SO₂ Tolerance of Rice Hull Ash Based Fe-Cu Catalysts for Low-Temperature CO-SCR of NO

Abstract

Rice husk ash (RHA) has potential as a supporter of catalysts. In this research, we studied the activity and SO₂ tolerance of RHA-based Fe-Cu oxide in the reduction of NO by CO. Characterization methods were employed to study the properties of the catalysts and their SO₂ tolerance. Activity and SO₂ resistance were also tested at different temperatures. We recommend two catalysts with SO₂ resistance ability: Fe_0.67Cu_0.33/RHA (the highest catalytic activity) and Fe_0.8Cu_0.2/RHA. The NO removal rate hardly changed with the addition of SO₂ and was kept at about 100%. However, the CO conversion rate decreased with increasing SO₂ at the lower reaction temperatures, which may be due to the formation of sulfites. Fortunately, the deactivation was reversible and can be reduced with an increase in the reaction temperature. The results of our research may help promote the application of CO-SCR.

1. Introduction

Selective catalytic reduction (SCR) of NO has been widely used in the exhaust gas treatment of many industrial processes because of its high efficiency. It has been reported that CO-SCR has the potential to replace the common NH₃-SCR. New catalysts have been studied in much research. Rice hull ash (RHA) has many advantages as an absorber and supporter of catalysts [1]. The inorganic substance in rice husk is about 20%, the rice husk having relatively higher ash content compared with other agricultural wastes. The inorganic substance in rice husk is mainly amorphous SiO₂, which can take up to 87–97% in rice husk ash (RHA). Amorphous SiO₂ has a specific surface area, and this feature makes it a good option for the catalysis industry. Fe/RHA and Cu/RHA were proven to be helpful to CO-SCR in our previous studies [2,3]. The chemical reaction process can be expressed as 2NO + 2CO →2CO₂ + N₂.

However, SO₂ is regarded as a deactivator of catalysts, and is universal in flue gas from the combustion of fossil fuel. SO₂ hinders the application and development of CO-SCR by transition metal oxide catalysts. Investigating SO₂ resistance is a continuing concern within SCR of NO_X.

According to the literature, the mechanism of catalyst deterioration caused by SO₂ over SCR may be divided into four main categories as follows. (1) As a result of the stronger oxidation of SO₂ than NO, the adsorption of NO may be hindered by reducing catalytic activity and SO₂ may be competitively adsorbed onto the active sites [4]. (2) SO₂ can occupy the acidic active sites on the catalyst surface, which has a remarkably negative influence on ammonia SCR [5,6,7]. (3) With water vapor in the flue gas, SO₂ can be changed to sulfate on the surface of catalysts, particularly on the surface of metal oxide catalysts such as Mn [5,6], Cu [8], and Co [9], among others. (4) With ammonia SCR, SO₂ may combine with NH₃ and produce substances such as (NH₄)₂SO₄ and NH₄HSO₄ occupying the pores, which may result in a sharp decrease in the specific surface area of the catalysts [10,11]. Therefore, the SO2 tolerance of catalysts should focus on the problem of competitive adsorption with NO and irreversible catalyst deactivation caused by the formation of sulfate.

Much research has been conducted on sulfate on the surface of the catalysts [12]. Yang reported that V2O5 could improve the SO₂ resistance of Fe₂O₃/AC Catalyst for NH3-SCR of NO in dried flue gas but had little effect with water vapor in the flue gas [13]. Jiang added Zr in Fe–Mn/Ti catalysts to alleviate SO₂ poisoning via the Langmuir-Hinshelwood (L–H) method and reported its optimal ratio for Zr/(Ti + Zr) [14]. Zhang studied the activity and temperature window (225~625 °C) of Fe-Cu-SSZ-13 in NH₃-SCR [15]. SO₂ tolerance in the presence of water vapor is still a problem for different varieties of catalysts. However, there have been fewer studies on the SO₂ resistance of RHA-based catalysts. Their rich pore properties and structural stability at high temperatures (<850 °C) may help enhance their SO₂ resistance. In our work, a bimetallic catalyst (Fe and Cu) supported on rice husk ash showed its efficiency in reduction of NO by CO. CO-SCR over Fe-Cu/RHA has the potential to be a replacement for currently industrial used NH₃-SCR. The effect of Fe/Cu on Fe-Cu/RHA and its SO₂ resistance needs further study [2,16,17].

In this work, a study on the performance of Fe-Cu/RHA on CO-SCR and its SO₂ resistance were studied by experiments. Characterization methods were employed to investigate the characteristics of the catalysts. The performance of catalysts with different Fe/Cu (the ratio of moles) was investigated and compared under conditions with different SO₂ concentrations.

2. Materials and Methods

2.1. Catalyst Preparation

The metals were loaded onto the RHA by an impregnation method and RHA was produced by the combustion of rice husk as described previously [2]. Ferric nitrate, copper nitrate, and cobalt nitrate with analytical purity (Aladdin Biochemical Technology Co., Ltd., Shanghai, China) were dissolved in 100 mL deionized water (the iron and cooper mole ratios Fe/Cu were 0.8:0.2, 0.67:0.33, and 0.5:0.5). This was the precursor solution for 20 wt.% Fe-Cu/RHA. Then, 6 g RHA was immersed in the precursor solution and stirred at 80 °C for 5 h. The resulting slurry was dried at 120 °C overnight (18 h) in an oven. Finally, the mixture was calcined at 350 °C for 8 h in a muffle roaster.

2.2. Instruments for Characterization

Brunauer-Emmett-Teller (BET) surface area and pore diameter of catalysts were studied and analyzed in a Max Surface Area and Porosity Analyzer (Microtrac BELSORP, Osaka, Japan).

To observe the morphologies of catalysts, a scanning electron microscope (SEM) was operated with MAIA3 LMH (TESCAN CHINA, Ltd., Shanghai, China) at 15 kV. To further study the particle size and surface discreteness, Cs-corrected scanning transmission electron microscopy (TEM) was carried out on a JEM-ARM200CF (JEOL Ltd., Tokyo, Japan).

The surface atomic states of the catalysts were analyzed by AXIS ULtrabld (Shimadzu CHINA, Ltd., Shanghai, China). The binding energies were calibrated using the C1 peak at 284 eV.

To investigate the active sites on the surface of catalysts, temperature-programmed reduction (TPR) studies were carried out on an automatic chemical adsorption instrument Autosorb-QC-TPX (Quantachrome, Boynton Beach, FL, USA). Before TPR analysis, the pre-dried catalyst samples were heated at 300 °C for 15 min in a stream of He and then cooled to 50 °C. Reduction occurred in the nitrogen stream with 5 mol.% hydrogen. Samples were analyzed from 25 °C to 800 °C at 10 °C/min and the system recorded the desorption data.

Fourier transform infrared reflectance (FTIR) was collected from 400 cm⁻¹ to 6000 cm⁻¹ on a iS50 FTIR spectrophotometer (Nicolet, Madison, USA). The FTIR spectra of samples were recorded using potassium bromide pellets (Sample: KBr = 1: 50).

2.3. Catalytic Activity Testing System

Figure 1 shows the catalyst activity testing system. The catalytic layer (height = 30 mm, diameter = 20 mm) was located in a vertical tube furnace with a temperature-programmed controller. Gas flow was controlled by mass flow controllers. Water after evaporation was added to the N₂ flow as the carrier gas. The components of the flue gas were analyzed by an FTIR flue gas analyzer DX4000 (GASMET, Helsinki, Finland). The main experimental procedure was that simulated flue gas flowed through the reactor and the components of the flue gas at the inlet and outlet were analyzed. The gas hourly space velocity (GHSV) was 11,220 h⁻¹ (the flow rate was 2.4 mL/min). The test temperature ranged from room temperature to 500 °C. The simulated flue gas at the inlet was a mixture of N₂ (carrier gas), NO (560 ppm), CO (5600 ppm), SO₂ (0~280 ppm), and O₂ (6.1%). Repeated trials were conducted to make the results reliable.

3. Result and Discussion

3.1. Characterization Results of Catalysts

BET surface areas and pore volume information of prepared catalysts is listed in

Table 1. Particulate properties of prepared catalysts.

	Fe0.8Cu0.2/RHA	Fe0.67Cu0.33/RHA	Fe0.5Cu0.5/RHA
Surface area (m2g–1)	40.863	35.410	36.991
Average pore size (nm)	6.338	7.795	9.243
Pore volume (cm3g–1)	0.139	0.153	0.143

. All four catalysts had similar specific surface areas. It seems that the effect of Fe/Cu on pore characteristics is insignificant.

The N₂ adsorption/desorption isotherms of catalysts ae shown in Figure 2. According to the internationally recognized classification, the hysteresis characteristics of Fe_0.8Cu_0.2/RHA and Fe_0.67Cu_0.33/RHA show type II isotherms. The curves of the three catalysts are similar, indicating micropores and mesopores accompanied by some slits. The relatively gentle hysteresis loop indicates the existence of flat slit holes in the catalyst. The N₂ curve of Fe_0.5Cu_0.5/RHA isothermal adsorption and desorption has an ⅲ shape with an H3 hysteresis loop. At low pressure, the volume of absorption is very small, and only when the pressure approaches saturation vapor pressure does the volume of absorption increase rapidly due to capillary condensation, which usually occurs as a weak solid-gas interaction in the mesopore on the catalyst surface. It can be concluded that the surface pores of Fe_0.5Cu_0.5/RHA are mainly composed of disorderly wedge-shaped pores formed by flake particles loosely accumulated on the catalyst surface. Among all the tested catalysts, Fe_0.5Cu_0.5/RHA has the smallest hysteresis loop area and the steepest adsorption curve, indicating that it has the smallest adsorption capacity and the smallest desorption adsorption capacity difference.

The results of the SEM study are shown in Figure 3 with a magnification ruler of 5 μm or 2 μm, with small pictures in the lower right corner with a magnification ruler of 500 nm. It can be seen that all these catalyst samples have rough surface morphology. The characteristics of RHA, the dense skin, and the loosely aggregated amorphous SiO₂ spherical particles can be observed in Figure 3. The change of Fe/Cu has little effect on the microscopic morphology of the Fe-Cu/RHA.

The TEM study can be seen in Figure 4 with a magnification ruler of 10–20 nm. It can be seen that the crystalline metal oxide was distributed in the amorphous SiO₂ structure. However, in the three Fe-Cu/RHA catalyst samples, the dispersion of metal oxide crystals on the RHA carrier was different: Fe_0.67Cu_0.33/RHA had the most uniform dispersion, and almost no agglomerated crystals could be observed; the dispersibility of Fe_0.5Cu_0.5/RHA was relatively poor with some gathered clusters of crystals on the surface.

Figure 5 shows the H₂-TPR curves of catalysts. The reduction peaks of Fe-Cu/RHA were roughly in the same range, and all had a main reduction peak and a tail shoulder peak. The main peak was located at 277~320 °C. This reduction peak was due to the reduction process of Cu and Fe³⁺, that is, CuO to Cu⁺, Cu⁺ to Cu, Fe₂O₃ to FeO or Fe₃O_4, and the reduction process of CuFe₂O₄. The tail shoulder peak at 450~550 °C corresponds to the reduction process of FeO and Fe₃O₄ to elemental Fe.

The main peak strength of Fe_0.67Cu_0.33/RHA at 304 °C was the highest, and it can be estimated that Fe_0.67Cu_0.33/RHA had the strongest catalytic activity. According to the literature [18], part of Fe and Cu on the catalyst surface might form a complex with stronger catalytic activity of CuFe₂O₄, and the molar ratio of Fe and Cu in Fe_0.67Cu_0.33/RHA is exactly Fe:Cu = 2:1. From the results we can see that with the increase of Cu and decrease of Fe in the catalyst, the main peak temperature decreased and the shoulder peak weakened, which is consistent with the research results in reference [19]. These H₂-TPR results may indicate that the incorporation of Cu enhanced the reducibility of the Fe active site, and the complex CuFe₂O₄ may play an important role in this catalytic process.

The chemical state and relative concentration of the components on the surface of the catalysts were obtained from XPS (see Figure 6). The peaks at 723 eV and 710 eV indicate the spin orbits of Fe2p_1/2 and Fe2p_3/2, from which Fe was proved to be in the form of Fe₂O₃ in the catalysts. The peak at 933.6 eV suggests that Cu was in the form of CuO in catalysts. In the XPS results, the 933.6 eV peak of Fe_0.67Cu_0.33/RHA was the strongest, while the corresponding peak of Fe_0.5Cu_0.5/RHA was relatively weaker, which may indicate Fe_0.67Cu_0.33/RHA was most active in Fe-Cu/RHA. These test results were consistent with the H₂-TPR results.

Figure 7 shows the infra-red spectra of the fresh catalysts, catalysts after 18 h activity tests (Fe_0.8Cu_0.2/RHA-U, Fe_0.67Cu_0.33/RHA-U and Fe_0.5Cu_0.5/RHA-U) and catalysts after SO₂ resistance tests (Fe_0.8Cu_0.2/RHA-SO₂, Fe_0.67Cu_0.33/RHA-SO₂ and Fe_0.5Cu_0.5/RHA-SO₂). FTIR spectra of all tested samples indicated the characteristic peaks of silica [5,20,21,22]. In all cases there were five peaks at 467 cm⁻¹, 795 cm⁻¹, 1096 cm⁻¹, 1541 cm⁻¹ and 3448 cm⁻¹ associated with SiO₂ in asymmetric, symmetrical or curved modes. This indicates the stability of the supporter. The FTIR spectra of Fe-Cu/RHA catalysts were very similar, and the FTIR spectra of catalysts after 18 h activity tests were also very similar. The catalysts may likely have a similar catalytic process of CO reduction of NO.

There was a peak at 1384 cm⁻¹ in the FTIR spectra of Fe_0.8Cu_0.2/RHA, Fe_0.67Cu_0.33/RHA, and Fe_0.5Cu_0.5/RHA related to nitrite or nitric acid, which weakened in the FTIR spectra of Fe_0.8Cu_0.2/RHA-U, Fe_0.67Cu_0.33/RHA-U, and Fe_0.5Cu_0.5/RHA-U. A probable explanation is that carbonate was generated on the catalyst surface in the CO + NO reaction [19]. There were two remarkable peaks at 2341 cm⁻¹ and 2360cm⁻¹ corresponding to CO₂ and coordination group CO₃²⁻ [19]. It may be that the adsorbed NO reacted with the adsorbed carbon substance CO [23].

The peaks at 1384 cm^-1, associated with antisymmetric stretching vibration of the nitro group in the FTIR spectra of catalysts after SO2 resistance tests, were lower than those in the FTIR spectra of catalysts after the activity tests. It seems possible that the nitro material produced by adsorption of NO on the catalyst surface is decomposed directly. Peaks in connection with CO₂ and coordination group CO₃²⁻ in the FTIR spectra of catalysts after SO₂ resistance tests were similar to those in the FTIR spectra of catalysts after activity tests. It seems that the catalyst inactivation due to SO₄²⁻ and SO₃²⁻ is suppressed to some extent [23].

3.2. Catalytic Activity

Figure 8 presents the NO conversion rates of the catalysts. In the temperature range of room temperature to 500 °C, Fe_0.67Cu_0.33/RHA in this experiment had the best performance, and the NO removal rate hardly changed and remained at about 100%. In Fe_0.8Cu_0.2/RHA at 100 °C~500 °C, the NO removal rate remained at about 100%. When Fe_0.5Cu_0.5/RHA was at a temperature higher than 150 °C, the NO removal rate was about 100%.

This remarkable activity may be due to the effect of the bimetallic compounds CuFe₂O₄. These results are supported by the TEM images (Figure 4) in which Fe_0.67Cu_0.33/RHA had better dispersity of metal oxides on the supporter, the H₂-TPR profiles (Figure 5) in which Fe_0.67Cu_0.33/RHA had the strongest main peak intensity, and the FTIR spectra in which Fe_0.67Cu_0.33/RHA had the strongest peaks corresponding to CO₂ and coordination group CO₃²⁻.

The CO conversion results are shown in Figure 9. All catalysts showed activity in the oxidation of CO. The CO conversion rate changed with the reaction temperature under the action of Fe_0.8Cu_0.2/RHA and Fe_0.67Cu_0.33/RHA. When the temperature was higher than 100 °C, the conversion rate of CO over Fe_0.67Cu_0.33/RHA was about 100%. When the temperature was within 100 °C~500 °C, the conversion rate of CO over Fe_0.8Cu_0.2/RHA increased slightly with the increase in temperature. When the temperature was higher than 300 °C, the conversion rate of CO reached about 100%. The CO conversion rate under the action of Fe_0.5Cu_0.5/RHA increased significantly with the increase in temperature, from less than 40% at 100 °C to about 100% after 400 °C. The activity of Fe_0.5Cu_0.5/RHA was significantly lower than that of Fe_0.8Cu_0.2/RHA and Fe_0.67Cu_0.33/RHA. The temperature of the reaction section was 400 °C and above, making it similar in catalytic effect to the other two Fe-Cu/RHA catalysts. Generally speaking, the temperature of the reaction zone of SCR was higher than 100 °C. Due to the conversion rate of NO and CO, this study recommends Fe_0.8Cu_0.2/RHA and Fe_0.67Cu_0.33/RHA for CO-SCR. The catalysts were suitable for a wide reaction temperature range of 100 °C to ~500 °C, and the denitration rate and CO oxidation rate were close to 100%, resulting in flue gas purification.

The mechanism of NO reduction with CO includes two main steps. First, CO is adsorbed on the surface of the catalyst and the loss of intracrystalline oxygen results in oxygen vacancies, which makes the catalyst surface reductive. Then, NO is adsorbed on the reductive interface, followed by its reaction with another CO. The strong chemical adsorption between the reductive surface and NO decreases the strength of N-O bond in NO, which promotes the reaction between the chemisorbed NO and the gaseous CO. This causes the oxidation of CO along with the reduction of NO on the surface of the catalyst.

3.3. The Effect of SO₂

The SO₂ resistance of the catalysts can play an important role in addressing the issue of catalysts on CO-SCR. NO conversion in experiments on SO₂ resistance is displayed in Figure 10. Within the scope of these experiments, the NO removal rate under the action of Fe_0.8Cu_0.2/RHA and Fe_0.67Cu_0.33/RHA did not decrease with the addition of SO₂, and the increase in concentration and remained at about 100%. That is to say, SO₂ had little effect on the catalytic reduction of NO over Fe_0.8Cu_0.2/RHA and Fe_0.67Cu_0.33/RHA. However, the effect of SO₂ on Fe_0.5Cu_0.5/RHA catalytic reduction of NO was very conspicuous. As Figure 10c shows, when SO₂ was added (70 ppm SO₂), the NO removal rate dropped to below 40%, and as the reaction temperature increased, the extent of catalyst SO₂ poisoning increased. When the reaction temperature reached 500 °C, the NO removal rate almost dropped to 0%. Additionally, increase of SO₂ concentration could gradually increase the extent of catalyst SO₂ poisoning. A likely explanation is that SO₂ is more oxidizing than NO and competes with NO for occupying active sites. SO₂ preferentially occupies reducing active sites with oxygen vacancies. As the reaction temperature increases, the number of active sites occupied by SO₂ might also increase, which leads to a further decrease in the removal rate of NO. Fe-Cu/RHA catalysts have better SO₂ resistance than other catalysts, which may be due to the rich pore properties and structural stability at high temperatures of RHA. The results of our research may help the development of the CO-SCR industry.

Results of CO conversion experiments on SO₂ resistance are displayed in Figure 11a–c and show that over Fe_0.8Cu_0.2/RHA, Fe_0.67Cu_0.33/RHA, and Fe_0.5Cu_0.5/RHA, the CO conversion rate decreased to varying degrees with the addition and increase of SO₂. When Fe_0.8Cu_0.2/RHA and Fe_0.67Cu_0.33/RHA acted, the greatest effect of SO₂ occurred at 200 °C. When Fe_0.8Cu_0.2/RHA was employed, at a reaction temperature range of 100 °C to ~200 °C, catalyst activity decreased with increasing SO₂ concentration. When the SO₂ concentration was 280 ppm, the CO conversion rate was reduced to about 65%.

4. Conclusions

In this research, we studied the activity and SO₂ tolerance of RHA-based Fe-Cu oxide in the reduction of NO by CO. A series of Fe-Cu/RHA were tested by BET, SEM, TEM, H₂-TPR, XPS, and FTIR. Activity and SO₂ resistance were tested at different temperatures. We compared the influence of SO₂ concentration changes on catalytic activity at different reaction temperatures. The conclusions obtained are as follows:

(1)

The catalyst samples all have the characteristics of rough surface morphology and the crystalline metal oxides are distributed in the amorphous SiO₂ structure of RHA.

(2)

The incorporation of Cu may enhance the redox properties of Fe active sites. From the results of H₂-TPR detection, CuFe₂O₄ was detected. Carbonate may be formed and CO₂ and CO₃²⁻ in the two ligands formed in the active sites.

(3)

The prepared catalysts effectively catalyzed the complete removal of NO and the oxidation of CO. Fe_0.67Cu_0.33/RHA had the highest catalytic activity, consistent with the inference of catalyst characterization. The activity of Fe_0.8Cu_0.2/RHA was also very high.

(4)

Fe_0.8Cu_0.2/RHA and Fe_0.67Cu_0.33/RHA have good SO₂ resistance. The NO removal rate hardly changed. However, the CO conversion rate decreased with increasing SO₂ at the lower reaction temperatures. The effect of SO₂ reduced with increasing reaction temperature. The decrease in catalyst activity caused by SO₂ is reversible and mainly caused by the formation of sulfite. The decomposition of sulfites can be accelerated by increasing the temperature.