*2.3.1. Synthesis of superparamagnetic γ-Fe2O3 catalyst and characterization*

Slight modifications to the conventional co-precipitation method were done to obtain γ-Fe2O3 nanoparticles from Fe3O4 [45,46]. For the synthesis of Fe3O4 nanoparticles, predetermined amounts of FeCl2 and FeCl3 precursors were dissolved in an HCl solution (0.4 M, 25 mL) to prepare aqueous solutions of Fe2+/Fe3+ with 1:2 molar ratio. Sodium hydroxide (1.5 M, 250 mL) was then added rapidly and stirred. The addition of NaOH solution instantly produced a black precipitate, characteristic of Fe3O4. The surfactant used to control the particle size was citric acid (0.2 M, 50 mL). The synthesis was performed in a nitrogen atmosphere. The formed black precipitate was separated by placing a magnet and decanting the solution. The product was washed with water four times and centrifuged at 4000 rpm for 4 min. The final washing step was done using a 0.01 M HCl solution in order to neutralize the anionic charges on the nanoparticle surface. Calcination of the dried Fe3O4 powder at 210°C for 3 h induced the phase transformation into γ-Fe2O3 [12].

The particle morphology of the synthesized catalyst was observed by a 200 kV transmission electron microscope (TEM). As shown in Figure 11a, the particles had a size distribution in the range of 8–14 nm, with mean size of 10 nm. The BET surface area measured was 147 m<sup>2</sup> /g. The superparamagnetic property of the nanosized γ-Fe2O3 was confirmed by the magnetization curve obtained using a vibrating sample magnetometer (VSM). Figure 11b shows the magnetic behavior in the presence of a magnetic field, exhibiting a strong response with saturation magnetization reaching 47 emu/g. The curve does not have a hysteresis loop and retains no magnetization when the magnetic field is removed. The advantageous consequence of this property is redispersability of the catalyst particles when used in subsequent reactions [12].

The obtained X-ray diffraction (XRD) spectra of the catalyst suggest that the material is γ-Fe2O3. This is not conclusive, however, because the XRD patterns of Fe3O4 and γ-Fe2O3 are very similar. X-ray photoelectron spectroscopy (XPS) analysis of Fe2p cores was performed to distinguish the two phases. Higher binding energies before the calcination as shown in Figure 11c are indicative of Fe3O4. The shift to lower binding energies after the calcination step confirms the transformation of Fe3O4 to γ-Fe2O3 [12].

## *2.3.2. Catalytic activity, recoverability, and stability with repeated use*

The catalytic performance of the synthesized γ-Fe2O3 nanoparticles was compared to previously studied silica nanoparticle-supported metal oxide catalysts [10,40]. Under the same reaction conditions and catalyst/PET weight ratio, γ-Fe2O3 delivered comparable performance (Figure 12). The BHET yield reached higher than 90% in 70 min at 1.0% catalystto-PET loading. As with the supported nanocatalysts, the excellent catalytic performance of the γ-Fe2O3 nanoparticles may be attributed to the high surface area and greater accessibili‐ ty to active sites [12].

zero remanent magnetization. Iron oxides have further advantages being cheap, nontoxic, and

Slight modifications to the conventional co-precipitation method were done to obtain γ-Fe2O3 nanoparticles from Fe3O4 [45,46]. For the synthesis of Fe3O4 nanoparticles, predetermined amounts of FeCl2 and FeCl3 precursors were dissolved in an HCl solution (0.4 M, 25 mL) to prepare aqueous solutions of Fe2+/Fe3+ with 1:2 molar ratio. Sodium hydroxide (1.5 M, 250 mL) was then added rapidly and stirred. The addition of NaOH solution instantly produced a black precipitate, characteristic of Fe3O4. The surfactant used to control the particle size was citric acid (0.2 M, 50 mL). The synthesis was performed in a nitrogen atmosphere. The formed black precipitate was separated by placing a magnet and decanting the solution. The product was washed with water four times and centrifuged at 4000 rpm for 4 min. The final washing step was done using a 0.01 M HCl solution in order to neutralize the anionic charges on the nanoparticle surface. Calcination of the dried Fe3O4 powder at 210°C for 3 h induced the phase

The particle morphology of the synthesized catalyst was observed by a 200 kV transmission electron microscope (TEM). As shown in Figure 11a, the particles had a size distribution in the range of 8–14 nm, with mean size of 10 nm. The BET surface area measured was 147 m<sup>2</sup>

superparamagnetic property of the nanosized γ-Fe2O3 was confirmed by the magnetization curve obtained using a vibrating sample magnetometer (VSM). Figure 11b shows the magnetic behavior in the presence of a magnetic field, exhibiting a strong response with saturation magnetization reaching 47 emu/g. The curve does not have a hysteresis loop and retains no magnetization when the magnetic field is removed. The advantageous consequence of this property is redispersability of the catalyst particles when used in subsequent reactions [12].

The obtained X-ray diffraction (XRD) spectra of the catalyst suggest that the material is γ-Fe2O3. This is not conclusive, however, because the XRD patterns of Fe3O4 and γ-Fe2O3 are very similar. X-ray photoelectron spectroscopy (XPS) analysis of Fe2p cores was performed to distinguish the two phases. Higher binding energies before the calcination as shown in Figure 11c are indicative of Fe3O4. The shift to lower binding energies after the calcination step

The catalytic performance of the synthesized γ-Fe2O3 nanoparticles was compared to previously studied silica nanoparticle-supported metal oxide catalysts [10,40]. Under the same reaction conditions and catalyst/PET weight ratio, γ-Fe2O3 delivered comparable performance (Figure 12). The BHET yield reached higher than 90% in 70 min at 1.0% catalystto-PET loading. As with the supported nanocatalysts, the excellent catalytic performance of the γ-Fe2O3 nanoparticles may be attributed to the high surface area and greater accessibili‐

/g. The

*2.3.1. Synthesis of superparamagnetic γ-Fe2O3 catalyst and characterization*

152 Advanced Catalytic Materials - Photocatalysis and Other Current Trends

abundant [12].

transformation into γ-Fe2O3 [12].

confirms the transformation of Fe3O4 to γ-Fe2O3 [12].

ty to active sites [12].

*2.3.2. Catalytic activity, recoverability, and stability with repeated use*

**Figure 11.** (a) Particle morphology, (b) superparamagnetic property, and (c) phase identification of the synthesized γ-Fe2O3 nanoparticles. Figure from [12].

**Figure 12.** Catalytic performance of γ-Fe2O3 nanoparticles at various loadings and comparison to other metal oxide cat‐ alysts. Figure from [12].

The main potential of using γ-Fe2O3 as candidate for industrial glycolysis catalyst is the easy separation method and stability. In addition to the superparamagnetic property of γ-Fe2O3 nanoparticles that was beneficial to catalytic performance and separation, its stability with repeated use was also successfully demonstrated. The catalyst was reused in 10 reaction repeat cycles and did not significantly affect the BHET monomer yield, as shown in Figure 13. The thermal stability was supported by thermogravimetric analysis (TGA), while the XRD spectra of the used catalyst also proved phase stability [12]. These demonstrate the robust character‐ istics of the catalyst, withstanding repeated use at elevated temperature and pressure without deterioration in performance. Given comparable performance to a homogenous catalyst as shown in Figure 13b, the superparamagnetic catalyst provides a more practical approach to catalyst separation by application of a magnetic field.

**Figure 13.** Assessment of catalyst stability with repeated use and comparison to a recoverable homogeneous catalyst at the same reaction conditions. Figure from [12].

## **2.4. Mesoporous spinel oxide catalysts**

Due to several advantages such as high mechanical strength, possibility of regeneration, easier separation, and robust process integration, metal oxides can be considered superior to conventional PET glycolysis catalysts. Moreover, there are numerous possibilities to tailor their physical and chemical properties for the desired catalytic performance and functionality [47]. For the enhancement of catalytic activity, for example, altering the metal composition by introduction of another metal could result in higher catalytic activity [48–50]. The same principle is used to potentially enhance the performance of metal oxide catalysts for PET glycolysis. In this section, pure oxides and mixed-oxide spinel oxides of zinc, manganese, and cobalt were synthesised by simple precipitation or co-precipitation methods.

Simple precipitation method was used to synthesize the pure metal oxides. A 1.0 M solution of the salt precursors (Mn(NO3)2 xH2O, Zn(NO3)2 6H2O, or Co(NO3)2 6H2O) was mixed with 0.1 M ammonium hydroxide to set the pH at 9.0. Precipitates of the corresponding metal hydroxides formed, which were filtered, washed with water, and dried at 100°C for 8 h. The oxide form was obtained by calcination of the dried powder at 600°C for 4 h. For the mixed metal oxides, a modified co-precipitation method was implemented [51–53]. The molar ratios of the metal precursors were fixed to be 1:2. Similar to the previous synthesis, a 0.1 M ammo‐ nium hydroxide solution was stirred into the bimetallic precursor solutions, setting the pH value at 9.0. The same procedures for filtering, washing, drying, and calcining were performed on the mixed metal oxides [13]. The physical and chemical properties of the synthesized catalysts are summarized in Table 3. The analysis of atomic composition of the oxides via EDS was performed with results shown in Table 4.

The main potential of using γ-Fe2O3 as candidate for industrial glycolysis catalyst is the easy separation method and stability. In addition to the superparamagnetic property of γ-Fe2O3 nanoparticles that was beneficial to catalytic performance and separation, its stability with repeated use was also successfully demonstrated. The catalyst was reused in 10 reaction repeat cycles and did not significantly affect the BHET monomer yield, as shown in Figure 13. The thermal stability was supported by thermogravimetric analysis (TGA), while the XRD spectra of the used catalyst also proved phase stability [12]. These demonstrate the robust character‐ istics of the catalyst, withstanding repeated use at elevated temperature and pressure without deterioration in performance. Given comparable performance to a homogenous catalyst as shown in Figure 13b, the superparamagnetic catalyst provides a more practical approach to

**Figure 13.** Assessment of catalyst stability with repeated use and comparison to a recoverable homogeneous catalyst at

Due to several advantages such as high mechanical strength, possibility of regeneration, easier separation, and robust process integration, metal oxides can be considered superior to conventional PET glycolysis catalysts. Moreover, there are numerous possibilities to tailor their physical and chemical properties for the desired catalytic performance and functionality [47]. For the enhancement of catalytic activity, for example, altering the metal composition by introduction of another metal could result in higher catalytic activity [48–50]. The same principle is used to potentially enhance the performance of metal oxide catalysts for PET glycolysis. In this section, pure oxides and mixed-oxide spinel oxides of zinc, manganese, and

Simple precipitation method was used to synthesize the pure metal oxides. A 1.0 M solution of the salt precursors (Mn(NO3)2 xH2O, Zn(NO3)2 6H2O, or Co(NO3)2 6H2O) was mixed with 0.1 M ammonium hydroxide to set the pH at 9.0. Precipitates of the corresponding metal

cobalt were synthesised by simple precipitation or co-precipitation methods.

catalyst separation by application of a magnetic field.

154 Advanced Catalytic Materials - Photocatalysis and Other Current Trends

the same reaction conditions. Figure from [12].

**2.4. Mesoporous spinel oxide catalysts**


**Table 3.** Surface area, pore dimensions, and acid site concentration of the pure and mixed oxide catalysts [53]


**Table 4.** Atomic analysis of the synthesized oxide catalysts by EDX [53]

The catalytic activity of metal oxides in PET glycolysis is influenced by the interaction of the metal cation and the carbonyl oxygen in the polyester. The nature of the metal, oxidation state, and crystal structure affect this interaction. In spinel oxides, the metal cations can be located in tetrahedral and octahedral sites. The metal covalency of two or more different metals in the mixed oxides can result in a beneficial interaction that could enhance its redox properties and catalytic activity. In this study, the best catalytic activity for glycolysis was demonstrated by ZnMn2O4 catalyst (Figure 14). This catalyst has the ion pair Zn2+/Mn3+ in its crystal lattice compared to Co2+/Mn3+ and Zn2+/Co2+ in the other mixed spinels. The high activity of the catalyst was attributed to the nature of the manganese ions combined with structural effects in the spinel crystal [13].

**Figure 14.** Comparison of BHET yields among the pure and mixed oxide catalysts [13].

## **2.5. Purity of recovered BHET monomer**

For all the glycolysis reactions above using various catalysts, the recycled BHET was recovered via simple recrystallization. Several characterization methods have been employed to verify the structure and purity of the recovered monomer. The FT-IR spectra of the recycled BHET matched that of the standard sample [10], without extra peaks characteristic of contamination. As it is also possible that dimers and oligomers are not effectively separated, thermal and structure analyses have been performed using thermogravimetric analysis (TGA) and nuclear magnetic resonance spectroscopy (NMR). The proton and carbon NMR spectra identified peaks corresponding to distinct groups in the monomer and dimer backbones [13]. Along with TGA thermogram profiles, the structure characterization verified the good separation of the BHET from its oligomers.
