BiVO₄-assisted photocatalytic ozonation for efficient cyanide degradation in synthetic silver post-leaching effluents

TL;DR

This study shows that combining BiVO₄ photocatalysis with ozone efficiently removes cyanide from mining effluents, offering a sustainable solution.

Contribution

The novel contribution is the development of a BiVO₄-assisted photocatalytic ozonation system for efficient cyanide degradation in metallurgical effluents.

Findings

The BiVO₄/ozone system achieved faster cyanide abatement and improved efficiency compared to individual processes.

Complete cyanide removal was achieved in two cycles, though partial deactivation occurred in the third due to surface poisoning.

The process reduced ozone demand and consistently removed both free and complexed cyanide.

Abstract

Cyanide is extensively used in hydrometallurgical leaching due to its strong metal-chelating ability, yet its acute toxicity raises major environmental concerns. This study evaluates photoactivated BiVO₄ synthesized by the sol–gel method and combined with ozone for the degradation of metal–cyanide complexes in alkaline matrices (pH 10.5) containing silver, copper, iron, lead, and zinc, representative of effluents from silver leaching in the Mexican mining industry. Synthetic solutions were prepared and characterized to reproduce these conditions. Thermodynamic stability diagrams indicated that free cyanide and complexes such as Ag(CN)₂⁻, Cu(CN)₃2⁻, and Fe(CN)₆4⁻ persist at alkaline pH, while lead and zinc showed no tendency to form stable cyanide complexes. Oxidative treatments favored complex dissociation, with cyanate predicted as the main by-product; experimentally, ammonium was identified as the dominant degradation product. The BiVO₄/ozone system under irradiation demonstrated superior performance compared to individual processes, achieving faster cyanide abatement and improved efficiency. Kinetics followed a modified Langmuir–Hinshelwood model that accounted for both surface-mediated photocatalysis and bulk-phase ozone reactions. Complete cyanide removal was achieved in two consecutive cycles, though partial deactivation occurred in the third cycle due to BiVO₄ surface poisoning by adsorbed metals, confirmed by SEM/EDS and atomic absorption. Despite gradual loss of activity, the coupled process consistently removed both free and complexed cyanide while reducing ozone demand compared with conventional ozonation. These findings highlight photocatalytic ozonation with BiVO₄ as a promising and sustainable strategy for the treatment of metallurgical effluents, offering efficient pollutant degradation and lower oxidant consumption in comparison with conventional approaches.

The online version contains supplementary material available at 10.1007/s11356-026-37456-x.

Figures

Pourbaix diagrams used for the pre-degradation thermodynamic analysis of the chemical species in solution: a AgNO₃ ([Ag⁺] = 2.66 ppm), b CuSO₄·5H₂O ([Cu^2^⁺] = 12.00 ppm), c FeSO₄·7H₂O ([Fe^2^⁺] = 1.99 ppm), d Pb(NO₃)₂ ([Pb^2^⁺] = 2.05 ppm), e ZnSO₄·7H₂O ([Zn.^2^⁺] = 15.99 ppm), and f NaCN ([CN⁻] = 100 ppm)

Post-degradation thermodynamic analysis at an ORP equal to 0.75 V showing the predominance zone diagrams of: a AgNO₃ ([Ag⁺] = 2.66 ppm), b CuSO₄·5H₂O ([Cu^2^⁺] = 12.00 ppm), c FeSO₄·7H₂O ([Fe^2^⁺] = 1.99 ppm), d Pb(NO₃)₂ ([Pb^2^⁺] = 2.05 ppm), e ZnSO₄·7H₂O ([Zn^2^⁺] = 15.99 ppm), and f NaCN ([CN⁻] = 100 ppm)

UV–Vis spectra of [CN^─^]Total solution, explore the evolution of [Cu(CN)_3_]^2─^ complexes during degradation process by a reaction involving only ozone and b photocatalytic ozonation reaction

Reaction possible mechanism in the degradation of total cyanide in presence of photoactivated BiVO_4_ and ozone

Full text

Introduction

Cyanide is an extremely toxic chemical compound widely used in gold and silver mining. Owing to the availability of empty p and d orbitals, cyanide can undergo oxidation to form cyanate or thiocyanate species. In addition, the cyanide group exhibits strong nucleophilic character, allowing it to participate in additional reactions. As an anionic ligand, cyanide can donate an electron pair to metal d orbitals, forming stable σ bonds and coordination complexes with transition metals. Furthermore, the cyanide group can engage in conjugation with adjacent π systems, influencing electron distribution and thereby contributing to the stability and reactivity of the molecule (Berkinbayeva et al. 2025). Due to these properties, the cyanide ion shows a strong affinity for metals such as silver, gold, copper, iron, zinc, aluminum, and nickel, and it can also bind to highly toxic metals including lead, mercury, chromium, arsenic, manganese, and tin (Rashesh et al. 2009). In alkaline conditions (pH > 10), Au and Ag extraction can be efficiently performed using sodium cyanide (NaCN) due to its high selectivity. However, the resulting wastewater contains not only free cyanide but also trace concentrations of heavy and non-heavy metals, which are highly detrimental to the environment. Consequently, the resulting wastewater containing these metals and trace concentrations of heavy and non-heavy metals is highly detrimental to the environment (Pan et al. 2021).

A central issue is the pollution of surface water bodies caused by mining wastewater. Metal–cyanide complexes such as \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${(Fe,Co,Au)}_{y}{CN}_{x}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${(Zn,Ag,Cd, Ni, Cu, Cr, Mn)}_{y}{CN}_{x}$$\end{document} are middle to moderately toxic compared to free cyanide, which is the most toxic form (Donato et al. 2007). The contamination of water bodies can lead to severe health effects and environmental damage. Compounds containing cyanide ions and metal–cyanide complexes act rapidly as poisons, blocking the process of cellular respiration. Hypoxia disrupts the functioning of all body cells. Tissues with the highest oxygen metabolism, such as the brain and heart muscle, are the most sensitive to the toxic effects of cyanide. The estimated lethal dose is 1.5 mg of CN^−^ per kg of body weight for an adult human (Jaszczak et al. 2017). Upon ingestion, cyanide salts such as potassium and sodium cyanide release hydrogen cyanide in the acidic gastric environment, facilitating systemic absorption (Guo et al. 2025).

The concentrations of metallic elements in water matrices have been documented in wastewater from post-leaching processes. Quantified concentrations of metals such as transition metals (d-block), metalloids, so-called heavy metals, and rare earth elements in different water bodies across Mexico have been reported (Godwyn-Paulson et al. 2022; Martínez-Ayala et al. 2022; Ochoa-Contreras et al. 2023; Roman Ahumada-Mexía et al. 2021). The treatment of wastewater containing cyanide and metal–cyanide complexes involves oxidizing processes such as chlorination, sulfur dioxide, ozonation, hydrogen peroxide, electrochemical oxidation, biological treatments, and adsorption processes (Liu et al. 2023). An effective treatment technology for the removal of organic pollutants and wastewater remediation by advanced oxidation processes (AOPs) is the photocatalytic ozonation. AOPs are justified for treating cyanide-containing solutions because they generate highly reactive species, such as hydroxyl radicals (•OH), capable of rapidly and non-selectively degrading both free cyanide and metal–cyanide complexes. Unlike conventional oxidation methods, AOPs achieve higher mineralization efficiency and minimize the formation of toxic by-products. The combination of both photocatalysis and ozone (O_3_) has shown substantial improvement in efficiency and reduction in reaction time compared to individual processes (Aydin et al. 2016). O_3_ is a strong oxidant with an oxidation potential of 2.07 V vs. SCE, whereas the hydroxyl radical is the most reactive oxidizing agent, with an oxidation potential between 2.8 V (pH 0) and 1.95 V (pH 14) vs. SCE (Deng & Zhao 2015). The synergistic interaction between ozone and photocatalysis enables rapid and effective generation of •OH radicals, governed by favorable thermodynamics and reaction kinetics. Studies have shown that an efficient single-electron transfer pathway between conduction band electrons (CB-e⁻) and O₃ is crucial for •OH formation. Moreover, continuous ozone injection during the reaction further enhances the overall reaction kinetics (Chen et al. 2025).

Under alkaline conditions, O₃ undergoes decomposition into various reactive oxygen species, thereby accelerating pollutant degradation (Eqs. (1)–(6)) (Černigoj et al. 2007; Mecha & Chollom 2020).

\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$O_3+OH^-\rightarrow O_3^{\bullet-}+HO^\bullet$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$O_3^{\bullet-}\rightarrow O^{\bullet-}+O_2$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$O^{\bullet-}+H^+\rightarrow HO^\bullet$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$O_3+HO^\bullet\rightarrow O_2+HO_2^\bullet$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathrm O}_3+\mathrm{HO}^\bullet\rightarrow{\mathrm O}_2+\mathrm{HO}_2^\bullet$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$O_3+{HO}_2^\bullet\rightarrow2O_2+{HO}^\bullet$$\end{document}

On the other hand, the photocatalytic system generates reactive oxygen species through activation of the photocatalyst under irradiation, where different radicals drive the degradation process. The following pathways are proposed (Palomares-Reyna et al. 2023):

\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${BiVO}_{4}+h\nu \to {({BiVO}_{4})}^{e-}+{({BiVO}_{4})}^{h+}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${({BiVO}_4)}^{h+}+H_2O\rightarrow{BiVO}_4+{HO}^\bullet+H^+$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${{O}_{3}}_{(g)}+{H}_{2}O+h\nu \to {O}_{2}+{H}_{2}{O}_{2}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$H_2O_2+h\nu\rightarrow2{HO}^\bullet$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${({BiVO}_4)}^{e-}+O_3\rightarrow O_3^{\bullet-}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$O_3^{\bullet-}+H^+\rightarrow{HO}_3^\bullet$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${HO}_3^\bullet\rightarrow{HO}^\bullet+O_2$$\end{document}

Thus, photocatalytic ozonation has the capacity to break down organic pollutants and transform products into mineral acids and carbon dioxide (Pang et al. 2023). While ozone readily reacts with organic and inorganic compounds, its effectiveness is often constrained by limited ozone mass transfer into water, its thermodynamic instability, and the presence of ozone-resistant pollutants. Similarly, the photocatalysis process can be hindered by intrinsic electron–hole recombination, which slows down the reaction. Despite these challenges, photocatalytic ozonation remains a highly effective and environmentally friendly treatment technology for removing organic pollutants and remediating wastewater (Mehrjouei et al. 2015).

Different strategies have been implemented to enhance the degradation efficiency of cyanide-containing wastewater, considering aspects related to the design of novel photocatalysts to advanced materials-engineering approaches, including doping, defect control, morphology tailoring, particle-size reduction, porosity modulation, composite formation, selective co-catalyst deposition, conductive supports, and plasmonic surface modification, among others (Zhang et al. 2025). These strategies share a common goal: to improve light absorption, enhance charge separation, and maximize oxidative pathways, which are critical for cyanide degradation (Betancourt-Buitrago et al. 2019; Ren et al. 2021). Building upon these concepts, composite photocatalysts combining graphene oxide (GO), nano-titanium dioxide (TiO₂), and ZSM-5 zeolite have been developed, achieving enhanced photocatalytic oxidation of real cyanide-containing wastewater. The formulation containing 60 wt% TiO₂ exhibited the highest total-cyanide degradation efficiency (Pan et al. 2022). Expanding on this line of research, Zn(II)- and Fe(III)-oxide nanostructures supported on activated carbon have also been investigated for cyanide degradation through the H₂O₂/UV/catalyst advanced oxidation process (AOP). Under these conditions, a Fe^3^⁺/ZnO molar ratio of 6% delivered the highest degradation efficiency (Eskandari et al. 2017). In parallel, TiO₂–SiO₂ aerogels and V₂O₅–SiO₂ xerogels, as well as ozone alone assisted by UV irradiation, have been tested for free-cyanide removal. A comparative analysis of material and energy costs revealed that aerogel and xerogel systems are economically favorable, as their equipment costs are negligible compared to the ozone-based approach, which requires an expensive ozone generator (Ibrahim et al. 2003). More recently, attention has shifted toward multifunctional nanocomposites such as those derived from iron-porphyrin photosensitizers (Fe-TCPP), sulfur-doped TiO₂ (S–TiO₂), and reduced graphene oxide (rGO). The S–TiO₂@rGO–FeTCPP photocatalytic system, immobilized on polyurethane foam (PUF), exhibited remarkable performance, achieving over 91% free-cyanide degradation and 88% toxicity removal. In this case, sulfur doping induced a bandgap shift into the visible region, enhancing solar light utilization, while rGO facilitated charge transport and suppressed electron–hole recombination. Moreover, the Fe-TCPP photosensitizer efficiently absorbed photons, further promoting overall activity (Pattanayak et al. 2021). Another effective strategy involves the synthesis of AgVO₃/PAN via a combined electrospinning and hydrothermal process. Through an ion-exchange approach, AgVO₃ is subsequently transformed into a BiVO₄/AgI composite, leading to the construction of a ternary AgI–BiVO₄–AgVO₃/PAN heterojunction with a hierarchical architecture. This synergistic configuration provides an optimized redox potential distribution under visible-light irradiation, thereby imparting the composite with enhanced photocatalytic activity and broad-spectrum degradation capability. Notably, during rhodamine B photodegradation experiments, the composite catalyst maintained a degradation efficiency of 80.4%, while the flexible fibrous support exhibited excellent reusability (Liu et al. 2026).

Despite these advances, to the best of our knowledge, BiVO₄-assisted photocatalytic ozonation has not yet been reported for efficient cyanide removal from post-leaching effluents. This gap is particularly relevant since BiVO₄ is a visible-light-active photocatalyst with favorable band alignment for oxidative reactions, and its combination with ozone could potentially overcome the limitations of conventional UV-driven AOPs. Accordingly, the initial phase of our study was dedicated to elucidating how the synthesis route affects the structural, morphological, and photocatalytic properties of bismuth vanadate in the degradation of free cyanide. BiVO₄ was prepared via microwave-assisted hydrothermal and sol–gel routes. Both catalysts exhibited high free-cyanide degradation efficiencies; the principal difference lay in the reaction kinetics, underscoring the critical influence of synthesis method on photocatalytic performance (Vargas-Rueda et al. 2024). In this study, a comprehensive thermodynamic analysis of the pre- and post-degradation stages of free cyanide and metal–cyanide complexes was conducted using predominance zone diagrams, providing valuable insights into the reaction pathways in this multicomponent system. As a study case, the complete degradation of cyanide was examined in water matrices containing metal species such as Ag, Cu, Fe, Pb, and Zn, simulating a residual aqueous solution from silver leaching processes in the Mexican mining industry. The efficiency of total cyanide degradation by photocatalytic ozonation was evaluated, demonstrating that combining heterogeneous photocatalysis and ozonation is an effective and sustainable approach for the removal of cyanide species. In these reactions, BiVO₄ photocatalysts synthesized via the sol–gel method were employed based on previous studies demonstrating their superior photocatalytic performance (Vargas-Rueda et al. 2024). Therefore, the present work does not focus on an extensive characterization of the photocatalyst; instead, it aims to elucidate the reaction mechanism through thermodynamic analysis, kinetic modeling, and post-reaction characterization of both the treated solutions and the photocatalyst. However, detailed compositional and structural characterization of the BiVO₄ photocatalyst is provided in the Supplementary Information (Text S1). To analyze the degradation of total cyanide, considering ozone as a complementary oxidizing agent, the modified Langmuir–Hinshelwood kinetic model was applied. Reaction monitoring also included redox potential measurements to correlate the progression of cyanide degradation. The generation of ammonium as a degradation product of free cyanide was confirmed using the Nessler method. Finally, the stability of the BiVO₄ photocatalyst was assessed over multiple reaction cycles through elemental analysis (EDS) and metal quantification (AAS), while UV–Vis spectroscopy was used to track the degradation of copper–cyanide complexes as reference species to better understand the reaction mechanism.

Experimental

Synthesis of BiVO4 by sol gel technique

A 0.03 M solution of Bi(NO₃)₃∙5H₂O was prepared by dissolving it in 50 mL of 4 M nitric acid (HNO₃). Similarly, 0.03 M NH₄VO₃ was dissolved in 50 mL of 4 M ammonium hydroxide (NH₄OH). These precursor solutions were mixed and magnetically stirred for 30 min at room temperature. Following this, 100 mL of ethanol was added, and the mixture was heated at 70 °C. At this temperature, the solution stirred for 1 h, and 5 mL of 1 M acetic acid (CH₃COOH) and 50 mL of deionized water were added. The resulting tangerine-colored gel was dried in an oven at 100 °C for 48 h, followed by a calcination at 400 °C for 2 h (Vargas-Rueda et al. 2024).

Pre- and post-degradation analysis

The thermodynamic analysis was performed with Medusa-Hydra chemical equilibrium software. This algorithm is based on the free energy minimization and simultaneous equilibrium principles reported by Ericsson (Eriksson 1979). The metallic species were determined using an Agilent SpectrAA 220FS atomic absorption spectrophotometer (AAS). The surface morphology and elemental characterization were examined using a SUPRA 55VP field-emission scanning electron microscope (FESEM) coupled with energy-dispersive X-ray spectroscopy (EDX). The [Cu(CN)3]^2─^ complexes, used as reference species, and ammonium quantification by the Nessler method were analyzed using a PerkinElmer Lambda 35 UV/Vis spectrometer.

Photocatalytic ozonation test

The photocatalytic ozonation activity of BiVO₄ was evaluated by quantifying free cyanide using the 4500-CN⁻ D titrimetric method (American Public Health Association [APHA] et al. 2017). Sodium cyanide (NaCN, Sigma-Aldrich, ≥98%) was used as the free cyanide source. Metallic species of silver, copper, iron, lead, and zinc were introduced as silver nitrate (AgNO₃, Sigma-Aldrich, ≥99%), copper(II) sulfate pentahydrate (CuSO₄·5H₂O, Sigma-Aldrich, ≥98%), iron(II) sulfate heptahydrate (FeSO₄·7H₂O, Sigma-Aldrich, ≥99%), lead(II) nitrate (Pb(NO₃)₂, Sigma-Aldrich, ≥99%), and zinc sulfate heptahydrate (ZnSO₄·7H₂O, Sigma-Aldrich, ≥99%), respectively. The water matrix containing these chemical species was sonicated for 30 min in the dark. The solution was continuously sparged with ozone at a volumetric gas flow of 1 L min⁻^1^ and an ozone mass flow of 1000 mg h⁻^1^, generated by a Biozon OZC-0.5 GB ozone generator and irradiated with UV/Vis light from a UVP 99-0055−01 Model PS-1 Pen-Ray lamp (254 nm, 115 V/60 Hz). An external water recirculation system was employed to maintain a constant reactor temperature throughout the experiment.

The photocatalytic ozonation removal efficiency (η) of the free cyanide solutions was calculated using the following equation:

\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\eta =\frac{{C}_{o}-C}{{C}_{o}}*100$$\end{document}

where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${C}_{o}$$\end{document} is the concentration of free cyanide in solution before the illumination and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$C$$\end{document} is the concentration of free cyanide in solutions at given time intervals.

Results and discussion

Synthesis and characterization of the BiVO4 via sol–gel

The synthesis and characterization of the BiVO₄ photocatalyst have been described in detail in a previous publication (Vargas-Rueda et al. 2024), where two different synthesis routes were comparatively analyzed. That study concluded that the most photoactive BiVO₄ phase was obtained via the sol–gel method, primarily due to its higher crystallinity, improved phase purity, and more favorable electronic structure, which enhance light absorption and charge carrier separation compared to alternative synthesis routes; whence it is herein used to evaluate the cyanide degradation in synthetic silver post-leaching effluents. Supplementary Material 1 (SM 1) presents the X-ray diffraction (XRD) pattern of the as-synthesized BiVO₄ obtained by the sol–gel method. The diffraction pattern exhibits well-defined and sharp reflections, indicating a high degree of crystallinity. All diffraction peaks can be indexed to the monoclinic scheelite structure of BiVO₄ (m-BiVO₄), in good agreement with the standard reference pattern (ICDD PDF No. 96–901−3438). The most intense reflections correspond to the characteristic planes of monoclinic BiVO₄, confirming the successful formation of the desired photoactive phase. The surface morphology and microstructural features of the sol–gel-synthesized BiVO₄ were further examined by scanning electron microscopy (SEM), while the elemental composition was assessed by energy-dispersive X-ray spectroscopy (EDS), confirming the homogeneous distribution of Bi, V, and O across the material. The SEM micrographs (SM2) reveal a predominantly cuboid-like morphology, accompanied by irregular polyhedral crystallites, with cuboidal structures being the most abundant. The corresponding EDS analysis (SM3) indicates that the Bi/V atomic ratio is very close to the theoretical stoichiometric value of BiVO₄, demonstrating that the sol–gel method yields a highly stoichiometrically accurate composition.

In addition, the surface chemical composition of the sol–gel-synthesized BiVO₄ was analyzed by X-ray photoelectron spectroscopy (XPS) in terms of mass and atomic percentages (see SM 4). The XPS results confirm the presence of Bi, V, O, C, and N on the catalyst surface. The Bi/V atomic ratio is close to unity, indicating the successful formation of stoichiometric BiVO₄, while oxygen is the most abundant element, consistent with the oxide nature of the material. Carbon is attributed to adventitious surface contamination and residual organic species from the sol–gel precursors, whereas nitrogen is detected in trace amounts (≈0.8 at.%), likely originating from nitrate- or ammonium-based reagents used during synthesis. Overall, these results confirm the successful synthesis of phase-pure and stoichiometric BiVO₄ via the sol–gel route, with adequate structural and chemical properties to support its application as an efficient photocatalyst for cyanide degradation.

Thermodynamic analysis: pre-degradation of total cyanide

As mentioned previously, cyanide exhibits a strong tendency to form coordination complexes with transition metals, which are stable under neutral and alkaline conditions but may undergo dissociation in acidic media. In weakly acidic solutions, partial destabilization can occur, whereas in strongly acidic environments, cyanide ligands are protonated to HCN. In addition, displacement of cyanide may occur in the presence of competing ligands with higher binding affinities, as reflected in thermodynamic stability constants. Due to the toxicity and biological risks associated with free cyanide and metal–cyanide complexes, the total cyanide concentration in effluents discharged into the environment is strictly regulated worldwide.

As a case study, the composition of residual solutions from silver post-leaching processes in the Mexican mining industry was determined by atomic absorption spectroscopy. The results, summarized in Table S1, show the concentrations of metallic ions present in the residual solutions. According to Table S1, elemental chemical analysis revealed that the residual concentration of free cyanide far exceeds the permissible limits for human consumption, since the World Health Organization (WHO) has established a health-based guideline of 0.5 ppm for drinking water and short-term exposure (World Health Organization [WHO], 2017). Furthermore, the presence of metals such as Ag and Cu was detected; both are known to form highly toxic cyanide complexes that stabilize cyanide ions in solution. In addition, the detection of heavy metals such as lead poses a significant risk due to its bioaccumulation in the human body. The WHO has set a maximum limit of 0.01 ppm for lead in drinking water, and the measured values in these residual solutions exceed this threshold (WHO, 2017).

Based on these results (Table S1), the initial conditions for the thermodynamic analysis and the concentrations of the synthetic solutions in a complex water matrix were established, as reported in Table S2. Table S2 summarizes the concentrations of metallic species in the synthetic solutions designed to represent effluents from the silver post-leaching process. For these calculations, complete dissociation of the chemical precursors was assumed, thereby defining the ionic concentrations of the metallic species. Moreover, given that the concentration of cyanide was considerably higher than that of the counter-ions, their effect on the displacement of metal–cyanide coordination equilibria was considered negligible.

In Fig. 1, the Pourbaix diagrams for silver, copper, iron, lead, zinc, and cyanide are presented. From the Ag(I) predominance diagram (Fig. 1a), the stable solid species are Ag_2_O_3_(cr), AgO(s), AgCN(s), and crystalline elemental silver. For the Cu(II), Fe(II), and Zn(II) components (refer to Fig. 1), Cu(cr)/CuO(cr)/CuCN(s), Fe_2_O_3_(cr), and ZnO(cr) are the predominant solids.Fig. 1. Pourbaix diagrams used for the pre-degradation thermodynamic analysis of the chemical species in solution: a AgNO₃ ([Ag⁺] = 2.66 ppm), b CuSO₄·5H₂O ([Cu^2^⁺] = 12.00 ppm), c FeSO₄·7H₂O ([Fe^2^⁺] = 1.99 ppm), d Pb(NO₃)₂ ([Pb^2^⁺] = 2.05 ppm), e ZnSO₄·7H₂O ([Zn.^2^⁺] = 15.99 ppm), and f NaCN ([CN⁻] = 100 ppm)

Under the physicochemical conditions prior to the degradation stage with 100 ppm CN⁻, pH 10.5, and an experimental redox potential close to zero, some solid species and metal hydroxide coordination compounds of lead and zinc predominate in the solution (Fig. 1d, e).

Likewise, some metal–cyanide complexes are the thermodynamically stable species under these conditions, with the exception of Pb and Zn, which do not readily form stable cyanide complexes. Specifically, for Ag(I), Cu(I), and Fe(II), the predominant cyanide complexes are [Ag(CN)₂]⁻, [Cu(CN)₂]⁻/[Cu(CN)₃]^2^⁻/[Cu(CN)₄]^3^⁻ and [Fe(CN)₆]^4^⁻, respectively. In contrast, although Zn can form the stable tetracyanido complex [Zn(CN)₄]^2^⁻, under strongly alkaline conditions hydroxide complexation often prevails; lead cyanide complexes are comparatively weak, and Pb is more likely to occur as hydroxide/solid phases rather than as cyanide complexes. So, the Pb(OH)2(s), and ZnO(cr) solid phases or Pb(OH)4/Pb(OH)4^2─^, and Zn(OH)3^─^/Zn(OH)4^2─^ hydroxide complexes are stable at alkaline pH.

The chemical speciation diagram of total cyanide indicates that hydrogen cyanide (HCN) predominates at acidic pH and negative potential. While free cyanide (CN−) is stable at pH > 9 and Eh < 0.25 V. In contrast, within a limited range of oxidizing potentials, the formation of (CN)2, CuCN(s), and AgCN(s) is favored.

At this stage, thermodynamic analysis serves as a valuable tool for determining the chemical stability of the species present. In this context, the predominant species were characterized as a starting point to identify the potential metal species complexed with cyanide and their associated risk of dissociation and release of free cyanide if the residual solutions remain untreated.

Catalytic evaluation and reaction kinetics of the free cyanide and total cyanide

Ozonation process

The results of the ozonation process for free and total cyanide are shown in Figure S1. As observed in Figure S1a, ozonation of free cyanide alone achieves extensive removal within ~55 min. During ozonation, the free cyanide concentration remains approximately constant between 25 and 40 min. At this stage, the probability of effective O₃–CN⁻ encounters decreases, rendering direct ozonation progressively less probable. Consequently, competitive pathways, most notably base-catalyzed ozone self-decomposition and reactions with matrix constituents—increasingly account for ozone consumption. Hydroxide (OH⁻) accelerates O₃ decay; however, the resulting decomposition generates reactive oxygen species (e.g., •OH, O₂•⁻) that continue to oxidize residual cyanide. Consistent with this behavior, a kinetic transition is observed after ~40 min: the reaction rate becomes approximately independent of [CN⁻], indicating a shift from predominantly molecular-ozone control at early times to radical-mediated pathways at later times.

By contrast, ozonation of total cyanide—i.e., free cyanide in the presence of dissolved metal species (Figure S1b)—achieves an efficiency comparable to that observed for free cyanide alone, with reaction times of the same order. Notably, a transient induction period (during which [CN⁻] remains constant) is again observed but over a shorter window: for free cyanide, it spans ~25–40 min, whereas in the presence of dissolved metals, it arises earlier, at ~ 20–30 min. This shift suggests that aqueous metal species may, to some extent, promote free-cyanide degradation—as metal–cyanide complexes progressively dissociate—by catalyzing ozone decomposition into reactive oxygen species and/or facilitating ligand-exchange pathways that increase the effective oxidant flux. At the same time, certain metal ions can act as ozone sinks; however, under the conditions tested, their net effect appears to shorten the induction period and advance the onset of the decay phase rather than diminish overall ozonation efficiency.

The experimental data were fitted with either a pseudo–first-order or a zero-order kinetic model, depending on the dominant reaction pathway during ozonation of free and total cyanide (Figure S1). The free cyanide ozonation kinetics exhibited a multi-regime behavior (Figure S1a). Initially (0–25 min), the decay followed first-order kinetics with \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${k}_{1}=0.0006 {\mathrm{min}}^{-1}$$\end{document} . Subsequently, the free-cyanide concentration reached an apparent plateau at ∼27 ppm. Beyond ∼40 min, the process transitioned to zero-order kinetics with \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${k}_{0}=1.818 {\text{mg L}}^{-1}$$\end{document} . In contrast, for total-cyanide ozonation (free cyanide in the presence of dissolved metals), the pseudo–first-order rate constant was \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${k}_{1}=0.095 {\mathrm{min}}^{-1}$$\end{document} . The kinetic analysis indicates that, despite a modest difference in characteristic times (~5 min), dissolved metal cations originating from metal–cyanide complexes may catalyze cyanide degradation and/or promote ozone decomposition, thereby generating additional oxidizing species that sustain the reaction.

Photocatalytic ozonation process

The experimental results of both free and total cyanide degradation are shown in Fig. 2. During the first 30 min, the reaction was conducted in the dark to establish adsorption equilibrium. The degradation system corresponds to a hybrid process of heterogeneous photocatalytic ozonation, employing BiVO_4_ synthesized by the sol–gel method as the photocatalyst. As noted above, the physicochemical properties of the obtained catalyst were investigated in depth. Accordingly, the discussion of its characterization is provided in Ref. (Vargas-Rueda et al. 2024). The degradation of only free cyanide achieves 100% after 14 min of treatment. Also, the analysis of the water matrix containing aqueous metallic species and cyanide revealed high efficiency in the degradation of total cyanide ([CN^─^]TOTAL). Monitoring the cyanide levels demonstrated that total cyanide degradation was achieved within a 20 min period. During the first 8 min of reaction, a significant degradation percentage, close to 80%, was experienced. The changes in slope observed within the first 8 min may be attributed to the formation of a new intermediate or to the stepwise conversion of metal–cyanide complexes after 8 min. As cyanide concentration declines, the photocatalytic ozonation rate decreases due to substrate depletion, and the emergence of new oxidizable intermediates, distinct from the original cyanide species, as coordinated compounds less stable with cyanide compared to the initial complexes.Fig. 2% Degradation and oxidation–reduction potential (ORP) profiles for free cyanide ([CN⁻]) and total cyanide ([CN⁻]TOTAL) during photocatalytic ozonation under UV/Vis irradiation. Initial cyanide concentration in both cases was 100 ppm at pH 10.5. The ozone gas flow rate was maintained constant at 1 L·h⁻^1^. For the total-cyanide condition, the aqueous matrix contained dissolved metal species Ag, Cu, Fe, Pb, and Zn

It should be noted that there is a marked difference in activity between the degradation of free cyanide and that of total cyanide by the photocatalytic ozonation process, which contains metal–cyanide complexes, indicating that the overall degradation behavior is strongly influenced by the chemical nature of these species.

It is worthwhile to mention that free cyanide undergoes rapid degradation either through molecular ozone or through reactive oxidizing species generated in the system. In contrast, the presence of metal–cyanide complexes increases the multi-component interactions within the system, as these species may act as interfering agents. Even so, certain metallic cations can promote ozone decomposition and catalyze the formation of reactive radicals. However, the reaction efficiency ultimately depends on the thermodynamic stability of the cyano-metallic complexes and their dissociation capacity to release free cyanide, which limits the overall degradation efficiency.

Additionally, the presence of competing species (e.g., hydroxide ions) can interfere with ozone availability. As shown in Supplementary Figure S2, the impact of hydroxide ions was demonstrated: in the solution containing only free cyanide, a progressive acidification was observed as the reaction proceeded, indicating that hydroxide ions were consumed by ozone molecules. Consequently, the system exhibits dynamic interactions in which both metallic species and hydroxide ions may act as catalytic promoters of ozone decomposition or as inhibitors, depending on chemical species.

This dual behavior explains the observed differences in kinetic constants between free and total cyanide ozonation and underscores the advantages of photocatalytic ozonation over ozonation alone, particularly in matrices rich in metal–cyanide complexes, as it enhances the formation of reactive species capable of overcoming these thermodynamic barriers. Thus, the combined system—visible-light irradiation, a BiVO₄ photocatalyst, and ozone—outperforms each treatment applied separately under identical operating conditions. It achieves higher conversion of inorganic pollutants in shorter times and improves efficiency of ozone utilization. Mechanistically, photogenerated electrons in BiVO₄ are scavenged by ozone, which suppresses e⁻/h⁺ recombination and enhances the in-situ formation of reactive oxygen species (ROS; e.g., •OH, O₂•⁻), thereby increasing the effective oxidant flux. Besides, in the presence of dissolved metals, oxidation of free cyanide shifts coordination equilibria and promotes the stepwise dissociation of metal–cyanide complexes, releasing CN⁻ and enabling more efficient oxidation pathways.

Additionally, the monitoring of the oxidation–reduction potential (ORP) as a function of degradation time for free cyanide and total cyanide is also shown in Fig. 2. The concentration of photogenerated holes (h⁺), hydroxyl radicals, and superoxide radicals significantly increased the ORP in both systems. Moreover, the presence of Ag, Cu, Fe, Pb, and Zn species with cyanide, along with their oxidation/reduction capacities, contributed to a slight rise in ORP, with the maximum potential recorded at 554.2 mV vs Ag/AgCl in the degradation of total cyanide.

An analysis of the multivariate reaction rate of cyanide degradation is proposed. Figure S3 presents the reaction kinetics of both free cyanide and total cyanide. For the degradation of free cyanide, as shown in Figure S3(a), the rate constant was 6.512 mg⋅L⁻^1^ min⁻^1^, best described by a zero-order kinetic model. In contrast, the first-order kinetics evaluation did not show a strong linear correlation.

For the degradation of total cyanide, as shown in Figure S3(b), the reaction follows first-order kinetics (k₁ = 0.167 min⁻^1^), contingent upon the total cyanide concentration. However, after 10 min, the reaction appears to transition to zero-order kinetics, where the concentration of the organic pollutant is not a factor. This change in reaction order is notable due to the total cyanide concentration at 10 min being 21.43 ppm, decreasing more slowly during the degradation reaction in the complex water matrix. The reaction kinetics change order as the cyanide complexes are completely degraded. This suggests that the cyanide complexes initially engage in reaction, contributing to an increase in ORP and reaction kinetics. Consequently, the ORP tends to stabilize as the cyanide concentration decreases, which is also attributed to the constant presence of solubilized ozone in the solution.

With the objective of deepening the understanding of the reaction mechanism, a kinetic analysis was conducted to determine whether the degradation of total cyanide was limited by the photocatalyst surface or by the adsorption process. In this approach, it was assumed that the kinetic model based on the Langmuir–Hinshelwood mechanism accounts for the influence of solubilized ozone, in which aqueous ozone reacts directly without requiring a significant adsorption step on the surface of bismuth vanadate. This led to the consideration of the following mathematical expression (Smith 1991):

\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$-\frac{dC_{{CN}^-}}{dt}=\frac{k'\bullet K_{ads}\bullet C_{{CN}^-}{\bullet C}_{O_3}}{\left(1+K_{ads}\bullet C_{{CN}^-}+C_{O_3}\right)^2}$$\end{document}

Where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${C}_{{O}_{3}}={k}_{{O}_{3}}\left[{O}_{3}\right]$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${k}^{\prime}={k}_{heterogeneous}^{\acute{i} ntrinsic}\bullet {C}_{{O}_{3}}$$\end{document} . Specifically, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${k}^{\prime}$$\end{document} is an apparent kinetic constant that includes the interaction of ozone in solution, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${k}_{{O}_{3}}$$\end{document} is the ozone solubility coefficient, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\left[{O}_{3}\right]$$\end{document} is the ozone concentration in the gas phase, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${k}_{heterogeneous}^{\acute{i} ntrinsic}$$\end{document} is the intrinsic rate constant of the heterogeneous reaction. A constant value of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${C}_{{O}_{3}}$$\end{document} in the aqueous solution implies that the system operates under controlled conditions, where ozone solubility does not limit the reaction.

From the fitting of experimental results, as shown in Figure S4, the modified Langmuir–Hinshelwood kinetic model estimated an adsorption rate constant ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${K}_{ads})$$\end{document} equal to \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$0.052 {\mathrm{ppm}}^{-1}$$\end{document} , and an intrinsic heterogeneous rate constant ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${k}_{heterogeneous}^{\acute{i} ntrinsic}$$\end{document} ) of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$3.82 {\mathrm{min}}^{-1}$$\end{document} . Since bismuth vanadate exhibits a low adsorption capacity, a low \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${K}_{ads}$$\end{document} value was expected, which is consistent with the calculated result, indicating that the photocatalyst has a low affinity for the chemical species in solution. As a consequence, a low \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${K}_{ads}$$\end{document} value suggests that the catalyst surface does not become easily saturated, potentially allowing a continuous transfer of the pollutant from the bulk solution to the photocatalyst surface. Despite this limitation, if the catalyst exhibits a high intrinsic rate constant for the heterogeneous reaction, the system may compensate for the low adsorption rate. In such a case, cyanide conversion can still reach significant levels, as the surface reaction occurs rapidly, favoring pollutant degradation before the adsorption limitation critically affects the overall reaction kinetics.

Thermodynamic analysis: post-degradation of total cyanide

Based on the ORP at the end of the post-degradation process of the complex water matrix, fraction diagrams at a final potential of 751.2 mV vs SHE (ORP Measurement = 554.2 mV vs Ag/AgCl) were constructed. The post-degradation chemical speciation diagrams for Ag(I), Cu(II), Fe(II), Pb(II), Zn(II), and [CN⁻] are shown in Fig. 3. For Ag(I), the cationic form Ag⁺ predominates at a pH below 11, while the chemical species such as Ag(OH)(aq) and Ag_2_O_3_(cr) become predominant near pH 11. Regarding Pb(II), pre-degradation thermodynamic predictions indicate the formation of solid Pb(OH)₂. However, under the established oxidizing conditions, Pb(OH)₂ is solubilized, suggesting that the Pb(OH)₄ coordinated compound is now stable at pH levels above 10. For Cu(II), Fe(II), and Zn(II) at pH 10.5, the predominant species are the metal oxides Cu_2_O(s), CuO(cr), Fe₂O₃(cr), and ZnO(cr), respectively. However, in the case of Zn(II), approximately 10% of the species corresponds to the aqueous complex Zn(OH)2. In the case of total cyanide, the main degradation byproduct is cyanate (OCN⁻), which predominates at pH > 6 regardless of the initial cyanide fraction.Fig. 3. Post-degradation thermodynamic analysis at an ORP equal to 0.75 V showing the predominance zone diagrams of: a AgNO₃ ([Ag⁺] = 2.66 ppm), b CuSO₄·5H₂O ([Cu^2^⁺] = 12.00 ppm), c FeSO₄·7H₂O ([Fe^2^⁺] = 1.99 ppm), d Pb(NO₃)₂ ([Pb^2^⁺] = 2.05 ppm), e ZnSO₄·7H₂O ([Zn^2^⁺] = 15.99 ppm), and f NaCN ([CN⁻] = 100 ppm)

Mass balance and metal quantification to validate thermodynamic predictions

Tables 1 and 2 show the mass balance for the pre-degradation processes, the concentration of dissolved metallic ionic species analyzed by absorption spectroscopy after the degradation process, and the mass balance for the post-degradation stage. The initial concentrations of Ag(I), Cu(II), Fe(II), Pb(II), Zn(II), and CN⁻ aqueous species promoted the formation of cyanide-metal complexes and solid species during the pre-degradation stage. The formation of these complex species in synthetic solutions is analogous to the residual solutions from silver leaching, in which the pre-degradation chemical equilibrium established that the moderately strong complex cyanides such as [Ag(CN)2] ^─^ and [Cu(CN)3] ^2─^, as well as strong complex cyanides such as [Fe(CN)6] ^4─^ are formed. Table 1. Initial metal speciation of the synthetic silver post-leaching solution prior to photocatalytic ozonation at pH 10.5 and room temperature, calculated under thermodynamic equilibrium conditions. The table includes the predominant aqueous and solid metal species and their corresponding stability constants (log K). Initial physicochemical conditions: [CN⁻] = 2041 µM, [Ag⁺] = 15.67 µM, [Cu^2^⁺] = 48.06 µM, [Fe^2^⁺] = 7.19 µM, [Pb^2^⁺] = 6.20 µM, and [Zn^2^⁺] = 55.64 µMChemical speciesInitial concentration (µM)log K[Ag(CN)₂]⁻10.220.38[Ag(CN)OH]⁻5.47 −0.56[Cu(CN)₃]^2^⁻48.0632.03[Fe(CN)₆]^4^⁻5.4745.61Fe(aq), non-speciated1.72—Pb(OH)₂(s)6.19 −8.15ZnO(cr)55.64 −11.20Table 2Metal distribution and mass balance of the synthetic silver post-leaching solution after photocatalytic ozonation at pH 10.5 and room temperature, determined by atomic absorption spectroscopy and mass balance analysis. The table reports the remaining dissolved metallic species and the corresponding solid phases formed after cyanide degradationMetal****Dissolved species (µM) †Solid speciesSolid concentration (µM) ††log KspAg5.83AgO(s)9.84 −29.95Cu3.42CuO(cr)44.64 −7.68Fe0Fe₂O₃(cr)7.19 −26.45Pb1.78Pb(OH)₂(s)4.41 −53.13Zn33.96ZnO(cr)21.68 −11.20†The concentration of dissolved metallic ionic species after the degradation process was analyzed by absorption spectroscopy††The concentration of solid species, considered the predominant form in the post-degradation stage, was determined by subtracting the concentration of dissolved metallic ionic species measured after degradation from the initial concentration calculated based on chemical equilibrium

Also, thermodynamic predictions in the pre-degradation stage indicated that some solid compounds in the form of hydroxides or oxides initially predominated in the solution (see Table 1). In the post-degradation stage, based on species distribution diagrams and under oxidizing conditions, the chemical equilibrium favors the breakdown of both free cyanide and various metal–cyanide complexes, leading primarily to the formation of solid metal species from those initially present in solution. However, the quantification of metals after the degradation process in a water matrix containing cyanide and metal–cyanide complexes revealed a significant concentration of dissolved metallic species, including Ag, Cu, Pb, and Zn, as shown in Table 2, which presents the concentration of aqueous metallic species measured by atomic absorption spectroscopy in the post-degradation stage of the post-leaching silver synthetic solution.

An analysis based on thermodynamic prediction indicated that the formation of zinc-cyanide complexes in solution is limited by the concentrations of cyanide/zinc reactants and its low log K value, which may lead to the formation of weak complex cyanides. However, based on thermodynamic analysis, chemical equilibrium established that under the concentrations studied, zinc-cyanide complexes did not form. Instead, the predominant species is ZnO(cr) in the pre-degradation stage. Under the oxidizing conditions analyzed in the chemical stability diagrams after the degradation process, the predominant crystalline solid remains ZnO(cr). However, the quantification of Zn in the post-degradation system revealed a higher concentration of Zn. Thus, it can be inferred that part of ZnO(cr) is dissolved, and zinc(II) hydroxide complexes such as [Zn(OH)3] ^─^ and [Zn(OH)4] ^2─^ could be formed at pH 10.5.

For Ag(I) and Cu(II), the dissociation of these moderately strong complex cyanides will result in the release of Ag^+^ ionic species and formation of CuO crystalline, respectively. It was expected that the entire Ag^+^ concentration would correspond to its ionic form. However, at pH 10.5, the predominance zone of Ag^+^ is close to that of AgO(cr). Therefore, it is possible that part of the Ag^+^ forms AgO(cr). Regarding Cu(II), the amount of copper measured in solution is low compared to the concentration of CuO(cr). Thus, the residual copper in solution could be bound to hydroxide species.

Even though iron-cyanide compounds are strong complexes, the complete dissociation of [Fe(CN)6] ^4─^ into Fe_2_O_3_(cr) is achieved. In this case, the thermodynamic prediction aligns closely with the experimental results, as the quantified iron concentration in solution was effectively zero.

Also, the treatment of tailing wastewater containing heavy metals such as lead (Pb) presents a significant challenge in environmental remediation. Although lead cyanide compounds are considered weak complexes ([Pb(CN)4] ^2─^, Log K = 10.3), the presence of dissolved Pb in solution poses a serious problem for water bodies. The World Health Organization recommends a strict limit of 0.01 mg L⁻^1^ (10 µg L⁻^1^) for lead in drinking water (WHO, 2017). While a portion of Pb forms Pb(OH)2(s), the residual Pb could be bound to hydroxyl groups, forming Pb(OH)4(aq), indicating a potential risk to human health.

Post-degradation solution characterization

[Cu(CN)3]2─ complex evolution as an indicator species

In Fig. 4, UV–Vis spectra of solutions containing free cyanide and metal–cyanide complexes at different reaction times were recorded to explore the evolution of α-[Cu(CN)3]^2─^ complexes during both the ozonation process and the photocatalytic ozonation process. This evolution is attributed to electron transitions in α- [Cu(CN)3]^2─^ complexes, particularly ligand-to-metal charge transfer (LMCT). In the UV–Vis spectrum of copper–cyanide complexes, characteristic transitions have been reported at 197.5 nm (HCN), 208.6 nm (CN⁻), 214 nm ([Cu(CN)₄]^3^⁻), 224.2 nm (α-[Cu(CN)₃]^2^⁻), 235.3 nm (β-[Cu(CN)₃]^2^⁻), and 242 nm ([Cu(CN)₂]⁻) (Noblitt 1973). Additionally, the high initial concentration of copper in synthetic [CN^─^]Total solutions allowed the study of the dissociation of copper-cyanide complexes. Copper was selected as the indicator species because the iron concentration was much lower and its UV–Vis signal under our conditions was negligible. Moreover, Fe(III) species display a characteristic absorption band at ~420 nm (Murton et al. 2025), clearly distinct from the spectra of copper–cyanide complexes. Furthermore, thermodynamic speciation at the operating conditions indicates that Zn and Pb do not form (or do not predominate as) stable cyanide complexes, instead favoring hydroxide/oxide species. In addition, the Supplementary Information (Fig. S6) provides the standard curve used for the detection of cyanide–copper complexes.Fig. 4UV–Vis spectra of [CN^─^]Total solution, explore the evolution of [Cu(CN)3]^2─^ complexes during degradation process by a reaction involving only ozone and b photocatalytic ozonation reaction

Prior to the reaction involving only ozone (seen Fig. 4a), two peaks were identified at 203 nm and 238 nm, corresponding to the absorption of the [CN]^─^ and β-[Cu(CN)3]^2─^ complex. After 10 and 20 min of reaction, a new peak emerged at 223 nm, attributed to the absorption of the [Cu(CN)2]^─^ complex. These results suggest that the [Cu(CN)3]^2─^ complex partially dissociates into [Cu(CN)2]^─^ within the first 10 min, releasing additional free cyanide ions. As the reaction progresses, these complexes dissociate completely, and a broad absorption peak appears around 260 nm, along with a sharp absorption peak at 213 nm, both attributed to aqueous ozone (Levanov et al. 2016).

In the photocatalytic ozonation process (seen in Fig. 4b), before the reaction commenced, two peaks were observed at 223 nm and 238 nm, corresponding to the α- [Cu(CN)3]^2─^ and [Cu(CN)2]^─^ complexes, respectively. The interaction between the [Cu(CN)3]^2─^ complex and other species such as (Ag, Fe, Pb, and Zn)-cyanide complexes with the BiVO_4_ photocatalyst during the dark reaction established a different chemical equilibrium compared to the reaction with ozone alone. Specifically, a change in the concentration of the [Cu(CN)3]^2─^ complex was observed, with an initial concentration of 12.5 ppm, which decreased to 4.34 ppm after the dark reaction. This indicates that some metal–cyanide complexes were adsorbed on the catalyst surface, and that a minor proportion of these complexes dissociated.

Once the reaction started, the concentration of the [Cu(CN)3]^2─^ complex decreased significantly to 0.1 ppm within 3 min, with both [Cu(CN)3]^2─^ and [Cu(CN)2]^─^ complexes still present at that time. After 6 min, the copper–cyanide complexes were completely dissociated. As the reaction continued, a broad absorption peak around 260 nm was identified, which was assigned to the absorption of ozone. Additionally, two contiguous absorption peaks at 203 nm and 213 nm were observed, both also attributed to aqueous ozone (Levanov et al. 2016). This suggests that other species, likely reactive ozone derivatives, are contributing to the growth of these peaks.

Ammonia quantification: Nessler's reagent method

The photocatalytic ozonation process for both free cyanide ([CN^─^]) and total cyanide ([CN^─^]TOTAL) can lead to the formation of basic inorganic molecules as degradation byproducts. Although thermodynamic predictions establish that the cyanate ion (CNO^─^) is the predominant species after the degradation process, the complete mineralization of free cyanide into carbon dioxide (CO₂), water (H₂O), or nitrogen (N₂) can be expected. Therefore, the quantification of ammonia (NH₃) in the aqueous medium as a possible degradation byproduct was carried out using the Nessler’s reagent method. The overall reaction for the conversion of cyanide to ammonia can be represented as follows:

\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${CN}^{-}+\text{Reactive Species}\to {CO}_{3}^{ 2-}+{NH}_{3}+{H}^{+}+\mathrm{byproducts}$$\end{document}

The ammonia concentrations determined for both free cyanide and total cyanide after the ozonation and photocatalytic ozonation processes are shown in Figure S7, respectively. Additionally, the standard curve for the detection of ammonia is provided in supplementary information (Figure S7). For the ozonation process, aqueous ammonia was not detected in either the free cyanide or total cyanide systems. In contrast, during the photocatalytic ozonation process, the highest concentration of ammonia was recorded in the system containing only free cyanide, with a value of 86.80 ppm. Meanwhile, the ammonia concentration measured in the total cyanide system was 22.70 ppm.

The median lethal dose (LD_50_ oral for rats) for free cyanide is between 2 and 5 mg kg^−1^. Conversely, the median lethal dose (LD_50_ oral for rats) for ammonia is 350 mg/kg, and for cyanate, it ranges between 160 mg/kg and 320 mg/kg. Although both ammonia and cyanate are less toxic than free cyanide, the ammonia concentration remains high. According to the Mexican Official Standard NOM-127-SSA1-2021, “Water for Human Use and Consumption,” the permissible limit for ammoniacal nitrogen (N-NH_3_) is 0.5 ppm, indicating that the residual solutions will require post-treatment before wastewater disposal.

These results demonstrate that, under photocatalytic ozonation, free cyanide is efficiently transformed into ammonia, whereas the presence of dissolved metal ions ([CN^─^]TOTAL) significantly modifies nitrogen transformation pathways, leading to a markedly lower accumulation of ammonia in the total cyanide system. From a thermodynamic perspective, the further oxidation of ammonium to nitrate is favored under strong oxidative conditions. This behavior is consistent with the nitrogen speciation diagram shown in Figure S8, which describes the fractional distribution of nitrogen-containing species as a function of pH at E_H_ = 0.75 V and 25 °C—a redox potential that matches the experimental conditions applied in this study. However, despite being thermodynamically feasible, the oxidation of ammonium to nitrate is kinetically less favorable than cyanide oxidation and therefore does not dominate within the investigated reaction timeframe. Consequently, ammonia remains a detectable nitrogen-containing product, particularly in the free cyanide system.

Re-cycle test: stability and elemental chemical characterization

The stability of bismuth vanadate in the presence of total cyanide ([CN^─^]Total) was evaluated through consecutive photocatalytic ozonation reactions. The photocatalyst was recovered and reused in four subsequent reactions. As shown in Fig. 5a, the photocatalytic ozonation activity decreased after the third cycling, indicating deactivation of BiVO_4_. During the first two cycles, the reaction efficiency remained constant, achieving 100% degradation of total cyanide. However, in the third run, efficiency decreased by 20%, and in the fourth run, it further reduced to 40% of the initial efficiency. This decline is linked to the recycling and collection process of BiVO_4_ catalysts, which becomes less efficient with each subsequent reaction. Consequently, the concentration of BiVO_4_ photocatalyst decreases over time, reducing the available active sites and surface adsorption capacity. Also, the photocatalyst concentration is not the only parameter that limits the stability of BiVO_4_.Fig. 5a Cycling runs for oxidation of total cyanide. b Elemental chemical analysis by EDS of the BiVO_4_ photocatalyst in the presence of both free cyanide and metal–cyanide complexes after the first cycle and the fourth recycling

An analysis of the elemental chemical composition by EDX was conducted for the photocatalyst in both the first run and the last run of the BiVO_4_ recycle test. As shown in Fig. 5b, the results revealed the presence of metals such as silver, iron, copper, and zinc in the first cycle. Although lead was not detected initially, this does not imply that Pb is absent on the catalyst surface. In the fourth cycling, lead was detected in addition to the previously mentioned metals. It was also observed that the concentration of metals increased with each reuse of the photocatalyst, indicating that more metals were adsorbed onto the surface as the photocatalyst was reused. The highest concentrations of metallic traces were copper and zinc in both the first and fourth runs. Additionally, sodium was quantified, originating from the 0.1 M sodium hydroxide used to adjust the pH of the solution to 10.5, and its concentration remained constant throughout the process.

Likewise, the total concentrations of dissolved silver, copper, iron, lead, and zinc species in the synthetic post-leaching silver solution were quantified by atomic absorption spectroscopy at the end of each cycle of the process. Thermodynamically, BiVO₄ is regarded as essentially insoluble in aqueous media and remains stable over a wide electrochemical potential window under near-neutral conditions (pH 4–11), as predicted by the calculated Pourbaix diagram. Due to its high thermodynamic stability and the strong Bi–O and V–O bonds, metal dissolution under the experimental conditions employed is expected to be negligible. Although metal leaching from BiVO₄ has been reported under photoelectrochemical water oxidation conditions, such behavior is typically associated with anodic polarization and does not necessarily indicate catalyst degradation under mild photocatalytic environments. In the present study, the redox potential was limited to approximately ≈550 mV vs Ag/AgCl (≈0.75 V vs. SHE), a regime in which BiVO₄ is expected to remain structurally and chemically stable (Toma et al. 2016; Zhang et al. 2020).

As observed in Table 3, the concentrations of the metals are highest in the first and second reaction cycles compared to the third and fourth. Even so, the concentration of zinc remains higher in each cycle compared to the other metals. The presence of residual Pb in the solutions after each consecutive photocatalytic ozonation reaction exceeds the maximum allowed by the Mexican Official Standard. In contrast, iron was not detected in any of the cycles. An opposite trend was observed in the quantification of % weight elemental composition on the surface of the photocatalyst after subsequent reactions. As the BiVO_4_ catalyst is reused, the concentration of metal species in the aqueous solution decreases after the degradation process, leading to increased precipitation of metallic species on the surface of the photocatalyst. It can also be inferred that all iron present precipitates in some form of metal oxide according to thermodynamic prediction. In the case of lead, the precipitation of lead (II) hydroxide is desired rather than having lead remain dissolved in the solution after the degradation process. To confirm the above, an elemental mapping was carried out. Table 3. Concentration of metallic species analyzed by absorption spectroscopy after the degradation process in each reuse of the photocatalystCycleConcentration [ppm]AgCuFePbZn 1 st run0.9910.85400.599.7662nd run0.9181.09900.887.8133rd run0.5960.732005.8604th run0.210.97700.591.953

Elemental chemical mapping was performed to show the spatial distribution of metallic elements such as silver, copper, iron, and zinc in the BiVO₄ photocatalyst during both the first and fourth recycling. The weight percentage of lead was so low that it was not detected in the analysis. This indicates that the average point analysis by EDX showed the weight percentage of lead was below the detection limit and therefore not identified in the analysis. As shown in Figure S9(a), the electron image of the studied area after the first reaction cycle shows a uniform distribution of bismuth, vanadium, and oxygen. However, larger crystals corresponding to zinc around the bismuth vanadate were identified. Based on the oxygen distribution in the sample, these crystals are consistent with thermodynamic predictions after degradation, indicating the formation of zinc oxide. The precipitation of metallic elements such as silver, copper, iron, and zinc on the photocatalyst surface was also confirmed. For the fourth recycling of BiVO₄, as seen in Figure S9(b), a higher agglomeration of particles is observed. These particles correspond not only to bismuth, vanadium, and oxygen, but also to silver, copper, iron, and zinc. The above results indicate that as consecutive photocatalytic ozonation reactions proceed, solids precipitation on the catalyst surface progressively increases, as supported by the previously discussed elemental chemical analysis.

Although the degradation of total cyanide with BiVO_4_ produced effective results in the first and second recycle reactions, the photocatalyst may tend to agglomerate, decreasing the effective surface area and promoting the recombination of electrons and holes, which results in a reduction in catalytic efficiency (Li et al. 2010). Furthermore, it can be assumed that the photocatalyst becomes increasingly poisoned, progressively inhibiting the adsorption of free cyanide and metal–cyanide complexes on the active sites and thereby losing its photocatalytic activity.

Cyanide degradation reaction mechanism

The schematic illustration of the total cyanide degradation mechanism via photocatalytic ozonation is shown in Fig. 6. The valence band and conduction band potentials for bismuth vanadate were calculated as +2.16 V and −0.32 V, respectively, based on the valence band spectrum, as shown in Supplementary Figure S10. The complexity of chemical reactions in a multicomponent system containing not only free cyanide but also metallic ions results in a synergistic interaction between metallic reactive species, the photocatalyst, and the ozone in solution. The role of oxidative chemical species was studied previously (refer to Reference (Vargas-Rueda et al. 2024)). Free cyanide degradation was tested with scavengers: IPA, p-BQ, and EDTA at 1 mM. Degradation decreased to 82%, 64%, and 45%, respectively, showing that O₂•⁻ radicals and h⁺ are the main contributors, while OH• plays a minor role. However, the degradation process of a complex water matrix containing free cyanide, metal–cyanide complexes, and metal-oxyhydroxide species involves not only photogenerated holes or superoxide radicals, but also additional reactive oxygen species that contribute to the overall degradation of total cyanide. It is worth noting that ozone is a more effective electron scavenger than oxygen due to its higher electrophilicity compared to O_2_ towards electrons generated on the catalyst surface (Hernández-Alonso et al. 2002). This limits the recombination of charge carriers on the surface of bismuth vanadate, resulting in faster kinetics for cyanide degradation and lower ozone consumption compared to the ozone-only process. Therefore, a series of chemical reactions is proposed to describe the photocatalytic ozonation degradation of total cyanide, as outlined below.Fig. 6. Reaction possible mechanism in the degradation of total cyanide in presence of photoactivated BiVO_4_ and ozone

So, in a photocatalytic ozonation reaction, BiVO_4_ particles are excited by a radiation source to produce positive holes in the valence band ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${h}^{+}$$\end{document} ) and negative electrons at the conduction band ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${e}^{-}$$\end{document} ):

\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$BiV{O}_{4}+h\nu \to {e}^{-}+{h}^{+}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${e}^{-}+{h}^{+}\to \mathrm{recombination}$$\end{document}

Photogenerated holes and electrons can further form hydroxyl radicals from species such as OH^−^, H_2_O, and O_2_ (Nosaka & Nosaka 2016).

\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$h^++{OH}_{(surface)}^-\rightarrow{OH}^\bullet$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$h^++H_2O_{(surface)}\rightarrow{OH}^\bullet+H^+$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$e^-+{O_2}_{(surface)}\rightarrow O_2^{\bullet-}$$\end{document}

Additionally, it is suggested that the photogenerated holes can directly promote the oxidation of cyanide, as shown in Equation (21), or by hydroxyl radicals (Eq. 22).

\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${CN}^{-}+2{h}^{+}+{H}_{2}O\to {CNO}^{-}+2{H}^{+}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${CN}^-+2{HO}^\bullet\rightarrow{CNO}^-+H_2O$$\end{document}

Depending on the solution physiochemical conditions, ozone can react directly with cyanide, or it can produce intermediates byproducts such as \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$O_2^{\bullet-}/HO_2^\bullet/OH^\bullet/H_2\;O_2/O_3^{\bullet-}$$\end{document}   (Achisa & Chollom 2020) that may subsequently react with cyanide, as shown in Eqs. 23–28.

The oxidation of cyanide to isocyanate by direct reaction with superoxide radicals (•O₂⁻) is thermodynamically and kinetically plausible under acidic conditions, where a high concentration of hydronium ions favors proton-coupled pathways, as represented in Eq. (26). Under alkaline conditions, superoxide remains predominantly in its anionic radical form (•O₂⁻), and its direct oxidative capability toward cyanide is significantly reduced. Instead, superoxide primarily participates in coupled reaction pathways that promote the generation of more strongly oxidizing species, such as hydroxyl radicals (•OH) or other secondary reactive oxygen species, and facilitates indirect oxidation routes under highly oxidative environments. Consequently, the synergistic interplay among photogenerated holes, hydroxyl radicals, ozone-derived reactive species, and superoxide radicals collectively drives the oxidation of cyanide into isocyanate.

Also, at pH alkaline, the reaction rate of ozone decomposition in water is enhanced due to the catalytic effect of hydroxide ions. Consequently, at higher pH values, the decay of ozone occurs more rapidly (Yershov et al. 2009).

\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$3{CN}^{-}+{O}_{3} \to {3CNO}^{-}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${O_3}_{(surface)}+e^-\rightarrow{\bullet O}_3^-$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$O_3+{OH}^-\rightarrow O_2^{\bullet-}+{HO}_2^\bullet$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${CN}^-+{\bullet O}_2^-+H^+\rightarrow{CNO}^-+{HO}^\bullet$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${HO}_2^\bullet+{CN}^-\rightarrow{CNO}^-+{OH}^-$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${CN}^-+{2HO}_3^\bullet\rightarrow{CNO}^-+H_2O+{2O}_2$$\end{document}

Given that the degradation process involves the presence of a photocatalyst activated by UV/Vis radiation, the photolysis of ozone could be initiated (as shown in Eq. 29), leading to the photodecomposition of ozone and the formation of hydrogen peroxide (Masjidin et al. 2024).

\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${O}_{3}+{H}_{2}O+h\nu \to {H}_{2}{O}_{2}+{O}_{2}$$\end{document}

Besides, thermodynamic predictions established that iron(II), silver(I), and copper(II) cyanide complexes are formed. Prior to free cyanide degradation, these metal–cyanide complexes can decompose into intermediate byproducts as follows.

For Iron (aq):

\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\left[Fe{(CN)}_{6}\right]}^{4-}+{h}^{+}\to {\left[Fe{(CN)}_{6}\right]}^{3-}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\left[Fe{(CN)}_{6}\right]}^{4-}+{O}_{3}+{H}_{2}O+{e}^{-}\to {\left[Fe{(CN)}_{6}\right]}^{3-}+{O}_{2}+{2OH}^{-}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\left[Fe{(CN)}_6\right]^{4-}+O_2^{\bullet-}+H_2O\rightarrow\left[Fe{(CN)}_6\right]^{3-}+O_2+2{OH}^-$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\left[Fe{(CN)}_6\right]^{4-}+{OH}^\bullet\rightarrow\left[Fe{(CN)}_6\right]^{3-}+{OH}^-$$\end{document}

The oxidation state change of iron (II) to iron (III) can be facilitated by oxidizing agents that form in the solution, such as photogenerated holes and superoxide radicals. Tied to the above, the formation of mixed-ligand cyanides, as shown in Eq. 34, can be considered in the formation of iron aquapentacyano ferrate (III) ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\left[{Fe}^{III}{(CN)}_{5}{H}_{2}O\right]}^{2-}$$\end{document} ) and iron hydroxopentacyano ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\left[{Fe}^{III}{(CN)}_{5}OH\right]}^{3-}$$\end{document} ), leading to the subsequent release of cyanides ions (Van Grieken et al. 2005).

\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\left[Fe{\left(CN\right)}_{6}\right]}^{m-}+L \leftrightarrow {\left[Fe{\left(CN\right)}_{5}L\right]}^{1-m}+{CN}^{-}$$\end{document}

Thus, the degradation of iron-cyanide complexes by photocatalytic ozonation can imply the breakage of these complexes, forming less stable iron-cyanide complexes. As the reaction progresses, the redox potential of the solution increases towards positive values, directly oxidizing the cyanide and further breaking the complexes (as shown in Eqs. 35–37).

\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\left[{Fe}^{II}{(CN)}_{5}{H}_{2}O\right]}^{2-}+2{h}^{+}\to {\left[{Fe}^{III}{(CN)}_{4}\right]}^{-}+{CNO}^{-}+2{H}^{+}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\left[{Fe}^{II}{(CN)}_{5}{H}_{2}O\right]}^{2-}+{O}_{3}+{4e}^{-}+{H}^{+}\to {\left[{Fe}^{III}{(CN)}_{4}\right]}^{-}+{CNO}^{-}+3O{H}^{-}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\left[{Fe}^{II}{(CN)}_5H_2O\right]^{2-}+{OH}^\bullet+{4O}_2\rightarrow\left[{Fe}^{III}{(CN)}_4\right]^-+{CNO}^-+3{HO}_3^\bullet$$\end{document}

In the same way, more cyanide ions were released and subsequently oxidized, as follows.

\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\left[Fe{(CN)}_{4}\right]}^{-}+{8h}^{+}+{4H}_{2}O\to {Fe}^{3+}+4{CNO}^{-}+8{H}^{+}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\left[Fe{(CN)}_{4}\right]}^{-}+{O}_{3}\to {Fe}^{3+}+4{CNO}^{-}+{O}_{2}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\left[Fe{(CN)}_4\right]^-+{2OH}\bullet\rightarrow{Fe}^{3+}+4{CNO}^-+H_2O$$\end{document}

For silver in aqueous solution, the decomposition of silver-cyanide complexes is described by Eqs. 41–43.

\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\left[Ag{(CN)}_{2}\right]}^{-}+{4h}^{+}+{2OH}^{-}\to {Ag}^{+}+2{CNO}^{-}+{2H}^{+}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\left[Ag{(CN)}_{2}\right]}^{-}+{2O}_{3}\to {Ag}^{+}+2{CNO}^{-}+{2O}_{2}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\left[Ag{(CN)}_2\right]^-+{OH}^\bullet+{2O}_2\rightarrow{Ag}^++2{CNO}^-+{HO}_3^\bullet$$\end{document}

Regarding copper (aq), a similar mechanism has been proposed (as shown in Eqs. 44–45) based on the UV–Vis spectrum of [CN^─^]Total solution, as was shown in Fig. 4. Although Cu(II) is more stable than Cu(I), the copper (I)-cyanide complexes predominate in solution due to their chemical equilibrium (Log K).

\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${{Cu}^{2+}+4{CN}^{-}\to \left[{Cu}^{II}{(CN)}_{4}\right]}^{2-}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${{Cu}^{+}+3{CN}^{-}\to \left[{Cu}^{I}{(CN)}_{3}\right]}^{2-}$$\end{document}

So, the direct oxidation of cyanide and the subsequent breakdown of the complexes take place, as follows.

\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\left[{Cu}^{I}{(CN)}_{3}\right]}^{2-}+{2h}^{+}+{H}_{2}O\to {\left[{Cu}^{I}{(CN)}_{2}\right]}^{-}+{CNO}^{-}+2{H}^{+}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\left[{Cu}^{I}{(CN)}_{3}\right]}^{2-}+{O}_{3}\to {\left[{Cu}^{I}{(CN)}_{2}\right]}^{-}+{CNO}^{-}+{O}_{2}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\left[{Cu}^I{(CN)}_3\right]^{2-}+{OH}^\bullet+O_2\rightarrow\left[{Cu}^I{(CN)}_2\right]^-+{CNO}^-+{HO}_2^\bullet$$\end{document}

As the chemical reaction progresses, the \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\left[{Cu}^{I}{(CN)}_{2}\right]}^{-}$$\end{document} specie is decomposed, as shown in Eqs. 49–51.

\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\left[{Cu}^{I}{(CN)}_{2}\right]}^{-}+{2h}^{+}+{H}_{2}O\to {Cu}^{+}+2{CNO}^{-}+2{H}^{+}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\left[{Cu}^{I}{(CN)}_{2}\right]}^{-}+{O}_{3}\to {Cu}^{+}+2{CNO}^{-}+{O}_{2}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\left[{Cu}^I{(CN)}_2\right]^-+{OH}^\bullet+O_2\rightarrow{Cu}^++2{CNO}^-+{HO}_2^\bullet$$\end{document}

As metal–cyanide complexes decompose, metallic ionic species and free cyanide are released. Eventually, the free cyanide ions are degraded by various reactive species. Additionally, the presence of these metal ions raises the possibility that the kinetics of ozone decomposition could be affected. An increase in the concentration of metallic ionic species can accelerate ozone decomposition. B. G. Ershov et al. demonstrated that increasing concentrations of silver (I) and copper (II) enhance the kinetics of ozone decomposition, indicating that the presence of transition metal cations can catalyze these reactions (Ershov et al. 2012).

Although, it is possible that high concentrations of metal ions act as inhibitors by forming secondary radicals that do not generate additional oxidizing species, thereby terminating the chain reaction and inhibiting ozone decay. In other words, the oxidation of inorganic compounds through ozonation is favored because these compounds can react much faster with ozone than organic compounds (Gottschalk et al. 2009). Nevertheless, the presence of metallic ions can also serve as catalysts, enhancing the overall efficiency of ozonation. C. H. Ni et al. studied various metallic ions at low concentrations ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\sim$$\end{document} 1 ppm), where iron (II) demonstrated the highest catalytic efficiency compared to zinc (II), copper (II), and lead (II) (Ni et al. 2003). This suggests that the rate of ozone decay depends not only on the concentration of metallic ions but also on the specific type of metal ions involved. Initially, as metal ions are released due to the decomposition of metal–cyanide complexes, the degradation process is favored by ozone decay. However, as the concentration of metallic ionic species increases, this catalytic effect becomes limited.

Cyanate was not directly quantified in this study. However, ammonia was experimentally detected as a degradation byproduct during photocatalytic ozonation, as discussed previously in the “Ammonia quantification: Nessler’s reagent method” section. The significantly higher ammonia formation observed for free cyanide compared to total cyanide systems indicates that ammonia is a major nitrogen-containing product under the applied reaction conditions. Once metal–cyanide complexes dissociate and free cyanide is oxidized, it is proposed that a fraction of the resulting cyanate undergoes further transformation to ammonium and carbonates through cyanate hydrolysis and oxidative pathways.

\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${CNO}^{-}+{4h}^{+}+{4OH}^{-}\to {CO}_{3}^{ 2-}+{NH}_{3}+{O}_{2}+{H}^{+}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${CNO}^{-}+{O}_{3}+2{H}_{2}O+2{e}^{-}\to {CO}_{3}^{ 2-}+{NH}_{3}+{O}_{2}+{OH}^{-}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${CNO}^{-}+{2H}_{2}O\to {CO}_{3}^{ 2-}+{NH}_{3}+{H}^{+}$$\end{document}

Furthermore, ozone readily oxidizes dissolved metal ions, yielding oxides or hydroxides that may occur either as soluble hydroxo complexes or as precipitated solid phases. The Supplementary Information (Note S1) compiles the relevant redox reactions between ozone and representative metal species, with supporting literature references, and proposes a mechanistic framework consistent with those reports (Ershov et al. 2012; Gottschalk et al. 2009; Ni et al. 2003; Logager et al. 1992). These considerations support the experimental results, which suggest that metallic species can be oxidized by reactive oxygen species or stabilized as coordination compounds with hydroxyl ions or oxygen, thus remaining in solution as aqueous complexes.

Conclusion

A thermodynamic analysis of the pre- and post-degradation stages of free cyanide and metal–cyanide complexes was conducted using chemical stability diagrams, which provided a theoretical framework to clarify the underlying reaction mechanisms. As a case study, the complete degradation of cyanide in synthetic water matrices containing Ag, Cu, Fe, Pb, and Zn—all species commonly found in effluents from the Mexican mining industry—was evaluated. The degradation efficiency of both free cyanide and total cyanide was investigated under ozonation and photocatalytic ozonation conditions.

The results indicate that complex water matrices containing free cyanide and metal ions exhibit synergistic interactions among various reactive species, enhancing total cyanide degradation. The accelerated kinetics observed during total cyanide degradation are attributed to the high electrophilicity of ozone and the additional contribution of reactive oxygen species. Kinetic analysis confirmed that photocatalytic ozonation significantly increases the reaction rate and reduces ozone consumption compared to ozonation alone.

Among the byproducts of cyanide degradation, ammonia was identified. The highest concentration (86.80 ppm) was observed in the system containing only free cyanide, while the total cyanide system showed a lower concentration of 22.70 ppm. These findings highlight a variation in degradation pathways depending on the chemical nature of the matrix.

The stability of the BiVO₄ photocatalyst was demonstrated during the initial reaction cycles. However, as the number of cycles increased, the progressive accumulation of metallic precipitates on its surface inhibited the adsorption of cyanide species, resulting in a gradual loss of photocatalytic activity. This is supported by the higher concentrations of adsorbed metallic species observed in the first and fourth cycles by EDS analysis, indicating precipitation and progressive surface fouling of the catalyst over time.

Supplementary Information

Below is the link to the electronic supplementary material.ESM 1(DOCX 2.47 MB)