Core-Sheath Pt-CeO2/Mesoporous SiO2 Electrospun Nanofibers as Catalysts for the Reverse Water Gas Shift Reaction

One-dimensional (1D) core-sheath nanofibers, platinum (Pt)-loaded ceria (CeO2) sheath on mesoporous silica (SiO2) core were fabricated, characterized, and used as catalysts for the reverse water gas shift reaction (RWGS). CeO2 nanofibers (NFs) were first prepared by electrospinning (ES), and then Pt nanoparticles were loaded on the CeO2 NFs using two different deposition methods: wet impregnation and solvothermal. A mesoporous SiO2 sheath layer was then deposited by sol-gel process. The phase composition, structural, and morphological properties of synthesized materials were investigated by scanning electron microscope (SEM), scanning transmission electron microscopy (STEM), X-ray diffraction (XRD), nitrogen adsorption/desorption method, X-ray photoelectron spectroscopy (XPS), inductively coupled plasma—optical emission spectrometry (ICP-OES) analysis, and CO2 temperature programmed desorption (CO2-TPD). The results of these characterization techniques revealed that the core-sheath NFs with a core diameter between 100 and 300 nm and a sheath thickness of about 40–100 nm with a Pt loading of around 0.5 wt.% were successfully obtained. The impregnated catalyst, Pt-CeO2 NF@mesoporous SiO2, showed the best catalytic performance with a CO2 conversion of 8.9% at 350 °C, as compared to the sample prepared by the Solvothermal method. More than 99% selectivity of CO was achieved for all core-sheath NF-catalysts.


Introduction
Depending on the desired structural properties, one-dimensional (1D) structures could be synthesized using various methods such as electrospinning (ES), hot-filament metal-oxide vapor deposition, sacrificial-template method, etc., [1][2][3][4]. For nanofibers (NFs) as a promising 1D structure, achieving a high surface-to-volume ratio is always desirable. Among these synthesis techniques, ES is a facile method to produce several NFs materials. Moreover, ES is a straightforward and cost-effective technique to create 1D fiber structures on the scale of nanometers to several micrometers and with various shapes, including solid, hollow, core-sheath, and hierarchical structures [5][6][7]. Such flexibility and potentials are owed to the controllable parameters of the ES process. In a typical ES process, a polymeric viscous-enough solution loaded in a needle (spinneret) is exposed to homogenous sheath layer challenging. Here, to obtain a uniform sheath layer around CeO 2 , nonporous smooth CeO 2 NFs were fabricated using the ES technique. To the best of our knowledge, electrospun NF core-sheath structures have not been investigated for RWGS so far. Moreover, two different methods were used to deposit Pt on CeO 2 NFs, including wet impregnation and solvothermal. Deposition of pre-synthesized Pt NPs on CeO 2 NFs using the solvothermal method have not been reported before.
In this study, we designed a systematic multi-step synthesis method for developing the core-sheath structure of Pt-CeO 2 NFs@mSiO 2 . Non-porous smooth CeO 2 NFs were first produced by ES technique, then two different approaches i.e., wet impregnation and solvothermal, were applied and compared for deposition of Pt NPs, and at the last step, a mesoporous SiO 2 was homogeneously formed around the core using the sol-gel process. The obtained catalysts were tested in the RWGS. A comprehensive characterization of the materials was further carried out to investigate the Pt distribution on NFs, core-sheath morphology, and the chemical properties of metals.

Preparation of Spinnable Solution
In a typical synthesis, 1.4 mmol of Ce(NO 3 ) 3 ·6H 2 O and 1.4 mmol of PVP (M w = 1,300,000 gmol −1 ) were separately dissolved in 2 and 3 mL of DMF, respectively. The polymer solution was stirred at 50 • C to facilitate PVP dissolution. After 2 h, Ce(NO 3 ) 3 ·6H 2 O solution was added dropwise into the polymeric solution under vigorous stirring. A yellowish spinnable solution was obtained after stirring overnight and was used for electrospinning.

Axial Electrospinning of CeO 2
Electrospun CeO 2 NFs were prepared using an electrospinning apparatus, Yflow ® 2.2.D-300. The axial electrospinning equipment consists of a high-voltage supply, a peristaltic pump, a syringe with a 21-gauge silver-coated needle, and a ground collector. The spinnable solution was loaded into the syringe and pumped to the needle at a constant flow rate. NFs were obtained by adjusting the operating parameters with a voltage of 15 kV, a needle-to-collector distance of 15 cm, and a flow rate of 0.5 mL/h. The as-spun NFs were peeled off from the collector and dried at 80 • C in an oven for 24 h. Then the dried NFs were calcined at 600 • C for 2 h with a heating rate of 1 • C/min in air to remove PVP and obtain CeO 2 NFs. The commercial CeO 2 powder was used for comparison and was labeled as CeO 2 -Com. separated powder-like CeO 2 NFs were further used in sol-gel synthesis to obtain the core-sheath structure.

Sol-Gel Synthesis of CeO 2 NF@SiO 2 (CeSi)
A sol-gel method was implemented to synthesize the core-sheath structure, according to the previous work [49]. About 20 mg of ultrasonicated CeO 2 NFs were dispersed in 45 mL of DI water. Then, a solution of 0.62 mmol of CTAB in 30 mL ethanol was added to the CeO 2 NFs aqueous solution. A total of 0.2 mL of ammonia solution was added dropwise to the above solution. Subsequently, 100 µL of diluted TEOS in ethanol (1 vol%) was slowly added into the solution, followed by stirring at room temperature for 6 h. A centrifugation was performed at 4000 rpm for 5 min to collect the as-synthesized core-sheaths. The solid was calcined at 360 • C for 2 h to obtain CeO 2 NF@SiO 2 (labeled as CeSi) .

Preparation of Pre-Synthesized Pt NPs
Preparation of Pt NPs was carried out as follows.
Briefly, 21.8 mg PVP (M w = 29,000 gmol −1 ) and 36.8 mg TTAB were dissolved in 10 mL EG and transferred to an argon-protected three-necked flask equipped with a condenser in Argon protection. A total of 21.9 mmol (9.1 mg) K 2 PtCl 4 was dissolved in 4 mL EG using sonication for 20 min. Then, the Pt solution was injected into the flask and stirred for 15 min with a stirring rate of 250 rpm at room temperature. Afterwards, the reaction temperature was increased to 175 • C, kept for 30 min, and then cooled to room temperature naturally. The Pt NPs dispersed in EG were used for the solvothermal method.

Loading of Pre-Synthesized Pt on CeO 2 NF by Solvothermal Method
To prepare the solution, a certain amount of CeO 2 NFs was ultrasonically dispersed in 20 mL of ethanol. Then, the pre-synthesized Pt NPs dispersed in EG solution were added dropwise into the CeO 2 NF/ethanol suspension to prepare a mixture solution of Pt/CeO 2 NF/ethanol. The mixture was transferred into a Teflon-lined stainless steel container and placed in an oven at 140 • C for 6 h. Afterwards, the obtained solution was centrifuged at 6000 rpm for 15 min to separate the Pt-loaded CeO 2 NFs and then, dried overnight at 150 • C. This sample was labeled as ST-PtCe and was used for further synthesis.

Sol-Gel Synthesis of Pt-CeO 2 @SiO 2
The same sol-gel procedure (Section 2.3.2) was used to prepare the core-sheath of Pt-CeO 2 @SiO 2 . In a typical synthesis, 20 mg of Pt-CeO 2 NFs were dispersed in 45 mL of DI water. Then as a separate solution, 225 mg of CTAB was dissolved in 30 mL of ethanol. The CTAB/Ethanol was subsequently added to Pt-CeO 2 suspension and stirred for some minutes. About 0.2 mL of ammonia solution was added to the above solution to control the pH between 9 and 11. Subsequently, 100 µL of TEOS (diluted with ethanol) was slowly added to this solution under stirring. The core-sheath of Pt-CeO 2 @SiO 2 was then separated by centrifuging at 4000 rpm for 5 min and dried overnight at 80 • C. Finally, the solid products were kept at 360 • C for 2 h to remove the CTAB template and to get solid oxides. Figure 1a-g schematically shows the fabrication and synthesis procedures performed in this study. All used samples described in the above sections are shown in Table 1.
Nanomaterials 2023, 13, 485 5 by centrifuging at 4000 rpm for 5 min and dried overnight at 80 °C. Finally, the solid ucts were kept at 360 °C for 2 h to remove the CTAB template and to get solid o Figure 1a-g schematically shows the fabrication and synthesis procedures perform this study. All used samples described in the above sections are shown in Table 1.

Characterization
The crystalline phase composition of CeO 2 NFs and core-sheath NFs after synthesis and calcination was assessed by X-ray diffraction (XRD, Bruker D8 Advance, Germany) in the reflection mode using Co K α radiation (λ = 1.789 Å) in the 2θ range of 10-90 • with the step size and time of 0.019 • and 192 s, respectively. The indexing of crystalline phases was performed based on powder diffraction data distributed from the International Centre for Diffraction Data (ICDD ® ) [50,51]. Rietveld refinement was implemented using FullProf Suit software [52]. The refinement of all samples was performed by the profile function 7. The resolution of the instrument was provided from the structure refinement of LaB 6 as standard. The parameters corresponding to the refinement consisted of the scale factor, zero-point of the detector, background parameters, lattice parameters, isotropic atomic displacement parameters (B iso ), asymmetric parameters, and the fractions of side phases.
The identification of functional groups was implemented by Fourier-transform infrared spectroscopy (FTIR) in Vertex 70 (Bruker, Germany) in the wavenumber range of 400-4000 cm −1 . The surface composition of the materials, as well as the chemical state of the corresponding elements, was analyzed by X-ray photoelectron spectroscopy (XPS). The XPS measurement was implemented with a source gun type of Al K α , a spot size of 400 µm, an energy step size of 0.1 eV, and energy steps of 601 (Thermo Fischer Scientific, Waltham, MA, USA). The fitting of curves was performed using Origin 2018, and the deconvolution of the curves was performed by adjusting a shared full width at half maximum (FWHM) in a Gaussian function. All XPS spectra were corrected based on C1s binding energy of 284.8 eV.
The specific surface area, pore size, and pore volume of NFs were investigated using N 2 adsorption-desorption at a cryogenic temperature of 77K by QuadraSorb SI device (Quantachrome Instruments, Boynton Beach, FL, USA). The NFs were outgassed for 12 h at the temperature of 150 • C. Brunauer-Emmett-Teller (BET) theory was employed to assess the surface area of NFs. The QuadraWin software (Quantachrome Instruments, USA) was used to explore the BET data.
The microstructure of NFs and core-sheath products were investigated by scanning electron microscopy (SEM, LEOGEMINI 1530, Zeiss, Jena, Germany). The elemental analysis was performed using energy-dispersive X-ray spectroscopy (EDS). The samples were prepared by scattering a layer of carbon to inhibit the charging during characterization.
Morphology and fine microstructure of core-sheath structures were investigated by transmission electron microscopy (TEM), using a 200 kV LaB6 TECNAI from FEI company, operated at 200 kV and a high-resolution scanning electron microscopy (STEM), using a 300 kV cold FEG and probe-corrected JEM-ARM300F2 from JEOL Ltd., Freising, Germany, operated at 300 kV. Samples were prepared by dispersing a certain amount of electrospun and synthesized solids in ethanol using ultrasonication. Mapping analysis was implemented to evaluate the distribution of the elements in core-sheath NFs. The microscope was operated at 300 kV, equipped with a dual SDD EDX System (JEOL Ltd.) with a detection area of 2 × 158 mm 2 and an energy resolution of 134 eV. STEM Images were acquired with a camera length of 8 cm, which corresponds to a HAADF detection angle of 68-280 mrad.
The amount of Pt loading on CeO 2 NFs was determined using inductively coupled plasma measurement by a Horiba Scientific ICP Ultima2 (Horiba, Kyoto, Japan).

Catalytic Activity Test
Reverse water gas shift (RWGS, Equation (1)) is one of the common reactions in the industry in which CO 2 reacts with hydrogen (H 2 ) to produce carbon monoxide (CO) and water (H 2 O). Due to its endothermic nature, the RWGS is favored at high temperatures [19,53]. Low-temperature CO 2 and CO methanation (Equations (2) and (3), respectively) are side reactions of RWGS.
In this study, RWGS was carried out in a stainless-steel fixed-bed tubular column reactor (inner diameter = 4 mm and length = 70 cm). A 50 mg amount of catalyst was first diluted in 450 mg of SiC with the mesh sieve of 100-200 µm, then loaded into the reactor. The bottom of the column was packed with a layer of pure SiC (400-500 µm) and quartz wool on which the diluted catalyst was placed. The bed temperature of the reactor was measured by an installed thermocouple inside the center of the column. Before operating the catalytic activity test, the catalyst was in situ reduced at 350 • C for 2 h in the flow rates of H 2 (40 mL/min) and N 2 (30 mL/min) at atmospheric pressure. Next, the reactor was cooled down to the reaction temperature, and pressure was increased to 6.2 bar. The catalyst was tested under the gas flows of H 2 (30 mL/min), N 2 (15 mL/min), and CO 2 (10 mL/min) with the molar ratio of H 2 :N 2 :CO 2 (3:1.5:1) and a gas hour space velocity (GHSV) of 66,000 mL gcat −1 h −1 . The operating pressure was set to 6.2 bar in all reaction temperatures. The catalyst performance was measured at three different temperatures of 250, 300, and 350 • C. The reaction was set by heating the reactor to the desired temperature at a rate of 10 • C/min. The concentrations of gas products were analyzed online by a gas chromatography instrument (Schimatzu 7890A) equipped with a thermal conductivity detector (TCD) and a flame ionized detector (FID). CO and CH 4 were the main products of the process. The CO 2 conversion (X CO 2 ) and selectivity of CO (S CO ) and CH 4 (S CH 4 ) were calculated through the following Equations (4)-(6): where (in) and (out) are denoted for the mole of reactant and effluent corresponding gases respectively.

Results and Discussion
CeO 2 NFs were fabricated using ES technique, followed by a calcination step at 600 • C. The NFs were further used for synthesizing the core-sheath structure of CeO 2 @SiO 2 via a sol-gel route. Figure 2 shows XRD patterns of all samples, including electrospun CeO 2 NFs, CeSi core-sheath, ST-PtCeSi, and IM-PtCeSi NFs. The measured pattern for CeO 2 NFs can be attributed to a pure CeO 2 crystal structure based on JCPDS 34-0394 representing the fluorite cubic structure. The reflections at 28.5 • , 33.14 • , 47. Although well-designed core-sheath structures have been achieved, a few extra spherical SiO2 particles can be seen in some regions. These particles might have formed first separately during sol-gel synthesis and then attached on the silica sheath. To verify the crystallinity of structure, SAED analyses were performed for both core and sheath layers (Figure 3e,f). The SAED images confirm the presence of a polycrystalline CeO2 material with a fluorite cubic structure in the core (coded by 1) of CeO2 NFs and an amorphous silica sheath (coded by 2).  As mentioned before, the ultrasonicated electrospun CeO2 NFs have been used to support Pt nanoparticles by two different methods: wet impregnation and solvothermal deposition. In the impregnation method, Pt was directly deposited on CeO2 NFs, while PVP and TTAB were used as capping agent and template for synthesis of Pt NPs and its deposition on CeO2. After the deposition of Pt, a silica sheath was grown around the Pt-CeO2 NFs to obtain the core-sheath structure. STEM was used to investigate the Pt distribution and core- To gain further insight about lattice parameters, crystallite size, microstrain, and preferred orientation in the samples, Rietveld refinement was performed on the diffraction patterns. The preferred orientation was determined with the March-Dollase model [54], as implemented in the FULLPROF program. Figure S1 shows the observed, calculated, and difference profile for the final cycle of the structure refinement. The results of Rietveld refinement reveal that the lattice parameter of CeO 2 does not change significantly with synthesizing a silica layer around the electrospun CeO 2 NFs or Pt loading, suggesting that the silica and Pt are mainly added to the surface of the CeO 2 NFs without altering the lattice CeO 2 phase. In contrast, the crystallite size of CeO 2 increases from 14.9 nm in CeO 2 NFs to 21.3 nm and 36.5 nm in IM-PtCeSi and ST-PtCeSi, respectively, with the addition of silica and Pt. The March-Dollase (MD) parameter r along the <110> directions was found to be higher than unity in all samples, with increasing from 1.1 CeO 2 NFs to 1.3 and 1.4 for IM-PtCeSi and ST-PtCeSi samples, respectively. This MD parameter r defines the crystallites' habit distribution and is unity for an ideal random-orientation (i.e., no preferred orientation), greater than one for needle-habit crystals and less than one for platy crystals pack along the diffraction vector. Thus, the CeO 2 crystals in the NFs are grown as needlehabit along the <110> directions in all samples. However, the values of MD parameter r indicate the percentage of excess crystallites with preferential orientation in comparison with randomly oriented crystallites in the NFs, which means that ST-PtCeSi sample has crystallites with the highest total preferential orientation along the <110>, followed by the IM-PtCeSi sample. Moreover, the weight fraction of nanocrystalline Pt metals (~4.4 nm) was found to be 2.5 (0.2) wt.% and 19.2 (0.5) wt.% in IM-PtCeSi and ST-PtCeSi samples, respectively. The higher amounts of Pt detected in the samples from XRD compared to the experimental value (7 wt.%) can be explained by the amount of amorphous silica excluded in the Reitveld refinement analysis. These results suggest that the Solvothermal deposition of Pt on the CeO 2 NFs enhances the growth of CeO 2 crystallites along <110> direction and the formation of a high amount of Pt nanocrystallites (4.4 nm). In contrast, although wetimpregnation of Pt increases the preferred orientation of CeO 2 along the <110> direction, a slight increase in the crystallite size of CeO 2 and a low amount of metallic Pt are observed. Since the same amount of Pt loading was used for both impregnation methods, the low weight fraction of Pt in the IM-PtCeSi suggests the presence of a high amount of metallic Pt with a very small crystallites size to be detected by XRD (<2 nm).
The FTIR spectra of CeSi from 4000 to 430 cm −1 are illustrated in Figure S2. All absorption bands corresponding to Si-O and Ce-O groups can be seen in the FTIR spectra.
The band at about 440 cm −1 corresponds to the Ce-O vibration. The two absorption bands at 1063 cm −1 and about 810 cm −1 can be ascribed as symmetric and asymmetric Si-O-Si bonding groups, respectively [55]. The small band at 1650 cm −1 can be related to O-H stretching bond. Likewise, a small absorption band can be observed at 3750 cm −1, which is attributed to the OH vibrations of free silanol groups [56,57].
SEM images of the electrospun NFs are shown in Figure 3. As-spun PVP/Ce(NO 3 ) 3 ·6H 2 O NFs are illustrated in Figure 3a. The average diameter of NFs before calcination is about 200-250 nm, while the diameters have been reduced to 90-100 nm after heat treatment of the NFs at 600 • C due to the removal of PVP along with other organic moieties and oxidation of Ce(NO 3 ) 3 ·6H 2 O into CeO 2 , as can be seen in Figure 3b. This shows that the fibrous shape of CeO 2 NFs remains intact while removing PVP from the structure. Figure 3b shows that the surface of produced NFs is smooth with a relatively uniform average diameter of 143 nm. Prior to introducing the SiO 2 sheath, the mat of CeO 2 NFs was ultrasonicated for better accessibility of the entire CeO 2 surface. Figure 3c shows that after ultrasonication, the average length of the NFs decreased to 500-1000 µm. TEM image of the CeSi core-sheath NFs confirms the successful formation of core-sheath structure (Figure 3d). A clear interface between CeO 2 NFs as core and SiO 2 as sheath can be identified, with a core diameter of about 340 nm and a sheath thickness of about 70 nm. Although well-designed core-sheath structures have been achieved, a few extra spherical SiO 2 particles can be seen in some regions. These particles might have formed first separately during sol-gel synthesis and then attached on the silica sheath. To verify the crystallinity of structure, SAED analyses were performed for both core and sheath layers (Figure 3e,f). The SAED images confirm the presence of a polycrystalline CeO 2 material with a fluorite cubic structure in the core (coded by 1) of CeO 2 NFs and an amorphous silica sheath (coded by 2). Although well-designed core-sheath structures have been achieved, a few extra spherical SiO2 particles can be seen in some regions. These particles might have formed first separately during sol-gel synthesis and then attached on the silica sheath. To verify the crystallinity of structure, SAED analyses were performed for both core and sheath layers (Figure 3e,f). The SAED images confirm the presence of a polycrystalline CeO2 material with a fluorite cubic structure in the core (coded by 1) of CeO2 NFs and an amorphous silica sheath (coded by 2).  As mentioned before, the ultrasonicated electrospun CeO2 NFs have been used to support Pt nanoparticles by two different methods: wet impregnation and solvothermal deposition. In the impregnation method, Pt was directly deposited on CeO2 NFs, while PVP and TTAB were used as capping agent and template for synthesis of Pt NPs and its deposition on CeO2. After the deposition of Pt, a silica sheath was grown around the Pt-CeO2 NFs to obtain the core-sheath structure. STEM was used to investigate the Pt distribution and core- As mentioned before, the ultrasonicated electrospun CeO 2 NFs have been used to support Pt nanoparticles by two different methods: wet impregnation and solvothermal deposition. In the impregnation method, Pt was directly deposited on CeO 2 NFs, while PVP and TTAB were used as capping agent and template for synthesis of Pt NPs and its deposition on CeO 2 . After the deposition of Pt, a silica sheath was grown around the Pt-CeO 2 NFs to obtain the core-sheath structure. STEM was used to investigate the Pt distribution and core-sheath morphology. Distinguishable interfaces between core and sheath can be observed for IM-PtCeSi with the theoretical 10 wt.% of Pt loading on CeO 2 NF. The core diameter is about 100 nm, and the sheath thickness is around 40 nm ( Figure S3). Elemental mapping of Ce, Si, O, and Pt was also performed and is shown in Figure S3. As can be seen, Pt presents in some areas of the sheath layer. To improve the Pt stabilization on the CeO 2 core, a heat treatment at 150 • C for 24 h was carried out after wet impregnation. Furthermore, the theoretical loading of Pt on CeO 2 NF was decreased from 10 wt.% to 7 wt.% for IM-PtCeSi. As shown in Figure 4, the successful formation of the core-sheath structure and the well distribution and stabilization of Pt on CeO 2 were obtained. It can be concluded that no migration of Pt into the sheath has taken place. The elemental mapping analysis of IM-PtCeSi confirms as well that Pt is well distributed on CeO 2 . The diameter of core-sheath NF and the sheath thickness were about 160 nm and 55 nm, respectively. Solvothermal deposition of pre-synthesized NPs on metal oxides has been utilized to increase the interaction between metal and support [58]. So, in another approach, solvothermal deposition of Pt on CeO 2 NFs was adopted to further enhance the distribution and stabilization of Pt on the substrate. Figure S4 illustrates STEM images of ST-PtCeSi sample which also shows the migration of Pt into the SiO 2 sheath. To improve the Pt stabilization and prevent its detachment from the surface of CeO 2 NFs, a heat treatment at 150 • C for 24 h was applied after the solvothermal method, Figures 5 and S5. A welldefined core-sheath structure is obtained with a core diameter of about 110 nm and a sheath thickness of about 42 nm.
The particle size of Pt in ST-PtCeSi was measured to be 3.8 nm, consistence with crystallite size determined from XRD analysis (4.4 nm). The particle size of the pre-synthesized Pt NPs used in the solvothermal method is larger than that used in the wet impregnation method, Figure S5, and it agrees with XRD results. Considering Figure 4, it can be concluded that the Pt particle size in IM-PtCeSi is less than 3.8 nm.
Nanomaterials 2023, 13, 485 10 As mentioned before, the ultrasonicated electrospun CeO2 NFs have been used to port Pt nanoparticles by two different methods: wet impregnation and solvothermal d sition. In the impregnation method, Pt was directly deposited on CeO2 NFs, while PVP TTAB were used as capping agent and template for synthesis of Pt NPs and its depos on CeO2. After the deposition of Pt, a silica sheath was grown around the Pt-CeO2 NF obtain the core-sheath structure. STEM was used to investigate the Pt distribution and c sheath morphology. Distinguishable interfaces between core and sheath can be observe IM-PtCeSi with the theoretical 10 wt.% of Pt loading on CeO2 NF. The core diameter is a 100 nm, and the sheath thickness is around 40 nm ( Figure S3). Elemental mapping of C O, and Pt was also performed and is shown in Figure S3. As can be seen, Pt presents in s areas of the sheath layer. To improve the Pt stabilization on the CeO2 core, a heat treatm at 150 °C for 24 h was carried out after wet impregnation. Furthermore, the theoretical l ing of Pt on CeO2 NF was decreased from 10 wt.% to 7 wt.% for IM-PtCeSi. As show Figure 4, the successful formation of the core-sheath structure and the well distribution stabilization of Pt on CeO2 were obtained. It can be concluded that no migration of Pt the sheath has taken place. The elemental mapping analysis of IM-PtCeSi confirms as that Pt is well distributed on CeO2. The diameter of core-sheath NF and the sheath thick were about 160 nm and 55 nm, respectively. Solvothermal deposition of pre-synthes NPs on metal oxides has been utilized to increase the interaction between metal and sup [58]. So, in another approach, solvothermal deposition of Pt on CeO2 NFs was adopte further enhance the distribution and stabilization of Pt on the substrate. Figure S4 illust STEM images of ST-PtCeSi sample which also shows the migration of Pt into the SiO2 she To improve the Pt stabilization and prevent its detachment from the surface of CeO2 N heat treatment at 150 °C for 24 h was applied after the solvothermal method,    The particle size of Pt in ST-PtCeSi was measured to be 3.8 nm, consistence with tallite size determined from XRD analysis (4.4 nm). The particle size of the pre-synthe Pt NPs used in the solvothermal method is larger than that used in the wet impregn method, Figure S5, and it agrees with XRD results. Considering Figure 4, it can be cluded that the Pt particle size in IM-PtCeSi is less than 3.8 nm. Figure 6 shows the nitrogen adsorption-desorption isotherms for electrospun NFs with a specific surface area of 18 m 2 ·g −1 . A type IV BET isotherm is obtained for (Figure 6a), from which a surface area of 476.5 m 2 ·g −1 can be calculated. While a type BET isotherm appeared for the core-sheath sample of CeSi with a hysteresis loop, F 6b. The significant increase in surface area can be attributed to the presence of the p SiO2 layer. The inset graph in Figure 6b shows the Barret-Joyner-Halenda (BJH) size distribution curve for CeSi, which gives a pore diameter of about 2.8 nm for the layer.
For a better understanding of the chemical states of samples, the XPS analysi performed and compared with CeO2-Com. Figure 7 and Figure S6 show the XPS sp of Ce 3d and O 1s [59][60][61][62][63][64], respectively, for CeO2-Com, CeO2 NF, IM-PtCe, SN-PtCe PtCeSi, and ST-PtCeSi. Multiplets of u and v corresponding to the spin-orbital splitt Ce 3d3/2 and 3d5/2 are observed in Figure 7. The spin-orbit splitting of Ce 3d is repor be about 18.4 eV [65]. Typically, the Ce 3d spectrum displays five doublet pairs. The blet pairs of (u0-v0), (u′-v′) ascribes to Ce(III), whereas (u-v), (u″-v″), and (u‴-v‴) a cribed to Ce(IV) [66,67]. The doublets with corresponding oxidation states are indi in the XPS spectrum for each sample. In the case of CeO2 NFs, in addition to the Ce 4+ p the peaks with the binding energies (BEs) of about 903 and 885 eV indicate the pre of Ce 3+ species in the electrospun CeO2 NFs. Both oxidation states are also observ CeO2-Com.  Figure 6 shows the nitrogen adsorption-desorption isotherms for electrospun CeO 2 NFs with a specific surface area of 18 m 2 ·g −1 . A type IV BET isotherm is obtained for CeSi (Figure 6a), from which a surface area of 476.5 m 2 ·g −1 can be calculated. While a type IV of BET isotherm appeared for the core-sheath sample of CeSi with a hysteresis loop, Figure 6b. The significant increase in surface area can be attributed to the presence of the porous SiO 2 layer. The inset graph in Figure 6b shows the Barret-Joyner-Halenda (BJH) pore size distribution curve for CeSi, which gives a pore diameter of about 2.8 nm for the silica layer. The fractional amount of Ce 4+ and Ce 3+ was calculated using Equations (S1)-(S3) [66]. As shown in Figure S7, the commercial CeO2 powder consists of 24.4% of Ce 3+ and 75.6% of Ce 4+ , while a Ce 3+ concentration of 13.6% was obtained for electrospun CeO2-NFs. During wet impregnation of Pt on the CeO2-NFs, the amount of Ce 3+ species increases to 20.4%, Figure 6. Nitrogen adsorption-desorption isotherm of (a) electrospun CeO 2 NFs, and (b) core-sheath of CeSi, insert graph shows pore diameter of CeSi.
For a better understanding of the chemical states of samples, the XPS analysis was performed and compared with CeO 2 -Com. Figures 7 and S6 show the XPS spectra of Ce 3d and O 1s [59][60][61][62][63][64], respectively, for CeO 2 -Com, CeO 2 NF, IM-PtCe, SN-PtCe, IM-PtCeSi, and ST-PtCeSi. Multiplets of u and v corresponding to the spin-orbital splitting of Ce 3d 3/2 and 3d 5/2 are observed in Figure 7. The spin-orbit splitting of Ce 3d is reported to be about 18.4 eV [65]. Typically, the Ce 3d spectrum displays five doublet pairs. The doublet pairs of (u 0 -v 0 ), (u -v ) ascribes to Ce(III), whereas (u-v), (u"-v"), and (u"'-v"') are ascribed to Ce(IV) [66,67]. The doublets with corresponding oxidation states are indicated in the XPS spectrum for each sample. In the case of CeO 2 NFs, in addition to the Ce 4+ peaks, the peaks with the binding energies (BEs) of about 903 and 885 eV indicate the presence of Ce 3+ species in the electrospun CeO 2 NFs. Both oxidation states are also observed in CeO 2 -Com.
Nanomaterials 2023, 13,485 13 of 20 causes a peak shift to lower BE, as also has been reported in the literature [80]. The peak positions, oxidation states, and corresponding integrated areas for Ce and Pt are summarized in Tables S1 and S2.  To investigate the basicity of the materials, CO2-TPD was performed for IM-PtCeSi and ST-PtCeSi and presented in Figure S8. The high-temperature desorption peaks observed at 550 and 650 °C for ST-PtCeSi and IM-PtCeSi, respectively, reveal relatively strong basicity in both catalysts. The higher desorption temperature for IM-PtCeSi confirms its stronger basicity compared to ST-PtCeSi at high temperature. Moreover, the desorption peak for IM-PtCeSi is much more intense than ST-PtCeSi, revealing a higher amount of adsorption centers for CO2 molecules. The fractional amount of Ce 4+ and Ce 3+ was calculated using Equations (S1)-(S3) [66]. As shown in Figure S7, the commercial CeO 2 powder consists of 24.4% of Ce 3+ and 75.6% of Ce 4+ , while a Ce 3+ concentration of 13.6% was obtained for electrospun CeO 2 -NFs. During wet impregnation of Pt on the CeO 2 -NFs, the amount of Ce 3+ species increases to 20.4%, indicating that more oxygen vacancies have been generated. Similarly, the fraction of Ce 3+ species in the ST-PtCe sample rises to 17.5%. The higher oxygen vacancies observed in IM-PtCe might be related to the more uniform distribution of Pt on the surface of CeO 2 NFs compared to ST-PtCe (see Figures 4 and 5). Since XPS is a surface-sensitive technique, a corelevel etching XPS of the core-sheath NFs was performed to obtain precise information for Ce species. As can be seen in the spectra of IM-PtCeSi and ST-PtCeSi (Figure 7), the intensity of the peaks at 881.8 eV, 900.3 eV, 916.5 eV related to Ce 4+ significantly decreases, whereas for Ce 3+ increases, indicating that the oxidation state of Ce in the vicinity of SiO 2 sheath was changed to non-stoichiometric (CeO 2−X ) [68][69][70]. The observed reduction of CeO 2 on the surface can be likely attributed to the presence of free OH − groups in the synthesis medium during the SiO 2 sheath formation. As discussed in Rietveld refinement analysis ( Figure S1), the lattice parameter of CeO 2 was not significantly changed in all samples meaning that the reduction of Ce 4+ to Ce 3+ did not occur in the bulk of the materials and considering to the XPS analysis, it can be concluded that Ce 3+ species was merely attributed to the surface.
The Pt 4f region (Figure 8) is deconvoluted into two spin-orbit split doublets of  [78,79]. The larger particle size in ST-PtCe and ST-PtCeSi causes a peak shift to lower BE, as also has been reported in the literature [80]. The peak positions, oxidation states, and corresponding integrated areas for Ce and Pt are summarized in Tables S1 and S2. Both core-sheath catalysts, IM-PtCeSi and ST-PtCeSi, were tested for the RWG tion at different temperatures, 250, 300, and 350 °C. Figure 9 shows the temperature and CO2 conversion during the reaction. The selectivity of CO and CH4 for both cata illustrated in Figure S9. As RWGS is an endothermic reaction, the CO2 conversion inc To investigate the basicity of the materials, CO 2 -TPD was performed for IM-PtCeSi and ST-PtCeSi and presented in Figure S8. The high-temperature desorption peaks observed at 550 and 650 • C for ST-PtCeSi and IM-PtCeSi, respectively, reveal relatively strong basicity in both catalysts. The higher desorption temperature for IM-PtCeSi confirms its stronger basicity compared to ST-PtCeSi at high temperature. Moreover, the desorption peak for IM-PtCeSi is much more intense than ST-PtCeSi, revealing a higher amount of adsorption centers for CO 2 molecules.
Both core-sheath catalysts, IM-PtCeSi and ST-PtCeSi, were tested for the RWGS reaction at different temperatures, 250, 300, and 350 • C. Figure 9 shows the temperature profile and CO 2 conversion during the reaction. The selectivity of CO and CH 4 for both catalysts is illustrated in Figure S9. As RWGS is an endothermic reaction, the CO 2 conversion increases with increasing temperature [81][82][83]. The highest CO 2 conversion of 9% and 6.7% was obtained for IM-PtCeSi and ST-PtCeSi, respectively, at 350 • C. Figure S10 shows the equilibrium conversion of CO 2 between 200 and 600 • C. The experimental CO 2 conversions obtained in this study, are less than the equilibrium one, for example, at 350 • C a CO 2 conversion of 30% can be achieved at the equilibrium conditions. The CO 2 hydrogenation mechanism has been reported previously [35], in which CO 2 can be converted through the redox mechanism and directly reacts with the oxygen vacancies in CeO 2 , producing CO. In this case, Pt mainly contributes to generate oxygen vacancies by activating H 2 , and then the oxygen vacancies migrate to CeO 2 . XPS spectra of core-sheath samples (Figure 7) confirmed the existence of oxygen vacancies in the corresponding interface of ceria core and SiO 2 sheath. Moreover, IM-PtCe possesses higher Ce 3+ species ( Figure S7), so more oxygen vacancies are generated in this sample, therefore it can be expected that the catalytic performance of IM-PtCeSi becomes higher than ST-PtCeSi. In addition, due to the fact that Pt can spill over hydrogen to the neighboring CeO 2 [84], smaller particle size and higher distribution of Pt can lead to more hydrogen placing in its coordination to CeO 2 , which might enhance the catalytic performance. The smaller Pt particle size and its high distribution present in IM-PtCeSi sample compared to ST-PtCeSi can provide a higher surface area for Pt particles on CeO 2 NF that improves corresponding catalytic efficiency. With regard to CO 2 -TPD results, IM-PtCeSi shows higher CO 2 uptake compared to St-PtCeSi ( Figure S8), thus CO 2 has more opportunity to be reduced to CO in IM-PtCeSi catalyst leading to its higher catalytic performance. Moreover, the presence of a smaller crystallite size of CeO 2 , grown along the <110> direction, in IM-PtCeSi catalyst compared to ST-PtCeSi could be another reason for higher catalytic performance. Tan et al. reported that the exposed (110) facets and the existence of abundant oxygen vacancies in CeO 2 nanotube-supported Cu-Ni contribute to its higher catalytic performance for CO 2 hydrogenation [46]. Table 2 summarizes the performance of different tested catalysts for RWGS with various catalysts and the Pt loadings% of 0.025 to 2%wt, showing that increasing the loading amount of Pt enhances the CO 2 conversion. Considering Table 2, the obtained CO 2 conversions in our study are comparable to the catalysts with the same Pt loading tested at a similar range of temperatures [23,24,36,37,[85][86][87][88]. Besides the catalyst type, several parameters corresponding to reactor configuration (reactor type, reactor dimension, etc.,) and operation conditions (temperature, pressure, GHSV, feed ratio) can also affect the catalytic performance; therefore, a precise comparison study is challenging in RWGS. A conversion of 6.7% has been reported by Chen et al. for Pt-CeO 2 nanoparticles with the Pt loading of 1 wt.% at 300 • C [36], whereas in this study, core-sheath NFs showed less conversion with 0.5 wt % Pt loading in a higher GHSV, at the same temperature. In another study performed by Zhao et al. [37], 8.5% CO 2 conversion with more than 98% CO selectivity had been reported approximately at similar testing conditions, while the core-sheath catalysts tested in our study showed nearly the same conversion, 9%, and a higher CO selectivity of 99%. Considering this comparison, it can be assumed that the access of the reactants to the core is not completely prohibited by silica sheath and Pt active sites are accessible. Pt-CeO 2 @SiO 2 -Co spherical NPs have been studied for CO 2 hydrogenation tandem systems to olefins. A CO 2 conversion of 25% and more than 98% selectivity could be achieved at 350 • C over Pt-CeO 2 @SiO 2 nanoparticles for RWGS reaction with the loading of 4.4% Pt [49]. Whereases in this study, the core-sheath NFs showed conversions of 9% and 6.8% with IM-PtCe and ST-PtCeSi, respectively, with around nine times lower Pt loading. Although the obtained conversion in this study is not very high considering the thermodynamic conversion, the synthesis approach can be optimized in future to improve the catalytic performance. The proposed catalysts can be potentially used in a tandem system as a basic structure by introducing a second active metal on the sheath layer. It should be noticed that in a tandem system such as CO 2 hydrogenation, the obtained conversion of RWGS should be optimized considering the second active site to reach the best ratio of subsequent reactions. Based on the systematic synthesis approach developed in this article, the proposed method can be used to synthesize different catalysts for a variety of tandem application.
temperature. In another study performed by Zhao et al. [37], 8.5% CO2 conversion with more than 98% CO selectivity had been reported approximately at similar testing conditions, while the core-sheath catalysts tested in our study showed nearly the same conversion, 9%, and a higher CO selectivity of 99%. Considering this comparison, it can be assumed that the access of the reactants to the core is not completely prohibited by silica sheath and Pt active sites are accessible. Pt-CeO2@SiO2-Co spherical NPs have been studied for CO2 hydrogenation tandem systems to olefins. A CO2 conversion of 25% and more than 98% selectivity could be achieved at 350 °C over Pt-CeO2@SiO2 nanoparticles for RWGS reaction with the loading of 4.4% Pt [49]. Whereases in this study, the core-sheath NFs showed conversions of 9% and 6.8% with IM-PtCe and ST-PtCeSi, respectively, with around nine times lower Pt loading. Although the obtained conversion in this study is not very high considering the thermodynamic conversion, the synthesis approach can be optimized in future to improve the catalytic performance. The proposed catalysts can be potentially used in a tandem system as a basic structure by introducing a second active metal on the sheath layer. It should be noticed that in a tandem system such as CO2 hydrogenation, the obtained conversion of RWGS should be optimized considering the second active site to reach the best ratio of subsequent reactions. Based on the systematic synthesis approach developed in this article, the proposed method can be used to synthesize different catalysts for a variety of tandem application.   In order to investigate the morphology of the tested catalysts, a TEM analysis was performed after reaction, Figure S11. As seen, the morphology of catalysts is not changed, and a clear interface between the core and sheath can be observed in both catalysts. This indicates that the core-sheath structures remained intact under the catalytic testing conditions. However, Pt aggregation has been observed for ST-PtCeSi, Figure S11d.

Conclusions
The electrospinning technique has the potential to produce NFs on a large scale, compared to the chemical syntheses in which the product might be yielded on limited scales. The NFs morphology can also be tuned using this technique so that potentially a bulk texture of nonwoven NFs can be directly utilized for further surface modifications and tested in a reactor for catalysis. The electrospun NFs can provide a higher surface area which might improve the catalytic performance. In this work, core-sheath NFs of Pt-CeO 2 NF@mSiO 2 were successfully fabricated using the electrospinning technique and further sol-gel synthesis. The sheath layer of SiO 2 was grown directly on electrospun NFs without using a capping agent. Pt was deposited on CeO 2 NFs by two different methods, including wet-impregnation and solvothermal deposition of pre-synthesized Pt. Structural and morphological studies revealed that Pt was more homogeneously dispersed on CeO 2 NF in IM-PtCeSi compared to ST-PtCeSi. The results of Rietveld refinement of XRD data revealed that the Pt impregnation led to the formation of small weight fractions of metallic Pt nanoparticles (≤4.4 nm) without inducing a significant change in the lattice parameter of CeO 2 , indicating that the Pt is mainly added to the surface of the CeO 2 NFs without altering the lattice CeO 2 phase. Although the CeO 2 crystals in the NFs were grown as needle-habit along the <110> directions in all samples, Pt impregnation increased the preferred orientation of CeO 2 along this direction. The XPS results regarding Ce 3d showed that the oxygen vacancies were increased by reducing Ce 4+ species to Ce 3+ in IM-PtCe after wet impregnation of Pt on CeO 2 NF. A uniform and intact porous SiO 2 sheath layer were obtained in both catalyst samples. Both catalysts were then tested for the RWGS reaction. IM-PtCeSi showed better performance compared to ST-PtCeSi, with a CO 2 conversion of 8.9% and a CO selectivity of 98.9% (at P = 6.2 bar, 350 • C, GHSV = 66,000 mLgcat −1 h −1 ). It is also demonstrated that by increasing the operating temperature, the catalytic performance is enhanced in all catalysts. Such a porous silica layer can be potentially used as a second substrate and interlayer to design tandem bifunctional catalysts as well.
Supplementary Materials: The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/nano13030485/s1. Figure S1: XRD Rietveld refinement of (a) CeO 2 -NF with the lattice parameter of 5.411(15) Å, (b) IM-PtCeSi with the lattice parameter of 5.410(30) Å and 3.922(85) Å for CeO 2 and Pt, respectively, and Pt weight fraction of 2.5(0.2) wt.% and (c) ST-PtCeSi with the lattice parameter of 5.410(14) Å and 3.923(47) Å for CeO 2 and Pt, respectively, and Pt weight fraction of 19.2(0.5) wt.%. The lattice parameter of CeO 2 was kept intact after introducing Pt, meaning the reduction of CeO 2 occurred only on the surface, not in bulk. Figure S2: FTIR spectra of CeSi. Figure S3: STEM and EDS elemental mapping of IM-PtCeSi with 10% loading of CeO 2 with Pt in bright field (BF) mode and mapping of elements, red color is Pt, green is Si and Blue is Ce. Figure S4: STEM and EDS elemental mapping of ST-PtCeSi with 10% loading of CeO 2 with Pt. Figure  S5: STEM of ST-PtCeSi after optimization of Pt with the Pt particle size distribution in ST-PtCeSi histogram plot (the average particle size is 3.8 nm); Equations (S1)-(S3): The fractions of Ce 4+ and Ce 3+ . Figure S6: XPS for O 1s for all samples. Figure S7: Ce 3+ species are present in the samples of CeO 2 -Com, CeO 2 -NF, IM-PtCe, and ST-PtCe which all are without the SiO 2 sheath. Table S1: Peak positions, Ce oxidation state, and integrated area of all samples showed in Figure 7, compared to the reference peak positions and corresponded oxidation states. Table S2: Pt peaks with corresponded oxidation states occurred in Figure 8. Figure S8: CO 2 -TPD profile of IM-PtCeSi and ST-PtCeSi coresheath NF catalysts. Figure S9: CO selectivity and CH 4 selectivity at different reaction temperatures for two tested catalysts. Figure S10: Equilibrium conversion of CO 2 at different temperature with different H 2 :CO 2 compositions. Figure