Multivalent cobalt nanoparticles supported on silica for efficient and sustainable methanolysis of commercial polyethylene terephthalate waste bottles†

Kamlesh Kumari , Priyanka Choudhary and Venkata Krishnan *
School of Chemical Sciences and Advanced Materials Research Center, Indian Institute of Technology Mandi, Kamand, Mandi 175075, Himachal Pradesh, India. E-mail: vkn@iitmandi.ac.in

First published on 2nd August 2024

---

---

1. Introduction

Plastics have now become an important part of our daily life and our reliance on this wonder material is ever increasing. Every year, almost 380 million tons of plastic is manufactured and used in every industry, including packaging, construction, textile, electronics and so forth, which results in the generation of humungous amounts of plastic waste.¹ Commonly, polyethylene (PE),² polyethylene terephthalate (PET),³ polypropylene (PP),⁴ polyvinyl chloride (PVC)⁵ and polystyrene (PS)⁶ are the main components of plastic waste. With an annual production of about 70 million tons, polyethylene terephthalate (PET), is a significant class of plastics that finds widespread application in a variety of applications like in the production of containers, packaging materials, and textiles.⁷ PET produces a lot of plastic waste since it does not break down naturally and has a low recycling rate, particularly if it is colored or mixed with other plastics.⁸ To deal with this issue, intense work is being done to create efficient upcycling or recycling techniques for waste and contaminated plastics. Currently, mechanical recycling is the primary method used to recycle post-consumer PET.^9,10 Nevertheless, this may deteriorate the material's performance and structure and the resultant items produced are of low value. The process of mechanical recycling also generates microplastic waste.^11,12 As compared to mechanical recycling, chemical upcycling of PET is of great importance as it converts the polymer into some value-added products.^13–15 Common methods used for the chemical upcycling of PET waste include glycolysis,^16–18 hydrolysis (neutral, acidic, alkaline)^19–21 and methanolysis.²² Glycolysis is the process of chemical upcycling where the PET waste is treated with ethylene glycol to form bis (2-hydroxyethyl) terephthalate (BHET).²³ In the hydrolysis process, the PET waste is treated with pure water or steam to produce terephthalic acid (TPA) and ethylene glycol under relatively high temperature (190–400 °C) and pressure (1–35 MPa) conditions.²⁴ The primary benefit of neutral hydrolysis of PET is that it is an environmentally benign process with no directly associated pollution, but the main disadvantage is that the impurities present in PET remain in TPA even after the completion of the process. Furthermore, strong acids or strong bases are used as solvents in both acid and alkaline hydrolysis, which can seriously corrode the equipment and generate a lot of waste during the post-processing steps. Due to these shortcomings in the present chemical upcycling procedures, colored PET bottles and PET fabrics are not considered sufficiently recyclable to be converted back into premium grade PET.²⁵ Conversely, high purity DMT can be obtained via alcoholysis of PET with methanol since DMT has a lower boiling point (284 °C) than BHET (446 °C), which facilitates its isolation and refinement. Therefore, it is important to create a sustainable and efficient process for upcycling PET waste.

Research on chemical upcycling of PET has increased because of recent developments and widespread attention towards plastic waste management. Hongkailers et al.²⁶ developed an alternate method for recycling PET trash through hydrodeoxygenation utilizing a Co/TiO₂ catalyst which involves transforming polyethylene terephthalate (PET) into arenes in the presence of hydrogen with 79% yield. Methanolysis of waste PET is performed at a wide range of temperatures and pressures to produce dimethyl terephthalate (DMT) and ethylene glycol. DMT can be easily refined and converted into other feedstocks for the manufacturing of PET or it can be hydrogenated to produce higher value chemicals like dimethyl cyclohexane dicarboxylate (DMCD),^27,28 and p-xylene.²⁹ Sako et al. reported the depolymerization of PET using supercritical methanol with 94% DMT yield at 400 °C and 40 MPa pressure.³⁰ Kurokawa et al. reported the methanolysis of PET using an aluminum triisopropoxide (AIP) catalyst at 200 °C using a mixture of toluene and methanol as a solvent which yields 88% and 87% of DMT and ethylene glycol, respectively.³¹ Use of ionic liquid as a cosolvent for the alcoholysis of PET with isooctyl alcohol at 200 °C for 5 h to produce 93.1% dioctyl terephthalate by using zinc acetate as a catalyst was reported by Chen et al.³² A report of Mishra et al. presented the use of ZnC₄H₆O₄ and Pb(C₂H₃O₂)₂ as catalysts to produce 97.8% and 100% DMT yield, respectively, at a temperature range of 120 °C–140 °C.³³ Similarly, Lalhmangaihzuala et al. reported synthesized orange peel ash coated Fe₃O₄ nanoparticles for the depolymerization of PET at 200 °C.³⁴ Tang et al. showed the methanolysis of PET using a MgO/NaY catalyst where the DMT and ethylene glycol yield of 99% and 91% was obtained at 200 °C.³⁵ Nevertheless, the majority of metal inorganic and metallic oxide catalysts suffer from inadequate stability or insufficient catalytic activity. Ionic liquids show good catalytic activity but are difficult to separate from the reaction mixture which ultimately reduces the products' purity. Thus, there exists an essential and urgent need to design a catalyst with high stability, high catalytic activity, and simple separation.

The above-mentioned reports suggest that to obtain a high yield of DMT, the reactions were performed under very high temperature and pressure conditions. Therefore, to shorten the reaction time, and to lower the temperature and pressure conditions, metal-based catalysts have gained a lot of attention recently. In this work, highly efficient heterogeneous Co–SiO₂ catalysts with different cobalt loadings having high surface area and high catalytic activity were synthesized and tested for the methanolysis of PET under mild conditions. Herein, multivalent Co nanoparticles were synthesized and supported on SiO₂ using a facile hydrothermal method. As it is well known, transition metal catalysts are essential components of nearly every catalysis application and are considered as an affordable alternative to expensive metal-based catalysts. Among them, cobalt is extensively used in almost all fields of catalysis. The partially filled d-orbitals of cobalt provides an optimum interaction to the reactant arriving on the surface of the catalyst by making bonds which are neither too strong nor too weak to facilitate the easy adsorption of reactants and desorption of products. In addition to this, cobalt exhibits multiple valence states (elemental, Co²⁺, Co³⁺) which facilitate easier composite formation with supports. Here, SiO₂ was chosen as a support due to its high surface area, cost effectiveness, ease of availability and porosity. To confirm the successful synthesis of the catalyst, various characterization techniques were used. Subsequently, the catalyst was used for the methanolysis of PET. Detailed optimization studies of the reaction were carried out using Co–SiO₂ catalysts and it was observed that the best Co–SiO₂ catalyst gives 100% PET conversion and 97% DMT yield. Another important objective of this work was to upcycle waste PET under greener reaction conditions. The reactions were carried out using a minimum amount of catalyst, mild temperature and pressure conditions. In addition, a gram scale reaction was also performed using 5 g of PET to check its potential industrial applicability. Furthermore, the recyclability studies show the good stability and high reactivity of the catalyst up to five cycles. Therefore, Co–SiO₂ being a non-noble metal based heterogeneous catalyst provides a sustainable and greener approach for performing methanolysis of PET.

2. Experimental section

2.1. Chemicals

Cobalt(II) chloride hexahydrate (99%, SRL), hydrazine hydrate (80%, SRL), ethanol (≥99.5%, Merck), sodium hydroxide pellets (98%, SRL), silicon dioxide (extra pure, Loba Chemie), methanol (Rankem), ethyl acetate (Merck), and CDCl₃ (99.80% D, Eurisotop) were purchased and used without any further purification. A double-stage water purifier (ELGA PURELAB Option-R7) was used to obtain deionized (DI) water (18.2 MΩ cm).

2.2. Material preparation

2.3. Catalytic activity studies

Methanolysis of PET was performed to test the catalytic activity of the as-synthesized Co–SiO₂ catalysts. Commercially available post-consumption plastic bottles were washed with DI water and dried at 60 °C overnight and were cut into small pieces for the methanolysis process. The methanolysis reaction was performed using 1 g of PET, 100 mg of Co–SiO₂ catalyst and 7.5 mL of methanol at 170 °C for 6 h in a hydrothermal reactor to obtain DMT. After 6 h, the reaction was allowed to cool down to room temperature. During the cooling process, the DMT monomers get precipitated. The obtained product was dissolved by adding ethyl acetate to the reaction mixture. The catalyst was then separated from the reaction mixture by centrifugation and the reaction mixture was dried using a rotary evaporator forming white crystals of DMT. The product formed was further characterized using ¹H and ¹³C NMR spectroscopy. The yield and conversion percentage were calculated by using the formula presented in section S2 (refer to the ESI†) and the NMR data obtained are shown in S4 (refer to the ESI†).

2.4. Compound characterization

Dimethyl terephthalate.³⁷ White solid, 97%; ¹H NMR (500 MHz, CDCl₃) δ (ppm) 8.08 (s, 4H), 3.93 (s, 6H). ¹³C NMR (125 MHz, CDCl₃) δ (ppm) 166.4, 133.9, 129.6, 52.6, 52.5.

3. Results and discussion

3.1. Synthesis and characterization

The cobalt nanoparticles (Co NPs) were synthesized using CoCl₂·6H₂O as a precursor and hydrazine hydrate as a reducing agent. In order to optimize the amount of hydrazine hydrate, control reactions were performed by varying the amount of hydrazine hydrate (5 mL, 10 mL, 15 mL, 20 mL) and it was observed that 20 mL of hydrazine hydrate was the optimum amount required to convert Co(OH)₂ to metallic cobalt as evidenced from the structural analysis (refer to ESI† Fig. S1). After the successful synthesis of Co NPs, the Co NPs were supported on SiO₂via a hydrothermal method as shown in Scheme 1 and were characterized using various techniques to confirm the successful synthesis of Co NPs and Co–SiO₂.

To examine the crystal structure of the catalysts, powder X-ray diffraction (PXRD) measurements were carried out. Fig. 1a shows the characteristic PXRD patterns for Co NPs, SiO₂ and Co–SiO₂. The diffraction peaks (2θ) at 41.74°, 44.44°, 47.44°, 75.88°, and 84.06° correspond to the (100), (002), (101), (110) and (103) facets of metallic cobalt in the hexagonal phase (P63/mmc, a = b = 2.50 Å, c = 4.06 Å, JCPDS no. 00-001-1278).³⁸ The weak diffraction peak (2θ) at 62.70° corresponds to the (103) facet of the hexagonal phase of CoO (P63/mc, a = b = 3.21 Å, c = 5.24 Å, JCPDS no. 00-089-2803).³⁸ The broad peak located at 2θ = 22.68° corresponds to amorphous SiO₂.³⁹ After the incorporation of Co NPs onto the surface of SiO₂, Co–SiO₂ shows a broad diffraction peak for SiO₂ at 22.68° and all peaks for hcp-Co at 41.74°, 44.44°, 47.44°, 75.88°, and 84.06° and hcp-CoO at 62.70°. The XRD pattern with different cobalt loadings of the Co–SiO₂ catalyst are shown in Fig. S2 (refer to the ESI†). All the four catalysts show a diffraction peak at 2θ = 22.68° which corresponds to the amorphous silica and the peaks at 41.74°, 44.44°, 47.44°, 75.88°, and 84.06° correspond to the hexagonal phase of cobalt and 62.70° for hcp CoO; all of them remain the same after the incorporation of Co NPs onto the surface of SiO₂. It is also observed that upon increasing the amount of Co NP loading on silica from 5% to 20%, the intensity of cobalt peaks has also increased. Fig. 1b shows the FTIR spectra of SiO₂ and Co–SiO₂ catalysts. The peaks between 1300 and 400 cm⁻¹ clearly show the presence of Si–O–Si and Si–OH vibration modes.⁴⁰ The peaks at 1084 cm⁻¹ and 462 cm⁻¹ signifies the presence of asymmetric and symmetric stretching vibration modes of Si–O–Si bonds, respectively, and the peak at 796 cm⁻¹ corresponds to the Si–OH vibration mode of SiO₂ and Co–SiO₂ catalysts.⁴¹ The thermal stability of the as-synthesized catalysts was investigated using TGA analysis and the obtained results are shown in Fig. S3 (refer to the ESI†). Co NPs show good thermal stability up to 400 °C. The initial weight loss at 200 °C is due to the physisorbed water, after which a slight increase in weight is observed which is attributed to the oxidation of metallic cobalt which was confirmed by examining the residue collected from TGA using PXRD analysis. The PXRD plot of the TGA residue is shown in Fig. S4 (refer to the ESI†) which confirms the formation of CoO (Fm3m, a = b = c = 4.25 Å, JCPDS no. 01-074-2392) due to the oxidation of cobalt at higher temperature.⁴² The TGA plot of SiO₂ shows initial weight loss up to 100 °C, which is due to the evaporation of adsorbed water molecules. Furthermore, Co–SiO₂ also shows high thermal stability up to 550 °C, after that the weight gain is observed which is due to the oxidation of cobalt content present in the Co–SiO₂ catalyst. The weight loss for Co–SiO₂ starting from room temperature is due to the dehydration of physisorbed water from the surface of SiO₂. It is observed that the oxidation temperature of Co NPs in Co–SiO₂ is substantially higher than that of the pure Co NPs. This suggests that SiO₂ significantly enhances the thermal stability of Co NPs.

Using scanning electron microscopy (SEM) and transmission emission spectroscopy (TEM), the morphology of the as-synthesized catalyst was examined. The SEM images of Co–SiO₂ show an aggregate-like morphology as depicted in Fig. 2a–c. The TEM images of Co–SiO₂ are shown in Fig. 2d–g which show the particle-like morphology of SiO₂ and the distribution of Co NPs over the surface of SiO₂. It was observed that the supported Co NPs were present in the form of agglomerated nanoparticles as no capping agent was used in the synthesis procedure. The lattice fringe spacing measurements are shown in Fig. 2g which reveals a d-spacing of 0.23 nm corresponding to the (100) lattice plane, which is in close agreement with the PXRD studies. Fig. 2h and i show the inverse fast Fourier transform (IFFT) image and line profile image of Co–SiO₂ which were used to calculate the lattice fringes for Co NPs present on the surface of SiO₂, which also indicates a d-spacing value of 0.23 nm. The average particle sizes of Co NPs before and after their deposition onto the surface of SiO₂ were calculated to be 25.30 nm and 25.67 nm, respectively, as shown in Fig. S5 (refer to the ESI†) which are similar.

XPS studies were carried out to examine the chemical species present on the surface of the catalyst. The atomic composition of the constituent elements present in the catalyst are shown in Table S1 (refer to the ESI†). The survey spectra of XPS (Fig. 3a) clearly show the presence of Co, Si, and O elements in the sample. High resolution spectra of cobalt, silicon and oxygen were investigated to know the chemical states of all the elements. The binding energy measured for each element has been corrected using the adventitious C 1s peak at 284.8 eV as a reference and the C 1s spectra for Co NP, SiO₂ and Co–SiO₂ are provided in Fig. S6 (refer to the ESI†). High resolution Co 2p spectra of Co NPs are split into two spin-orbit coupled state peaks, which are Co 2p_3/2 and Co 2p_1/2 (as shown in Fig. 3b). The peak at 778.2 eV is assigned to the Co(0) 2p_3/2 state and the peak at 793.1 eV is assigned to the Co(0) 2p_1/2 state which matches well with the results of PXRD analysis.^43,44 The peaks at 779.4 eV and 794.2 eV are attributed to the Co(III) state whereas the peaks at 782.1 eV and 796.8 eV are assigned to the Co(II) state.⁴⁵ The two satellite peaks are observed at 786.5 eV and 802.6 eV. High resolution Co 2p spectra of Co–SiO₂ also split into two spin-orbit coupled state peaks of Co 2p_3/2 and Co 2p_1/2. The peaks at 778.3 eV and 793.2 eV are assigned to the Co(0) 2p_3/2 and Co(0) 2p_1/2 state, respectively. The peaks at 781.2 eV and 796.7 eV are attributed to the Co(III) state whereas the peaks at 785.1 eV and 801.1 eV are assigned to the Co(II) state and the two satellite peaks are also observed at 788.6 eV and 804.2 eV. After the incorporation of Co NPs onto the surface of silica, the peaks of Co 2p spectra of Co–SiO₂ are shifted to a higher binding energy. The electron deficient environment of Co atoms supported on silica surface accounts for the peak shift to a higher binding energy. For further clarification, the peak table showing different peaks of Co 2p in Co NPs and Co–SiO₂ catalysts is presented in Table S2 (refer to the ESI†). The Si 2p high resolution spectra for SiO₂ and Co–SiO₂ are shown in Fig. 3c. A broad peak at a binding energy of 103.3 eV is observed in the case of SiO₂, which takes into consideration the two spin-orbit coupled peaks, which are Si 2p_3/2 and Si 2p_1/2.⁴⁶ The obtained spectrum can be fitted into a single peak as the coupled peaks could not be unequivocally separated due to their close proximity. With the incorporation of cobalt onto the surface of SiO₂ the Si 2p peak is slightly shifted to a higher binding energy i.e., 103.6 eV. This could be due to the electronic interaction of Co NPs with SiO₂.⁴⁷ The O 1s spectra of SiO₂ are deconvoluted into two peaks having binding energies of 532.3 eV and 532.9 eV (Fig. 3d). The peak at 532.3 eV is attributed to the surface lattice oxygen species and the peak at 532.9 eV is assigned to the surface adsorbed oxygen species.⁴⁸ With the incorporation of cobalt onto the surface of SiO₂, the O 1s peak shifts towards higher binding energy values. This might be due to the change in the bonding environment of the oxygen atom as the cobalt interacts with the outer surface layer of the SiO₂ support.

The Brunauer–Emmett–Teller (BET) measurements were performed to determine the surface area and pore size of the developed catalysts and the obtained results are shown in Fig. 4. In the case of SiO₂, the isotherm can be classified as a type-IV isotherm having an H3 hysteresis loop as per the International Union of Pure and Applied Chemistry (IUPAC) classification.^45,49 The Barrett–Joyner–Halenda (BJH) plot of the material describes its pore size distribution, and the linear fitting of multipoint BET data provides the surface area of the material. SiO₂ shows a surface area of about 170.5 m² g⁻¹. Co–SiO₂ also shows a type-IV isotherm having an H3 hysteresis loop which indicates that the SiO₂ structure is retained after the incorporation of cobalt onto its surface. The BJH plots showing the average pore size distribution of SiO₂ and Co–SiO₂ were within the range of 2–25 nm which signifies the presence of mesopores in the catalysts. The multipoint BET of Co–SiO₂ obtained from the linear fitting of nitrogen adsorption–desorption data and its surface area turned out to be 127.1 m² g⁻¹. The surface area of Co–SiO₂ is decreased as compared to the support, SiO₂, due to the incorporation of Co NPs.

To investigate the surface acidic and basic properties of the Co–SiO₂ catalyst, NH₃-TPD and CO₂-TPD analysis were performed, and the obtained plots are shown in Fig. S7.†²⁹ The basic sites were examined using a CO₂ gas probe, and the acidic sites were examined using an NH₃ gas probe. The gases were adsorbed onto the surface of the catalyst and as the temperature increased the desorption of gases occurred. These desorbed gases were then quantified to calculate the amount of acidic and basic sites present in the catalyst. The amounts of acidic and basic sites present in SiO₂ were found to be 1.247 and 1.627 mmol g⁻¹, respectively. The amount of basic sites present on the surface is due to the presence of hydroxyl groups on the surface and the amount of acidic sites is due to the vacant d-orbital present in the Si atoms.⁵⁰ The overall acidity and basicity of the catalyst increases after the incorporation of Co NP onto the surface of SiO₂. The amounts of acidic and basic sites present in Co–SiO₂ catalyst were found to be 2.292 and 2.616 mmol g⁻¹, respectively. For the NH₃-TPD profile of SiO₂ and Co–SiO₂, two desorption peaks in the temperature range of 450–600 °C and 650–750 °C were assigned to the desorption of NH₃ on medium and strong acidic sites. For the CO₂-TPD profile of SiO₂ and Co–SiO₂, three desorption peaks were observed in the temperature range of 150–250 °C, 420–630 °C, and 650–750 °C, and were assigned to the weak, medium, and strong basic sites, respectively.

3.2. Catalytic activity

The activity of the as-synthesized Co–SiO₂ catalyst was investigated for the methanolysis of PET. The primary objective of this work was to carry out the reaction at mild temperature for an optimal period of time using a minimum amount of catalyst. Therefore, several reaction parameters were varied, such as metal loading, catalyst amount, temperature, and reaction time in order to optimize the reaction conditions. A model reaction was conducted using PET (1 g) and methanol (7.5 mL) using 100 mg of Co–SiO₂ catalyst at 170 °C and 6 h to obtain white crystals of DMT as shown in Scheme 2. The resultant product was characterized using ¹H and ¹³C NMR spectroscopic methods.

Various optimization studies as mentioned above were performed to arrive at the optimal reaction conditions. The best optimized reaction conditions show 100% conversion of PET (substrate) having a 97% yield of DMT (product). Fig. 5a shows the effect of metal loading on the conversion of PET and DMT yield of the reaction. During the optimization process, in the first step, the reaction was carried out using different amounts (5 wt%, 10 wt%, 15 wt% and 20 wt%) of Co NP loading on SiO₂ to obtain the corresponding Co–SiO₂ catalysts. It was observed that with the increase in Co NP loading, both the PET conversion and the DMT yield increased substantially. As shown in Fig. 5a, the conversion of PET with 5 wt% Co loading on SiO₂ is 80% and then increases to 100% with 10 wt%, 15 wt% and 20 wt% Co NP loading. Simultaneously, the yield of DMT was determined to be 73% (with 5 wt% loading), 92% (with 10 wt% loading), 93% (with 15 wt% loading) and 97% (with 20 wt% loading). This enhanced activity is due to the fact that with the increase in the amount of Co NP loading, the number of active sites on the catalyst increases. This indicates that 20 wt% loading is considered as the optimum metal loading on silica and this sample was utilized for subsequent optimization studies. The effect of catalyst amount (Fig. 5b) on the reaction was studied by varying the amount of catalyst in the reaction. The catalyst amount was varied between 70 mg and 100 mg, where 100 mg of catalyst gives 100% PET conversion with 97% DMT yield. When the amount of catalyst was 70 mg, the PET conversion and DMT yield were 90% and 76%, respectively. The amount of catalyst was then increased to 80 mg, and the PET conversion and DMT yield were also increased to 92% and 88%, respectively. With a further increase in the catalyst amount (90 mg), the PET conversion and DMT yield was also increased to 97% and 90%, respectively, and finally with 100 mg of catalyst, the PET conversion and DMT yield increased to 100% and 97%, respectively.

Later, to study the effect of temperature on the reaction, the temperature was varied between 150 and 180 °C (Fig. 5c), and it was observed that at 170 °C, PET was completely converted to DMT. At 150 °C, the conversion of PET was only 80% which drastically increased to 100% at 170 °C. This is because methanolysis is an endothermic process which causes the reaction rate to increase substantially as the temperature rises. At lower temperature, the PET conversion and DMT yield were low and increased when the temperature increased to 160 °C where the PET conversion was 82% with 72% DMT yield. The PET conversion and DMT yield increased to 100% and 97%, respectively, when the temperature was increased to 170 °C. However, when the temperature was further increased to 180 °C, there was no significant increase in DMT yield. The aforementioned results indicate that a mildly high temperature is helpful to increase the PET conversion and eventually the DMT yield.

The reaction time also plays a crucial role in the reaction. To study its impact, the reaction was carried out at different time periods between 2 h and 8 h with 100 mg of Co–SiO₂ catalyst at 170 °C temperature (Fig. 5d). A DMT yield of 55% was obtained within 2 h and the yield increased from 55% to 97% within the next few hours. When the reaction was performed for 4 h, the PET conversion was 95% and 81% DMT was obtained. With a further increase in reaction time to 6 h and 8 h, full conversion of PET was observed with DMT yields of 97% and 96%, respectively. From the obtained results, one can conclude that 100 mg of 20 wt% Co–SiO₂ catalyst and 6 h reaction time at 170 °C are the best suited reaction conditions for the methanolysis of PET. In addition, control reactions were also performed using Co NPs and SiO₂ and without using any catalyst (refer to ESI;† section S3). With Co NPs alone, although the PET conversion was 100%, only 92% DMT yield was observed. When the reaction was carried out with SiO₂, the PET conversion and DMT yield were only 26% and 15%, respectively, indicating that SiO₂ primarily serves as a support which provides a high surface area for the active reaction sites. The reaction did not proceed when it was performed without using any catalyst. This implies that the catalyst plays a crucial role in the catalytic conversion of PET to DMT.

The developed reaction protocols were tested for plastics obtained from other types of PET bottles, including colored ones, under optimal reaction conditions to check the applicability of the designed reaction parameters. It was observed that the colorless PET plastics gave higher DMT yield as compared to the colored PET plastics as shown in Table 1. The highest yield of 97% was obtained in the case of colorless plastic which decreases to 94% when the plastic is green colored. Similarly, the DMT yield was 93% and 94% when methanolysis of yellow-colored plastics and mixed plastics were carried out, respectively. This could be attributed to the presence of additives in the colored plastic, which remain in the liquid phase and do not alter the DMT purity and catalyst reusability. This is also advantageous from a techno-economic perspective as there is no need to separate colored and colorless plastics beforehand which decreases the cost of separation of plastic waste. The reaction scheme for colored PET bottles is shown in Fig. S8 (refer to the ESI†) to provide valuable insights into the process applicability and catalyst reusability.

---

To check its potential industrial application, a gram scale reaction was also performed as shown in Scheme 3. The reaction was performed by using 5 g of PET wherein a DMT yield of 97% was observed with 100% PET conversion which exactly matches the DMT yield obtained using 1 g of PET. This result shows the applicability of the Co–SiO₂ catalyst for large scale conversion of PET.

3.3. Mechanism and structure activity correlation

According to previous studies, metal oxides in transesterification reactions can function as basic or acidic catalysts.⁵¹ As a basic catalyst, Co–SiO₂ provides enough adsorption sites for the alcohol to change into a more nucleophilic and catalytically active species which makes the oxygen in methanol attack the carbonyl carbon of PET. Thereafter, the ester bond breaks, creating a new C–O bond between the carbonyl in PET and oxygen in methanol which results in the formation of DMT. However, when Co–SiO₂ acts as an acidic catalyst, it binds with the ester's carbonyl oxygen which increases the electrophilicity of the carbonyl carbon and vulnerability to nucleophilic attack by methanol. Therefore, by determining whether Co–SiO₂ functions as an acidic or basic catalyst, the reaction mechanism of PET methanolysis can be suggested.

The control reactions were carried out to comprehend the role of the acidic and basic sites present in the Co–SiO₂ catalyst. Pyridine and benzoic acid were used as quenchers for the acidic and basic sites, respectively, as shown in Fig. S9 (refer to the ESI†). As shown in TPD investigations, Co–SiO₂ contains both acidic and basic sites. The model reaction was severely hindered by the addition of benzoic acid, which quenched the basic sites present in the catalyst yielding only 8% DMT. The addition of pyridine also had an impact on the progress of the reaction and because it quenched the acidic sites, the DMT yield was reduced to 40%. This shows that Co–SiO₂ predominantly acts as a mildly basic catalyst for the methanolysis of PET. It is well known that PET first gets converted into its oligomers, which then gets converted into 2-hydroxyethyl methyl terephthalate (MHET) and finally to dimethyl terephthalate (DMT).^52–55 Based on relevant literature reports^52–55 and experimental results, a plausible reaction mechanism for PET methanolysis using Co–SiO₂ as a basic catalyst is proposed in Scheme 4. Co–SiO₂ functions as a base catalyst, providing methanol sufficient adsorption sites to transform into a more nucleophilic and catalytically active species. This allows methanol's oxygen to attack the carbonyl carbon of PET. After that the ester bond breaks, forming a new C–O bond between the methanol's oxygen and the carbonyl carbon of PET which leads to the formation of oligomers which are finally converted into a DMT monomer as the major product.

3.4. Recyclability studies

Recyclability studies were performed to check the heterogeneous nature of the catalyst and its stability so that it can be utilized for several reaction cycles. In this regard, the reaction was carried out using 5 g of PET, 500 mg of Co–SiO₂ and 37.5 mL of methanol. The catalyst was then separated from the reaction mixture using centrifugation and washed several times using DI water. The catalyst was then dried for 24 h at 80 °C and was further utilized for the next cycle. Following each cycle, the PET conversion and DMT yield were calculated (Fig. 6a). It is noteworthy to mention that there was no decrease in the PET conversion, which always remained at 100% and there was only a very slight decrease in the DMT yield from 97% to 96%. Subsequently, to check the structural integrity of the catalyst, the PXRD measurements were carried out on the recovered catalyst obtained after five catalytic cycles (Fig. 6b). Only a slight decrease in the intensity of SiO₂ was observed which could be due to the loss of SiO₂ during washing and recovery of the catalyst after each cycle as it is a fluffy material. In addition to this, XPS measurements were also performed to determine the atomic composition of the recovered catalyst. Here, the data obtained for the recovered catalyst were compared with those of the fresh catalyst, which indicates no remarkable change in the atomic composition of elements in the recovered Co–SiO₂ catalyst as shown in Table S3 (refer to the ESI†). This reveals that there is no leaching of Co NPs from the SiO₂ support.

4. Conclusion

Through this work we have presented a catalytic system having high efficiency and sustainability for the methanolysis of PET by supporting cobalt nanoparticles on silica. The silica support provides increased surface area and high thermal strength to the catalyst leading to its higher stability and better catalytic activity for the methanolysis reaction with high PET conversion (100%) and high yield of DMT (97%) under optimized reaction conditions. The reactions were performed using a minimal amount of catalyst under mild conditions of temperature and pressure in an environmentally benign solvent. Furthermore, the reactions were carried out in a shorter reaction time. It was also observed that the high purity DMT was obtained from both colored and colorless PET plastics, which is very crucial for the large-scale chemical upcycling of PET from a techno-economic viewpoint. In addition, the catalyst is very stable and can be reused for multiple cycles without any remarkable decrease in the product yield. Based on the current study, it can be inferred that methanolysis of PET waste can be performed using Co–SiO₂ catalysts under mild reaction conditions to achieve good product yield. This work is anticipated to pave the way for the chemical upcycling of PET waste in a green and sustainable manner.

Data availability

The data supporting this article have been included as part of the ESI.†

Conflicts of interest

The authors declare no conflict of interest.

Acknowledgements

The Advanced Materials Research Center (AMRC) at the Indian Institute of Technology Mandi is gratefully acknowledged for providing laboratory and characterization facilities. KK acknowledges the doctoral scholarship from the Ministry of Education (MoE), Government of India.

References

1. X. Zhao, M. Korey, K. Li, K. Copenhaver, H. Tekinalp, S. Celik, K. Kalaitzidou, R. Ruan, A. J. Ragauskas and S. Ozcan, Chem. Eng. J., 2022, 428, 131928 CrossRef CAS .
2. S. L. Aggarwal and O. J. Sweeting, Chem. Rev., 1957, 57, 665–742 CrossRef CAS .
3. R. Nisticò, Polym. Test., 2020, 90, 106707 CrossRef .
4. H. A. Maddah, Am. J. Polym. Sci., 2016, 6, 1–11 CAS .
5. Y. Wu, G. Wang, Z. Wang, Y. Liu, P. Gu and D. Sun, Front. Environ. Sci. Eng., 2014, 8, 451–462 CrossRef CAS .
6. P. Mishra, S. Vinayagam, K. Duraisamy, S. R. Patil, J. Godbole, A. Mohan, A. Mukherjee and N. Chandrasekaran, Environ. Sci. Pollut. Res., 2019, 26, 1537–1547 CrossRef CAS PubMed .
7. V. Tournier, C. Topham, A. Gilles, B. David, C. Folgoas, E. Moya-Leclair, E. Kamionka, M.-L. Desrousseaux, H. Texier and S. Gavalda, Nature, 2020, 580, 216–219 CrossRef CAS PubMed .
8. C.-C. Chen, X. Han, X. Li, P. Jiang, D. Niu, L. Ma, W. Liu, S. Li, Y. Qu and H. Hu, Nat. Catal., 2021, 4, 425–430 CrossRef CAS .
9. D. Loaeza, J. Cailloux, O. Santana Pérez, M. Sánchez-Soto and M. L. Maspoch, Polymer, 2021, 13, 310 CAS .
10. J. Zhou, T. G. Hsu and J. Wang, Angew. Chem., Int. Ed., 2023, 62, e202300768 CrossRef CAS PubMed .
11. G. Suzuki, N. Uchida, K. Tanaka, H. Matsukami, T. Kunisue, S. Takahashi, P. H. Viet, H. Kuramochi and M. Osako, Environ. Pollut., 2022, 303, 119114 CrossRef CAS PubMed .
12. M. Jalilian, A. S. Joshi, M. R. Coleman, Y.-W. Kim and J. G. Lawrence, Polym. Degrad. Stab., 2021, 194, 109748 CrossRef CAS .
13. L. G. Schaerer, R. Wu, L. I. Putman, J. M. Pearce, T. Lu, D. R. Shonnard, R. G. Ong and S. M. Techtmann, Trends Biotechnol., 2023, 41, 184–196 CrossRef CAS PubMed .
14. M. D. de Dios Caputto, R. Navarro, J. L. Valentín and Á. Marcos-Fernández, J. Polym. Sci., 2022, 60, 3269–3283 CrossRef CAS .
15. Y. Ogiwara and K. Nomura, ACS Org. Inorg. Au, 2023, 3, 377–383 CrossRef CAS PubMed .
16. L. Wang, G. A. Nelson, J. Toland and J. D. Holbrey, ACS Sustainable Chem. Eng., 2020, 8, 13362–13368 CrossRef CAS .
17. R. López-Fonseca, I. Duque-Ingunza, B. de Rivas, L. Flores-Giraldo and J. I. Gutiérrez-Ortiz, Chem. Eng. J., 2011, 168, 312–320 CrossRef .
18. Y. Jo, E. J. Kim, J. Kim and K. An, Green Chem., 2023, 25, 8160–8171 RSC .
19. J. J. R. Arias and W. Thielemans, Green Chem., 2021, 23, 9945–9956 RSC .
20. H. Abedsoltan, I. S. Omodolor, A. C. Alba-Rubio and M. R. Coleman, Polymer, 2021, 222, 123620 CrossRef CAS .
21. A. Barredo, A. Asueta, I. Amundarain, J. Leivar, R. Miguel-Fernández, S. Arnaiz, E. Epelde, R. López-Fonseca and J. I. Gutiérrez-Ortiz, J. Environ. Chem. Eng., 2023, 11, 109823 CrossRef CAS .
22. Q. Liu, R. Li and T. Fang, Chem. Eng. J., 2015, 270, 535–541 CrossRef CAS .
23. K. Lee, Y. Jing, Y. Wang and N. Yan, Nat. Rev. Chem., 2022, 6, 635–652 CrossRef PubMed .
24. P. Pereira, P. E. Savage and C. W. Pester, ACS Sustainable Chem. Eng., 2023, 11, 7203–7209 CrossRef CAS .
25. A. Singh, N. A. Rorrer, S. R. Nicholson, E. Erickson, J. S. DesVeaux, A. F. Avelino, P. Lamers, A. Bhatt, Y. Zhang and G. Avery, Joule, 2021, 5, 2479–2503 CrossRef CAS .
26. S. Hongkailers, Y. Jing, Y. Wang, N. Hinchiranan and N. Yan, ChemSusChem, 2021, 14, 4330–4339 CrossRef CAS PubMed .
27. Y. Huang, Y. Ma, Y. Cheng, L. Wang and X. Li, Ind. Eng. Chem. Res., 2014, 53, 4604–4613 CrossRef CAS .
28. Y. Li, M. Wang, X. Liu, C. Hu, D. Xiao and D. Ma, Angew. Chem., 2022, 134, e202117205 CrossRef .
29. Z. Gao, B. Ma, S. Chen, J. Tian and C. Zhao, Nat. Commun., 2022, 13, 3343 CrossRef CAS PubMed .
30. T. Sako, I. Okajima, T. Sugeta, K. Otake, S. Yoda, Y. Takebayashi and C. Kamizawa, Polym. J., 2000, 32, 178–181 CrossRef CAS .
31. H. Kurokawa, M.-A. Ohshima, K. Sugiyama and H. Miura, Polym. Degrad. Stab., 2003, 79, 529–533 CrossRef CAS .
32. J. Chen, J. Lv, Y. Ji, J. Ding, X. Yang, M. Zou and L. Xing, Polym. Degrad. Stab., 2014, 107, 178–183 CrossRef CAS .
33. S. Mishra and A. S. Goje, Polym. Int., 2003, 52, 337–342 CrossRef CAS .
34. S. Lalhmangaihzuala, Z. Laldinpuii, V. Khiangte, G. Lallawmzuali and K. Vanlaldinpuia, Adv. Powder Technol., 2023, 34, 104076 CrossRef CAS .
35. S. Tang, F. Li, J. Liu, B. Guo, Z. Tian and J. Lv, J. Environ. Chem. Eng., 2022, 10, 107927 CrossRef CAS .
36. F. Guo, H. Zheng, Z. Yang and Y. Qian, Mater. Lett., 2002, 56, 906–909 CrossRef CAS .
37. S. Zekriardehani, A. Joshi, S. Jabarin, D. Gidley and M. Coleman, Macromol., 2018, 51, 456–467 CrossRef CAS .
38. S. Lyu, L. Wang, J. Zhang, C. Liu, J. Sun, B. Peng, Y. Wang, K. G. Rappé, Y. Zhang and J. Li, ACS Catal., 2018, 8, 7787–7798 CrossRef .
39. P. Reif, N. K. Gupta and M. Rose, Green Chem., 2023, 25, 1588–1596 RSC .
40. B. Shi, L. Xie, B. Ma, Z. Zhou, B. Xu and L. Qu, Gels, 2022, 8, 744 CrossRef CAS PubMed .
41. Y. Liang, J. Ouyang, H. Wang, W. Wang, P. Chui and K. Sun, Appl. Surf. Sci., 2012, 258, 3689–3694 CrossRef CAS .
42. W. Chen, Z. Bao and Z. Zhou, React. Kinet., Mech. Catal., 2022, 135, 2533–2550 CrossRef CAS .
43. Y. Liu, K. Zhou, M. Lu, L. Wang, Z. Wei and X. Li, Ind. Eng. Chem. Res., 2015, 54, 9124–9132 CrossRef CAS .
44. X. Guo, M. Zhang, J. Zheng, J. Xu, T. Hayat, N. S. Alharbi, B. Xi and S. Xiong, Dalton Trans., 2017, 46, 11598–11607 RSC .
45. T. Li, F. Li, J. Nginyo, W. Cai and B. Yu, Mol. Catal., 2023, 547, 113307 CrossRef CAS .
46. X. Yu, Z. Zhao, Y. Wei, L. Zhao and J. Liu, Catal. Sci. Technol., 2019, 9, 1372–1386 RSC .
47. J. Cañón and A. V. Teplyakov, Surf. Interface Anal., 2021, 53, 475–481 CrossRef .
48. M. Ojeda, F. J. Pérez-Alonso, P. Terreros, S. Rojas, T. Herranz, M. López Granados and J. L. G. Fierro, Langmuir, 2006, 22, 3131–3137 CrossRef CAS PubMed .
49. M. Thommes, K. Kaneko, A. V. Neimark, J. P. Olivier, F. Rodriguez-Reinoso, J. Rouquerol and K. S. Sing, Pure Appl. Chem., 2015, 87, 1051–1069 CrossRef CAS .
50. M. Mihet, M. Dan, L. Barbu-Tudoran and M. D. Lazar, Catalysts, 2021, 11, 443 CrossRef CAS .
51. L. Bartolome, M. Imran, K. G. Lee, A. Sangalang and J. K. Ahn, Green Chem., 2014, 16, 279–286 RSC .
52. Z. Laldinpuii, V. Khiangte, S. Lalhmangaihzuala, C. Lalmuanpuia, Z. Pachuau, C. Lalhriatpuia and K. Vanlaldinpuia, J. Polym. Environ., 2022, 30, 1600–1614 CrossRef CAS .
53. Z. Laldinpuii, C. Lalmuanpuia, S. Lalhmangaihzuala, V. Khiangte, Z. Pachuau and K. Vanlaldinpuia, New J. Chem., 2021, 45, 19542–19552 RSC .
54. D. D. Pham and J. Cho, Green Chem., 2021, 23, 511–525 RSC .
55. Y. Zhang, J. Gao, C. Jiang, G. Luo, J. Fan, J. H. Clark and S. Zhang, Green Chem., 2024, 26, 6748–6759 RSC .

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4cy00468j

This journal is © The Royal Society of Chemistry 2024