Water-Splitting Hydrogen Production by an Iron-Modified Bentonite Clay (BC) Laced with Ag Nanoclusters

Abstract

This study examines the use of natural bentonite clay (BC) modified with aluminum and iron as low-cost electrocatalysts for the hydrogen evolution reaction (HER) in an acidic medium. The successful modification process was accompanied by changes in the basal spacing, an increase in the BET surface area and pore volume, and promotion of mass transfer for enhanced hydrogen production. After mechanically blending with silver (Ag) nanoclusters, the more sustainable and higher surface-area exhibiting Fe-clay composite showed about 21% increase in hydrogen production and a 134% increase in turnover frequency (TOF) relative to the reversible hydrogen electrode (RHE). The Ag–Fe–Clay composite achieved a current density of 274 mA/cm², similar to that of the commercial palladium black catalyst at −1.20 V (vs RHE). This work demonstrates the potential of Ag-laced Fe-modified bentonite clay as a nonexpensive sustainable alternative to the state-of-the-art highly expensive (HER) catalysts with an industrial output of about 1018 mA/cm² at −2.22 V vs RHE applied potential. Further work is ongoing to evaluate the performance of the clay with other metallic catalytic agents.

Introduction

Increasing energy demands and environmental pollution associated with fossil fuels have led to a focus on developing alternative technological solutions. Hydrogen has been considered as the future replacement of fossil fuels. , It is attractive based on the fact that it can be produced from abundant resources such as biomass and water and produces zero-carbon emissions after combustion. This has led to increasing efforts in its production and storage for a complete economy. In addition, the hydrogen evolution reaction (HER) through electrochemical water splitting can be performed at room temperature and atmospheric pressure, making it the most efficient, less expensive, and eco-friendly process, among most other hydrogen generation processes (e.g., coal gasification and methane reforming). ,

The nature of the electrocatalyst employed during the water splitting for hydrogen production is very crucial due to its impact on production efficiency. Noble metal platinum (Pt) and its related metals, such as palladium (Pd), exhibit excellent HER electrocatalysis performance (i.e., high reaction rates at very small overpotential). However, these materials are very expensive due to their scarce raw materials and cost of production. As a result, the development and use of highly efficient and inexpensive catalysts become a necessity to achieve a sustainable economy. − Non-Pt metals, such as nanostructured materials (e.g., carbon nanomaterials), and transition-metal compounds (e.g., 2D transition-metal chalcogenides nanomaterials) are considered some of the most attractive alternative catalysts − in terms of lower cost, electrocatalytic activity, and stability. Nanomaterials offer a high surface area-to-volume ratio. This makes them exhibit a high density of exposed active sites. With reduced diffusion paths, electron transfer increases, effectively promoting the enhancement of electrocatalytic activity.

Clays are fascinating naturally available materials in many parts of the world with several beneficial properties such as high-surface areas, plasticity, swelling, ion exchange capacities, surface tunability, and adsorption. Many clays exhibit different morphologies, particle sizes, and compositions. As a result, they have been exploited for many functions, including water treatment, − drug delivery, − flame retardance, and as fillers in papers, cosmetics, paints and food packaging. For example, a halloysite clay possesses a nanotubular structure, − while bentonite displays a sheet-like structural feature. Another fascinating thing about clay structure is the fact that it can be functionalized or tuned to achieve better functional performance. For example, bentonite can be modified to enhance its catalytic functions. − For example, in 2017, Zhang et al. reported the significantly superior performance of Au–Ti–Ce-modified bentonite clay toward hydrogen production through methanol steam reforming with a 72% methanol conversion and 99% of hydrogen selectivity at 350 °C, compared to other tested catalysts. Modification involves the replacement of some cationic guest ions within the clay interlayers with some transition-metal ions, such as those of Al, Au, Ti, Cu, Ce, Cr, and Fe. Subjection of the modified clay to a thermal environment causes a transformation with enhanced catalytic properties. In terms of the water splitting process, it has been reported that clay materials inherently favor the alkaline medium oxygen evolution reaction (OER) better than the acidic hydrogen evolution process (HER) counterparts due to their intrinsic catalytic compositional make-ups and stability in the respective media. However, doping the clay materials through the inclusion of transition metals within their structure, alongside the use of polymeric proton exchange membrane materials, may provide a solution for stability and further enhancement of efficiency in the acidic medium.

This report exploits the potential of bentonite clay successfully as an electrocatalyst for water-splitting hydrogen production in an acidic medium. Previously, many modified clay studies for hydrogen production have focused on high-energy-demanding methanol steam reforming, biomass pyrolysis, NaBH₄ hydrolysis, and acetic acid reforming. However, our present work demonstrates a relatively lower-energy-demanding electrochemical water splitting in an acidic medium using Fe-modified bentonite clay (Fe-BC). In addition, unlike many Ag-modified clay reports that focused on photocatalysis or antibacterial applications, our work presents an enhancement of the hydrogen evolution reaction activity of Fe-BC through composting with silver nanoparticles, achieving better overpotential, lower Tafel slopes, and better charge transfer efficiency. Furthermore, the composite material was benchmarked against the commercial Pd black product, providing information about the overpotential and, thus, energy input relativity.

Initially, the clay was modified with Al³⁺ as well as Fe²⁺ ions to evaluate their effects on surface area properties, thermal stability, morphology, as well as electrocatalytic performance. However, for sustainability purposes, in terms of a cheaper alternative coupled with a higher surface area and thermal stability, Fe-modified clay was found to be better and, therefore, was chosen over that of the Al counterpart for blending with the noble metal silver nanoclusters for performance comparison and relatively benchmarking efficiency with the commercially available palladium black.

Experimental Section

Chemicals

All chemicals were of analytical grade, purchased from Sigma-Aldrich, and used without further purification. Commercially available bentonite clay was purchased from Sigma-Aldrich and used as a precursor for the preparation of the modified clays (BCs). Aluminum sulfate hexadecahydrate (Al₂(SO₄)₃·16H₂O), iron­(II) sulfate heptahydrate (FeSO₄·7H₂O), and sodium hydroxide (NaOH) were used to prepare the metal solutions.

Preparation of Al and Fe Solutions

Modified materials were prepared from natural bentonite with a structural formula of IV­[Si_7.43Al_0.57]­VI­[Al_2.96Fe_0.73Mg_0.24]. (Na_0.71K_0.21Ca_0.01; O₂₂). BCs were prepared by cationic exchange reactions between the natural bentonite and hydroxy-oligomeric solutions of Al³⁺ or Fe²⁺. The direct addition of dry clay to the metal ion solution and the use of ultrasonication to speed up the homogenization and intercalation steps were adopted from previous studies. The precursor salts, namely, aluminum sulfate hexadecahydrate (Al₂(SO₄)₃·16H₂O) and iron­(II) sulfate heptahydrate (FeSO₄·7H₂O), were used as sources of Al and Fe, respectively.

The Al and Fe solutions were prepared by slowly adding 0.4 M NaOH solution to either 0.4 M Al or Fe solutions, using Al₂(SO₄)₃·16H₂O and FeSO₄·7H₂O as sources, at a NaOH/Al or NaOH/Fe molar ratios of 2:1 (post reaction pH: Al = 3.32, Fe = 7.61). The solutions underwent continuous vigorous stirring at room temperature for 2 h, followed by ultrasonication for 30 min. The intercalation procedure was performed using the dry modification method, i.e., without prior swelling of the clay: dry bentonite clay (1.0 g per 25 mL modification solution), equivalent to a metal/clay ratio of 10 mmol/g, was subjected to constant stirring at room temperature for 2 h, followed by 30 min of ultrasonication. The resultant intercalated clay underwent centrifugation at 4500 rpm, followed by washing with distilled H₂O, five times in 50 mL quantities. The recovered clay was then oven-dried for 3 h at 100 °C and finally calcined at 500 °C for 2 h.

Ag–Fe–BC Preparation

The silver particles were synthesized by heating a mixture of 50 mL of 0.01 M AgNO₃ solution with a solution of 50 mL (0.1 M NaOH solution in ethanol) from room temperature (22 °C) to boiling phase. This was followed by boiling of the resulting solution for an additional 10 min. The particles were separated from the supernatant solution via decantation and washed several times with water before being dried under ambient conditions for 12 h. Afterward, 13.8 mg of the as-synthesized Ag was mechanically mixed with 4.4 mg of Fe-BC to make the Ag–Fe–BC composite.

Characterization

The textural properties, morphological characteristics, and elemental compositions of the bentonite clay and the prepared metal-modified BCs were analyzed using BET surface probing system and field-emission scanning electron microscope with an energy-dispersive X-ray spectrometer (FESEM–EDS, JEOL JSM-7800F, Japan). The morphology and the clay layer spacing were evaluated with a transmission electron microscope (TEM, JEOL 2100 F) operating at an accelerating voltage of 200 kV. Ethanol was used to disperse the clay samples, which were sonicated in a bath sonicator for 5 min. Subsequently, a drop of the clay dispersion was deposited on a carbon-coated copper grid and left to air-dry before analysis. The Brunauer–Emmett–Teller (BET) nitrogen adsorption–desorption isotherms, at −196 °C, of raw bentonite clay and BCs were measured using a Micromeritics 3030 TriStar II Plus instrument. The sample preparation procedure began by allowing an inert gas to flow over the samples (∼0.2 g) at 90 °C for 1 h, followed by degassing under vacuum at 150 °C overnight to eliminate any adsorbed gases and moisture. The specific surface areas (SSA) of the samples were obtained by the BET method, micropore volume by the t-method, and the total pore volume from the amount of nitrogen adsorbed at a relative pressure of 0.95.

X-ray diffraction (XRD) was performed using a PAN analytical X’Pert PRO instrument (Netherlands) equipped with nickel filtered Cu Kα radiation (λ = 1.5406 Å), and an adjustable slit operating at 45 kV/40 mA, recorded over 5°–80° of 2θ range at the scanning speed of 0.02°/s, to determine the basal spacing of the clay and BCs samples, using the Bragg equation, as depicted in equation 1

$$2d\sin \theta =n\lambda$$ 1

where d is the basal spacing within the clay structure (nm), θ is the Bragg diffraction angle, n = 1, as the order of reflection, and λ = 1.5405980 is the wavelength of the incident X-ray.

Thermogravimetric measurements of the natural clay and BCs were conducted using a Perkin–Elmer Thermogravimetric analyzer (Pyris 1 TGA), under a nitrogen atmosphere at a gas flow rate of 20 mL/min and heating rate of 20 °C/min, from room temperature to 900 °C. TGA experiments allowed us to determine the thermogravimetric (TG) and differential thermogravimetric (DTG) curves of bentonite clay and BCs.

Electrochemical Characterization

Linear sweep voltammetry (LSV) studies were performed in 0.5 M H₂SO₄ within a potential window of +0.0 V and −0.80 V vs Ag/AgCl (3 M KCl) at a scan rate of 10 mV/s. The electrochemical impedance spectroscopy (EIS) analysis was carried out at 1 × 10⁵ Hz to 0.1 Hz frequency, and 10 mA amplitude. All experiments were carried out using a three-electrode cell (TEC) connected to an iR-compensated Metrohm Autolab Potentiostat (PGSTAT302) equipped with Nova software version 2.1.8. The TEC consists of platinum wire, Ag/AgCl (3 M KCl) and as-synthesized materials on graphite/Nafion composite as counter, reference, and working electrodes, respectively. All data were converted to RHE data using the relation in eq .

$${\mathrm{E}}_{\mathrm{R}\mathrm{H}\mathrm{E}}={\mathrm{E}}_{\mathrm{A}\mathrm{g}/\mathrm{A}\mathrm{g}\mathrm{C}\mathrm{l}}+0.0591\mathrm{p}\mathrm{H}+0.197$$ 2

The working electrode was prepared by mixing 35 mg of catalysts (BC, Al-BC, and Fe-BC) or 4.4 mg of Fe-BC/13.8 mg of Ag nanocluster in Nafion solution (0.1 mL), followed by casting on a graphite rod (geometrical diameter: 6 mm) as a paste, drying in an oven at 80 °C (5 min) and subsequently cooling to room temperature (22 °C) before analysis. The analysis was done in an unstirred solution. Diluted H₂SO₄ (0.5 M, 100 mL, pH 0.0) was used as the electrolyte. The effective capacitance, C_eff, electrochemical surface area (ECSA), roughness factor, Tafel slope, mass activity, specific activity, and turnover frequency (TOF) were estimated using the following relations (eqs –).

$${\mathrm{C}}_{eff}=Y^{0(1/n)}{(1/Rs+1/R_p)}^{(n-1)/n}$$ 3

$$\mathrm{E}\mathrm{C}\mathrm{S}\mathrm{A}={\mathrm{C}}_{\mathrm{e}\mathrm{f}\mathrm{f}}/(\mathrm{C}\mathrm{s}=40\mu \mathrm{F})$$ 4

$${\mathrm{R}}_{\mathrm{f}}=\mathrm{E}\mathrm{C}\mathrm{S}\mathrm{A}/{\mathrm{A}}_{\mathrm{G}\mathrm{e}\mathrm{o}.}$$ 5

$$\mathrm{T}\mathrm{a}\mathrm{f}\mathrm{e}\mathrm{l}\mathrm{s}\mathrm{l}\mathrm{o}\mathrm{p}\mathrm{e}:\eta =\mathrm{a}\pm \mathrm{b}\log (-j)$$ 6

$$\mathrm{m}\mathrm{a}\mathrm{s}\mathrm{s}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{i}\mathrm{t}\mathrm{y}={\mathrm{I}/\mathrm{m}}_{\mathrm{c}\mathrm{a}\mathrm{t}.}$$ 7

$$\mathrm{s}\mathrm{p}\mathrm{e}\mathrm{c}\mathrm{i}\mathrm{f}\mathrm{i}\mathrm{c}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{i}\mathrm{t}\mathrm{y}=\mathrm{I}/\mathrm{E}\mathrm{C}\mathrm{S}\mathrm{A}$$ 8

$$\mathrm{T}\mathrm{O}\mathrm{F}=\mathrm{J}\times {\mathrm{A}}_{\mathrm{G}\mathrm{e}\mathrm{o}.}/{2\mathrm{n}}_{\mathrm{c}\mathrm{a}\mathrm{t}}\mathrm{F}$$ 9

where Y⁰ = CPE element, Rs = solution resistance, Rp = polarization resistance, n = CPE constant, Cs = specific capacitance, A_Geo. = Geometrical area of the electrode, η = overpotential, a = Tafel constant relating to exchange current density, b = Tafel slope, j = current density, I = current, m_cat. = mass of active catalyst, n_cat. = mole of active catalyst, and F = Faraday’s constant (96,485 C/mol).

Results and Discussion

XRD analysis Figure presents the XRD results of the bentonite and modified clays (BCs) in the 2θ angle range of 5°–80°. The analysis of XRD spectra provided the corresponding basal spacing values (Table ).

1.

Crystalline, textural, and thermal stability characterizations of the as-synthesized materials. (a) XRD patterns and (b) nitrogen adsorption/desorption; (c) thermogravimetric and (d) differential thermogravimetric curves.

1. Physicochemical Characteristics and Textural Properties of Natural and Modified Bentonite Clay Samples by XRD and BET.

sample	BET SA (m2/g)	Langmuir SA (m2/g)	average pore diameter, dpore (nm)	average pore volume, Vpore (cm3/g)	micropore volume, Vmic (cm3/g)	basal spacing (d–001) (nm)	basal spacing (d–001) (Å)
bentonite	9	45	6.06 ± 0.29	0.028 ± 0.005	0.0005	1.19	12
Al-BC (500 °C)	12	60	6.16 ± 0.43	0.031 ± 0.01	0.07	0.96	10
Fe-BC (500 °C)	37	208	3.72 ± 0.21	0.083 ± 0.02	0.003	0.96	10

The natural clay exhibits typical peaks at 2θ = 7.5° (001), 19.2° (020) (montmorillonite component), 26.6° (101), 35.2° (110), and 62.08° (211), corresponding to quartz, and 28.33°(104) and 54.4° (116) indicating presence of calcite, according to JCPDS card numbers 00–013–0135, 00–046–1045, and 00–005–0586, respectively. No distinct crystalline Al₂O₃ or Fe₂O₃ phases were detected, suggesting the modifiers exist as amorphous or are present at a concentration below the detection limit of the characterization instrument. These observations were similar to a previous literature report. Similarity in the XRD peaks was observed for all materials, except for the peak at 7.5 (2θ), which shifted to a higher (2θ) site (8.20) for Al-BC and 8.90 for Fe-BC, respectively. This shift in the crystalline structure caused about 19% change in the basal spacing from 1.2 nm (12 Å) to about 1.0 (10 Å), as shown in Table . The decrease in the basal spacing associated with the BCs, compared to that of the natural clay, indicates collapsed pillars within the clay-modified structures. This collapse of the pillared clay structure is often associated with the thermal synthesis condition at higher temperature, typically ≥500 °C. During high heat treatment, inherent thermally induced mechanical pressure may occur, lowering the intensity and causing a shift in the basal space spot to the left rather than to the right of the 2-theta crystallographic plane. This causes changes to the typical structural ordering (collapse of the pillar) as explained by Pichowicz and Mokaya in their report.

BET Analysis

The textural properties of natural and modified bentonite clays are presented in Figure b and Table . The textural properties include BET-specific surface area (S _BET), porosity, total pore volume (V_p), micropore volume (V_mic), and average pore diameter (d_pore). A type IV­(a) isotherm was observed for all tested materials according to the IUPAC classification. This confirms that all samples were porous in nature. Also, all materials revealed an H3 hysteresis loop similar to the one depicted in Denohue’s report (Donohue and Aranovich, 1998). However, the modification of the bentonite clays (with Al and Fe) increased the S _BET to some extent, suggesting the modified clays exhibited better surface properties. The lower N₂ adsorption by natural clay denotes less contribution from the micropores than that of the BCs. Desai et al. observed a type IVa and an H2 hysteresis loop for natural bentonite for conical pores, together with an H4 hysteresis with slit-like pores for Cr–BCs. The natural clay and BCs exhibit mesoporous features, with pore diameters of 6.06 nm for natural bentonite and as low as 3.7 nm for Fe-BC. Al-clay shows little change in the pore dimension relative to that shown by the natural clay, causing a slight change in the specific surface area (about 37% vs 315% for Fe-clay). Upon modification, the surface area of bentonite increased by 4.2 times for Fe–BCs post calcination, higher than that of Al-BC by 204%. The micropore and total volumes of the BCs increased with an increase in surface area, while the pore diameters decreased. Increase in the total pore volume (V_pore) ranged from 13–199%, and this was from 0.028 cm³/g for the natural bentonite to 0.031 and 0.083 cm³/g for Al- and Fe–BCs, respectively. A good porosity distribution of both micropores (<2 nm) and mesopores (2–50 nm) is highly beneficial toward mass transfer of reactants and products. Overall, the Fe-clay exhibited the highest surface area among the tested materials. Compared to other previous reports, the relatively smaller BET surface areas reported in this study were due to a number of reasons. First, the commercial bentonite used in this study had a low initial surface area of 9 m²/g due to dense particle aggregation, unlike purified montmorillonites that typically start at 50–100 m²/g. Also, the use of higher calcination temperature, viz., 500 °C, may be suboptimal for the modified BCs; lower temperatures (350–400 °C) might preserve more microporosity. Furthermore, the presence of some metal oxide particles may necessitate blocking of micropores through deposition on external surfaces rather than intercalating. Nonetheless, despite the modest BET increase (from 9 to 37 m²/g for Fe-BC), the material showed significant improvement in the HER performance, suggesting other factors, in addition to the surface area, played important roles.

TGA Analysis

Figure c,d shows the TG and DTG curves of the natural and modified clays. As a general result, the materials exhibited three mass losses. The first mass loss was detected from room temperature to ca. 150 °C due to the removal of physically adsorbed water from the surface of the clay sheets, as well as water weakly attached either within the BC galleries or bonded to the polycations or modifiers of the BCs. The second mass loss was observed between 382 and 500 °C, resulting from the removal of H₂O molecules aligned with the modifiers. This indicates that the clay’s structure of the material was maintained, suggesting good thermal stability of the BCs due to the clay’s crystalline structure. However, at this stage, the natural clay underwent higher mass loss than the BCs. The third mass loss (after 500 °C) is typically associated with the destruction of the clay’s crystalline structure alongside dehydroxylation processes. , However, while the clay and Al-BC showed a loss in mass at 600 and 800, respectively, the Fe-BC exhibit negligible loss at this amorphization stage, suggesting its relatively stronger thermal stability compared to those of the clay and Al-BC.

SEM and EDS Analyses

The morphology and elemental composition of the clay samples are presented in Figure a–c and Table , respectively. The Al modification process partially transformed the natural clay, while the Fe counterpart showed some full surface transformation (Figure a–c). As revealed in Figure a–c, the bentonite clay exhibits a flaky appearance, indicating a typical layered structure with clustered sheets that have a few layers protruding from the surface of the clusters. After modification, the samples exhibit a more porous morphology and loosely packed clusters with rough surfaces, probably resulting from the deposition of metal oxide crystallite particles on the clay sheet surfaces or within the sheets, the effect being largely found with Fe-clay.

2.

Morphology analysis of the clay samples. (a) Pristine bentonite clay; (b) Al-BC (500 °C); (c) Fe-BC (500 °C). BC = modified clay.

2. Elemental Composition of the Natural Clay and the BC Catalysts.

sample	Na (%)	Mg (%)	Al (%)	Si (%)	K (%)	Ca (%)	Ti (%)	Cu (%)	Fe (%)	Si/Al
bentonite	1.5	1.1	8.5	23	0	0	0	0	1.0	2.7
Al-BC	1.0	0.9	12	14	0	0.2	0	0	0	1.1
Fe-BC	0.1	0.5	5.9	15	0.1	0.1	0	0	14	2.6

The near similarity in morphology of the Al-clay and the pristine clay may stem from the fact that a slight increase in the Al content was observed (12.4 vs 8.5%) after modification, compared to that of the Fe-Clay (13.5 vs 1%), as shown in the EDS elemental composition analysis results (Table ). The modification process is expected to lead to the replacement of some exchangeable cations within the natural clay, namely, Na⁺, Mg²⁺, Al³⁺, Si⁴⁺, K⁺, Ca²⁺, and Fe³⁺, with the intercalated cations (Al³⁺, Fe^2/3+). According to Table , both the Al and Fe contents increased after modification, while the contents of the exchangeable cations, such as Na, Mg, and Si, decreased. This indicates a successful exchange between the exchangeable interlayer cations of the natural bentonite and the Fe and Al metal oxide cations from the modification solution. The slight presence of Ca and K in the clay after modification may be as a result of their introduction from the modification solution precursors.

Hydrogen Production Activity

The results of the linear sweep voltammetric analyses of the tested electrocatalysts are listed in Figure a. As shown in Figure a, the onset potential occurred at different overpotentials, viz: 0.4 < 0.42 < 0.43 < 0.44 V (vs RHE) for all tested materials, i.e., graphite/Nafion/Clay-Al < graphite/Nafion/Clay-Fe < graphite/Nafion/Clay < graphite/Nafion, respectively. However, variation in current densities occurred as the applied potential increased from about −0.40 to −0.60 V vs RHE. According to Figure a, the Al-clay produced the highest current density at both −0.4 V (−2.2 mA/cm²) and −0.5 V (3.6 mA/cm²) (vs RHE), while the bare electrode (graphite/Nafion) displayed the least value (−1 mA/cm² and −2.2 mA/cm², respectively) (Table ). In contrast, at the current density of −5 mA/cm², the Fe-clay material overtook the Al-clay system, displaying an overpotential of 0.5500 vs 0.5525 V (vs RHE). At this point onward, the Fe-clay continued to sustain the lead in terms of higher hydrogen current outputs compared to the rest of the other tested materials. For example, at −0.6 V (vs RHE), the Fe-clay system exhibited a higher current density of about 7.32, 29.41, and 20.78% relative to those of Al-clay, clay, and bare electrode (graphite/Nafion), respectively (Table ). Overall, the Fe- or Al-modified clay exhibited superior hydrogen production compared to the native clay. For example, Al or Fe-clay showed 37.5, 12.5, 28, 57% or 21.43, 20.59, and 29.41% increase over that of the clay-system at −0.4, −0.5, and −0.6 V vs RHE, respectively.

3.

Hydrogen evolution reaction characterization of bentonite clay and its Al and Fe composites on graphite surface. (a) LSV; (b) Tafel plots; (c) Nyquist plots; (d) EIS circuit diagrams: (i) graphite/Nafion, (ii) graphite/Nafion/clay, (iii) graphite/Nafion/Al-clay; (iv) graphite/Nafion/Fe-clay.

3. LSV Data from −0.4 to −0.6 V (vs. RHE) and Corresponding Current Density Values.

material	overpotential (V) vs RHE	current density (mA/cm2)
graphite/Nafion/clay/Al	–0,4	–2,2
graphite/Nafion/clay/Fe	–0,4	–1,8
graphite/Nafion/clay	–0,4	–1,6
graphite/Nafion	–0,4	–1
graphite/Nafion/clay/Al	–0,5	–3,6
graphite/Nafion/clay/Fe	–0,5	–3,4
graphite/Nafion/clay	–0,5	–2,8
graphite/Nafion	–0,5	–2,2
graphite/Nafion/clay/Al	–0,6	–7,175
graphite/Nafion/clay/Fe	–0,6	–7,7
graphite/Nafion/clay	–0,6	–5,95
graphite/Nafion	–0,6	–6,375

In terms of corrosion resistance robustness, Al-clay showed a smaller Tafel slope value and thus, higher corrosion resistance characteristics than either Fe or native clay, as shown in Figure b. The order of increasing Tafel slope is graphite/Nafion/Al-clay (82 mV/dec) < graphite/Nafion/Fe-clay (149.01 mV/dec) < graphite/Nafion/clay (158.72 mV/dec) < graphite/Nafion (161.97 mV/dec). The superiority of the Al-clay over that of the Fe or pristine clay counterpart lies in the fact that Al₂O₃ naturally has a higher surface passivation property. In the context of charge transfer characteristics, the order of increasing resistance is graphite/Nafion (5.48 kΩ) < graphite/Nafion/Fe-clay (5.77 kΩ) < graphite/Nafion/Al-clay (11.1 kΩ) < graphite/Nafion/clay (13.2 kΩ), making graphite/Nafion/Fe-clay the most robust in terms of electron exchange among the tested electrocatalysts besides the control.

Benchmarking with Pd Black and Ag-Composite

Figure 4a–d reveals the comparative LSV, Tafel plots, and charge transfer resistance (Rp) data of the Fe-modified clay system, commercial Pd black, and Fe-modified clay laced with silver. As shown in Figure a, an enhancement in current density, about 21% increase, was observed after lacing the Fe-clay with silver particles (1:3 w/w Fe/Ag) at 0.6 V overpotential (vs RHE) (J_Fe = 7.5 mA/cm², J_Ag = 9.10 mA/cm²). This value corresponds to about 3.30% that of the commercial Pd black catalyst (274 mA/cm² at 0.6 V vs RHE). Silver was added to see any possible improvement on the Fe-clay, which exhibited about 2.7% current density output of that shown by the Pd black catalyst. To achieve equal performance as the commercial Pd catalyst, the potential scan was increased for the Fe/Ag system up to −1.2 V vs RHE. This potential range falls within the typical voltage employed in a commercial water-splitting electrolyzer.

4.

Comparative electrochemical data. (a,b) LSVs; (c) Tafel plots; (d) Nyquist plots. Support = graphite/Nafion, scan rate = 10 mV/s.

It was notably observed that the Fe/Ag system showed comparative performance with the commercial Pd black catalyst at −1.2 V (vs RHE), exhibiting a current density of 274 mA/cm². This is approximately equivalent to about 50% increase in energy/electricity consumed relative to that consumed by the commercial Pd black catalyst (Table ).

4. Estimation of Energy Consumed (E) per Hour at Various Comparable Overpotentials.

catalyst system	time/h	current (A)	voltage (V)	time/s	P = I × V (W)	E = P (W) × t (s) (J)	E = P (kW) × t (h) (kWh)
Pd	1	0.274	0.6	3600	0.1644	591.84	0.0001644
Fe	1	0.0075	0.6	3600	0.0045	16.20	0.0000045
Fe/Ag	1	0.0091	0.6	3600	0.0055	19.66	0.0000055
Fe/Ag	1	0.274	1.2	3600	0.3288	1183.68	0.00033

Thus, further tuning of either the Fe-clay or its Ag nanocomposite structure may be necessary to improve the efficiency of HER activity, in terms of shifting the energy input, i.e., the electricity consumed, to a very low potential. This is important for achieving more sustainability advantages during industrial upscaling. In the context of Tafel plot analysis, the order of increasing Tafel slope was graphite/Nafion/Pd < graphite/Nafion/Fe-clay/Ag < graphite/Nafion/Fe-clay/Ag_ext < graphite/Nafion/Fe-clay (Figure c). This means, in terms of reaction rate kinetics, the Pd system exhibited the fastest HER activity, while the Fe-clay electrode showed the least, supporting the information obtained from the LSV analysis. However, this also means a higher corrosion resistance for the Fe-clay electrode than for the Pd or Ag composite system. In a similar vein, the results of the polarization resistance analysis obtained from the Nyquist plot are listed in Figure d. Typically, the higher the polarization resistance, the greater the resistance to charge transfer, implying a higher corrosion resistance. As revealed in Figure d, the order of increasing polarization resistance was graphite/Nafion/Pd < graphite/Nafion/Fe-clay/Ag < graphite/Nafion/Fe-clay, further supporting the Tafel analysis reports.

Current at Constant Applied Potential (Bubble Effects)

To make an industrial application feasible, the behavior of the current at a constant applied potential should be evaluated. Figure a–d shows the chronoamperometric analysis (CAA) of the graphite/Nafion/Fe-clay/Ag electrode at −0.55 and −2.30 V vs RHE. These potential values were chosen due to their capability to generate bubbles during HER activity. CAA is a powerful tool which enables elucidation of current stability or kinetic reaction information. Thus, it reveals bubble effects on the stability of the current or its dynamics. For example, a reaction with a faster rate produces gas bubbles which have a negligible time of residence at the electrode surface, whereas a slow kinetic reaction rate tends to produce bubbles with significant residence time. The latter stays longer at the electrode’s surface from its adsorption point, through growth and finally desorption. This causes the current to produce a zigzag wave-like signal. The larger the space between two wave amplitudes, the slower the rate of reaction at the electrode surface and vice versa. Thus, a stable current wave signal represents an electrode surface saturated with bubbles experiencing simultaneously spontaneous adsorption (including growth) and desorption mechanisms in such a way that all active sites are shielded from further production of hydrogen. However, deviation in the current stability signals implies a kinetic process where the rate of adsorption and desorption leaves some space for hydrogen production as the reaction proceeds with time. This means that at a fixed applied potential, the hydrogen current can increase with time depending on the behavior of the bubbles (adsorption, growth, and desorption) at the surface of the electrode.

5.

Impact of bubbles on current and reaction kinetics. (a) Chronoamperometric analysis at −0.55 V vs RHE; (b) enlargement of (a) within 300 and 600 time frame; (c) chronoamperometric analysis at −2.3 V vs RHE; (d) enlargement of (c) within 300 and 310 time frame −0.05 A.

In Figure a–d, it is obvious that hydrogen currents increased with time at both −0.55 V and −2,3 V vs RHE on the Fe-Clay/Ag surface. According to Figure b–d, six bubbles were produced in 5 min at the lower potential (−0.55 V), while 14 bubbles were produced in 10 s at a higher applied potential (−2.3 V), implying the direct dependency of the rate of reaction on the applied potential.

Mechanism of Hydrogen Formation on the Fe-Clay/Ag Surface

The results of redox properties and surface electron exchange kinetics are presented in Figure a–d for Pd black, Fe-clay, as well as Fe-Clay/Ag. All tested materials exhibited quasi-reversible surface electron exchange properties since the difference between the oxidation and reduction peak potentials deviated from 59 mV at 25 °C for the expected one-electron exchange surface transfer (Figure a–c). Also, since the peak potential increases with the scan rate, the electrocatalysts favored the diffusion-controlled electrochemical interactions at their surfaces (Figure a–c). During calcination of the clay with the Fe­(II) precursor, Fe²⁺ was converted to Fe³⁺ (evidence of its existence was observed during the forward scan of the CV, which revealed a reduction peak rather than an oxidation peak (Figure d)). The signal was obtained after the background (support) signal subtraction. In terms of hydrogen evolution efficiency, this means that the reduction of these Fe³⁺ species will compete with that of the hydrogen-ion species, causing a higher onset potential for hydrogen production. This explains why Fe-Clay exhibited a higher overpotential than that of the commercial Pd Black, which was in a zerovalent metal state (Figure a–b). To support this claim, the Fe-clay was laced with the zerovalent Ag metal particles (Fe/Ag = 1:3 w/w). As shown in Figure d, the silver particles dominated the surfaces of the composite. As revealed in Figure a, the Fe-clay/Ag composite electrode showed higher hydrogen production (21% increase in production at −0.60 V vs RHE) than that produced by the pristine Fe-clay electrode.

6.

Redox and surface electron exchange properties of the tested electrocatalysts. (a) Pd black; (b) Fe-clay; (c) Ag; (d) combined CV of all catalysts at 10 mVs^–1. Support: graphite/NaFION, electrolyte: 0.55 M H₂SO₄.

Typically, hydrogen production involves two major pathways, the adsorption of the hydrogen ions on the surface of the catalyst followed by the reduction of the adsorbed hydrogen ions to adsorbed hydrogen atoms (Vomer step), the interaction of protons (H⁺) with adsorbed hydrogen atoms (H) to form molecular hydrogen molecules (H₂ gases) which desorbs from the surface (Volmer-Heyrovsky step), or combination of the adsorbed hydrogen atoms to form molecular hydrogen molecules (Vomer-Tafel step) (eqs –).

$${2\mathrm{H}}^{+}+Cat.{[]}^{Surface}\to Cat.{[{2H}^{+}]}_{Ads}^{Surface}(Volmer)$$ 10

$$\mathrm{C}\mathrm{a}\mathrm{t}.{[{2\mathrm{H}}^{+}]}_{\mathrm{A}\mathrm{d}\mathrm{s}}^{\mathrm{S}\mathrm{u}\mathrm{r}\mathrm{f}\mathrm{a}\mathrm{c}\mathrm{e}}{+2\mathrm{e}}^{-}\to \mathrm{C}\mathrm{a}\mathrm{t}.{[2\mathrm{H}]}_{\mathrm{a}\mathrm{b}\mathrm{s}\mathrm{o}\mathrm{r}\mathrm{b}}^{\mathrm{S}\mathrm{u}\mathrm{r}\mathrm{f}\mathrm{a}\mathrm{c}\mathrm{e}}(\mathrm{V}\mathrm{o}\mathrm{l}\mathrm{m}\mathrm{e}\mathrm{r})$$ 11

$$\mathrm{C}\mathrm{a}\mathrm{t}.{[2\mathrm{H}]}_{\mathrm{A}\mathrm{d}\mathrm{s}}^{\mathrm{S}\mathrm{u}\mathrm{r}\mathrm{f}\mathrm{a}\mathrm{c}\mathrm{e}}{+2\mathrm{H}}^{+}+{2\mathrm{e}}^{-}\to \mathrm{C}\mathrm{a}\mathrm{t}.{[]}_{\mathrm{D}\mathrm{e}\mathrm{s}\mathrm{o}\mathrm{r}\mathrm{b}}^{\mathrm{S}\mathrm{u}\mathrm{r}\mathrm{f}\mathrm{a}\mathrm{c}\mathrm{e}}+{2\mathrm{H}}_2(\mathrm{H}\mathrm{e}\mathrm{y}\mathrm{r}\mathrm{o}\mathrm{v}\mathrm{s}\mathrm{k}\mathrm{y})$$ 12

$$\mathrm{C}\mathrm{a}\mathrm{t}.{[2\mathrm{H}]}_{\mathrm{A}\mathrm{d}\mathrm{s}}^{\mathrm{S}\mathrm{u}\mathrm{r}\mathrm{f}\mathrm{a}\mathrm{c}\mathrm{e}}\to \mathrm{C}\mathrm{a}\mathrm{t}.{[\mathrm{H}..\mathrm{H}]}_{\mathrm{A}\mathrm{d}\mathrm{s}}^{\mathrm{S}\mathrm{u}\mathrm{r}\mathrm{f}\mathrm{a}\mathrm{c}\mathrm{e}}\to \mathrm{C}\mathrm{a}\mathrm{t}.{[]}_{\mathrm{d}\mathrm{e}\mathrm{s}\mathrm{o}\mathrm{r}\mathrm{b}}^{\mathrm{S}\mathrm{u}\mathrm{r}\mathrm{f}\mathrm{a}\mathrm{c}\mathrm{e}}{+\mathrm{H}}_2(\mathrm{T}\mathrm{a}\mathrm{f}\mathrm{e}\mathrm{l})$$ 13

The adsorption pathway involves both Heyrovsky or Tafel steps after the Volmer process, as shown in eqs –. Typically, hydrogen adsorption occurs better on some metallic surfaces than on the nonmetallic counterparts, such as those of the ionic state, due to the absence of the cationic species repulsion effect. Also, certain metallic structures, such as those of Fe and Pd, have the tendency to absorb hydrogen molecules, causing embrittlement or the formation of different hydride states, which can affect the overall efficiencies of the catalysts. Using the Tafel slope to predict the pathway of hydrogen evolution displayed by any catalytic material, it has been reported that the rate-determining Volmer-Heyrovsky route exhibited 120 mV/dec, while the Heyrovsky and Tafel pathways showed 40 and 30 mV/dec, respectively. Thus, as shown in Figure c, the Pd catalyst showed the lowest and a negative Tafel slope of −25 mV/dec, implying it to be the most reactive in terms of the hydrogen evolution reaction compared to other tested working electrodes, i.e., the Fe-Clay and Fe-Clay/Ag catalysts. This is in support of the LSV findings presented in Figure a–b. Also, being the most reactive but with a negative slope suggested that it was the most prone to corrosion. In terms of its hydrogen production rate-determining step (r.d.s), its lowest Tafel slope value suggests it favors the Volmer–Tafel route. However, Fe-clay or its Ag composite exhibited 158.72 or 92.60 mV/dec, respectively, suggesting their r.d.s to be the Volmer-Heyrovsky pathway.

Role of Fe and Silver in HER Enhancement

The natural bentonite basal spacing of 12 Å (1.19 nm) includes the 9.6 Å clay layer plus ∼2.4 Å interlayer space containing exchangeable cations and water molecules. Upon modification and calcination at 500 °C, the interlayer water was removed, leading to the complete collapse of the space. The observed decrease in basal spacing from 12 Å to 10 Å indicates this complete collapse of the clay interlayer structure, suggesting a possible clustering of iron oxide particles outside the collapsed structure. This fact was clearly seen in the results of the comparative TEM image analysis of the pristine BC and Fe-BC shown in Figure a–b. In Figure a, the estimated d₀₀₁ spacing was 1.06 ± 0.14 nm (range = 0.92–1.2 nm) in agreement with the XRD results in Table (12 Å = 1.20 nm). However, evidence of the collapsed interlayer structure of the BC can be seen in Figure b after thermal interaction with the Fe modifier (d₀₀₁ = 0.77 ± 0.04 nm (range = 0.73–0.80 nm)). As expected, the residues of the collapsed pillared structure, in the form of α -Fe₂O₃ particles (d₀₂₄ = 0.185 ± 0.019 nm (range = 0.17–0.2 nm)) were seen uniformly distributed on top of the collapsed BC structure (Figure b). The existence of Ag within the collapsed Fe-BC structure was as nanoparticle aggregates as seen in Figure c–d. The presence of clusters means the mechanical mixing of Ag contributes slightly more to the already collapsed structure. Figure d and its inset (Figure d­(i) reveal the respective presence of nanoparticle and microscopic particle conductive network. Thus, the silver aggregates act as metallic “islands” dispersed across the collapsed Fe-BC “sheets”. This mechanical distribution is the primary driver behind the improved performance shown in Figure a.

7.

TEM characterization of the interlayer spacing. (a) Native BC; (b) Fe-BC; (c) Ag–Fe–BC; (d) lower resolution of (c) showing the Ag nanoclusters within the collapsed Fe-BC structure: Inset (i) = SEM image of (d).

The clustering positioning of the Ag noble metal particles within the Fe-BC matrix explained the reason behind the higher polarization resistance the composite exhibited (Table ). In the Ag–Fe–BC system, the presence of silver clusters created a metal–semiconductor Schottky barrier heterojunction, , which caused a difficult flow of electrons at equilibrium. This explains the subsequent drop in the exchange current density from about 16 to 6 μA/cm² (Table ). This increase in resistance at rest is closely tied to the poor performance of the catalyst at 0 V vs RHE (onset potential = −0.43 V vs RHE) (Figure a). This also applies to the Fe-BC which exhibits an onset potential of −0.42 V (vs RHE).

5. Charge Transfer Parameters and Inherent Catalytic Properties.

catalyst system	Y0 (mF)	n	Rs (Ω)	Rp (kΩ)	Ceff* (mF)	ECSA** (cm2)	Rf***
Pd	0.829	0.789	4.77	3.55	0.1887	16.68	66.73
Fe-BC	0.209	0.836	15.20	5.77	0.0676	5.97	23.90
Ag–Fe–BC	0.416	0.776	4.50	15.30	0.0679	6.00	24.00
catalyst system	j @‑0.6 V vs RHE (mA/cm2)	Tafel slope	mass activity (A/g)	Log j 0	ΔGH* (eV)	specific activity (A/cm2)ECSA	TOF (s–1)
Pd	274	25.94	22.15	–4.60	–0.01	0.016	0.012
Fe-BC	7.5	158.72	0.61	–4.81	–0.42	0.0013	0.00018
Ag–Fe–BC	9.1	92.60	0.74	–5.23	+0.43	0.0015	0.00041
catalyst system	η@Industrial (V) vs RHE	j 0@Industrial η (mA/cm2)	mass activity (A/g)	Log j 0	ΔGH* (eV)	specific activity (A/cm2)	TOF (s–1)
Ag–Fe–BC	–1.23	274	22.15	–5.23	+0.43	0.046	0.012
Ag–Fe–BC	–1.54	–575	46.48	–5.23	+0.43	0.096	0.026
Ag–Fe–BC	–2.22	–1018	82.28	–5.23	+0.43	0.170	0.046

Typically, Ag exhibits a hydrogen subrepelling surface due to its inherent weak binding characteristics. − Thus, the energy required to initiate the first proton discharge (the Volmer step) is usually very high (bulk’s Ag ΔG_H* about +1.0 eV), explaining the no hydrogen production of the Ag nanoparticle composite at 0 V vs RHE (Rp = 15.3 kΩ). However, applying an overpotential of ≥0.43 V vs RHE) beyond the equilibrium state lowered the activation energy, consequently activating the Volmer’s step (the rate-determining step (r.d.s.), Tafel slope = 92.60 mV/dec) and subsequent faster Tafel or Heyrovsky pathway, promoting rapid desorption of molecular hydrogen from the surface (Figure c). The production of hydrogen on a silver surface at a lower overpotential of 0.43 V vs RHE occurs through surface restructuring such as existence in nanoscale dimensions (for example, in this study, the average particle size = 65 ± 18 nm) (Figure d). This size reduction state caused some changes in the d band center position of Ag, moving it closer to the Fermi level, promoting some moderate binding characteristics to proton (onset at −0.43 V vs RHE), initiating its reduction. In contrast, Fe-BC are rich in surface oxygen groups due to the presence of α-Fe₂O₃ species (Figure b–d) in the BC’s collapsed structure. These species are naturally proton (H⁺) binders. At pH 0, they bind strongly to H⁺, favoring its reduction to a hydrogen atom (H*) (fast Volmer step). However, the Fe-(0001) facet of the α- Fe₂O₃ has been reported to show strong affinity (ΔG_H* = −7.11 eV) to binding molecular H₂. This could reduce the overall amount of hydrogen generated overtime, explaining the lower current density and TOF associated with Fe-BC compared to that produced by the Ag–Fe–BC counterpart at 0.60 V vs RHE, despite having a better intrinsic catalytic property of exchange current density and overpotential (Figure a, Table ). However, some of the atomic hydrogen formed on the Fe-BC system may spillover to the Ag sites in Ag–Fe–BC, promoting their combination and desorption (Tafel or Heyrovsky step). Thus, overall, the iron oxide on the collapsed clay structure acted as proton harvesters from the solution, promoting the Volmer step, while the silver aided in synergy, providing extra active sites, together with, enhancing faster recombination (Tafel step) and desorption rate (Heyrovsky step) (eqs –) (Scheme ) (Table ).

1. Proposed Mechanism .

^a Proton source: 0.5 M H₂SO₄.

In Table , the addition of Ag caused a slight change in the ECSA value, about 0.5% increase. This slight increase in the ECSA led to the lowering of the Tafel slope from 158.72 to 92.60 mV/dec, about 21% increase in mass and specific activities and 134% increment in turnover frequency (TOF). Besides the laboratory-scale performance at room temperature, it is essential to see how the Ag–Fe–BC catalyst behaves under industrial conditions. As shown in Table , under the industrial applied potential condition, between −1.0 and −2.3 V vs RHE, a current density range of about 270–1020 (mA/cm²) was achieved.

Conclusion

In this work, bentonite natural clay (BC) was successfully modified with Al and Fe, resulting in Al- and Fe-modified BCs with improved textural and thermal stability properties. The shift in basal spacing confirmed the success of the modification process. Enhancement in the BET surface area upon calcination, along with a decrease in pore size and an increase in pore volume, resulted in the more exposure of the Fe-BC active sites and the formation of loosely packed clay layers, thereby facilitating the mass transfer of reactants during hydrogen production. After lacing with Ag, the performance of the Fe-clay material improved significantly (about 21% increase in hydrogen production and 134% increase in TOF at −0.60 V applied potential vs RHE). The Ag@Fe-Clay composite exhibited the same current density of 274 mA/cm² as that of the commercial Pd black catalysts at −1.23 V vs RHE. However, the nanocomposite catalyst displayed a current density of about 1000 mA/cm² at an industrially applied potential of −2.2 V vs RHE. The study reveals the potential of Ag-laced Fe-modified bentonite clay as a low-cost catalyst in the water-splitting hydrogen production process. Evaluation of other clay types is currently ongoing, the results of which will be communicated in due course.

The authors declare no competing financial interest.