Efficient Charge Transfer of p-n Heterojunction UiO-66-NH₂/CuFe₂O₄ Composite for Photocatalytic Hydrogen Production

Abstract

Using a p-n heterojunction is one of the efficient methods to increase charge transfer in photocatalysis applications. So, herein, p-type UiO-66 (NH₂) and n-type CuFe₂O₄ (CFO) are used to form an effective p-n heterojunction. Due to their poor charge separation in their pristine form, both UiO-66 (NH₂) and CFO materials cannot produce hydrogen; however, the composite p-n heterojunction formed between these materials makes fast charge separation and so hydrogen is efficiently produced. The optimized catalyst UCFO 25% produces a maximum of 62.5 µmol/g/h hydrogen in an aqueous methanol solution. The formation of a p-n heterojunction is confirmed by Mott–Schottky analysis and optical properties, crystallinity and the local atomic environment of the material was analyzed by various analytical tools like UV-Vis spectroscopy, XRD, and XANES.

1. Introduction

The development of science and technology has dramatically impacted the betterment of human civilization. To persist with the use of technology, energy is needed. The existing energy is derived from fossil fuels, which cause global warming that leads to climate change [1,2,3]. Hydrogen is a green fuel with remarkable features. It can be produced from many methods, is easy to store and transport, and has a high fuel cell efficiency [4]. Hydrogen may be a green fuel, but its methods are not environmentally friendly. Steam reforming, gasification, and thermochemical methods are the traditional methods that are used to generate hydrogen on a large scale. These methods have their drawbacks. For instance, steam reforming of methane emits 9–12 tons of CO₂ to generate 1 ton of H₂ [5]. Photocatalysis is one such way that effectively uses renewable resources to produce hydrogen. In this method, hydrogen is generated by water splitting using a photocatalyst in sunlight. First time Fujishima and Honda was reported, using the TiO₂ as semiconductor electrode to produce H₂ and O₂. This is the step up for the field of heterogeneous photocatalysts began to develop. Since, a variety of semiconductor materials has been employed for the utilization of visible light to generate hydrogen [6,7]. The single-component system has drawbacks like poor charge separation efficiency, limited visible light-harvesting ability, and rapid charge recombination rate, making it a poor material for photocatalytic activity [8,9]. To overcome this issue, various steps were carried out, among which the creation of p-n-heterojunction-based material is one of the effective methods to separate the electrons and holes to improve photocatalysis. This scheme looks like a type II mechanism, but, in this method, the internal electric field is built between both p and n materials due to the opposite charges, which are helpful for effective charge separation. When light falls on the catalyst’s surface, the holes in the p-type material begin to move toward the n-type material, and the electrons in the n-type material begin to move toward the p-type material. Various p-n heterojunction materials were used in photocatalysis, like NiSe/Cd_0.5Zn_0.5S for photocatalytic hydrogen production, with a remarkable activity of 78.01 mmol·h⁻¹·g⁻¹ [10], g-C₃N₄/CaFe₂O₄ used to completely degrade the organic pollutant ciprofloxacin by 30 min and phenol by 130 min with effectiveness [11], and NiO/g-C₃N₄, which was used to reduce CO₂ to produce CO with a yield of 4.7 μmol g⁻¹ h⁻¹ [12]. With these results, we have chosen a p-n heterojunction catalyst system for the photocatalytic hydrogen production reaction.

In recent years, metal–organic frameworks (MOF), emerging metal ions, and organic ligand material with high surface area, crystalline structure, bandgap tunability, and charge transport have created widespread attention among researchers around the world. The high porosity of the MOFs, which promotes the diffusion rate of the products and reactants, contributes to these attractive properties [13,14]. Research has shown that MOFs can act as excellent semiconductors, and their bandgap tunability property makes them ideal candidates for photocatalysis [15]. The advantage of using MOFs as a catalyst is that they have a metal–oxo cluster and a comprehensive, varied combination of bridged organic ligands, which forms a high surface area with a porous structure [16,17]. The challenge is to find the ideal MOF from the vast library. There are specific criteria to be satisfied by the MOF in order to utilize it for effective photocatalytic hydrogen production. The MOF must have a bandgap of less than 3 eV to perform in visible light, practical charge separation ability, and chemical stability. As conventional semiconductors, MOFs have the problem of a rapid recombination rate after photoexcitation. The charge transfer in the MOF exists between the ligand and metal as the organic linker is keen to have a π-conjugated system [18]. A heterostructure is formed between the MOF and other semiconductor materials to enhance photocatalytic activity. Also, the replacement of metal or ligands in the MOF is performed in order to achieve the optimal bandgap and high light-harvesting ability to improve photocatalytic activity [15].

UiO-66, a Zirconium-based n-type MOF, has gained massive attractiveness in photocatalytic hydrogen production since Silva et al. (2010) reported water splitting to generate hydrogen using UiO-66 and UiO-66-NH₂ [19]. The advantage of UiO-66 is its excellent chemical and hydrothermal stability, with a high coordination number between the organic linker and metal cluster compared to other MOFs [20]. It also has remarkable thermal stability; up to 300 °C, it does not make any change in its structure [21]. This stability makes UiO-66 a competent material for photocatalytic hydrogen generation [22,23,24]. However, the bandgap of UiO-66 (4 eV) is not favorable for visible light absorption. It can be optimized by functionalization of the linker and forming a heterostructure by making a composite with a semiconductor that effectively absorbs visible light [25]. CuFe₂O₄, a spinel ferrite and p-type compound, is suitable for photocatalysis owing to its stability, low bandgap, and environmental compatibility. Yang et al. (2009) reported photocatalytic hydrogen production using CuFe₂O₄ [26]. The utilization of this ferrite is less due to the challenges faced, like phase separation during the synthesis [27]. The low bandgap of this material can absorb visible light effectively, but it cannot perform photocatalytic activity on its own due to the rapid charge recombination rate. So, when a heterostructure is formed with this compound with a potential MOF, it can enhance the photocatalytic activity and its light absorption efficiency [28].

Herein, we show a simple ex situ method for forming a novel p-n heterojunction composite, UiO-66-NH₂/CuFe₂O₄ The photocatalytic experiments are carried out using different sacrificial agents under direct sunlight irradiation, and the generated hydrogen is quantified using Gas Chromatography (GC). The composite material UCFO was synthesized with different wt.% of CuFe₂O₄ (UCFO 50 and UCFO 25) and it was characterized using XRD, UV-DRS, XPS, SEM, and XANES. The bandgap is calculated using a Tauc plot. The result shows that heterostructure formation enhances the photocatalytic activity compared to pristine UiO-66-NH₂ and CuFe₂O₄, that is producing the highest amount, 62.5 µmol/g/h. This study offers a novel understanding of the heterostructure UiO-66-NH₂/CuFe₂O₄ and explores its potential use in solar energy conversion.

2. Result and Discussion

The crystallinity of the material was confirmed through XRD. The observed PXRD pattern of UiO-66 demonstrated sharp peaks with diffraction planes at 7.37° and 8.5°, corresponding to the (111) and (200) crystalline planes, respectively [22]. A characteristic peak was observed at 29°, 35°, and 38° for CuF₂O₄, which corresponds to the (220), (311), and (222) crystalline planes, respectively [29], and it is well matched with corresponding card numbers 4512072 and 034-0425. In the composite, both materials’ peaks are evidently shown in the graph, and, in UCFO 25, UiO-66 (NH₂) peaks dominated compared to CFO and were slightly shifted compared to its pristine material, which confirms the successful formation of the composite. In UCFO 50, the UiO-66 (NH₂) peak intensity was decreased. This is due to CFO covering the surface of UiO-66 (NH₂), which is a masking effect, as shown in Figure 1a. Furthermore, the Copper valence states and local arrangement of the synthesized catalyst were determined using the Cu K-edge of XANES analysis. The Cu K-edge profiles of the UCFO nanocomposites were compared with different Cu references like Cu₂O, CuO, Cu foil, and Cu standard [30]. Figure 1b shows that UCFO nanocomposites were well matched to pristine CFO and CuO standard, which denotes that it was in a +2 oxidation state, and, based on the linear combination fitting (LCF) analysis, the synthesized UCFO was mainly formed, 99%, as a Cu (II) state. The optical property of the materials UiO-66 (NH₂), CFO, UCFO 25, and UCFO 50 was investigated using the UV–visible absorption spectra shown in Figure 2a. The absorption edge for UiO-66 (NH₂) was observed at 320 nm, which indicates that the pristine MOF did not have significant absorption in the visible light region with a bandgap of 3.20 eV [31]. The CFO exhibited a wide absorption range of 200–700 nm and had a bandgap of 1.50 eV, indicating efficient absorption of visible light. The integration of CFO with the MOF exhibited significant light absorption in the visible spectrum. The composite materials UCFO 25 and UCFO 50 had bandgaps of 2.60 and 2.70 eV, respectively. The estimated bandgap energy (Eg) of the materials was determined by applying Tauc’s equation [32].

$$\alpha h\vartheta =A{\left(h\vartheta -Eg\right)}^{n/2}$$

(1)

where the variables α, h, $\vartheta$, A, and Eg denote the absorption coefficient, Planck constant, the frequency of light, proportionality constant, and the bandgap energy, respectively. The “n” value is 1 for the direct bandgap semiconductor and 4 for an indirect bandgap [33].

The material’s morphology was analyzed using SEM. From Figure 3a, the synthesized CuFe₂O₄ had an irregular plate-like morphology, and UiO-66 (NH₂) was in an agglomerated nanosphere, as shown in Figure 3b,c, which confirms that UiO-66 (NH₂) was present in the top of CuFe₂O₄. It confirms that the formation of the UCFO 25 composite and corresponding elemental map (Figure 3d–i) denotes that all the elements present throughout the material and the atomic percentage from the energy-dispersive X-ray spectroscopy through to SEM was given in Figure S1.

The chemical states and functional groups of UCFO 25 composites were analyzed using the XPS spectra shown in Figure 4a–f. The C 1s spectra exhibit three distinct peaks at 284.4, 285.9, and 288.2 eV, which correspond to the sp² hybridized C=C bonds found in aromatic carbon, the C-N bonds present in triazine rings, and the C=O bonds characteristic of carboxylic acids, respectively [34]. The N 1s spectra are split into two, attributed to the amino group (-NH₂) at 399 eV and the protonated amidogen (-NH₃⁺) at 401.7 eV [35]. The Zr 3d spectrum has two peaks at 182.4 and 184.7 eV, which are assigned to the Zr 3d_5/2 and Zr 3d_3/2 orbitals, respectively [36]. These peaks suggest the existence of Zr⁴⁺ chemical states. The Cu 2p spectra are composed of Cu 2p_3/2 and Cu 2p_1/2 at 933.5 and 953 eV, respectively, and the satellite peak between them stands for the existence of the Cu²⁺ chemical state in the UCFO composites. For Fe 2p spectra, the peaks at 711.3 and 724.5, and the satellite peak at 718.4 eV, evidence the presence of Fe 2p_3/2, Fe 2p_1/2, and Fe³⁺ ions in the UCFO [37,38]. Finally, the O 1s spectra are composed of peaks at 530, 531.3, and 533.4 eV, which is attributed to the metal–oxygen bond, C=O bonds, and the lattice oxygen, respectively, in the UCFO. The CFO spectra (Figure S2) were compared with the above XPS results and confirmed the linkage between UiO-66 (NH₂) and CFO and the successful formation of UCFO composites [39]. The pristine CFO XPS peaks given in Figure S3, UiO-66 (NH₂) in Figure S4, and UCFO 50 in Figure S5, attached in the Supplementary Materials.

Mott–Schottky (MS) plots were used to identify the conductivity type and flat-band potentials of the as-synthesized photocatalysts [40,41]. Figure 5a demonstrates that the positive slope of UiO-66 (NH₂) confirms its classification as an n-type semiconductor. The CFO exhibited negative slopes, indicating p-type conductivity, which confirms Figure 5b. The graphs indicate that the flat-band potentials for UiO-66 (NH₂) and CFO relative to Ag/AgCl are −1.45 V and 0.1 V, respectively. Therefore, the conduction band (CB) of the material can be determined and converted to the reversible hydrogen electrode (RHE) scale using the Nernst equation, which is represented as Equation (2) [42].

$$E_{RHE}=E_{Ag/AgCl}+{E°}_{Ag/AgCl}+(0.059\ast pH)$$

(2)

Utilizing the aforementioned equation, the specific values for UiO-66 (NH₂) (CBM) and CFO (VBM) are −0.79 V and 0.71 V, respectively. The detailed calculation is given in the supporting information. Significantly, the combination of UiO-66 (NH₂) and CFO resulted of an inverted “V-shape” MS plot. This indicating that the formation of a p-n heterojunction between UiO-66 (NH₂) and CFO, as depicted in Figure 5c, and UCFO 50 (Figure S6) [43].

To study the charge separation and conductivity of the material, electrochemical impedance spectroscopy was used [44]. The lowest arc radius of UCFO 25 denotes that it has higher conductivity and better charge separation compared with other material (Figure 5d). The graph depicts that the MOF has a lower charge separation property because the localized nature of charge carriers in MOFs can hinder overall electron flow. While MOFs combine with ferrite, a p-n heterojunction forms between both materials, which will increase the charge separation and improve the photocatalytic hydrogen production [45]. The photocatalytic hydrogen production mechanism shown in Figure 5e was confirmed through an MS plot. The photocatalytic hydrogen production was carried out in direct sunlight. The optimized weight ratio of the UCFO 25 nanocomposite produced the highest amount, 62.5 µmol/g/h (Figure 6a), with a TON of 50,000 (Figure S2, and both pristine materials could not produce the hydrogen due to the poor charge transfer in the MOF, which resists the conductivity and lower bandgap of ferrite, easily allowing the electron hole recombination. The p-n heterojunction formation between the MOF and ferrite was efficiently helpful for charge transfer, and it created an impact on photocatalytic hydrogen production. Different scavengers are used to quench the holes in the valence band to improve the hydrogen yield. In that, methanol is efficiently worked to produce a significant amount of hydrogen compared to other scavengers like ascorbic acid and TEOA (Figure 6b) [46].

Ascorbic acid slightly helped to quench the holes to produce 45 µmol/g of hydrogen, but, in TEOA, there was no activity observed because the more basic nature of the solution broke the bonding between zirconium and terephthalic acid so the MOF dissolved in the solution. The recyclability of the material was tested. For stability up to four cycles it maintained its activity; afterward, the activity significantly decreased. In each cycle, the catalyst was tested for four hours for photocatalytic activity (Figure 6c). After four catalytic cycles, the crystallinity of the material did not change, which was confirmed by XRD (Figure S7), and the activity was compared with a similar kind of photocatalyst (

Table 1. A comparison table of hydrogen efficiency with various photocatalytic materials.

S. No	Materials	Experimental Conditions	Hydrogen Efficiency	Ref.
1.	Cu2O/TiO2	200 W Hg-Xe arc lamp (λ > 420 nm), 20 mg catalyst, 50 mL distilled water with 0.5 M Na2SO3 as a sacrificial agent	60.96 μmol g−1 h−1	[30]
2.	NH2-UiO-66/CoFe2O4/CdIn2S4	20 mg of catalyst, TEOA as a sacrificial agent with Pt cocatalyst in presence of visible light λ > 420 nm	55 μmol h−1	[45]
3.	NiFe2O4/Cu2O	250 W metal halides lamp, model Philips. Methanol at 2% v/v was employed as a scavenger	2.8 μmol g−1 h−1	[47]
4.	MoS2 QDs/UiO-66-NH2/G	100 mL of 10% TEOA solution with Eosin Y dye kept in 300-W Xe lamp	62 μmol h−1	[48]
5.	Fe2O3/CuxO/TiO2	20 mg of catalyst was used in 50 mL of 5 wt.% of MeOH solution in the presence of a 500 W Xenon lamp	65.82 μmol g−1 h−1	[49]
6.	Cu2O/Fe2O3/rGO	50 mg catalyst was dispersed into 140 mL of distilled water in the presence of 300-W Xe lamp	4.86 μmol g−1 h−1	[50]
7.	rGO/ UiO-66-NH2/Co-Mo-S	10 mg catalyst was added to 30 mL TEOA (15%) solution. After 10 min of ultrasonic dispersion, 20 mg of Eosin Y was added and kept in 300-W Xe lamp	67.8 μmol h−1	[51]
8.	Cubic Cu2O/nano fiberTiO2	20 mg of catalyst was used in 50 mL of 5 wt.% of MeOH solution in the presence of a 500 W Xenon lamp	48 μmol g−1 h−1	[52]
9.	UiO-66-NH2@Au@CdS	A 300 W xenon lamp, 10 mg photocatalysts was dispersed in 20 mL aqueous solution of L-ascorbic acid (pH = 4.0; 0.1 M). In particular, Pt (0.25 wt.%) was the cocatalyst for hydrogen production	39.5 μmol h−1	[53]
10.	UiO-66 (NH2)/CuFe2O4	5 mg catalyst dispersed in 10 wt.% of MeOH solution kept in direct sunlight	62.5 μmol g−1 h−1	This work

). The compared results denote that the synthesized p-n heterojunction UiO-66NH₂/CFO effectively improved the photocatalysis.

3. Experimental Details

3.1. Chemicals

Zirconium chloride (ZrCl₄) (purchased from Nacalai Tesque, Kyoto, Japan), Amino terephthalic acid (purchased from TCI, Hyderabad, India), Acetic acid (purchased from Fischer Scientific, Maharashtra, India.), Methanol (purchased from molychem, Mumbai, India) N, N-Dimethylformamide (DMF), Copper nitrate trihydrate (Cu(NO₃)₂.3H₂O), Ferric nitrate nonahydrate (Fe(NO₃)₃.9H₂O), Citric acid anhydrous, Ethylene glycol, Ethanol (all purchased from SRL, Mumbai, India), and DI water.

3.2. Synthesis of UiO-66-NH₂

Amine-functionalized UiO-66 was synthesized as per the previously reported literature with slight modification [54]. Here, zirconium chloride was used as a precursor for this reaction because its high hydrolysis capability achieves better bonding with the ligand. The Amine-functioned ligand has better visible absorption than usual terephthalic acid. Acetic acid is one of the modulators which regulates the structure of MOFs and competes with the ligand to accelerate the formation of MOF. A 1:1 equimolar ratio of zirconium chloride and amino terephthalic acid was dissolved in an equal amount of DMF separately. The metal and ligand mixtures were added together under stirring at room temperature. After 20 min of stirring, a certain amount of acetic acid was added. Following 2 h of stirring, the mixture was relocated to a 100 mL autoclave lined with Teflon and subjected to a temperature of 120 °C in a hot air oven for a duration of 24 h. The resulting product was subjected to centrifugation, followed by washing with DMF and methanol, and subsequently dried overnight at a temperature of 60 °C. The resulting product was designated as UiO-66 (NH₂) (Scheme 1).

3.3. Synthesis of CuFe₂O₄

Copper ferrite was synthesized by a simple sol–gel method followed by a calcination method. A 1:1 equimolar ratio of copper nitrate trihydrate and ferric nitrate nonahydrate was dissolved in equal DI water separately. In copper nitrate solution, Ethylene glycol and Citric acid anhydrous were added and dissolved, and ferric nitrate solution was added to this solution. The addition of citric acid acted as a reducing agent, in this reaction and ethylene glycol was made a link between both metals. At room temperature, the mixture was stirred for two hours. Following stirring, the mixture underwent heating in an oven at 130 °C for a duration of 2 h, and, subsequently, in a muffle furnace at 300 °C for a period of 20 h. The product was ground into a fine powder using a mortar and pestle. The resulting powder was heated for three hours at 800 °C in a tube furnace. A brown-colored substance was acquired and named CFO.

3.4. Synthesis of UiO-66-NH₂/CuFe₂O₄

The composite material was synthesized by an ex situ method using methanol as a solvent. After dispersing a specific amount of UiO-66 (NH₂) in 20 mL of ethanol and sonicating it for a few minutes, a specific amount of CuFe₂O₄ was added and sonicated for thirty minutes. Following that, the mixture was gently stirred while being heated at 60 °C for 2 h, and, subsequently, the solvent was evaporated. The sample was washed three times with DMFand ethanol then dried at 60 °C overnight. The reaction was conducted using different weight percentages (25 and 50 wt.%) of CuFe₂O₄ named UCFO 25 and UCFO 50.

3.5. Characterizations

The structure and composition of the materials were investigated using powder X-ray diffraction (PXRD) patterns measured by Cu Kα radiation in the 2θ range 0 to 90° (Malvern PANalytical, Malvern, UK), X-ray photoelectron spectroscopy (XPS) spectra (Physical Electronics, Chanhassen, MN, USA), and X-ray absorption near-edge spectroscopy (XANES) spectra obtained in the Kyushu Synchrotron Light Research Centre, also known as SAGA-LS, in Tosu, Japan. The optical property of the materials was analyzed by UV–visible spectrometer, and the spectrum was recorded in the wavelength region between 200 and 800 nm using Shimadzu (Tokyo, Japan) UV 3600 Plus. The morphologies and micrographs of the materials were observed by scanning electron microscopy (HR SEM, Thermo Scientific Apreo S, Waltham, MA, USA).

3.6. Photocatalytic Experiment

The photocatalytic hydrogen generation reaction was performed on the rooftop of Sir C. V. Raman Research Park, SRM IST, Chennai, India. The reaction was conducted in a Kjeldahl flask containing 5 mg of the photocatalyst suspended in 50 mL of 5% scavenger (Methanol, TEOA, and 0.1 M Ascorbic acid) solution. The reaction solution was purged with nitrogen gas for 20 min to expel dissolved gases. The reaction setup was exposed to sunlight, and an offline gas chromatograph (Shimadzu GC-2014 with Molecular Sieve/5 °A column) analyzed hydrogen generation with a TCD detector. All the experiments were conducted from 11 am to 2 pm Indian Standard Time.

3.7. Electrochemical and Photoelectrochemical Experiment

The electrochemical properties of the synthesized catalysts were studied using a modified glassy carbon as a working electrode, an Ag/AgCl reference electrode, and a Pt counter electrode. In the experimental procedure, 2 mg catalyst was dispersed in 1 mL of water and then ultrasonicated for half an hour. Then, 25 mL of 0.25 M Na₂SO₄ solution (~7 pH) was prepared and used as an electrolyte for all reactions. Subsequently, the slurry was applied onto a glassy carbon (GC) electrode using drop casting technique, and the electrochemical examination was conducted using an electrochemical workstation (CH Instruments 760). The photoelectrochemical experiments were carried out in presence of a 100 W Hg lamp.

4. Conclusions

In conclusion, a straightforward ex situ technique was employed to produce p-n-type UiO-66(NH₂)/CuFe₂O₄ composites. The charge separation efficiency and charge carrier of electrons and holes were enhanced due to the creation of the p-n heterojunction interface effect between UiO-66(NH₂) and CuFe₂O₄. Therefore, composites generate a substantial quantity of hydrogen in comparison to their original form. Scavenger investigations have shown that methanol serves as an effective hole quencher, enhancing the photocatalytic activity. The Mott–Schottky measurements provide evidence of a p-n heterojunction formed by the UiO-66(NH₂) and the CFO. This indicates that the combination of spinel ferrites with an MOF-based p-n heterojunction photocatalyst has new possibilities for effectively harnessing solar energy to produce hydrogen through photocatalysis.