Effects of Preparation Method on the Structure and Catalytic Activity of Ag–Fe₂O₃ Catalysts Derived from MOFs

Abstract

In this work, Ag–Fe₂O₃ catalysts were successfully prepared using several different methods. Our main intention was to investigate the effect of the preparation methods on the catalysts’ structure and their catalytic performance for CO oxidation. The catalysts were characterized by X-ray diffraction (XRD), N₂ adsorption–desorption, transmission electron microscopy (TEM), X-ray photoelectron spectroscopy (XPS), H₂-temperature program reduction (H₂-TPR) and inductively coupled plasma optical emission spectroscopy (ICP-OES). Ag–Fe catalysts prepared by impregnating Ag into MIL-100 (Fe) presented the best catalytic activity, over which CO could be completely oxidized at 160 °C. Based on the characterization, it was found that more metallic Ag species and porosity existed on Ag–Fe catalysts, which could efficiently absorb atmospheric oxygen and, thus, enhance the CO oxidation.

1. Introduction

Carbon monoxide catalytic oxidation is a typical model for catalyst studies due to its relatively simple surface bimolecular reaction. It has also been used in many industrial reactions, such as low-temperature water–gas shift reaction and fuel cell reaction [1,2]. There are mainly two types of catalysts for CO oxidation, including metal oxides and noble metal catalysts.

In the past few decades, many researchers have paid much attention to developing metal oxide catalysts because of their low economic costs [3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18]. Among the inexpensive transition metal oxides, hematite (α–Fe₂O₃) is an extremely fascinating candidate for potential application values, due to its high resistance to corrosion, abundant resources, low cost, and environmentally friendly properties [19,20]. Therefore, many researchers have focused on Fe₂O₃ nanocatalysts with various structures for CO oxidation [21,22,23].

Metal–organic frameworks (MOFs) are a kind of porous network material composed of metal atoms and organic ligands, which have high surface areas, tuneable porosity, and a well-defined structure [24,25,26]. MOFs have attracted much attention as precursors for nanoporous metal oxides [27,28,29,30]. Abney et al. [31] levered the advantageous properties of a sacrificial MOF template to form a series of CuO–CeO₂ catalyst systems displaying enhanced CO oxidation performance. Raja et al. [32] also prepared metal and metal oxide nanoparticles, such as Cu–CuO, Co–Co₃O₄, ZnO, Mn₂O₃, MgO, and CdS–CdO, by thermolysis of MOFs. In our previous study, α–Fe₂O₃ catalysts derived from MIL-100 (Fe) achieved a complete conversion for CO being oxidized to CO₂ at 255 °C [30], and its activity can be further improved.

The noble metal catalysts containing Ag [33,34,35], Au [36,37], Pd [38], In [39] or Pt [40,41] have been used in the catalytic oxidation of CO to CO₂ in the past few decades. Ag is a relatively inexpensive catalyst compared to Pt and Au, showing excellent low-temperature activities for many catalytic reactions, such as NOx abatement [42], ethylene epoxidation [43,44], and methane oxidation [45]. It is generally accepted that supporting Ag on transition metal oxide is beneficial to catalytic performance. Therefore, the exploration focusing on various oxides-supported Ag catalysts has attracted increasing interest.

It is well established that the preparation method has a significant influence on the structural properties and performance of the catalyst. The modifications of the surface area, particle size, dispersion of the active component, and strength of the interaction between the active phase and the support determine the redox properties and the reactivity of the final catalysts [32,46,47,48]. To the best of our knowledge, very little attention has been focused on comparing the effect of preparation methods for Ag–Fe₂O₃ catalysts derived from MOFs. In this work, we synthesized Ag–Fe₂O₃ catalysts by several different methods. Simultaneously, the Ag–Fe₂O₃ catalytic performances for CO oxidation were evaluated. The mutual relationship between the activity of catalysts and their physicochemical properties will also be discussed in detail.

2. Results and Discussion

2.1. Characterization of Catalysts

XRD patterns of pure α–Fe₂O₃ and Ag–Fe₂O₃ catalysts prepared by various methods are shown in Figure 1 (α–Fe₂O₃ was obtained by calcination of MIL-100 (Fe) at 430 °C for 2 h) [30]. It was found that all samples present two major diffraction lines at 2θ = 33.1° and 35.6°, which correspond to (104) and (110) lattice planes of α–Fe₂O₃ [3] (PDF No. 330664), In the case of the Ag–Fe₂O₃ catalysts being prepared by several different ways, the addition of silver did not change the crystalline structure. In addition, the XRD patterns of a series of Ag–Fe₂O₃ catalysts show diffraction lines at 38.2°, 44.2°, 64.4°, 77.1°, corresponding to the (111), (200), (220), and (311) lattice planes of the cubic structure of Ag, respectively [49], indicating that metal silver particles formed over the supports.

However, weak diffraction lines of silver were detected on the Ag–PB sample, which is probably due to the small particle size of Ag. To investigate the influence of the preparation process on the variation of crystallite size, the mean crystallite sizes of Fe₂O₃ and Ag for all samples were calculated using the Scherrer Formula based on the XRD results and listed in

Table 1. Textural properties, crystallite size, and composition of the samples.

Catalysts	Textural Properties			Crystallite Size (nm)			Chemical Composition
	SBET (m2 g−1) a	V (cc g−1) b	D (nm) c	XRD (nm) d			TEM (nm) e			ICP-OES (%) f		XPS(%) g
				Fe	Ag	Ag	Fe	Ag	Ag–Fe2O3	Fe	Ag	Ag–Fe
Fe2O3 [30]	27	0.21	2.2	26	-	-	69.1	0	0	-	-	-
Ag–Fe	68	0.26	1.5–4, 12–25	21	10	4.7 ± 1.2	65.5	3.41	3.64	6.81	1.78	0.26
Ag–MIL	17	0.06	1.4	26	9	4.5 ± 1.4	62.2	8.26	9.29	6.07	1.72	0.28
Ag–Fe2O3	21	0.06	1.7	26	12	5.3 ± 1.0	60.6	5.03	5.81	7.27	1.35	0.19
Ag–PB	21	0.19	1.9	27	8	3.8 ± 0.6	53.2	5.09	6.70	7.76	2.38	0.31

^a Brunauer–Emmett–Teller (BET) specific surface area; ^b Total pore volume measured at P/P₀ = 0.99; ^c The pore diameter calculated from the desorption branch of the isotherm using the Barrett–Joyner–Halenda (BJH) method; ^d Calculated by Scherrer formulation from the FWHM of Fe₂O₃ (104) and Ag (111) diffraction peak; ^e Measured by transmission electron microscopy (TEM); ^f Element ratio determined by inductively coupled plasma optical emission spectroscopy (ICP-OES); ^g Surface atomic ratio calculated from the X-ray photoelectron spectroscopy (XPS) results.

. It is well known that the smallest average crystallite size of Fe₂O₃ is observed for Ag–Fe (21 nm), while it is ca. 26–27 nm in all other methods. The mean silver grain sizes of Ag–Fe, Ag–MIL, Ag–Fe₂O₃, and Ag–PB samples, calculated from the Ag (111) half-peak broadening (FWHM), were 10, 9, 12, and 8 nm, respectively.

Figure 2 shows the N₂ adsorption–desorption isotherms (Figure 2A) and the pore size distribution curves (Figure 2B) of Ag catalysts. The isotherms of Fe₂O₃, Ag–Fe, and Ag–PB show the type IV isotherms with an unclosed hysteresis loop at P/P₀ > 0.8, indicating that there are interparticle pores or that some impure phases exist. The isotherms of Ag–MIL and Ag–Fe₂O₃ can be classified as type I according to the IUPAC classification. The structural parameters of the catalysts, including BET surface areas, total pore volumes, and the pore diameters are summarized in Table 1.

As a first observation, compared with the pure Fe₂O₃ catalyst, the pore size of Ag–Fe lies mainly in the range of 1.5–4 nm and 12–25 nm, indicating that the catalyst produced by this method can produce many large pores, which results in an increase of BET surface areas from 27 to 68 m² g⁻¹. The relatively high surface area might be attributed to the decomposition of nitrates (AgNO₃) and organic ligands (MIL-100 (Fe)) simultaneously during preprocessing; a lot of pores are formed in the formation of gases. However, for Ag–MIL, Ag–Fe₂O₃, and Ag–PB, the surface areas and pore diameters decreased with the additional Ag component, which is consistent with most literature reports [50,51].

The dispersion of noble metal can be significantly affected by the preparation process, implying the diversity in the size distribution of the noble metal particles, which can be confirmed by the TEM images (Figure 3). In the TEM images, highly dispersed silver nanoparticles can be observed on the surface of Ag–MIL and Ag–Fe catalysts. The Ag particle size distributions established, based on the TEM images, are also presented in Figure 3. It can be seen that the mean Ag mean particle size in Ag–Fe is about 4.7 ± 1.2 nm (Figure 3a’), and the average sizes of Ag nanoparticles on Ag–MIL, Ag–Fe₂O₃ and Ag–PB are about 4.5 ± 1.4 nm (Figure 3b’), 5.3 ± 1 nm (Figure 3c’) and 3.8 ± 0.6 nm (Figure 3d’), respectively. The size and distribution of silver nanoparticles were further provided by TEM images, rending it possible to observe the particles with sizes lower than the results obtained using the Scherrer method.

The results of TEM and XRD show that the sizes and size distribution of Ag particles could be affected by the preparation method. In addition, it can be seen from Figure 3 that the morphologies of the Fe₂O₃ supporters are slightly different shapes due to the different preparation methods. The Ag–Fe catalyst maintains the brick-like shape of MIL-100 (Fe), Ag–MIL and Ag–Fe₂O₃ present the irregular appearance of different sizes, and Ag–PB shows a smaller block or polyhedron morphology.

The results of inductively coupled plasma optical emission spectroscopy (ICP-OES) analysis for the samples synthesized by different methods are listed in Table 1. It can be found that the chemical compositions of materials are not very close to the theoretically expected ones, except for Ag–MIL. The results showed that the Ag–Fe₂O₃ actual weight loading determined for Ag–Fe, Ag–MIL, Ag–Fe₂O₃, and Ag–PB materials are of 3.64%, 9.29%, 5.81%, and 6.70%, respectively. These results show that the in-situ synthesis can minimize the loss of silver, and three other methods of introducing silver can result in a big loss, especially for Ag–Fe catalyst. Moreover, bulk Ag to Fe₂O₃ ratios (from ICP-OES) seem to be higher than surface Ag to Fe ratios (determined from XPS), which indicates the inclusion of the Ag particle into the Fe₂O₃ matrix.

In order to investigate the redox properties of the Ag catalyst prepared by different methods, H₂-temperature program reduction (H₂-TPR) measurements were carried out, as shown in Figure 4. There are two main H₂-consumption peaks in the experimental temperature range for all catalysts; the peak at lower temperatures is attributed to the reduction process of Fe₂O₃ to Fe₃O₄ and the other broad peak at higher temperatures is associated with the transformations of Fe₃O₄→FeO→Fe⁰ [33]. It is worth noting that the temperature of reduction peaks of Fe₂O₃ to Fe₃O₄ over the Ag–Fe catalyst (at 170 °C) is much lower than the other three catalysts, indicating that Fe₂O₃ can be more easily reduced by introducing Ag impregnated in MIL-100 (Fe). Meanwhile, the temperature of reduction peaks of Fe₂O₃ to Fe₃O₄ over Ag–MIL, Ag–Fe₂O₃, and Ag–PB are 384 °C, 362 °C, and 502 °C, respectively. Among all the Ag catalysts, Ag–Fe shows the lowest temperature, indicating that its reducibility is stronger. Thus, this catalyst seems to be beneficial to the CO oxidation reaction.

The XPS characterizations were carried out to further investigate the relationships between surface chemical species and catalytic performance on the catalyst. As shown in Figure 5 and

Table 2. XPS species molar ratio of the synthesized Ag–Fe₂O₃ catalysts with different preparation methods.

Catalysts	Ag0/Agδ+	Fe2+/Fe3+	Olatt/Oads
Fe2O3	-	1.24	1.27
Ag–Fe	2.00	1.38	2.78
Ag–MIL	0.97	0.71	6.05
Ag–Fe2O3	1.17	0.51	1.32
Ag–PB	1.37	0.68	1.15

, The high-resolution Ag-3d XPS spectrum shows two individual broad peaks located at about 368.1 eV and 374.4 eV (Figure 5A), which are assigned to metallic Ag 3d_5/2 and Ag 3d_3/2 levels, respectively. The 3d_5/2 peak can be resolved into two peaks located at about 368.2 eV and 367.6 eV, which can be attributed to metallic Ag⁰ and silver oxide species, respectively [51,52]. A quantitative analysis of surface Ag⁰/Ag^δ+ molar ratios on the Ag 3d_5/2 XPS spectra of Ag–Fe₂O₃ catalysts with different preparation methods is summarized in Table 2. As the preparation method changes, Ag⁰/Ag^δ+ molar ratios of Ag–Fe, Ag–MIL, Ag–Fe₂O₃, and Ag–PB are 2.00, 0.97, 1.17, and 1.37, respectively.

As shown in Figure 5B, the Fe-2p spectrum is comprised of two characteristic binding energy peaks at 710.9 eV for Fe2p_3/2 and 724.6 eV for Fe2p_1/2. The peak separation between Fe 2p_3/2 and Fe 2p_1/2 peaks is about 13.7 eV, indicating the formation of the α–Fe₂O₃. The peaks at 710.9 eV can be fitted with two peaks of 712.8 and 710.7 eV, which are attributed to Fe (II) and Fe (III) species [53]. Fe²⁺/Fe³⁺ molar ratios also change with the preparation methods changes. As Table 2 shows, the Fe²⁺/Fe³⁺ molar ratios of Ag–Fe, Ag–MIL, Ag–Fe₂O_3, and Ag–PB are 1.38, 0.71, 0.51, and 0.68, respectively.

Figure 5C shows the XPS spectrum of O-1s, fitted by two bending energy peaks of about 530.1 eV and 532.1 eV. The peaks at 530.1 eV can be attributed to oxygen species corresponding to Fe₂O₃, while the other peak at 532.1 eV can be related to oxygen in the electrophilic state adsorbed on Ag [54]. As Table 2 shows, the O_latt/O_ads molar ratios of Ag–Fe, Ag–MIL, Ag–Fe₂O₃ and Ag–PB are 2.78, 6.05, 1.32, and 1.15, respectively.

2.2. Catalytic Performance

The catalytic behaviors of the Ag–Fe₂O₃ catalysts with various preparation methods toward CO oxidation are shown in Figure 6. All samples were calcined for 2 h at 430 °C in muffle furnaces before testing. It was found that the preparation method markedly influenced the catalytic activity. It is obvious that the Ag–Fe catalyst shows outstanding catalytic activity for the CO oxidation (T₁₀₀ = 160 °C). However, the total conversion temperatures of Ag–MIL, Ag–Fe₂O₃, and Ag–PB toward CO oxidation are 275 °C, 220 °C, and 480 °C, respectively. In order to get more detailed comparison about the activity of CO oxidation for different silver catalysts, several catalysts have been listed in

Table 3. Comparison of the activity for CO oxidation on different catalysts.

Catalysts	Ag Loading (wt %)	Flow Rate (mL min−1)	T10 (°C)	T50 (°C)	T98 (°C)	References
Fe2O3	0	30	110	230	253	[30]
Ag–Fe	3.64	30	<100	132	157	This work
Ag–MIL	9.29	30	190	230	275	This work
Ag–Fe2O3	5.81	30	120	≈180	215	This work
Ag–PB	6.70	30	290	>350	>350	This work
Ag–Fe2O3	≈20	100	≈110	≈150	≈270	[50]
Ag–Fe2O3	6.7	30	<50	80	≈200	[51]
Ag–Co3O4	-	30	≈50	≈85	100	[52]
Pd/α–Fe2O3	1.04	200	-	-	20	[55]
Au/α–Fe2O3	0.74	200	≈50	>75	>75	[56]

. The Ag–Fe catalysts prepared in this work show higher activity for complete conversion than those reported ones [50,51,52,53,54,55,56].

It is worth noting that the preparation methods show a great influence on the structure and activity of the catalysts, according to the results of the catalyst characterization and the activity test results. The smaller size of crystalline Fe₂O₃ (XRD), as well as its relatively high surface area, are observed in Ag–Fe catalyst. It can also be found that Ag–Fe prepared in this way has a highly dispersed, uniform silver particle on its surface (TEM). According to the H₂-TPR results, Ag–Fe is more likely to be reduced at lower temperatures. Moreover, the Ag–Fe catalyst shows the optimal Ag⁰/Ag^δ+ ratios. The addition of Ag changes the distribution of original Fe²⁺/Fe³⁺, showing that a larger ratio of Fe²⁺/Fe³⁺ is favorable to form more oxygen vacancies [57] and is also much more abundant in surface lattice oxygen species, which are favorable to the adsorption and activation of CO molecules [57].

In addition, the Ag–Fe catalyst absorbs more atmospheric oxygen, which most likely represents the oxygen supply during the oxidation reaction. The increase in activity has often been ascribed to the oxygen donation from the surface transition metal oxide sites to the silver species [58]. When the catalyst was prepared by impregnating Ag into MIL-100 (Fe) and pretreated in a muffle furnace at 430 °C (Ag–Fe), the carrier MIL-100 (Fe) is decomposed into small size Fe₂O₃ with the formation of large pores and larger specific surface areas, which is likely due to the decomposition of nitrate. In the meantime, the organic ligands of MIL-100 (Fe) were decomposed to produce reducing gases, which promote the formation of Ag⁰.

3. Materials and Methods

3.1. Catalyst Preparation

MIL-100 (Fe) was prepared by a traditional hydrothermal method [59] according to a given amount of iron powder (0.654 g), 1, 3, 5-benzenetricarboxylic acid (1, 3, 5-BTC,1.76 g), a given volume of hydrofluoric acid (1.02 mL), nitric acid (0.48 mL) and pure water (56.9 mL), where the component ratio of n (Fe):n (1,3,5-BTC):n (HF):n (HNO₃):n (H₂O) = 1:0.67:2:0.6:277. After continuous stirring for five minutes, the mixture was transferred to a polypropylene bottle and subsequently placed into an oven at 150 °C for 12 h. After cooling to room temperature naturally, the resulting solid was collected by filtration, washed with deionized water and dried under 80 °C for 12 h to get light orange bulk crystals. The dried samples were further purified with ethanol under the condition of 70 °C for 4 h, then washed with deionized water and dried under 80 °C for 12 h to obtain high purity MIL-100 (Fe).

Scheme 1 illustrates the specific process for generating the Ag–Fe₂O₃ composite catalysts (Fe/Ag ratio of 15) used in this study: (i) Silver catalysts supported on MIL-100 (Fe) support were synthesized by a conventional incipient wetness impregnation method [2]. A given amount of AgNO₃ precursor solution was dropped slowly to MIL-100 (Fe) support with constant stirring, then the sample was aged at room temperature overnight and dried at 70 °C for 12 h. The final product was named as Ag–Fe.

In addition, we also used three methods to prepare the Ag–Fe₂O₃ catalysts. (ii) The synthesis process is the same as the synthesis of MIL-100 (Fe), except that a given amount of silver is added during the agitation process before hydrothermal treatment; the final product was named as Ag-MIL. (iii) The Fe₂O₃ was used as the carrier, which derived from MIL-100 (Fe) with calcination for 2 h at 430 °C in a muffle furnace [30], and the silver was impregnated into the Fe₂O₃ as before, while the final catalyst was denoted as Ag–Fe₂O₃. iv) Commercial Prussian blue (C₁₈Fe₇N₁₈) was used as the support, and the silver was loaded by the impregnation method as before which was denoted as Ag–PB. All as-prepared samples were calcined at 430 °C for 2 h in a muffle furnace and kept in a drying vessel. Before any characterization and tests, the samples were purged at 100 °C for 30 min with Ar atmosphere.

3.2. Catalysts Characterization

X-ray powder diffraction (XRD) characterizations were performed with a Bruker D8 Advance instrument, employing a Copper Kα radiation monochromatic detector, which with λ = 1.5418 Å at an accelerating voltage of 40 kV and an emission current 40 mA, the sample was scanned in the range between 10° and 80° with a scan rate of 5°·min⁻¹. N₂ adsorption–desorption isotherms were obtained on a Quantachrome autosorb-iQ-2MP. The Brunauer–Emmett–Teller (BET) method was used to calculate the specific surface areas; pore size distribution curves were determined from the adsorption branches calculated by the Barrett–Joyner–Halenda (BJH) method. Transmission electron microscopy (TEM) images were obtained by using a JEM 2100 instrument, and the acceleration voltage was 200 kV. Temperature programmed reduction (TPR) experiments were carried out with a ChemBET TPR/TPD instrument, where the flow was fed with 10 vol.% H₂ and balanced with Ar. 0.05 g samples, which were reduced in a flow of H₂–Ar mixture. The temperature increased from room temperature to 850 °C with a heating rate of 10 °C·min⁻¹. X-ray photoelectron spectroscopy (XPS) relevant data were recorded by ThermoFisher with Al Kα source, and the binding energies were calibrated internally by the carbon deposit C-1s binding energy (BE) at 284.6 eV. Metal content was carried out by an inductively coupled plasma optical emission spectroscopy (ICP-OES), which was performed on Optima 8000 PE.

3.3. Catalytic Activity Measurements

The performance of catalysts for CO oxidation was evaluated in a continuous flow fixed-bed reactor system; the outer diameter of quartz microreactor is 6 mm. The 0.1 g catalysts (20–40 mesh) were used for CO oxidation testing at the atmospheric pressure. The reactants were fed with a volume ratio of He/CO/O₂ = 79/1/20 (a total flow rate of 30 mL min⁻¹) controlled by independent thermal mass flow controllers. The concentrations of CO were analyzed after the reaction process became stable by using on-line gas chromatograph (GC 2060) with a thermal conductivity detector (TCD). In this paper, the CO conversion was calculated from the CO concentration of inlet and outlet gases:

$$\mathrm{CO}\mathrm{conversion}=\left(\left[\mathrm{CO}\right]\mathrm{in}-\left[\mathrm{CO}\right]\mathrm{out}\right)/\left[\mathrm{CO}\right]\mathrm{in}\times 100\%$$

where [CO]in is the inlet CO concentration and [CO]out is the outlet CO concentration. In previous experiments, we measured the conversion rates by detecting CO₂, and the results are consistent with the CO calculation, and there was no hydrogen produced during the catalytic process by mass spectrometry.

4. Conclusions

According to the structural characterizations and catalytic activity evaluation results, it can be found that the preparation method has a great influence on the structure of catalysts and catalytic performance. It can be reasonably assumed that Ag⁰ species and porosity are the main factors for excellent catalytic performance, which was established via impregnating Ag into MIL-100 (Fe) and calcination for 2 h at 430 °C in a muffle furnace. It is worth noting that the simultaneous decomposition of organic ligands and nitrates is beneficial to the formation of Ag⁰ species and larger specific surface areas. All these advantages are attributed to the Ag–Fe catalyst prepared by impregnating Ag to MIL-100 (Fe), a route which could be employed as a general method for preparing other noble-metal supported catalysts.