Tailoring of poly[Ni(OH)₂salen] nanoparticle-based electrocatalysts for effective urea remediation

Abstract

There is no universal recipe for the proper structure tuning of Ni(OH)₂ nanoparticle (NP)-based catalysts for efficient urea electrooxidation (UOR) in alkaline media. However, it is known that fast generation of Ni³⁺OOH-type catalytic centers that are sustained and resilient during the overall catalytic process is crucial. Towards this, we report how we optimized and compared operating conditions and structural tuning of poly[NP-Ni(OH)₂SaltMe] and poly[meso-NP-Ni(OH)₂SaldMe] electrocatalysts active in alkaline media towards UOR. We started with studies of morphological differences evoked by the use of different NaOH_aq concentrations for catalyst fabrication by SEM and TEM. Then, we distinguished the most promising molecular structures of fabricated catalysts featuring the highest poisoning resistance and in situ generation of poly(NP-Ni³⁺OOHsalen) electrocatalytic centers for UOR. Furthermore, we found the best conditions for operation of both structured UOR catalysts using a comprehensive electrochemical approach. This approach involved multiple scan rate, Tafel slope, and activation energy (E_ac) analysis to finally compare which structured poly[NP-Ni(OH)₂salen] catalyst produces catalytic current more efficiently in response to a change in applied potential. Ultimately, we performed a longevity/durability test under real-system mimicking conditions. The fabricated catalysts constituted good platforms for studying the surface-remaining and bulk-remaining types of catalytically active sites of poly[NP-Ni(OH)₂salen]s for UOR activity. Our findings point to the bulk-structure-reactivity requirements of poly[NP-Ni(OH)₂salen]s, emphasizing their catalytic durability and effectiveness.

---

Introduction

Nowadays, there is a need to find an effective method for simultaneous wastewater treatment and electricity generation. The proper treatment of urea-containing wastewater is essential from an energy as well as an environmental point of view. Urea comprises ∼45% nitrogen and is a vital crop fertilizer.¹ Part of urea from agricultural fields enters surface water. Another source of urea introduction into the environment is mammalian excretion. The widespread release of non-treated urea-rich wastewater is potentially dangerous for the aquatic system. Moreover, urea pollution triggers sea algae to produce a deadly toxin called domoic acid.² The natural decomposition of urea into nitrogen-based pollutants like toxic ammonia harms air quality and eventually causes health risks.^3,4 Using urea-rich wastewater as an alternative energy source has a double benefit. Hence, it can help with wastewater remediation simultaneously with electricity generation.^1,4,5 However, the catalyst cost, efficiency, and problems with scale-up experiments to industrial scale limit the transfer of laboratory experiments to real application.^3,6

There were many attempts to devise and fabricate heterogeneous electrocatalysts for urea electrooxidation.^1,7,8 Most of them included the involvement of noble metals, such as Pt, Pd, and Rh.^9–13 The high cost, low abundance, and low electrochemical stability limit the commercialization of these noble metal-based catalysts.¹⁴

Nickel (Ni)-based materials operating in alkaline media are considered the best cost-efficient alternative of catalysts active towards efficient urea electrooxidation.^15–22 The most popular ones are Ni(OH)₂-based electrocatalysts, which can oxidize urea in an appropriate alkaline medium concentration to carbon dioxide (in the form of CO₃²⁻), nitrogen, and water with simultaneous energy release.²³ However, these catalysts still suffer from (i) poor accessibility of active centers in the catalyst bulk and thus slow diffusion of urea^24–26 and (ii) strong adsorption of electrooxidation products to active sites.^27,28 Thus, providing high durability is still one of the challenges restricting Ni(OH)₂-based catalysts' large-scale application.

The kinetics of urea oxidation in the absence of a catalyst is slow. On the Ni(OH)₂-based electrocatalysts, urea oxidation becomes thermodynamically favourable and can follow the so-called indirect/catalyst regeneration mechanism.^23,29 The indirect/regeneration mechanism involves the electrochemical oxidation of Ni²⁺(OH)₂ to Ni³⁺OOH, following which Ni³⁺OOH acts as an oxidizer that becomes reduced to Ni²⁺(OH)₂ and oxidizes urea. This regeneration mechanism indicates that urea oxidation proceeds only on OH⁻ and urea-accessible catalyst centers of Ni³⁺OOH electrogenerated from Ni(OH)₂ forms at thermodynamically favourable potentials in both forward and backward CV scans. However, the balance between OH⁻ doping ions responsible for active center production in the first step and urea reactant diffusion to those centers must be preserved during the electrocatalytic process. Moreover, based on the indirect/regeneration mechanism, one can deduce the possible adsorption/accumulation of reaction products as well as trapping of OH⁻ dopant, i.e., the amount of blocked/deactivated Ni³⁺OOH catalytically active sites. The CV-monitored urea oxidation reveals the formation and stabilization of Ni³⁺OOH catalytically active sites. The forward scan process of electrocatalytic urea oxidation influences the catalytic peak current height, which appears at a backward scan in the potential range where urea oxidation still proceeds. Therefore, the catalytic performance of Ni(OH)₂-based electrocatalysts can be followed and adjusted by comparing forward and backward scan electrocatalytic responses.³⁰

The most desired Ni(OH)₂-based catalyst operating in alkaline media requires fast charge transfer, causing the generation of Ni³⁺OOH active centers from Ni(OH)₂ pristine forms.³¹ Hence, the presence and accessibility of these catalytically active Ni³⁺OOH forms after OH⁻ doping, to which urea diffuses as fast as possible, determines its efficiency.²⁴ By increasing accessibility to the bulk structure of Ni(OH)₂-based catalysts for OH⁻ and urea, one can increase the catalyst activity by increasing the amount of accessible active sites.^3,27,28 Furthermore, a proper concentration of OH⁻ counterions forcing the change in the binding affinity of urea oxidation products tends to increase their durability.^{3,23,27,32,33} However, too high concentrations of OH⁻ are not preferred from an ecological point of view and may lead to the chemical decomposition of Ni³⁺OOH active centers,^24,34 according to the Pourbaix diagram. This diagram shows that the Ni(OH)₂-based NPs undergo transformation into HNiO₂ in a too-high pH.³⁵ Reported DFT calculations suggested that urea and OH⁻ when present in the solution compete during diffusion to active centers because of their similar affinity.³³ The OH⁻ concentration has to be sufficiently high to desorb urea oxidation products from the active sites.^3,32 However, a too high concentration of OH⁻ in comparison to the concentration of urea can block urea access to active centers .^33,36–38 Therefore, this concentration needs to be optimized to achieve the best operation conditions for the new type of structured Ni(OH)₂-based catalyst.

The electrochemical studies performed during the catalytic oxidation of urea on the new Ni²⁺-based catalysts operating in alkaline media according to the two-step indirect mechanism are aimed at monitoring the impact of urea diffusional limitations.^23,25–27 The electrocatalytic urea oxidation process occurring in the second step on Ni³⁺OOH centers of properly structured Ni(OH)₂-based catalysts is expected to be under urea oxidation control, i.e., under charge transfer control (kinetically controlled). In that case, the slowest step has to be the chemical oxidation of urea on catalytically active centers, not the diffusion of urea to these centers.^1,24,26

Recently, a few research groups reported the fabrication of Ni(OH)₂-type NP electrocatalysts derived from poly[Ni²⁺(salen)] precursors differing in chemical structure and morphology.^31,39–41 Those Ni(OH)₂-type NP catalysts were studied towards electrooxidation of small molecules, e.g., methanol, ethanol, and glycerol.^31,40 The fabrication method of Ni(OH)₂-type NPs is called potential-driven generation,³¹ and, importantly, it is conducted in a water-based alkaline medium. The uniform distribution of Ni(OH)₂-type NPs in the catalyst bulk was preserved by their 3D-templating inside the poly(salen) matrix. Hence, this advantageous arrangement of NPs was found to be dependent on conditions of poly[Ni(salen)] precursor electrodeposition and its molecular structure.³¹ However, the effect of alkaline medium concentration on NP fabrication by potential-driven generation still needs to be studied. The Ni(OH)₂-type NP catalyst derived from poly[Ni(salen)] precursor electrodeposited under potentiostatic conditions indicated the highest amount of electrocatalytically active centers engaged in ethanol oxidation. Importantly, no active center blocking by ethanol electrooxidation product was observed.³¹ In other words, due to the high accessibility of active centers in the catalyst bulk, fast diffusion of ethanol was achieved. Furthermore, we noticed that the use of Ni(OH)₂-type NPs embedded in poly(salen) matrixes as a catalyst for urea electrooxidation demands additional structuring and adjusting of catalytic conditions.

Experimental section

Chemicals

Anhydrous acetonitrile (ACN, 99.8%), electrochemical grade tetra(n-butyl)ammonium hexafluorophosphate ((TBA)PF₆, ≥99% electrolyte salt), and BioReagent grade urea (NH₂CONH₂, ≥98%) were purchased from Sigma-Aldrich. Analytical grade sodium hydroxide (NaOH) was purchased from POCH. All of these reagents were used without further purification. Carbon paper (CP FuelCellsETc) was purchased from College Station, TX. Two phenyl ring unsubstituted Ni²⁺salen monomers differing in the configuration of methyl-substituted imine bridge, (i) the 2,3-dimethyl-N,N′-bis-(salicylidene)-2,3-butanediaminonickel] (Ni²⁺SaltMe)⁴² and (ii) meso-N,N′-bis(salicylidene)-2,3-butanediaminonickel,⁴³ were synthesized according to an earlier described methodology (Scheme 1).⁴³ Milli-Q water (resistivity 18.2 MΩ cm at 25 °C) was used for measurements in aqueous conditions. Artificial urine containing 0.25 M urea was prepared according to the procedure mentioned by Sarigul et al.⁴⁴

Scheme 1 Structural formulas of (a) meso-Ni²⁺SaldMe and (b) Ni²⁺SaltMe.

Electrochemical cells and apparatus

A BioLogic SP300 potentiostat, controlled by EC-Lab software, coupled with a classical three-electrode glass electrochemical cell, was used for all electrochemical measurements, including poly[Ni(salen)] precursor electrodeposition, potential-driven transformation of these polymer precursors into Ni(OH)₂-type NPs embedded in poly(salen) matrixes, and their electrochemical characterization. A 5 mm diameter (∼0.196 cm² geometrical area) laminated CP electrode,⁴⁵ Pt coil, and Ag/Ag⁺ pseudoreference (used for experiments performed in organic solvent electrolyte solutions), or Ag/AgCl (3 M KCl) electrode (used for experiments performed under aqueous conditions) served as working, counter, and reference electrodes, respectively.

Electrodeposition of Ni(OH)₂-type NP catalyst precursors

The precursor films were deposited by electropolymerization under potentiostatic conditions on a laminated CP electrode from an ACN solution containing 1 mM of Ni²⁺salen monomer (Ni²⁺SaltMe or meso-Ni²⁺SaldMe) and 0.1 M (TBA)PF₆. A constant potential of 1.30 V vs. Ag/Ag⁺ was kept until 130 mC cm⁻² polymerization charge passed. Consequently, the resulting precursor films were abbreviated as poly(NiSaltMe)₁₃₀ and poly(meso-NiSaldMe)₁₃₀.

The conditions of precursor film electrodeposition were adapted based on our previous work, where we observed that poly(meso-NiSaldMe) precursor that was potentiostatically electrodeposited by passing 130 mC cm⁻² was the best performing towards ethanol electrooxidation after its transformation into Ni(OH)₂-type NPs.³¹

Optimization of NaOH_aq concentration for the potential-driven generation of Ni(OH)₂-type NPs

Poly(NiSaltMe)₁₃₀ and poly(meso-NiSaldMe)₁₃₀ precursor films were transferred separately to 0.2, 0.5, 1.0, and 2.0 M NaOH_aq to fabricate eight different catalysts. These precursor layers were abbreviated as poly[NP-Ni(OH)₂SaltMe]_{0.2 M}, poly[NP-Ni(OH)₂SaltMe]_{0.5 M}, poly[NP-Ni(OH)₂SaltMe]_{1 M}, poly[NP-Ni(OH)₂SaltMe]_{2 M}], poly[meso-NP-Ni(OH)₂SaldMe]_{0.2 M}, poly[meso-NP-Ni(OH)₂SaldMe]_{0.5 M}, poly[meso-NP-Ni(OH)₂SaldMe]_{1 M}, and poly[meso-NP-Ni(OH)₂SaldMe]_{2 M}. The potential-driven precursor transformation into the Ni(OH)₂-type NPs was conducted in the potential range of 0.00 to 1.20 V vs. Ag/AgCl (3 M KCl) at 20 mV s⁻¹ till the Ni²⁺/Ni³⁺ oxidation and respective reduction current peaks were no longer growing. After NP preparation, electrodes were rinsed with Milli-Q water.

Next, the electrocatalytic responses of fabricated NP catalysts were measured using the same NaOH_aq concentration as applied for their potential-driven generation. The appropriate urea volume was added separately or continuously to the electrochemical cell to obtain the following urea concentrations: 0.01, 0.05, 0.1, 0.3, and 0.5 M, and to determine the in situ generation ability of active sites and the loss of concentration linearity range.

Morphological studies of the Ni(OH)₂-type NPs

Poly[NP-Ni(OH)₂salen]s fabricated under different NaOH_aq concentrations were imaged by scanning electron microscopy (SEM) and transmission electron microscopy (TEM). SEM imaging was performed with an FEI Nova NanoSEM 450 microscope. TEM imaging was carried out using a Talos F200X microscope with Super X energy-dispersive X-ray and an EDX detector. NPs embedded in poly(salen) matrixes were scratched from the CP electrode and placed on a copper grid for TEM imaging.

Optimization of NaOH_aq concentration during electrocatalytic oxidation of urea

Poly[NP-Ni(OH)₂SaltMe]_{1 M} and poly[meso-NP-Ni(OH)₂SaldMe]_{1 M}, generated in 1.0 M NaOH_aq, were further examined toward electrooxidation of 0.3 M urea in 0.2, 0.5, 1.0, and 2.0 M NaOH_aq. The electrocatalytic measurements were carried out in the potential range of 0.00 to 1.00 V vs. Ag/AgCl (3 M KCl) at 50 mV s⁻¹. After each catalytic measurement, catalysts were cycled in pure electrolyte solution containing the same concentration of NaOH_aq that was used during urea electrooxidation.

Electrocatalytic performance comparison

Multiple scan rate experiments were performed for poly[NP-Ni(OH)₂SaltMe]_{1 M} and poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} (i) in 1.0 and 2.0 M NaOH_aq (bare electrolyte) and then (ii) in 1.0 M NaOH_aq containing 0.3 M urea. These CV responses were registered at the scan rate of 2, 5, 10, 20, 50, 100, and 200 mV s⁻¹. Furthermore, the electron transfer coefficient α and diffusion coefficient D were calculated based on the slope values determined from the peak potential vs. the logarithm of the scan rate⁴⁶ and the current vs. the square root of the scan rate dependencies.

α was calculated according to eqn (1):

where n₀ is the number of electrons exchanged at the interface of the CP electrode|NP-Ni(OH)₂salen (Scheme 2). The slope was determined from the linear part of the curve E_pa = slope (log v) + b, where b is the Ni²⁺/Ni³⁺ oxidation rate constant.

Scheme 2 Illustration of the two-stage catalytic process of urea electrooxidation at the poly[NP-Ni(OH)₂salen] catalyst (U: urea, P – product).

D was calculated according to eqn (2):⁴⁷

where n is the number of electrons involved in urea oxidation (n = 6) at the NP-Ni³⁺OOH(salen)|solution interface, A is a geometric area of the working electrode (0.196 cm²), and C is the added urea concentration (0.3 M). The diffusion coefficient of urea in 1.0 M NaOH_aq was determined from the linear part of the curve I_pa = slope v^1/2.

Linear sweep voltammetry (LSV) curves were measured at 5 mV s⁻¹ at the potential range of 0.0 to 1.0 V vs. Ag/AgCl (3 M KCl) in a solution containing 0.3 M urea in 1.0 M NaOH_aq for poly[NP-Ni(OH)₂SaltMe]_{1 M} and poly[meso-NP-Ni(OH)₂SaldMe]_{1 M}. These conditions ensured kinetic control of urea electrooxidation, i.e., the slowest process was urea electrooxidation after its diffusion to accessible catalyst active centers.⁴⁸ For these conditions, Tafel slopes for both catalysts were determined from the potential vs. the logarithm of the current density, i.e., E vs. log(I) plots.

The activation energy (E_ac) values were determined for 0.3 M urea electrooxidation in 1.0 M NaOH_aq on poly[NP-Ni(OH)₂SaltMe]_{1 M} and poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} at 50 mV s⁻¹ in the potential range of 0.0 to 1.0 V vs. Ag/AgCl (3 M KCl), measured separately at different temperatures of 16, 19, 24, 28, 33, 38, and 43 °C. The temperature was controlled by two thermocouples placed inside and outside the cell. The E_ac values were determined for potentials corresponding to the urea electrooxidation: 0.45 and 0.5 V from the Arrhenius plots. Those plots displayed the current density's logarithm against the temperature's reciprocal, i.e., log(I) vs. 1/T. In the absence of urea, the control experiment was prepared under the same conditions.

Durability test of electrocatalyst in real-system mimicking conditions

To mimic the actual operating conditions of urea fuel cell (UFC) and artificial urine fuel cell (AUFC), a constant potential of 0.55 V vs. Ag/AgCl (3 M KCl) was applied to poly[NP-Ni(OH)₂SaltMe]_{1 M} and poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} catalysts for 17 h in 1.0 M NaOH_aq in the absence and presence of 0.3 M urea or artificial urine, and the resulting current was recorded.

Results and discussion

Optimization of NaOH_aq concentration for potential-driven transformation of poly[Ni(salen)]s

Two phenyl ring unsubstituted Ni²⁺(salen) complexes, namely Ni²⁺SaltMe⁴⁹ and meso-Ni²⁺SaldMe,⁴³ differing in methyl substituent configuration on the imine bridge, were deposited on laminated CP electrodes⁴⁵ under potentiostatic conditions as poly(Ni²⁺SaltMe)₁₃₀ (Fig. S1a†) and poly(meso-Ni²⁺SaldMe)₁₃₀ (Fig. S1b†). In this moderately electron-donating medium, Ni²⁺(salen) monomers electropolymerize in an oxidative manner, forming a conducting film at the electrode surface.⁴³ These electrodeposited poly[(Ni²⁺(salen)]s, during the conduction process in moderately electron-donating electrolyte solution, behave like a polyphenylene compound, with the Ni²⁺ acting as a bridge between biphenylene moieties.⁵⁰ In such a system, the electron is released from the salen ligand parts.^49–51 However, during the potential-driven transformation of poly[Ni²⁺(salen)]s into Ni(OH)₂-type NPs conducted in a strongly electron-donating medium, e.g., NaOH_aq, the place of electron release is changed from the salen ligand center to the Ni²⁺.³¹ The formation of Ni(OH)₂-type NPs is manifested by a consistent growth of new oxidation and reduction peaks.

During the potential-driven generation process, the poly[Ni²⁺(salen)] precursors are gradually transformed into Ni(OH)₂-type NPs embedded into a non-conducting poly(salen) matrix (Fig. S2 and S3†). The axial coordination of OH⁻ to the Ni²⁺ centers is crucial for potential-induced cleavage of the bonds between nitrogen belonging to imine moieties and Ni²⁺, thus resulting in Ni(OH)₂-type NP formation inside the poly(salen) polymer matrix. Therefore, one might suppose that an increased NaOH_aq concentration would result in faster and more efficient NP generation. However, when we performed NP generation using 0.2, 0.5, 1.0, and 2.0 M NaOH_aq by keeping the same scan rate and potential range for poly(NiSaltMe)₁₃₀ and poly(meso-NiSaldMe)₁₃₀ precursors deposited in the same manner, we did not observe the time shortening of complete NP generation (Table S1†). However, the highest amount of poly(Ni³⁺OOHsalen) active centers were generated from both precursors in 1.0 M NaOH_aq (Fig. 1a and b, blue curves), indicated by the highest current density. Furthermore, at higher NaOH_aq concentrations, the poly[Ni²⁺(OH)₂salen]/poly(Ni³⁺OOHsalen) electrooxidation and poly(Ni³⁺OOHsalen)/poly[Ni²⁺(OH)₂salen] electroreduction occurred at lower potential values (Fig. 1a and b). We anticipate that the differences observed in the last CV cycle patterns of the potential-driven transformation of (a) poly(NiSaltMe)₁₃₀ and (b) poly(meso-NiSaldMe)₁₃₀ into Ni(OH)₂-type NPs, performed in different NaOH_aq concentrations are related to potential-induced cleavage of the bonds between nitrogen belonging to imine moieties and Ni²⁺ occurring in bulk and on the surface for poly(meso-NiSaldMe)₁₃₀ and poly(NiSaltMe)₁₃₀, respectively.

Fig. 1 The last cycle of the potential-driven transformation of (a) poly(NiSaltMe)₁₃₀ and (b) poly(meso-NiSaldMe)₁₃₀ into the Ni(OH)₂-type NPs, performed at different NaOH_aq concentrations.

Morphological characterization of Ni(OH)₂-type nanoparticles fabricated at different NaOH_aq concentrations

Morphological differences of poly(SaltMe)₁₃₀ and poly(meso-SaldMe)₁₃₀ matrixes embedding Ni(OH)₂-type NPs evoked by the use of different NaOH_aq concentrations in their fabrication process were studied by SEM. This imaging indicated that matrixes of Ni(OH)₂-type NPs derived from poly(Ni²⁺SaltMe)₁₃₀ and poly(meso-Ni²⁺SaldMe)₁₃₀ were differently spatially arranged (Fig. 2). Previously, we observed that the morphology of these poly[Ni²⁺(salen)] precursors governed by their monomer polymerization conditions and chemical structure influenced the arrangement of Ni(OH)₂-type NPs in poly(salen) matrixes when NP-based catalyst fabrication was conducted in 0.2 M NaOH_aq.³¹ Now we have observed that regardless of the NaOH_aq concentration used for fabrication, catalysts derived from potentiostatically deposited poly(Ni²⁺SaltMe)₁₃₀ featured the planar and more condensed arrangement of matrix templating Ni(OH)₂-type NPs (Fig. 2a–d). However, the poly(meso-SaldMe) matrix arrangement was much more spatially expanded (Fig. 2e–h).

Fig. 2 SEM images of Ni(OH)₂-type NPs derived from poly(NiSaldMe)₁₃₀ (a–d) and poly(meso-NiSaldMe)₁₃₀ (e–h) fabricated in (a and e) 0.2 M NaOH_aq, (b and f) 0.5 M NaOH_aq, (c and g) 1.0 M NaOH_aq, and (d and h) 2.0 M NaOH_aq.

As a result of catalyst fabrication from poly(Ni²⁺SaltMe)₁₃₀ in the low concentrations of NaOH_aq, carbon fibers were coated with 2D planar poly(SaltMe) matrix templating Ni(OH)₂-type NPs (Fig. 2a and b). In high concentrations of NaOH_aq (1.0 and 2.0 M), the flake-like 2D poly(SaltMe) matrix covered carbon fibers entirely (Fig. 2c and d). In contrast, the poly(meso-SaldMe) matrix resulting from catalyst fabrication in low concentrations of NaOH_aq was spatially diversified in 3D and loosely packed (Fig. 2e and f). Fabrication using higher concentrations of NaOH_aq caused more uniform coverage of carbon fibers by the poly(meso-SaldMe) matrix (Fig. 2g and h). The spacing between neighboring poly(meso-SaldMe) matrix parts was significantly bigger when fabrication was conducted in 1.0 M NaOH_aq (Fig. 2g).

TEM (Fig. 3 and 4) imaging was used to confirm the presence of nanoscale objects of Ni(OH)₂-type NPs embedded in poly(SaltMe) and poly(meso-SaldMe) matrixes. The presence of Ni(OH)₂-type NPs was additionally confirmed by EDX element mapping (Fig. 3 and 4). Nickel (Ni) and carbon (C) were mapped in red and blue, respectively. The differences in Ni(OH)₂-type NP distribution and separation originated from different spatial arrangements of poly(SaltMe) and poly(meso-SaldMe) matrixes as indicated by SEM.

Fig. 3 TEM (a–d), HAADF images (a′–d′) and corresponding EDX mapping (a′′–d′′) of Ni(OH)₂-type NPs derived from poly(NiSaltMe)-PS₁₃₀ in various NaOH_aq concentrations. (a) 0.2 M, (b) 0.5 M, (c) 1.0 M, and (d) 2.0 M. EDX profiles for Ni (red) and carbon (blue).

Fig. 4 TEM (a–d), HAADF images (a′–d′) and corresponding EDX mapping (a′′–d′′) of Ni(OH)₂-type NPs derived from poly(meso-NiSaldMe)-PS₁₃₀ in various NaOH_aq concentrations. (a) 0.2 M, (b) 0.5 M, (c) 1.0 M, and (d) 2.0 M. EDX profiles for Ni (red) and carbon (blue).

The size of NPs in the poly(SaltMe) matrix was below 5 nm (Fig. 3a, a′ and a′′). However, a false impression of NPs aggregates was visible when two matrix fragments overlapped, as visible on the TEM images (Fig. 3a, a′ and a′′). Sizes of NPs generated in 0.5 M (Fig. 3b, b′ and b′′) and 1.0 M NaOH_aq (Fig. 3c, c′ and c′′) were slightly smaller than NPs generated in 0.2 M NaOH_aq (Fig. 3a, a′ and a′′), while the size of NPs was substantially lower (below 2 nm) in the catalyst generated in 2.0 M NaOH_aq (Fig. 3d, d′ and d′′). This suggested that NPs generated from poly(NiSaltMe) at such high NaOH_aq were chemically not stable.

Opposite to this, Ni(OH)₂-type NPs generated from poly(meso-Ni²⁺SaldMe)₁₃₀, in different NaOH_aq solutions had different sizes. NPs generated from a lower concentration of NaOH_aq (0.2 and 0.5 M NaOH_aq) were slightly bigger in size than NPs generated in a higher concentration of NaOH_aq (Fig. 4). More importantly, the size of NPs generated from poly(meso-Ni²⁺SaldMe), in different concentrations of NaOH_aq, were higher than the size of NPs generated from poly(Ni²⁺SaltMe). Even at 2.0 M NaOH_aq concentration, NPs were visible (Fig. 4). It appears that NP embedment in poly(salen) matrixes allowed stabilizing Ni(OH)₂ subnanoclusters down to the nanometer.

The preliminary understanding of the electrocatalytic performance of Ni(OH)₂-type NPs fabricated under different alkaline medium concentrations towards urea oxidation

The electrocatalytic responses of poly[NP-Ni(OH)₂SaltMe]_{0.2 M}, poly[NP-Ni(OH)₂SaltMe]_{0.5 M}, poly[NP-Ni(OH)₂SaltMe]_{1 M}, poly[NP-Ni(OH)₂SaltMe]_{2 M}, poly[meso-NP-Ni(OH)₂SaldMe]_{0.2 M}, poly[meso-NP-Ni(OH)₂SaldMe]_{0.5 M}, poly[meso-NP-Ni(OH)₂SaldMe]_{1 M}, and poly[meso-NP-Ni(OH)₂SaldMe]_{2 M} were registered in the solutions containing the same concentration of NaOH_aq as that used for their potential-driven generation. To these solutions, urea portions were added separately (Fig. 5a–d and 6a–d) or continuously (Fig. S4a–d and S5a–d†) to obtain final urea concentrations of 0.01, 0.05, 0.1, 0.3, and 0.5 M. However, during continuous addition, urea is consumed during electrooxidation, resulting in a final concentration that is not as intended. Moreover, continuous urea addition leads to the accumulation of electrooxidation products in the system. In contrast, there is no continuous accumulation of reaction products when urea portions are added separately. Hence, the final concentration added is as intended. The comparison of results obtained after these two different urea addition ways can indicate which of the studied catalysts is more resistant to the presence of oxidation products in the electrocatalytic system and poisoning.

Fig. 5 The electrocatalytic CV responses of Ni(OH)₂-type NPs fabricated at (a) 0.2 M, (b) 0.5 M, (c) 1.0 M and (d) 2.0 M NaOH_aq from poly(NiSaltMe)₁₃₀ towards the electrooxidation of 0.01, 0.05, 0.1, 0.2, and 0.5 M urea. Urea portions were added separately to the cell to reach the final concentration values. The electrocatalytic CV responses were performed at 50 mV s⁻¹ in the solution containing (a) 0.2 M, (b) 0.5 M, (c) 1.0 M, and (d) 2.0 M NaOH_aq.

Fig. 6 The electrocatalytic CV responses of Ni(OH)₂-type NPs fabricated at (a) 0.2 M, (b) 0.5 M, (c) 1.0 M and (d) 2.0 M NaOH_aq from poly(meso-NiSaldMe)₁₃₀ towards the electrooxidation of 0.01, 0.05, 0.1, 0.2, and 0.5 M urea. Urea portions were added separately to the cell to reach the final concentration values. The electrocatalytic CV responses were performed at 50 mV s⁻¹ in the solution containing (a) 0.2 M, (b) 0.5 M, (c) 1.0 M, and (d) 2.0 M NaOH_aq.

The preparation of the above experiments aimed at a preliminary understanding of the electrocatalytic mechanism of urea electrooxidation in alkaline media on the Ni(OH)₂-type NPs embedded in poly(SaltMe) and poly(meso-SaldMe) matrixes fabricated in different concentrations of NaOH_aq. This step is crucial for further identifying the best fabrication conditions of poly[NP-Ni(OH)₂salen] catalysts. Above all, it directs the further optimization of the best conditions of urea electrooxidation on our structured catalysts.

In the absence of urea, all of the Ni(OH)₂-type NP-based catalysts fabricated in different concentrations of NaOH_aq exhibited a typical Ni²⁺/Ni^3+ 31,40 electrooxidation peak of poly[Ni²⁺(OH)₂salen]/poly(Ni³⁺OOHsalen) in the forward scan and its respective poly(Ni³⁺OOHsalen)/poly[Ni²⁺(OH)₂salen] reduction peak in the backward scan (Fig. 5, 6, S4 and S5,† black curves). Furthermore, at higher NaOH_aq concentrations, the electrooxidation processes of Ni²⁺/Ni³⁺ electrooxidation and Ni³⁺/Ni²⁺ electroreduction occurred at lower potentials (Fig. 5, 6, S4, and S5,† black curves). At potentials exceeding ∼0.75 V, the forward anodic current increased further because of the oxygen evolution reaction (OER).⁵²

After urea additions, all of the studied poly[NP-Ni(OH)₂salen] catalysts followed the so-called indirect mechanism^28,29,38 of urea oxidation. This mechanism is indirect because urea oxidation occurs only at accessible and catalytically active poly(NP-Ni³⁺OOHsalen) centers. Gradually electrogenerated poly(NP-Ni³⁺OOHsalen) centers mediate the chemical oxidation of urea to CO₂ (in the form CO₃²⁻) and N₂ ^3,38 (eqn (4)) with simultaneous reduction of poly(NP-Ni³⁺OOHsalen) centers to poly[NP-Ni²⁺(OH)₂salen] in between forward and backward scan, defined as ‘catalyst regeneration’. What is important is that this ‘catalyst regeneration’ to poly[NP-Ni²⁺(OH)₂salen] can decrease the amount of poly(NP-Ni³⁺OOHsalen) catalytically active sites during the overall process.

Analyzing electrocatalytic CV responses registered for separately added urea (Fig. 5 and 6), it is visible that the current increases in the forward scan until the number of poly(NP-Ni³⁺OOHsalen) accessible active sites is not reduced to poly[NP-Ni²⁺(OH)₂salen] by chemical oxidation of urea. This results in a current density decrease after the peak when products are formed. The second peak of urea oxidation on accessible poly(NP-Ni³⁺OOHsalen) centers increases in the backward scan at the potentials where urea can be electrooxidized. The higher the backward peak, the better the electrocatalytic performance due to the greater durability and availability of poly(NP-Ni³⁺OOHsalen) centers.

In situ generation/retention ability of poly(NP-Ni³⁺OOHsalen) centers

According to the identified indirect mechanism of urea electrooxidation on poly[NP-Ni(OH)₂salen], the amount of charge corresponding to urea electrooxidation on accessible poly(NP-Ni³⁺OOHsalen) centers in the forward scan (Q₁) is equal to the integrated area under the I(t) curve. Meanwhile, Q₂ integrated from the I(t) in the backward scan is the amount of charge associated with urea oxidation on poly(NP-Ni³⁺OOHsalen) centers that remain accessible after a chemical reaction. Knowing the values of Q₁ and Q₂, we calculated the percentage of active centers that were still accessible for urea electrooxidation in the backward scan. In this way, the in situ generation abilities of catalytically active centers for differently structured poly[NP-Ni(OH)₂salen] catalysts in the presence of different concentrations of urea added either separately or continuously were compared.

We did not observe common systematic relations in the generation of active centers on poly[NP-Ni(OH)₂SaltMe]_{0.2 M} and poly[meso-NP-Ni(OH)₂SaldMe]_{0.2 M} in 0.2 M NaOH_aq during continuous or separate addition of increased concentrations of urea (Table S2†). In the presence of 0.3 M urea in 0.2 M NaOH_aq, poly[meso-NP-Ni(OH)₂SaldMe]_{0.2 M} catalyst, in situ generation ability was higher in comparison to other concentrations of urea (Table S2†). The % Q_s and % Q_c were 84.7% and 69.9% for separately and continuously added urea, respectively. We noticed much lower in situ generation abilities (45–55%) for poly[NP-Ni(OH)₂SaltMe]_{0.2 M} operating in 0.2 M NaOH_aq for all added urea. In 0.5 M NaOH_aq with the increasing additions of urea for poly[NP-Ni(OH)₂SaltMe]_{0.5 M} and poly[meso-NP-Ni(OH)₂SaldMe]_{0.5 M}, the in situ generation ability increased regardless of whether urea was added separately or continuously (Fig. 5b, 6b, S4b, and S5b and Table S3†). Meanwhile, for separately added urea portions exceeding 0.3 M, added to 1.0 M NaOH_aq, for poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} and poly[NP-Ni(OH)₂SaltMe]_{1 M}, in situ generation abilities were 83.9–84.6% and 80.3–81.7%, respectively (Table S4†). The higher in situ generation ability observed for poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} indicated that this catalyst retained more active sites during the backward scan. In situ generation abilities for continuous urea additions after reaching 0.3 M were 83.5% and 75.1% for poly[meso-NP-Ni(OH)₂SaldMe]_{1 M}, and poly[NP-Ni(OH)₂SaltMe]_{1 M}, respectively. This showed a significant effect of the electrooxidation product accumulation in the poly[NP-Ni(OH)₂SaltMe]_{1 M} catalyst.

The smaller differences in Q% values obtained for poly[meso-NP-Ni(OH)₂SaldMe]_{1 M}, regardless of whether urea was added separately or continuously indicated that this catalyst is resistant to the presence of catalytic reaction products. Consequently, oxidation products are more effectively desorbing from the poly(meso-NP-Ni³⁺OOHSaldMe)_{1 M} active sites than from poly(NP-Ni³⁺OOHSaltMe)_{1 M} centers.

Forward scan electrocatalytic peak currents of poly[NP-Ni(OH)₂SaltMe]_{2 M} and poly[meso-NP-Ni(OH)₂SaldMe]_{2 M} were shifting towards higher potential values in 2.0 M NaOH_aq (Fig. 5d, 6d, S4d and S5d†). This behavior can be attributed to poly(NP-Ni³⁺OOHsalen) active center blocking by OH⁻ ions when considering separate urea additions (Fig. 5d and 6d). Under these conditions, OH⁻ ions and urea molecules compete to access active centers. However, when urea was added continuously, the potential shift in 2 M NaOH_aq was more prominent (Fig. S4d and S5d†). Thus, this points to gradual active center blocking by the accumulation of reaction products that was most probably coupled with the influence of too-high concentrations of OH⁻ ions. Additionally, it suggests electrocatalyst chemical stability issues in 2.0 M NaOH_aq. This can be explained with the use of the potential/pH diagram. At 2.0 M NaOH_aq, the Ni(OH)₂-type NPs start to be chemically unstable and probably convert to HNiO₂⁻ according to eqn (5):³⁵

Ni(OH)₂ + OH⁻ ⇌ HNiO₂⁻ + H₂O (5)

Moreover, for poly[NP-Ni(OH)₂SaltMe]_{2 M} electrocatalytic responses register during separate (Fig. 5d) and continuous urea additions (Fig. S4d†), generation ability could not be determined because of overlapping of urea oxidation and OER processes.

The best results were obtained for both catalysts fabricated in 1.0 M NaOH_aq. These catalysts indicated the highest ability of active center generation in 1.0 M NaOH_aq (Table S4†). Furthermore, the higher in situ generation abilities obtained for poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} revealed that this catalyst is more resistant to the presence of electrooxidation products.

Knowing that increasing the urea concentration increases the anodic peak current density if the catalytic reaction is under diffusion control,²³ we analyzed the electrocatalytic responses of poly[NP-Ni(OH)₂salen] catalysts fabricated in different NaOH_aq concentrations, focusing on the linearity ranges obtained after the separate urea additions (Fig. 5 and 6). For all poly[NP-Ni(OH)₂salen] catalysts fabricated in different NaOH_aq concentrations, forward scan peak current density increase was observed after separate urea additions until 0.3 M (Fig. 5 and 6). This observation is connected with competing interactions of urea with OH⁻ counterions to reach the accessible active centers. When the urea concentration was too low to overcome this competition, the diffusion of urea molecules to the accessible active sites was the slowest process. However, for urea concentrations greater than 0.3 M (Fig. 5 and 6), the amount of urea was sufficient to overcome the binding affinity competition with OH⁻, and the urea electrooxidation on the poly(NP-Ni³⁺OOHsalen) active sites became the slowest step. Because of that, for further experiments, we chose a urea concentration of 0.3 M.

Optimization of NaOH_aq concentration during electrocatalytic oxidation of urea on structurally tuned catalysts

The onset potential is one of the most important parameters describing the studied material's electrocatalytic performance. It is defined as the lowest potential value at which the reaction products start to appear.^53,54 A simultaneous decrease in the onset potential and increased anodic peak current density in the forward scan indicate higher catalytic activity for urea electrooxidation catalysts operating in alkaline electrolytes.³ Consecutively, the higher backward anodic peak current density indicates higher resistance against poisoning/product accumulation according to the indirect mechanism.^23,29 Additionally, it was shown that an increase in an alkaline electrolyte concentration may result in shifting the onset potential to lower values and increasing the anodic peak current density, thus forcing oxidation of intermediates and desorption of final reaction products.³² Encouraged by these facts, two NP catalysts, poly[NP-Ni(OH)₂SaltMe]_{1 M} and poly[meso-NP-Ni(OH)₂SaldMe]_{1 M}, fabricated in 1.0 M NaOH_aq, revealing the best in situ generation ability, were chosen to optimize the best NaOH_aq concentration for the electrocatalytic oxidation of 0.3 M urea. Both catalysts were electrochemically examined in the absence (Fig. S6†) and presence of 0.3 M urea (Fig. 7) in 0.2, 0.5, 1.0, and 2.0 M NaOH_aq.

Fig. 7 The catalytic CV responses of (a) poly[NP-Ni(OH)₂SaltMe]_{1 M}, and (b) poly[meso-NP-Ni(OH)₂ SaldMe]_{1 M} toward 0.3 M urea in different NaOHaq concentrations. The CVs were performed at 50 mV s⁻¹. Insets display CV responses of (a) poly[NP-Ni(OH)₂SaltMe]_{1 M}, and (b) poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} registered at 50 mV s⁻¹ in the presence and absence of 0.3 M urea in 1.0 M NaOH_aq.

In the urea-free electrolyte solution, a shift of Ni²⁺/Ni³⁺ oxidation/reduction potentials to lower values was observed with increasing NaOH_aq concentration (Fig. S6a and b†). Moreover, there were visible differences in the peak current heights. The highest current density was obtained for CV responses registered in 1.0 and 0.5 M NaOH_aq for poly[NP-Ni(OH)₂SaltMe]_{1 M} (Fig. S6,† red and blue curves) and for 1.0 M NaOH_aq for poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} (Fig. S6b,† red curve), indicating that the highest amount of poly(NP-Ni³⁺OOHsalen) active centers was electrogenerated for those NaOH_aq concentrations. Furthermore, slightly higher current density obtained for poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} indicated that more active sites were electrogenerated for this catalyst. The lower current density observed in the presence of 2.0 M NaOH_aq suggested that in this concentration, both catalysts might be chemically unstable and side reactions occur (the example of such a side reaction is presented in eqn (5)).

In the presence of urea (Fig. 7), a shift in the onset potential of the urea electrooxidation to lower values was observed with the increase of the NaOH_aq concentration for both poly[NP-Ni(OH)₂salen]_{1 M} structured catalysts. Comparing these two catalysts, lower onset potential values were obtained for poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} when comparing 0.3 M urea electrooxidation for all NaOH_aq concentrations (Table 1). Most likely, it is because of the more efficient generation of poly(meso-NP-Ni³⁺OOHSaldMe) catalytically active centers and their bigger spatial accessibility for urea and OH⁻. This accessibility seemed disrupted when catalytic oxidation was conducted in 2.0 M NaOH_aq (Fig. 7a and b, black curves). Hence, it is directly reflected in the lack of simultaneous anodic peak growth and onset potential decrease because of too strong adsorption of OH⁻ resulting in the merging of the OER⁵⁵ with urea oxidation in 2.0 M NaOH_aq. These results revealed that 2.0 M NaOH_aq was not appropriate for urea oxidation on poly[NP-Ni(OH)₂SaltMe]_{1 M} and poly[meso-NP-Ni(OH)₂SaldMe]_{1 M}. The origin and comparison of the impact of active center blocking in 2.0 M NaOH_aq for structured catalysts is described in detail in the part describing multiple scan rate experiments conducted in blank NaOH_aq.

Table 1 Onset potential measured in different concentrations of NaOH_aq in the presence of 0.3 M urea for poly[NP-Ni(OH)₂SaltMe]_{1 M} and poly[meso-NP-Ni(OH)₂SaldMe]_{1 M}

C NaOH (M)	Poly[NP-Ni(OH)2SaltMe]1 M onset potential (V)	Poly[meso-NP-Ni(OH)2SaldMe]1 M onset potential (V)
0.2	0.474	0.466
0.5	0.445	0.438
1.0	0.424	0.416
2.0	0.406	0.399

---

Electrocatalytic performance comparison of poly[NP-Ni(OH)₂SaltMe]_{1 M} and poly[meso-NP-Ni(OH)₂SaldMe]_{1 M}

Multiple scan rate experiments were registered for poly[NP-Ni(OH)₂SaltMe]_{1 M} and poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} in two different concentrations of pure electrolyte: 1.0 M NaOH_aq (Fig. 8 and 9) and 2.0 M NaOH_aq (Fig. S7 and S8†). These experiments aimed to determine differences in the limiting conditions of the electrochemical redox process of poly[NP-Ni²⁺(OH)₂salen] ⇌ poly[NP-Ni³⁺OOHsalen] influenced by the electrochemical dopant concentration for both differently structured materials.

Fig. 8 (a) CV curves measured over poly[NP-Ni(OH)₂SaltMe]_{1 M} at 2, 5, 10, 30, 50, 100, 200 mV s⁻¹ scan rate in 1.0 M NaOH_aq. (b) The anodic (red curve) and cathodic (blue curve) peak current dependence on the scan rate. (c) The anodic and cathodic peak currents vs. the square root of the scan rate. (d) The anodic and cathodic peak potentials vs. the logarithm of the scan rate. (e) The logarithm of the anodic and cathodic peak currents vs. the logarithm of the scan rate.

Fig. 9 (a) CV curves of poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} at 2, 5, 10, 30, 50, 100, and 200 mV s⁻¹ scan rate in 1.0 M NaOH_aq. (b) The anodic (red curve) and cathodic (blue curve) peak current dependence on the scan rate. (c) The anodic and cathodic peak currents vs. the square root of the scan rate. (d) The anodic and cathodic peak potentials vs. the logarithm of the scan rate. (e) The logarithm of the anodic and cathodic peak currents vs. the logarithm of the scan rate.

Multiple scan rate CV responses registered for poly[NP-Ni(OH)₂SaltMe]_{1 M} and poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} in 1.0 M NaOH_aq (Fig. 8 and 9) and 2.0 M NaOH_aq (Fig. S7a and S8a†) revealed that the characteristic peak currents corresponding to electrooxidation and electroreduction of nickel hydroxide-type redox centers, poly[NP-Ni²⁺(OH)₂salen] ⇌ poly[NP-Ni³⁺OOHsalen], were linearly dependent on the scan rate up to 20 mV s⁻¹ (Fig. 8b, 9b, S7b and S8b), indicating that considered electrochemical redox processes occurring in 1.0 M and 2.0 M NaOH_aq for both materials were under finite diffusion control⁵⁶ of charge maintaining OH⁻ counterions. However, at scan rates above 20 mV s⁻¹, it seems that the characteristic peak currents (Fig. 8c, 9c, S7c, and S8c†) were linearly dependent on the square root of scan rates, thus suggesting that diffusion of OH⁻ was close to entering a semi-infinitive diffusion regime⁴⁸ in 1.0 M and 2.0 M NaOH_aq for both materials.

Comparing multiple scan rate responses within one material, i.e., separately poly[NP-Ni(OH)₂SaltMe]_{1 M} and poly[meso-NP-Ni(OH)₂SaldMe]_{1 M}, operating either in 1.0 M or in 2.0 M NaOH_aq one can notice the differences in the current level and reversibility of the redox processes. Higher current levels coupled with the Nernstian nature of electrochemical responses were observed in 1.0 M NaOH_aq for both analyzed materials. Hence, the peak potential separation observed for poly[NP-Ni(OH)₂SaltMe]_{1 M} (Fig. 8d) and poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} (Fig. 9d) in 1.0 M NaOH_aq was low, up to 20 mV s⁻¹, and then it increased because the OH⁻ diffusion cannot keep the rate of charge transfer. The lower current level responses (Fig. S7a and S8a†), increased peak potential separations (Fig. S7d and S8d†), and lower values of the slopes determined from the dependence of the logarithm of the peak currents vs. the logarithm of the scan rate observed for both materials operating in 2.0 M NaOH_aq pointed out hindered OH⁻ diffusion to redox centers. This suggests that redox centers of both materials, poly[NP-Ni(OH)₂SaltMe]_{1 M} and poly[meso-NP-Ni(OH)₂SaldMe]_{1 M}, operating in 2.0 M NaOH_aq were blocked by too strong adsorption of counterions that most probably resulted in chemical degradation of centers.

Meanwhile, the differences between the Nernstian multiple scan rate responses registered in 1.0 M NaOH_aq for poly[NP-Ni(OH)₂SaltMe]_{1 M} and poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} are indicative of the nature of the OH⁻ finite diffusion control electrochemical redox process. The slope value determined from the dependence of the logarithm of the peak currents vs. the logarithm of the scan rate for poly[NP-Ni(OH)₂SaltMe]_{1 M} was 0.79 (Fig. 8e), which is characteristic of an electrochemical redox process occurring on the surface of the active material,⁵⁷ while the lower slope value (0.65) obtained for poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} (Fig. 9e) indicates that the fast electrochemical redox process occurred in the bulk of the active material. This signifies that the poly(SaltMe) matrix is relatively flat and possesses active sites close to the surface. In contrast, the poly(meso-SaldMe) matrix is spatially diversified, and Ni³⁺OOH active sites are embedded in the bulk of the matrix.

The multiple scan rate electrocatalytic responses of the poly[NP-Ni(OH)₂SaltMe]_{1 M} and poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} catalysts were registered in 1.0 M NaOH_aq solution containing 0.3 M urea to distinguish which of these two structured catalyst indicated better catalytic performance. To obtain the origin of the differences in the electrocatalytic oxidation of urea on the poly[NP-Ni(OH)₂SaltMe]_{1 M} and poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} at different scan rates, we coupled previously obtained conclusions considering active center generation in pure 1.0 M NaOH_aq with new findings from electrocatalytic oxidation of urea experiments probing its active center accessibility.

The electrochemical responses obtained for both catalysts in the presence of 0.3 M urea did not show the electrocatalytic peak current heights remaining at the same level for all studied scan rates (Fig. 10a and 11a). Fig. 10b and 11b show that the peak currents were linearly dependent on the square root of the scan rate below 20 mV s⁻¹, thus indicating that only for those scan rates the desired charge transfer control of urea oxidation on fast generated poly[NP-Ni³⁺OOHsalen] was achieved. Above 20 mV s⁻¹, the generation of active centers was slow, thus reflecting OH⁻diffusion limitations recognized in the experiment conducted in 1.0 M NaOH_aq (Fig. 8 and 9). Additionally, the dependence of catalytic peak currents normalized towards the square root of the scan rate plotted vs. scan rate confirmed that the desired chemical oxidation of urea was the controlling step, i.e., the slowest step of the overall electrocatalytic process occurring below 20 mV s⁻¹ (Fig. 10e and 11e). At scan rates higher than 20 mV s⁻¹, the generation of poly[NP-Ni³⁺OOHsalen] forms was slowed down because of OH⁻ diffusional limitations; thus, the normalized current values on Fig. 10e and 11e did not change much. Moreover, the slope of the logarithm of the current vs. the logarithm of the scan rate determined for 2 to 20 mV s⁻¹ was 0.25 (Fig. 10d) and 0.24 (Fig. 11d) for poly[NP-Ni(OH)₂SaltMe]_{1 M} and poly[meso-NP-Ni(OH)₂SaldMe]_{1 M}, respectively, indicating that the accessibility of the poly[meso-NP-Ni³⁺OOHSaldMe] electrocatalytically active sites for urea was bigger.

Fig. 10 (a) CV curves of poly[NP-Ni(OH)₂SaltMe]_{1 M} at 2, 5, 10, 20, 50, 100, 200 mV s⁻¹ scan rate in the presence of 0.3 M urea in 1.0 M NaOH_aq. (b) The anodic and cathodic peak currents vs. the square root of the scan rate. (c) The anodic and cathodic peak potentials vs. the logarithm of the scan rate. (d) The logarithm of the anodic and cathodic peak currents vs. the logarithm of the scan rate. (e) The scan rate normalized current (I_pav^−1/2) vs. the square root of the scan rate.

Fig. 11 (a) CV curves of poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} at 2, 5, 10, 20, 50, 100, 200 mV s⁻¹ scan rates in the presence of 0.3 M urea in 1.0 M NaOH_aq. (b) The anodic and cathodic peak currents vs. the square root of the scan rate. (c) The anodic and cathodic peak potentials vs. the logarithm of the scan rate. (d) The logarithm of the anodic and cathodic peak currents vs. the logarithm of the scan rate. (e) The scan rate normalized current (I_pav^−1/2) vs. the square root of the scan rate.

Furthermore, to parameterize the poly[NP-Ni(OH)₂SaltMe]_{1 M} and poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} molecular structure differences and their implication for the electrocatalytic urea oxidation process, α and D were calculated based on data obtained from the electrocatalytic multiple scan rate experiment. These parameters are related to the electrocatalytic process occurring at the interface of the NP-Ni³⁺OOHsalen|urea containing solution – 2nd interface (Scheme 2). α is associated with the electrocatalytic charge transfer occurring at the NP-Ni³⁺OOHsalen|urea interface, i.e., the second interface (Scheme 2). α can range from 0 to 1; α <0.5 means a one-electron process, and α >0.5 indicates a multistep system. The α values were calculated according to eqn (1) based on the slopes determined for data obtained in the scan rate range of 2 to 20 mV s⁻¹ (Fig. 10c and 11c). The α values were 0.59 and 0.60 for poly[NP-Ni(OH)₂SaltMe]_{1 M} and poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} catalysts, respectively. These values confirmed that urea undergoes oxidation according to the two-step indirect mechanism where six electrons are exchanged within electrocatalytic charge transfer at the 2nd interface (Scheme 2).

D describes how fast urea is able to reach active centers.²³D values were calculated from eqn (2) based on the slopes determined from dependencies presented in Fig. 10b and 11b. For poly[NP-Ni(OH)₂SaltMe]_{1 M} and poly[meso-NP-Ni(OH)₂SaldMe]_{1 M}, the calculated D values were: 5.87 × 10⁻⁷ cm² s⁻¹ and 7.49 × 10⁻⁷ cm² s⁻¹, respectively. A higher D value obtained for urea reaching the poly(meso-NP-Ni³⁺OOHSaldMe)_{1 M} active sites indicated that a spatially diversified molecular structure of poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} facilitated the electrocatalytic process.

The Tafel slopes were determined to distinguish which structured poly[NP-Ni(OH)₂salen] catalyst produces catalytic current more efficiently in response to change in applied potential. The Tafel slopes determined for poly[NP-Ni(OH)₂SaltMe]_{1 M} and poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} catalysts (Fig. 12) operating in the solution containing 0.3 M urea and 1.0 M NaOH_aq were 25 mV dec⁻¹ and 20 mV dec⁻¹, respectively. A lower Tafel slope obtained for poly[meso-NP-Ni³⁺OOHSaldMe]_{1 M} indicates that for this catalyst there is a significant current density increment as a function of the overpotential change, or in other words, this catalyst produces catalytic current more efficiently in response to change in applied potential than poly[NP-Ni(OH)₂SaltMe]_{1 M}. Furthermore, the obtained Tafel slopes are lower than those reported for other well-performing urea electrooxidation catalysts.^18,19,58

Fig. 12 Tafel slopes determined for (a) poly[NP-Ni(OH)₂SaltMe]_{1 M} and (b) poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} catalysts operating in the solution containing 0.3 M urea and 1.0 M NaOH_aq.

To compare the E_ac values of urea electrooxidation on poly[NP-Ni(OH)₂SaltMe]_{1 M} and poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} and study the influence of temperature on the overall two-step catalytic process, the CV responses in 1.0 M NaOH_aq were registered in the absence (Fig. S10†) and the presence of 0.3 urea (Fig. 13) at different temperatures (16, 19, 24, 28, 33, 38, and 43 °C). The control experiment performed in the absence of urea showed that with the temperature increase, the peak current slightly increased together with the peak potential decrease for poly[NP-Ni(OH)₂SaltMe]_{1 M} (Fig. S10a†), indicating more efficient generation of active sites at higher temperatures.

Fig. 13 Electrocatalytic CV responses of (a) poly[NP-Ni(OH)₂SaltMe]_{1 M} and (b) poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} performed at different temperatures in the solution containing 0.3 M urea and 1.0 M NaOH_aq.

A similar behavior was observed for poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} but only for temperatures below 28 °C (Fig. S10b†). Further temperature increase led to a decrease in peak current coupled with the shift of peak potential to a higher value, thus suggesting hindered dedoping of OH⁻ from poly(meso-NP-Ni³⁺OOHSaldMe)_{1 M} active sites (Fig. S10b†). This hindered dedoping of OH⁻ counterions was associated with a more spatially diversified structure of matrix templating active centers in 3D. In the presence of 0.3 M urea, with the temperature increase, the catalytic peak currents and the onset potentials decreased for both studied catalysts (Fig. 13), thus indicating the facilitated electrooxidation of urea at higher temperatures. A shift of the catalytic peak potential from 0.49 V to 0.66 V for poly[NP-Ni(OH)₂SaltMe]_{1 M} (Fig. 13a) and from 0.55 V to 0.67 V for poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} (Fig. 13b) was observed. Because in the control experiment, we did not observe hindered desorption of OH⁻ for poly[NP-Ni(OH)₂SaltMe]_{1 M} for all studied temperatures, thus the catalytic peak potential shift is undeniably attributed to the electrooxidation product accumulation. However, in the case of poly(meso-NP-Ni³⁺OOHSaldMe)_{1 M}, the catalytic peak potential shift is attributed to hindered desorption of OH⁻, which is supported by the control experiment (Fig. S10b†). The observed hindered dedoping of OH⁻ ions from spatially expanded poly(meso-NP-Ni³⁺OOHSaldMe)_{1 M} positively influences the final desorption of catalysis products. This result remains consistent with earlier analysis of the in situ generation/retention ability of poly(NP-Ni³⁺OOHsalen) centers.

Arrhenius plots were prepared for potentials 0.45 and 0.5 V at different temperatures for urea electrooxidation. These plots showed a linear relation (Fig. 14), indicating no change in the two-step indirect urea oxidation mechanism for all studied temperatures.⁵⁹E_ac was determined from Arrhenius plots according to eqn (6):

where I is the current density, T is the temperature, and R is the gas constant. The calculated E_ac was lower for poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} (20.82 and 21.35 kJ mol⁻¹ for potentials 0.45 and 0.50 V, respectively) than for poly[NP-Ni(OH)₂SaltMe]_{1 M} (21.60 and 33.36 kJ mol⁻¹ for potentials 0.45 and 0.50 V, respectively). These determined E_ac values indicated that poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} exhibits higher catalytic activity than poly[NP-Ni(OH)₂SaltMe]_{1 M} toward urea electrooxidation.

Fig. 14 Arrhenius plots prepared for (a) poly[NP-Ni(OH)₂SaltMe]1 M and (b) poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} based on data from Fig. 13.

Long-term stability test under real-system mimicking conditions

Chronoamperometric experiments were performed to mimic the operating conditions of the anode of the artificial urine and urea fuel cells. During these experiments, a constant potential was kept for 17 h in 1.0 M NaOH_aq medium, ensuring progressing electrocatalytic oxidation of urea (i) from its pristine form (Fig. 15) or (ii) from artificial urine (Fig. 16) on the poly[NP-Ni(OH)₂SaltMe]_{1 M} and poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} anodes. The same conditions were applied in the control experiment, where poly[NP-Ni(OH)₂SaltMe]_{1 M} and poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} were continuously electrooxidized to poly(NP-Ni³⁺OOHSaldMe)_{1 M} and poly(meso-NP-Ni³⁺OOHSaldMe)_{1 M}, respectively. Under these conditions, the amount of electrogenerated poly(NP-Ni³⁺OOHsalen) centers was constant and unchanged for both studied materials (Fig. 15, red curves). After urea addition, the sustenance of active poly(NP-Ni³⁺OOHSaldMe) species throughout the pending reaction remains unaccomplished because of the indirect mechanism that involves the reduction of poly(NP-Ni³⁺OOHsalen) into poly[NP-Ni²⁺(OH)₂salen] during electrocatalysis. The current retention values obtained after 15 h of pure urea electrooxidation on poly[NP-Ni(OH)₂SaltMe]_{1 M} and poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} (Fig. 15) were ∼44% and ∼56% (Table S6†), respectively.

Fig. 15 Chronoamperometric measurements performed in 1.0 M NaOH_aq in the absence (red) and presence (black) of 0.3 M urea for (a) poly[NP-Ni(OH)₂SaltMe]_{1 M} and (b) poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} at a constant potential of 0.55 V vs. Ag/AgCl.

Fig. 16 Chronoamperometric measurements performed in 1.0 M NaOH_aq in the absence (red) and the presence (black) of artificial urine containing 0.25 M urea for (a) poly[NP-Ni(OH)₂SaltMe]_{1 M} and (b) poly[meso-NP Ni(OH)₂SaldMe]_{1 M} at the constant potential of 0.55 V vs. Ag/AgCl.

The respective retention values obtained for these catalysts operating in the presence of artificial urine were ∼65% and ∼70% (Fig. 16 and Table S6†). Poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} indicated higher current retention in both analyzed cases. The high conductivity of artificial urine solution dissolved in 1 M NaOH_aq enhanced the electrocatalytic current output (Fig. 16) compared to the current measured at 0.3 M urea in 1 M NaOH_aq (Fig. 15).⁶⁰ Measurements in artificial urine helped check the proof of using the poly[NP-Ni(OH)₂salen] catalysts for urea oxidation from urea-rich wastewater-like systems.

Conclusions

We optimized the best NaOH_aq concentration used for the fabrication of two differently structured catalysts named poly[NP-Ni(OH)₂SaltMe]_{1 M} and poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} active towards urea electrooxidation. A combination of SEM and TEM imaging supplemented by evaluating the poly(NP-Ni³⁺OOHsalen) catalytic centers' in situ generation abilities according to the indirect urea electrooxidation mechanism was our criteria for choosing the best catalyst fabrication conditions. Recognition of the two-step indirect mechanism of electrocatalytic urea oxidation on poly[NP-Ni(OH)₂SaltMe]_{1 M} and poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} imposed a need for finding the best possible catalytic conditions. Hence we focused on finding the best NaOH_aq and urea concentration for electrocatalysis. Both studied catalysts indicated the best performance in 1.0 M NaOH_aq containing 0.3 M urea. Moreover, we elucidated the differences in the electrocatalytic performance of the fabricated catalysts using comprehensive electrochemical methodology. This methodology primarily involved electrochemical control studies differentiating the indirect two-step process of urea oxidation, determination of urea diffusion coefficients, Tafel slope analysis, and activation energy studies. Finally, we elucidated the differences in structure-related performance of poly[NP-Ni(OH)₂SaltMe]_{1 M} and poly[meso-NP-Ni(OH)₂SaldMe]_{1 M}. These catalysts featured different placements of active center arrangements, i.e., poly[NP-Ni(OH)₂SaltMe]_{1 M} active centers remained mainly on the catalyst surface, while poly[meso-NP-Ni(OH)₂SaldMe]_{1 M} active centers were distributed in the catalyst bulk. It turned out that the high spatial distribution of poly(NP-Ni³⁺OOHSaldMe)_{1 M} catalytically active sites caused their better sustenance during electrocatalytic urea oxidation. This was also reflected in the high retention of catalytic current obtained for poly(NP-Ni³⁺OOHSaldMe)_{1 M} when operating in real-system mimicking conditions. Moreover, the Tafel slope of 20 mV dec⁻¹ and E_a of 20.82 kJ mol⁻¹ confirmed that this bulk-structured catalyst is efficient and highly active towards UOR.

Data availability

The authors confirm that the data supporting the findings of this study are available within the article and/or in its ESI.†

Author contributions

Monika Mierzejewska: electrochemical experiments, writing. Kamila Łępicka: conceptualization, Ni-salen monomer synthesis, writing. Jakub Kalecki: SEM imaging. Piyush Sindhu Sharma: project management, editing and writing.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

The authors thank Dr Kamil Sobczak (Biological and Chemical Research Centre, University of Warsaw) for TEM analysis. The present research was financially supported by the National Science Center (Poland) through Grant No. 2018/29/B/ST5/02335 to PSS.

References

1. R. K. Singh, K. Rajavelu, M. Montag and A. Schechter, Energy Technol., 2021, 9, 2100017.
2. A. Coombs, Nature, 2008, 2578.
3. E. T. Sayed, T. Eisa, H. O. Mohamed, M. A. Abdelkareem, A. Allagui, H. Alawadhi and K.-J. Chae, J. Power Sources, 2019, 417, 159–175.
4. J. Li, S. Wang, S. Sun, X. Wu, B. Zhang and L. Feng, J. Mater. Chem. A, 2022, 10, 9308–9326.
5. Y. M. T. A. Putri, J. Gunlazuardi, Y. Yulizar, R. Wibowo, Y. Einaga and T. A. Ivandini, Open Chem., 2021, 19, 1116–1133.
6. S. Lu, M. Hummel, Z. Gu, Y. Wang, K. Wang, R. Pathak, Y. Zhou, H. Jia, X. Qi, X. Zhao, B. B. Xu and X. Liu, ACS Sustainable Chem. Eng., 2021, 9, 1703–1713.
7. B. Zhu, Z. Liang and R. Zou, Small, 2020, 16, 1906133.
8. E. Urbanczyk, M. Sowa and W. Simka, J. Appl. Electrochem., 2016, 46, 1011–1029.
9. W. Simka, J. Piotrowski and G. Nawrat, Electrochim. Acta, 2007, 52, 5696–5703.
10. A. T. Miller, B. L. Hassler and G. G. Botte, J. Appl. Electrochem., 2012, 42, 925–934.
11. R. L. King and G. G. Botte, J. Power Sources, 2011, 196, 9579–9584.
12. A. Kumar, X. Liu, J. Lee, B. Debnath, A. R. Jadhav, X. Shao, V. Q. Bui, Y. Hwang, Y. Liu, M. G. Kim and H. Lee, Energy Environ. Sci., 2021, 14, 6494–6505.
13. B. K. Boggs, R. L. King and G. G. Botte, Chem. Commun., 2009, 4859–4861.
14. H. Kim, T. Y. Yoo, M. S. Bootharaju, J. H. Kim, D. Y. Chung and T. Hyeon, Adv. Sci., 2022, 9, 2104054.
15. M. A. Goda, M. G. A. El-Moghny and M. S. El-Deabz, J. Electrochem. Soc., 2020, 167, 064522.
16. F. Wu, G. Ou, J. Yang, H. Li, Y. Gao, F. Chen, Y. Wang and Y. Shi, Chem. Commun., 2019, 55, 6555–6558.
17. L. Xia, Y. Liao, Y. Qing, H. Xu, Z. Gao, W. Li and Y. Wu, ACS Appl. Energy Mater., 2020, 3, 2996–3004.
18. L. Jin, R. Ji, H. Wan, J. He, P. Gu, H. Lin, Q. Xu and J. Lu, ACS Catal., 2023, 13, 837–847.
19. N. N. Rao, C. Alex, M. Mukherjee, S. Roy, A. Tayal, A. Datta and N. S. John, ACS Catal., 2024, 14, 981–993.
20. Y. Tong, P. Chen, M. Zhang, T. Zhou, L. Zhang, W. Chu, C. Wu and Y. Xie, ACS Catal., 2018, 8, 1–7.
21. R. P. Forslund, C. T. Alexander, A. M. Abakumov, K. P. Johnston and K. J. Stevenson, ACS Catal., 2019, 9, 2664–2673.
22. R. P. Forslund, J. T. Mefford, W. G. Hardin, C. T. Alexander, K. P. Johnston and K. J. Stevenson, ACS Catal., 2016, 6, 5044–5051.
23. V. Vedharathinam and G. G. Botte, Electrochim. Acta, 2012, 81, 292–300.
24. G. Hopsort, D. P. D. Carmo, L. Latapie, K. Loubière, K. G. Serrano and T. Tzedakis, Electrochim. Acta, 2023, 442, 141898.
25. R. K. Singh and A. Schechter, Electrochim. Acta, 2018, 278, 405–411.
26. R. M. Abdel Hameed and S. S. Medany, Int. J. Hydrogen Energy, 2019, 44, 3636–3648.
27. S. Wang, P. Xu, J. Tian, Z. Liu and L. Feng, Electrochim. Acta, 2021, 370, 137755.
28. C. Alex, G. Shukla and N. S. John, Electrochim. Acta, 2021, 385, 138425.
29. V. Vedharathinam and G. G. Botte, Electrochim. Acta, 2013, 108, 660–665.
30. M. Mazloum-Ardakani, V. Eslami and A. Khoshroo, Mater. Sci. Eng., B, 2018, 229, 201–205.
31. M. Mierzejewska, K. Łępicka, J. Kalecki, W. Lisowski and P. S. Sharma, ACS Appl. Mater. Interfaces, 2022, 14, 33768–33786.
32. W. Wang, D. Chai, J. Zhang, S. Xue, Y. Wang and Z. Lei, J. Taiwan Inst. Chem. Eng., 2017, 80, 326–332.
33. D. A. Daramola, D. Singh and G. G. Botte, J. Phys. Chem. A, 2010, 114, 11513–11521.
34. L. F. Huang, M. J. Hutchison, R. J. Santucci Jr., J. R. Scully and J. M. Rondinelli, J. Phys. Chem. C, 2017, 121, 9782–9789.
35. M. Pourbaix, Atlas of Electrochemical Equilibria, National Association of Corrosion Engineers, Houston, Texas, USA, 1974.
36. S. Lu, X. Zheng, L. Fang, F. Yin and H. Liu, Electrochem. Commun., 2023, 107599.
37. X. Hu, J. Zhu, J. Li and Q. Wu, ChemElectroChem, 2020, 7, 3211–3228.
38. J. Ge, J. Kuang, Y. Xiao, M. Guan and C. Yang, Surf. Interfaces, 2023, 41, 103230.
39. N. Kuznetsov, P. Yang, G. Gorislov, Y. Zhukov, V. Bocharov, V. Malev and O. Levin, Electrochim. Acta, 2018, 271, 190–202.
40. J. L. Bott-Neto, T. S. Martins, S. R. A. S. Machado and E. A. Ticianelli, ACS Appl. Mater. Interfaces, 2019, 11, 30810–30818.
41. A. Olean-Oliveira, C. F. Pereira, D. N. David-Parra and M. F. S. Teixeira, ChemElectroChem, 2018, 5, 3557–3565.
42. M. Vilas-Boas, C. Freire, B. de Castro and A. R. Hillman, J. Phys. Chem. B, 1998, 102, 8533–8540.
43. K. Łępicka, P. Pieta, A. Shkurenko, P. Borowicz, M. Majewska, M. Rosenkranz, S. Avdoshenko, A. A. Popov and W. Kutner, J. Phys. Chem. C, 2017, 121, 16710–16720.
44. N. Sarigul, F. Korkmaz and İ. Kurultak, Sci. Rep., 2019, 9, 20159.
45. K. Łępicka, P. Pieta, R. Gupta, M. Dabrowski and W. Kutner, Electrochim. Acta, 2018, 268, 111–120.
46. E. Laviron, J. Electroanal. Chem. Interfacial Electrochem., 1979, 101, 19–28.
47. L. M. F. Dantas, A. P. D. Reis, S. M. C. N. Tanaka, J. H. Zagal, Y.-Y. Chen and A. A. Tanaka, J. Braz. Chem. Soc., 2008, 19, 720–726.
48. A. Bard and L. Faulkner, Electrochemical Methods. Fundamentals and Applications, John Wiley & Sons, New York, 2nd edn, 2001.
49. M. Vilas-Boas, C. Freire, B. de Castro and A. R. Hillman, J. Phys. Chem. B, 1998, 102, 8533–8540.
50. M. Vilas-Boas, C. Freire, B. de Castro, P. A. Christensen and A. R. Hillman, Inorg. Chem., 1997, 36, 4919–4929.
51. M. Vilas-Boas, C. Freire, B. de Castro, P. A. Christensen and A. R. Hillman, Chem. – Eur. J., 2001, 7, 139–150.
52. R. Lin, L. Kang, T. Zhao, J. Feng, V. Celorrio, G. Zhang, G. Cibin, A. Kucernak, D. J. L. Brett, F. Cora, I. P. Parkin and G. He, Energy Environ. Sci., 2022, 15, 2386–2396.
53. C. Batchelor-McAuley, Curr. Opin. Electrochem., 2023, 37, 101176.
54. R. Sanchis-Gual, A. Seijas-Da Silva, M. Coronado-Puchau, T. F. Otero, G. Abellán and E. Coronado, Electrochim. Acta, 2021, 388, 138613.
55. X. Gao, X. Bai, P. Wang, Y. Jiao, K. Davey, Y. Zheng and S.-Z. Qiao, Nat. Commun., 2023, 14, 5842.
56. K. Aoki, K. Tokuda and H. Matsuda, J. Electroanal. Chem. Interfacial Electrochem., 1983, 146, 417–424.
57. J. O. M. Bockris and S. U. M. Khan, Surface Electrochemistry, A Molecular Level Approach, Springer New York, Plenum Press, New York and London, 1993.
58. Z. Ji, Y. Song, S. Zhao, Y. Li, J. Liu and W. Hu, ACS Catal., 2022, 12, 569–579.
59. J. Masa, S. Barwe, C. Andronescu and W. Schuhmann, Chem. – Eur. J., 2019, 25, 158–166.
60. C. Santoro, M. J. S. Garcia, X. A. Walter, J. You, P. Theodosiou, I. Gajda, O. Obata, J. Winfield, J. Greenman and I. Ieropoulos, ChemElectroChem, 2020, 7, 1312–1331.

---

Footnote

† Electronic supplementary information (ESI) available: Additional experimental details, results and supporting tables. See DOI: https://doi.org/10.1039/d4cy01139b

---