The Catalytic Performance of Metal-Oxide-Based Catalysts in the Synthesis of Glycerol Carbonate: Toward the Green Valorization of Glycerol

Abstract

The rising concern over carbon dioxide (CO₂) emissions has led to increased research on its conversion into value-added chemicals. Glycerol carbonate (GC), a versatile and eco-friendly compound, can be synthesized via the catalytic carbonylation of glycerol with CO₂. This study investigates the catalytic performance of three novel mixed metal oxide catalysts, Ti-Al-Mg, Ti-Cr-Mg, and Ti-Fe-Mg, synthesized via co-precipitation. The catalysts were characterized using XRD, SEM, XPS, CO₂-TPD, FTIR, TGA-DTG, and nitrogen adsorption–desorption isotherms. Among the tested systems, Ti-Al-Mg demonstrated the highest surface area, optimal porosity, and a balanced acid–base profile, resulting in superior catalytic activity. Under optimized conditions (175 °C, 10 bar CO₂, 4 h), Ti-Al-Mg achieved a maximum GC yield of 36.1%, outperforming Ti-Cr-Mg and Ti-Fe-Mg. The improved performance was attributed to the synergistic effects of its physicochemical properties, including high magnesium content and lower CO₂ binding energy, which favored CO₂ activation and glycerol conversion while minimizing side reactions. These findings highlight the potential of tailored mixed metal oxide systems for efficient CO₂ immobilization and sustainable glycerol valorization.

1. Introduction

Carbon dioxide (CO₂) is recognized as one of the main greenhouse gases due to its significant and lasting impact on global climate change. In 2024, the atmospheric concentration of CO₂ reached approximately 422.5 parts per million (ppm), reflecting an increase of about 3 ppm compared to 2023 and a 50% rise over pre-industrial levels [1]. As a result, developing effective CO₂ mitigation strategies has become essential, including its capture and use as both an extraction solvent [2] and a feedstock for producing value-added chemicals such as methanol, methane, mono-, di-, and tri-acetone, and 1,2-propanediol [3,4]. Among these, glycerol carbonate (GC) stands out as a particularly valuable compound, not only for its industrial applications but also for its strategic role in valorizing bio-based glycerol [5].

To this end, glycerol carbonate is typically synthesized through the carbonylation of glycerol with CO₂ in the presence of heterogeneous catalytic systems, mainly based on metal oxides, due to their ability to adsorb CO₂ on their surfaces. Heterogeneous catalysis offers significant advantages in this process, particularly in terms of catalyst separation and recyclability [6]. However, direct carbonylation of glycerol with carbon dioxide is very challenging due to thermodynamic limitations. The chemical reaction equilibrium constant is quite small, and increasing pressure is favorable to the chemical equilibrium [7]. In addition to raising the pressure, extracting water from the system drives the equilibrium in favor of GC formation, resulting in higher yields [8].

Gao et al. [5] reported the use of zeolite-based ETS-10 titanosilicate catalysts impregnated with active transition metals (Cu, Zn, Ni, Zr, Ce, Fe) for the efficient production of GC. Under reaction conditions at 170 °C, using Co/ETS-10 in the presence of CH₃CN as a dehydrating agent, a glycerol conversion of 35.0% and a GC yield of 12.7% were achieved. Similarly, Yi-Hu Ke et al. [9] prepared CuO, NiO, Co₃O₄, ZrO₂, and Al₂O₃ via a hydrothermal method. Among these, CuO showed the best performance, achieving 89.0% glycerol conversion and 69.4% selectivity toward glycerol carbonate at 120 °C and 3.0 MPa CO₂ over 5 h. In a separate study, Al-Kurdhani et al. [10] used a CuO/Al₂O₃ catalyst prepared by impregnating activated alumina with an aqueous solution of Cu(NO₃)₂·3H₂O. The 30% CuO/Al₂O₃ catalyst, calcined at 700 °C, showed notable catalytic activity, reaching 41.3% glycerol conversion and a 17.5% GC yield at 150 °C and 4.0 MPa CO₂ over 5 h. 2-Cyanopyridine was employed as both a dehydrating agent and a co-catalyst to enhance CO₂ activation.

In addition, glycerol carbonation with CO₂ was also investigated using various metal oxides (Fe, Zn, La, Ce, and Sn) prepared via the sol–gel method at 180 °C and 150 bar. ZnO showed the highest performance, yielding 8.1% glycerol carbonate after 12 h of reaction [11].

Furthermore, the use of mixed metal oxide catalysts has been shown to increase the quantity and activity of moderate basic sites. Hongguang Li et al. [12] synthesized a series of La/Zn mixed oxide catalysts with varying molar ratios. The La₂O₂CO₃ catalyst containing ZnO at an La/Zn atomic ratio of 0.25, calcined at 500 °C, achieved 30.3% glycerol conversion and a 14.3% GC yield. The improved performance was attributed to ZnO’s role in increasing the concentration of moderate basic sites, facilitating the activation of both glycerol and CO₂.

In a related study, Yajin Li et al. [13] developed a series of xLa₂O₂CO₃-ZnO catalysts via hydrothermal synthesis to convert glycerol and CO₂ into glycerol carbonate through photothermal catalysis. The synergistic effect of light and heat, along with the interaction between ZnO and La₂O₂CO₃, enhanced catalytic performance. The 20% La₂O₂CO₃-ZnO catalyst achieved 6.9% glycerol conversion under reaction conditions of 150 °C, 5.5 MPa CO₂, and 6 h.

Moreover, cerium–zirconium mixed oxide catalysts with varying Zr doping levels of Ce_(1-x)Zr_xO₂ (x = 0–1) [14] were studied to assess the effect of Zr incorporation on the catalytic activity. Increasing the Zr molar fraction from x = 0 to x = 0.02 significantly improved glycerol conversion (from 25.5% to 40.9%) and GC yield (from 21.4% to 36.3%). However, further zirconium addition reduced catalytic performance, with conversion and yield dropping to 25.9% and 21.7%, respectively, at x = 0.2.

Finally, Hongguang Li et al. [15] synthesized a series of Zn/La/Al/M catalysts (M = Li, Mg, Zr) derived from hydrotalcite-like compounds via co-precipitation. Incorporating Li, Mg, and Zr significantly enhanced catalytic activity. Under optimized conditions (7.0 mL CH₃CN, 0.14 g catalyst, 170 °C, 14 h, CO₂ pressure = 6.0 MPa), the Zn/La/Al/Li catalyst achieved a glycerol conversion of 39.5% and a glycerol carbonate yield of 18.7%.

In this study, three mixed metal oxide catalysts, Ti-Al-Mg, Ti-Cr-Mg, and Ti-Fe-Mg, were synthesized via co-precipitation using the same metal ratio of 2:1:1 to limit the number of variables, and evaluated for their performance in converting glycerol and CO₂ into glycerol carbonate in an attempt to tune the acid–base functions of the catalytic surface. The deliberate selection of Al, Cr, and Fe as the third component in the Ti-Mg-Me oxide systems constitutes a distinctive element of our research. These trivalent metal ions were specifically chosen for their capacity to modify both the acid–base characteristics and the electronic environment around magnesium on the catalyst surface. Reports in the literature have demonstrated that incorporation of such ions into mixed metal oxides can significantly modulate surface acidity/basicity, and the distribution of active sites. Specifically, Al³⁺ is widely reported to reduce surface acidity and enhance basicity, creating a favorable environment for CO₂ adsorption and glycerol activation, while minimizing undesired side reactions [16]. Cr³⁺ and Fe³⁺, in contrast, tend to increase Lewis acidity, which may either promote or hinder catalytic performance, depending on their dispersion and interactions within the oxide matrix [17,18]. By integrating these ions into the Ti-Mg matrix, our aim was to fine-tune the physicochemical properties of the catalysts and systematically evaluate their influence on glycerol carbonate synthesis. Structural and surface characterizations (XRD, SEM, BET, XPS, TPD) revealed that, among the synthesized catalysts, Ti-Al-Mg had the highest surface area, well-developed porosity, and optimal acid–base site distribution. Among the catalysts tested, Ti-Al-Mg demonstrated superior catalytic performance, reflected in the highest glycerol carbonate yield of 36.1% at 175 °C and 10 bar CO₂ over 4 h, which was attributed to its homogeneous structure and well-balanced acid–base properties.

2. Catalyst Characterization Techniques

The catalysts were thoroughly characterized using various analytical techniques, including X-ray diffraction (XRD), scanning electron microscopy (SEM), thermogravimetric analysis (TGA-DTG), textural analysis (N₂ adsorption/desorption), X-ray photoelectron spectroscopy (XPS), and Fourier-transform infrared spectroscopy (FTIR).

XRD measurements were conducted at room temperature using a Bruker X-ray diffractometer (Bruker, Karlsruhe, Germany; θ-θ type) equipped with a Cu-Kα radiation source (λ = 1.5418 Å) operating at 40 kV and 5 mA. The scans covered a 2θ range of 5–80° at a rate of 10°/min.

The microstructural morphology was examined using a scanning electron microscope (SEM) from FEI Company (Hillsboro, OR, USA).

Thermogravimetric and derivative thermogravimetric analyses (TGA-DTG) were performed on a METTLER TOLEDO TGA/IST Thermal Analysis System (Mettler Toledo, Nänikon, Switzerland) within the temperature range of 25–600 °C under a nitrogen atmosphere with a heating rate of 10 °C/min. The nitrogen adsorption–desorption isotherms were recorded at 77.35 K across a relative pressure (p/p0) range of 0.005 to 1.0 using a Quantachrome Nova 2200e instrument (Boynton Beach, FL, USA). Data processing was carried out using the NovaWin version 11.03 software. The specific surface area was calculated using the BET (Brunauer–Emmett–Teller) method, while the total pore volume was estimated from the desorbed volume at a relative pressure near unity, utilizing the BJH (Barrett–Joyner–Halenda) method.

The surface chemistry was studied by X-ray photoelectron spectroscopy (XPS) using a K-Alpha instrument from Thermo Scientific (Thermo Fisher Scientific, Waltham, MA, USA) with a monochromatic Al Kα source (1486.6 eV) at a bass pressure of 2 × 10⁻⁹ mbar. Charging effects were compensated by a flood gun and binding energies were calibrated by placing the C1s peak at 284.4 eV as the internal standard. A pass energy of 200 eV and 20 eV was used for survey and high-resolution spectra acquisition, respectively.

The basic properties of the catalysts were determined by CO₂ Temperature-Programmed Desorption (TPD) using a Porotec instrument (Frankfurt, Germany) equipped with a thermal conductivity detector (TCD). The samples (0.1 g) were pretreated for 1h at 200 °C under He (20 mL min⁻¹) and a heating rate 10° C min⁻¹ to clean the surface before the experiment, and then cooled down to RT. Further, the samples were exposed to a gas stream composed of 10% CO₂/He (20 mL min⁻¹) for 1 h at RT to saturate the catalyst with CO₂. Subsequently, the samples were exposed to an He flow (20 mL min⁻¹) for 60 min to remove the physically adsorbed CO₂, and, finally, they were heated from RT to 800 °C at 10 °C min⁻¹, and the CO₂ released was measured by a TCD.

3. Results and Discussions

3.1. Catalyst Characterization

The crystalline phases of the active components in the nano-catalysts Ti-Al-Mg, Ti-Cr-Mg, and Ti-Fe-Mg were identified using X-ray diffraction (XRD), as depicted in Figure 1. The resulting diffraction patterns were analyzed with the aid of the “Match!” software Version 3.1.0 Build 50, and phase identification was validated through cross-referencing with reliable crystallographic databases and the peer-reviewed literature.

All catalysts exhibited characteristic diffraction peaks at 2θ = 25.37°, 38.25°, 48.80°, 54.65°, and 62.75°, corresponding to the crystalline phases of MgTi₂O₄ [19,20], Ti₃O₅ [21,22,23], TiO₂ [21,24] and MgO [25,26], respectively.

In the case of Ti-Al-Mg, a distinct peak was observed at 2θ ≈ 47.45°, which was attributed to the formation of the ternary oxide phase MgAl₂Ti₃O₁₀ [16]. Notably, no reflections corresponding to binary aluminum–magnesium oxides, such as MgAl₂O₄, or to aluminum oxide (Al₂O₃) were detected, indicating that the ternary phase is the dominant crystalline form in this system.

Conversely, the Ti-Fe-Mg and Ti-Cr-Mg catalysts exhibited additional diffraction peaks at 2θ = 64.10° and 36.17°, which were assigned to the spinel-type binary oxides MgFe₂O₄ [26,27,28] and Cr₂MgO₄ [26], respectively. No evidence of ternary oxide phase formation was observed in either system.

Furthermore, the XRD diffractograms revealed the presence of monometallic oxide phases. In Ti-Fe-Mg, a peak at 2θ = 33.27° was consistent with the hematite phase of Fe₂O₃ [29] while the Ti-Cr-Mg catalyst exhibited a peak at 2θ = 41.23°, corresponding to the crystalline phase of Cr₂O₃ [30].

The crystallite size (D) of the catalysts was calculated using the Debye–Scherrer Equation (1) [31]:

$$\mathrm{D}={\displaystyle \frac{k\cdot \lambda }{\beta \cdot \cos \theta }}$$

(1)

In the equation employed for crystal size estimation, λ represents the wavelength of the X-ray radiation used (Cu Kα = 0.15406 nm), β denotes the full width at half maximum (FWHM) of the diffraction peak, and θ corresponds to the Bragg diffraction angle. The constant k, known as the shape factor, is typically assumed to have a value of 0.98. Based on this equation, the average crystallite sizes of the phases formed during the preparation of the Ti-Al-Mg, Ti-Cr-Mg, and Ti-Fe-Mg catalysts were calculated, as presented in

Table 1. Oxide types and calculated crystallite sizes of the phases in Ti-Al-Mg, Ti-Fe-Mg, and Ti-Cr-Mg catalysts based on X-ray diffraction (XRD) patterns.

2θ	Ti-Al-Mg		Ti-Fe-Mg		Ti-Cr-Mg
	Oxide Type	Crystallite Size (nm)	Oxide Type	Crystallite Size (nm)	Oxide Type	Crystallite Size (nm)
25.37	MgTi2O4	0.087	MgTi2O4	0.143	MgTi2O4	0.116
33.27	-	-	Fe2O3	0.249	-	-
36.17	-	-	-	-	Cr2MgO4	0.092
38.25	TiO2	0.062	TiO2	0.276	TiO2	0.074
41.23	-	-	-	-	Cr2O3	0.135
47.45	MgAl2Ti3O10	0.108	-	-	-	-
48.80	-	-	Ti3O5	0.268	Ti3O5	0.149
54.65	TiO2	0.088	TiO2	0.224	TiO2	0.086
62.75	MgO	0.090	MgO	0.172	MgO	0.106
64.10	-	-	MgFe2O4	0.237	-	-

.

The XPS spectra of the prepared catalysts, as shown in Figure 2a, confirm the presence of Mg, Ti, and O in all samples, as evidenced by the characteristic peaks of Mg 1s, Mg 2p, Ti 2p_3/2, Ti 2p_1/2, and O 1s. The Mg 1s peak at approximately 1304.1 eV indicates the successful incorporation of magnesium through Mg-O bonding, while the Mg 2p signal around 50.1 eV suggests the presence of Mg-OH groups [32]. High-resolution Ti 2p spectra reveal a well-defined Ti 2p_3/2 peak at ~458 eV and a weaker Ti 2p_1/2 peak at ~464 eV, consistent with Ti⁴⁺ species forming Ti-O bonds [19]. The O 1s peak at 530.70 eV corresponds to lattice oxygen associated with Mg²⁺, Al³⁺, and Ti⁴⁺, further confirming the formation of Ti-O, Al-O, and Mg-O bonds [16,33].

These observations are in agreement with the XRD analysis, which shows the formation of crystalline phases such as TiO₂, MgO, and the binary oxide MgTi₂O₅ with an orthorhombic crystal structure [34]. The Ti 2p binding energies confirm the presence of Ti⁴⁺, as expected in TiO₂ and MgTi₂O₅, while the Mg 1s signal indicates Mg²⁺, in accordance with the known crystal structures.

In the case of the Ti-Al-Mg catalyst, the Al 2p and Al 2s peaks at 73.9 and 117.8 eV, respectively, indicate oxidized aluminum forming Al-O bonds [16]. These findings are supported by the XRD results, which confirm the formation of the ternary oxide phase MgAl₂Ti₃O₁₀, which arises due to cation substitution phenomena, specifically the replacement of Al³⁺ by Ti⁴⁺ within the oxide lattice, leading to a homogenous crystalline network [16,35]. Such substitutions induce changes in the crystal structure and are consistent with reports in the literature describing Ti-Al-Mg as a well-structured mixed metal oxide catalyst with strong metal–oxygen interactions, which contribute to its enhanced performance in glycerol carbonate synthesis. This substitution is confirmed by the 1 eV red shift of the Mg 1s peak observed for this sample at 1303.1 eV compared to the other two, as depicted in Figure 2b. Thus, because Mg²⁺ ions have different neighboring cations (Al³⁺ and Ti⁴⁺), the surrounding environment is more electron rich and donates electron density towards Mg²⁺ ions through inductive effects. Therefore, in this complex crystalline environment, the Mg²⁺ ions are expected to have a weaker base character compared to the other crystalline structures identified in the other two samples, as indicated in Figure 2b.

In contrast, no ternary oxide formation was observed for Ti-Cr-Mg and Ti-Fe-Mg catalysts. The XPS spectrum of Ti-Cr-Mg confirms the presence of chromium through the Cr 2p peak at ~579 eV, indicating the formation of Cr-O bonds within a mixed metal oxide matrix. The O 1s peak is associated with lattice oxygen bonded to Mg²⁺, Cr³⁺, and Ti⁴⁺, supporting the presence of Ti-O, Cr-O, and Mg-O bonds. These results are in line with the XRD findings, showing the formation of Cr₂MgO₄ with a classical AB₂O₄ spinel structure and Cr₂O₃ [36,37] reinforcing the presence of a well-defined oxide network with strong metal–oxygen interactions [33,38].

Similarly, the XPS spectrum of the Ti-Fe-Mg catalyst reveals Mg, Ti, O, and Fe species. The Fe 2p signal at approximately 712.20 eV corresponds to Fe³⁺ species forming Fe-O bonds (Fe₂O₃) [33,39] XRD analysis confirms the formation of MgFe₂O₄, a binary oxide with a classical spinel structure of the AB₂O₄ type [36,40,41].

Figure 3 presents SEM images of the porous catalysts, revealing distinct morphological features across the prepared catalysts. The Ti-Al-Mg catalyst shows a porous, rough, and agglomerated surface composed of irregularly shaped particles with layered and compact structures. This morphology suggests the presence of an extended network of pores and interparticle voids, which facilitates effective molecular diffusion. The roughness and porosity provide a large number of exposed active sites, allowing for efficient adsorption and activation of reactants. These features indicate minimal diffusion resistance and are generally associated with high catalytic activity due to the enhanced accessibility of internal active regions [42,43].

In contrast, the Ti-Cr-Mg catalyst exhibits a more agglomerated morphology with fewer visible pores and irregular particle shapes. The agglomeration reduces the external surface area and blocks the internal porosity, creating diffusion constraints. Such a morphology limits the penetration of glycerol and CO₂ into the catalyst structure, reducing the contact between reactants and catalytic sites. This structural feature is consistent with the findings of Granados-Reyes et al. [17], who observed that Cr-containing Mg-based catalysts exhibited decreased catalytic performance due to agglomeration-induced diffusion resistance and reduced site availability.

The Ti-Fe-Mg catalyst further emphasizes this limitation. SEM images show a fragmented, dense, and heterogeneous surface with large agglomerates and significantly restricted porosity. The compact arrangement of particles and lack of interconnected pores severely hinder internal diffusion, especially under liquid-phase reaction conditions where glycerol’s high viscosity further reduces mass transport. These morphological features impose substantial diffusion limitations, restricting the effective utilization of catalytic sites. Singh et al. [18] reported similar observations in Fe-based catalysts for glycerol carboxylation, where particle agglomeration and low porosity resulted in low CO₂ activation and reduced catalytic activity.

In summary, the SEM images of Ti-Cr-Mg and Ti-Fe-Mg catalysts reveal morphologies that significantly restrict reactant diffusion. The agglomerated, dense structures with limited porosity create diffusion barriers, leading to poor catalyst performance due to the reduced accessibility of active sites.

The EDS analysis in Figure 4 confirms the successful formation of a Ti-Al-Mg oxide structure in the Ti-Al-Mg catalyst, as evidenced by the presence of Ti, Al, Mg, O, C, and N, with significant contributions from oxygen (64.09%), aluminum (23.51%), and magnesium (6.61%). Elemental mapping further reveals a uniform distribution of these key elements, which supports improved catalytic efficiency and thermal stability [16].

In comparison, the EDX spectrum of Ti-Cr-Mg shows the presence of Ti, Cr, Mg, and O, with chromium contributing 21.99%. While Cr incorporation enhances moderate Lewis acidity and structural stability, the elemental mapping displays a slightly less homogeneous distribution than in Ti-Al-Mg. This lower uniformity may hinder the synergetic effects of the acid–base sites ensured by the different metal ions in proximity, behaving more like mechanical mixtures of different oxide phases. These results are in line with the findings of Granados-Reyes et al., who reported that Cr-based mixed oxides, though moderately effective, were limited by uneven elemental dispersion and lower acid site availability [17].

The EDX analysis of the Ti-Fe-Mg catalyst confirms the presence of Ti, Fe, Mg, and O, with Fe comprising 28.97%. Although Fe contributes weak acid sites, its high content may lead to the formation of less active species, likely due to a reduced pore volume. As shown in

Table 2. Textural properties of prepared catalysts.

Catalysts	BET Surface Area (m2/g)	Pore Volume (cm3/g)	Average Pore Width (nm)
Ti-Al-Mg	119.43	0.345	6.81
Ti-Cr-Mg	54.47	0.193	7.12
Ti-Fe-Mg	28.20	0.166	10.68

, the pore volume decreases to 0.166 cm³/g compared to 0.345 cm³/g for Al-Ti-Mg, which may hinder diffusion and catalytic activity. Additionally, elemental mapping reveals a non-uniform distribution of Fe, further reducing the number of effective catalytic sites. These findings are consistent with those reported by Singh et al., where Fe-containing catalysts exhibited low performance due to uneven elemental dispersion and weak acidity [18].

The uneven elemental dispersion in the Cr- and Fe-containing catalysts resulted in the formation of single metal oxide clusters, leading to a reduction in surface area to 54.47 m²/g and 28.20 m²/g, respectively, as shown in Table 2. This decline in surface area consequently reduced the number of available active sites compared to the Al-Ti-Mg catalyst, which exhibited a surface area of 119.43 m²/g.

The nitrogen adsorption–desorption isotherms of the Ti-Al-Mg, Ti-Cr-Mg, and Ti-Fe-Mg catalysts (Figure 5) all exhibit type IV isotherms with H₂ hysteresis loops, which are characteristic of mesoporous structures with ink-bottle-shaped pores according to IUPAC classification [44,45]. In all cases, the gradual increase in adsorption at low relative pressure (P/P₀ < 0.3) confirms the presence of micropores, while the sharp rise at higher pressures (P/P₀ > 0.7) indicates the existence of larger mesopores and some macropores.

However, differences in pore size distributions are observed. For Ti-Al-Mg and Ti-Cr-Mg, the inset graphs show sharp peaks around 3–5 nm, reflecting a dominant mesoporous region with limited expansion beyond 10 nm. This indicates rather uniform porous textures, in line with reports in the literature [16,46], and confirms a well-structured pore network that supports catalytic performance [33].

In contrast, Ti-Fe-Mg exhibits a broader mesoporous distribution with a sharp peak between 8 and 12 nm and minimal extension beyond 20 nm. Despite the larger average pore size, the distribution remains relatively narrow, indicating a fairly uniform structure. These observations are consistent with the reported findings in the literature [16,46], further supporting the classification and textural analysis of these catalysts.

Table 2 summarizes the surface area, pore volume, and pore width of the as-prepared catalysts. Among metal oxides, Ti-Al-Mg has the highest BET surface area (119.43 m²/g) and pore volume (0.345 cm³/g), confirming a highly mesoporous structure (6.81 nm pore width) that supports good adsorption and catalytic activity. Ti-Cr-Mg, with a moderate surface area (54.47 m²/g) and slightly larger pores (7.12 nm), offers a balance between porosity and stability, making it suitable for catalysis but less efficient than Ti-Al-Mg. Ti-Fe-Mg has the largest pores (10.68 nm), allowing better mass transfer, but its lower surface area (28.20 m²/g) and moderate pore volume (0.166 cm³/g) indicate a trade-off between adsorption capacity and catalytic efficiency, which aligns with studies in the literature [16,46].

The thermal properties of the Ti-Al-Mg, Ti-Cr-Mg, and Ti-Fe-Mg catalysts were evaluated using thermogravimetric analysis (TGA) under nitrogen (N₂) flow. As shown in the TGA curves (Figure 6), all three samples exhibit an initial weight loss below 300 °C, which is attributed to the dehydration of adsorbed water and interlayer crystal water. These materials display minimal weight loss during heating, indicating high thermal stability across the series.

Between 300 °C and 600 °C, a gradual weight reduction was observed in all samples, corresponding to the removal of interlayered crystal water and hydroxyl (-OH) layers [4]. This steady mass loss reflects a stable thermal profile with no abrupt structural breakdown. The corresponding DTG curves for each catalyst further confirm this behavior, showing broad, low-intensity peaks that suggest a slow and continuous dehydration process rather than a sudden decomposition event [13].

The basicity of the three samples was evaluated by CO₂ TPD, and the desorption curves are shown in Figure 7. As predicted by the XPS spectra, the CO₂ desorption temperature from the Ti-Al-Mg catalysts is lower than those of the other two catalysts, indicating a weaker binding energy for carbon dioxide molecules on the catalyst containing aluminum. This is most likely caused by the transfer of electron density from neighboring Ti⁴⁺ and Al³⁺ ions in the ternary oxide crystalline structure of this catalyst. By contrast, in the catalysts with Cr and Fe, because a ternary oxide has not formed, such an electron density transfer is not possible; thus, the Mg²⁺ ions behave similarly to the pure MgO phase, which binds carbon dioxide more strongly, and the surface reaction is slowed down.

The Ti-Al-Mg catalyst shows a broad CO₂ desorption peak with a maximum at approximately 286 °C, and two higher temperatures with weak desorption features. While the large and broad peak at lower temperatures most likely corresponds to the desorption of CO₂ chemisorbed on the ternary oxide phase, with some contributions from the Ti-Mg spinel structure, the weaker ones at higher temperatures likely belong to carbonate species formed with the pure MgO not incorporated in the mixed oxide phases evidenced in the XRD analysis.

In contrast, the Ti-Cr-Mg catalyst shows two well-defined and weak desorption features at low temperatures near 300 °C and 390 °C, respectively, which can likely be attributed to the CO₂ adsorbed on the Mg²⁺ sites in the Ti-Mg and Cr-Mg spinel structures observed in the XRD analysis, and a dominant higher-temperature desorption most probably corresponding to the decomposition of magnesium carbonate formed on the pure MgO surface. This is consistent with the XRD results and EDX results indicating the presence of a high concentration of MgO in this sample.

The Ti-Fe-Mg catalyst also shows two low-temperature CO₂ desorption peaks near 240 °C and 365 °C that are assigned to the Ti-Mg and Fe-Mg spinel structures, respectively, as observed in the XRD analysis. This sample also shows the high-temperature strong desorption peak that is attributed to the decomposition of magnesium carbonate formed by the pure MgO phase in the sample.

Comparing the catalysts with Cr and Fe, the TPD profiles suggest that the one with Cr contains the highest concentration of pure MgO phase since the ratio between the area of the high-temperature CO₂ desorption peak and the area of the lower-temperature features is considerably higher than in the case of the Ti-Fe-Mg catalyst. This interpretation is also consistent with the EDX investigation that showed the highest Mg concentration for the Ti-Cr-Mg, most likely caused by crystalline phase segregation. However, among the three catalysts, the Ti-Al-Mg has the highest low-temperature CO₂ adsorption capacity, and most likely the lowest CO₂ binding energy on the surface.

3.2. Catalysts’ Effect on Glycerol Carbonate Synthesis

Figure 8 shows the effect of temperature on glycerol carbonate yield when using Al-Ti-Mg, Fe-Ti-Mg, and Cr-Ti-Mg catalysts at 5% catalyst-to-glycerol loading, 10 bar pressure, and a 4 h reaction time.

For Al-Ti-Mg and Fe-Ti-Mg catalysts, the yield increased as the temperature rose from 130 °C to 175 °C. Beyond 175 °C, the yield with Al-Ti-Mg remained stable, while that with Fe-Ti-Mg began to decline. Specifically, Al-Ti-Mg produced a yield of 11.3% at 130 °C, which increased to 36.10% at 175 °C. In comparison, Fe-Ti-Mg showed a yield increase from 8.55% to 25.19% over the same temperature range, followed by a decrease to 17.35% at 200 °C. This decrease is attributed to the formation of undesired by-products resulting from side reactions, including glycerol dehydration and polymerization [47,48,49].

Cr-Ti-Mg, however, exhibited a significantly lower yield at 175 °C and 10 bar CO₂ pressure, over four times lower than Al-Ti-Mg and less than half that of Fe-Ti-Mg.

Based on Figure 8, for the Ti-Al-Mg catalyst, the glycerol carbonate yield at 200 °C is approximately 35.9%, which is only marginally lower than the maximum yield of 36.1% observed at 175 °C. This indicates that the yield remains relatively stable beyond 175 °C, suggesting good thermal resistance of the catalyst.

However, 175 °C is considered the optimal temperature not only because it corresponds to the highest yield, but also because further increases in temperature may promote side reactions such as glycerol dehydration and polymerization, leading to undesired by-products, as previously reported in the literature [47,48,49]. Therefore, operating at 175 °C ensures high selectivity toward glycerol carbonate while minimizing secondary reactions. In contrast, for the Fe-Ti-Mg catalyst, the yield begins to decline significantly beyond 175 °C (from 25.2% to 17.4% at 200 °C), supporting the conclusion that 175 °C is indeed the optimal condition for both selectivity and conversion.

The Al-Ti-Mg catalyst exhibits high efficiency due to its large specific surface area (119.43 m²/g), which is more than four times greater than that of Fe-Ti-Mg (28.20 m²/g), enhancing its capacity to adsorb reactants, as shown in Figure 5. Its CO₂ chemisorption properties are also optimal, as shown in the TPD experiments above.

XRD and XPS analysis reveal the presence of iron oxides (Fe₂O₃) in Fe-Ti-Mg, which increase its acidity [50] and reduce its effectiveness in converting CO₂ and glycerol to glycerol carbonate. Additionally, SEM-EDX analysis shows that the magnesium content in Al-Ti-Mg is 6.51%, nearly double that of Fe-Ti-Mg (3.36%).

Although Cr-Ti-Mg has a moderate surface area (54.47 m²/g) and high magnesium content, its catalytic performance is poor due to the strong acidity of chromium oxide, which is higher than that of iron and aluminum oxides, and also because of its poor performance in CO₂ activation, with most of the carbon dioxide binding irreversibly onto its surface under our reaction conditions, as shown by the TPD experiments. These findings align with the previously reported literature, further validating the acidity-based classification of these metal oxide catalysts [51,52,53] thus hindering its activity in this reaction.

As shown in Figure 9, increasing the temperature enhances glycerol conversion. However, the glycerol carbonate yield remains relatively stable above 175 °C for Al-Ti-Mg, while it decreases for Fe-Ti-Mg beyond this point. This decline is attributed to the formation of undesired by-products from side reactions such as glycerol polymerization and thermal degradation [47,48,49].

The sharp drop in glycerol carbonate yield with Fe-Ti-Mg at temperatures above 200 °C is likely due to iron-catalyzed degradation of glycerol carbonate, which reduces reaction efficiency and leads to product loss to by-products.

Figure 10 shows that increasing the catalyst loading up to 10% improves glycerol carbonate yield for all catalysts. Al-Ti-Mg delivered the highest yield, followed by Fe-Ti-Mg and then Cr-Ti-Mg. A significant improvement was observed when increasing the catalyst loading between 2.5% and 5%, while the increase at 10% gave only a marginal increase in glycerol carbonate yield, suggesting the reaction may be approaching a catalytic efficiency limit at higher loadings.

Similarly, as shown in Figure 11, glycerol conversion increased steadily with higher catalyst loading. Al-Ti-Mg again showed the highest conversion, followed by Fe-Ti-Mg and Cr-Ti-Mg. At 10% concentration, the conversion was significantly higher than at 2.5%, due to the increased number of active sites available at higher catalyst loadings, which enhances both glycerol conversion and glycerol carbonate formation.

The superior performance of Al-Ti-Mg is attributed to its balanced acid–base properties, which play a key role in CO₂ activation and glycerol adsorption.

The catalytic performance follows the order Al-Ti-Mg > Fe-Ti-Mg > Cr-Ti-Mg. This is attributed to Al-Ti-Mg’s higher surface area, better porosity, and uniform magnesium content at the surface, which improves base sites at the surface. Its lower acidity also minimizes over-adsorption and suppresses undesired side reactions.

At 10% catalyst loading, performance tends to stabilize, likely eliminating mass transfer limitations that might occur at lower catalyst loading resulting in limited active sites being available for the reaction. Therefore, increasing catalyst loading from 2.5% to 10% at 175 °C and 10 bar CO₂ pressure significantly improves both glycerol carbonate yield and glycerol conversion. However, the benefit beyond 5% is less pronounced, emphasizing the need to optimize catalyst loading for maximum efficiency without excess material use.

The catalytic performance of the synthesized mixed metal oxide catalysts (Ti-Al-Mg, Ti-Fe-Mg, and Ti-Cr-Mg) is governed by a complex interplay of surface composition, homogeneity, acid–base character, and textural properties. Among the three catalysts, Ti-Al-Mg exhibited the highest catalytic activity, with a maximum glycerol carbonate yield of 36.1%. This superior performance is attributed to the formation of a well-defined ternary oxide phase (MgAl₂Ti₃O₁₀), confirmed by XRD and XPS analyses. This phase reflects strong metal–oxygen interactions and a homogeneous distribution of Ti, Al, and Mg, as revealed by EDX mapping. The structural uniformity promotes a balanced distribution of active sites and reduces phase segregation, which are essential for efficient CO₂ activation and glycerol conversion.

As illustrated in

Table 3. Comparison of catalytic performance for various catalysts in glycerol carbonate synthesis from glycerol and CO₂.

Catalyst	Parameters	Glycerol Conversion (%)	GC Yield (%)	Selectivity (%)	Ref
20%La2O2CO3-ZnO	Temperature: 150 °C; PCO2: 55 bar; Time: 6 h;	6.9	6.1	88.40	[13]
Ce0.98Zr0·02O2	Temperature: 150 °C; PCO2: 40 bar; Time: 5 h;	40.9	36.3	88.75	[14]
La2O2CO3/ZnO (La/Zn = 1:4)	Temperature: 170 °C; PCO2: 30 bar; Time: 7 h;	30.3	14.3	47.20	[12]
Cu/Mg-Al-Zr	Temperature: 150 °C; PCO2: 40 bar; Time: 12 h;	11.5	2.4	20.9	[54]
30%,CuO/Al2O3	Temperature: 150 °C; PCO2: 40 bar; Time: 5 h;	37.8	15.0	39.8	[10]
Co-ETS-10	Temperature: 170 °C; PCO2: 40 bar; Time: 6 h;	35.0	12.7	36.3	[5]
Zn-ETS-10	Temperature: 170 °C; PCO2: 40 bar; Time: 6 h;	32.7	11.7	35.6	[5]
Ti-Al-Mg	Temperature: 175 °C; PCO2: 10 bar; Time: 4 h;	58.3	36.1	61.9	Present work

, the Ti-Al-Mg catalyst developed in this study demonstrates a superior glycerol conversion (58.3%) and GC yield (36.1%) under milder CO₂ pressure (10 bar), outperforming most catalysts reported in the literature under comparable or even harsher conditions.

In contrast, Ti-Fe-Mg and Ti-Cr-Mg formed binary spinel-type oxides (MgFe₂O₄, Cr₂MgO₄) and segregated metal oxide phases (Fe₂O₃, Cr₂O₃), without evidence of ternary phase formation. These structures led to surface heterogeneity and less uniform elemental dispersion, which limited the accessibility and effectiveness of active sites. The SEM images confirmed that both catalysts exhibited more agglomerated and fragmented morphologies, contributing to reduced surface exposure.

The acid–base character of the catalysts further explains the observed differences in performance. Ti-Al-Mg displayed a good acid–base balance due to the presence of weak Lewis acid sites from Al³⁺ and basic sites from Mg²⁺. This balance favors the GC formation by facilitating CO₂ adsorption and minimizing side reactions. In contrast, Fe³⁺ and Cr³⁺ introduced stronger acidic sites, as supported by the literature and indirectly confirmed by the catalytic behavior. The acidity in Ti-Fe-Mg and Ti-Cr-Mg promoted glycerol dehydration and polymerization side reactions, reducing the GC yield.

Textural analysis reinforced these trends. Ti-Al-Mg had the highest BET surface area (119.43 m²/g) and a well-developed mesoporous structure, providing more accessible sites for reactants. Ti-Cr-Mg had a moderate surface area (54.47 m²/g), while Ti-Fe-Mg exhibited the lowest (28.20 m²/g), limiting its adsorption capacity despite a larger average pore size.

In summary, the findings confirm that optimal catalytic performance is achieved when the catalyst exhibits structural homogeneity, uniform dispersion of metal ions capable of forming stable ternary oxide phases, and a well-balanced acid–base surface. Ti-Al-Mg meets these conditions, resulting in superior activity. In contrast, the less organized structures and higher surface acidity of Ti-Cr-Mg and Ti-Fe-Mg limit their effectiveness. Consequently, the observed activity trend Ti-Al-Mg > Ti-Fe-Mg > Ti-Cr-Mg is directly correlated with these parameters.

Figure 12 presents the catalytic stability behavior of the Ti-Al-Mg mixed oxide catalyst across four consecutive reaction cycles, each lasting 4 h, conducted under constant reaction conditions: 175 °C, 10 bar CO₂ pressure, and 5 wt% catalyst loading relative to glycerol. Prior to each cycle, the catalyst was thoroughly washed with ethanol and subsequently dried at 90 °C for one hour to ensure proper cleaning.

The catalyst exhibited high initial activity, achieving a glycerol conversion of 58.3% and a glycerol carbonate yield of 36.1% in the first cycle, demonstrating excellent selectivity toward the desired product under mild operating conditions. Upon repeated use, a gradual decline in performance was observed; the GC yield dropped to 33.4% and 33.0% in the second and third cycles, respectively, and reached 31.6% by the end of the fourth cycle. This trend suggests a slight deactivation, likely due to contamination of surface active sites with heavy by-products adsorbed irreversibly.

Nevertheless, the maintained activity across multiple cycles indicates the catalyst’s robust structural integrity and thermal stability. The observed resistance to deactivation mechanisms such as surface fouling or sintering reinforces the suitability of Ti-Al-Mg as a promising heterogeneous catalyst for semi-batch processes.

4. Materials and Methods

4.1. Materials

Titanium (IV) butoxide (Merck, 97%), Al(NO₃)₃·9H₂O (Merck, 99%), Mg(NO₃)₂·6H₂O (Merck (Rahway, NJ, USA) ≥99.0%), Cr(NO₃)₃·9H₂O, (Merck, 99%), FeCl₃·6H₂O (Carl Roth, Karlsruhe, Germany, min. 98%), Carbon dioxide CO₂ (SIAD, Bucharest, Romania, Purity: ≥99.5%, Glycerol (Gly) (Merck, ≥99.5%), Acetonitrile (Amex-lab, Bucharest, Romania, ≥99.9%), and Glycerol carbonate (GC) (Merck-Sigma-Aldrich, Burlington, MA, USA, ≥99%).

4.2. Catalysts Synthesis

The nanostructured Ti-(Me)-Mg catalysts, where Me represents metal ions such as Cr³⁺, Al³⁺, and Fe³⁺, were synthesized via a co-precipitation method. This approach involved the reaction of titanium (IV) butoxide with nitrate salts of Al³⁺ and Mg²⁺, and with a chloride salt of Fe³⁺, maintaining a fixed molar ratio of Ti:Me:Mg = 2:1:1. To prepare Ti-Al-Mg, 27.23 g of titanium (IV) butoxide, 15.00 g of Al(NO₃)₃·9H₂O, and 10.62 g of Mg(NO₃)₂·6H₂O were used. In the case of Ti-Fe-Mg, 27.23 g of titanium (IV) butoxide, 10.80 g of FeCl₃·6H₂O, and 10.62 g of Mg(NO₃)₂·6H₂O were used. For Ti-Cr-Mg, the synthesis involved 27.23 g of titanium (IV) butoxide, 11.65 g of Cr(NO₃)₃·9H₂O, and 10.62 g of Mg(NO₃)₂·6H₂O.

For each catalyst, the precursors were dissolved in 500 mL of absolute ethanol. To ensure complete dissolution, the pH was adjusted to 3.0 using concentrated nitric acid. The solution was stirred at 60 °C, and ammonia solution (25 wt%) was added dropwise until the pH reached 10–11, leading to the precipitation of nanoparticles. The mixture was stirred for an additional 45 min. After the reaction, the nanoparticles were separated by centrifugation and washed with distilled water. The nanoparticles were then washed with ethanol, dried for 2 h at 150 °C, and calcined at 450 °C for 4 h with a heating rate of 10 °C/min. The calcination temperature was selected based on the thermal stability of the nanoparticles before calcination.

4.3. Experimental Setup and Procedure for Catalyst Testing in Glycerol Carbonate Synthesis

The experimental program for the synthesis of glycerol carbonate was carried out using a batch-type BERGHOF installation. The main component is a stainless-steel autoclave-type reactor with a working volume of 600 mL equipped with a turbine-type mechanical stirrer (up to 2000 rpm) and sealed at the reactor lid using a packing gland system. The setup includes sensors for temperature, pressure, and rotational speed. Heating is provided by an electric mantle with a temperature control system. The reactor lid is equipped with the following ports: a gas inlet for volatile reactants or inert gases, a controlled gas outlet connected to a purge system for depressurization, and a safety valve, also linked to the purge line.

Using this setup, the experimental procedure was carried out as follows: A mixture of 25 g of glycerol and 50 mL of acetonitrile serving as a dehydrating agent to shift the unfavorable equilibrium towards the product side and enhance the selectivity for glycerol carbonate [39,55] was introduced into the reaction vessel. Dried catalyst was subsequently added at room temperature and the reactor was immediately closed and purged for 5 min with nitrogen gas. Subsequently, the reaction temperature was gradually increased while continuous mechanical stirring of the glycerol–catalyst–acetonitrile mixture was maintained. Once the desired temperature was reached, carbon dioxide gas was introduced into the system until the desired pressure was reached, initiating the reaction, which proceeded for 4 h.

A reaction time of 4 h was selected based on commonly reported durations for similar catalytic systems. As referenced in the manuscript [5,9,10], reactions for glycerol carbonylation over metal oxide catalysts are typically conducted for 4–5 h to reach the optimal yield. In our study, the reaction was stopped at four hours.

Upon completion of the reaction time, the autoclave was allowed to cool naturally to room temperature, and then the pressure was released in a controlled manner. The liquid phase was then separated from the catalyst through centrifugation. The recovered catalyst was washed with ethanol and deionized water and then dried at 80 °C overnight to ensure complete removal of residual solvents. The liquid phase was subjected to analysis using gas chromatography with a flame ionization detector (GC-FID, Varian 3800, Palo Alto, CA, USA). The gas chromatograph was equipped with a ZB-5ms column (L = 30 m, D = 50 μm, d = 0.25 μm) and Helium (He) was used as the carrier gas at a flow rate of 1 mL/min.

The oven thermal program was initiated at 100 °C with a heating rate of 5 °C/min until it reached 250 °C, where it was maintained for 1 min. The injector temperature was set to 250 °C, and the injected sample volume was 1 μL.

To quantify the concentration of glycerol carbonate in the final product, a calibration curve was established, as shown in Figure 13, using acetonitrile as the solvent. The calibration curve was constructed over a glycerol carbonate concentration range of 1 to 90% by weight.

The yield was calculated based on the following equation:

$$\%Yield of GC={\displaystyle \frac{moles of \mathrm{GC}}{moles of \mathrm{Gly} \mathrm{introduced}}}$$

5. Conclusions

This study demonstrates the successful synthesis, characterization, and performance evaluation of three mixed metal oxide catalysts, Ti-Al-Mg, Ti-Cr-Mg, and Ti-Fe-Mg, for the conversion of glycerol and CO₂ into glycerol carbonate via heterogeneous catalysis. Physicochemical analyses highlighted the superior textural properties of Ti-Al-Mg, including ternary and binary oxide phases formation, high surface area, and uniform porosity. The crystalline surface composition was identified, and its base character were inferred from XPS and CO₂ TPD characterizations, as well as from the literature, all indicating that its enhanced catalytic activity is most likely due to the weaker chemisorption of CO₂ molecules on magnesium active sites with weaker base character resulting from electron density transfer from surrounding Al³⁺ and Ti⁴⁺ ions in the ternary oxide crystalline phase. In contrast, in the case of the other two catalytic systems, CO₂ molecules likely bound more strongly to the surface, leading to a lower surface glycerol carbonylation reaction rate. Strong chemisorption of CO₂ to the surface blocks the magnesium sites in the pure MgO phase, and likely slows down the surface reaction rate on double oxides with a spinel structure such as Ti-Mg, Fe-Mg, and Cr-Mg. This also explains why the Ti:Cr:Mg, which has the highest surface Mg concentration evidenced by EDX, but in the form of pure MgO or Cr₂MgO₄/MgTi₂O₄ phases, performs more weakly than the catalyst with aluminum and the one with iron, which shows a higher concentration of double oxides.

Among the catalysts tested, Ti-Al-Mg exhibited the highest glycerol carbonate yield of 36.1% at 175 °C and 10 bar CO₂ over 4 h, outperforming Ti-Fe-Mg and Ti-Cr-Mg. The high efficiency of Ti-Al-Mg is attributed to its optimal magnesium surface content, lower acidity, and improved CO₂ activation capacity. Conversely, Ti-Cr-Mg showed the lowest catalytic performance, mainly due to stronger CO₂ bonding onto the surface and limited active site concentration. Since the Ti-Fe-Mg and Ti-Cr-Mg systems did not form a ternary oxide phase, the use of different metal ratios should be explored for obtaining such systems, and to assess the potential of fine-tuning the chemisorption properties of CO₂ for glycerol carbonylation.

Catalytic testing under varying temperatures and catalyst loadings further confirmed the superior performance of Ti-Al-Mg. Yield and conversion were enhanced up to a 10% catalyst loading, although marginal improvements beyond 5% suggest a performance plateau likely limited by the surface reaction kinetics limitations under the tested reaction conditions.

Overall, Ti-Al-Mg is a promising catalyst for sustainable glycerol valorization through CO₂ utilization, providing an effective route toward green chemical synthesis.