Improving the activity of Mn/TiO₂ catalysts through control of the pH and valence state of Mn during their preparation

Abstract

In this study, the authors investigated the influence of the valence state of Mn on the efficacy of selective catalytic reduction using a Mn-based catalyst. The nitrogen oxides (NO_x) conversion rate of the catalyst was found to be dependent on the type of TiO₂ support employed and on the temperature, as the catalyst showed an excellent conversion of > 80% at a space velocity of 60,000 hr⁻¹ when the temperature was above 200 °C. Brunauer-Emmett-Teller, X-ray diffraction, and X-ray photoelectron spectroscopy analyses confirmed that catalyst displaying the highest activity contained the Mn⁴⁺ species and that its valence state was highly dependent on the pH during the catalyst preparation.

Implications

Recently, various Mn catalysts have been evaluated as selective catalyst reduction (SCR) catalysts. However, in these previous studies, only the reaction characteristics and catalytic activity on the NH₃ SCR over Mn catalysts were evaluated. There have been no studies on the effect of pH during catalyst preparation. Therefore, in this study, the effect of pH during the catalyst preparation process was examined and a new application of the Mn catalysts was proposed based on the current findings.

View correction statement:

Erratum

Introduction

Emission of nitrogen oxides (NO_x) produced during the combustion of fossil fuels such as natural gas, oil, and coal can cause diminished visibility, green house effects, and acid rain. Combustion also produces two of the most detrimental environmental pollutants, O₃ and peroxyacetyl nitrate (PAN), which are components of photochemical smog that bond with oxygen under ultraviolet (UV) light (Bosch and Janssen, 1988). Accordingly, the demand for more eco-friendly and cost-effective technologies to remove NO_x from fixed sources of nitrogen oxides is increasing, and selective catalyst reduction (SCR) is the most prevalent method developed to meet these demands in terms of reliability and affordability.

SCR involves the installation of a catalyst coated with reducing agents such as NH₃ inside a chimney through which exhaust gases are expelled, which selectively decomposes NO_x into harmless substances such as N₂ and H₂O₂. Catalyst development is an essential aspect of this process, since the position of the catalyst within the chimney and the operating temperature affect the catalytic activity. Currently, V₂O₅/TiO₂ catalysts are the most widely used commercial catalysts for SCR, with WO₃ or MoO₃ frequently added as co-catalysts. In order to maximize SCR activity, the catalyst requires a sufficiently high temperature (above 300 °C) and low space velocity. Because reheating the exhaust gas to maintain sufficient catalytic activity would be enormously expensive (Kang et al., 2007; Peña et al., 2004), the development of low-temperature SCR catalysts is essential to addressing the current problems of commercial catalysts and the efficient removal of NO_x. An additional consideration is that SCR catalysts are often deactivated by SO₂, so it must be utilized after the flue gas desulfurization (FGD) process has removed SO₂ from the gas emissions.

Toward this end, various transition metals such as Mn, Cu, Cr, Co, and Fe have been placed onto a variety of supports such as Al₂O₃, SiO₂, and TiO₂ to develop improved catalysts for SCR, which have shown excellent activity in the low temperature range. In particular, Mn-based catalysts placed on supports such as TiO₂ (Qi and Yang, 2003; Smirniotis et al., 2001), Al₂O₃ (Singoredjo et al., 1992), NaY (Bentrup et al., 2001), Ultra Stable Y zeolite (USY) (Qi et al., 2003), Activated Carbon (AC) (Marbán and Fuertes, 2001; Yosikawa et al., 1998), and natural manganese oxides (Kim et al., 2011) have been extensively studied, as they readily undergo the redox cycle and the oxidation state of manganese ions can be easily changed (Kim and Hong, 2007).

Sol-gel studies conducted on Mn-based catalysts have shown that structure has the largest effect on activity, which is dependent on the manufacturing method but independent of Mn content (Bai et al., 2009; Mohamed et al., 2007; Schmpf et al., 2002).

Singoredjo et al. (Singoredjo et al., 1992) observed that Mn-based catalysts displayed the highest activity in the temperature range of 107–297 °C and that the activity was dependent on the oxidation state of manganese. In addition, Peña et al. (Peña et al., 2004) showed that the catalysts containing Mn produced the highest NO_x conversion in SCR activity tests. Using X-ray diffraction (XRD) and X-ray photoelectron spectroscopy (XPS) analysis, these authors found that Mn⁴⁺ was the predominant Mn species on the catalyst surface. In contrast, Li et al. (Kim et al., 2011) reported that the superior activity of the catalyst was caused by the high Mn/Ti ratio on the catalyst surface as well as the presence of amorphous Mn₂O₃ species.

Several studies examining the effects of the valence state of manganese and the morphology of Mn/TiO₂ catalysts have been reported (Ettireddy et al., 2007; Kapteijn et al., 1994; Li et al., 2007; Park et al., 2001; Peña et al., 2004; Singoredjo et al., 1992,; Yamashita and Vannice, 1996). However, no previous study has evaluated the nature of the Mn oxidation state, which we addressed in this study along with the relationship between Mn valence and catalytic activity.

Experimental

Catalyst Preparation

The physical characteristics of the eight types of commercial TiO₂ used as supports are presented in Table 1; Table 2 lists their compositions. The solution-impregnation method was used to prepare the Mn/TiO₂ catalysts. Manganese(II) nitrate hydrate (Mn(NO₃)₂·xH₂O, 98%) was dissolved in 50 mL of distilled water to produce a bright red solution. TiO₂ was mixed in with gentle stirring, maintaining the content of Mn in TiO₂ at 10 wt% for all samples. After 1 hr of agitation, the water was evaporated under vacuum at 0.086 atm and 70 °C using a rotary vacuum evaporator. Evaporated samples were dried for 24 hr at 110 °C in a dry oven to remove residual water. The samples were then calcined and prepared in a tubular furnace by increasing the velocity from 10 to 400 °C/min over 4 hr.

Table 1. Physical properties of various TiO₂ supports tested

TiO2	Name	Manufacturer	Anatase:Rutile (%:%)	BET Area (m^2/g)	Particle Size (μm)	Crystallite Size (Å)
TiO2(A)	MC-50	Ishihara	100:0	70.964	1.2450	210
TiO2(B)	MC-90	Ishihara	100:0	77.14	1.1900	185
TiO2(C)	DT-51	Millennium	100:0	84.053	1.1950	206
TiO2(D)	P-25	Degussa	75:25	45.431	3.6100	246
TiO2(E)	UV100	Hombikat	100:0	196.42	2.3350	85.33
TiO2(F)	ST-01	Ishihara	100:0	220.68	1.1750	155.5
TiO2(G)	G-5	Millennium	100:0	220.13	1.3650	83.66
TiO2(H)	Sigma-Aldrich's anatase TiO2*	Sigma-Aldrich	100:0	9.9804	1.7600	419

Table 2. The composition of various TiO₂ supports tested

	Ca	Mg	Na	K	W	Nb	Fe	S	SO4
TiO2(A)	0.005	0.002	ND	ND	ND	ND	ND	0.0007	2 ppm
TiO2(B)	0.013	0.012	2.97	0.018	ND	0.065	0.005	0.7	2.1
TiO2(C)	0.14	0.11	0.035	0.017	0.1	0.12	0.0043	0.37	1.11
TiO2(D)	0.0099	0.0027	0.13	0.057	ND	0.012	0.004	0.0019	59 ppm
TiO2(E)	0.0128	0.0024	0.035	0.0053	ND	0.0031	0.0013	0.014	421 ppm
TiO2(F)	0.008	0.0013	0.061	0.0064	ND	0.0068	0.0052	0.0357	1071 ppm
TiO2(G)	0.008	0.001	ND	ND	ND	ND	ND	0.0008	1 ppm
TiO2(H)	0.0177	0.0023	0.031	0.083	ND	0.024	0.0018	0.0045	135 ppm
Note: Unit: wt% except where indicated otherwise.

To adjust the pH of TiO₂, nitric acid (HNO₃, 60–62%) or ammonium hydroxide (NH₄OH, 25–28%) was added to the TiO₂ slurry. The pH of the slurry was continuously monitored during adjustment using a pH meter (Ecoscan pH 6). Manganese(II) nitrate hydrate was then added into the pH-adjusted TiO₂ slurry using the same process described above.

Characterization

The specific surface area of TiO₂ was measured using an ASAP 2010C and the Brunauer Emmett-Teller (BET) formula. For this analysis, the samples were degassed under vacuum at 110 °C for 3–5 hr prior to analysis.

The structure of the materials was examined by XRD using Cu Kα radiation. Diffraction peaks, which were recorded in a 2θ range of 10° to 90°, were used to identify the structure of the sample.

XPS analysis was performed using an instrument equipped with an Al Kα monochromate (1486.6 eV) excitation source. After the catalysts were completely desiccated by drying at 100 °C for 24 hr, they were analyzed without surface sputtering and etching under vacuum maintained at 10–12 mm Hg. Wide-scan spectra were taken to determine the binding energies and intensities of Mn, Ti, O, and C in the samples.

Inductively coupled plasma (ICP) analysis was performed using a Perkin-Elmer Optima 3000XL (PerkinElmer Inc., USA) with a Radio Frequency (RF) power level of 1300 watts, 15 L/min plasma flow, 0.5 L/min coolant flow, and 0.8 L/min nebulizer flow. In a Teflon bottle, 0.1 g of the sample was decomposed with 2 mL of acid (HF, HNO₃, and HClO₄, 4:4:1 v/v), then diluted with distilled water prior to ICP analysis. Ion chromatography (IC) was conducted using a DIONEX-120 Automated Dual Column IC (Dionex Co., USA) equipped with a AS4A SC 4-mm column using a 1 mL/min flow rate, 1.8 mM Na₂CO₃/1.7mM NaHCO₃ eluent, and a ASRS-I 4-mm anion self-regenerating suppressor (Dionex Co., USA).

Activity Test

The activity test apparatus consisted of a gas injector, a reactor, and a reaction gas analyzer. The amount of gas supplied to the reactor was regulated using a mass flow controller (MFC; MKS Co., USA) on individual cylinders of N₂, O₂, NO (1% NO/N₂), and NH₃ (1% NH₃/N₂). The reaction conditions were as follows: 200 ppm NO, 200 ppm NH₃, 8% O₂, 8% H₂O, balance N₂, and 500 mL/min total flow rate. Water was supplied in conjunction with N₂ injection through a bubbler to the reactor. Water at 45 °C was circulated in a double jacket outside of the bubbler to maintain the temperature of the reactor using a circulator (CW-05G; JEIO TECH. Co., Korea). The gas supply line was made of stainless steel and maintained at 180 °C using a heating band wrapped around the line to prevent the formation of salts such as NH₄NO₃, NH₄NO₂, and NH₄HSO₄, which are produced by the reaction between NO and NH₃ and condensation in the reaction gases.

The continuous-flow reactor had a diameter of 8 mm and height of 65 cm. A quartz tube was used to fix the catalyst layer. The temperature of the reactor was regulated with a PID temperature controller using a K-type thermocouple installed on the lower part of the fixed layer, and the temperature above and below the catalyst layer was measured by the same type of thermocouple installed on the upper catalyst layer near the gas injector.

The concentration of NO was measured using an Non Dispersive Infrared absorption (NDIR) gas analyzer (ZKJ-2; Fuji Electric Co., Ltd., Japan), and the NO₂ (9L: NO₂ [0–300 ppm]) and ammonia concentrations (3M: NH₃ [0–800 ppm]; 3La: NH₃ [0–200 ppm]; 3L: NH₃ [0–60 ppm]) were measured using a detector tube (GASTEC Co., Japan), respectively. In addition, water from all gases was removed using a water trap in the chiller, which was a cooling bath maintained at −40 °C (CTB-10; JEIO TECH. Co., Korea) that was positioned before the analyzer and detector tube.

Results and Discussions

Effect of the Support Type on Catalytic Activity

Activity tests of Mn/TiO₂ catalysts were carried out at a space velocity of 60,000 hr⁻¹ to investigate the influence of different TiO₂ supports. The activity was found to vary depending on the type of TiO₂ support employed (Figure 1). The highest conversion (>80%) was observed for Mn/TiO₂ (A) at 200 °C, whereas the Mn/TiO₂ (H) displayed the lowest activity (<20%). The order of catalytic activity was as follows: A > B > C > D > F > E > G > H.

Figure 1. Effect of reaction temperature on NO_x conversion over 10 wt% Mn/TiO₂.

Figure 2 shows the results of the XRD analysis of the TiO₂ supports and Mn/TiO₂ catalysts. Except for the TiO₂ (D) sample, which contained 25% rutile phase, all catalysts were in the anatase phase. The intensity and sharpness of the TiO₂ peak did not vary significantly, indicating that the deposited Mn was highly dispersed on the surface of the support and existed in the amorphous phase (Peña et al., 2004).

Figure 2. X-ray diffraction spectra of the catalysts.

Effect of Valence State on Catalyst Characteristics

Mn is a multivalent metal that adopts different valence states based on the preparation conditions and temperature, which is relevant because SCR reaction efficiencies are dependent on oxidation state (Kapteijn et al., 1994; Park et al., 2001; Singoredjo et al., 1992,). XPS analysis was performed to assess the effect of the oxidation state on the catalytic activity (Figure 3). The spectra were background-corrected based on the C–C bonding (284.6 eV) of the C 1s peak. Figure 3a shows the wide-scan spectrum of the Mn/TiO₂ (A) catalyst, which displayed excellent activity. These peaks were typical for the Mn/TiO₂ catalyst and indicated that the sample contained the component atoms C, O, Mn, and Ti. Ti can be identified from the unique bonding energy of the internal electron. The Mn 2p spectrum of Mn/TiO₂ (A) is shown in Figure 3b. The peaks corresponding to MnO₂ (Mn⁴⁺), Mn₂O₃ (Mn³⁺), and Mn(NO₂)₂ revealed bond energies of 642.2 ± 0.4 (Peña et al., 2004; Song et al., 2009), 641.2 ± 0.2 (Peña et al., 2004; Song et al., 2009), and 643.8 (Peña et al., 2004; Park et al., 2001) eV, respectively. When manganese nitrate was used as the Mn precursor in the preparation process and the sample was calcined at a low temperature, it was not completely removed (Peña et al., 2004). Pêna et al. (Peña et al., 2004) performed XPS analysis on a catalyst prepared using manganese nitrate and determined that calcination reduced the intensity of the manganese nitrate peak, in agreement with our results.

Figure 3. (a) Survey spectra measured for 10 wt% Mn/TiO₂ (A) catalysts by XPS analysis. (b) Deconvoluted XPS result of Mn 2p for 10 wt% Mn/TiO₂ (A).

The atomic surface concentrations of Mn/TiO₂ catalysts are shown in Table 3. The Mn/TiO₂ (A), Mn/TiO₂ (B), and Mn/TiO₂ (C), which all displayed excellent activity, had Mn⁴⁺/Mn³⁺ ratios ranging from 1.86 to 2.66, whereas the Mn⁴⁺/Mn³⁺ ratios of the other, poorer catalysts ranged from 0.77 to 0.96. The correlation between the Mn⁴⁺/Mn³⁺ ratio and catalytic activity is shown in Figure 4. Higher Mn⁴⁺ content also correlated with higher NO_x conversion rates, meaning that catalysts with the Mn⁴⁺ species present on the surface displayed excellent SCR activity. Kapteijn et al. (Kapteijn et al., 1994), who investigated SCR activity of manganese oxides on γ-Al₂O₃ supports, reported that MnO₂ (Mn⁴⁺) was an excellent catalytic metal oxide in terms of its ability to reduce NO. Pêna et al. (Peña et al., 2004) observed similar results, providing further support for our observed result that valence state affects the activity of Mn-based catalysts and that the Mn⁴⁺ species provided the highest activity.

Table 3. Atomic surface concentrations of the catalysts determined by XPS

	TiO2(A)	TiO2(B)	TiO2(C)	TiO2(D)	TiO2(E)	TiO2(F)	TiO2(G)	TiO2(H)
MnO2 (642.2 eV)	7108.0	4864.3	4102.1	2538.2	3120.9	1340.9	2855.2	3648.8
Mn2O3 (641.2 eV)	2667.5	2199.9	2203.0	2903.3	3480.3	1713.5	2952.4	4707.8
Mn-nit (643.8 eV)	2228.1	—	—	1751.9	—	1664.9	1662.0	568.05
Mn^4+/Mn^3+ (atomic ratio)	2.6646	2.2112	1.8620	0.8743	0.8968	0.7836	0.9671	0.774
Note: Unit: number of atoms/cm^3.

Figure 4. Correlation between the ratio of Mn oxidation species and NO_x conversion over 10 wt% Mn/TiO₂.

Effect of pH on the Mn Valence State

It has been shown that the pH of the manganese solution used to prepare the catalyst affects the stability of the Mn oxidation state (Stone, 1983). Accordingly, we examined the effect of the pH of the TiO₂ aqueous slurry during the catalyst preparation process on the Mn valence state. The pH of each TiO₂ aqueous slurry is listed in Table 4, and the correlation between the pH of the TiO₂ aqueous slurry and the rate of NO_x conversion is shown in Figure 5a. The latter were shown to be linearly related, with catalysts prepared at low pH values displaying excellent NO_x conversion rates. As shown in Figure 5b, the Mn⁴⁺/Mn³⁺ ratio decreased with a decrease in the pH of the slurry.

Table 4. The potential of hydrogen over various TiO₂ supports

	TiO2(A)	TiO2(B)	TiO2(C)	TiO2(D)	TiO2(E)	TiO2(F)	TiO2(G)	TiO2(H)
pH	2.51	2.50	3.99	3.87	6.30	5.42	6.06	7.03

Figure 5. (a) Correlation between TiO₂ pH and NO_x conversion over 10 wt% Mn/TiO₂. (b) Correlation between TiO₂ pH and Mn valence state over 10 wt% Mn/TiO₂.

In order to better understand these results, one TiO₂ support was selected and 10 wt% Mn/TiO₂ catalysts were prepared at 7 different pH values ranging from 1.5 to 9.15. TiO₂ (G), which displayed a medium activity level, was selected for these experiments. Before the Mn was impregnated, the pH was adjusted by adding nitric acid or ammonium hydroxide to the TiO₂ slurry. Table 5 shows the pH values used during the preparation of Mn/TiO₂ (G); 4 represents the condition used without adjusting the pH, 1–3 represent the addition of nitric acid, and 5–7 represent the addition of ammonium hydroxide. The results of the SCR activity tests of the catalysts prepared under the above conditions are shown in Figure 6. The catalysts prepared at lower pH values displayed higher activities. XPS was used to determine the Mn valence state of catalysts 1–7 prepared with varying pH conditions. The atomic surface concentrations of each peak corresponding to catalysts 1–7 are shown in Table 6. As shown in Figure 5, the valence state of Mn was found to be dependent on the pH of the TiO₂ aqueous slurry. Thus, the Mn/TiO₂ catalytic activity was closely correlated to the oxidation state of Mn and to the pH during the catalyst preparation. These finding clearly demonstrate that the activity can be enhanced by adjusting the pH of the TiO₂ aqueous slurry.

Figure 6. Effect of TiO₂ pH on NO_x conversion over 10 wt% Mn/TiO₂ (G).

Table 5. The potential of hydrogen TiO₂(G)

	1	2	3	4	5	6	7
pH	1.53	2.77	4.3	6.06	6.87	8.01	9.17

Table 6. Atomic surface concentrations of the catalysts determined by XPS

	1	2	3	4	5	6	7
MnO2 (642.2 eV)	3655.58	2898.51	2284.69	1794.30	1979.57	1600.69	2384.12
Mn2O3 (641.2 eV)	1737.80	1404.06	1847.14	1564.38	1989.82	2231.49	2998.73
Mn-nit (643.8 eV)	1063.47	1164.31	1164.31	1164.31	1542.48	351.24	1117.61
Mn^4+/Mn^3+	2.10	2.06	1.24	1.15	1.99	0.72	0.80
Note: Unit: number of atoms/cm^3.

Conclusions

We found that the catalytic activity of the Mn/TiO₂ catalysts studied was dependent on the type of TiO₂ support used. XPS and XRD analysis confirmed that high catalytic activity was achieved when the Mn⁴⁺ species was predominant on the catalyst surface, and that the Mn valence state was dependent on the pH of the TiO₂ aqueous slurry during the catalyst preparation process.

Acknowledgments

This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (2010-0023963).