Converting waste PET plastics into automobile fuels and antifreeze components

With the aim to solve the serious problem of white plastic pollution, we report herein a low-cost process to quantitatively convert polyethylene terephthalate (PET) into p-xylene (PX) and ethylene glycol (EG) over modified Cu/SiO2 catalyst using methanol as both solvent and hydrogen donor. Kinetic and in-situ Fourier-transform infrared spectroscopy (FTIR) studies demonstrate that the degradation of PET into PX involves tandem PET methanolysis and dimethyl terephthalate (DMT) selective hydro-deoxygenation (HDO) steps with the in-situ produced H2 from methanol decomposition at 210 °C. The overall high activities are attributed to the high Cu+/Cu0 ratio derived from the dense and granular copper silicate precursor, as formed by the induction of proper NaCl addition during the hydrothermal synthesis. This hydrogen-free one-pot approach allows to directly produce gasoline fuels and antifreeze components from waste poly-ester plastic, providing a feasible solution to the plastic problem in islands.

24. SI line 105: ICP calibration method used? 25. Figure S2: Please add the initial HT sample result as well for comparison.
Reviewer #2: Remarks to the Author: PET contributes significantly to plastic waste generation. In this work, Zhao and co-workers developed a new H2 free method using Cu based catalyst to convert PET into xylene and ethylene glycol. The conversion efficiency is high and the reaction pathway has been well studied. The catalyst structureactivity correlation has also been convincingly described. Overall it is a nice piece of study that deserves to be published in Nat. Commun. after proper revision. 1) The stability of the catalyst. I feel this is a major limitation of the work. There is no adequate information on the stability of the Cu catalyst. Detailed characterizations of the spent Cu catalyst may be provided. Further, the reusability of the catalyst may be studied in more detail. If direct reuse is not possible, what is the reason for catalyst deactivation and whether it is possible to recover catalyst activity by certain treatment.
2) The authors may also comment on the applicability of the catalytic system beyond PET. If it is only applicable to PET, then sorting strategies have to be applied.
3) Methanol is used as solvent and hydrogen donor. How much methanol is decomposed during the process? Does the consumption of methanol match that of PET conversion? 4) It would also be good if the authors comment on how to purify products and reuse unreacted methanol. 5) It is interesting to see the example of Phuket Island. Did the authors really use samples collected there, or it is just based on literature report? This needs to be made clear. If the sample from Phuket Island is not used, then it should be removed from Figure 5. Related parts in MS should be modified as well.
Reviewer #3: Remarks to the Author: The manuscript "Converting waste PET plastics into automobile fuel and antifreeze components" by Zhao and coworkers describes a novel methodology for the depolymerization of PET waste using methanol as the solvent and hydrogen source, and a Cu-based catalyst. This new method is very important due to the use of an alcohol as reducing agent and also a non-toxic and earth abundant metal catalyst. The work is well written and the discussion of the results appropriate, involving an extensive study of reaction conditions. Some aspects of the reaction mechanism were also included in this study.
I recommend the publication of this manuscript in nature communications, after major revisions: -Authors should test this method using other alcohols, for example ethanol and isopropanol, as the hydrogen source and verify that PET depolymerization and p-xylene formation also occur. -Authors should also indicate whether the PET alcoholysis reaction can be carried out at temperatures below 210 °C and what yields are obtained.
-To study the applicability of this method, it would be very interesting if the authors tested this method from the second most used polyester, polybutylene terephthalate (PBT), and checked if it is also possible to obtain p-xylene with good yields. We are grateful for the constructive comments from three reviewers, and based on these suggestions, we have modified and improved our manuscript. The detailed responses to the comments of the reviewers are listed in blue together with the text of the original material.

Manuscript ID: NCOMMS-21-40636A
Title: Converting waste PET plastics into automobile fuel and antifreeze components

Reviewer #1:
The article describes hydrogen gas-free conversion of PET plastic to para-xylene and glycol. In itself the work is interesting, but many questions remain regarding the catalytic compounds, rationale, comparison with competing technology, as well as experimental details that I would like the authors to address. I have listed them below.
1. The rationale for converting PET waste into fuel baffles me a bit. In this work PET is converted first to glycol and DMT, after which DMT is converted to PX for fuel applications. This would mean that valuable resources, especially aromatics, will just be burnt rather than reused. Given the limited amount of resources available on the planet one should focus on reusing resources rather than burning them. Please discuss the rationale of the work in light of this in the introduction.
Reply: Thanks for the Reviewer's suggestion. We agree with you that plastic waste is one of the most valuable wastes and can be considered as a potentially cheap source for the production of industrial fuels and chemicals. In land cities, it is feasible to converted plastics into chemical raw materials through effective chemical 2 recycling methods. However, in some islands, especially those with developed tourism, due to the lack of industry on the island, a large amount of abandoned plastic waste can only be disposed by landfill or incineration. Recycled plastics are also generally limited to feasible applications where low-quality materials are collected, resulting in minimal economic incentives for waste recycling, sorting and processing. At this time, if it can be used as a raw material to output gasoline energy and antifreeze components on the island through a simple process, it will be a very practical way.
2. PET depolymerization has recently been reviewed by Barnard, Rubio and Thielemans in Green Chemistry. This work also introduces a methodology to quantifiably compare different processes based on materials use (using Sheldon's factor) and energy consumption. Please compare the efficiency of this work with the existing literature using this (or a similar) methodology.
Reply: Thanks for the Reviewer's suggestion. We tried to use the environmental factor and environmental energy impact in the Barnard, Rubio and Thielemans's work to evaluate the efficiency of several parallel works. Firstly, energy economy coefficient (ε) is proposed to enable objective comparison on the influence of para meters such as temperature, catalyst type, or proportion of starting materials, where t is the reaction time (in minutes), T is the reaction temperature in degrees celsius, and Y is the yield of the main monomer in mass fraction (which containing the aromatic moiety) in eqn (1). Barnard et al. improved the environmental factor (Efactor) in eqn (4) which took the effect of materials input that results in waste generation into consideration. The environmental energy impact factor (ξ) results from the combination of the two factors above as presented in eqn (5). The best processes would tend to present low values of Efactor and ξ factor and high ε values.
Table R1 clearly showed that this work has the highest ε (1.323E-5°C -1 *min -1 ) due to its excellent product yield (100%) and low reaction temperature (210 °C) and time (360 min). High solvent/PET ratio (197.5) resulted in the high Efactor (37.19). But ξ (2811035°C*min) is still the smallest by combining the above two coefficients, and about twice time than other works. We also tried to redouble the PET and catalyst at the same time, and still get 100% PX yield, greatly reduced Efactor as well as ξ. 3 In addition, only this work uses non-noble metal catalysts, and the obtained products are highly selective. 3. PET recycling in the introduction is heavily focused on conversion to fuel, whereas depolymerization into monomers would economically make more sense (less steps and no loss of resources) yet is not covered. Please add this to the introduction.
Reply: Thanks for the Reviewer's suggestion. We have added the depolymerization of PET into monomers in introduction.
Chemical depolymerization methods, mainly include hydrolysis, glycolysis, and ammonolysis [3][4][5][6] , can reverse the chemical composition of plastics and turn into stable monomer molecules again. However, these methods still face limitations of harsh reaction conditions, low product yield, and purification difficulties.   Reply: Thanks for the Reviewer's suggestion. We repeated each experiment three times, and the error bars are added in following figures.  Reply: Thanks for the Reviewer's suggestion. Due to the small particle size, poor crystallinity and weak intensity, XRD results maybe not suitable for Rietveld refinement. We compared the results with the standard spectrum (PDF#27-0188) in the database [7][8] , and the results are shown in Figure 2a. Ref.  The O1s XPS of reduced catalysts was conducted, as shown in Figure R1. The peak at 530.3 eV was observed in the reduced CuNa/SiO2 and Cu/SiO2, ascribed to Cu2O or CuSiO3 as reported.
[9-10] The peak of 532.5 eV was attributed to the O 1s of SiO2 support [10] . However, we cannot distinguish the Cu2O and CuSiO3 species through the O1s XPS analysis.

Ref.
[9] Ding, J. 7. Line 99: How was partial reduction of Cu + to Cu 0 measured? How much is partial reduction? Please quantify.
Reply: Thanks for the Reviewer's suggestion. The compositions of reduced catalysts were calculated from XPS and XAES analysis [12][13] , as shown in Supplementary to the higher difficulty in reduction. The X-ray induced Auger spectra (XAES) Cu LMM were employed to distinguish between the Cu 0 and Cu + and the deconvolution results are listed. The higher Cu + /Cu 0 ratio (1.87) of CuNa/SiO2 confirmed that after the addition of Na + , copper silicate with a low crystallinity and a dense texture was less likely to be reduced to Cu 0 . A higher ratio of Cu + /Cu 0 was indicative of a higher tendency to both methanol dehydrogenation and DMT hydrodeoxygenation.
This part has been modified in the revised version. 8. Line 112: XRD data in Figure 2S do not say anything about surface species, only bulk species. Please provide and describe surface analysis info as additional info.

Supplementary
Reply: Thanks for the Reviewer's suggestion. To investigate the surface information of Cu/SiO2 samples prepared by different methods, we used XPS and Auger Cu 11 LMM analysis to analyze the distributions of copper species on their surfaces ( Supplementary Fig. 3). On the XPS profiles of HT and DPA samples, there are obvious Cu 2+ satellite peaks (940-950 eV) ( Supplementary Fig. 3a), indicating that Cu 2+ was incompletely reduced. However, the same phenomenon does not appear in the XPS of DPU and IM Cu/SiO2 samples. The Cu LMM X-ray induced Auger spectra (XAES) were employed to distinguish between the Cu 0 and Cu + species ( Supplementary   Fig. 3b). The results showed that the ratios of Cu + /Cu 0 in DPU, DPA and IM samples were significantly lower than such ratio in the HT sample, which corresponds to a significant decrease in reactivity of methanol dehydrogenation and PET hydrodeoxygenation (Table 1). This part has been revised in the renewed version. 10. Line 167: The way this is written seems like Cu + has a peak at 952.2 eV and Cu 0 at 932.1 eV. Please rewrite and also specify the orbital assignment.
Reply: Thanks for the Reviewer's suggestion. We have already fixed the typographical error and rewrite it as below.
13. Line 210: "poor crystallinity". Please quantify Reply: Thanks for the Reviewer's suggestion. We used the software MDI-Jade to fit the XRD data of samples firstly, and then calculated the crystallinity of each sample.
The crystallinity and R-values (fitting error) are listed below.
14 Supplementary Line 231: "large interface area" How was this determined? Quantification?
Reply: In order to roughly quantify the interface areas of the formed copper silicate with SiO2, we supplemented the measurement of IR spectroscopy in vacuum and determined the remaining silanols groups on SiO2 (Supplementary Fig. 14). The peak at 3740 cm -1 in the SiO2 sample is attributed to Si-OH groups ( Supplementary   Fig. 14) [16][17] . With the traditional hydrothermal method, Cu 2+ in the solution combined with the silanol on the SiO2 surface to form copper silicate, which accelerated the layered copper silicate nucleation and growth significantly. Thus only a small amount of Si-OH can still be detected. Upon addition of 5 NaCl, a large amount of Na + occupied the silanol on the surface of the SiO2, Cu 2+ in the solution could only be combined with the remaining silanol on the SiO2 surface to form scattered and isolated copper silicate particles (Figures 2g and 2i), which was attached to the surface of the carrier. Thus, the formed granular copper silicate showed a large interface area with SiO2, since no remaining Si-OH on CuNa/SiO2 was detected by IR spectra in vacuum ( Supplementary Fig. 14). When 15 Na + was introduced, Na + occupied almost all the Si-OH on the surface, and Cu 2+ can only combined with SiO3 2− in the solution to form granular copper silicate and then deposited on the SiO2. After washed with deionized water, some Si-OH groups are exposed, the peak at 3740 cm -1 is reserved. Therefore, upon adding 15 Na + , this type of copper silicate showed 15 better crystallinity (Supplementary Table 6) and small interface areas with SiO2 ( Supplementary Fig. 14).
This part has been modified accordingly. 16. Line 227: "inhibiting nucleation and growth" How was this determined? 16 Reply: Thanks for the Reviewer's suggestion. Transmission electron microscopy (TEM) images intuitively showed the different morphologies of the two copper silicates formed with and without NaCl introduction during the hydrothermal process. Thus, while the dried precursor of Cu/SiO2 showed a layered copper silicate structure (Figure 2g), the dried precursor of CuNa/SiO2 showed a special state of granular particle accumulation (Figure 2i). Therefore, it is inferred that Na + introduction indeed inhibits the nucleation and growth of layered copper silicate precursor. 17. What is special about the 5:1 ratio Na + /Cu 2+ that this would give rise to the optimal Cu + /Cu 0 ratio? Reply: Thanks for the Reviewer's suggestion.
When the Na + /Cu 2+ ratio is 5:1, the obtained granular copper silicate has the smallest specific surface area, the lowest water content and the poorest crystallinity ( Supplementary Fig. 10-13). TGA tests of the CuNa/SiO2 precursor showed that physisorbed water (2.41%) and crystal water (6.75%) upon addition of 5 NaCl was the lowest among all the samples tested ( Supplementary Fig. 11). This also confirmed that the copper silicate structure was densest at this ratio. N2 adsorption-desorption 17 ( Supplementary Fig. 12) revealed that CuNa/SiO2 had the lowest surface area (46.9 m 2 /g) upon addition of 5 NaCl, indicating that the formed structure was the most compact among the samples tested herein.
A large amount of Na + occupied the silanol on the surface of the SiO2 upon addition of 5 NaCl, thereby inhibiting nucleation and growth of layered copper silicate (Figures 2g and 2i). Cu 2+ in the solution could only be combined with the remaining silanol on the SiO2 surface to form scattered and isolated copper silicate particles, and the compact structure had a small surface area and poor crystallinity. This granular copper silicate grown from the surface of SiO2 is more stable and difficult to be reduced, resulting in a higher Cu + /Cu 0 ratio and benefiting the further methanol dehydrogenation and DMT hydrodeoxygenation.
However, when the amount of added NaCl was too high, Na + occupied all the silanol sites on SiO2, resulting in the precipitation of Cu 2+ with SiO3 2− in solution to form copper silicate, which was then deposited on the SiO2 surface. Compared to the catalyst with 5 NaCl introduced during hydrothermal treatment, this type of copper silicate showed better crystallinity (Supplementary Table 6) and was relatively easier to be reduced to Cu/Cu2O·SiO2 with a low ratio of Cu + /Cu 0 ( Supplementary   Fig. 9d). In general, the addition of NaCl in the hydrothermal treatment resulted in the formation of granular copper silicate with a lower crystallinity, smaller specific surface area, and denser texture, which leads to form an optimal Cu + /Cu 0 ratio after reduction. Reply: Thanks for the Reviewer's suggestion. A recent survey of beach sediment along the coastline of the Phuket Island showed that PET (mainly containing beverage bottles, plastic films, and microwave packaging) accounted for ca. 33.1% of the overall plastic sediment (Supplementary Fig. 17). Several common PET plastics that are available on the tourist island were chosen to convert, such as Coca-Cola bottles, McDonald's drink caps, disposable lunch boxes, packaging bags and even some polyester clothes ( Supplementary Fig. 18a). After simple treatment with the raw materials by scissors ( Supplementary Fig. 18b), we obtained 100% yield of p-xylene from different sources of PET plastics at the same catalytic conditions.
Considering that lots of plastic wastes on the island sediments landfill are mixed together, in this context, we used CuNa/SiO2 to catalyze the mixtures of PET and another plastic PBT at 210 °C, and the results showed that mixed plastics can be completely converted to p-xylene as well.
Supplementary Fig. 17 Compositions in sediments along the coast of Phuket island. Reply: Thanks for the Reviewer's suggestion. We added all the chemical purities. Reply: Thanks for the Reviewer's suggestion. Inductively coupled plasma atomic emission spectroscopy (ICP-AES): The content of each element in the catalyst sample was determined by a PerkinElmer Optima 8300 inductively coupled plasma atomic emission spectrometer. The test process was as follows, Firstly, the catalyst sample was dissolved in hydrofluoric acid to ensure that it was completely dissolved and in a clear state. Finally, the solution was diluted to a suitable test range. The five standard solutions were prepared to construct the external standard curve. The content of elements in the samples was determined by external standard curve. We have 23 added this part into the revised manuscript.
25. Figure S2: Please add the initial HT sample result as well for comparison.
Reply: Thanks for the Reviewer's suggestion. We have added the initial HT sample result in the revised version. Comments: PET contributes significantly to plastic waste generation. In this work, Zhao and co-workers developed a new H2 free method using Cu based catalyst to convert PET into xylene and ethylene glycol. The conversion efficiency is high and the reaction pathway has been well studied. The catalyst structure-activity correlation has also been convincingly described. Overall it is a nice piece of study that deserves to be published in Nat. Commun. after proper revision.
1) The stability of the catalyst. I feel this is a major limitation of the work. There is no adequate information on the stability of the Cu catalyst. Detailed characterizations of the spent Cu catalyst may be provided. Further, the reusability of the catalyst may be studied in more detail. If direct reuse is not possible, what is the reason for catalyst deactivation and whether it is possible to recover catalyst activity by certain treatment.
Reply: Thanks for the Reviewer's suggestion. After we tested the catalyst for four bathes, the catalyst deactivated obviously (Supplementary Table 9). According to XRD results, a large amount of Cu 0 was reduced and the size of particles was increased ( Supplementary Fig. 16a). In line with XRD results, TEM image showed that the Cu particles size was also increased after recycling tests ( Supplementary Fig. 16b-c). XAES analysis proved that the ratio of Cu + /Cu 0 drastically decreased from 1.80 to 0.57 ( Supplementary Fig. 16f), which hindered the synergistic effect of Cu 0 and Cu + in the catalytic process. In addition, Cu 2+ was partially reduced by the excess hydrogen produced by methanol dehydrogenation, as shown in XPS of Supplementary Fig. 16d-e.
Concerning on the recovery of catalyst activity, we tried to calcinate the used catalyst at 450°C in an air atmosphere for 4 h, and then reduced it at 450 °C in a hydrogen atmosphere for 4 h. The obtained catalyst continued for cycle testing and the results showed that PX yield attained 74.8%, 62.5% in the runs 1 and 2, while such recovered catalyst totally deactivated in the third run, probably due to the difficulty in rebalancing the ratio of Cu 2+ , Cu + and Cu 0 species in copper silicate catalysts.  Fig. 16 (a) XRD patterns of used CuNa/SiO2 catalyst and CuNa/SiO2 (reduced); TEM image of (b) used CuNa/SiO2 catalyst and (c) CuNa/SiO2 (reduced);
2) The authors may also comment on the applicability of the catalytic system beyond PET. If it is only applicable to PET, then sorting strategies have to be applied. 3) Methanol is used as solvent and hydrogen donor. How much methanol is decomposed during the process? Does the consumption of methanol match that of PET conversion?
Reply: Thanks for the Reviewer's suggestion. By calculating the consumption of methanol and the production of hydrogen, the experimental results showed that the values basically match. The following is the calculation process： Reactions involved in the PET conversion process: Real methanol consumption (according to liquid phase)： 28 Experimental Method: 30 mL methanol was diluted 10 times in ethyl acetate with 0.1 mL tetrahydronaphthalene as the internal standard before the reaction and determined the response factor (f) by GC-MS using following formula.
The reacted solution was diluted 10 times with ethyl acetate, the residual amount of methanol is calculated by the following formula subsequently. According to the reduction of peak area, the residual methanol content was 97.6%. 4) It would also be good if the authors comment on how to purify products and reuse unreacted methanol.
Reply: Thanks for the Reviewer's suggestion. Both the products and methanol can be separated by simple distillation based on their different boiling points (see Supplementary Table 10 below).  Fig. 18a). After simple treatment with the raw materials by scissors ( Supplementary Fig. 18b), we obtained 100% yield of p-xylene from different sources of PET plastics at the same catalytic conditions. Considering that most plastic wastes on the island sediments landfill are mixture together, in this context, we used CuNa/SiO2 to catalyze the mixture of PET and PBT at 210 °C, and the results showed that mixed plastics can be completely converted to p-xylene as well. The manuscript "Converting waste PET plastics into automobile fuel and antifreeze components" by Zhao and coworkers describes a novel methodology for the depolymerization of PET waste using methanol as the solvent and hydrogen source, and a Cu-based catalyst. This new method is very important due to the use of an alcohol as reducing agent and also a non-toxic and earth abundant metal catalyst.

Supplementary
The work is well written and the discussion of the results appropriate, involving an extensive study of reaction conditions. Some aspects of the reaction mechanism were also included in this study. I recommend the publication of this manuscript in nature communications, after major revisions: 1.Authors should test this method using other alcohols, for example ethanol and isopropanol, as the hydrogen source and verify that PET depolymerization and p-xylene formation also occur.
Reply: Thanks for the Reviewer's suggestion. We have supplemented the experiments on the PET conversion in ethanol and isopropanol in Supplementary Table 3.
Experimental data showed that PET can be well alcoholyzed in both ethanol and isopropanol, obtain 80.3% and 73.5% yields of monomers after 0.5 h, respectively.
However, the hydrogen released from ethanol and isopropanol decomposition over CuNa/SiO2 was not sufficient, which attained only 0.7 and 0.8 MPa incremental pressure at ambient temperature, respectively. In comparison, methanol released as high as 3.8 MPa gases at identical conditions. In the further PET alcoholysis and hydrodeoxygenation tests in ethanol and isopropanol, the gained DMT monomers from PET were not hydrogenated and no p-xylene was formed in ethanol or isopropsanol (Supplementary Table 3  3. To study the applicability of this method, it would be very interesting if the authors tested this method from the second most used polyester, polybutylene terephthalate (PBT), and checked if it is also possible to obtain p-xylene with good yields.
Reply: Thanks for the Reviewer's suggestion. We tested PBT conversion in methanol over CuNa/SiO2 at different temperatures in the same catalytic system, and the results are very similar to PET conversion (Supplementary and 72.4%. DMT was not further converted due to the low gas pressures (0.5 and 0 MPa) at these two temperatures.