Effect of Support on Steam Reforming of Ethanol for H₂ Production with Copper-Based Catalysts

Abstract

Catalytic studies hydrogen production via steam reforming of ethanol (SRE) are essential for process optimization. Likewise, selecting the ideal support for the active phase can be critical to achieve high conversion rates during the catalytic steam reforming process. In this work, copper-based catalysts were synthesized using two different supports, NaY zeolite and Nb₂O₅/Al₂O₃ mixed oxides. The materials were prepared using wet impregnation and characterized for their physicochemical properties using different analytical techniques. Differences in the catalyst morphologies were readily attributed to the characteristics of the support. The Cu/NaY catalyst exhibited a higher specific surface area (210.40 m² g⁻¹) compared to the Cu/Nb₂O₅/Al₂O₃ catalyst (26.00 m² g⁻¹), resulting in a homogeneous metal dispersion over the support surface. The obtained results showed that, at 300 °C, both the Cu/Nb₂O₅/Al₂O₃ and Cu/NaY catalysts produced approximately 50% hydrogen and 40% acetaldehyde, but with significant differences in conversion (6% and 56%, respectively). At 450 °C, a greater product distribution and a 10% higher conversion were observed when the catalyst was supported on NaY compared to Nb₂O₅/Al₂O₃. Hence, the performance of copper-based catalysts was influenced significantly by the textural properties of the support.

1. Introduction

The increasing global demand for energy, coupled with the socioenvironmental impacts of an energy matrix that is still heavily reliant on traditional fossil fuels such as coal, crude oil, and natural gas [1,2], requires the development of sustainable production processes to support the transition to a new and more diversified energy matrix. Among the sustainable alternatives for this energy transition, hydrogen (H₂) stands out as one of the most prominent.

H₂ can be obtained through different pathways such as electrolysis [3,4], biological reactions [5], biomass gasification [6], steam reforming [7,8], aqueous phase reforming [9], and partial oxidation of hydrocarbons and alcohols [10]. Among these possibilities, the use of ethanol as feedstock for H₂ production in fuel cells has considerable advantages. These include easier storage, handling, and safe transportation due to its low toxicity and volatility. Additionally, ethanol is a renewable feedstock when obtained through biomass fermentation, is rich in H₂, and has a nearly closed carbon cycle that helps in the abatement of greenhouse gas emissions [11,12]. Thus, the steam reforming of ethanol (SRE) emerges as an attractive solution for H₂ production due to its high H₂ yield and thermodynamic feasibility [13].

SRE for H₂ production is a catalytic process. Therefore, H₂ yield depends on the properties of the catalyst to be employed [14]. This includes the catalytic support for the active phase [15] and the method used for catalyst preparation [16]. In general, the catalyst design is crucial for a successful SRE process. Different catalytic systems have been investigated for SRE using noble and non-noble metal-based catalysts [17]. Among them, copper-based catalysts [18,19,20] have the advantage of being cost-effective and widely available compared to other metals. Copper as an active phase is widely used in commercial catalysts for methanol reforming. Additionally, the presence of copper active sites promotes ethanol steam reforming to produce H₂ and CO or its dehydrogenation to acetaldehyde followed by decarbonylation, producing CH₄ and CO [21,22,23]. Hence, copper-based catalysts have potential for SRE applications.

The choice of support for the active phase is extremely important for SRE because it plays a significant role in H₂ selectivity and catalyst stability [24,25], as already demonstrated in previous studies [26,27]. In general, efficient SRE supports must have favorable textural properties and moderate acidity, in addition to being relatively cheap, readily available, and easily accessible. In this study, two supports were selected and evaluated for their effect on H₂ production by SRE using copper as the catalytic active phase.

The first selected support was NaY, a commercially available zeolite that is known for its high heat resistance, unique ordered three-dimensional porous structure, and larger pores compared to the dimensions of the ethanol molecule, as well as low production costs [28,29]. For instance, NaY can be synthesized from alternative, abundant, and inexpensive materials such as rice husks [30,31] and wheat straw [32] ashes. Several studies have already demonstrated the application of NaY as a catalyst support for SRE [33,34,35]. The other selected support was a Nb₂O₅/Al₂O₃ mixed oxide. Alumina is widely used as a support in heterogeneous catalysis [36,37,38,39,40] due to its large surface area, good stability, and wide commercial availability [41]. Nb₂O₅ is also a notable material in the field of catalysis, known for its non-toxic nature, suitable acid properties [42,43], excellent chemical stability, high thermodynamic stability, low cost, and high commercial availability [44,45]. The combination of Nb₂O₅ with alumina is favorable because, being an n-type semiconductor, Nb₂O₅ can interact with copper in catalytic active reaction sites [46,47]. Additionally, Nb₂O₅ is structurally similar to commercial catalysts for methanol reforming (Cu/ZnO/Al₂O₃). Since ZnO is also an n-type semiconductor oxide, Nb₂O₅ may have similar catalytic properties.

This study aimed at evaluating the effect of NaY and Nb₂O₅/Al₂O₃ as the catalyst support for SRE. H₂ production was carried out using copper as the active phase under the same reaction conditions. To the best of our knowledge, these supports have never been compared in SRE reactions with copper used as the active phase. Copper was anchored on the support surface by wet impregnation. Then, the resulting catalytic systems were characterized using various analytical techniques and subjected to SRE using an experimental reaction module.

2. Materials and Methods

2.1. Material

Materials were synthesized using copper nitrate (Cu(NO₃)2·3H₂O, 98%) from Sigma-Aldrich, commercial alumina (Al₂O₃, 90%) from Merck, NaY zeolite from Sigma-Aldrich, and niobic acid (HY-340) from the Brazilian Metallurgy and Mining Company (CBMM). HY-340 was heat-treated to obtain niobium pentoxide (Nb₂O₅).

2.2. Methods

2.2.1. Catalyst Preparation

The catalysts were prepared using a simple wet impregnation methodology under solvent excess [48]. For the Cu/NaY catalyst, 57.25 g Cu(NO₃)₂·3H₂O were dissolved in water and mixed with 14.94 g of the NaY support. For the Cu/Nb₂O₅/Al₂O₃ catalyst, 57.25 g Cu(NO₃)₂·3H₂O were dissolved in water and mixed with 9.27 g Nb₂O₅ and 5.67 g Al₂O₃. The mixtures were placed in a rotary evaporator and evaporated for 2 h at 343 K for the complete elimination of water. After this, the materials were placed in an oven at 353 K for about 10 h. Then, the dried materials were crushed and placed in a muffle furnace in ambient atmosphere. The temperature was raised to 773.15 K using a 10 K min⁻¹ heating rate, and the catalysts remained at this plateau for 5 h. At the end of the process, two catalysts, named Cu/NaY and Cu/ Nb₂O₅/Al₂O_3, were obtained.

2.2.2. Catalyst Characterization

Scanning Electron Microscopy (SEM) was performed using a Zeiss EVO MA15 microscope coupled with an X-Max 20 mm² Energy Dispersive Spectrometer (EDS). X-ray Diffraction (XRD) analyses were conducted on a Shimadzu XDR-7000 diffractometer using CuKα radiation. Measurements were taken at 40 kV and 30 mA using a Cu tube with a wavelength of 1.54 nm, with a scan rate of 2° min⁻¹ and an interval of 5° ≤ 2θ ≤ 80°. The FWHM of the XRD peaks was used to estimate the average particle size using the Scherrer equation (Equation (1)).

$${\mathrm{D}}_{\left(\mathrm{hkl}\right)}=\frac{0.9\times \mathsf{\lambda }}{{\mathsf{\beta }}_{\left(\mathrm{hkl}\right)}\times \mathsf{\theta }},$$

(1)

The textural properties of the catalysts were determined by N₂ adsorption/desorption at 77 K using a NOVA-4000-Quantachrome adsorption analyzer. Infrared spectroscopy analyses (FTIR) were performed using a Varian 640-IR spectrometer with potassium bromide (KBr) as the dispersing agent in the region from 4000 to 400 cm⁻¹. Temperature-programmed desorption of NH₃ (TPD-NH₃) and temperature-programmed reduction (TPR) were carried out using a Quantachrome Chembet-3000 multi-use unit coupled with a ThermoStar-GSD 301 mass spectrometer. In both analyses, 0.1 g sample was placed in a “U”-shaped quartz reactor, which was first subjected to a 20 cm³ min⁻¹ N₂ flow at 300 °C for 1 h to remove humidity and possibly adsorbed materials. For TPD-NH₃ analysis, the samples were reduced with 1.75% H₂ diluted in N2 for 1 h using a heating rate of 10 °C min⁻¹ from room temperature to 500 °C and remaining at this temperature for another 1 h. NH₃ adsorption was performed at 100 °C for 30 min with a flow rate of 15 cm³ min⁻¹ of 5% NH₃ diluted in N₂. Subsequently, the system was purged for 2 h with a flow rate of 20 cm³ min⁻¹ N2. Finally, the sample was heated to 700 °C at a heating rate of 10 °C min⁻¹ under N2 flow for NH₃ desorption. TPR was performed with a reducing gas feed containing 1.75% H₂ in N2 at a flow rate of 20 cm³ min⁻¹, progressing from room temperature to 1000 °C with a heating rate of 10 °C min⁻¹.

2.2.3. Catalytic Performance Evaluation

SRE was carried out using two catalysts, Cu/NaY and Cu/Nb₂O₅/Al₂O₃. Tests were performed in an experimental unit consisting of a preheating system, a 20 cm-long stainless-steel reactor with an internal diameter of 2.54 cm, a condenser, and a phase collector/separator. The reactant mixture was introduced through the system inlet using a peristaltic pump.

Before the catalytic tests, the catalysts were activated in situ with an 85 cm³ min⁻¹ N₂ flow rate containing 40% H₂ by volume using the following heating steps: 30 min at 100 °C, 1 h at 200 °C, and 4 h at 500 °C. After this, the H₂ flow was stopped, and the N₂ flow was adjusted to 85 mL min⁻¹ for 4 h to purge H₂ from the entire reaction system. Subsequently, catalytic tests were conducted at 300 °C and 450 °C using a mass hourly space velocity of 40 dm³ h⁻¹ g_cat⁻¹, 5 g catalyst (20–40 mesh, positioned at the center with the reactor ends filled with silica of the same particle size), and a H₂O/C₂H₅OH molar ratio of 10/1 without the presence of an inert. The pellet size distribution was determined according to Trimm [49] to minimize both mass and heat diffusion effects in packed bed reactors (L/d_part ≥ 100 and D/d_part ≥ 30), resulting in a mean particle size distribution of 0.6 mm. The carbon balance was determined as described in Equation (2),

$$\mathrm{C}(\%)=\left(\frac{{\sum }_{\mathrm{i}}{\mathrm{F}}_{\mathrm{i}}{\mathrm{nc}}_{\mathrm{i}}}{{\mathrm{F}}_{\mathrm{in}}^{\mathrm{EtOH}}{\mathrm{nc}}_{\mathrm{i}}}\right)\times 100$$

(2)

where ${\mathrm{F}}_{\mathrm{i}}$ is the molar flow rate of carbon in the products, ${\mathrm{nc}}_{\mathrm{i}}$ is the corresponding number of carbons, and ${\mathrm{F}}_{\mathrm{in}}^{\mathrm{EtOH}}$ is the molar flow rate at the inlet.

The gaseous products were analyzed in a Trace GC ThermoQuest gas chromatograph with a Carboxen 1010 PLOT column, with argon as carrier gas, detection by TCD, and the following temperature program: 7 min at 45 °C, 25 °C min⁻¹ to 180 °C, and 5 min at 180 °C. The liquid phase was analyzed in a Varian 3300 gas chromatograph using a 10% Carbowax 20M CHR W HP column, helium as carrier gas, detection by TCD, and the following temperature program: 2 min at 50 °C, 25 °C min⁻¹ to 100 °C, and 2 min at 100 °C.

Evaluation of catalytic performance for H₂ production was based on ethanol conversion (${\mathrm{C}}_{\mathrm{EtOH}}$) and product selectivity (${\mathrm{S}}_{\mathrm{i}}$) on dry basis following Equations (3) and (4),

$${\mathrm{C}}_{\mathrm{EtOH}}(\%)=\left(\frac{{\mathrm{F}}_{\mathrm{in}}^{\mathrm{EtOH}}-{\mathrm{F}}_{\mathrm{out}}^{\mathrm{EtOH}}}{{\mathrm{F}}_{\mathrm{in}}^{\mathrm{EtOH}}}\right)\times 100$$

(3)

$${\mathrm{S}}_{\mathrm{i}}(\%)=\left(\frac{{\mathrm{F}}_{\mathrm{i}}}{\sum {\mathrm{F}}_{\mathrm{i}}}\right)\times 100$$

(4)

where ${\mathrm{F}}_{\mathrm{in}}^{\mathrm{EtOH}}$ is the molar flow rate of ethanol at the inlet, ${\mathrm{F}}_{\mathrm{out}}^{\mathrm{EtOH}}$ is the molar flow rate of ethanol at the outlet, and ${\mathrm{F}}_{\mathrm{i}}$ is the molar flow rate of i-species.

3. Results and Discussion

3.1. Catalyst Characterization

3.1.1. Morphology

SEM was used to analyze the microstructural morphology of the catalysts. Figure 1a shows the NaY support, and Figure 1b shows the synthesized Cu/NaY catalyst. Even after the wet impregnation of copper on the NaY support, the polyhedral shapes of the zeolite remained regular and homogeneous, indicating no change in its morphological structure. By contrast, Figure 1c shows the Nb₂O₅ support, while Figure 1d shows the synthesized Cu/Nb₂O₅/Al₂O₃ catalyst. Unlike Cu/NaY, the particles are randomly distributed with various geometries and distinct sizes. These morphological properties were attributed to the nature of the support used, which does not present a well-defined microstructural morphology and evenly distributed particle sizes.

Elemental mapping was performed by EDS to identify chemical elements at the catalyst surface using a magnification of 4000×. Figure 2 shows that all proposed elements were detected, indicating the effectiveness of the proposed wet impregnation process. Additionally, Cu seemed to be better dispersed on NaY compared to Nb₂O₅/Al₂O₃, which presented higher Cu concentrations in some regions. This is strongly linked to the superior textural properties of the zeolite, which allowed for a better distribution of the metal particles.

Table 1. Elemental analysis of Cu/NaY and Cu/Nb₂O₅/Al₂O₃.

Sample	Element (%)
	Cu	Na	Al	Si	O	Nb	C
Cu/NAY	11.90	3.87	3.00	7.76	26.75	-	3.01
Cu/Nb2O5/Al2O3	11.17	-	2.84	-	15.72	10.82	0.88

presents the EDS mass composition of the catalysts. Cu was present in both catalysts in very similar percentages. Furthermore, a Si/Al ratio of 2.58 confirmed the presence of an unmodified NaY zeolite after the wet impregnation synthesis.

3.1.2. Crystallinity

The XRD technique was employed to determine the crystallinity and purity of the synthesized catalysts. The experimental XRD patterns obtained for Cu/NaY and Cu/Nb₂O₅/Al₂O₃ are shown in Figure 3. For the Cu/NaY catalyst, the diffraction peaks at 2θ = 6.18°, 15.61°, 18.66°, 23.64°, 26.98°, and 31.36° are indexed to the cubic crystal structure of NaY, which corresponds to the Fd-3m space group in card #00-039-1380. The diffraction peaks at 2θ = 35.57°, 38.72°, 48.81°, 58.30°, 61.60°, 66.32°, and 68.12° are indexed to the monoclinic crystal system of CuO, referring to card #01-089-5895. The X-ray diffractogram of the Cu/NaY catalyst indicated that there were no modifications in the original crystal structure of the zeolite, probably due to the fine distribution of copper on the zeolite structure [50].

The Cu/Nb₂O₅/Al₂O₃ catalyst presented diffraction peaks at 2θ = 35.57°, 38.72°, 48.81°, and 56.72°, which are indexed to the monoclinic crystal system of CuO in card #01-089-5895. The diffraction peak at 2θ = 67.03° is indexed to the hexagonal crystal system of Al₂O₃ in card #00-013-0373, while the diffraction peaks at 2θ = 22.60°, 28.58°, 36.71°, and 46.23° are indexed to the hexagonal crystal system of Nb₂O₅ in card #00-028-0317. Both copper oxide and niobium pentoxide are present in the diffractogram with significant intensities, indicating that the precursors maintained their defined crystal lattice even after the catalyst synthesis. Also, alumina diffraction peaks are almost imperceptible, indicating a high dispersion of alumina in niobium oxide [51].

The crystallite size has important implications for the rate of molecular diffusion and the contribution of the external surface area to adsorption and desorption rates. For the calculation of the crystallite size of the active phase in both catalysts, the (−111) diffraction peak at 2θ = 35.57° of copper oxide was considered. The result was approximately 55 nm for both catalysts, indicating that the change of support did not cause an alteration in the crystallite size of the active phase. For the calculation of the support crystallite sizes, the (533) diffraction peak at 2θ = 23.64° of NaY and the (100) diffraction peak at 2θ = 28.58° of Nb₂O₅ were considered, resulting in 114 nm and 43 nm, respectively. Kugai et al. [52] concluded that the smaller the crystallite size of the support, the greater the dispersion of the active phase and consequently the higher the catalytic activity, indicating that the catalyst performance strongly depends on the surface area and crystallite sizes.

3.1.3. Textural Parameters

The determination of the textural parameters of the synthesized materials and their precursors is relevant to understand their SRE catalytic performance. The parameters obtained through N₂ physisorption are presented in

Table 2. Textural parameters of the catalysts and precursors.

Sample	Surface Area (m2 g−1)	Pore Volume (cm3 g−1)	Micropore Volume (cm3 g−1)	Pore Diameter (nm)
Nb2O5	71.73	0.3700	0.3200	8.24
NaY	588.49	0.1479	0.0281	1.74
Al2O3	99.26	0.1799	0,0389	7.24
Cu/NaY	210.40	0.1180	0.1114	2.24
Cu/Nb2O5/Al2O3	26.00	0.0617	0.0106	9.56

, while the N₂ adsorption/desorption isotherms are shown in Figure 4. For the Cu/NaY catalyst, the isotherm exhibits characteristics of type IV according to I.U.P.A.C. [53], which is attributed to the presence of micropores associated with mesopores. The formation of a round knee-like feature at the beginning of the isotherm is related to the formation of adsorbed N₂ monolayers inside micropores [54], while the increase in relative pressure improves adsorption as the material mesopores are filled. For the Cu/Nb₂O₅/Al₂O₃ catalyst, the obtained isotherm is of type V, which is also associated with mesoporous materials with weak adsorbate–adsorbent interactions. In both isotherms, there was hysteresis in N₂ desorption, which is characteristic of capillary condensation in mesoporous materials [55]. For Cu/NaY, the hysteresis of type H4 is associated with narrow slit-like pores, while for Cu/Nb₂O₅/Al₂O₃, the hysteresis of type H3 refers to non-rigid aggregates of plate-like particles forming slit pores, typical of non-uniform pore sizes and shapes as observed by SEM.

Concerning the obtained values for the textural parameters, the surface areas of the catalytic supports were similar to other studies involving NaY [56,57], Nb₂O₅ [11,45], and Al₂O₃ [58,59]. For the catalysts, a noticeable reduction in surface area and pore volume was observed, compared to their corresponding precursors. This effect is strongly associated with the possible obstruction of smaller pore diameters (micropores) by deposition of copper oxides inside the catalyst structure, as evidenced by the reduction in micropore volume. However, despite this reduction, a significant variation in textural properties was detected among our catalysts. This reinforces the possible influence of the support type on the catalyst performance, as it plays a fundamental role in the reaction. Cu/NaY has a significantly higher surface area than Cu/Nb₂O₅/Al₂O₃, which is expected to result in a superior catalytic performance of the former compared to the latter. This advantageous characteristic of Cu/NaY has been attributed to the three-dimensional pore structure of zeolite NaY [60].

The FTIR spectra of the synthesized catalysts and both Nb₂O₅ and NaY supports are shown in Figure 5a and Figure 5b, respectively. The band at approximately 3500 cm⁻¹, observed in all spectra, reveals the presence of surface hydroxyl groups in these materials [29,61]. The main NaY vibration modes were identified as the strong band located at 1029 cm⁻¹ and the lower intensity band at 469 cm⁻¹, which correspond to the internal vibrations of the tetrahedral units of the zeolite, while the bands identified at 1150, 794, and 570 cm⁻¹ were attributed to the external linkages between the (Si/Al)O₄ tetrahedra [62]. Sensitive bands did not show significant changes in the Cu/NaY spectrum compared to those of NaY. On the other hand, the band at 570 cm⁻¹, attributed to the polyhedral ring in the zeolite structure, showed a slight alteration that may be associated with the binding of copper. According to previous studies, copper oxide (CuO) bands are present between the wavenumbers of 610 and 500 cm⁻¹ [62,63].

The FTIR spectrum of Nb₂O₅ (Figure 5b) shows strong and broad bands in the region between 500 and 900 cm⁻¹. The band centered at 896 cm⁻¹ is attributed to the Nb-O stretching vibration and the band at 760 cm⁻¹ to the Nb-O-Nb vibration [64]. The narrower band at 1622 cm⁻¹ is attributed to water molecules adsorbed on the Nb₂O₅ surface [65]. The presence of niobium pentoxide was confirmed by the presence of its main vibration modes in the FTIR spectrum of the Cu/Nb₂O₅/Al₂O₃ catalyst (Figure 5a). Alterations were also noted in the intensity of the two bands between 500 and 900 cm⁻¹, and in the resolution between them, when compared to the Nb₂O₅ spectrum. These changes are attributed to the binding with copper, as we are once again referring to the region of copper oxide bands.

3.1.4. Temperature-Programmed Desorption (TPD)

Analyses of both catalysts by TPD-NH₃ are shown in Figure 6. The NH₃ desorption profiles for Cu/NaY and Cu/Nb₂O₅/Al₂O₃ were confined to the 140 °C to 460 °C and 150 °C to 620 °C temperatures ranges, respectively, showing that the support type influenced the acidity of the catalyst. The peak location and wide temperature range for Cu/NaY suggest the presence of sites of weak and intermediate acid strength. By contrast, the higher temperature range for Cu/Nb₂O₅/Al₂O₃ is attributable to the presence of intermediate to strong acid sites. Also, Cu/NaY had a higher concentration of acid sites compared to Cu/Nb₂O₅/Al₂O₃ (

Table 3. Acidity of the synthesized catalysts by TPD-NH₃.

Sample	Chemisorbed NH3 (mmol g−1)	Temperature (°C)
Cu/NaY	1.598	247
Cu/Nb2O5/Al2O3	0.059	336

). In general, the higher acidity of Cu/NaY may be justified by its higher surface area.

The maximum NH₃ desorption temperature for the Cu/NaY catalyst was slightly higher than that for NaY alone, which typically ranges between 150 °C and 250 °C [66,67,68]. This is an indication that the incorporation of Cu into the zeolite structure increased the strength of its acid sites. On the other hand, the impregnation of Cu onto Nb₂O₅/Al₂O₃ did not have a significant influence on the desorption temperature range.

3.1.5. Temperature-Programmed Reduction (TPR)

Both catalysts displayed a wide reduction range in their TPR profiles (Figure 7). However, the position of the maximum reduction temperature of CuO for Cu/Nb₂O₅/Al₂O₃ was shifted to higher values compared to Cu/NaY, which may indicate a greater interaction of Cu with Nb₂O₅/Al₂O₃ compared to NaY. Reduction at temperatures below 300 °C indicated that CuO was dispersed on the catalyst surface with little interaction with the support. Above this temperature, total copper reduction was prevented in both catalytic systems by the interaction between CuO and the support surface [18,69]. Additionally, Cu/Nb₂O₅/Al₂O₃ showed a reduction peak with a maximum around 928 °C, which was attributed to the partial reduction of Nb₂O₅ to NbO₂ [46,70]. The hydrogen consumption of the Cu/NaY catalyst was 6.16 mmol g_cat⁻¹, while that of Cu/Nb₂O₅/Al₂O₃ was 6.38 mmol g_cat⁻¹—a slightly higher value due to the partial reduction of Nb₂O_5.

3.2. Catalytic Performance Evaluation

The main products generated with the Cu/Nb₂O₅/Al₂O₃ catalyst at 300 °C were hydrogen and acetaldehyde, but with a conversion below 10%. Increasing the temperature resulted in an increase in the average conversion, reaching approximately 55%, but the selectivity for hydrogen and acetaldehyde decreased while the selectivity for ethylene increased. For the Cu/NaY catalyst, a similar result was obtained at 300 °C regarding the selectivity for hydrogen and acetaldehyde, but with a higher average conversion close to 55%. Finally, with an increase in the reaction temperature to 450 °C, the conversion increased to 65% but the product distribution was greater, with CO, CO₂, CH₄, and C₂H₄ in detectable quantities.

Figure 8 shows the selectivity data for both Cu/Nb₂O₅/Al₂O₃ and Cu/NaY at 300 °C and 450 °C. The differences between the catalytic performances show that both support and temperature influenced the reaction efficiency. Conversion increased with increasing temperature for both catalysts, but the effect was more pronounced for the Cu/Nb₂O₅/Al₂O₃ catalyst. On the other hand, the selectivity for H₂ and acetaldehyde decreased with increasing temperature, and the quantity of by-products (especially ethylene) increased, with this effect being greater for the Cu/Nb₂O₅/Al₂O₃ catalyst.

For both catalysts at 300 °C, the main reaction was ethanol dehydrogenation (Equation (5)) forming acetaldehyde and H₂, while higher temperatures favored dehydration forming ethylene (Equation (6)). Also, the use of NaY at 450 °C favored parallel reactions forming CO and CH₄ [33]. This is probably due to the presence of more pronounced acid sites in the zeolite at lower temperatures, as observed by TPD-NH₃ analysis of the Cu/NaY catalyst, where acid sites are found between 100 and 450 °C (Figure 6). Also, the high surface area of this catalyst contributed to the greater exposure of acid sites, as well as the catalytically active copper sites on the catalyst surface, promoting parallel reactions that are part of the ethanol reforming reaction pathway. This explains the detection of reaction intermediates such as acetaldehyde.

C₂H₅OH → C₂H₄O + H₂

(5)

C₂H₅OH → C₂H₄ + H₂O

(6)

The selectivity of Cu/Nb₂O₅/Al₂O₃ to C₂H₄ was higher than that of Cu/NaY. The TPD-NH₃ of this catalyst showed acid sites of greater strength due to the desorption of NH₃ at higher temperatures, which may be connected to the higher formation of C₂H₄ [70]. Lorenzut et al. [71] observed the same low selectivity behavior for Cu catalysts supported on ZnO/Al₂O₃. Also, the support seemed to have influenced the reaction pathway, since Nb₂O₅ is an n-type semiconductor that may have driven the selectivity toward ethylene. For the Cu/Nb₂O₅/Al₂O₃ catalyst, the formation of CO₂ was limited at both temperatures, while for Cu/NaY, CO₂ formation was more pronounced at 450 °C, indicating the occurrence of ethanol reforming.

Both catalysts exhibited more stability at 300 °C (Figure 9) because, at 450 °C, C₂H₄ formation accelerated deactivation by coke deposition [72]. For the Cu/NaY catalyst, the carbon balance at 300 °C and 450 °C was 87% and 79%, respectively, whereas for Cu/Nb₂O₅/Al₂O₃, the carbon balance was 99% at 300 °C and 83% at 450 °C. In general, values below 100% may suggest the formation of coke on the catalyst surface [73]. Overall, the results highlight the strong influence of the catalytic support on the steam reforming of ethanol for H₂ production.

The Cu/NaY catalyst showed a higher ethanol conversion rate compared to Cu/Nb₂O₅/Al₂O₃, suggesting that support properties such as surface area, pore volume, and pore size significantly influenced the catalytic performance. The porous structure and higher surface area of NaY may have facilitated copper dispersion and interaction with the reactants, enhancing the catalytic activity. Additionally, zeolites have a structure with a regular arrangement of uniform micropores that promote greater selectivity for H₂ formation. Furthermore, the acidity measurements revealed that the incorporation of copper into NaY increases the strength of the acid sites, as indicated by the higher desorption temperature found for Cu/NaY compared to pure NaY in other studies [74,75,76]. This increased acidity may have contributed to the improvement of the catalytic activity by influencing the adsorption and breakdown of ethanol molecules.

The catalytic performance of Cu/SiO [77] and CuO/ZrO₂ [73] in SRE reactions was evaluated at lower temperatures, e.g., 300 °C and 350 °C. The former showed no catalytic activity, while the latter produced hydrogen (Y = 36%), acetaldehyde (S = 58%), and acetic acid (S = 24%) as the main products, achieving a conversion of 90%. Only at 500 °C did the Cu/SiO catalyst produce acetaldehyde with a conversion of 12%. Cu/Nb₂O₅ resulted in catalytic conversions below 10% at 450 °C [18], producing 60% H₂ and 35% C₂H₄O, on average, along with co-products such as ethylene and ethyl ether (<5%). These results, along with those obtained in this work, highlight the role of the catalytic support in SRE. Therefore, a careful selection of the support is crucial for optimizing the performance of copper catalysts in H₂ production by SRE.

4. Conclusions

The steam reforming of ethanol was viable using copper-based catalysts supported on a NaY zeolite (Cu/NaY) and niobium-aluminum oxides (Cu/Nb₂O₅/Al₂O₃). The physicochemical properties of the two catalysts were different due to differences in the support properties. Nevertheless, copper particles were well dispersed in both catalysts, contributing to achieving better catalytic performances. Both catalysts were active in the steam reforming of ethanol, but Cu/NaY was best for H₂ production at 450 °C, with CO₂ formation remaining constant throughout the reaction course. Therefore, this catalyst has potential for large-scale operations and, with the addition of a small amount of acidity dopants, it may become even more selective for H₂ production by ethanol reforming. Finally, the support effect was demonstrated as a relevant parameter for optimal SRE catalytic performance. The greatest limitation of this work was catalyst shortage, and this leads to the emergence of synthesizing larger quantities of the most efficient NaY-supported copper catalyst for further optimization studies including the addition of acidity dopants to enhance the selectivity for hydrogen production. Additionally, studies on catalyst recovery and reuse must be performed, as well as a detailed economic analysis for process development on an industrial scale.