Pd‑Loaded Zr‑MOF for Visible‑Light Photocatalytic Hydrogen Generation

Background

The growing energy demand and environmental concerns have intensified the search for renewable, low‑cost alternatives to fossil fuels. Photocatalytic water splitting powered by sunlight is a particularly attractive approach, yet it requires catalysts that are highly active, stable, and responsive to visible light. Traditional photocatalysts—such as TiO₂, NiO, CuS, and C₃N₄—often suffer from limited light absorption and rapid charge recombination.

Metal–organic frameworks (MOFs) have emerged as a versatile class of porous materials with exceptional surface areas, tunable pore sizes, and structural diversity. Their crystalline frameworks can host catalytic sites and facilitate charge transport, making them compelling candidates for photocatalysis. However, most MOFs absorb only in the ultraviolet, restricting their practical use.

Recent studies have demonstrated that incorporating metal nanoparticles (e.g., Pt, Pd) and dye sensitizers (e.g., rhodamine B, erythrosine B) can extend the spectral response of MOFs into the visible region. In particular, the Zr‑based UiO‑66 framework, known for its remarkable thermal and chemical stability, has been combined with Pd nanoparticles to produce efficient catalysts for the Suzuki–Miyaura reaction and, more recently, for photocatalytic hydrogen evolution.

Despite these advances, achieving high hydrogen production rates under visible‑light irradiation remains challenging. This work addresses that gap by integrating a Pd‑loaded UiO‑66 with eosin Y, thereby harnessing the benefits of a robust framework, active metal sites, and a broad‑band photosensitizer.

Methods

Synthesis of UiO‑66

UiO‑66 was synthesized via a solvothermal route. In a typical synthesis, ZrCl₄ (1.89 g) and terephthalic acid (2.79 g) were dissolved in 48.7 mL DMF containing 1.43 mL HCl. The solution was transferred to a 100 mL Teflon‑lined autoclave and heated at 220 °C for 20 h. After natural cooling, the product was collected by centrifugation, washed sequentially with DMF and methanol to remove residual solvents, and dried under vacuum (90 °C, 6 h).

Preparation of Pd/MOF Catalysts

Pd nanoparticles were deposited onto UiO‑66 by impregnation‑reduction. 0.2 g of UiO‑66 was dispersed in 200 mL deionized water and mixed with an appropriate amount of H₂PdCl₄. The mixture was stirred for 1 h, then NaBH₄ was added dropwise (n(H₂PdCl₄):n(NaBH₄) = 1:3). After 3 h of stirring, the black granules were washed with water and dried in a vacuum oven at 70 °C for 6 h. Pd loadings of 1%, 3%, and 5% (by weight) yielded Pd/MOF 1%, Pd/MOF 3%, and Pd/MOF 5%, respectively.

Characterization

Morphology and particle distribution were examined by FESEM (JSM‑6701F) and TEM (FEI Tecnai TF20). XRD (Rigaku RINT‑2000, Cu Kα) confirmed the retention of the UiO‑66 framework. UV‑Vis diffuse reflectance spectra were recorded on a Shimadzu UV‑2550. XPS (ESCALAB 250Xi) quantified elemental composition and oxidation states. Nitrogen adsorption–desorption isotherms (ASAP 2020M) provided BET surface areas and pore size distributions (BJH method).

Photocatalytic H₂ Evolution

Photocatalytic tests were performed in a quartz reactor (62 cm³). 10 mg of catalyst was dispersed in 30 mL of 15% v/v TEOA aqueous solution containing 20 mg EY. The mixture was sonicated for 15 min, degassed with N₂, and irradiated with a 5‑W LED lamp (λ ≥ 420 nm) while stirring. H₂ evolution was quantified by GC (Tianmei GC7900, TCD). Photocurrent and photo‑electrochemical measurements were conducted in a three‑electrode setup using FTO working electrodes, Pt counter, and SCE reference, with a 300‑W xenon lamp (λ ≥ 420 nm) and 0.2 M Na₂SO₄ electrolyte.

Results and Discussion

Morphology and Structure

FESEM images (Fig. 1) show that pristine UiO‑66 consists of uniform, smooth particles, while Pd/MOF displays a roughened surface due to well‑dispersed Pd nanoparticles (~6 nm, TEM in Fig. 2). The HRTEM lattice spacing of 0.223 nm corresponds to the (111) plane of metallic Pd, confirming successful reduction. XRD patterns (Fig. 3) reveal that the UiO‑66 crystalline framework remains intact after Pd loading; no new Pd peaks are observed, indicating excellent dispersion and minimal structural disruption.

Surface Area and Porosity

N₂ adsorption–desorption isotherms (Fig. 5) demonstrate that the BET surface area of UiO‑66 (791.6 m² g⁻¹) increases slightly upon Pd incorporation, peaking at Pd/MOF 3% (≈840 m² g⁻¹). This enhanced surface area facilitates dye adsorption (39.7 µmol g⁻¹ for Pd/MOF 3%), which is critical for efficient sensitization. Excessive Pd loading (5%) reduces surface area, likely due to pore blockage.

Optical Properties

UV‑Vis diffuse reflectance (Fig. 7) shows that both UiO‑66 and Pd/MOF exhibit absorption edges around 300 nm, unsuitable for visible‑light activation. Introducing EY as a photosensitizer extends the absorption into the visible region, enabling effective charge excitation under λ ≥ 420 nm irradiation.

Photocatalytic Performance

Under visible‑light irradiation (λ ≥ 420 nm) with EY and TEOA, only 0.03 mmol g⁻¹ h⁻¹ H₂ was produced over pristine UiO‑66. Loading 1% Pd increased production to 3.24 mmol g⁻¹ h⁻¹, while 3% Pd yielded 9.43 mmol g⁻¹ h⁻¹. At 5% Pd, the rate dropped to 6.06 mmol g⁻¹ h⁻¹, attributed to pore blockage and reduced dye adsorption. Thus, Pd/MOF 3% delivers the highest activity, surpassing previous reports by two orders of magnitude.

Effect of pH

Hydrogen evolution peaked at pH 7 (18.10 mmol h⁻¹). At lower pH, protonation of TEOA shortens the EY lifetime; at higher pH, reduced proton concentration limits H⁺ availability. The optimal pH balances sacrificial donor activity and proton supply.

Stability

Four‑cycle stability tests (Fig. 10) show sustained H₂ evolution after re‑addition of EY, confirming the robustness of Pd/MOF 3% under repeated use, though dye degradation remains a challenge for long‑term operation.

Photoluminescence Quenching

Fluorescence measurements (Fig. 11) reveal that EY emission intensity is progressively quenched by Pd/MOF, with the strongest quenching at 3% Pd. This indicates efficient electron transfer from EY⁻· to the MOF and then to Pd, supporting the proposed charge‑separation mechanism.

Photo‑Electrochemical Analysis

Transient photocurrent responses (Fig. 12a) show that Pd/MOF 3% generates the highest photocurrent density under visible light, correlating with its superior H₂ production. Linear sweep voltammetry (Fig. 12b) confirms enhanced charge transfer kinetics for Pd‑loaded samples.

Mechanistic Insight

The proposed mechanism (Scheme 1) involves adsorption of EY on UiO‑66, excitation to EY¹* → EY³*, reductive quenching by TEOA to EY⁻·, electron transfer to the MOF CB (−0.6 V vs. NHE), followed by rapid migration to Pd nanoparticles where protons are reduced to H₂. The large surface area and well‑ordered pores of UiO‑66 facilitate efficient electron transport and prolong charge carrier lifetimes.

Conclusions

We have demonstrated that Pd‑loaded UiO‑66, when sensitized with eosin Y, achieves a record hydrogen evolution rate of 9.43 mmol g⁻¹ h⁻¹ under visible‑light irradiation (λ ≥ 420 nm). Comprehensive characterization confirms that the synergy between Pd nanoparticles, EY, and the Zr‑MOF lattice yields superior light harvesting, charge separation, and catalytic activity. This work establishes a robust, scalable strategy for visible‑light photocatalytic hydrogen production using MOF‑based materials.