A Study on the Reaction Characteristics of Vanadium-Impregnated Natural Manganese Oxide in Ammonia Selective Catalytic Reduction

ABSTRACT

This study investigated the effect of adding vanadium (V) to natural manganese oxide (NMO) in ammonia (NH₃) selective catalytic reduction (SCR). The addition of V to NMO decreased the catalytic activity at low temperatures by blocking the active site. However, the enhancement of catalytic activity was achieved by controlling NH₃ oxidation at high temperatures. From the NH₃ temperature programmed desorption and oxygen on/off test, it was confirmed that the amount of Lewis acid site and active lattice oxygen of the catalyst affects the catalytic performance at low temperature

IMPLICATIONS

Recently, NMO and manganese oxide have been reported as SCR catalysts. They usually have only reported the reaction characteristics and catalytic activity on the NH₃ SCR over NMO or manganese/metal oxide catalysts. There are no studies about the effect of addition of V to NMO. Therefore, this study investigates the catalytic activity and reaction characteristics on the NH₃ SCR over NMO and V/NMO, and a new application is proposed based on the conclusions of this study.

INTRODUCTION

Recent attempts to solve some of the numerous environmental problems caused by rapid industrialization have been unsuccessful because of ever-increasing energy consumption.1 These environmental problems have various classifications, but air pollution has emerged as the focus of international regulation because it is not contained in its country of origin. Indeed, air pollution with its unique ability to diffuse into adjacent areas and countries is now threatening the global environment. Oxides of nitrogen (NO_x) and sulfur (SO_x) are two of the main source materials causing air pollution. The amount of SO_x exhaust produced annually is being reduced by the removal of sulfur content in fuels, installation of flue-gas desulfurization facilities, use of clean fuels, etc. More than 70% of NO_x originates from a fixed source such as a thermal power station or industrial facility. Such NO_x species are a continuing problem because their removal is more diffi-cult than the SO_x. At the moment, selective catalytic reduction (SCR) using ammonia (NH₃) as a reductant for NO_x diffusion from a fixed source is recognized as the most technically and economically efficient removal method,2 for which vanadium(V) oxide (V₂O₅) is the most widely used catalyst. The most essential technology of an SCR process is the catalyst.

Bosch and Janssen3 classified the basic metallic oxides for SCR with NH₃ (NH₃ SCR) as V₂O₅, iron(III) oxide (Fe₂O₃), copper(II) oxide (CuO), cobalt oxide (Co₃O₄), chromium(III) oxide (Cr₂O₃), nickel(II) oxide (NiO), cerium(IV) oxide (CeO₂), lanthanum(III) oxide (La₂O₃), praseodymium oxide (Pr₆O₁₁), neodymium(III) oxide (Nd₂O₃), gadolinium oxide (Gd₂O₃), ytterbium oxide (Yb₂O₃), and so on, all of which have been studied extensively. Among these, V₂O₅ was found to have the highest catalytic activity. Furthermore, the catalyst support used in SCR also plays an important role in the efficiency of reduction.4 Bauerle et al.4a reported that V₂O₅ has the highest catalytic activity when titanium dioxide (TiO₂) or aluminum oxide (Al₂O₃) is used as support. Shikada et al.5 found through a systematic study that catalytic activity increases with the use of the TiO₂ support and that the catalytic activities using various supports are in the following order: TiO₂-silicon dioxide (SiO₂) >γ-Al₂O₃ > SiO₂. Additionally, Pearson et al.6 showed that a TiO₂ support in the form of anatase rather than rutile has higher catalytic activity.

Previously, Markvart et al.7 and Choi et al.8 reported that manganese oxide showed excellent catalytic activity in NH₃ SCR at low temperatures, and its reactivity was verified using manganese oxide itself. However, there is almost no report of manganese oxide being positioned as a support but used as a catalyst. Therefore, the role of a support in SCR is very important and can potentially be the basis for manufacture of future catalysts.

Accordingly, this study examined the reaction characteristics and catalytic activity of manganese oxide impregnated with vanadium (V) in NH₃ SCR at low temperatures. The authors further attempted to investigate the factors controlling the catalytic activity and reaction characteristics.

EXPERIMENTAL PROCEDURES

Catalysts Preparation

Natural manganese oxide (NMO) exists mostly in two forms—pyrolusite and psilomelane—and in various other forms such as manganite, braunite, and hausmanite. The NMO used in this study was pyrolusite, composed mainly of β-MnO₂, and its physicochemical properties are listed in Table 1.

Table 1a. Chemical properties of NMO

Component	Mn	SiO2	Al2O3	Fe	CaO	MgO	Balance O2 of Mn and Fe
wt %	51.85	3.13	2.51	3.86	0.11	0.25	38.33

Table 1b. Physical properties of NMO

Mean particle size	0.359 mm
Density	3980 kg/m^3
Pore volume	0.0369 cm^3/g
Surface area	23.29 m^2/g

To increase the catalytic activity of NMO, the amount of V to be impregnated was calculated according to the desired composition ratio, and the amount of calculated ammonium metavanadate (NH₄VO₃; Aldrich Chemical) was dissolved in distilled water heated to 60 °C. To determine the effect of pH on the sample, V/NMO was prepared in the presence of oxalic acid (OA; Aldrich Chemical). OA, corresponding to a 1:1 molar ratio of the calculated quantity of V, was added. After the mixture was prepared in slurry state, it was agitated for more than 1 hr, and its moisture content was evaporated at 70 °C using a rotary vacuum evaporator (NN Series, Eyela).

After drying the material in a dry oven at 110 °C for 24 hr, the samples were prepared by calcination by heating to the desired temperature at a rate of 5 °C/min in a tubular furnace under air for 4 hr. The samples prepared by the addition of V are denoted as “1% V/NMO and 2% V/NMO” and the sample prepared with the addition of OA is denoted as “1% V/NMO OA.”

Catalytic Activity Measurement

The activity test equipment included a gas injector, reactor, and reaction gas analyzer. The flow rate of gas (nitric oxide [NO], nitrogen [N₂], oxygen [O₂], and NH₃) into the reactor from each cylinder was controlled using a mass flow controller (MFC; MKS). Additionally, moisture was supplied to the reactor by injection of N₂ containing water vapor through the bubbler, and this level of moisture was constantly maintained by the circulation of water at a fixed temperature of 50 °C by a circulator outside of the jacket-type bubbler. The entire gas supply pipe was made of stainless steel and was wrapped with a heating band set at 180 °C so that the moisture would not condense.

The continuous-flow fixed-bed reactor that was used had a quartz pipe with an inner diameter of 8 mm and height of 60 cm and quartz wool for fixing the sample layer. The temperature of the reactor was controlled by a proportional-integral-derivative (PID) temperature controller using a K-type thermocouple positioned at the lower part of the fixed bed. Furthermore, the difference between the before and after sample layer temperatures was measured by installing another thermocouple on the upper part of the sample layer to measure the temperature of the gas inlet.

To measure the concentrations of the reactants and products of NO, Uras 10E (Hartman & Braun) and 42C HL (Thermo Instruments) were used. In the case of nitrogen dioxide (NO₂), a detector tube (9L, Gastech) was used when more than 5 parts per million (ppm) of NH₃ was exhausted, and the abovementioned 42C analyzer was used when less than 5 ppm of NH₃ was exhausted. The concentration of NH₃ was measured using the detector tube (3M, 3La, 3L, Gastech).

Characterizations

The specific surface area of the samples was measured with ASAP 2010C from Micromeritics. Here, each specimen was analyzed after being degassed into a vacuum state at 110 °C for 3–5 hr.

To observe the crystal structure of NMO and V/NMO, X-ray diffraction (XRD) analysis was performed and XRD patterns were analyzed with a D/Max-III (3 kW) diffractometer (Rigaku). Cu Kα (λ= 0.1506 nm) was used as a radiation source, and measurements were made in the range of 2θ= 10–90° at a scanning rate of 4°/min.

NH₃ temperature programmed desorption (TPD) measurements were carried out with 50 mg of the sample from room temperature to 400 °C at a rate of 10 °C in flowing argon (Ar) gas (50 ml/min) after exposure of the NH₃/Ar gas (1:99 v/v mixtures with a total flow of 50 ml/min).

X-ray photoelectron spectroscopy (XPS) analysis was performed using ESCALAB 210 (VG Scientific), and Al Kα monochromate (1486.6 eV) was used as an excitation source. After moisture contained in the samples was removed completely through drying at approximately 100 °C for 24 hr, the samples were analyzed without surface sputtering and etching to maintain the vacuum pressure of the XPS equipment at 10⁻⁶ Pa.

RESULTS AND DISCUSSION

Characterizations and Catalytic Activities of NMO and V/NMO

The analysis results of NMO and V/NMO used in this study are listed in Table 2. The pH of the solution used during preparation of the samples ranged from 3.8 to 6.8 as per the amount of added OA. The pH of the 1% V/NMO OA sample solution had a decreasing tendency, whereas that of the V/MNO samples showed a higher pH than that of NMO. Additionally, among OA-added samples, the more the content of V increased, the less the pH decreased. As pointed out earlier, the amount of OA added depends on the amount of V impregnated. A greater amount of impregnated V requires the addition of a greater amount of OA, with pH decreasing drastically. With increasing V contents, specific surface area shows a little decreasing tendency. In Figure 1, the crystalline structure of NMO and V/NMO used in this test, determined by XRD analysis, is shown; all of the samples showed the structure of typical β-MnO₂.

Table 2. Physicochemical properties and NO_x conversion over NMO and V/NMO

	NMO	1% V/NMO	1% V/NMO OA	2% V/NMO	2% V/NMO OA
Mn atom (atoms/cm^3)	78.14	69.53	59.09	59.38	52.52
Vatom (atoms/cm^3)	–	9.76	9.98	10.63	11.22
V/Mn atomic ratio	–	0.14	0.17	0.17	0.21
Surface area (m^2/g)	23.29	22.13	22.40	17.30	15.60
pH in solution (NMO or NMO +V)	5.48	6.74	4.48	6.85	3.82
NOx conversion (%) at 175 °C	62.87	54.45	46.53	49.50	43.56

Figure 1. XRD patterns of NMO and V/NMO.

In Figure 2, the activities of NMO, V/NMO, and V/NMO OA are shown. Although NMO shows the highest catalytic activity in the low-temperature range of less than 350 °C, this activity drastically decreases at 300 °C. In the case of V/NMO, the more the content of V increases, the less the activity decreases and the less activity the OA-added samples show. The cause of these phenomena can be explained by the XPS results in Table 2. The OA-added samples have a greater number of V atoms exposed on the surface than the samples without any OA added. The number of atoms exposed on the surface is used in the same sense as a degree of dispersion, which is often used for representing the turnover frequency (TOF) of catalyst.9 Accordingly, the above result implies that the degree of dispersion increased by the low pH. The samples by V content also show that an increase in the V content results in more V atoms being exposed on the surface. The increased number of V atoms exposed on the surface controls the exposure of NMO on the surface. Accordingly, because V/NMO shows lower activity than NMO in the range of low temperatures because of V atoms having blocked the active site of the NMO surface, it can be concluded that the activity of V/NMO samples at a low temperature is determined by the degree of dispersion of V

Figure 2. Effect of V loading and OA addition on NO_x conversion over NMO (space velocity [SV] = 60,000 hr⁻¹, NO = 190 ppm, NO₂ = 20 ppm, water vapor = 8%, O₂ = 15%, NO_x/NH₃ = 1).

NH₃ and NO Oxidation

To evaluate the SCR reactive characteristics of NMO and V/NMO in the high-temperature range, NH₃ and NO oxidation tests were conducted. The NH₃ oxidation is shown in Figure 3. The experimental conditions were the same as those in the catalytic activity test, except that there was no injection of NO in the oxidation test. The results show that NMO promoted the highest rate of NH₃ oxidation throughout the temperature range, and a higher quantity of V resulted in decreased NH₃ oxidation. NMO showed complete oxidation of NH₃ at 300 °C, whereas the V/NMO samples exhibited incomplete NH₃ oxidation at 350 °C. All samples showed the specific phenomena (lack of nitrogen components regarding balance) in an aspect of nitrogen balance when regarding the NH₃ oxidation, generating NO and NO₂. Therefore, an experiment using mass spectroscopy was performed to explain the phenomena, as presented in Figure 4. With increasing temperature, the amount of NH₃ decreased and N₂, water, and NO increased. Therefore, it was assumed that the generated N₂ corresponded to the lack of nitrogen components in Figure 3.

Figure 3. Effect of temperature on NH₃ oxidation over NMO and V/NMO (SV = 60,000 hr⁻¹, NH₃ = 225 ppm, water vapor = 8%, O₂ = 15%, NO = 0, NO₂ = 0).

Figure 4. Effect of temperature on NH₃ oxidation over NMO as detected by mass spectroscopy (SV = 60,000 hr⁻¹, NH₃ = 225 ppm, water vapor = 8%, O₂ = 15%, NO = 0, NO₂ = 0).

The NO oxidation is shown in Figure 5. The NO oxidation test was carried out under the same conditions as the NH₃ oxidation test (Figure 5). NMO had the highest conversion rate of NO to NO₂ at all reaction temperatures, and a higher quantity of impregnated V resulted in a reduced conversion rate. A higher ratio of NO₂ content in the reactant gas caused an increase in the SCR rate, as in eq 2 of standard SCR (eq 1), which in turn resulted in increased catalytic activity.10 However, the SCR reaction over NMO at high temperature in this study was very slow because there was a shortage of NH₃ because of the conversion of NO to NO₂ (by the fast SCR). Therefore, it was assumed that the low catalytic activity of NMO at high temperature (300 °C), as shown in Figure 2, was because of the fast rate of NO oxidation and a shortage of NH₃ as reductant. Additionally, in the case of V/NMO, the addition of V blocked the active site on NMO and resulted in low activity at low temperature.

Figure 5. Effect of temperature on NO oxidation over NMO and V/NMO (SV = 60,000 hr⁻¹, NO = 190 ppm, NO₂ = 20 ppm, water vapor = 8%, O₂ = 15%).

1

2

Characteristics of NH₃ Adsorption and Desorption

Determining the characteristics of NH₃ adsorption to and desorption from the sample surface is integral to understanding SCR with NH₃. Therefore, a TPD test was performed on NMO impregnated with V.

Results of a typical NH₃ TPD test with NMO are shown in Figure 6. According to Lietti et al.,11 the ad sorbed NH₃ species present on the surface of V₂O₅/TiO₂ are of two types: the molecular form of NH₃ adsorbed at Lewis acid sites (hereafter referred to as “L acid sites”) and NH₄ ⁺ adsorbed at Bronsted acid sites (hereafter referred to as “B acid sites”). As previously shown by Busca et al.,12,13 the NH₃ species adsorbed at L acid sites is more stable than that adsorbed at B acid sites at high temperatures. In the results presented here, the characteristics of the two types of NH₃ species absorbed to NMO were the same as those for V₂O₅/TiO₂. Specifically, NH₃ that desorbed at temperatures lower than 100 °C was adsorbed at B acid sites, which implies that the stable form adsorbed at L acid sites at temperatures higher than 100 °C. The NH₃ species adsorbed at L acid sites required a high temperature for desorption on the basis of the strong interaction with the surface. In contrast, the species desorbed from the weak acid sites was in the form of NO, nitrous oxide (N₂O), or N₂. The oxygen of the produced material originated from the active lattice oxygen or the oxygen adsorbed to the surface because no oxygen was injected.

Figure 6. NH₃ TPD profiles of NMO.

Qingya et al.14 reported that the most important factor in NH₃ SCR is the amount of NH₃ adsorbed at L acid sites. Accordingly, the correlation between the quantity of NH₃ adsorbed at the various acid sites and catalytic activity was surveyed. The values that indicate the amount of NH₃ adsorbed to the weak and strong acid sites as well as the amount of NH₃ adsorbed to the weak acid sites of NMO and V/NMO are shown in Figure 7, along with the conversion rates of NO_x. It was difficult to observe any correlation between NO_x conversion rates at low temperatures and the amount of NH₃ adsorbed to the weak and strong acid sites. Accordingly, the correlation characteristics between the amount of NH₃ adsorbed at L acid sites and catalytic activity are shown in Figure 8. The amount of NH₃ adsorbed at L acid sites significantly affects the catalytic activity at low temperatures but does not follow a linear relationship at high temperatures. This is because the catalytic activity of NMO decreased rapidly. As mentioned previously, the catalytic activity of NMO decreased rapidly at temperatures higher than 250 °C upon the fast oxidation of NH₃. These characteristics seem to suggest that the amount of NH₃ adsorbed at L acid sites is correlated only with catalytic activity at low temperatures. Additionally, the above results imply that the rate of NO_x conversion at high temperatures is more dependent on the catalytic oxidation rate than on the amount of adsorbed NH₃.

Figure 7. Comparison of NO_x conversion with all desorbed nitrogen compounds from (a) B and (b) total acid sites over various samples.

Figure 8. Comparison of NO_x conversion with all desorbed nitrogen compounds from L acid sites over various samples.

Redox Characteristics

The main factors affecting SCR are the redox and adsorption properties of the sample, especially adsorption to the acid sites. As such, the redox properties at low temperatures play an important role in SCR.15–17 These redox properties depend on the lattice oxygen of the sample, the reductant NH₃, and gaseous oxygen. However, the role of oxygen is controversial. Absence of NO from the reaction because of decomposition to N₂ and O₂ encourages the participation of the other two types of oxygen—the active lattice oxygen and gaseous oxygen—in the reaction. Therefore, the reaction mechanism can be categorized into the following three types according to the different types of oxygen.

The first category involves the lattice oxygen supplemented with gaseous oxygen. The second involves direct participation of gaseous oxygen in the reaction. Finally, the last category comprises oxygen changing the oxidation state of the sample without participation in the reaction.

Regarding the reaction mechanism of oxygen, Lietti et al.17 stated that SCR occurs by a redox mechanism and that the oxygen participates in the reoxidation of the reduced catalyst.

To determine the degree of participation of the lattice oxygen in the sample reaction, an O₂ on/off test was performed, and the results are shown in Figure 9. Upon ceasing the administration of oxygen, the rate of NO_x conversion decreased instantaneously, which was followed by increases at 200 °C and 180 °C as well as another decrease as time passed. This result is attributed to the gaseous oxygen and the lattice oxygen participating in the SCR reaction. Additionally, the higher the reaction temperature, the higher was the conversion rate. Therefore, the degree of participation of the lattice oxygen in the reaction increased with increasing temperature.

Figure 9. Decrease in NO_x conversion with time after O₂ on-off over NMO (SV = 60,000 hr⁻¹, NO = 190 ppm, NO₂ = 20 ppm, water vapor = 8%, O₂ = 15%, NO_x/NH₃ = 1).

The O₂ on/off test was performed to determine the degree of participation of the lattice oxygen of NMO and V/NMO in the SCR reaction (Figure 10). Upon ceasing the administration of oxygen, NMO impregnated with V experienced a significant decrease in the rate of NO_x conversion that continued to decrease over time. On the other hand, although simple NMO also initially experienced a significant decrease in the rate of NO_x conversion, this rate remained constant for approximately 30 min.

Figure 10. Decrease in NO_x conversion with time after O₂ on-off over NMO, 1%V/NMO, and 2%V/NMO (SV = 60,000 hr⁻¹, NO = 190 ppm, NO₂ = 20 ppm, water vapor = 8%, O₂ = 15%, NO_x/NH₃ =1 at 200 °C).

Thereafter, the rate of NO_x conversion showed the same gradually decreasing trend as observed for V/NMO. The active lattice oxygen participates in the SCR reaction when no other oxygen is available. However, impregnation of the sample with V results in a decrease in lattice oxygen participation. The amount of active lattice oxygen was calculated relatively, and the calculation result is listed in Table 3. Choi et al.8 and Seo et al.18 reported that the higher the amount of activated lattice oxygen, as calculated by the O₂ on/off test, the higher is the catalytic activity at low temperatures. In this study, the high catalytic activity of NMO at low temperatures was therefore due to a large amount of active lattice oxygen.

Table 3. Relative amount of active lattice oxygen in NMO and V/NMO

	NMO	1%V/NMO	2%V/NMO
Relative amount of active lattice oxygen	1	0.34	0.33

CONCLUSIONS

The following conclusions were drawn based on results of various tests of NMO impregnated with V: • Samples impregnated with V resulted in decreased catalytic activity at low temperatures. However, V/NMO exhibited higher activity than NMO at high temperatures. This result is closely linked with the oxidation rate of NH₃ and is mainly attributed to decreased catalytic activity at high temperatures because of the shortage of reductant NH₃.

•

The activity of the SCR catalyst at lower temperatures is closely linked to the amount of NH3 adsorbed at L acid sites on the catalyst bed and to the amount of active lattice oxygen. Therefore, the reaction characteristics and activity of SCR at low temperatures using NMO impregnated with V are important factors affecting the adsorption characteristics of NH3 and the participation of lattice oxygen in the reaction.

•

Upon addition of V to NMO, catalytic activity was reduced through the limitation of available lattice oxygen on the sample bed and the adsorption of NH3 at L acid sites that also blocks the active site. However, in the case of V/NMO without any OA added, relatively excellent activity was observed even in the low-temperature range. Therefore, V/NMO can potentially be used for the development of a catalyst functional over a wide temperature range.