Retrieved MP sizes and their O/C molar ratios. Figure 1 shows an SEM image and EDX analysis of a sample of tiny plastic particles retrieved from the sea (sampling station: S1-B) that have a particle size of around 450 nm and an O/C molar ratio of 0.15. This low ratio suggests that the particles are composed of artificial plastic material. The particles may be some phytoplankton; however, their body is composed of polysaccharides such as cellulose, the O/C ratio of which is much high, e.g., 0.83. The results reveal that marine MP can be distinguished from the phytoplankton according to the O/C molar ratio. Figure 2(a) shows the relationship between the O/C molar ratio and the long diameter of MP samples retrieved from the sea. PE, PP, PS, polyvinylchloride (PVC), and polyethylene terephthalate (PET) are produced globally on a huge scale. The total amount of these commercially produced plastics constitutes around 80% of all thermoplastics23, 24. For example, PP and PE represent 22% and 23% of the Japanese manufacturing of resins (in 2018), respectively. Consequently, most marine MP is composed of PE, PP, PS, PVC, and PET25. As shown in Fig. 2(a), most of the MP samples show O/C molar ratios of < 0.4. The MP contained almost no PET with the 0.4 O/C molar ratio. Similarly, only a small proportion of the MP particles were identified as PVC with a high chlorine content. The O/C molar ratios indicate that the MP samples retrieved from the sea mainly composed of PE, PP and PS. Most of the MP particles exhibit diameters of < 20-µm long and O/C molar ratios of 0.1–0.2. These results thus provide an index of size and the degree of degradation (degree of oxidation) for modeling. Figure 2(b) shows the relationship between the AOP degradation time and the O/C molar ratio of a PP sample in distilled water. The ratios gradually increased with up down. This behavior can be attributed to repeated oxidation and peeling off15, 26. After 75 days, an O/C of 0.1 was achieved, indicating that it takes a long time using the AOP degradation method to achieve the same ratio for PP as that of marine MP. To investigate the effect that MPs have on marine organisms, it is thus necessary to develop a method to more quickly establish an MP model.
Inhibiting and accelerating effects on autoxidation. Salinity lowers the degradation level of polyolefins such as PP and PE17, 27, 28. The refractive index of seawater increases due to salinity, and the utilization rate of UV light decreases in the degradation27, 28. Wu et al. reported that aqueous Cl− acts as an inhibitor in the photooxidation of PP in seawater17. Figure 3 shows the transformation of radical species from SO4•− to OH• in seawater. In the case of PP photodegradation in seawater, Cl− reacts with OH• generated by solar irradiation and converts to ClOH•−, which is a less reactive molecule17. To avoid the inhibitory effect of Cl−, it is thus necessary to change the initiator of the oxidation degradation (autoxidation) reaction from OH• to another radical species. Considering the reactivity of radical species, SO4•− is suitable for autoxidation in seawater. As shown in Fig. 3, the SO4•− initiator is transformed by Cl− 18. A large amount of the SO4•− is converted into OH•, with some of the OH• converted into ClOH•−. However, since the production of ClOH•− requires re-reaction with Cl−, the residual amount of OH• increases. The initiation efficiency of autoxidation is greatly improved in seawater due to the reactivity of OH• being higher than that of SO4•−. Moreover, the reaction between Cl− and OH• also occurs and produces Cl•. As shown in Fig. 3, two Cl• atoms couple to produce Cl2, which then reacts with H2O and forms ClOH, with the equilibria of the two reactions dependent on pH19. Since the pH of seawater is around 8, the equilibrium is biased toward the less reactive ClO−, which suppresses the autoxidation of PP in seawater. It is noted here that as the SO4•− gradually converts to SO42−, the pH value of the K2S2O8 in seawater solution decreases from ca. 8 to ca. 3 by the time of daily exchange (see Materials and methods section). This procedure ensures that there is a bias in the ClOH-rich equilibrium19 for a period of time before the exchange of fresh K2S2O8 seawater solution. The ClOH has a longer lifetime18 and migrates deeply into the polymer matrix before dissociating into radicals and then initiates autoxidation19. The autoxidation proceeds from the PP interior as well as the surface, and the rate of MP formation is synergistically accelerated. The usage of SO4•− generates both OH• and ClOH, overcoming the inhibiting effect and accelerating the autoxidation process.
Comparison of fragmentation behavior. Figure 4 shows SEM images of PP degraded via AOP for 15 days in seawater and pure water. Numerous micro pits can be observed on the surface of the seawater sample, which are traces of chemi-crystallization that is related to the autoxidation process29, 30, showing that the PP degradation rapidly progresses. However, the surface of the pure water sample exhibits a lattice-like texture formed by cracks and is relatively smooth. The degree of degradation of the PP in pure water is thus considerably less than that of the seawater sample. Figure 5 shows SEM images and EDX analysis around the peeling-off location on the PP sample degraded via AOP for 15 days in seawater. Much microsized peeling-off marks can be observed, and the O/C molar ratios are around 0.17, 0.13, and 0.13 (see arrows in Fig. 5). These values are similar to those of the MP samples retrieved from the sea, indicating that a MP sample with the same degree of degradation can be prepared in a short degradation time of 15 days. Moreover, it was determined that the combination of seawater and K2S2O8 initiator promotes excellent accelerated degradation of the plastic.
In our previous study, PP film degradation tests were performed in water with a specific photocatalyst under visible light irradiation26 or using the same AOP treatment15. The results revealed that MP particles are generated by planar exfoliation via autoxidation in the presence of water. Therefore, it is believed that the size of the MP is dependent on the degree of degradation, i.e., the degradation time. Figure 6 shows the long diameter distributions of PP degraded via AOP over 9, 12, and 15 days. After 9, 12 and 15 days of degradation, 79, 136 and 279 MP particles were recovered by filtration, respectively, indicating that the size is dependent on the AOP degradation time. The size distribution narrows with increasing AOP degradation time, with a bias toward smaller sizes. As shown in Fig. 7, it is confirmed that nanosized PP particles are obtained over 15 days of AOP degradation. These results suggest that the size of the MP particles can be controlled according to the degradation time.