Photoelectrocatalytic degradation of high-density polyethylene microplastics on TiO2-modified boron-doped diamond photoanode

Summary Microplastic (MP) accumulation in the environment is accelerating rapidly, which has led to their effects on both the ecosystem and human life garnering much attention. This study is the first to examine the degradation of high-density polyethylene (HDPE) MPs via photoelectrocatalysis (PEC) using a TiO2-modified boron-doped diamond (BDD/TiO2) photoanode. This study was divided into three stages: (i) preparation of the photoanode through electrophoretic deposition of synthetic TiO2 nanoparticles on a BDD electrode; (ii) characterization of the modified photoanode using electrochemical, structural, and optical techniques; and (iii) degradation of HDPE MPs by electrochemical oxidation and photoelectrocatalysis on bare and modified BDD electrodes under dark and UV light conditions. The results indicate that the PEC technique degraded 89.91 ± 0.08% of HDPE MPs in a 10-h reaction and was more efficient at a lower current density (6.89 mA cm−1) with the BDD/TiO2 photoanode compared to electrochemical oxidation on bare BDD.


INTRODUCTION
The amount of plastic products used reached almost 370 million tons in 2019. 1 Perhaps more alarmingly, is that approximately 10 million tons of plastics are released into the oceans every year as pollution, and an estimated 2.7 3 10 5 tons of plastic are currently floating in the ocean. 2 Plastics begin to slowly degrade into microplastics (MPs) mainly as a result of the environmental conditions to which they are exposed (e.g., solar radiation, microorganisms, salinity, humidity, and abrasive effects). 2Researchers' have taken steps to study MPs (<5 mm) as water source contaminants because synthetic polymers are considered hazardous waste, and thus studies should be extended to soils, biota, and the earth's atmosphere. 3,4Polyethylene (PE) is indeed one of the most widely used commercial plastics in the world. 5Its versatility, durability, and low cost make it a popular choice for various applications, and it is also frequently found in marine environments. 6,7However, the monitoring of MP pollution has been so challenging that it has been necessary to create the concept of a ''microplastic cycle.'' 80][11] Specific research on MPs' consequences for the health of humans and other organisms is in the early stages.For instance, Feckelman et al. demonstrated that current levels of MPs pollution are leaving certain seabirds susceptible to their ingestion, finding that MPs affect the gut biomes of these birds. 12In humans, MPs made from PE and polystyrene have been shown to induce cytotoxicity due to oxidative stress for both cerebral and epithelial (surface lining) human cells. 13Moreover, MPs enter the body via ingestion and inhalation, and their harmful effects are likely related to cellular damage and immune and inflammatory reactions. 14Despite such evidence, environmental remediation by catalysis is only a recent and minor effort in the face of a major environmental dilemma.Modern technologies adapted for removal or remediation is based on the application of physical, chemical, and biological processes or some combination.][17][18] Previous research has shown that AOPs may be promising for the efficient removal of a wide variety of contaminants. 17,19Among the AOPs, photoelectrocatalysis (PEC) stands out as an environmentally friendly and efficient technology, as it strikes a balance between the energy costs of conventional heterogeneous photocatalysis and advanced electrochemical oxidation (EO). 20,21As a simple explanation, PEC can be divided into the following stages.First, light is absorbed on a crystal-lattice, photocatalytic thin film supported on a conductive substrate (photoelectrode) (light energy > photocatalyst band gap), generating electron-hole charge carriers ðe --

XRD and EDS analysis
In order to confirm that the BDD/TiO 2 photoanode significantly impacted the PEC and catalytic capabilities of the material and that the anatase phase of TiO 2 had higher photocatalytic activity than the rutile phase, the surface features of the photoelectrode were evaluated. 29,38he X-ray diffraction (XRD) patterns of TiO 2 Degussa P25 and the BDD/TiO 2 photoanode produced by electrophoretic deposition is presented in Figure 1A.The various peaks in the XRD pattern in part (b) correspond to the anatase phase and the tetragonal crystal structure according to ICDD no.900-9086.On the other hand, the peaks at 25.3 , 36.9 , 37.8 , 48.0 , 53.9 , 62.1 , 68.8 , 74.1 , and 82.2 were indexed as crystal planes (101), ( 103), ( 200), ( 105), ( 213), ( 116), (107), and (303), respectively. 32In part (c), the two distinct peaks at 43.9 and 75.3 correspond to crystal planes (111) and (220) of the diamond (ICDD no.01-089-3441) 39 ; the remaining peaks correspond to the niobium (Nb) substrate where the diamond layer was deposited.The XRD pattern of the BDD/TiO 2 photoanode in part (a) shows that the BDD diffraction peaks at 38 , 54 , and 69 are sharp, indicating an overlap or interplay between the TiO 2 and BDD crystal planes is possible. 40However, the peaks are consistent with the diffraction patterns, confirming the predominant presence of anatase crystalline phases on BDD.As shown in Figure 1, energy-dispersive X-ray spectroscopy (EDS) analysis confirmed the presence of adequately deposited TiO 2 , as well as the elemental composition.The EDS spectrum revealed the presence of C, O, and Ti in the photoanode, suggesting that the material is fairly pure, since no unwanted elements were observed.The average weight percentages were 83 %C, 15 %O, and 2 %Ti, which is consistent with the results obtained in another study. 41This was confirmed by scanning electron microscopy (SEM) images, presented in Figures 1C-1E.The micrograph in Figure 1C reveals the size and shape of the TiO 2 nanoparticles; it shows a rather polydisperse grain size distribution, with particle sizes ranging from 97 to 550 nm (Figure S1A) and an irregular spherical shape.The SEM image in 1D shows the bare BDD electrode composed of randomly disordered, angular polycrystalline grains, with grain size ranging from 1.4 to 13.4 mm (Figure S1B) and uniform distribution.Finally, in Figure 1E, the BDD/TiO 2 photoanode shows the heterojunction is well dispersed in interstitial spaces, while the photocatalyst is monodispersed.

PEC and optical analyses Electrochemical and photoelectrochemical characterization
The electro-and photoelectrochemical characterization of the bare BDD electrode and BDD/TiO 2 photoanode was done with cyclic voltammetry (CV) using the redox couple [Fe(CN) 6 ] 3-/4-.Information on the charge transfer kinetics of the bare and TiO 2 -modified BDD electrodes under both dark conditions and UV light exposure was obtained. 42In Figure 2A, the CV profile of the BDD electrode under UV light exhibits a slight increase in current compared to BDD under no light; moreover, the slope of the charge transfer region of the oxidation-reduction curves also increased due to the light's effect.This further demonstrates that due to the potential region where the ferri/ferro-cyanide redox reaction occurs and the role of high-energy light such as UV, bare BDD exhibits photocurrent without needing a photocatalyst to modify its surface.4][45] Similarly, the BDD/TiO 2 heterojunction exhibits well-defined redox peaks in both UV light and dark conditions.The anodic peak current intensity was 0.15 mA cm À2 for the modified BDD electrode, but in UV light, it increased to 0.21 mA cm À2 .Furthermore, in UV light, the anodic peak potential of BDD/TiO 2 shifted to 0.30 V (vs.Ag/AgCl), which was 0.35 V (vs.Ag/AgCl) for the bare BDD electrode; the heterogeneous rate constant (k ) for the BDD/TiO 2 photoanode was 5.42 3 10 À4 cm s À1 , which was approximately 1.7 times faster than that of the bare BDD electrode (3.97 3 10 À4 cm s À1 ) (See also Table S1).This may be attributed to the TiO 2 nanoparticles having a high surface area, which favors the adsorption of reagents on the photoelectrode surface, which enhances the charge transfer process. 41,46,47dditionally, the CV profile in Figure 2B demonstrates HDPE-MPs' influence on the photoelectrochemical response of the BDD/TiO 2 photoanode in the electrolyte medium.However, there is no extra oxidation peak, suggesting that MPs do not undergo direct oxidation; therefore, their oxidation is expected to occur mainly through PEC.Conversely, the HDPE MPs' high hydrophobicity requires a surfactant to decrease the surface tension between the aqueous medium and polymer to achieve a proper dispersion.Accordingly, the controlled addition of tween 20 (8 mmol L À1 ) is needed, which causes a subtle increase in the medium's resistance (a slight decrease in conductivity). 48,49However, because Tween 20 is also sensitive to oxidation at the electrode surface, a faradaic current contribution corresponding to this reaction occurs, which is indicated by a slight current increase in the voltammetric profile, shown in Figure 2B. 50,51Furthermore, the behavior of the BDD/TiO 2 under UV light improved the current response of the oxidation peak, because the onset potential (V onset ) for the HDPE MPs and Tween 20 mixture ($ 1.65 V (vs.Ag/AgCl)) was lower than that for HDPE MPs alone ($ 1.73 V (vs.Ag/AgCl)).
The inset of Figure 2B shows the Tafel plot analysis from the voltamperometric data; two slopes are visible, attributed to independent charge transfer processes.At an overpotential greater than 1.8 V (vs.Ag/AgCl), the oxygen evolution reaction (OER) is expected to be dominant; accordingly, prior to the OER, d OH generation was assumed to dominate, which was further verified via radical trapping analysis (see the '' d OH trapping study'' section).The Tafel analysis is critical as it allows the proper selection of the anodic overpotential value with the highest expected catalysis yield of d OH radicals and thus an effective degradation of organic matter. 23However, because of changes in cell configuration, such as increased volume or distance between electrodes in the PEC treatment, it is not uncommon to use overpotentials even higher than those predicted by Tafel analysis.Na 2 SO 4 (concentration 0.1 mol L À1 ) was ultimately selected for the electrochemical tests given its stability and its use in other MPs degradations. 49,52,53n this regard, the transient photocurrent response at a bias potential of +1.8 V (vs.Ag/AgCl) was evaluated according to the CV profile under UV light (Figure 2C).The highest photocurrent density was 214 mA cm À2 , which corresponded to the BDD/TiO 2 photoanode; this was significantly higher than that of the bare BDD (1.8 mA cm À2 ), indicating that the construction of the p-n heterojunction of BDD/TiO 2 favored the charge transfer across the interfaces of the two semiconductors and greatly inhibited the rapid recombination of photogenerated e - CB = h + VB pairs. 54o determine the drop in charge transfer resistance (R ct ) from electrode modification, Nyquist analysis was used (Figure 2D). 55Hence, in the Nyquist plot, the radius of the semicircle indicated a significant decrease in the photoanode's R ct (23 Ohms) compared to the bare electrode, and a decrease still occurred under UV light conditions; in other words, the presence of UV light increased the photoanode's internal electron mobility, which enhanced the charge dynamics and led to a decrease in the instantaneous recombination rate of e --CB =h + VB pairs.To assess the interfacial charge transfer and recombination, Mott-Schottky plots for BDD and TiO 2 were calculated.Figure 2E indicates the respective p-type and n-type semiconducting natures of BDD and TiO 2 .The capacitance-potential curve also demonstrates that a p-n heterojunction successfully formed between TiO 2 and BDD; conversely, the flat band potential (E fb ) was calculated from the extrapolation of the linear region of the x axis, and donor or acceptor density (N D or N A ) was calculated from the slope with Equation 1: 56 1) where C sc is the space charge capacitance; e is the elementary charge; N is the carrier concentration for the acceptor or donor (cm À3 ); E app is the applied bias potential; ε is the dielectric constant of the material and has a value of 55 and 5.6 for TiO 2 and BDD, respectively; ε 0 is the vacuum dielectric constant (8.85 3 10 À14 F cm À1 ); k B is the Boltzmann constant (1.38 3 10 À23 J K À1 ); T is the absolute temperature; and A is a material-specific constant.The E fb of TiO 2 was 0.32 V (vs.Ag/AgCl) with a positive slope (n-type semiconductor), and the E fb of the BDD was 1.73 V (vs.Ag/AgCl) with a negative slope (p-type semiconductor), which are consistent with previously reported values. 29,56,57This observation was further validated by the carrier density of BDD/TiO 2 (4.6 3 10 15 cm À3 ).Further, the anticipated upward shift of the Fermi level caused by the elevated charge carrier density may lead to a noticeably greater degree of band bending at the heterostructure's surface, which would encourage charge separation at the interface between the heterostructure and electrolyte. 57

Optical properties
Figure 3 shows the diffuse reflectance spectra (inset) and Tauc plot for determining the band gap energy of the bare BDD and BDD/TiO 2 photoanode according to Equation 2: 32 Where F(R) is the Kubelka-Munk function, hv is the photon energy, A is a constant, n is an integer that depends on the transition (1 for direct and 4 for indirect), and E g is the band gap of the semiconductor.TiO 2 and BDD are indirect transition semiconductors; therefore, (F(R) hv) 1/2 vs. hv was plotted, from which the values of E g for the TiO 2 and BDD electrodes were estimated as 3.32 eV and 4.93 eV, respectively.[60] Photoelectrocatalytic degradation of HDPE-MPs

FTIR analysis
Fourier transform infrared spectroscopy (FTIR) was used to quantify the carbonyl content of HDPE MPs in a spectral ranging range from 4000 to 650 cm À1 .The signal was averaged over 10 scans at a resolution of 4 cm À1 .As shown in Figure 4A, the characteristic peaks of -CH 2 in the HDPE MPs are not missed after EO and PEC treatments.The original polymer (spectrum c, Figure 4A) shows typical peaks at 2850 cm À1 and 2912 cm À1 corresponding to C-H stretching in -CH 2 groups.In addition, at 1470 cm À1 there is a bending band of CH 2 groups. 61,62dditionally, to monitor the oxidation and deterioration of the mechanical properties, we concentrated on the stretching vibrations of the carbonyl and hydroxyl groups.The carbonyl index (CI) is used to determine the degree of oxidation or degradation of certain materials, such as PE MPs, based on the absorbance ratio between the carbonyl group at 1714 cm À1 to that of a reference band (1500 and 1400 cm À1 ). 63omparing the spectra before and after electrochemical oxidation, no peak at 1714 cm À1 was observed that corresponded to the (-C = O) bond, and it is possible the increase of the hydroxyl region or hydroxyl index (HI) that appeared at 3460 cm À1 was due to a typical intermolecularly bonded -OH and free OH, which has been found by previous studies. 49,52egarding the PEC treatment, as shown in Figure 4A part (a) and in the magnified spectrum in Figure 4B, a significant increase in the CI was determined, from 0.006 to 0.08, indicating the generation of excited states in the HDPE MPs, leading to the cleavage of the polymer chains and the formation of (-C = O), (-C-O), and (-COOH) bonds. 63Moreover, an increase was highlighted, from 0.48 to 0.61, which facilitated the formation of new groups (-OH and/or -OOH), which were subject to further conversion of carbonyl groups (Figure 4C). 64This was confirmed by gravimetric (Figure S3) and thermogravimetric (Figures S4A and S4B) analysis when applying a current density of 6.89 mA cm À2 .The percentage degradation of MPs-PE was 58.0%G 0.1 and 89.9% G 0.1, by electrooxidation and PEC, respectively (Table 1).In contrast, thermogravimetric analysis of Figure S4A shows the change in polymer mass before and after treatment by PEC. 65The decomposition weight loss of HDPE-MPs occurred in a first stage at approximately 424 C, and had the highest weight loss at approximately 466 C with complete decomposition of the polymer microspheres, 66 (See also Tables S2 and S3; Figure S4A).In summary, the weight loss after HDPE MP degradation via PEC occurred at a lower temperature than for standard MPs owing to the presence of ketones/carboxylic acids in polymer chains, which are first to leave in thermal decomposition.

SEM
Based on the aforementioned tests, SEM analysis was performed to determine the change in MP morphology during the electrochemical oxidation and PEC processes.As seen in Figure 5A, the standard MP was a compact sphere with a smooth surface with little roughness and an approximate diameter of 250 mm.After electrochemical oxidation (Figure 5B), the MPs showed an amorphous spherical shape; the presence of surface cracks and cavities indicated that the degree of deterioration increased, as d OH could enter the surface of the HDPE-MPs at a greater depth. 67During PEC, visible changes in the surface microstructure occurred (Figure 5), which decomposed into hemispheres and fragments, possibly due to the interaction (collisions) between the BDD/TiO 2 photoanode and HDPE-MPs under UV light.Furthermore, a honeycomb-like porosity with a network of interconnected voids was visible (Figure 5C, scale 45 mm), indicating PEC achieved maximum degradation of the HDPE-MPs.

TOC and COD analysis
Total organic carbon (TOC) and chemical oxygen demand (COD) analyses were carried out on the remaining solution that comes from degradation process once the HDPE-MPs were filtered.Thus, TOC analysis insights into the breakdown and the transformation of organic  compounds associated with microplastic particles after the degradation process. 68In the same way, COD analysis measures the amount of oxygen required to oxidize the organic matter released by microplastics degradation and can be used to evaluate the organic pollution load. 69These analyses can provide useful information regarding degradation methods effectiveness and the extent to which MPs are breaking down into smaller compounds.Therefore, TOC and COD were carried out using three different current densities (Figure 6); the highest amount of organic matter from polymer chain cleavage corresponded to 6.89 mA cm À2 with 250 and 396.2 mg L À2 O 2 for EO and PEC, respectively.Using the same current values for COD, 633 and 1003 mg L À2 O 2 were obtained for EO and PEC, respectively, which were the largest amounts of oxygen from the treatment (See also Table S4).Namely, the formation of oxidized products, such as aldehydes and ketones in steps 4.1 and 4.2 (suggests that HDPE-MPs chains are breaking down.This highlights a transformative process in which the HDPE molecules suffer chemical changes leading to the generation of oxygen-containing functional groups.

d OH trapping study
Since photoelectrodegradation of pollutants is usually mediated by ROS, mainly d OH is important for all analysis. 70In this regard, the quantification of hydroxyl radicals though spectroscopy was used to determine the maximum level of HDPE-MPs degradation by this highly oxidizing reactive species.Figure 7A shows the d OH generated in the absence of HDPE-MPs at the current densities of 3.48, 6.89,  and 9.52 mA cm À2 during 15 min of PEC; the inset presents the same current densities as well as that of 0.57 mA cm À2 , which corresponds to a 1.8 V vs. Ag/AgCl and 60 min duration and was the reference initial potential of d OH production.These values were consistent with the Tafel analysis.The N, N-dimethyl-4-nitroso-aniline (RNO) disappearance rate follows first-order kinetics that is the maximum amount of d OH occurs at 6.89 mA cm À2 , with a kinetic constant of 0.143 min À1 .At 9.52 mA cm À2 , the reaction rate slows, and O 2 evolution is predominant, which does not favor the degradation process and implies a higher energy consumption (Figure 7B). 71,72he amount of d OH photoelectrogenerated in the presence and absence of HDPE-MPs was compared at a current density of 6.89 mA cm À2 .According to Figure 7C, with MPs, the d OH concentration decreased by 53% in 100 min compared to when MPs were absent, that is, a fraction of the generated d OH was used for MPs degradation.Conversely, when Tween 20 was evaluated with RNO, the d OH production deceased because a fraction was used for surfactant degradation, and it is possible that solution resistance also increased.

Proposed mechanism
Based on this study's analyses, we propose a possible degradation pathway of HDPE MPs (Figure 8).As the first step, the C-H bond of the HDPE molecule is broken by the d OH generated by PEC on the BDD/TiO 2 surface, causing the production of a free radical centered on the carbon atom of the HDPE molecule. 73Thereafter, each reaction step was computationally evaluated and the free energy change for each was estimated.Steps 1 and 2 involve the abstraction of the H from the polymer and the incorporation of the O 2 molecule, resulting in free energy changes of À21.3 and À60.2 kcal mol À1 , respectively.These negative values indicate highly favorable and exergonic reactions. 74In step 3, three reaction paths for the formation of the hydroperoxide species are suggested and analyzed.Computational results reveal that step 3.1 is not plausible, considering a free energy change of 38.4 kcal mol À1 . 75However, for steps 3.2 and 3.3, comparable values for the free energy change (approximately 17 kcal mol À1 ) were obtained, suggesting that both processes are favorable and may occur as parallel reactions.A free energy change of about 17 kcal mol À1 is reasonable for an oxidation process performed at room temperature.In step 4, four competitive reactions are proposed for the decomposition of the hydroperoxide species. 76Examining the free energy change values, it becomes evident that steps 4.1 and 4.2 represent the most feasible pathways, with values of À36.0 and À75.7 kcal mol À1 , respectively.These results indicate that the direct formation of aldehydes and ketones from the hydroperoxide species is more likely, and no homolytic dissociation of the hydroperoxide is necessary (as suggested in step 4.4). 77In detail, step 4.1 involves a beta cleavage leading to the formation of an aldehyde or ketone, with a subsequent release of a hydroxyl radical.Similarly, a beta-cleavage aided by another radical in the system is suggested for step 4.2.Table S5 lists the Cartesian coordinate data used in the modeling.

Conclusion
PEC using a BDD/TiO 2 photoanode is an efficient technology for the degradation of HDPE-MPs in aqueous media.Photoelectrocatalytic performance in an electrolyte medium of 0.1 mol L À1 Na 2 SO 4 and an optimum current density setting of 6.89 mA cm À2 during 10 h reaction achieved a high degradation efficiently (89.91 G 0.08).FTIR results showed that a partial oxidation of the microplastics leads to the formation of organic species with carbonyl groups, while SEM results show the polymer with an amorphous spherical shape with surface cracks and cavities.Furthermore, evaluating the production of d OH radicals at the optimum current density showed that the oxidation of the microplastics was mediated by this oxidizing species, which was supported by density functional theory (DFT) calculations of the proposed degradation mechanism.Ultimately, the photoelectrocatalytic efficiency with BBD/TiO 2 for the degradation of HDPE-MPs proved to be more efficient than electrochemical oxidation.

Figure 4 .
Figure 4. Performance degradation of high-density polyethylene microplastics on BDD/TiO 2 (A) FTIR spectra of HDPE-MPs before and after 10 h reaction by EO using bare BDD and PEC using BDD/TiO 2 photoanode.(B and C) Magnified spectra of HDPE-MPs after PEC treatment.

Figure 6 .
Figure 6.Analysis of the filtered solution of the samples treated at different current densities (A) Total organic carbon obtained (TOC).(B) Chemical oxygen demand (COD).

Figure 7 .
Figure 7. Analysis of d OH generation in the absence and presence of HDPE MPs (A) d OH at different current densities in absence of high-density polyethylene microplastics (HDPE MPs).(B) First-order reaction kinetics analysis of RNO (0.1 mol,L -1 Na 2 SO 4 ).(C) Effect of the presence of HDPE MPs and Tween 20 on the production of d OH.Data are represented as mean G SEM.

Figure 8 .
Figure 8. Proposed mechanism of degradation of high-density polyethylene microplastics by photoelectrocatalysis Values above the arrows are the free energy change of each case in kcal$mol -1 .