Coordination engineering of Pd single-atom catalysts for non-radical organohalide degradation†

Zhenjie Li ^abc, Chaohuang Chen ^ac, Kaijian Sang ^ac, Xunheng Jiang ^ac, Xinyue Wu ^ac, Jiang Xu ^ac, Kun Yang ^ac and Daohui Lin *^abc
aZhejiang Provincial Key Laboratory of Organic Pollution Process and Control, Department of Environmental Science, Zhejiang University, Hangzhou 310058, China. E-mail: lindaohui@zju.edu.cn
^bZhejiang Ecological Civilization Academy, Anji 313300, China
^cState Key Laboratory of Soil Pollution Control and Safety, Zhejiang University, Hangzhou 310058, China

First published on 25th June 2025

---

---

1. Introduction

Halogenated organic carbons (HOCs) have attracted global health and ecological concerns,^1–3 as they represent a group of pollutants that are carcinogenic to humans and persistent in the environment.⁴ Because of the exceptional activity of Pd atoms for carbon–halogen bond (C–X) activation,⁵ substantial research has focused on the catalytic dehalogenation of HOCs by Pd-based nanoparticle (NP) catalysts. However, the high cost and easy deactivation of Pd NPs in reductive dehalogenation catalysis hamper their application in HOC degradation. For example, NPs and bulk catalysts are usually limited in utilizing surface atoms,⁶ thereby wasting much of the expensive Pd atoms in the particles. In addition, the heterogeneous nature of NPs generally results in slower reaction kinetics compared to their homogeneous counterparts.⁷ Furthermore, due to the strong adsorption of halide ions and other species (e.g., S-containing ions,^8,9 and H₂O¹⁰) on Pd NPs, the reactive sites on Pd NPs are prone to being occupied and poisoned. Thus, designing highly efficient Pd-based catalysts for HOC degradation remains challenging.

Recently developed SACs are gaining prominence in environmental remediation due to their maximized metal atom utilization and unique capability to amalgamate the complementary advantages of homogeneous and heterogeneous catalysts.^11–13 Nevertheless, the catalytic reductive activity of Pd SACs towards HOC dehalogenation is considerably inferior to that of their NP counterparts owing to the decreased H₂ activation capability in many cases.^14–16 However, the tunable electronic structure and unique properties of SACs may pave the way to achieve the efficient degradation of HOCs through PMS activation.^17–19 Thus, catalytic oxidative degradation of HOCs by fabricating Pd SACs with suitable electronic structures may tackle the above challenges of Pd NPs and become a practical application scenario for Pd-based catalysts. This hypothesis is highly logical because efficient degradation of persistent organic pollutants by PMS activation with Pd-based materials has been reported.^19–23 However, corresponding research is still in its infancy. For example, the effects of the electron density of Pd SACs on their catalytic performance for PMS activation have not been clearly elucidated, and the PMS activation pathway is poorly understood on the surface of Pd atoms. Divergent catalytic pathways have been identified in Pd-based systems, encompassing surface-bound radicals,^19,20,23 nonradical singlet oxygen (¹O₂) species,²² and direct electron transfer mediation pathways,^21,24,25 despite the structural similarity of these catalysts. Therefore, more studies are warranted to fill knowledge gaps regarding the PMS activation mechanism of Pd SACs. Tailoring the electronic structure of Pd SACs by modulating the coordination environment and elucidating the PMS activation pathway is of great significance, and it can guide the fabrication of highly efficient Pd SACs for HOC degradation.

The electronic configuration of transition metal SACs can be precisely engineered through targeted modifications of their short-range^26,27 or long-range²⁸ coordination environments, and doping p-block heteroatoms with electron-withdrawing or donating capability offers an effective way to modulate the coordination environment of SACs. For example, B can extract electrons through a vacant 2p_z orbital conjugated with the carbon π system after being doped into the carbon support,²⁹ which can result in the depletion of the electron density on the supported SAC. In contrast, doped P has a high potential to donate its readily available lone pair of electrons, which can increase the electron density of the supported SAC.²⁹ Notably, manipulating the electronic structure of some other SACs has been evidenced to significantly regulate their intrinsic activities in PMS activation.^30–33 Inspired by this, doping specific amounts of particular p-block heteroatoms into the support may be a reasonable way to tune the electron density of the supported Pd SAC and optimize its catalytic performance for PMS activation and HOC degradation.

Herein, we designed a series of carbon nitride (CN) supported Pd SACs with different valence states and investigated the effect of Pd electronic density modulation on PMS activation for HOC oxidation. By doping specific amounts of B or P into CN supports through a hydrogen bonding assisted self-assembling method, Pd SACs with high valence states (denoted as Pd₁/BCN-1 and Pd₁/BCN-5 with the molar ratios of B/Pd being 1 and 5, respectively) and low valence states (denoted as Pd₁/PCN-1 and Pd₁/PCN-5 with the molar ratios of P/Pd being 1 and 5, respectively) were synthesized. Characterization and partial density of states (PDOS) calculations were carried out to verify the modulation of the density of Pd SACs by B or P doping. Differences in PMS activation and 4-CP degradation by the Pd SACs were examined and discussed, and the potential effects of coexisting environmental ions and natural organic matter (NOM) on 4-CP degradation were tested. The results will open an avenue for regulating the electron density of Pd SACs to accelerate PMS activation and HOC degradation.

2. Experimental section

2.1. Chemicals and materials

Detailed information on the chemicals and materials employed in this study is accessible in Text S1.†

2.2. Catalyst synthesis

By adapting the H-bonding-assisted self-assembly and pyrolysis approach,^29,34,35 Pd SACs doped with specific contents of electron-deficient B atoms (denoted as Pd₁/BCN) and electron-rich P atoms (denoted as Pd₁/PCN) were successfully synthesized (Fig. 1a). Briefly, the self-assembled CN-supported Pd SAC precursor was obtained by chelating ligands with Pd ions and was then coordinated intramolecularly with melamine and boric/phosphoric acid through H-bonding. The CN support and a control material (denoted as Pd₁/CN) without decorated atoms were also synthesized by similar methods for comparison. The synthesis details are documented in Text S2.†

2.3. Characterization

Quantification of Pd mass loading in the catalysts was performed using inductively coupled plasma-mass spectrometry (ICP-MS, Thermo Fisher iCAP RQ, USA). Structural properties including Brunauer–Emmett–Teller (BET) surface area and Barrett–Joyner–Halenda (BJH) pore volume were characterized via gas physisorption measurements on a Nova 2000e system (Quatachrome, USA). The morphology of the CN support and CN supported Pd SACs was characterized using scanning electron microscopy (SEM, Hitachi Zeiss S-3400N, Japan) along with energy-dispersive X-ray spectroscopy (EDS). Atomic-level elemental distribution of Pd atoms was acquired via an aberration-corrected high-angle annular dark field scanning transmission electron microscopy (AC-HAADF-STEM, JEOL JEM-ARM200F, Japan). The dispersion fidelity of CN supported Pd atoms was quantified through statistical comparison of the nearest neighbor (NN) distance histograms with Monte Carlo-simulated distribution models (theoretical random dispersion).^36,37 The crystal phase was evaluated by X-ray diffraction analysis (XRD, PANalytical X’PERT Pro, Netherlands). The oxidation state of Pd atoms was analyzed using high-resolution X-ray photoelectron spectroscopy (XPS, Thermo Fisher ESCALab-200i-XL, USA). The X-ray absorption near edge structure (XANES) and extended X-ray absorption fine structure (EXAFS) spectra were collected at the BL08U1A and BL11B beamlines in the Shanghai Synchrotron Radiation Facility (SSRF), respectively. Electrochemical characterization including linear sweep voltammetry (LSV) and chronoamperometry measurements was conducted with an electrochemical workstation (CHI760E, Chenhua Instrument, China). The details of these characterization methods are present in Text S3.†

2.4. Catalytic experiments

The catalytic reactions were carried out in a magnetically stirred beaker (100 mL) at 25 °C. The reaction systems were formulated by adding 50 mL DI water, catalyst (1 g L⁻¹) and 4-CP (100 mg L⁻¹) into batch reactors. The heterogeneous mixture was stirred vigorously for 30 min to achieve adsorption–desorption equilibrium. Then, the 4-CP degradation reaction was initiated by spiking 0.2 mL of PMS stock solution (0.5 mol L⁻¹) into the reactor. At every predetermined time point, 100 μL of the sample was aspirated and immediately quenched with 900 μL of methanol solution (50 mM), followed by filtering through a polyethersulfone filter (0.22 μm) for the measurement of residual 4-CP in the filtrate. The kinetic rate constants of 4-CP degradation were calculated as follows:

kobs = −ln(C/C0)/t	(1)

kSA = kobs/SA	(2)

kper-site = kobs × M/(m × wt%)	(3)

where k_obs (min⁻¹) is the apparent rate constants, C is the concentration of 4-CP at time t, C₀ is the initial concentration of 4-CP, k_SA (L min⁻¹ m⁻²) is surface area (SA) normalized rate constants, k_per-site (min⁻¹ mol⁻¹) is the catalyst molar number normalized rate constants, M denotes the molar mass of Pd, m represents the mass of the catalyst, and wt% corresponds to the Pd mass loading.

The catalytic durability assessment of Pd/CN comprised five consecutive cycles, each initiated by spiking 0.5 mL of the 4-CP solution into the reaction system. 100 μL samples were withdrawn at selected intervals and quenched for the determination of residual 4-CP concentration. To clarify the contribution of reactive oxygen species (ROS) to 4-CP degradation, various scavengers for radicals and nonradicals were used in the 4-CP degradation experiments. Typically, specified quantities of scavengers, as calibrated by the molar ratio to PMS (200 mM, 100 times the PMS dosage), including methanol (MeOH), ethanol (EtOH), tert-butanol (TBA), and furfuryl alcohol (FFA), were added before PMS injection to distinguish the potential participation of ·OH,³⁸ SO₄˙⁻,³⁹ O₂˙⁻,⁴⁰ and ¹O₂,^41,42 respectively. To test the effects of coexisting ions and NOM on 4-CP degradation, NO₃⁻, SO₄²⁻, Cl⁻, and HCO₃⁻ anions (100 mg L⁻¹) from their corresponding sodium salts and humic acid (HA, 10–50 mM) were added into the reaction solution before the reaction started. The influence of pH on the reactivity of Pd₁/BCN-5 for 4-CP degradation was also investigated, by adjusting the reaction solution pH within 3.5–9.2 using HCl and NaOH solutions. Natural water samples were collected from tap water, a groundwater well, and the Yuhangtang River in Hangzhou city (China), located at (120°00′42′′E, 30°17′43′′ N), (120°02′20′′ E, 30°14′57′′ N), and (120°11′45′′ E, 30°15′3′′ N), respectively. The physicochemical properties of the water samples were analyzed and are provided in Table S1.† The 4-CP degradation performances of the catalyst-activated PMS in the water samples were evaluated following the same procedures described above in the reaction solution. All the above reactions were carried out in triplicate and the average values with their standard deviations are reported.

2.5. Analysis methods

The PMS concentration was determined using the 2,2′-azino-bis(3-ethylbenzothiazoline)-6-sulfonic (ABTS) acid colorimetric method.⁴³ The concentrations of 4-CP were measured with a high-performance liquid chromatography (HPLC, SHIMADZU LC-20AD, Japan) equipped with a C18 reverse-phase column and an ultraviolet (UV) detector. The concentration of anions in the solution was detected by an Ion Chromatography system (IC, Metrohm IC 883, Switzerland) fitted with an anion-exchange column (Metrosep A Supp 5-250/4.0). Radical species formation was identified by electron paramagnetic resonance spectroscopy (EPR, Bruker, E500, Germany) employing 5,5-dimethyl-1-pyrroline N-oxide (DMPO) as the spin-trapping agent at 298 K. Full details of these analytical methods are available in Text S4.†

2.6. Density functional theory (DFT) calculations

DFT calculations were performed to elucidate the effect of B or P doping on the catalytic performance of Pd SACs in PMS activation. We chose Pd₁/CN, Pd₁/PCN-5, and Pd₁/BCN-5 as representative catalysts. The local structures of the Pd SACs were constructed based on their characterization results, and their corresponding PDOS were calculated. Geometric and energy calculations were performed via an iterative DFT procedure implemented within the VASP's plane-wave basis framework.⁴⁴ The Generalized Gradient Approximation (GGA) with the Perdew–Burke–Ernzerhof functional (PBE) was used to describe the exchange–correlation potential.⁴⁵ The core–valence interactions were modeled by the projector-augmented wave method.^46,47 The plane wave basis expansion employed a kinetic energy cutoff of 520 eV. The energy convergence criterion was set to 10⁻⁵ eV in the iterative solution of the Kohn–Sham (K–S) equation (eqn (4)):

Eads = EM/PMS − EM − EPMS	(4)

where E_M/PMS represents the energy of the catalyst (Pd–P, Pd, and Pd–B) after the adsorption of PMS, E_M represents the energy of the catalyst, and E_PMS represents the energy of PMS. All structures were relaxed until the residual forces on the atoms were reduced to less than 0.02 eV Å⁻¹. Charge density differences were employed to analyze the electronic structures of the selected catalysts. The charge density difference is defined as eqn (5):

Δρ = ρA–B − ρA − ρB	(5)

where ρ_A–B represents the optimized structural charge density at the interface of the catalyst–PMS complex; ρ_A and ρ_B represent the charge densities of the catalyst and PMS, respectively. The atomic positions were fixed when calculating the latter two ρ.

3. Results and discussion

3.1. Morphologies and structures of the CN-supported Pd SAC

The CN support and Pd₁/CN catalysts display rod-like structures (Fig. S1a and b†), while Pd₁/BCN and Pd₁/PCN catalysts exhibit graphene-like layered structures (Fig. S1c–f†). The morphology changes of the CN support in Pd₁/BCN and Pd₁/PCN could be attributed to the introduction of ethanol and dopant ligands (e.g., phosphoric acid and DMAB) during the catalyst synthesis process, which may change the supramolecular morphology and structure of the CN support.⁴⁸ The Pd mass contents in all five catalysts were about 1.0 wt% (Fig. S2†), while B in Pd₁/BCN and P in Pd₁/PCN reached 0.15–0.91 wt% (Table S2†) and 0.61–3.26 wt% (Table S3†), respectively. The SEM and EDS images of Pd₁/PCN-5 and Pd₁/BCN-5 (Fig. S3†) confirm the coincident distribution of Pd atoms and the doped heteroatoms (P or B) on the CN support. These results provide robust evidence for the anchoring of Pd atoms and heteroatom doping in the CN support. N₂ physisorption analysis revealed that although the morphology of CN supports in Pd₁/CN, Pd₁/PCN, and Pd₁/BCN changed after B and P doping, they still had comparable specific surface areas and pore size distributions (Fig. S4†), ensuring that the catalytic performance originates exclusively from the deposited Pd sites on these heteroatom doped CN supports.

AC-HAADF-STEM analysis showed isolated Pd atoms (the bright spots marked with yellow circles) present on the CN support (Fig. 1b–f), without any Pd nanoparticles. The atomic-scale dispersion of Pd atoms in different catalysts was statistically characterized using NN distance frequency matrices (Fig. 1g–k). The measured NN distances of Pd atoms in Pd₁/CN, Pd₁/PCN-1, Pd₁/PCN-5, Pd₁/BCN-1, and Pd₁/BCN-5 reached 0.38, 0.37, 0.35, 0.38 and 0.38 nm, respectively, which substantially exceed the average metallic Pd–Pd bond length of 0.27 nm (Table S4†). The experimental NN distance histograms of Pd atoms demonstrated exceptional agreement with the Rayleigh distributions (Fig. 1g–k), further signifying the random and atomic dispersion of Pd atoms in the catalysts. No typical peaks ascribed to crystalline Pd species were observed in the XRD patterns (Fig. S5†), further verifying the atomic dispersion of Pd atoms. Notably, the characteristic peaks corresponding to the (002) lattice plane of CN exhibited a −0.1° and 0.15° shift after the doping of B and P, respectively (Fig. 2a), suggesting the modulation of the coordination environment of the Pd SACs on the B or P doped CN supports as discussed below.

3.2. Chemical state and atomic coordination configurations of the CN-supported Pd SAC

The XPS spectra of the catalysts are shown in Fig. S6–S10.† Similar C 1s spectra (Fig. S7†) and N 1s spectra (Fig. S8†) with their corresponding deconvoluted peaks suggest that the incorporation of B and P did not alter the main tri-s-triazine units in the CN support. Compared with Pd₁/CN, both the N–CN/CN₃ ratio in the C 1s spectra (Fig. 2b) and the C–NC/NC₃ ratio in the N 1s spectra (Fig. 2c) increased in the heteroatom doped CN supports. The changes in N–CN/CN₃ and C–NC/NC₃ ratios could be associated with the substitution of N by B or P atoms at the junction of the triazine ring in the CN scaffold,^15,29,49 as proved by the deconvoluted peaks in the P 2p spectra of Pd₁/PCN (Fig. S9a and b†) and the B 1s spectra of Pd₁/BCN (Fig. S9c and d†). Due to the electron-deficient nature of B and the electron-rich nature of P atoms, the doped heteroatoms could alter the electronic structure of the CN support.²⁹ This was confirmed by the positive binding energy shifts in the C 1s (Fig. 2d) and N 1s XPS spectra (Fig. 2e) of the B-doped CN support, and the negative shifts in the C 1s and N 1s XPS spectra of the P-doped CN support. As a rule of thumb, changes in the electronic configuration of the CN support would lead to variations in the coordination environment of the Pd active sites,¹⁵ thus altering their corresponding chemical states. As revealed in Fig. S10,† the prominent Pd atoms in Pd₁/BCN catalysts exist as Pd⁴⁺ (338.5 eV) and Pd²⁺ (337.0 eV), while a greater proportion of Pd atoms are present as Pd⁰ at 335.0 eV in Pd₁/PCN. The valence distributions of Pd⁰, Pd²⁺, and Pd⁴⁺ in all catalysts are shown in Fig. 2f, and we find that the proportion of Pd⁴⁺ species in Pd₁/BCN-5 is 11.9% and 62.9% higher than that in Pd₁/CN and Pd₁/PCN-5, respectively. The average valence states of Pd atoms in each catalysts followed the trend of Pd₁/PCN-5 < Pd₁/PCN-1 < Pd₁/CN < Pd₁/BCN-1 < Pd₁/BCN-5 (Fig. S11†), consistent with the binding energy shifts observed in C 1s (Fig. 2d) and N 1s XPS (Fig. 2e) spectra. This result indicates the vital role of the coordination configuration of the CN support in determining the valence state of the deposited Pd atoms.

The XANES analyses of C, N, and Pd atoms further revealed the role of the coordination environment of the CN support in determining the oxidation state of the deposited Pd SAC. As shown in Fig. 3a, compared with Pd₁/BCN, the peaks attributed to the π* excitations of C–N–C (285.1 eV) and –CN₃ (288.1 eV)⁵⁰ shift to higher photon energy in Pd₁/PCN, suggesting a lower valence state of C atoms in Pd₁/PCN catalysts. Similarly, the higher energy shifting of the peaks attributed to the π* excitations of C–NC (399.8 eV) and –NC₃ (402.4 eV)⁵¹ in the N K-edge XANES spectra of Pd₁/PCN indicates the lower valence state of N atoms in the CN support of Pd₁/PCN catalysts (Fig. 3b). These results imply that the doping of electron-rich P atoms tailors a reducing coordination geometry for Pd active sites, while the incorporation of electron-deficient B resulted in an oxidizing coordination configuration. In line with the results of C K-edge and N K-edge XANES analyses of the catalysts, the pre-edge absorption peak thresholds in the K-edge XANES spectra of Pd demonstrated a systematic modulation of Pd valence states, with Pd₁/BCN-5 exhibiting the most oxidized Pd atoms and Pd₁/PCN-5 containing the most reduced Pd atoms (Fig. 3c). This valence state progression is also consistent with the corresponding Pd XPS results (Fig. 2f).

EXAFS spectra were further used to demonstrate the fine coordination environment of the deposited Pd atoms, and the quantitative fitting analysis results demonstrated excellent consistency with experimental spectral data (Fig. S12,†3d, and Table S4†). The Fourier transform EXAFS (FT-EXAFS) spectra of the catalysts exhibit one major peak ascribed to the first shell of Pd–N coordination at about 1.5 Å (Fig. 3d). The coordination number of all catalysts was close to 4 (Table S4†), suggesting a Pd–N₄ configuration in these catalysts. No obvious peak corresponding to Pd–Pd coordination at 2.5 Å was found, further verifying the mononuclear dispersion of Pd species in the catalysts. It is noteworthy that the Pd–N coordination peak shifted to 1.4 Å for Pd₁/PCN and 1.6 Å for Pd₁/BCN, which could be caused by the incorporation of B and P atoms, respectively, consistent with the above AC-HAADF-STEM and XRD analysis results. Since wavelet transform provides powerful resolution in both K and R spaces and discriminates the backscattering atoms, the EXAFS oscillation of the Pd K-edge EXAFS is further discussed. The wavelet transform contour plot of Pd₁/CN showed only one intensity maximum at 6.9 Å attributed to the Pd–N bond,¹⁵ which shifted to 7.2 Å in Pd₁/PCN-5 and 5.2 Å in Pd₁/BCN-5 (Fig. 3e). This further confirms that the doping of B and P atoms in CN supports could alter the coordination environment of CN for the deposited Pd SAC, thus affecting the Pd chemical state and the catalytic performance as discussed below.

3.3. The effect of the electronic structure of Pd SACs on PMS activation and 4-CP degradation

Both the bare CN and heteroatom doped CN supports could not catalyze PMS for 4-CP degradation (Fig. S13†), implying that the observed 4-CP removal in the catalyst systems could only be caused by the deposited Pd SAC sites. Despite all these Pd SAC-based catalysts possessing comparable surface areas (Fig. S3†) and Pd mass loadings (Fig. S4†), their activities in PMS activation and 4-CP degradation vary significantly (Fig. 4a). Compared with Pd₁/CN, Pd₁/PCN displayed the lowest catalytic activity for PMS activation and 4-CP degradation, with only about 40% of 4-CP removed after more than 1 h, demonstrating the low catalytic performance of low valent Pd SACs in PMS activation and the sequential 4-CP degradation. In contrast, Pd₁/BCN exhibited much higher activity in 4-CP degradation, as indicated by more than 95% 4-CP removal by Pd₁/BCN-5 within less than 5 min (Fig. 4a), indicating that the high valent Pd SAC in Pd₁/BCN-5 played a key role in PMS activation and 4-CP degradation. The degradation kinetics of 4-CP were analyzed by using pseudo-first-order reaction kinetics, enabling quantitative comparison of the catalytic activities of the catalysts. As depicted in Fig. 4b, Pd₁/BCN-5 achieved a surface area normalized 4-CP degradation rate of 2.6 × 10⁻³ L min⁻¹ m⁻², which is 13.0 folds higher than that of Pd₁/CN and 43.3–86.7 folds higher than that of Pd₁/PCN. Additionally, the surface area normalized kinetic rates of these deposited Pd SACs for 4-CP removal correlate well with the valence states of Pd atoms (Fig. 4c); the B-doped Pd SACs with relatively higher valent Pd showed higher 4-CP degradation rates, while P doping lowered the valence of Pd and resulted in a lower 4-CP removal rate. The enhancement in PMS activation and 4-CP degradation by Pd₁/BCN catalysts confirms that the valence state of Pd SACs triggers a considerable impact on oxidative catalysis. Furthermore, the catalytic performance of Pd₁/BCN-5 in PMS activation and HOC degradation is superior to that of other state-of-art catalysts (Fig. S14 and Table S5†). Therefore, targeted electronic structure engineering of Pd SACs provides a promising approach to optimize their performance in catalytic degradation of HOCs.

The catalytic performance of Pd₁/BCN-5 was quite stable, with no observable decrease in the 4-CP degradation rate (remaining at 0.47 h⁻¹) throughout the 5-cycle consecutive experiment (Fig. 4d). The leaching of Pd atoms in the Pd₁/BCN-5 catalyst was negligible after the 5-cycle experiment (Fig. S15†), and both the coordination structure (Fig. S16a†) and the average Pd valence (Fig. S16b†) of Pd₁/BCN-5 remained largely unchanged after the recycling experiment. These results verify the high stability of high valent Pd SACs, which can be a major factor contributing to the stable catalytic performance of the Pd₁/BCN-5 catalysts in the recycling experiment. In addition, Pd₁/BCN-5 exhibited satisfactory catalytic performance in the presence of a wide range of ions and NOM (Fig. S17†), and under varying pH (Fig. S18†), indicating that the effects of Cl⁻ and other ions on the catalytic performance of Pd₁/BCN-5 in PMS activation and 4-CP degradation could be negligible, if any. As anticipated, Pd₁/BCN-5 also showed comparable reactivities in DI water, tap water, groundwater, and river water (Fig. S19† and 4e). These results demonstrate considerable application potential and reactivity robustness of Pd₁/BCN under real environmental conditions.

3.4. Mechanism of high valent Pd SACs in PMS activation and 4-CP degradation

Neither radicals nor nonradical ROS were generated in the Pd₁/BCN-5 activation system, as indicated by the unaffected 4-CP degradation kinetics regardless of the addition of MeOH (for ·SO₄⁻ quenching³⁹), EtOH (for ·OH quenching³⁸), TBA (for O₂˙⁻ quenching⁴⁰), or FFA (for ¹O₂ quenching^41,42) at high dosages (200 mM, 100 times the PMS dosage) (Fig. S20† and 4f). This suggests that the non-ROS reaction pathway might be responsible for the 4-CP degradation in the Pd₁/BCN activation system. Notably, the consumption of PMS is negligible without the addition of 4-CP, and a decrease in PMS concentration was only observed in the presence of 4-CP (Fig. S21†), indicating that the PMS was exclusively consumed for the degradation of 4-CP and further indicating the non-ROS pathway of PMS activation and 4-CP degradation by Pd1/BCN-5.^52–54 Negligible EPR signals of DMPO–·OH (Fig. 4g) and DMPO–O₂˙⁻ (Fig. 4h) were detected for Pd₁/BCN-5 in water or methanol solution, providing another strong evidence for the non-ROS pathway.⁵⁵ Since ¹O₂ is commonly generated via O₂˙⁻,^32,56 the absence of O₂˙⁻ also implies that ¹O₂ was not the major reactive species either. It has been reported that the presence of HCO₃⁻ could inhibit the formation of ¹O₂ and the degradation of pollutants by O₂˙⁻,^53,57 while the negligible influence of HCO₃⁻ on 4-CP degradation (Fig. S16†) suggests that O₂˙⁻ was also irrelevant in the Pd₁/BCN-PMS activation system. Thus, the above results ruled out all possible radical or nonradical ROS reaction pathways, implying that the degradation of 4-CP could be ascribed to the direct electron transfer (DET) between 4-CP and PMS on the catalyst surfaces. The DET reaction pathway on the catalysts enables highly selective oxidation of the 4-CP and other electron-rich pollutants,²⁴ which is helpful for PMS utilization, HOC degradation, and catalytic stability of Pd SACs. Based on the results of PMS consumption (Fig. S21a†)and total organic carbon removal (Fig. S21b†), and using established methods for quantifying PMS utilization,^58,59 we found that Pd₁/BCN-5 achieved a PMS utilization efficiency of 83.3% ± 3.7%, exceeding the efficiency of most state-of-the-art transition metal catalysts reported to activate PMS via the radical pathway.⁵⁸ Collectively, these results highlight the potential of Pd SACs, especially the B-doped Pd SACs, as efficient Fenton-like catalysts for the degradation of HOCs.

The electrochemical analysis results also highlight the important role of the direct oxidation reaction pathway (DET) in the 4-CP degradation in the Pd₁/BCN-PMS system. The LSV result (Fig. 5a) shows that the current of the Pd₁/BCN-5 electrode increased significantly in the presence of PMS, while the current decreased in the presence of 4-CP, indicating the electron transfer between PMS and 4-CP in the Pd₁/BCN-PMS system, which results in the rapid 4-CP degradation. Consistent with the results of LSV analysis, the chronopotentiometry measurements show that the current decreased instantaneously after PMS addition and then rebounded after the addition of 4-CP on the Pd₁/BCN-5 electrode (Fig. 5b), with a pattern similar to that of the DET reaction process on the surface of carbon nanotubes.²¹ Moreover, the increase in current on the catalyst electrode follows the trend of Pd₁/PCN-5 < Pd₁/CN < Pd₁/BCN-5, confirming the faster transfer of electrons on the surfaces of high valence-state Pd SACs.

DFT calculations further reveal the modulation effect of the Pd valence state on the activity of Pd atoms on PMS activation. In contrast to the Pd atoms in Pd₁/CN, the PDOS of Pd atoms in Pd₁/BCN-5 displays a negative shift (Fig. 5c), suggesting a decrease in the d-band center (Fig. 5d) and higher electronic coupling between Pd atoms and CN support. This could increase d-electrons at the energy level near the Fermi level and offer more active sites for PMS adsorption and activation.³² The charge density difference analysis also shows the higher electron accumulation around the PMS adsorbed on the surface of high valent Pd atoms in Pd₁/BCN-5 (Fig. 5e). Consistently, the Bader charge analysis shows that more electrons accumulated (i.e., yellow color) along PMS adsorbed on Pd atoms in Pd₁/BCN-5 (0.88e) compared with that in Pd₁/CN (0.86e). These results suggest that the electron transfer between PMS and Pd atoms was accelerated upon B doping in the CN support, thereby accelerating the 4-CP degradation process at the high valent Pd sites in the Pd₁/BCN-5 catalyst. In contrast, the PDOS of the low valent Pd atom in Pd₁/PCN-5 shows the opposite trend (Fig. 5c), corresponding to an up-shift of the d-band center (Fig. 5d) and a comparatively smaller Bader charge accumulation (0.85e) for the adsorbed PMS (Fig. 5e), consistent with the lower reactivity for PMS activation on the Pd atoms deposited on the P-doped CN support. This is further supported by the corresponding adsorption energies of PMS at different adsorption sites on each catalyst (Fig. S22 and Table S6†). These DFT calculation results confirm that PMS is preferably adsorbed on the B-doping sites with the lower adsorption energy in the catalysts (Fig. S22†). Pd₁/BCN-5 possessed the lowest adsorption energy, indicating that the Pd atoms in the B-doped CN support had optimal adsorption energy for PMS activation, which could lead to its superior catalytic performance in 4-CP degradation (Fig. 4a). Pd₁/PCN-5 displayed the strongest adsorption of PMS, leaving fewer reactive sites available for continued PMS activation and resulting in the poisoning of Pd sites and the deactivation of Pd₁/PCN-5 for 4-CP degradation.

4. Conclusion

A series of Pd SACs with tailored electronic structures were fabricated in this study, revealing that high-valent Pd species in the B-doped CN support had excellent reactivity in PMS activation and 4-CP degradation. Pd₁/BCN-5, possessing an 11.9% and 62.9% higher proportion of Pd⁴⁺ species compared to Pd₁/CN and Pd₁/PCN-5, respectively, demonstrated superior catalytic catalytic activity and stability in PMS activation and HOC degradation. Mechanistic studies identified the DET pathway between PMS and 4-CP on the catalyst surfaces as the dominant process, rather than the conventional radical routes. This work establishes that Pd sites are the primary active centers for PMS activation; the optimal adsorption behavior of high valent Pd atoms in Pd₁/BCN-5 for PMS contributes to superior PMS activation, facilitating the degradation of 4-CP. Therefore, targeted electronic structure engineering of high valent Pd SACs via p-block doping offers an effective strategy for improving their application in PMS-mediated catalytic degradation of 4-CP. The unaffected reactivity of B-doped Pd SACs for PMS activation and 4-CP degradation in the presence of ions and NOM, as well as under different pH conditions and in various real water samples highlights their potential in real application scenarios. This study elucidates the pivotal influence of Pd oxidation states in governing PMS activation and HOC degradation kinetics, providing an atomically precise design principle for engineering high-performance Pd – based catalysts for environmental remediation applications.

Data availability

The data supporting this article have been included as part of the ESI.†

Author contributions

Zhenjie Li: conceptualization, investigation, methodology, writing – original draft. Chaohuang Chen: data curation, investigation, validation. Kaijian Sang: data curation, validation. Xunheng Jiang: data curation, validation. Xinyue Wu: data curation, validation. Jiang Xu: data curation, resources. Kun Yang: resources. Daohui Lin: funding acquisition, supervision, validation, writing – review & editing.

Conflicts of interest

The authors have no competing financial interests.

Acknowledgements

This work was supported by the Key Research and Development Program of Zhejiang Province (2024C03228), the National Key Research and Development Program of China (2022YFC3702100), the National Natural Science Foundation of China (U21A20163), the Natural Science Foundation of Zhejiang Province (LMS25B070001), and the Postdoctoral Fellowship Program of CPSF (GZC20241474). The authors acknowledge Beijing PARATERA Tech Co., Ltd for providing HPC resources.

References

1. Y. Liu, E. S. Richardson, A. E. Derocher, N. J. Lunn, H. J. Lehmler, X. Li, Y. Zhang, J. Y. Cui, L. Cheng and J. W. Martin, Angew. Chem., Int. Ed., 2018, 57, 16401–16406 CrossRef CAS PubMed .
2. T. Ruan, P. Li, H. T. Wang, T. Li and G. Jiang, Chem. Rev., 2023, 123, 10584–10640 CrossRef CAS PubMed .
3. Y. Guo, Y. Li and Z. Wang, Water Res., 2023, 234, 119810 CrossRef CAS PubMed .
4. N. Jadbabaei, T. Ye, D. Shuai and H. Zhang, Appl. Catal., B, 2017, 205, 576–586 CrossRef CAS .
5. R. Liu, H. M. Chen, L. P. Fang, C. Xu, Z. He, Y. Lai, H. Zhao, D. Bekana and J. F. Liu, Environ. Sci. Technol., 2018, 52, 4244–4255 CrossRef CAS PubMed .
6. M. O. Nutt, K. N. Heck, P. Alvarez and M. S. Wong, Appl. Catal., B, 2006, 69, 115–125 CrossRef CAS .
7. G. C. Bond, Surf. Sci., 1985, 156, 966–981 CrossRef CAS .
8. W. Xu, J. Ni, Q. Zhang, F. Feng, Y. Xiang and X. Li, J. Mater. Chem. A, 2013, 1, 12811–12817 RSC .
9. H. N. Sharma, V. Sharma, A. B. Mhadeshwar and R. Ramprasad, J. Phys. Chem. Lett., 2015, 6, 1140–1148 CrossRef CAS PubMed .
10. X. Hua, W. Wang, W. Ran, G. He, J. Liu and R. Liu, Langmuir, 2023, 39, 4692–4700 CrossRef CAS PubMed .
11. L. Liu and A. Corma, Chem. Rev., 2018, 118, 4981–5079 CrossRef CAS PubMed .
12. J. Li, M. F. Stephanopoulos and Y. Xia, Chem. Rev., 2020, 120, 11699–11702 CrossRef CAS PubMed .
13. S. Weon, D. Huang, K. Rigby, C. Chu, X. Wu and J. Kim, ACS ES&T Eng., 2021, 1, 157–172 Search PubMed .
14. C. Chu, D. Huang, S. Gupta, S. Weon, J. Niu, E. Stavitski, C. Muhich and J. Kim, Nat. Commun., 2021, 12, 5179 CrossRef CAS PubMed .
15. Z. Li, Z. Guo, X. Wu, X. Jiang, H. Li, J. Xu, K. Yang and D. Lin, ACS Nano, 2023, 17, 22859–22871 CrossRef CAS PubMed .
16. Y. Luo, Y. Cai, X. Long, D. Zhou, C. Zhou and B. E. Rittmann, Environ. Sci. Technol., 2022, 56, 4447–4456 CrossRef CAS PubMed .
17. X. Jiang, L. Zhang, H. Liu, D. Wu, F. Wu, L. Tian, L. Liu, J. Zou, S. Luo and B. Chen, Angew. Chem., Int. Ed., 2020, 59, 23112–23116 CrossRef CAS PubMed .
18. L. Zhang, X. Jiang, Z. Zhong, L. Tian, Q. Sun, Y. Cui, X. Lu, J. Zou and S. Luo, Angew. Chem., Int. Ed., 2021, 60, 21751–21755 CrossRef CAS PubMed .
19. M. Gao, Y. Ren, W. Zheng, M. Liu, X. Wang, N. Nanayakkara and Y. Liu, Appl. Catal., B, 2025, 371, 125218 CrossRef CAS .
20. Y. Feng, P. Lee, D. Wu and K. Shih, Water Res., 2017, 120, 12–21 CrossRef CAS PubMed .
21. Y. Ahn, E. Yun, J. Seo, C. Lee, S. Kim, J. Kim and J. Lee, Environ. Sci. Technol., 2016, 50, 10187–10197 CrossRef CAS PubMed .
22. Y. Wu, L. Chen, P. Sun, Q. Zhang, Y. Chen and X. Zheng, Appl. Catal., B, 2024, 340, 123213 CrossRef CAS .
23. J. Deng, F. Li, Z. Qi, W. Huang, Z. Wan, L. Zhang, D. Zheng, G. Li and F. Zhang, Appl. Catal., B, 2024, 359, 124499 CrossRef CAS .
24. Y. Si, Z. Guo, Y. Meng, H. Li, L. Chen, A. Zhang, C. Gu, W. Li and H. Yu, Environ. Sci. Technol., 2022, 56, 564–574 CrossRef CAS PubMed .
25. Y. Ahn, H. Bae, H. Kim, S. Kim, J. Kim, S. Lee and J. Lee, Appl. Catal., B, 2019, 241, 561–569 CrossRef CAS .
26. J. Wan, Z. Zhao, H. Shang, B. Peng, W. Chen, J. Pei, L. Zheng, J. Dong, R. Cao, R. Sarangi, Z. Jiang, D. Zhou, Z. Zhuang, J. Zhang, D. Wang and Y. Li, J. Am. Chem. Soc., 2020, 142, 8431–8439 CrossRef CAS PubMed .
27. Y. Wang, R. Shi, L. Shang, G. I. N. Waterhouse, J. Q. Zhao, Q. Zhang, L. Gu and T. Zhang, Angew. Chem., Int. Ed., 2020, 59, 13057–13062 CrossRef CAS PubMed .
28. H. Shen, E. Gracia-Espino, J. Ma, K. Zang, J. Luo, L. Wang, S. Gao, X. Mamat, G. Hu, T. Wagberg and S. Guo, Angew. Chem., Int. Ed., 2017, 56, 13800–13804 CrossRef CAS PubMed .
29. X. Zhou, M. Ke, G. Huang, C. Chen, W. Chen, K. Liang, Y. Qu, J. Yang, Y. Wang, F. Li, H. Yu and Y. Wu, Proc. Natl. Acad. Sci. U. S. A., 2022, 119, 2119492119 CrossRef PubMed .
30. J. S. Song, N. N. Hou, X. C. Liu, G. Y. Bi, Y. Wang and Y. Mu, Adv. Mater., 2024, 36, e2405832 CrossRef PubMed .
31. Y. Lin, Y. Wang, Z. Weng, Y. Zhou, S. Liu, X. Ou, X. Xu, Y. Cai, J. Jiang, B. Han and Z. Yang, Nat. Commun., 2024, 15, 10032 CrossRef CAS PubMed .
32. X. Jiang, B. Zhou, W. Yang, J. Chen, C. Miao, Z. Guo, H. Li, Y. Hou, X. Xu, L. Zhu, D. Lin and J. Xu, Proc. Natl. Acad. Sci. U. S. A., 2024, 121, 2309102121 CrossRef PubMed .
33. Q. Wu, Z. Yang, Z. Wang and W. Wang, Proc. Natl. Acad. Sci. U. S. A., 2023, 120, 2219923120 CrossRef PubMed .
34. C. Hu, J. Hu, Z. Zhu, Y. Lu, S. Chu, T. Ma, Y. Zhang and H. Huang, Angew. Chem., Int. Ed., 2022, 61, 202212397 CrossRef PubMed .
35. Y. Liu, J. Sun, H. Huang, L. Bai, X. Zhao, B. Qu, L. Xiong, F. Bai, J. Tang and L. Jing, Nat. Commun., 2023, 14, 1457 CrossRef CAS PubMed .
36. E. Vorobyeva, E. Fako, Z. Chen, S. M. Collins, D. Johnstone, P. A. Midgley, R. Hauert, O. V. Safonova, G. Vilé, N. López, S. Mitchell and J. Pérez-Ramírez, Angew. Chem., Int. Ed., 2019, 58, 8724–8729 CrossRef CAS PubMed .
37. Z. Chen, S. Mitchell, E. Vorobyeva, R. K. Leary, R. Hauert, T. Furnival, Q. M. Ramasse, J. M. Thomas, P. A. Midgley, D. Dontsova, M. Antonietti, S. Pogodin, N. López and J. Pérez-Ramírez, Adv. Funct. Mater., 2017, 27, 1605785 CrossRef .
38. Y. Shang, X. Xu, Q. Yue, B. Gao and Y. Li, Environ. Sci.: Nano, 2020, 7, 1444–1453 RSC .
39. C. Guan, J. Jiang, Y. Shen, S. Pang, C. Luo and X. Zhao, Environ. Sci. Technol., 2019, 53, 9054–9062 CrossRef CAS PubMed .
40. X. Li, T. Hou, L. Yan, L. Shan, X. Meng and Y. Zhao, J. Hazard. Mater., 2020, 398, 122884 CrossRef CAS PubMed .
41. S. Mostafa and F. L. Rosario-Ortiz, Environ. Sci. Technol., 2013, 47, 8179–8186 CrossRef CAS PubMed .
42. W. Ren, L. Xiong, X. Yuan, Z. Yu, H. Zhang, X. Duan and S. Wang, Environ. Sci. Technol., 2019, 53, 14595–14603 CrossRef CAS PubMed .
43. P. Shao, J. Tian, F. Yang, X. Duan, S. Gao, W. Shi, X. Luo, F. Cui, S. Luo and S. Wang, Adv. Funct. Mater., 2018, 28, 1705295 CrossRef .
44. J. Hafner, Comput. Phys. Commun., 2007, 177, 6–13 CrossRef CAS .
45. G. Kresse and J. Furthmuller, Phys. Rev. B, 1996, 54, 11169–11186 CrossRef CAS PubMed .
46. J. P. Perdew, J. A. Chevary, S. H. Vosko, K. A. Jackson, M. R. Pederson, D. J. Singh and C. Fiolhais, Phys. Rev. B, 1992, 46, 6671–6687 CrossRef CAS PubMed .
47. J. P. Perdew, J. A. Chevary, S. H. Vosko, K. A. Jackson, M. R. Pederson, D. J. Singh and C. Fiolhais, Phys. Rev. B, 1993, 48, 4978 CrossRef CAS PubMed .
48. Z. X. Zhou, Y. Y. Zhang, Y. F. Shen, S. Q. Liu and Y. J. Zhang, Chem. Soc. Rev., 2018, 47, 2298–2321 RSC .
49. H. L. Poh and M. Pumera, Chemelectrochem, 2015, 2, 190–199 CrossRef CAS .
50. Y. Wang, G. Jia, X. Cui, X. Zhao, Q. Zhang, L. Gu, L. Zheng, L. Li, Q. Wu, D. J. Singh, D. Matsumura, T. Tsuji, Y. Cui, J. Zhao and W. Zheng, Chem, 2021, 7, 436–449 CAS .
51. P. Liu, Z. Huang, X. Gao, X. Hong, J. Zhu, G. Wang, Y. Wu, J. Zeng and X. Zheng, Adv. Mater., 2022, 34, 2200057 CrossRef CAS PubMed .
52. H. Lee, H. Kim, S. Weon, W. Choi, Y. Hwang, J. Seo, C. Lee and J. Kim, Environ. Sci. Technol., 2016, 50, 10134–10142 CrossRef CAS PubMed .
53. J. Lee, U. von Gunten and J. Kim, Environ. Sci. Technol., 2020, 54, 3064–3081 CrossRef CAS PubMed .
54. W. Ren, C. Cheng, P. Shao, X. Luo, H. Zhang, S. Wang and X. Duan, Environ. Sci. Technol., 2022, 56, 78–97 CrossRef CAS PubMed .
55. A. D. Bokare and W. Choi, Environ. Sci. Technol., 2015, 49, 14392–14400 CrossRef CAS PubMed .
56. Q. Yi, J. Ji, B. Shen, C. Dong, J. Liu, J. Zhang and M. Xing, Environ. Sci. Technol., 2019, 53, 9725–9733 CrossRef CAS PubMed .
57. S. Zhu, X. Li, J. Kang, X. Duan and S. Wang, Environ. Sci. Technol., 2019, 53, 307–315 CrossRef CAS PubMed .
58. Y. Meng, Y. Q. Liu, C. Wang, Y. Si, Y. J. Wang, W. Q. Xia, T. Liu, X. Cao, Z. Y. Guo, J. J. Chen and W. W. Li, Nat. Commun., 2024, 15, 5314 CrossRef CAS PubMed .
59. Y. Wei, J. Miao, J. Ge, J. Lang, C. Yu, L. Zhang, P. J. J. Alvarez and M. Long, Environ. Sci. Technol., 2022, 56, 8984–8992 CrossRef CAS PubMed .

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d5ta03050a

This journal is © The Royal Society of Chemistry 2025