CET vol 99 DOI: 10.3303/CET2399115 Paper Received: 12 February 2023; Revised: 6 March 2023; Accepted: 20 April 2023 Please cite this article as: Murakami Y., Kiyosugi A., Sakashita R., Yoshinaga H., Saito Y., Shono A., 2023, Thermal Reformation of Polystyrene Using Metal Oxide as Redox Catalyst, Chemical Engineering Transactions, 99, 685-690 DOI:10.3303/CET2399115 A publication of The Italian Association of Chemical Engineering Online at www.cetjournal.it CHEMICAL ENGINEERING TRANSACTIONS VOL. 99, 2023 Thermal Reformation of Polystyrene Using Metal Oxide as Redox Catalyst Yuya Murakamia,*, Ayumiko Kiyosugia, Ryosuke Sakashitab, Hideo Yoshinagac, Yasukazu Saitod, Atsushi Shonoa aDepartment of Industrial Chemistry, Faculty of Engineering, Tokyo University of Science, 6-3-1, Nijuku, Katsushika-ku, Tokyo, Japan, 125-8585 bResearch & Development Division, Negami Chemical Industrial Co., Ltd., Ro22, Dorin-machi, Nomi-shi, Ishikawa, Japan cAko Laboratory, Taiyo Koko Co., Ltd., 1603-1 Higashioki, Nakahiro-aza, Ako, Hyogo, Japan dNew Energy Institute Co., Ltd., 2-15-5, Kiyosumi, Koto-ku, Tokyo, Japan y.murakami@ci.tus.ac.jp While plastic has been regarded as a useful and cost-effective material, there is growing global concern about its disposal. Chemical recycling presents a promising solution to this issue. This study explores the utilization of vanadium oxide as a redox catalyst to effectively decompose polystyrene into industrially useful CO. By heating polystyrene with V2O5 under an inert gas atmosphere, CO was successfully produced accompanied with CO2 as the primary by-product. Adding 5 wt% iron to V2O5 improved the selectivity of CO production without compromising gas yield. X-ray diffraction analysis indicated that V2O5 acted as an oxygen source and turned into V2O4 and V6O13 after the reaction. This process enables polymer reformation at lower temperatures than conventional methods, making it an energy-efficient chemical recycling strategy. Additionally, V2O4 and V6O13 generated during the process were easily oxidized to V2O5 through heating under atmospheric conditions. As both polymer reformation and oxidation are exothermic reactions, the proposed reaction scheme can be used as a thermally efficient chemical recycling process. 1. Introduction Plastic is extensively utilized in diverse fields due to its exceptional versatility and cost-effectiveness. However, in recent years, the issue of plastic waste has emerged as a pressing environmental challenge (Yang et al.), which requires efforts to reduce plastic production and disposal through chemical, mechanical, physical, biological, and thermal recycling (Shamsuyeva and Endres, 2021). Chemical recycling, in particular, is a highly anticipated and versatile approach, given its ability to generate other substances from the resulting products (Lee and Liew et al.) However, most plastic is currently processed via incineration or landfill (Prajapati et al., 2021). For instance, the recycling rate for polystyrene (PS), recognized as one of the most versatile plastics, remains at only 0.9 % in the USA (Rahimi and Garcia, 2017). An innovative recycling method for PS is necessary to solve this global problem. Most of the chemical recycling methods for PS reported so far have used high-temperature reformation reactions. For example, A. Demirbaş (2005) reported a chemical recycling method without a catalyst at around 400-500 °C to convert PS into a liquid hydrocarbon consisting mainly of styrene. Patrick et al. (2006) proposed a method to convert PS into styrene at high selectivity using a fluidized bed reactor at 520 °C, which was then further recycled into a biodegradable polymer. Achilias et al. (2007) demonstrated the reformation of PS into high-purity styrene at 510 °C using a fluidized bed reactor, which can be recycled as a raw material for PS. Park et al. (2020) showed that PS could be thermally decomposed into aromatic compounds at 780 °C and used as fuel. Although these studies all showed high selectivity and usefulness of the obtained products, the reaction temperature was relatively high, at 400 °C or higher. High-temperature processes have low thermal efficiency, making it difficult to downsize the process and resulting in high implementation barriers in terms of both initial and operating costs. Guest Editors: Sauro Pierucci, Flavio Manenti Copyright © 2023, AIDIC Servizi S.r.l. ISBN 978-88-95608-98-3; ISSN 2283-9216 685 In this study, we investigated polystyrene (PS) reformation at low temperatures (300-400 °C) in light of the growing concern over plastic waste. To efficiently decompose PS, we employed V2O5 as a redox catalyst to provide an appropriate supply of oxygen to PS and aimed to decompose it into gases mainly composed of CO and H2O. These gases can be converted to hydrogen via the water-gas shift reaction and are expected to be utilized for producing various materials through C1 chemistry. By lowering the operating temperature, we can reduce the barriers to introducing the process and expect to utilize it as a versatile chemical recycling process. 2. Experimental 2.1 Materials Polystyrene (PS, Mw 260,000) and vanadium oxide (V2O5, purity > 99 %) were purchased from Kanto Chemical Co., Inc., Japan. All chemicals were used without further purification. 2.2 Thermal reformation of polystyrene The mixture of PS and V2O5 (total of 0.55 g) was placed in the reactor made of an SUS tube (I.D. 19.9 mm, length 100.0 mm), and 0.15 MPa of N2 was introduced as an inert gas. The reactor was sealed and heated by a furnace and controller (SU12, Chino Corporation, Japan) from room temperature to operating temperature (300 to 400 °C), which takes 30 min. After leaving the reactor in the furnace for a certain period (30 to 150 min), the reactor was air-cooled to room temperature. After cooling, the gas inside the reactor was collected and analyzed by gas chromatography (8860 GC System, Agilent Technologies, Inc., USA). Since the amount of N2 in the reactor is known, the amount of product gas can be calculated. The remaining solid was analyzed by X- ray diffraction (XRD) to characterize the compound in the residue using MiniFlex II (Rigaku Corporation, Japan). The angle was scanned in the range of 2θ = 5 to 100° at the rate of 5° min-1. The weight change of the vanadium oxide during the reaction was measured by thermogravimetry (TG) analysis (TG/DTA 6200, Seiko Instruments, Inc., Japan). 30 mg of sample was placed in a sample pan and heated from room temperature until 350 °C at 10 °C min-1 under atmospheric conditions. After reaching 350 °C, the sample was kept at this condition for 5 hours. 3. Results and discussion 3.1 Thermal reformation of polystyrene using redox catalyst CO and CO2 generation from the PS/V2O5 mixture was confirmed at 360 – 400 °C by GC analysis. To evaluate the efficacy of thermal reformation, the carbon yield (XC) and CO selectivity (SCO) were calculated by the following equations: 𝑋C and 𝑆CO at various operating temperature and time were presented in Figure 1. Since the reactor was heated from room temperature to operating temperature during the first 30 min of operation, no significant gas production could be observed at this point (XC < 10 %). On the other hand, gasification drastically proceeded after the reactor reached the operating temperature (i.e., after 30 min). XC increased with temperature because of the high reaction rate. Notably, no hydrocarbon or hydrogen was confirmed at any conditions, and therefore, the main product of the reaction should be CO, CO2, and H2O. Since the gas was sampled at room temperature in this work, the amount of water generated by the reaction could not be measured. At the beginning of the reaction (30 min), SCO was sufficiently high (40 %), although it significantly decreased with increasing operating time. For example, 95 % of the product was CO2 when the reformation was conducted at 400 °C for 150 min. Before the reaction, the mixture was orange, and it turned dark blue (or black) after the reaction, suggesting that V2O5 (orange color) changed into another compound during the reaction. The residue was analyzed by XRD to confirm the loss of V2O5 during the reaction (Figure 2a). As expected, the sharp peaks of V2O5 could not be seen in the reaction residue, and V2O4 seems to be the most dominant product. From the observation above, we concluded that the following two reactions mainly occur during the process: 𝑋C = 𝑛C, gas 𝑛C, polymer , (1) 𝑆CO = 𝑛CO 𝑛CO + 𝑛CO2 . (2) (C8H8)n + 12𝑛V2O5 → 8𝑛CO + 4𝑛H2O + 12𝑛V2O4, (3) 686 Figure 1: The carbon yield (a) and CO selectivity (b) of the reaction at various operating temperature and time Figure 2: The XRD patterns: (a) Reaction residue (V2O5 + PS, at 380 OC for 150 min), (b) Reaction residue (V2O5 + Fe + PS, at 380 °C for 150 min), (c) pristine V2O5, (d) V2O5 after recycling. Note that the V2O5/PS ratio was determined so that Eq(3) completes without any reactants left; therefore, 1.5 mol of V2O5 was added per 1 mol of carbon atoms in PS. If SCO is small, an insufficient O atom exists in the system to gasify carbon; thus, XC did not surpass 60 %. The ratio of V2O5 used during the reaction can be calculated as the following equation, under the assumption that the selectivity of Eq(3) over Eq(4) is precisely SCO: The calculated results were summarized in Table 1, along with the experimental results. As can be seen, most of V2O5 was consumed (XV2O5 ~ 90 %) at high temperature and a long operating time. In these conditions, most of V2O5 changed into V2O4, as suggested by XRD analysis (Figure 2a). The effect of temperature on the reaction was investigated in the range of 300 to 400 °C; the results are given in Figure 3. XC was clearly increased with temperature, while SCO decreased. Since the mixture of CO and H2O obtained in this reaction is variable in the engineering field, the ideal situation is that both XC and SCO should be sufficiently high. To understand the trade-off effect between XC and SCO, the following CO yield was defined: The value is given in Table 1. Interestingly, XCO decreases with operating time at 380 and 400 °C, suggesting that CO further reacted and changed into another compound during the process. Since only CO, CO2, and H2O were obtained in the process, the following reaction seems to proceed: (C8H8)n + 20𝑛V2O5 → 8𝑛CO2 + 4𝑛H2O + 20𝑛V2O4. (4) 𝑋V2O5 = 𝑋C ∙ (2 − 𝑆CO). (5) 𝑋CO = 𝑋C ∙ 𝑆CO. (6) 687 The results suggest for Eq(7) to proceed faster than Eq(3) at high temperatures. In order to establish a feasible process, the reaction Eq(7) should be suppressed. Table 1: CO gasification results with various operating temperature and time T / °C t / min XC / % SCO / % XV2O5 / % XCO / % 360 30 1.3 35.4 2.2 0.5 380 30 1.7 40.7 2.6 0.7 400 30 5.9 39.1 9.4 2.3 360 45 11.9 30.7 20.1 3.6 380 45 24.7 31.5 41.6 7.8 400 45 35.5 26.8 61.5 9.5 300 150 3.4 37.7 5.4 1.3 310 150 5.2 24.9 9.1 1.3 320 150 13.9 24.3 24.5 3.4 330 150 21.9 22.8 38.7 5.0 340 150 27.0 24.9 47.2 6.7 350 150 49.5 21.2 88.4 10.5 360 150 40.5 21.2 72.4 8.6 380 150 47.3 11.3 89.2 5.4 400 150 55.3 5.7 > 100.0 3.2 Figure 3: The carbon yield (a) and CO selectivity (b) of the reaction at various operating temperature with pure V2O5 or V2O5/Fe mixture as redox catalyst (operation time was fixed at 150 min.) In order to produce CO efficiently, V2O5/PS ratio was varied. The scaled weight ratio (w* V2O5/PS) was defined as follows: The scale factor ω was set to 0.0477. When w* V2O5/PS = 1, Eq(3) completes without any reactants left, while an excess amount of V2O5 exists when w* V2O5/PS > 1. The effect of w* V2O5/PS on XC and SCO (operating time = 45 min) was given in Figure 4. As can be seen, XC drastically increases when an excess amount of V2O5 exists in the system. Intriguingly, SCO was almost constant regardless of w* V2O5/PS, suggesting that the amount of O atoms and CO molecules does not affect the reaction rate of Eq(7). Therefore, Eq(7) seems to have a zero-order reaction rate in terms of the fugacity of O and CO. It should be noted that this result may be attributed to the batch-wise operation during the process. Therefore, the reaction kinetics should be carefully examined using a (semi-) continuous system to obtain a reliable conclusion. To further facilitate the production of CO, Fe was added to V2O5 as a co-catalyst by 5.0 wt%. While the existence of Fe did not affect XC, SCO increased at all tested operating temperature, as demonstrated in Figure 3. Since Fe only affects SCO, it is an excellent candidate to suppress Eq(7). The reaction residue was analyzed by XRD to investigate the effect of Fe (Figure 2b). As can be seen, the main product of the reaction was changed from CO + V2O5 → CO2 + V2O4. (7) 𝑤V2O5/PS ∗ = ω 𝑤V2O5 𝑤PS . (8) 688 V2O4 to V6O13 when Fe was added. The result implies that the existence of Fe changes the reactivity of vanadium oxide with CO and reduces the reaction rate of Eq(7). Figure 4: The carbon yield (a) and CO selectivity (b) of the reaction at various V2O5/PS ratios and temperature (operation time was fixed at 45 min.) 3.2 Chemical recycling of PS using redox catalyst In the previous subsection, it was demonstrated that CO/H2O mixture could be obtained by the thermal reformation of PS. Another advantage of V2O5 is that the residue (V2O4 and V6O13) can be easily reversed to V2O5 at mild conditions by the following reaction: The weight change of residue over time was measured at 350 °C by TG under an atmospheric condition, and the result is shown in Figure 5. The weight clearly increases with time, confirming the reaction of V2O4 and V6O13 with O2 in the atmosphere. Moreover, the XRD pattern of the remaining solid indicates the complete recovery of V2O5 (Figures 2c and 2d). These results suggest that V2O5 can be a redox catalyst for the system. Figure 5: Weight change of vanadium oxide at 350 °C under an atmospheric condition. PS reformation using V2O5/V2O4 as a redox catalyst is thermodynamically beneficial because the main reactions in the process are exothermic: ΔH = -54.5 kJ/mol-C for Eq(3) and ΔH = -247.0 kJ/mol-O for Eq(9). Therefore, with proper process design, no heat is to be added to this chemical recycling process during steady state. Moreover, the operating temperature is much lower than the conventional processes, thus, making energy efficient. These advantages make the proposed reaction system an effective chemical recycling method for PS. 4. Conclusions The increasing severity of environmental problems caused by plastics demands a drastic solution, and developing new recycling methods is crucial. In this study, we focused on the chemical recycling of plastics and 2V2O4 + O2 → 2V2O5. (9) 689 explored a novel CO-generation process using a redox catalyst. This work demonstrated that heating a mixture of polystyrene and vanadium oxide can effectively gasify plastics at lower temperatures than conventional methods. The produced CO has potential use in various industrial applications, while mitigating by-product generation, CO2, is identified as a challenge for this process. Our investigations into reaction temperature and time reveal a trade-off between total gas production and CO selectivity, as high temperature and CO partial pressure lead to the oxidation of CO into CO2. We successfully increased total gas production by introducing excess redox catalyst while maintaining CO selectivity. To enhance CO selectivity, we conducted a thermal reformation using 5.0 wt% of Fe as a co-catalyst, which successfully improved CO selectivity without compromising total gas production. XRD analysis of the residual redox catalyst confirmed an increase in V6O13 production, suggesting that Fe affects surrounding V2O5 and enhances polymer degradation in the system. Notably, the redox catalyst after the reaction (V2O4 and V6O13) was regenerated to V2O5 by heating in the presence of oxygen at the reaction temperature. The exothermic nature of CO production and catalyst regeneration suggests that this system presents a novel approach to producing CO without the addition of heat energy during steady state. Our findings suggest that this system could be a superior and cost-effective method for the chemical recycling of plastics in terms of initial and running costs. Nomenclature nC,gas – quantity of carbon atoms in the gas, mol nCO – quantity of CO in gas, mol nCO2 – quantity of CO2 in gas, mol nC,polymer – quantity of carbon atom in polymer, mol SCO – CO selectivity, % wPS – weight of PS, % wV2O5 – weight of V2O5, % w* V2O5/PS – scale weight rate of V2O5/PS, % XC – carbon yield, % XCO – CO yield, % XV2O5 – fraction of consumed V2O5, % XC – carbon conversion, % ΔH – enthalpy of reaction, kJ/mol ω – scaling factor of w* V2O5/PS, - References Achilias D.S., Kanellopoulou I., Megalokonomos P., Antonakou E., Lappas A.A., 2007, Chemical recycling of polystyrene by pyrolysis: Potential use of the liquid product for the reproduction of polymer, Macromolecular Materials and Engineering, 292, 923–934. Demirbaş A., 2005, Recovery of chemicals and gasoline-range fuels from plastic wastes via pyrolysis, Energy Sources, 27:14, 1313–1319. Lee A., Liew M.S., 2021, Tertiary recycling of plastics waste: an analysis of feedstock, chemical and biological degradation methods, Journal of Material Cycles and Waste Management, 23, 32–43. Park K.B., Jeong Y.S., Guzelciftci B., Kim J.S., 2020, Two-stage pyrolysis of polystyrene: Pyrolysis oil as a source of fuels or benzene, toluene, ethylbenzene, and xylenes, Applied Energy, 259, 114240. Prajapati R., Kohli K., Maity S.K., Sharma B.K., 2021, Potential Chemicals from Plastic Wastes, Molecules, 26, 3175–3196. Rahimi A., Garcia J.M., 2017, Chemical recycling of waste plastics for new materials production, Nature Reviews, 1, 46. Shamsuyeva M., Endres H.J., 2021, Plastics in the context of the circular economy and sustainable plastics recycling: Comprehensive review on research development, standardization and market, Composites Part C: Open Access, 6, 100168. Ward P.G., Goff M., Donner M., Kaminsky W. O’Connor K.E., 2006, A two step chemo-biotechnological conversion of polystyrene to a biodegradable thermoplastic, Environmental Science and Technology, 40, 2433–2437. Yang L., Gao J., Liu Y., Zhuang G., Peng X., Wu W.M., Zhuang X., 2021, Biodegradation of expanded polystyrene and low-density polyethylene foams in larvae of Tenebrio molitor Linnaeus (Coleoptera: Tenebrionidae): Broad versus limited extent depolymerization and microbe-dependence versus independence, Chemosphere, 262, 127818. 690 217Murakami-PAGATO.pdf Thermal Reformation of Polystyrene Using Metal Oxide as Redox Catalyst