Catalytic Transfer Hydrogenation Performance of Magnesium-Doped ZrO₂ Solid Solutions

Abstract

This is the first study to investigate the activity of a solid solution containing magnesium ions in a zirconia matrix in the catalytic transfer hydrogenation (CTH) of acetophenone with 2-pentanol. The results have shown that magnesium oxide is very highly active in CTH when physically mixed with zirconia. However, the same concentration of Mg²⁺ ions (Mg:Zr = 3:97) inserted into a zirconia lattice did not yield high activity in CTH. A higher concentration of Mg²⁺ ions (5%) was also tested in the two types of systems, i.e., a physical mixture of oxides and a solid solution. The increase in the concentration of Mg²⁺ ions in the physical mixture led to a pronounced increase in the activity of the system, whereas in the case of the solid solution it led to a slight decrease in activity. The impact of the zirconyl salt used in the synthesis was also examined, but showed little effect on the properties and activity of the systems. The study has also shown that the increase of the concentration of magnesium ions caused a decrease in the m-ZrO₂ to t-ZrO₂ ratio. Nevertheless, the rate of heating had an even bigger effect on this ratio.

1. Introduction

It is commonly known that zirconia can exhibit one of three structures: monoclinic, tetragonal and cubic, and that the stabilization of the latter two is common in both the ceramic industry (including fuel cell applications and dentistry purposes) and the jewelry industry, but has been scarcely explored in catalysis. The significance of the effect of doping of different systems on their properties is well established in e.g., metallurgy, where the benefits of doping steel with carbon are well known. However, recently, the impact of doping on other properties has been gaining interest, such as influencing the electrochemical properties of an oxide [1,2], the improvement of hole transport across an oxide layer [3] or even enhancing the photocatalytic performance of oxide systems by doping [4,5]. The main reason for doping zirconia is to stabilize either the tetragonal or cubic phase, which, without stabilization, would transform into monoclinic zirconia. Structural stability is also a benefit in catalysis. The two most commonly used ions are Mg²⁺ and Y³⁺ [6,7,8,9]. For Mg²⁺, this is known as “magnesia stabilized zirconia” and is typically obtained by the “skull-melting” technique, which employs high temperatures (1400 °C) and a subsequent cooling of the system once the dopant had been added. Since such a system contains magnesia as a separate phase, it could exhibit some of the properties typical for magnesia, including catalytic activity typical for this highly basic oxide. Hence, the motivation for the study was to determine whether the zirconia stabilized with Mg²⁺ ions would behave similarly to pure MgO physically mixed with zirconia in a selected catalytic reaction or not. The hypothesis was that the change in the environment of Mg²⁺ ions, e.g., an increase in the nearest Mg²⁺-O²⁻ distance, would result in activity similar to that of undoped zirconia rather than the high activity of MgO. There are different views in the literature regarding the phase diagram of a system containing Mg, Zr and O [10]. When synthesis of stabilized/partially stabilized zirconia is carried out at such high temperatures, the predominant phases are tetragonal or cubic zirconia, although monoclinic zirconia can be detected depending on the concentration of the dopant [6,10]. However, when zirconia is prepared for catalytic applications, either as a support or a catalyst, the oxide should be porous, unlike those used for e.g., dental purposes, in order to exhibit a higher surface area. The papers published on such low-temperature synthesis of mixed Zr–Mg systems include the use of soluble zirconium salts, such as zirconyl nitrate, zirconyl chloride or zirconium chloride [11,12,13,14] and magnesium nitrate or chloride, which are co-precipitated using ammonia water or sodium hydroxide. These are calcined at temperatures between 500 and 700 °C to obtain the oxide(s) or solid solution. There is some debate as to whether the stabilized form of zirconia is tetragonal or cubic [15,16,17,18,19], which is most likely due to the fact that the diffraction patters of these two phases are almost identical, differing only by the fact that, instead of two symmetrical single peaks in the diffraction pattern of the former, two asymmetric peaks are seen in the same place in the diffraction pattern of the latter: one between 34 and 35 degrees, the other at approx. 59 degrees. The asymmetry is caused by the overlapping of peaks from (0 0 2) and (1 1 0) planes, as well as those from (1 0 3) and (2 1 1) planes.

Magnesia has been applied in numerous reactions [20,21,22,23], including hydrogenation reactions, Fatty Acid Methyl Ester synthesis, double bond isomerization and catalytic transfer hydrogenation. ZrO₂ has also been applied as a catalyst, either alone in hydrogenation of carboxylic acids [24] or with other metals, such as nickel [25] in the hydrogenation of CO₂ or copper as a part of Cu_xZr_y oxide systems used for hydrogenation of CO and CO₂ [26], although it is more commonly applied as a support [27,28,29,30] for metals such as silver, nickel, cobalt, etc. Zirconia has also been applied as a catalyst in biomass conversion [31]. However, zirconia is most known for being a solid electrolyte for solid oxide fuels (SOFC) [32,33,34,35], especially in the form of yttria-stabilized zirconia, due to its high ionic conductivity and high chemical stability at elevated temperatures. Chemical stability, however, is limited when it comes to actual SOFCs, in which YSZ is the electrolyte and lanthanum strontium manganite is the cathode material. There have been multiple studies regarding the interaction of these two SOFC components [36,37,38,39] that indicate a migration of some of the electrode components into the electrolyte.

The excellent catalytic activity of MgO in catalytic transfer hydrogenation may be influenced by several factors, especially its purity [40]. Our previously published results indicate that the presence of impurities such as chloride or sulphate ions can lead to pronounced changes in the surface area of the obtained oxide, not to mention the acidity/basicity of the surface [40]. Taking this into account, special precautions need to be taken to eliminate chloride ions when they are present in synthesis as counter ions of either Mg²⁺ or Zr⁴⁺. Therefore, one of the factors investigated in the present research was the influence of the zirconium salt used as the precursor. Since using potassium or sodium hydroxide as the precipitating agent also leads to influencing the properties of the obtained magnesia [40], only ammonia water was used for the precipitation of the hydroxides/hydroxide mixture in this study. The two loadings of Mg²⁺ ions were chosen to be high enough for the peaks from MgO to be visible in diffraction patterns in such a concentration, but small enough so that all of the Mg²⁺ ions could be incorporated into the zirconia lattice. The expectation was that if the co-precipitated samples contained magnesia (MgO as a separate phase), it would be evidenced by the enhanced activity of the obtained zirconia in catalytic transfer hydrogenation of acetophenone with 2-pentanol, whereas if the Mg²⁺ ions were dispersed in the zirconia lattice, the activity of these systems would not be as active as physical mixtures containing the same concentration of Mg²⁺ ions in the form of MgO because their activity would only slightly modify the activity of undoped zirconia.

2. Results

Two types of catalytic systems were prepared: the first, physical MgO + ZrO₂ mixtures, and the second, solid solutions in which zirconia was doped with Mg²⁺ ions. Both contained the same concentration of Mg²⁺ ions. It can be seen in the Energy Dispersive X-Ray Spectroscopy elemental maps that the former (Figure 1a) exhibit separate domains of ZrO₂ and MgO. In contrast, the solid solutions with either 3 or 5% Mg²⁺ have a homogeneous composition, and the Mg²⁺ is uniformly distributed throughout (Figure 1b,c). This is a sharp contrast to the magnesia–zirconia and yttria–zirconia calcined at much higher temperatures that allowed the formation of separate phases. Figure 2 shows that pure ZrO₂ obtained from the commercial zirconyl nitrate (ZN) via precipitation with an aqueous ammonia solution, drying, heating at a rate of 15 K/min and calcining at 550 °C for 4 h contains two zirconia polymorphs (Figure 2a), namely monoclinic zirconia (m-ZrO₂; PDF: #861451) and tetragonal zirconia (t-ZrO₂; PDF: #800965). The two most dominant signals from the former appear at 2θ values of 28.22° (−1 1 1) and 31.51° (1 1 1), whereas the main signal from the latter is located at 30.27° (1 0 1). This is a similar result to that presented by Teterycz et al. [13] and Hill et al. [41], in which a major contribution of the monoclinic ZrO₂ phase was detected in the undoped sample. In contrast, this was not the case for Aramendia et al. [12], where the precipitated, dried and calcined hydroxide yielded a single-phase tetragonal zirconia. This result, however, may be due to the presence of chloride ions, which had not been completely washed away. This can be deduced from the fact that the MgO which was obtained in a similar way had a maximum decomposition temperature of 645 K, which is lower than that of very pure magnesia [40] and, hence, indicates that some impurities were present. The diffraction pattern of MgO shows only one phase, namely the cubic NaCl type structure, PDF: #711176 (Figure 2b). As expected, signals from all three phases, i.e., m-ZrO₂, t-ZrO₂ and MgO, are present in the diffraction patterns of the samples where the two oxides were physically mixed together (Figure 2c,d). It is noteworthy that even the mixture containing only 3% MgO shows the two most pronounced signals from MgO (marked with a red arrow and an asterisk).

It can be seen in Figure 3a and b that pure ZrO₂ obtained from the commercial zirconyl nitrate (ZN) and zirconyl chloride (ZC) via precipitation with an aqueous ammonia solution, which was heated at the faster heating rate contains both the monoclinic and tetragonal ZrO₂ phases. The arrows point to the most pronounced signals from each phase: the blue arrow corresponds to the (−1 1 1) plane of the monoclinic phase, whereas the red arrow corresponds to the signal from the (1 1 1) plane of the tetragonal phase. From the ratio of the heights of these signals, it can be seen that both diffraction patterns are very similar, though the latter has a lower m-ZrO₂ to t-ZrO₂ ratio, i.e., 12.5:1 (Figure 3a) vs. 8.8:1 (Figure 3b). The impact of a slower heating rate on the m-ZrO₂/t-ZrO₂ ratio was also explored. It can be seen that the zirconia made from zirconyl nitrate with a slow heating program contains much less of the monoclinic phase than that obtained with fast heating, i.e., 1 to 4.5 (Figure 3d). According to the literature, this is a metastable phase. Considering the visible asymmetry of the peaks at around 35 and 59 deg., it can be seen that the solid solution is tetragonal. The fact that this is a metastable phase has been demonstrated by Rong et al., who showed that even milling a co-precipitated and calcined Mg-doped tetragonal zirconia for 1 h can lead to the formation of a substantial amount of the monoclinic phase [42].

The solid solution containing 5% Mg²⁺ ions obtained with fast heating contains relatively less of the monoclinic phase than the undoped zirconia, though the m-ZrO₂:t-ZrO₂ is still high at 6.0 to 1 (Figure 3c). In contrast, the solid solutions formed with slow heating show that the insertion of 3% Mg²⁺ leads to a much smaller relative concentration (1 to 4.8) of the monoclinic zirconia (Figure 3e), whereas the slightly higher magnesium loading leads to full stabilization of the tetragonal zirconia and none of the monoclinic phase (Figure 3f). In both cases, the aforementioned peaks are asymmetrical and, hence, all of the formed solutions are considered tetragonal zirconia. A single tetragonal phase has previously been reported by Teterycz et al. [13] and Wang et al. [14], though in [12] there are two separate phases, namely MgO and tetragonal ZrO₂ in the sample which had also been co-precipitated under similar conditions. The difference in the results in the latter paper might have been caused by either or both of two things: (1) the precipitating agent was not ammonia water, but NaOH, and (2) the chloride ions from the zirconyl chloride might not have been completely eliminated. In the case of our studies, it cannot be ruled out on the basis of XRD alone that finely dispersed or amorphous magnesia does not exist on the surface, but the small loadings of magnesium were chosen in order to decrease the chances of the formation of a separate magnesium oxide phase. Moreover, the same amount of magnesium ions in the form of magnesia that was physically mixed with zirconia do produce XRD peaks of this phase in the diffraction pattern (Figure 2). The only difference observed between the tetragonal phase of pure zirconia and that in Zr_0.95Mg_0.05O_x are the lattice parameters; namely, for the former they are a = 3.5956 and c = 5.1885 Å, whereas for the latter they are a = 3.5979 and c = 5.1339 Å. Considering the fact that Mg²⁺ ions have the same radius as Zr⁴⁺ ions, the slight increase in parameter a of the unit cell (Δa = 0.0023 Å) and a substantial decrease of parameter c (Δc = 0.0546 Å) can be attributed to the oxygen vacancies forming as a result of Mg²⁺ ion incorporation into the zirconia lattice.

In situ XRD studies were carried out by heating the dry precipitate in 20 °C increments and using a holding time between each temperature long enough to acquire a diffraction pattern. It can be seen that such a slow heating rate results in a very high abundance of the tetragonal phase for both the undoped (Figure 4a), i.e., 1:5.9, and the Mg²⁺-doped zirconia (Figure 4b). In fact, in the case of the Mg-doped sample, the tetragonal phase is the only phase present in the diffraction pattern. It is noteworthy that they both start out as amorphous hydroxides. In the case of the undoped ZrO₂, both monoclinic and tetragonal zirconia appear at the same temperature, namely 300 °C, whereas the presence of the magnesium ions appears to shift the temperature to 320 °C. No further changes are observed in the heating range up to 420 °C.

The texture of the samples (Figure 5) is very similar to that of the previously published solid solutions prepared by us using a similar procedure, such as zirconium-doped ceria [43] and cerium-doped zirconia [44]. However, unlike for pure zirconia (Figure 5a) or zirconia doped with 5% Ce ions (Figure 5d), which fracture in large, irregular chunks using a mortar and pestle, both Zr_0.97Mg_0.03O_x (Figure 5b) and Zr_0.95Mg_0.05O_x (Figure 5c) show elongated structures with unusual fracture. It is not clear why these samples exhibit such fracture despite the presence of the stabilized tetragonal zirconia phase, which is also present in the Ce-doped samples. Both Mg²⁺ and Zr⁴⁺ have the same ionic radius, so it cannot be attributed to a change of this parameter. However, it could be due to the fact that magnesium ions have the +2 oxidation state, whereas zirconium is +4; oxygen vacancies form as a consequence of doping zirconia with magnesium ions. They might be responsible for the fracture of these materials, but more studies are needed to confirm or disprove this hypothesis.

The physicochemical properties of the studied systems are collected in

Table 1. Properties of selected oxides: MgO and zirconia-based systems (with 0, 3 and 5% Mg) obtained from zirconyl nitrate with slow calcination.

Sample	MgO	ZrO2	Zr0.97Mg0.03Ox	Zr0.95Mg0.05Ox
SSA [m2/g]	130	51	60	75
Pore diameter [Å]	25	70	55	40
CO2 desorption [cm3/g]
weak (<200 °C)	4.51	2.14	2.65	3.09
medium (200–400 °C)	5.66	2.24	2.49	3.09
Strong (>400 °C)	2.97	1.79	2.04	1.34
Total	13.14	6.17	7.18	7.52

. The surface area is the highest for pure magnesia, i.e., 130 m²/g, and the smallest for pure zirconia, i.e., 51 m²/g. The surface areas of the magnesium-doped zirconia samples are 60 m²/g and 75 m²/g for the solid solutions obtained from zirconyl nitrate and magnesium nitrate by co-precipitation with the heating rate of 15 K/min containing 3% and 5% Mg²⁺, respectively. This shows that the increase in Mg ions incorporated into the zirconia structure leads to an increase in the porosity. The average pore diameter is the smallest for pure MgO (Table 1), at 25 Å. It is the largest for ZrO₂, at 70 Å. The incorporation of 3% magnesium into the zirconia lattice leads to a decrease of the pore size by 21%. The increase of the magnesium content to 5% leads to a further decrease of the average pore diameter by another 21% (Table 1).

The BJH isotherms of pure ZrO₂ and MgO are presented in Figure 5a,b. It can be seen that the quantity of adsorbed nitrogen is much higher for the latter oxide, as the y-axis is twice as large for magnesia as for zirconia. The pore distribution (inserts) is also substantially different: in the case of zirconia, the majority of the pores have a diameter of 50–100 Å, whereas in magnesia they are predominantly below 50 Å in diameter. The isotherms obtained for the solid solutions with Mg²⁺ ions incorporated into the ZrO₂ matrix are shown in Figure 6c,d. It can be seen that they differ from the results of the pure phases, but the y-axis is the same as that for pure zirconia. The change in the magnesium ions content has a noticeable influence on the pore shape in the resulting materials. In the case of the higher magnesium ion content, the hysteresis loop is much wider than in the case of the lower one. Such a change indicates that Zr_0.97Mg_0.03O_x has an H1 type hysteresis with “spherical or cylindrical narrow mesopores” [45], whereas the isotherm of Zr_0.95Mg_0.05O_x resembles the H2 hysteresis loop whose shape is “common for inorganic oxides with a complex network of interconnected narrow pores”, which are characterized by so-called “ink bottle pores” [45].

One of the key parameters that makes MgO an excellent catalyst in transfer hydrogenation is the basicity of its surface. There are many methods of assessing the strength or concentration of basic sites found on the surface of a solid oxide. Some methods can assess only the former, e.g., the use of indicators [46], whereas other techniques can assess only the latter, e.g., titration with benzoic acid [46]. However, Temperature-Programmed Desorption of carbon dioxide can be used to determine both the strength of the sites, which is indicated by the temperature of the desorption peak, and the concentration of the basic sites, which can be obtained by integration of the area under the desorption peak [17,28,30,47,48,49]. It can be seen that zirconia has two main maxima: one at 106 °C, the other at 347 °C, after which it plateaus at approx. 500 °C and drops slowly but steadily above 600 °C (Figure 7a). Upon the addition of 3% Mg²⁺ ions, the desorption profile of the solid solution closely resembles that of the undoped zirconia (Figure 7b), with two small changes: the first maximum is larger (the same scale of the TCD signal is used for both profiles), and, secondly, the plateau of the desorption signal lasts until about 50 °C more than in the case of pure ZrO₂. García et al. [17] also found that the incorporation of Mg ions into the lattice of ZrO₂ increased the basicity of this oxide, which was applied as the support of nickel catalysts for methane reforming of carbon dioxide. Pure MgO exhibits a significantly different desorption profile than the zirconia-based systems (Figure 7d). In fact, MgO has much stronger basic sites, as seen in the large maximum. However, the strongest basic sites are not expected to take part in catalytic transfer hydrogenation, which has been shown in poisoning studies with phenol as the active site poison [50].

The amount of CO₂ desorbed from the samples has been categorized into desorption from one of three types of basic surface sites based on the temperature of desorption: CO₂ desorbing below 200 °C was bound to weak basic sites, CO₂desorbing between 200 and 400 °C was associated with medium-strength basic sites, and the CO₂ desorbing above 400 °C was bound to strong basic sites. Magnesia has the most of all three types of sites and much more basic sites overall (13.14 cm³/g). Zirconia has the least basic sites of the four tested samples. It can be seen that the main change upon the incorporation of the magnesium ions into the zirconia lattice is an increase in the first (low basic) desorption peak (Figure 7; Table 1). In a study where MgO and ZrO₂, obtained by calcining zirconium hydroxide, were both used [41], the authors determined that “MgO exhibited activated formation of surface carbonates from bicarbonate species, whereas ZrO₂ revealed the formation of a wide range of nonactivated bicarbonates and carbonates on basic sites”. The different nature of these basic sites might explain the different activity of the two oxides. The study by Teterycz et al. on the conversion of n-butyl alcohol condensation showed that the magnesia exhibited low conversion, whereas zirconia and the magnesium-doped zirconia have very high conversion [13].

The reaction of catalytic hydrogen transfer from 2-pentanol to acetophenone was chosen due to the very high activity and selectivity of MgO in this reaction [22]. Sample chromatograms after 1 and 3 h of reaction using a physical mixture containing 5% MgO and of Zr_0.95Mg_0.05O_x are provided in Supplementary Materials (Figure S1) and the peak areas used for calculations are provided in Table S1. This high activity can also be seen in Figure 8 when physical mixtures of MgO and ZrO₂ are used as the catalyst (Figure 8a). In the case of physical mixtures, the increase in the concentration of MgO from 3 to 5 at.% leads to a very large increase in the yield of the product, namely 1-phenylethanol, with 61% after only 3 h of reaction. In contrast, the activity of the solid solutions that contain magnesium trapped in a zirconia matrix is much smaller than that of mixtures containing the same amount of Mg²⁺ ions in the form of MgO physically mixed with ZrO₂. Moreover, the increase in the Mg²⁺ concentration has a slightly negative effect (Figure 8b), with only 12 and 8% yield of 1-phenylethanol after 3 h of reaction for the solid solution with 3 at.% and 5 at.% Mg²⁺, respectively. A change in the precursor zirconium salt from zirconyl nitrate to zirconyl chloride does not impact the catalytic activity of the mixture substantially. Considering the results of the activity studies and CO₂ TPD, it can be inferred that it is not the weak basic sites that determine the activity of the catalyst in CTH, because the concentration of weak basic sites increases in the order ZrO₂ > Zr_0.97Mg_0.03O_x > Zr_0.95Mg_0.05O_x, whereas the activity is the lowest for Zr_0.95Mg_0.05O_x. Since both finely dispersed and amorphous magnesia exhibit high activity in catalytic transfer hydrogenation [51] and no improvement in the activity of these systems in respect to the pure zirconia phase has been observed, it can therefore be deduced that there is no magnesia in these samples.

3. Materials and Methods

3.1. Synthesis of Samples

MgO was synthesized as described by us earlier [40]. In short, it was obtained from magnesium nitrate (puriss, p.a., Fluka Chemie GmbH, Buchs, Switzerland) dissolved in redistilled water and precipitated with an ammonia–water solution (25%, p.a., POCh, Gliwice, Poland) in three steps, with filtration between the individual steps. The first two precipitates were discarded, and the third was filtered, washed, dried, fractioned (0.16–0.40 mm) and used to obtain pure MgO via calcination of the hydroxide in a muffle furnace at 600 °C for 6 h.

ZrO₂ was synthesized using one of two precursors: zirconyl chloride, i.e., ZrOCl₂·8 H₂O (analytical grade, POCh Gliwice, Gliwice, Poland), or zirconyl nitrate, i.e., ZrO(NO₃)₂·2 H₂O (analytical grade, POCh Gliwice, Gliwice, Poland). In each case, the zirconyl salt was dissolved in redistilled water and zirconium hydroxide was precipitated with an ammonia–water solution (25%, p.a., POCh, Gliwice, Poland). The gel was filtered, washed with redistilled water, dried and calcined in a muffle furnace with a temperature ramp of 15 K/min to 550 °C (fast heating) or with 15 K/min to 400 °C, an hour at 400 °C, followed by 15 K/min to 550 °C (slow heating). In both cases, the oxides were calcined at 550 °C for 4 h. The obtained solid was crushed and sieved to obtain the 0.230–0.500 mm fraction that was used for the catalytic tests.

Solid solutions were prepared using either of the two zirconyl salts and Mg(NO₃)₂·6 H₂O (pure p.a., POCh Gliwice, Gliwice, Poland) in the appropriate amounts. For instance, the Zr_0.95Mg_0.05O_x solid solution from zirconyl chloride was prepared using 35 g of the zirconyl salt and 1.47 g of magnesium nitrate. They were dissolved in 400 mL of redistilled water, stirred under mixing using a magnetic stirrer. They were then precipitated using 50 mL of an ammonia water solution (25%, p.a., POCh, Gliwice, Poland), washed with redistilled water until the pH was neutral. The precipitate was dried and calcined in accordance with one of the abovementioned procedures.

3.2. Characterization

The obtained solids were imaged using a Secondary Emission Microscope (SEM) coupled with Energy Dispersive X-Ray Spectroscopy (EDX) with Prisma E from FEI (Field Electron and Ion Company, FEI, Hillsboro, OR, USA) with a working distance (WD) of 10 mm, 3.5 spot size and energy of 10 kV. Magnifications of 500, 1000, 2000, 5000, 10,000 and 20,000 times were applied. The elemental maps were acquired with the following parameter: 15 kV, spot size 6. Higher resolution images were recorded with Helios 5 DualBeam FIB-SEM (Thermo Scientific, Oxford, UK) with the following parameters: 3 kV voltage, 13 pA.

Nitrogen physisorption determinations (ASAP2020, Micromeritics Instrument Co., Norcross, GA, USA) were used to obtain both the specific surface area and pore distribution. The samples were degassed (50 °C, 2 h), the temperature was lowered to 77 K and the pressure was varied in the range of p/p₀ = 0–1. The BET (Brunauer–Emmett–Teller) model was used to calculate the S_BET values, whereas the hysteresis loops were drawn using the BJH (Barrett–Joyner–Halenda) method.

The carbon dioxide Temperature-Programmed Desorption (TPD) studies were carried out using an Autochem 2910 instrument (Micromeritics Instrument Corp., Norcross, GA, USA) with a thermal conductivity detector. Prior to the measurements, the sample was treated using the following procedure: heating the sample in a flowing He stream (5N; Multax, Stare Babice, Poland) to 500 °C with a flow rate of 40 mL/min, a 1 h hold at that temperature; cooling, adsorption of CO₂ (5N, Multax, Stare Babice, Poland) at 40 °C in a He/CO₂ stream (9:1 vol. ratio) with a flow rate of 40 mL/min, a hold time of 2 h; flushing with a He stream at 40 °C with a flow rate of 40 mL/min for 1 h. The Temperature Programmed Desorption of CO₂ was performed while heating the sample to 800 °C, 10 K/min in a He stream with a flow rate of 40 mL/min.

Both ex situ and in situ XRD measurements were performed in the scattering range 15–140 deg. with a 0.02 deg/s step, though, for clarity, some of the diffraction patterns are presented only in the 15–95 deg range. They were obtained using a Siemens D 5000 instrument (Bruker AXS GmbH, Karlsruhe, Germany). The measurements were performed with a Cu tube (40 kV, 40 mA), Ni filter and LynxEye detector, in a Bragg–Brentano geometry. For the in situ measurements, an environmental chamber was used. The details are described in [52]. Prior to each measurement, the sample was heated in 20 °C steps from room temperature to 420 °C. Between the measurements, the analyzed sample was held in a flowing He (5N, Multax, Stare Babice, Poland) stream with a flow rate of 20 mL/min. The diffraction patterns were used to calculate the relative ratio of the monoclinic and tetragonal phases using the RIR ratio of I(T)/I(M) = 2.069.

3.3. Activity Tests

The activity of the oxides in catalytic transfer hydrogenation of acetophenone (analytical grade, Sigma-Aldrich, St. Gallen, Switzerland; distilled) with 2-pentanol (analytical grade, Sigma-Aldrich, St. Gallen, Switzerland; distilled) with an 8:1 D:A ratio were performed using a glass reactor fused with a condenser. It was equipped with a side arm. Prior to experiments, approx. 0.2 g of each solid was used in the liquid phase CTH reaction. The reagents were dosed onto the catalyst in the following order: first the donor, then the acceptor. Next, the reactor was submerged in an oil bath preheated to approx. 140 °C. The temperature was maintained with samples taken after 30 min of reaction, as well as after 1, 2 and 3 h. The conversion and selectivity were monitored using gas chromatography (Agilent 6890N, Waldbronn, Germany) with a WAX capillary column. The chromatography program was 5 min at 120 °C, heating to 220 °C at a rate of 10 K/min., followed by an isothermal segment at 220 °C. There were either 4 or 5 signals in each chromatogram. The peaks were quantified using a five-point calibration curve for each compound. The conversion was calculated using the peak areas of acetophenone (t_R = 8.207 min), 1-phenylethanol (t_R = 10.359 min) and styrene (t_R = 3.060 min), if present. Since the donor was used in a large excess (D:A = 8:1), neither the 2-pentanol peak (t_R = 2.276 min) nor the peak of its oxidation product, 2-pentanone (t_R = 2.052 min), was used for quantification. The relative peak ratios of acetophenone, (1-phenylethanol) and styrene, as determined from the calibration curves, were 1, 1.10 and 0.62, respectively.

4. Conclusions

The activity of two types of solids containing magnesium ions in the form of a magnesia–zirconia physical mixture, as well as in a solid solution of zirconia, were tested in catalytic transfer hydrogenation (CTH) from 2-pentanol to acetophenone. The activity of magnesia in such reactions is well-documented, whereas those of the solid solutions have not been described yet. The results have clearly shown that the same concentration of magnesium ions leads to two drastically different effects, namely high activity of the physical mixtures, with the expected improvement of activity upon the increase in the MgO:ZrO₂ ratio and a lack of improvement of activity of zirconia upon the incorporation of magnesium ions into its lattice in the case of the solid solutions. In contrast to the result observed for the physical mixture, the increase in the concentration of Mg²⁺ ions in the solid solution led to a slight decrease of activity. The change of the zirconyl salt used in the synthesis exhibited little effect on the properties and activity of the systems, though it impacted parameters such as surface area, pore size/shape and the basicity of the surface. It was found that the incorporation of Mg²⁺ into the zirconia lattice had a slight effect on the m-ZrO₂:t-ZrO₂ ratio compared to the undoped system. However, the rate of heating had a much more pronounced effect on the ratio. The in situ XRD studies showed that the precipitated zirconium hydroxide, as well as the co-precipitated samples, are amorphous and that both zirconia phases form at the same temperature and that this ratio does not change when the temperature is slowly raised to 420 °C.