3.1. Morphology
The SEM images before and after photocatalysis for the PE and PS MPs are shown in Fig. 1. The original PE-FC-f had many granular bumps and folds as well as few pores on the surface. After 30 d of visible light irradiation, the PE-FC-f produced a more porous surface structure. The original PE-FC-s had a small number of folds on the surface structure and produced cracks and more folds after being irradiated, whereas the original PE-FC-a had folds and more debris on the surface structure and produced less debris and some folds and cracks after irradiation. The original PS-FC-f has a smoother surface structure with almost no cracks and bumps, which then became rougher and produced some surface cracks after irradiation. Compared to PS-FC-f, the PS MPs of PS-FC-s and PS-FC-a had a rougher surface and some folds before degradation. After 30 d of visible light irradiation, the surfaces of both PS-FC-s and PS-FC-a became more uneven and developed some cracks and uneven structures. The above results indicate that the MPs added with α-Fe2O3/g-C3N4 underwent structural changes under visible light irradiation, which were caused by the photodegradation of MPs. Visible light irradiation restructured the crystalline components of the MPs and removed the photodegradation by-products, leading to the formation of surface layer folds, cracks, and pores (Nabi et al., 2020; Zhu et al., 2019; Verma et al., 2017). The presence of these surface cracks and pores allowed oxygen to penetrate deeper into the MPs, which enhanced the extent of the subsequent photodegradation (Uheida et al., 2021).
3.2. Mass loss and kinetic analysis
The mass loss for the photocatalytic degradation of PE and PS MPs is shown in Fig. 2. After 30 d of visible light irradiation, there was almost no mass loss in PS-s or PS-a (Fig. 2a), while the mass loss rate of PS-f was 2.58%. There was also no significant change in the mass loss rates of PE-a or PE-s, and the mass loss of PE-f was 1.49% (Fig. 2b). Thus, PS and PE MPs were difficult to degrade without a photocatalyst. Compared to the treatment groups without the addition of photocatalyst, the mass loss rates of all PS and PE treatment groups increased significantly after the addition of α-Fe2O3/g-C3N4 photocatalyst. The highest mass loss rate was observed for PS-FC-f (9.94%), followed by PS-FC-a (7.81%) and PS-FC-s (1.41%), while PE-FC-f and PE-FC-a had mass loss rates of 3.48% and 3.75%, respectively, PE-FC-s had negligible mass loss rates. The films made of α-Fe2O3/g-C3N4 and MP particles and the composite system of MP particles with α-Fe2O3/g-C3N4 in water showed higher mass loss rates compared to the composite system of MP particles with α-Fe2O3/g-C3N4 exposed directly to visible light. Fitting the mass loss corves of both MPs to the pseudo-first-order kinetic equation revealed that the pseudo-first-order kinetic model can describe the mass loss of MPs due to photodegradation, with high correlation coefficients (R2) ranging from 0.980–0.998 for the six treatment groups with the addition of α-Fe2O3/g-C3N4, except for PE-s (0.789), and from 0.947–0.990 for the other treatment groups (Table 1). During the 30 d photocatalytic process, the mass loss of the plastic samples increased with the photocatalysis time.
Table 1
Parameters and determination coefficients (R2) of pseudo-first-order models used to describe MP mass loss.
Sample | Pseudo-first-order |
qe | k1 | R2 |
PE-f | -1.001 | -0.031 | 0.983 |
PE-s | -0.023 | -0.063 | 0.789 |
PE-a | -2.642 | -0.168 | 0.990 |
PE-FC-f | -949.095 | -1.213 | 0.987 |
PE-FC-s | 0.0139 | 0.134 | 0.989 |
PE-FC-a | 5.514 | 0.038 | 0.980 |
PS-f | 2.888 | 0.096 | 0.947 |
PS-s | 9.193 | 0.258 | 0.965 |
PS-a | 0.430 | 0.090 | 0.981 |
PS-FC-f | 9.682 | 0.358 | 0.998 |
PS-FC-s | 1.370 | 0.233 | 0.997 |
PS-FC-a | 8.397 | 0.096 | 0.997 |
Guadagno et al. (2001) showed that photocatalytic induction of free radicals in the polymerization chains of polymers is the main mechanism of polymer degradation. The photodegradation of polymers includes the photo-initiation, proliferation, and termination phases (Ter Halle et al., 2016). Photoinitiation occurs at the polymer surface, where chemical bonds in the polymer chains are broken to generate free radicals. In the proliferation phase, the polymer radicals react with oxygen to form peroxy radicals, while in the termination phase, the photodegradation of the polymer is terminated by the inert products generated by the radicals (Shang et al., 2003). The presence of a photocatalyst causes more complex radical reactions to occur with radicals generated on the polymer surface, resulting in the self-oxidization and chain breakage of the polymer. Therefore, PE and PS with the addition of α-Fe2O3/g-C3N4 exhibited relatively high mass loss rates compared to those without the photocatalyst (Fig. 2). PS and PE are difficult to photodegrade in the environment, whereas α-Fe2O3/g-C3N4 promotes the photodegradation of PS and PE. α-Fe2O3/g-C3N4 is more effective in photodegrading PS than PE, as the PE polymer backbone does not contain any unsaturated double bonds to facilitate photoinitiation. However, the main chain carbon of PS has a reactive C-H tertiary carbon bond and is more susceptible to photooxidative degradation and cleavage. PS undergoes a relatively rapid photodegradation process when exposed to oxygen and visible light. The benzene ring on the PS surface becomes excited and the excitation energy transfers to the nearest C-H bonds, resulting in the formation of free radicals and chain cleavage (Yousif and Haddad, 2013). Furthermore, the contact area between the polymer and photocatalyst is an important factor in the polymer degradation rate (Ter Halle et al., 2017). The contact area is greatly increased using a composite film consisting of MPs and photocatalysts, and the degradation of the MPs is accelerated. Similar to the present study results, Tian et al. (2019) found that photocatalytic degradation of MPs is more likely to occur in an aqueous environment. Hydroxyl radicals are mainly generated by water adsorbed on the surface of semiconductors and are generally used to carry out the photocatalytic degradation of polymers (Ariza-Tarazona et al., 2020). Therefore, plastics in water can be more readily degraded by photocatalysts.
It is worth noting that in most of the previous studies, UV lamps have been used as the light source for the photocatalytic degradation of MPs, and that UV irradiation resulted in a higher degree of degradation of MPs than visible light irradiation (Table 2). However, xenon lamps provide a light source that is more compatible with irradiation conditions in real environments (Andrade et al., 2019; Mao et al., 2020). A comparison revealed that the MP mass loss was lower in this study (Table 2). Researchers found that pH and temperature affect the degradation of photocatalysts (Ariza-Tarazona et al., 2020; Gan et al., 2022). A stronger light source better facilitates photocatalysis and the density of the MP also affects the photocatalytic degradation process (Liang et al., 2013; Lam et al., 2021). Therefore, the results of this study may be influenced by the experimental conditions as well as the degraded MPs themselves. In addition, optimized material preparation methods may also improve the photocatalytic effect of α-Fe2O3/g-C3N4 (Li et al., 2020; Wang et al., 2023). In a subsequent study, we will investigate the photocatalytic degradation effect under different experimental conditions and improve the preparation process of the photocatalytic materials to obtain better degradation results.
Table 2
Examples of photocatalyst-mediated MP degradation.
Material | Type | Light source | Weight percentage | Catalytic time | Weight loss | Reference |
TiO2 | PS | A 30 W UV-lamp (254nm) | 1.0 wt% | 300 h | 29% | Zan et al. (2004) |
TiO2/CuPc | PS | A 8 W UV-lamp (254nm) | 2.0 wt% | 250 h | 6.9% | Shang et al. (2003) |
Fe-TiO2 Ag-TiO2 Fe-Ag-TiO2 | PE | Two 6W UV-lamps (365 nm) | 1.0 wt% | 300 h | 13.49% 13.75% 14.34% | Asghar et al. (2011) |
PAM-g-TiO2 | LDPE | A 40 W UV-lamps (340 nm) | 5.0 wt% | 520 h | 39.85% | Liang et al. (2013) |
C, N-TiO2 | HDPE | A 50 W LED lamp (400–800 nm) | 0.4 w/v% | 50 h | 71.77% (pH = 3, 0°C) | Ariza-Tarazona et al. (2020) |
C, N-TiO2 | HDPE | A Slim LED IP65 50 W visible LED lamp | 0.4 w/v% | 50 h | 38% | Vital-Grappin et al. (2021) |
OMS-2 | PE | A 20 W UV-lamp (254 nm) | 1.0 wt% | 288 h | 16.5% | Liu et al. (2011) |
BiOCl | PE | A 250 W Xe lamp | 0.1 w/v% 1.0 w/v% | 5 h | 0.22% 5.38% | Jiang et al. (2021) |
Fe-ZnO | LDPE | Sunlight (271.6 W/m2) | 2.0 wt% | 120 h | 41.3% | Lam et al. (2021) |
α-Fe2O3/g-C3N4 | PE | a xenon lamp (35 W) | 0.2 w/v% 2.0 wt% | 720 h | 7.81% 9.94% | In this study |
PS | 0.2 w/v% 2.0 wt% | 3.75% 3.48% |
3.3. FTIR spectral properties
The ATR-FTIR spectra of the MPs before and after degradation are shown in Fig. 3. After 30 d of photocatalytic degradation under visible light irradiation, the characteristic tensile bands of the PE-FC-s and PE-FC-a treated groups increased significantly (Fig. 3d, e). The asymmetric stretching vibrations of carboxylic acids were observed around 1322–1409 cm− 1 and peaked around 1610 cm− 1. In addition, the stretching peak formed by the conjugation of C = O appeared around 1637 cm− 1 and peaked at about 1690 cm− 1. In comparison with the treatment of PE-FC-s, the carboxyl band of PE-FC-a was broader, which can be attributed to the enhanced photocatalytic effect of the increased production of radicals in aqueous environments. However, the diffraction peaks on the ATR-FTIR spectra of the PE-FC-f treatment group did not change significantly (Fig. 3a, b), indicating the lesser effect of α-Fe2O3/g-C3N4 on the photocatalytic degradation of the PE film compared to granular PE.
After 30 d of photocatalytic degradation under visible light irradiation, the ATR-FTIR spectrum of PS-FC-f showed changes in the peak intensities of different functional groups such as OH, OOH, C = O, and COOH (Fig. 3g). Among them, the formation and increased intensity of carbonyl (-C = O) stretching bands (1700–1740 cm− 1) and carbon-carbon double bond (-C = C-) stretching bands (1600–1680 cm− 1) in the polymer backbone chain indicated that a certain photodegradation reaction occurred on the PS-FC-f surface. In addition, the characteristic tensile bands of PS-FC-s and PS-FC-a noticeably increased after 30 d (Fig. 3i, 3j). The carbonyl signals generated by ketone carboxylic acids and ester functional groups were observed in the range of 1650–1850 cm− 1. Similar to PE, the carboxyl band of PS-FC-a was also wider than that of PS-FC-s, indicating that the α-Fe2O3/g-C3N4 photocatalyst more readily degraded MP particles in aqueous environments.
Carboxyl bands consisting of carboxyl products like ketones, carboxylic acids, and lactones were found in PE-FC-s, PE-FC-a, PS-FC-s, PS-FC-a, and PS-FC-f after 30 d of visible light irradiation. Peroxides formed on the surface of the MPs, which then decomposed into hydroxyl and alkoxy groups. The alkoxy groups can be further oxidized to form ketones, which can then be photolyzed into carboxyl groups such as carboxylic acids, esters, and lactones through the Norrish I and II reactions (Ter Halle et al., 2017). Alcohols, aldehydes, and ketones are the expected products of radical termination reactions. Due to their unsaturated double bonds, they are more susceptible to photodegradation by α-Fe2O3/g-C3N4, leading to the cleavage of polymer chains (Brandon et al., 2016). The enhancement of functional groups such as alcohols, aldehydes, and ketones, the spectral intensities of the stretching peaks, and the increased conjugation of -C = C- indicated that the PE and PS treatment groups, with the exception of PE-FC-f, underwent polymerization chain breaking under the photocatalytic action of α-Fe2O3/g-C3N4.
3.4. X-ray diffraction
Figure 4 shows the XRD patterns of PE and PS MPs before and after photocatalytic degradation. Before photocatalytic degradation, the crystalline peaks of PE-FC-f were significantly lower than those of PE-FC-s, while the difference between the crystalline peaks of PS-FC-f and PS-FC-s was not significant. The crystalline structure of the PE particles changed significantly after they were dissolved and recast into thin films, while the crystalline structure of the polymer crystals barely changed after the PS particles were recast into thin films. The crystallinity of PE-FC-s and PE-FC-a was significantly higher after 30 d of visible light irradiation, while the crystallinity of PE-FC-f did not change significantly. After 30 d of visible light irradiation, the crystallinity of PE-FC-f decreased by 2.57%, while the crystallinity of PE-FC-s and PE-FC-a increased from 33.01–42.57% and 50.02%, respectively. Compared to before degradation, the crystallinity of PS-FC-f increased slightly from 19.8–26.1% after 30 d of visible light irradiation, while the crystallinity of PS-FC-a increased significantly from 21.2–68.7% and the crystallinity of PS-FC-s barely changed from 21.2–23.1%.
The results show that the crystallinity of the MPs increased with the addition of α-Fe2O3/g-C3N4 via photocatalytic degradation, leading to further crystallization of the polymer surface due to the higher mobility of the amorphous fraction after photooxidation and chain breakage (Müller et al., 2018). The changes in surface crystallinity are mainly influenced by the post-crystallization process and the degradation of the amorphous phase. The degradation of the amorphous phase occurs, as the amorphous phase is an oxidation-sensitive structure on the polymer surface, while the post-crystallization process increases the crystallinity of the structure through the cleavage and rearrangement of the oxidative carbon chain. In the photocatalysis of polymers, the degradation of the amorphous phase and the increase in the crystalline phase may lead to an increase in surface crystallinity (Fayolle et al., 2008). Furthermore, the crystallinity of composites consisting of MP particles with α-Fe2O3/g-C3N4 after 30 d of visible light irradiation increased when placed in water. Therefore, photodegradation reactions such as the cleavage and rearrangement of the oxidative carbon chain and the degradation of the amorphous phase are more likely to occur when the composites are additionally placed in water than when they are simply mixed and directly exposed to visible light, resulting in significantly increased crystallinity on the surface of the MP particles.
3.5. Average molecular weight
Table 3 shows the average molecular weights of the treated MPs before and after degradation with the addition of photocatalysts. The average molecular weight (Mn) and weighted average molecular weight (Mw) of the PE groups were lower than those of the PS groups. After 30 d of visible light irradiation, the Mn and Mw of PE and PS MPs with α-Fe2O3/g-C3N4 decreased. Among them, PE-FC-f (17.8%, 5.30%), PE-FC-a (11.4%, 7.50%), PS-FC-s (11.2%, 6.80%), and PS-FC-a (11.5%, 8.72%) showed a greater decrease in Mn and Mw, followed by PE-FC-s (8.03%, 4.01%) and PS-FC-f (4.10%, 1.73%). The polymer dispersity index (PDI) was used to indicate the dispersibility index of polymers and describe the molecular weight distribution of polymers. After 30 d of exposure to visible light, the PDI increased in all treatments, indicating that the photocatalytic degradation of PE and PS occurred under visible light irradiation with the involvement of α-Fe2O3/g-C3N4, resulting in the broken polymer carbon chains.
In photooxidation, the polymer undergoes random or terminal chain cleavage, forming branching and various oxygen-containing functional groups that decrease the average molecular weight and molar mass of the polymer (Shang et al., 2003; Ter Halle et al., 2016; Ter Halle et al., 2017). The reduction rates of the Mn and Mw of PS-FC-s and PS-FC-a were greater than the corresponding PE structures, as the α-Fe2O3/g-C3N4 photocatalysts are more susceptible to the photocatalytic degradation of PS MP particles. However, the mean molecular weight change did not correspond to the weight change in the PE and PS partial treatment groups, which may be due to the benzene ring structure on the PS surface. Oxidizing radicals preferentially reacted with the benzene ring on the surface, protecting most of the substrate from further oxidation (Yousif and Haddad, 2013), resulting in a higher mean molecular weight reduction in PS-FC-f than in PE-FC-f. Both the Mn and Mw decreased less under visible light irradiation.
Table 3
Parameters calculated from the GPC data of the treatments of PE and PS.
Treatments | Mn (Da) | Mw (Da) | PDI (Mw/Mn) |
0d | 30d | 0d | 30d | 0d | 30d |
PE-FC-f | 19456 | 15986 | 54673 | 51774 | 2.810 | 3.239 |
PE-FC-s | 17669 | 16249 | 55701 | 53464 | 3.152 | 3.290 |
PE-FC-a | 17669 | 15644 | 55701 | 51523 | 3.152 | 3.293 |
PS-FC-f | 45524 | 43655 | 176072 | 173010 | 3.868 | 3.963 |
PS-FC-s | 47594 | 42253 | 181757 | 169380- | 3.819 | 4.009 |
PS-FC-a | 47594 | 41160 | 181757 | 165894 | 3.819 | 4.030 |