Hydrodynamic synthesis of Fe₂O₃@MoS₂ 0D/2D-nanocomposite material and its application as a catalyst in the glycolysis of polyethylene terephthalate

Abstract

We report a fast and simple synthesis of Fe₂O₃@MoS₂ 0D/2D-nanocomposite material using a Taylor–Couette flow reactor. A Taylor–Couette flow with high shear stress and mixing characteristics was used for fluid dynamic exfoliation of MoS₂ and deposition of uniform Fe₂O₃ nanoparticles, resulting in a Fe₂O₃@MoS₂ in the form of 0D/2D-nanocomposite material. Using Taylor–Couette flow reactor, we could synthesize Fe₂O₃@MoS₂ 0D/2D-nanocomposite material at a rate higher than 1000 mg h⁻¹ which is much higher than previously reported production rate of 0.2–116.7 mg h⁻¹. The synthesis of Fe₂O₃@MoS₂ nanocomposite was achieved in an aqueous solution without thermal or organic solvent treatment. Exfoliated MoS₂ nanosheets show an average thickness of 2.6 ± 2.3 nm (<6 layers) and a lateral size of 490 ± 494 nm. Fe₂O₃ nanoparticles have an average size of 7.4 ± 3.0 nm. Fe₂O₃ nanoparticles on chemically and thermally stable MoS₂ nanosheets show catalytic activity in the glycolysis of polyethylene terephthalate (PET). High conversion of PET (97%) and a high yield (90%) for bis(hydroxyethyl) terephthalate (BHET) were achieved in a reaction time of 3 h at the reaction temperature of 225 °C.

---

Introduction

Two-dimensional (2D) nanosheets such as graphene, transition metal dichalcogenides (TMD), and hexagonal boron nitride (hBN) may be promising for heterogeneous catalyst supports^1–3 because of their outstanding mechanical strength⁴ and high surface area.⁵ Among 2D layered materials, MoS₂ nanosheets exhibit thermal and chemical stability,^6–8 and antioxidation properties⁹ and thus were used as 2D substrates for the 0D/2D-nanocomposite material with noble metal^10,11 and metal oxide.¹² Among these, the combination of MoS₂ nanosheets with iron oxide nanoparticles enhanced electrochemical, optical, and catalytic properties by chemical bonding-induced synergistic effect such as heterojunction on the 0D/2D-nanomaterial platform.^13–15 To synthesize iron oxide nanoparticles and MoS₂ nanosheets composites, various methods have been reported.^13–21 Hydrothermal method is the most common method to synthesize either iron oxide nanoparticles or their composite with MoS₂ nanosheets.^13–17 It can synthesize uniform iron oxide nanoparticles on the MoS₂ nanosheets but it requires high temperature and pressure with long reaction time and has limited scale-up capacity due to energy inefficiency. Organic phase synthesis is similar to hydrothermal method, and it needs heating under inert gas conditions with organic solvents.^18–20 Most of previous methods have needed heat treatment, organic solvents, or specific conditions such as inert gas in the synthesis of iron oxide nanoparticles on the MoS₂ nanosheets which hinder mass production and further applications.

Fluid dynamic system is one of the promising alternatives for mass-production of 0D/2D nanocomposite. Thus, synthesis of nanoparticles,²² growth of polymers,²³ and exfoliation of 2D materials^24,25 were achieved using fluid dynamic system. Among the fluid dynamic systems, Taylor–Couette flow reactor with high shear stress and mass transfer characteristics was used for the exfoliation of 2D materials such as graphite, hBN, and MoS₂.^24–26 Our group reported deposition of noble metal (Pt, Ag, and Pd) nanoparticles on exfoliated hBN nanosheets in a Taylor–Couette flow reactor.²⁷ High shear stress and mixing characteristics of Taylor–Couette flow reactor allowed the operation under mild reaction conditions in short reaction time.

Polyethylene terephthalate, commonly called PET, is one of the most consumed polymers in more than 300 million tons annually due to its mechanical strength and thermoplastic-based property.^28,29 Enormous consumption of PET in a daily life causes serious environmental problems due to poor biodegradability of PET.^30,31 To solve the problem, the chemical recycling methods such as alcoholysis,³² aminolysis,³³ hydrolysis,³⁴ and glycolysis³⁵ were investigated. Among these, glycolysis of PET decomposes PET using ethylene glycol (EG) to produce bis(2-hydroxyethyl) terephthalate (BHET) monomers which can be used to synthesize polymers again providing the sustainability in the polymer ecosystem. The major drawback of PET glycolysis is slow reaction despite high temperature.³⁶ Heavy metal salts, known as trans-esterification catalysts, were used as catalysts for PET glycolysis to solve the problem.^37,38 Although the reaction temperature was lowered to around 200 °C, there was still an environmental problem due to the presence of heavy metals in recycled BHET monomers. To replace heavy metal salts, eco-friendly ionic liquids were used in PET glycolysis but their high cost was a concern.^39,40 Because of environmentally abundant and eco-friendly characteristics,^35,41,42 the use of cobalt oxide,⁴³ iron oxide,⁴⁴ and manganese oxide⁴⁵ was reported in the glycolysis of PET. However, needs for organic solvent-based processes and calcination make the synthetic processes of oxide material still formidable. Therefore, large-scale manufacturing of an effective catalyst material for PET recycling should be practiced in a more economical and eco-friendly routes.

Here, we report a fast and simple hydrodynamic synthesis method without heating for the synthesis of Fe₂O₃ nanoparticles on exfoliated MoS₂ nanosheets in the form of 0D/2D-nanocomposites in a Taylor–Couette flow reactor. To the best of our knowledge, synthesis of Fe₂O₃@MoS₂ 0D/2D nanocomposites without heating or organic solvents has not been reported. The synthesis of Fe₂O₃@MoS₂ nanocomposite was achieved in an hour without organic solvents. Fe₂O₃ nanoparticle is preferable for a catalyst because it is the most stable among iron oxides. However, it usually needs additional treatment such as calcination to convert Fe₂O₃ from Fe₃O₄. Fluid dynamic process enabled the direct synthesis of Fe₂O₃ nanoparticles on MoS₂ nanosheets. The synthetic Fe₂O₃@MoS₂ nanocomposite was used to validate its catalytic activity in PET glycolysis.

Experimental procedure

Materials and chemicals

Iron(III) chloride hexahydrate powder (FeCl₃·6H₂O, 99%), molybdenum disulfide powder (MoS₂, <2 μm), sodium hydroxide (NaOH) powder, ethylene glycol (EG, 99.8%), and bis(2-hydroxyethyl) terephthalate (BHET) powder were purchased from Sigma-Aldrich Corporation (USA). Polyethylene terephthalate powder (PET, <200 μm, weight average molecular weight: 56 780 g mol⁻¹) was obtained from Goodfellow Corporation (UK).

Preparation of Fe₂O₃@MoS₂ nanocomposite

The schematic for the synthesis of Fe₂O₃@MoS₂ nanocomposite using the Taylor–Couette flow reactor is illustrated in Fig. 1a. Pristine MoS₂ powder (1000 mg) was initially dispersed in deionized (DI) water at a concentration of 2 mg mL⁻¹ in a 500 mL volume. Next, FeCl₃ powder (181 mg–3940 mg) was dispersed in MoS₂ aqueous mixture at various Fe : Mo molar ratio from 1 : 9 to 7 : 3. The MoS₂/FeCl₃ mixtures were reacted for 1 h in the Taylor–Couette flow reactor with 2000 rotations per minute (rpm) of inner cylinder under room temperature. At the end of the reaction, 10 mL of 1 M NaOH aqueous solution was added to the Taylor–Couette flow reactor and reacted for 5 more minutes to convert FeCl₃ to Fe₂O₃ nanoparticles. To compare the effect of mass transfer in the size distribution of Fe₂O₃ nanoparticles, 3 different Fe₂O₃@MoS₂ nanocomposites with Fe : Mo molar ratio of 3 : 7 were synthesized at 3 different rotation speeds (500, 1500 and 2000 rpm) of inner cylinder in the Taylor–Couette flow reactor. Fe₂O₃@MoS₂ nanocomposite dispersed in the solution is shown in Fig. 1b. Synthetic Fe₂O₃@MoS₂ nanocomposite was centrifuged at 3200g for 1 h and separated from un-exfoliated MoS₂ followed by vacuum filtration using a cellulose nitrate filter (pore size: 100 nm) to extract extra ions (Na⁺ and Cl⁻) from the supernatant. The filtered Fe₂O₃@MoS₂ nanocomposite was re-dispersed in 200 mL DI water and freeze-dried at −50 °C with 20 Pa to obtain Fe₂O₃@MoS₂ nanocomposite powder at a rate higher than 1000 mg h⁻¹ as shown in Fig. 1c. Detailed specification of the Taylor–Couette reactor was denoted in Fig. S1.†

Fig. 1 (a) Schematic illustrations of the synthesis process of Fe₂O₃@MoS₂ nanocomposite by fluid dynamic exfoliation and deposition, (b) produced Fe₂O₃@MoS₂ nanocomposite in solution, (c) Fe₂O₃@MoS₂ nanocomposite powder.

Characterization

The spherical aberration transmission electron microscopy (Cs-TEM) images were obtained by Titan3™ G2 60-300 (FEI). Transmission electron microscopy (TEM), scanning transmission electron microscopy (STEM), fast Fourier transform (FFT) diffraction pattern, and energy-dispersive X-ray spectroscopy (EDS) analysis were carried out by Tecnai G² F30 S-TWIN (FEI) operating at 300 kV. Raman spectroscopy data were obtained using LabRAM HR Evolution Visible_NIR (HORIBA Co). Atomic force microscopy (AFM) data were obtained with INNOVA-LABRAM HR800 (Bruker Daltonik Co). X-ray photoelectron spectroscopy (XPS) data were acquired with a Kα (Thermo VG Scientific Co). Inductively coupled plasma-optical emission spectrometry (ICP-OES) analysis was carried out by Agilent ICP-OES 720 (Agilent Technologies). High performance liquid chromatography (HPLC) data were obtained by micrOTOF-QII (Bruker Daltonik Co) at 200–1000 parts per million (ppm) concentration in tetra hydro furan (THF). The thermogravimetric analysis (TGA)-differential scanning calorimetry (DSC) curves were obtained by LABSYS Evo (Setaram Co). Brunauer–Emmett–Teller (BET) analysis was carried out with Tristar II 3020 (Micromeritics Company). The ¹H and ¹³C nuclear magnetic resonance (NMR) spectra analyses were carried out with Agilent 400 MHz 54 mm NMR DD2 (Agilent Technologies) at room temperature. The chemical shifts were denoted in ppm using dimethyl sulfoxide-d₆ (DMSO-d₆) signal as a standard.

Glycolysis of PET

The condition for the glycolysis of PET was based on a previous report using iron oxide as a catalyst.⁴⁴ Prepared Fe₂O₃@MoS₂ nanocomposite (20 mg) was dispersed in 8 g of EG, and the mixture was placed in a Teflon reactor along with 2 g of PET powder (Goodfellow, <200 μm). The reactants in the Teflon reactor were heated in a furnace between 200 and 300 °C for 3 to 5 h to examine the effect of temperature and time on glycolysis. Reacted solution was cooled to 70 °C in an oven. Vacuum filtration was performed to separate unreacted PET and catalyst from the product. The filtered solution was cooled in a refrigerator at 4 °C for 24 h. BHET was filtered by vacuum filtration followed by washing with cool DI water 3 times to remove extra EG. Mass of BHET was calculated by conventional HPLC method with a calibration curve in the range of 200 to 1000 ppm.

Conversion of PET and the yield of BHET were calculated using eqn (1) and (2):²⁷

In the equations above, W_PET represents the initial weight of PET, and W is the weight of PET after reaction. M_PET is the molecular weight of the repeated unit of PET. W_BHET is the weight of BHET based on the HPLC calibration curve, and M_BHET is the molecular weight of BHET. Fe₂O₃@MoS₂ nanocomposite was recovered by vacuum filtration method. Due to more than 97% conversion of PET, the filtered precipitate was used for recycling as a recovered catalyst.

Results and discussion

Morphology of Fe₂O₃@MoS₂ nanocomposite

Morphology and crystalline structure of MoS₂ nanosheets were investigated using AFM and TEM. AFM image of MoS₂ nanosheets is presented in Fig. 2a. AFM data of 70 samples of Fe₂O₃@MoS₂ nanocomposite were used to determine the average thickness and average lateral size (Fig. S2†). MoS₂ nanosheets had an average lateral size of 493.19 ± 494.50 nm and an average thickness of 2.65 ± 2.38 nm, indicating successful exfoliation of MoS₂ (Fig. 2b and c). Cs-TEM and EDS analysis were performed to confirm the deposition of Fe₂O₃ nanoparticles on MoS₂ nanosheets. Fig. 2d is the Cs-TEM image of Fe₂O₃@MoS₂ nanocomposite. Specific lattice structures of iron oxide(III), denoted by red lines, were distinguished from the hexagonal lattice of the surrounding MoS₂. The characteristic lattice lengths of Fe₂O₃ and MoS₂ were 2.5 Å for (110) and 2.7 Å for (100), respectively.¹⁴ EDS analysis of the same sample was performed to confirm the deposition of Fe₂O₃ nanoparticles on the MoS₂ nanosheet (Fig. 2e). The yellow signal of sulfur and the purple signal of molybdenum indicate MoS₂ in the STEM image. The blue signal of Fe on the MoS₂ indicates successful synthesis of Fe₂O₃@MoS₂ nanocomposite.

Fig. 2 (a) AFM image of Fe₂O₃@MoS₂ nanocomposite. (b) Lateral size and (c) thickness distribution of Fe₂O₃@MoS₂ nanocomposite. (d) Cs-TEM image of Fe₂O₃@MoS₂ nanocomposite. (e) STEM image and elemental mapping images of sulfur, molybdenum, and iron of Fe₂O₃@MoS₂ nanocomposite.

BET surface area analysis was performed to further investigate the surface area of Fe₂O₃@MoS₂ nanocomposite. Fig. S3† shows the BET surface area data of pristine MoS₂, MoS₂ nanosheets, and Fe₂O₃@MoS₂ nanocomposite. In Fig. S3a, c and e,† BET surface area data show typical isotherm for 2D materials. The surface area of pristine MoS₂ and MoS₂ nanosheets was 3.18 m² g⁻¹ and 9.33 m² g⁻¹, respectively in Fig. S3b and d.† Fe₂O₃@MoS₂ nanocomposite had a surface area of 47.95 m² g⁻¹ and a higher slope isotherm due to Fe₂O₃ nanoparticles (Fig. S3f†) which are consistent with previous reports despite being a top-down method.⁴⁴ Since it is a 0D/2D hybrid material, the surface area per unit mass is smaller than that of Fe₂O₃ nanoparticles. Even with a less surface area, MoS₂ nanosheets represent a stable substrate to prevent aggregation of Fe₂O₃ nanoparticles and its lateral size facilitates the recovery of Fe₂O₃@MoS₂ nanocomposite after reaction.

Size distribution of Fe₂O₃ nanoparticles depending on flow regime in the Taylor–Couette flow reactor

TEM analysis was used to compare the distribution of Fe₂O₃ nanoparticles in Fe₂O₃@MoS₂ nanocomposite synthesized at various Fe : Mo ratios from 1 : 9 to 7 : 3. Fig. S4a and b† show Fe₂O₃@MoS₂ nanocomposite synthesized with Fe : Mo ratio of 1 : 9. Fe₂O₃ nanoparticles are rare on MoS₂ nanosheets in high resolution image. Fe₂O₃@MoS₂ nanocomposite synthesized with Fe : Mo ratio of 3 : 7 shows relatively more Fe₂O₃ nanoparticles on MoS₂ nanosheets in Fig. S4c and d.† The Fe₂O₃@MoS₂ nanocomposite with Fe : Mo ratio of 5 : 5 shows fairly dense Fe₂O₃ nanoparticles on MoS₂ nanosheets (Fig. S4e and f†). For Fe₂O₃@MoS₂ nanocomposite with Fe : Mo ratio of 7 : 3, Fe₂O₃ nanoparticles are distributed more densely on MoS₂ nanosheets than that with Fe : Mo ratio of 5 : 5 (Fig. S4g and h†). As expected, amount of Fe₂O₃ nanoparticles on MoS₂ nanosheets increased as ratio of Fe increased.

TEM images and FFT diffraction patterns of Fe₂O₃@MoS₂ nanocomposite synthesized at different rotation speeds of inner cylinder in the Taylor–Couette flow reactor are shown in Fig. 3. Fe₂O₃ nanoparticles are distributed well on MoS₂ nanosheets with the characteristic hexagonal FFT diffraction patterns marked with yellow circles. Flow regimes in the Taylor–Couette flow reactor varied depending on rotation speed of inner cylinder as laminar, wavy, and turbulent regimes in the order of increasing rotation speed. Previously, our group reported efficiency of exfoliation for 2D materials in various flow regimes of Taylor–Couette flow and found about 2000 rpm corresponding to wavy regime is optimum for the exfoliation of 2D materials.^24,25 In order to compare the synthesis of Fe₂O₃ nanoparticles at different rpm, Fe₂O₃@MoS₂ nanocomposite was synthesized in the Taylor–Couette flow reactor at various rotation speeds of inner cylinder. Fig. 3a and b show Fe₂O₃@MoS₂ nanocomposite synthesized at 500 rpm in the Taylor–Couette flow reactor. According to the TEM image and size distribution data, Fe₂O₃ nanoparticles showed an average size of about 13.73 ± 3.82 nm. As rpm increased, the average size of Fe₂O₃ nanoparticles decreased to 11.79 ± 4.21 nm at 1500 rpm of rotation speed (Fig. 3c and d). Uniformity of the Fe₂O₃ nanoparticles was improved at 2000 rpm of rotation speed, and the average size of Fe₂O₃ nanoparticles was also reduced to 7.41 ± 2.95 nm (Fig. 3e and f).

Fig. 3 Size distributions of Fe₂O₃ nanoparticles and TEM images of Fe₂O₃@MoS₂ nanocomposite synthesized at (a) 500 rpm, (c) 1500 rpm, and (e) 2000 rpm in the Taylor–Couette flow reactor, and their magnified TEM images with FFT diffraction patterns of Fe₂O₃@MoS₂ nanocomposite synthesized at (b) 500 rpm, (d) 1500 rpm, and (f) 2000 rpm.

Even in the top down method without heating, Fe₂O₃ nanoparticles were distributed well on MoS₂ nanosheets. Based on the TEM image, the size distribution of Fe₂O₃ nanoparticles on MoS₂ nanosheets was obtained, showing average size of the nanoparticles as 7.41 nm and mostly less than 10 nm at 2000 rpm of rotation speed. Fe₂O₃ nanoparticles usually require thermal treatment for their synthesis.^13–15,17 However, in our study, no thermal treatment was conducted. Increased mixing characteristics with increasing rpm of Taylor–Couette flow reactor⁴⁶ made Fe₂O₃ nanoparticles synthesized on MoS₂ nanosheets in a nanometer scale without thermal treatment. In addition, TEM grid was observed below the transparent MoS₂ nanosheets, indicating adequate exfoliation of MoS₂.

Chemical composition and synthetic mechanism of Fe₂O₃@MoS₂ nanocomposite

Raman spectroscopy and XPS were used to examine the chemical state of Fe₂O₃@MoS₂ nanocomposite by measuring its response to photons. E¹_2g mode reflects the in-plane vibration, and the A_1g mode reflects the out-plane vibration of MoS₂.⁴⁸ Fig. 4a shows the Raman spectroscopy data of Fe₂O₃@MoS₂ nanocomposite and pristine MoS₂, and peaks for MoS₂ and Fe₂O₃ were observed. In Fig. 4b, the E¹_2g mode of Fe₂O₃@MoS₂ nanocomposite was located at 375.7 cm⁻¹, while the E¹_2g mode of pristine MoS₂ was located at 379.9 cm⁻¹. Also the A_1g mode of Fe₂O₃@MoS₂ nanocomposite was located at 401.4 cm⁻¹, and A_1g mode of pristine MoS₂ was located at 404.2 cm⁻¹. A_1g mode of Fe₂O₃@MoS₂ nanocomposite were shifted less than E¹_2g mode. According to previous study, it means that out-plane of MoS₂ was affected by the external factor,⁴⁷ presumably Fe₂O₃ nanoparticles, which are bonded to the basal plane (out-plane) of MoS₂.

Fig. 4 (a) Raman spectroscopy data of pristine MoS₂ and Fe₂O₃@MoS₂ nanocomposite, (b) and its magnified data from 340 to 500 cm⁻¹.

XPS analysis was performed to confirm the interaction, and chemical composition of Fe₂O₃ nanoparticles and MoS₂ nanosheets.^20,25,42 Survey scan data show that both iron oxide and molybdenum disulfide exist in Fe₂O₃@MoS₂ nanocomposite (Fig. 5a). In Fig. 5b and c, Fe2p and O1s scan were used to compare the synthesis of Fe₂O₃ nanoparticles on MoS₂ and other 2D substrates. Since FeCl₃ with Fe³⁺ was used as the precursor, the Fe2p scan of all experimental groups shows the typical Fe³⁺ and its satellite peaks at 710.7 eV, and 719.2 eV, respectively, which are similar to pristine Fe₂O₃.⁴⁴ O–Fe bond and O–H bond in O1s scan represent Fe₂O₃ and Fe(OH)₃, respectively, and the ratio between O–H bond and O–Fe bond depends on the kind of substrates. The ratio of O–Fe bond to O–H bond was 70 : 30 in Fe₂O₃@MoS₂ nanocomposite. On the other hand, the ratios of O–Fe bond to O–H bond were 46 : 54 and 37 : 63 in graphite and hBN, respectively, with the Fe³⁺ precursor. The ratio of O–Fe bond to O–H bond was highest in Fe₂O₃@MoS₂ nanocomposite, and these results signify that, Fe₂O₃ is converted successfully from FeCl₃ with NaOH on MoS₂ substrate better than on graphite or hBN in the Taylor–Couette flow reactor. Previous research has reported that the interface between Fe₂O₃, and MoS₂ has a lower formation energy of −1.76 eV for chemical bonding.^49,50 To further confirm the chemical bonding between MoS₂ and Fe₂O₃, the binding energy of Fe₂O₃@MoS₂ nanocomposite was compared with the pristine Fe₂O₃ and pristine MoS₂. Fe2p scan data of Fe₂O₃@MoS₂ nanocomposite reveal the peak for Fe³⁺ at 711.7 eV, red shifted by 1 eV compared to pristine Fe₂O₃. On the other hand, Mo3d and S2p scans of the Fe₂O₃@MoS₂ nanocomposite show 0.7 eV and 0.2 eV blue shifted binding energies, respectively, compared to pristine MoS₂ in Fig. 5d and e. The shifts of binding energy in Fe₂O₃@MoS₂ nanocomposite indicate a chemical interaction between Fe₂O₃ nanoparticles and MoS₂ nanosheets as shown by Raman spectroscopy data. XPS and Raman spectroscopy data of Fe₂O₃@MoS₂ nanocomposite show that even in a fluid dynamic-based production method without heating, Fe₂O₃ nanoparticles were bound to MoS₂ nanosheets by chemical bonds, and MoS₂ nanosheet is effective 2D substrates to incorporate Fe₂O₃ nanoparticles.

Fig. 5 XPS (a) survey scan of Fe₂O₃@MoS₂ nanocomposite, pristine Fe₂O₃, and pristine MoS₂, (b) Fe2p and (c) O1s scan of pristine Fe₂O₃, Fe₂O₃@MoS₂ nanocomposite, and Fe(OH)₃ with various substrate reacted in the Taylor–Couette flow reactor, (d) Mo3d and (e) S2p scan of Fe₂O₃@MoS₂ nanocomposite and pristine MoS₂.

In XPS analysis, atomic ratio of Fe and Mo was 3 : 6.6, which is similar to that of the precursor ratio of 3 : 7. Further compositional analysis was conducted by ICP-OES, which confirmed successful conversion of FeCl₃ to Fe₂O₃ nanoparticles. The ICP-OES data show atomic ratio of 3 : 6.3, which is almost identical to the atomic ratio between Fe and Mo in the XPS data (Table S1†).

Glycolysis of PET by Fe₂O₃@MoS₂ nanocomposite

Because glycolysis reaction occurs at the temperature above 175 °C, it is necessary to confirm the thermal stability of Fe₂O₃@MoS₂ nanocomposite. Fig. S5a and b† present the TGA data and derivative thermogravimetry (DTG) data, which showed no reduction below 400 °C. There was small increase of mass at 400 °C because of nitrogen adsorption followed by the reduction at high temperature. Since the glycolysis reaction was performed below 300 °C, Fe₂O₃@MoS₂ nanocomposite will be thermally stable during the reaction.

Previously suggested mechanism of PET glycolysis using Fe₂O₃ is shown in Fig. 6a.^44,51 The yield of BHET in PET glycolysis was measured by conventional method using HPLC,²⁷ and the comparison with previous reports is presented in Table S2.† Fig. 6b shows the conversion of PET with Fe₂O₃@MoS₂ nanocomposite having various Fe : Mo ratios. MoS₂ nanosheets show only 7% conversion in the glycolysis of PET, and the conversion of PET using Fe₂O₃@MoS₂ nanocomposite was found to be saturated in the Fe : Mo ratio of 3 : 7. Fig. 6c shows the conversion of PET with the varying ratio between PET and Fe₂O₃@MoS₂ nanocomposite. A 98% conversion of PET was reached at the ratio higher than 1/100. In Fig. 6d, the yield of BHET was maintained at around 90% up to 250 °C. However, the yield of BHET suddenly decreased at 275 °C and continued to decrease to almost 50% at 300 °C. To analyze the decrease in yield, the process was carried out at 300 °C for 3 to 5 h (Fig. 6e) and showed further decrease in yield to 38%. Previous studies have reported that decreased yield of BHET in glycolysis of PET is caused by chemical equilibrium, and HPLC data of glycolysis product at 300 °C show by-product peaks in HPLC data different from HPLC data of commercial BHET and produced BHET at 225 °C (Fig. S6a–c†). There would be side reactions between BHET and EG or oligomer of PET, reducing the yield of BHET from 50% to less than 40% as reaction time increased.³⁵

Fig. 6 (a) Suggested mechanism for glycolysis of PET using Fe₂O₃ nanoparticles as a catalyst,^44,51 (b) conversion of PET with MoS₂ nanosheets and various Fe : Mo ratio of Fe₂O₃@MoS₂ nanocomposite, (c) conversion of PET with various ratio of PET and Fe₂O₃@MoS₂ nanocomposite, yield and conversion (d) with 3 h reaction at different reaction temperature using Fe₂O₃@MoS₂ nanocomposite, (e) at 300 °C with 3 different reaction time, and (f) with repeated recycling of catalyst at 225 °C for 3 h.

In Fig. S7a,† Fe₂O₃@MoS₂ nanocomposite catalyst was recovered by simple filtration, and the recycling reaction with recycled catalyst was carried out at 225 °C for 3 h. During 7 recycles, the yield of BHET remained around 85% (Fig. 6f). TEM and ICP-OES were used to characterize the recovered Fe₂O₃@MoS₂ nanocomposite. Fig. S7b and c† show the morphology of recovered Fe₂O₃@MoS₂ nanocomposite. Even though Fe₂O₃@MoS₂ nanocomposite was recycled repeatedly during PET glycolysis, Fe₂O₃ nanoparticles are clearly distributed on MoS₂ nanosheets in the recovered Fe₂O₃@MoS₂ nanocomposite as fresh Fe₂O₃@MoS₂ nanocomposite. ICP-OES data in Table S3† also show that the composition of Mo and Fe in recycled Fe₂O₃@MoS₂ nanocomposite was maintained after the recycling. Although the catalytic yield and conversion rate are not superior to previous work, the production rate of Fe₂O₃@MoS₂ nanocomposite was 1120 mg h⁻¹, and is much higher than 0.2 mg h⁻¹ to 116.7 mg h⁻¹ of other reports and stable without decomposition or separation (Table S4†).

¹H and ¹³C NMR analyses were performed to confirm the purity of BHET as shown in Fig. 7a. NMR δ values were as follows: white solid, ¹H NMR (DMSO-d₆, 400 MHz, δ/ppm), δ = 8.09 (s, 4H), δ = 4.95 (t, 2H), δ = 4.28 (t, 4H), δ = 3.69 (q, 4H). ¹³C NMR (DMSO-d₆, 400 MHz, δ/ppm) data were: δ = 165.06, 133.62, 129.42, 66.93, 58.86. Thus, the characteristic peaks of BHET consist of 5 carbons and 4 hydrogens indicated as A to E and 1 to 4 in NMR spectra data. Fig. 7b and c show DSC and TGA data of BHET. DSC curve shows a peak at 107 °C, and the TGA data show double decreases at 240 °C and 420 °C, showing the same thermal properties as reported previously for BHET.^35,39 Produced BHET is intended to be used as a monomer to synthesize polymeric material, and fewer impurity in BHET is more desirable. ICP-OES data indicating the residual amounts of Fe and Mo in the extracted BHET are shown in Table S5.† Only small amount of Fe and Mo, 10⁻⁴ and 10⁻⁵ weight percent, respectively, was detected in the extracted BHET.

Fig. 7 (a) ¹H and ¹³C NMR spectra data for produced BHET, (b) DSC data of crystalized BHET, (c) TGA data of crystalized BHET.

In order to check whether the catalyst works effectively in the scaled-up process, the glycolysis reaction was performed with 100 g of PET. 100 g of PET was dispersed in 400 mL of EG with 1 g of Fe₂O₃@MoS₂ nanocomposite as shown in Fig. S8a.† The reaction was proceeded at 196 °C for 3 h with simple stirring under ambient condition. The product of glycolysis and produced BHET had identical brown and white color, respectively, with the previous product of glycolysis (Fig. S8b and c†), and HPLC data of product showed identical peaks with that of product from the smaller scale glycolysis of PET (Fig. S8d†). However, the peak of retention time at 18.37 is higher than that of commercial BHET (Fig. S6a†) which means higher contents of dimer. The yield and conversion were 87% and 94%, respectively. Therefore, Fe₂O₃@MoS₂ nanocomposite can be used in the increased scale of PET glycolysis and can be considered as an attractive catalyst in the recycling of large amount of PET by glycolysis to produce high quality BHET.

Conclusions

We reported a method for the synthesis of uniform Fe₂O₃ nanoparticles on exfoliated MoS₂ nanosheets due to the high mixing characteristics of the Taylor–Couette flow reactor. Exfoliated MoS₂ nanosheets were produced in an aqueous solution with a thickness of few nanometers. Fe₂O₃ nanoparticles were synthesized with a size below 7 nm without thermal or organic solvent treatment. Fe₂O₃@MoS₂ nanocomposite was synthesized at high production rate (1120 mg h⁻¹) under mild conditions. Synthesized Fe₂O₃@MoS₂ nanocomposite was used for glycolysis of PET and showed an approximately 90% yield of high quality BHET. The aggregation of Fe₂O₃ nanoparticles during the reaction was prevented by forming Fe₂O₃ nanoparticles on MoS₂ nanosheets. Used Fe₂O₃@MoS₂ nanocomposite was recovered by simple filtration of glycolysis product without decomposition or separation of Fe₂O₃ nanoparticles and MoS₂ nanosheets. Recovered Fe₂O₃@MoS₂ nanocomposite showed 94.4% catalytic activity of fresh Fe₂O₃@MoS₂ nanocomposite. Our method for the synthesis of Fe₂O₃@MoS₂ nanocomposite would be a promising strategy to produce large amount of catalyst necessary for large-scale catalytic processes.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. 2018R1A2A3075668) and the R&D Center for Valuable Recycling (Global-Top R&BD Program) of the Ministry of Environment (Project No. 2016002240001).

Notes and references

1. A. Kumar and Q. Xu, Chem. Nanostruct. Mater., 2018, 4, 28–40.
2. R. I. Jafri, N. Rajalakshmi and S. Ramaprabhu, J. Mater. Chem., 2010, 20, 7114–7117.
3. X. Cai, Y. Luo, B. Liu and H.-M. Cheng, Chem. Soc. Rev., 2018, 47, 6224–6266.
4. Z. Fu, H. Zhang, C. Si, D. Legut, T. C. Germann, Q. Zhang, S. Du, J. S. Francisco and R. Zhang, J. Phys. Chem. C, 2018, 122, 4710–4722.
5. H. Long, A. Harley-Trochimczyk, T. Pham, Z. Tang, T. Shi, A. Zettl, C. Carraro, M. A. Worsley and R. Maboudian, Adv. Funct. Mater., 2016, 26, 5158–5165.
6. Y.-C. Chen, A.-Y. Lu, P. Lu, X. Yang, C.-M. Jiang, M. Mariano, B. Kaehr, O. Lin, A. Taylor, I. D. Sharp, L.-J. Li, S. S. Chou and V. Tung, Adv. Mater., 2017, 29, 1703863.
7. N. Bandaru, R. S. Kumar, D. Sneed, O. Tschauner, J. Baker, D. Antonio, S.-N. Luo, T. Hartmann, Y. Zhao and R. Venkat, J. Phys. Chem. C, 2014, 118, 3230–3235.
8. L. Nurdiwijayanto, R. Ma, N. Sakai and T. Sasaki, Inorg. Chem., 2017, 56, 7620–7623.
9. H. S. Sen, H. Sahin, F. M. Peeters and E. Durgun, J. Appl. Phys., 2014, 116, 083508.
10. S. Najmaei, A. Mlayah, A. Arbouet, C. Girard, J. Leotin and J. Lou, ACS Nano, 2014, 8, 12682–12689.
11. Y. Cheng, S. Lu, F. Liao, L. Liu, Y. Li and M. Shao, Adv. Funct. Mater., 2017, 27, 1700359.
12. J. Hu, C. Zhang, Y. Zhang, B. Yang, Q. Qi, M. Sun, F. Zi, M. K. H. Leung and B. Huang, Small, 2020, 16, 2002212.
13. M. B. Lejbini and P. Sangpour, Optik, 2019, 177, 112–117.
14. X. Yang, H. Sun, L. Zhang, L. Zhao, J. Lian and Q. Jiang, Sci. Rep., 2016, 6, 31591.
15. C.-Y. Wu, W.-E. Chang, Y.-G. Sun, J.-M. Wu and J.-G. Duh, Mater. Chem. Phys., 2018, 219, 311–317.
16. Y. Chen, B. Song, X. Tang, L. Lu and J. Xue, Small, 2014, 10, 1536–1543.
17. N. D. Chuong, T. D. Thanh, N. H. Kim and J. H. Lee, ACS Appl. Mater. Interfaces, 2018, 10, 24523–24532.
18. T. Liu, S. Shi, C. Liang, S. Shen, L. Cheng, C. Wang, X. Song, S. Goel, T. E. Barnhart, W. Cai and Z. Liu, ACS Nano, 2015, 9, 950–960.
19. Y. A. Kabachii, A. S. Golub, S. Y. Kochev, N. D. Lenenko, S. S. Abramchuk, M. Y. Antipin, P. M. Valetsky, B. D. Stein, W. E. Mahmoud, A. A. Al-Ghamdi and L. M. Bronstein, Chem. Mater., 2013, 25, 2434–2440.
20. V. Nandwana, W. Huang, Y. Li and V. P. Dravid, ACS Appl. Nano Mater., 2018, 1, 1949–1958.
21. A. S. K. Kumar, S.-J. Jiang and J. K. Warchol, ACS Omega, 2017, 2, 6187–6200.
22. A. K. Yadav, M. J. Barandiaran and J. C. Cal, Chem. Eng. J., 2012, 198–199, 191–200.
23. A. Igder, S. Pye, A. H. M. Al-Antaki, A. Keshavarz, C. L. Raston and A. Nosrati, RSC Adv., 2020, 10, 14761–14767.
24. D. Seo, W.-S. Kim and D. H. Kim, Chem. Eng. J., 2020, 399, 125726.
25. J.-M. Jeong, H. G. Kang, H.-J. Kim, S. B. Hong, H. Jeon, S. Y. Hwang, D. Seo, B. E. Kwak, Y.-K. Han, B. G. Choi and D. H. Kim, Adv. Funct. Mater., 2018, 28, 1802952.
26. H. Jeon, J.-M. Jeong, H. G. Kang, H.-J. Kim, J. Park, D. H. Kim, Y. M. Jung, S. Y. Hwang, Y.-K. Han and B. G. Choi, Adv. Energy Mater., 2018, 8, 1800227.
27. J.-M. Jeong, S. B. Jin, H. J. Park, S. H. Park, H. Jeon, H. Suh, Y.-J. Park, D. Seo, S. Y. Hwang, D. H. Kim and B. G. Choi, Adv. Mater. Interfaces, 2020, 7(16), 2000599.
28. X. Zhang, M. Fevre, G. O. Jones and R. M. Waymouth, Chem. Rev., 2018, 118, 839–885.
29. A. Rahimi and J. M. García, Nat. Rev. Chem., 2017, 1, 0046.
30. J. Zhang, X. Wang, J. Gong and Z. Gu, J. Appl. Polym. Sci., 2004, 93, 1089–1096.
31. F. Awaja and D. Pavel, Eur. Polym. J., 2005, 41, 1453–1477.
32. J. Chen, J. Lv, Y. Ji, J. Ding, X. Yang, M. Zou and L. Xing, Polym. Degrad. Stab., 2014, 107, 178–183.
33. S. R. Shukla and A. M. Harad, Polym. Degrad. Stab., 2006, 91, 1850–1854.
34. E. H. Acero, D. Ribitsch, G. Steinkellner, K. Gruber, K. Greimel, I. Eiteljoerg, E. Trotscha, R. Wei, W. Zimmermann, M. Zinn, A. Cavaco-Paulo, G. Freddi, H. Schwab and G. Guebitz, Macromolecules, 2011, 44, 4632–4640.
35. M. Imran, D. H. Kim, W. A. Al-Masry, A. Mahmood, A. Hassan, S. Haider and S. M. Ramay, Polym. Degrad. Stab., 2013, 98, 904–915.
36. A. B. Raheem, Z. Z. Noor, A. Hassan, M. K. A. Hamid, S. A. Samsudin and A. H. Sabeen, J. Cleaner Prod., 2019, 225, 1052–1064.
37. D. E. Nikles and M. S. Farahat, Macromol. Mater. Eng., 2005, 290, 13–30.
38. N. George and T. Kurian, Ind. Eng. Chem. Res., 2014, 53, 14185–14198.
39. A. M. Al-Sabagh, F. Z. Yehia, A. M. F. Eissa, M. E. Moustafa, G. Eshaq, A. M. Rabie and A. E. ElMetwally, Polym. Degrad. Stab., 2014, 110, 364–377.
40. A. M. Al-Sabagh, F. Z. Yehia, A.-M. M. F. Eissa, M. E. Moustafa, G. Eshaq, A.-R. M. Rabie and A. E. ElMetwally, Ind. Eng. Chem. Res., 2014, 53, 18443–18451.
41. M. Imran, K. G. Lee, Q. Imtiaz, B.-K. Kim, M. Han, B. G. Cho and D. H. Kim, J. Nanosci. Nanotechnol., 2011, 11, 824–828.
42. I. Cano, C. Martin, J. A. Fernandes, R. W. Lodge, J. Dupont, F. A. Casado-Carmona, R. Lucena, S. Cardenas, V. Sans and I. Pedro, Appl. Catal., B, 2020, 260, 118110.
43. F. R. Veregue, C. T. P. Silva, M. P. Moisés, J. G. Meneguin, M. R. Guilherme, P. A. Arroyo, S. L. Favaro, E. Radovanovic, E. M. Girotto and A. W. Rinaldi, ACS Sustainable Chem. Eng., 2018, 6, 12017–12024.
44. L. Bartolome, M. Imran, K. G. Lee, A. Sangalang, J. K. Ahn and D. H. Kim, Green Chem., 2014, 16, 279–286.
45. G. Park, L. Bartolome, K. G. Lee, S. J. Lee, D. H. Kim and T. J. Park, Nanoscale, 2012, 4, 3879–3885.
46. M. Nemri, S. Charton and E. Climent, Chem. Eng. Sci., 2016, 139, 109–124.
47. A. S. George, Z. Mutlu, R. Ionescu, R. J. Wu, J. S. Jeong, H. H. Bay, Y. Chai, K. A. Mkhoyan, M. Ozkan and C. S. Ozkan, Adv. Funct. Mater., 2014, 24, 7461–7466.
48. S. Wang, X. Wang and J. H. Warner, ACS Nano, 2015, 9, 5246–5254.
49. H. Pan, X. Meng, X. Qi and G. Qin, CrystEngComm, 2017, 19, 6333–6338.
50. Y. Zhao, W. Cai, Y. Shi, J. Tang, Y. Gong, M. Chen and Q. Zhong, ACS Sustainable Chem. Eng., 2020, 8, 12603–12611.
51. A. M. Al-Sabagh, F. Z. Yehia, D. R. K. Harding, G. Eshaq and A. E. ElMetwally, Green Chem., 2016, 18, 3997–4003.

---

Footnote

† Electronic supplementary information (ESI) available: Dimension of the Taylor–Couette flow reactor, AFM image of MoS₂ nanosheets, BET surface area of pristine MoS₂, MoS₂ nanosheets, and Fe₂O₃@MoS₂ nanocomposite, TEM image of Fe₂O₃@MoS₂ nanocomposite with different initial Fe : Mo ratio, TGA for Fe₂O₃@MoS₂ nanocomposite, HPLC data for glycolysis of PET, recovered Fe₂O₃@MoS₂ nanocomposite, comparison table for production of Fe₂O₃@MoS₂ nanocomposite and previous reports, ICP-OES for Fe₂O₃@MoS₂ nanocomposite, and BHET, comparison table for glycolysis reaction using Fe₂O₃@MoS₂ nanocomposite and previous reports. See DOI: 10.1039/d1ra02335g

---