naphtha dearomatization. The commercial nickel catalyst NiSAT 200 is reported to contain between 25 wt% and 50 wt% nickel monoxide, below 25 wt% calcium aluminate and more than 20 wt% silica [217]. The Johnson Matthey HTC Ni dearomatization [214,218] and solvent purification catalysts [219], from HTC Ni 200 to HTC Ni 900, are based on Ni/Al₂O₃ with Ni from 15 to 30 wt%; in particular, HTC Ni 500 is 21% Ni on alumina [220]. Also, the Axens LD 746 and AX 746 dearomatization catalysts are based on Ni/Al₂O₃ [216].

6.1.3. Low-Temperature Methanation of CO_x for Hydrogen Purification

Low-temperature methanation is usually realized in processes for preparing hydrogen for hydrogenation reactions or H₂/N₂ mixtures for ammonia synthesis [221,222]. They are realized to reduce as much as possible the concentration of CO_x in such gases because of the poisoning effect of CO_x towards hydrogenation catalysts. This reaction is realized in highly diluted CO_x in H₂ mixtures (CO_x concentrations most commonly < 1% mol), with CO being the predominant CO_x compound in the feed, allowing the CO_x concentration to be reduced to less than 5ppm in the process gas leaving the methanator. As a consequence of the small number of reacting molecules, the evolved heat is low, and adiabatic reactors can be used. The reaction temperature is typically 170–350 °C. Ruthenium-based catalysts are most active at low temperatures. However, they have found only a limited area of application due to their high cost. Due to their sufficient activity, stability and relatively low cost, Ni-based catalysts are predominantly used for conventional low-temperature methanation [221,223,224]. Typical nickel loadings are in the range 20–45 wt% as NiO, over γ-Al₂O₃ or metal aluminate supports, or Cr₂O₃ [223]. Clariant produces catalysts with 43% and 25% NiO on a support (likely alumina) METH 135 and METH 134 in the oxidized form [59]. Topsoe produces the prereduced PK-7R catalyst, constituted by >23 wt% Ni [225] on an alumina carrier, which ensures that CO and CO₂ are fully converted to methane at an operating temperature of 190 °C. Johnson Matthey sells catalyst precursors (Katalko 11 series) constituted by 35 wt% NiO supported by refractory oxides, such as alumina–silica–lime–magnesia, strengthened with calcium aluminate cement [226].

6.1.4. High-Temperature Methanation of CO_X for the Production of Substitute Natural Gas

High-temperature methanation is the conversion of a CO+CO₂-rich syngas that comes from coal gasification or biomass conversion to produce a methane-rich gas to be used as a Substitute Natural Gas (SNG [227,228]). Due to the much higher CO_x concentration in hydrogen (usually in the 25–35% mol range) and their high conversion degree, the heat evolved is very high, and the temperature, at least in the first reactor, is also definitely high. Although real industrial applications of this technology have been very few in the previous century, several plants appear to having been recently put in operation or are under construction now—in particular, in China [229]. Only Ni-based catalysts are apparently used, taking into account that the extent of the reaction under industrial conditions is more heat and mass transfer-limited, rather than kinetic-limited. With these catalysts, selectivity to methane in the full process is very high, although the carbon deposition and formation of higher hydrocarbons occur to a limited extent. In the Topsoe’s TREMP™ process, the catalyst (MCR2X) [230] is a supported nickel catalyst containing 22 wt% Ni on a stabilized support, possibly based on MgAl₂O₄ [231], with a surface area decreasing from 50 m²/g (fresh) to 30 m²/g (used) [232] and with stable activity up to 700 °C. Other nickel-based catalysts, MCR-8 and PK-7R, are used in medium and low temperature steps. Catalysts for this process need to be sufficiently active at low temperatures, resistant against sintering at high temperatures, and to other phenomena producing deactivation, i.e., gum and Ni(CO)₄ formation [233]. The Johnson Matthey CRG catalysts used in the process are apparently based on Ni and refractory supports such as alumina–lanthana–magnesia [234]. They catalyze the methanation reaction and, simultaneously, the Water–Gas Shift (WGS) reaction, thanks to water produced by the methanation reaction. Also, the catalysts provided by Clariant Catalysts (Munchen, Germany) [235], with a wide operating temperature range (230–700 °C), are based on nickel [236].

6.1.5. Methanation of CO₂

The methanation of the captured CO₂ using green hydrogen is considered a possible way to valorize and reuse it, in the frame of the so-called “power to gas” technologies [237,238]. Many projects and plants with various sizes have been undertaken in the last fifteen years [239,240]. The new processes currently under development for CO₂ methanation [241,242] are strictly related to the ones described above for SNG production from CO-rich syngases. Although an enormous number of projects are under development, information on real industrial applications is scarce. One of the earliest examples of power-to-gas technology was Audi’s e-gas plant in Werlte, Germany, which was constructed by Etogas to produce renewable synthetic methane from hydrogen and carbon dioxide emissions and to achieve low-carbon mobility for Audi’s A3 Sportback g-tron vehicles. Clariant provides the catalyst to be applied in such an application [243]. No detailed information is available on the Clariant’s SNG 100 DCARB methanation catalyst just developed for power-to-gas processes though it seems that it is based on nickel [236]. In fact, taking into account that practical reaction temperatures in the catalyst bed are relatively high, the low-temperature activity of noble metals such as ruthenium is not necessary, at least in main methanation reactors. Thus Ni-based catalysts are most promising [244]. High-temperature stability is certainly a key feature. A very large number of different catalyst compositions have been proposed in the literature for CO₂ methanation. A critical analysis of the scientific literature data concerning catalysts for CO₂ methanation allowed us to propose that Ni/Al₂O₃ catalysts with medium metal loading that allow for small metal particles to cover almost entirely the support surface, with additional ceria and/or basic oxide dopants as stabilizer and activating components, would be very active toward methanation and not as much prone to produce carbon residues and gums [185]. Takovite-derived coprecipitated Ni/Al₂O₃ catalysts can also applied to the 3D printing technique to fabricate commercial spherical CO₂ methanation catalysts [245].

6.1.6. Reverse Water–Gas Shift

The reverse water–gas shift reaction (rWGS) may represent an interesting approach to produce CO, or CO-containing syngases, from captured CO₂ as a first step for further conversion [246], e.g., to synthetic and renewable fuels from the Fischer–Tropsch process, acetic acid, oxo-alcohols or methanol syntheses and methanation. Being rWGS, an endothermic reaction, which is shifted to products only above around 830 °C, when its ΔG° starts to become negative, relatively high temperatures are needed for high conversion. Typical water–gas shift catalysts based on Cu/ZnO/Al₂O₃, Fe₃O₄ or ZnAl₂O₄, which are active also for rWGS at low temperatures, cannot be applied at high temperatures [185]. Also, Ni/γ-Al₂O₃ catalysts at low nickel loading were found to be active and selective in rWGS instead of the competitive methanation reaction at low temperatures [50]. In fact, several companies such as Topsoe (Lyngby, Denmark) [247], Clariant Catalysts (Munich, Germany) [248], Johnson Matthey (Billingham, UK) [249] and Axens (Salindres, France) [250] are developing catalysts for rWGS based on nickel on refractory supports, which will likely include corundum and alkali-earth aluminates. It is, however, supposed that the real reaction occurring is a sequence of methanation and methane steam reforming instead of direct rWGS over these catalysts.

6.2. Ni/Al₂O₃-Based Catalysts for Other Hydrogenation Reactions

6.2.1. Application in Amine Syntheses

In addition to the above applications, which have industrial relevance in refinery and petrochemistry, Ni/Al₂O₃-based catalysts may have applications in the hydrogenation field. In particular, Ni/Al₂O₃ systems are cited among the industrial catalysts for the synthesis of amines through the gas-phase catalytic hydrogenation of nitriles and reductive amination reactions [131,251]. In particular, Ni/γ-Al₂O₃, as such or also containing MgO, is reported to act as a good catalyst for the hydrogenation of acetonitrile to ethylamine [252] and for the reductive amination of adehydes (including purely aliphatic ones), aryl–alkyl, dialkyl and diaryl ketones to primary amines [253]. The Johnson Matthey HT 500 catalyst, based on 21 wt% Ni/Al₂O₃, is an active catalyst for the synthesis of aniline by nitrobenzene hydrogenation [254].

6.2.2. Hydrogenation of Oxygenates

Several studies report on the good catalytic activity of nickel-based catalysts in the hydrogenation of oxygenated compounds in fine industrial chemistry [255]. Recent studies report on the very high activity and selectivity of Ni/Al₂O₃ catalysts in the hydrogenation of n-octanal to n-octanol [256] and for the liquid-phase hydrogenation of hydroxymethyl furfural to tetrahydrofuran-2,5-diyldimethanol (THFDM) in water [257]. Similar catalysts also exhibited high catalytic activity in the maleic anhydride hydrogenation with 92% selectivity to succinic acid and nearly 100% conversion at 140 °C [258]. The industrial HT-500 catalyst from Johnson Matthey is active in the hydrogenation of a number of oxygenated compounds, but its stability is strongly improved in the presence of acid components by the addition of 0.1–0.25 wt% ruthenium [220].

A number of chemicals must be purified of impurities by the hydrogenation of unsaturated impurities. 2-ethylhexanol is an important industrial material, in particular, for the production of its phthalate esters as lubricants and plasticizers and is usually obtained by the hydrogenation of 2-ethylhexenal. However, the quality of the product is greatly affected by the existence of unsaturated by-products (2-ethylhexanal and 2-ethylhexenol) and residual reactant (2-ethylhexenal). Ni/Al₂O₃ catalysts can be applied to purify crude ethylhexanol by the hydrogenation of such impurity molecules [259].

6.2.3. Hydrogenations and Hydrodeoxygenation of Renewables

As discussed elsewhere [2], the production of biofuels is expected to grow in the near future to contribute to the defossilization of transport technologies, in particular, in the field of shipping and aviation. To produce biofuels, the hydrogenation, hydrodeoxygenation (HDO) and/or deoxygenation of biomass-derived organics is needed. Among the promising processes and products, hydrogenated vegetable oils (HVOs) are an already-growing market, such as green diesel and sustainable aviation fuel (green jet fuel). The theoretical reaction stoichiometry for tri-stearate full hydrogenation is the following:

(C₁₇H₃₅COO)₃C₃H₅ + 12H₂ → 3C₁₈H₃₈ + C₃H₈ + 6H₂O

where ΔH°₂₉₈ = −888 kJ (liquid water) [260], showing the strong exothermicity of the process. The full hydrogenation of vegetable oils, mainly non-edible ones such as jatropha oil, castor oil and spent frying oil waste, is mainly realized today using alumina-supported Ni-Mo sulfides, which need sulfur in the feed to maintain stability. The process produces a mixture of hydrocarbons, mainly linear paraffins, and also some CO₂ by decarboxylation. Nickel/alumina catalysts with high nickel loading (60 wt%) were found to have stable activity for this reaction at H₂ pressure, 40 bar and temperature 350 °C [58,261]. Interestingly, Ni/MgO-Al₂O₃ materials (up to 20 wt% Ni) were also found to be active for this reaction [262] but also in the H₂-free deoxygenation of vegetable oils, producing olefin-rich, diesel-like hydrocarbon mixtures by catalytic cracking at 450–500 °C [263]. The 10 wt% Ni/γ-Al₂O₃ catalyst exhibits good performance in the HDO of methyl laurate at 400 °C and 20 bar, mainly producing n-undecane [264].

Both 10% Ni/γ-Al₂O₃ and, even more, 10% Ni/α-Al₂O₃ were found to exhibit high activity in the hydrogenolysis of molecules similar to lignin monomers, such as diphenyl ether (DPE) and benzyl phenyl ether (BPE), but also show some activity in full lignin hydrolytic depolymerization [265]. In fact, an effective hydrolytic depolymerization of lignocellulosics was found using EDTA-activated Ni/Al₂O₃ [266] or by using Ni/Al₂O₃ in methanol [267].

6.2.4. Hydrogenolysis and Pyrolysis for the Conversion of Plastic Wastes to Useful Products

Another important point for the future is to convert wastes into chemicals, in particular, to recover organics from plastics [2]. Ni/Al₂O₃ catalysts were found active for the hydrotreating of plastic pyrolysis oils [111,268]. Similar catalysts are also reported to be very active for the pyrolysis and the production of hydrogen [269,270] from plastic wastes. Indeed, the pyrolysis of plastics will induce the dehydrogenation and hydrogenolysis of polyolefins together.

6.3. Catalysts for Hydrogen and Syngas Production

6.3.1. Industrial Steam Reforming of Methane, Natural Gas and Higher Hydrocarbons

Hydrogen is mostly produced today through the steam reforming of hydrocarbons, usually natural gas or methane [221,271,272,273], performed at 750–900 °C, 30–50 bar. Although Ni/γ-Al₂O₃ catalytic materials are even more active in short time frames in steam laboratory experiments [274], the industrial catalysts are mostly based on Ni supported on aluminate-based carriers, usually stabilized by the presence of alkali and/or alkali-earth cations to obtain the refractory character needed for stability at such a high temperature [275]. Typical methane steam-reforming catalysts contain 10–25 wt% Ni, 70–85 wt% Al₂O₃ or aluminate supports and up to 5 wt% K, Ba, Ca, etc., and have a limited surface area of 10–30 m²/g. The support phase is in fact a refractory oxide such as α-Al₂O₃, Mg aluminate spinel MgAl₂O₄, calcium aluminates such as CaAl₁₂O₁₉ and calcium-potassium aluminate CaK₂Al₂₂O₃₄ [59]. The formation of carbon nanotubes is a main difficulty in particular when natural gases containing high amounts of higher hydrocarbons are fed. In this case, over-pressions can appear, as well as catalyst deactivation results. To limit this phenomenon, cesium- or potassium- doped catalyst may be put at the beginning of the reactor. As an example, the fresh Johnson Matthey catalyst Katalco 57-4Q catalyst (29 m²/g) contains 18 wt% NiO supported over a mixture of calcium aluminate phases, while its alkalized analogue Katalco 25-4Q additionally contains 1.8 wt% of K₂O [276]. Similarly, the Clariant ReforMax 330 LDP Plus catalyst for methane SR is reported to contain 14 wt% NiO over CaAl₁₂O₉, while the alkalized ReforMax 210 LDP Plus catalyst contains 18 wt% NiO and 1.5 wt% K₂O over CaK₂Al₂₂O₃₄ and can also be used for the steam reforming of liquified petroleum gas, i.e., propane and butane. For naphtha steamreforming, Clariant commercializes the ReforMax 250 catalyst, with 23 wt% NiO and 6 wt% K₂O over Mg aluminate [59].

6.3.2. Industrial Pre-Reforming of Natural Gas

Modern large hydrogen production plants contain a previous adiabatic fixed bed pre-reforming reactor, where higher molecular weight hydrocarbons are pre-reformed. The steam reforming of propane and butane occurs at lower temperatures (up to 650 °C), and this allows one to avoid their decomposition to carbon and hydrogen in the main tubular reactor [271]. These catalysts contain much more Ni (NiO 45–60%), have very high nickel surface areas (typically 20–30 m²/g_cat, compared to <2 m²/g_cat for a primary reformer catalyst) and usually a Mg aluminate support with other additional components (some silica and chromia). The Clariant Reformax 100 pre-reforming catalyst contains 56 wt% NiO [59].

6.3.3. Methane Dry Reforming and Biogas Reforming Reaction

The so-called “dry reforming” reaction, i.e., the reaction between methane and carbon dioxide to produce syngas and hydrogen, represents a possible future new way to produce hydrogen, consuming two greenhouse gases. In particular, this reaction can be interesting to produce renewable hydrogen when applied to biogas [277]. While Ni/γ-Al₂O₃ catalysts are excellent for this reaction, they tend to deactivate by carbon deposition. Additionally, the relatively high reaction temperature needed (>700 °C) may cause catalyst instability for long times on steam. Both drawbacks can be approached and improved by adding a basic component to the catalyst, such as, e.g., MgO or CaO, and adding small amounts of noble metals (e.g., Pt) [278,279,280,281].

6.3.4. Methane Partial Oxidation, Secondary Reforming and Autothermal Reforming

The partial oxidation of methane, i.e., the oxidation of methane in defect oxygen or defect air, represents another way for producing syngas and hydrogen. It is an exothermic reaction and, for this reason, can be interesting with respect to steam and dry reforming. In practice, methane partial oxidation is commercially realized today, together with steam reforming in secondary reformers, with air and a methane content of 15 mol%, realized in the processes for producing hydrogen gas for the ammonia synthesis. In this case, the gas entering the catalyst bed, after a first thermal combustion and reaction step, is at 1050–1100 °C, where further endothermic reforming reactions take place to achieve an equilibrated gas of 925–975 °C at the exit of the reactor [272]. Additionally, this reaction is realized in autothermal reformers, e.g., the processes for producing syngases for methanol and synthesis where the steam reforming and partial oxidation of methane occur together. In this case, the temperature of the combusted gases entering the catalyst bed, after the non-catalytic step, is 1250–1300 °C.

Ni/γ-Al₂O₃ catalysts are active for this reaction. However, due to the high temperature during the reaction, the Ni/θ-Al₂O₃ catalysts and Ni/α-Al₂O₃ catalysts may have more stability on steam [282,283]. In fact, commercial catalysts for secondary reforming and autothermal reforming from Clariant are based on 11.5 wt% NiO on CaAl₁₂O₁₉ (ReforMax 410 LDP catalyst) but a top layer 11.5 wt% NiO on α-Al₂O₃ (ReforMax 400 GG) is recommended. For autothermal reforming, Clariant proposes approx. 5–10% of ReforMax 420 (7.4 wt% NiO on α-Al₂O₃) on top of ReforMax 330 LDP catalyst (14 wt% NiO on CaAl₁₂O₁₉) [59].

6.3.5. Methane and Hydrocarbon Decomposition Catalysts for the Production of Hydrogen and Carbon Nanomaterials

As already stated, Ni-based catalysts may react with hydrocarbons (usually methane) producing solid carbon species and C-free gaseous hydrogen. This is the so-called turquoise hydrogen [106,284]. Nickel-based materials are among the most active catalysts for this reaction. This reaction occurs at 800–900 °C over the 10% Ni/Al₂O₃ catalyst stabilized by calcium [285] with the formation of carbon nanotube-like deposit over the catalyst. The encapsulation of metallic nickel particles finally stops the reaction due to catalyst deactivation. In the case of the Ni/MgO-Al₂O₃ catalysts, the catalyst loaded with 25 wt% Ni has more prolonged activity than the 15 and 50 wt% Ni-loaded catalysts [286].

This activity can allow for the coproduction of hydrogen and useful carbonaceous materials, although the separation of carbon species from the catalyst is a problem. In practice, the production of carbon nanotubes [104,105] occurs over Ni/Al₂O₃ in the 450–700 °C temperature range. The amount of carbon nanotubes (CNTs) formed before the deactivation of the catalyst increases with the nickel content in the catalyst [43,287]. On the other hand, continuous nanofiber layers can be produced using very-low-loaded (0.04 wt%) Ni/α-Al₂O₃ catalysts [288]. The addition of other catalyst components, such as an alloying metal or an oxide component, can improve the activity and modify the morphology of the produced carbon material [284].

6.3.6. Ammonia Cracking Catalysts

Ammonia decomposition is commonly applied today on a small scale to produce forming gas that is used in metallurgical processes. In the near future, ammonia may be used as a hydrogen carrier molecule [289], thus large-scale ammonia cracking could be developed. While noble metal-based catalysts are needed for low-temperature ammonia cracking, nickel/alumina or aluminate-based catalysts [290] are commercial for high temperature ammonia cracking at temperatures up to 900 °C [291]. Johnson Matthey sells the Katalco™ 27–2 high-temperature ammonia cracking catalysts that typically operate in the range of 700–950 °C, giving high hydrogen recovery due to the low residual ammonia levels at this temperature range [292]. This catalyst is based on Ni/calcium aluminate, with a Ni content of about 15 wt% [293].

6.3.7. Steam Reforming of Biomass Tar for Purification of Biomass Gasification-Derived Syngas

The gasification of biomasses represents a promising and already quite established approach to produce syngases [294], whose combustion or conversion may give rise to renewable energy, bio-hydrogen and liquid biofuels. However, gasification technologies present a number of drawbacks, among which is the production of tar molecules polluting the syngas. Biomass tar is a complex mixture of quite-heavy organic molecules produced together with other low-volatility oxygenates. These molecules can condense in the cool sections of the plant (like on heat exchangers), producing fouling, or may cause the coking and deactivation of catalysts (such as for the water–gas shift) [295] and anodes of fuel cells [296]. Catalytic steam reforming appears to be one of the most interesting approaches to destroy such molecules. Ni-based steam-reforming catalysts similar to those discussed above for methane steam reforming [297] appear to be active for this purpose but sensitive to sulfur poisoning due to the small amount of H₂S usually present in the syngas. However, at relatively high temperatures, Ni-based tar steam-reforming catalysts maintain activity and are applied at the industrial level [298]. In particular, Ni/Al₂O₃ are active in the steam reforming of tar molecules, although care must be taken at controlling the temperature and during shut down and startups [299]. When doped with Mg or Ca, Ni/Al₂O₃ systems appear to be among the most active for the catalytic abatement of tars [300].

6.3.8. Steam Reforming, Partial Oxidation and Autothermal and Dry Reforming of Renewables for the Production of Renewable Hydrogen

Renewable hydrogen can be produced by the steam reforming of biomass-derived organic molecules. In particular, bioethanol steam reforming (ESR) [301,302] has been investigated deeply: it is an endothermic reaction and thus favored at relatively high temperatures and low pressures. It can be realized at 600–700 °C over metal catalysts such as supported Ni, Co, Pt or Rh, including Ni/Al₂O₃ [303], and is frequently improved by alloying [304] with excess water. It seems that this technology is still not realized at the industrial level but is seriously considered for commercial application [305]. It has been considered that ESR can be realized using existing hydrocarbon steam-reforming plants, provided effective catalysts have been developed [306]. Additionally, it has been shown that Ni/Ca aluminate commercial catalysts for natural gas steam reforming are excellent catalysts for ESR too in laboratory experiments [307].

Steam-reforming processes to produce hydrogen can also be applied to other biomass-derived substances [308] such as vegetable oils [309], wastes from agricultural manufacturing, glycerol and biomass-pyrolysis oil [310]. Also, the partial oxidation and autothermal reforming processes or dry reforming processes of biomass-derived organics have been investigated and can be realized over Ni/Al₂O₃ catalysts [311,312,313].

6.3.9. Ni/Al₂O₃ for Chemical Looping Processes

Chemical looping reforming (CLR) technologies imply the separation of the reforming step occurring over an “oxidized” catalyst surface by the reoxidation step of the catalyst using steam, CO₂ or oxygen [314]. Ni/Al₂O₃-based materials have been studied for ethanol, glycerol and methane CLR. In the case of glycerol CLR and in the case of glycerol CL dry reforming, the addition of alkali earth, in particular Sr, improves the performances [315]. NiO/LaAl₁₁O₁₈ was reported to be an excellent material for chemical looping combustion, although its structure was converted into NiO-NiAl₂O₄-αAl₂O₃ in the process [316].

6.3.10. Ni/Alumina Catalysts in Dehydrogenation and Aromatization Reactions

Reduced nickel catalysts have limited application in paraffin dehydrogenation reactions because of the high activity of metallic nickel particles in causing C-C bond cracking and deactivation by carbon deposition at high temperatures [317]. However, early studies showed excellent activities of Ni/Al₂O₃ catalysts for the aromatization of cyclohexanes, a reaction realized at low temperatures, e.g., at 300 °C. [318]. In particular, good activity has been reported for a 20 wt% Ni/Al₂O₃ [319] catalyst and of NiAl₂O₄ [320] in the dehydrogenation of methylcyclohexane to toluene at 380–450 °C. This shows that the aromatization and dearomatization activity of quite cheap Ni/Al₂O₃ catalysts can be interesting in view of the use of the methylcyclohexane/toluene (MCH-TOL) system as a Liquid Organic Hydrogen Carrier (LOHC) [321,322,323], which is already available commercially [324].

A recent work reported that a carefully prepared unreduced 1% Ni/Al₂O₃ catalyst has high activity and selectivity in propane dehydrogenation to propene at 580 °C, where the active site is supposed to be a highly isolated Ni²⁺ ion on alumina [196].

6.4. Applications of Unreduced NiO-Al₂O₃ Catalysts

6.4.1. NiO and NiO-Al₂O₃ as Combustion Catalysts

As already stated, NiO shows the highest catalytic activity for methane combustion, among simple transition metal oxides, with an ignition temperature in the range of 450–500 °C, attributed by reaction with surface oxide ions interacting with Ni³⁺ ions [189]. NiO is considered among the useful catalysts for the combustion of Volatile Organic Compounds (VOCs) [325] such as, e.g., in the total oxidation of toluene and formaldehyde [326]. Also, NiO/Al₂O₃ was reported to have good activity in methanol [327], toluene [328] and isopropanol full oxidation [329].

6.4.2. The Oxidation of Ammonia and Hydrogen

The combustion of hydrogen and the oxidation of ammonia are becoming more and more interesting in view of the large use of hydrogen as a green fuel and of ammonia as a hydrogen transport molecule. Among the non-noble, metal-based catalysts, NiO shows very high activity in hydrogen combustion [330]. Instead, NiO does not seem to have high activity in ammonia oxidation to NO, although some of its mixed oxides do have [331]. Interesting tests have been realized concerning the Ammonia Selective Oxidation (SCO) to nitrogen. Here, 10–20 wt% NiO/Al₂O₃ [328,332,333] and NiOMgO-Al₂O₃ [334] catalysts have high activity in the oxidation of ammonia, with a main production of N₂ (up to >90% yield) at about 550 °C but also some coproduction of NO.

Ni/Al₂O₃ has been tested in the combustion of H₂-NH₃ mixtures and shows the highest activity among tested transition metal catalysts. Over 5%Ni/Al₂O₃, the light-off of both H₂ combustion and NH₃ oxidation, is below 200 °C, with a limited coproduction of N₂O and NO_x [335].

6.4.3. Oxidation of Carbon Monoxide

The low-temperature oxidation of carbon monoxide is an important reaction for treating combustion and automotive flue gases. Bulk and supported NiO are among most active non-noble metal catalytic materials [336,337]. In particular, NiO/Al₂O₃ [338,339] shows good activity and stability. Interestingly, high activity in CO oxidation has been reported for a commercial NiO/Al₂O₃-based steam-reforming catalyst (G-65 from Clariant) whose composition is reported to be 34 wt% NiO on alumina at 180–210 °C with 7–25 kPa CO and 5–15 kPa O₂ partial pressure in nitrogen [340], which is in agreement with the activity of a 10% NiO/γ-Al₂O₃ laboratory catalyst [329].

6.4.4. Oxidative Dehydrogenation of Alkanes

The oxidative dehydrogenation (ODH) of light alkanes is a useful reaction to produce olefins starting from natural-gas-derived paraffins with an exothermic process. A totl of 15 wt% NiO/Al₂O₃ catalysts appear to be very promising for the production of ethylene from ethane ODH [341] at 400 °C with ethylene selectivity of about 95%. The catalyst works in the oxidized state, with the best activity near a monolayer NiO coverage. Similarly, propene can be produced from propane ODH over a NiO/Ca₁₂Al₁₄O₃₃ catalyst at 480–580 °C but with definitely lower selectivity and yield [342].

6.4.5. Ozone Decomposition and Use

NiO and NiO/Al₂O₃ are also reported among the most active ozone decomposition catalysts [329,343], as well as catalysts for ozonation reactions, i.e., the oxidation of CO and VOC with ozone [329].

7. Conclusions

The above review summarizes, hopefully in an almost complete way, the properties of the materials based on nickel and alumina and most of their actual and potential applications. As reported by Yu et al. [344], catalysts belonging to this system may also have a number of other less usual applications. It is evident that nickel and alumina represent together an extremely versatile system that is applicable in reducing and oxidizing conditions, offering stability in very different conditions, and is also cheap. A main drawback of this system is due to the toxicity of several nickel compounds that make cautions necessary in catalyst preparation and for recovery after their end of life.

The usefulness of Ni–alumina systems is due to the properties of nickel to act as a versatile catalytically active phase in many different conditions. A key property of nickel is to be quite easily reducible to metal but also quite easily oxidized to NiO up to Ni³⁺-containing NiO_1+x. Together, the unique ability of alumina to disperse ionic species at its surface in a stable way, without allowing nickel ions to penetrate the bulk in mild conditions, and to interact strongly with surface metal centers is the other key feature of this system. This allows one to produce small metallic particles with high stability, with moderate nickel loading, or larger particles with high loadings. By changing the alumina polymorphs from γ-Al₂O₃ or η-Al₂O₃ to δ-Al₂O₃, θ-Al₂O₃ and α-Al₂O₃, coprecipitating NiAl₂O₄ and/or NiO-NiAl₂O₄-Al₂O₃ solid solutions, or adding alkali-, alkali earth- or rare earth oxides, it is also possible to tune the thermal stability of the system from relatively low temperatures to very high ones and the acido-basicity as well.

Nickel alumina materials may find increased interest in the near future in the frame of hydrogen technologies as hydrogenation and dehydrogenation catalysts. In particular, they can act as optimal catalysts, e.g., for both the hydrogenation and dehydrogenation step in the case of the MCH-TOL system being used as an LOHC or for the decomposition step when ammonia is used as a circular hydrogen carrier.

As shown above, nickel–alumina catalysts can also be used for a number of reactions based on renewables and can be of interest in the energy sector, whose future application will grow significantly. Taking into account that noble metals more active in catalysis, palladium and platinum, in particular, are becoming more and more critical due to their limited availability and increasing applications, alternatives are needed. Although nickel, too, has been defined as a critical metal, its still significant abundance, the relative wide resource distribution and the application of new technologies for extraction from poor resources [345] and for recovery from used devices suggest that the application of nickel in the catalysis field (as well as in the parent electrocatalysis field) may grow very much in the near future. Thus, interest for the nickel–alumina material, very relevant in the past hydrocarbon-based industrial chemistry and energy sectors, will grow further in the future renewable-based industrial chemistry and energy sectors.