www.elsevier.com/locate/apcata

Applied Catalysis A: General 327 (2007) 261–269

Manganese oxide catalysts for NOx reduction with NH3

at low temperatures

Min Kang a, Eun Duck Park b, Ji Man Kim c, Jae Eui Yie d,*a R&D Center, EnD Solutions Co. Ltd., 223-235, Seoknam-Dong, Seo-Gu, Incheon 405-220, Republic of Korea

b Department of Chemical Engineering, Division of Chemical Engineering and Materials Engineering, Ajou University, Wonchun-Dong,

Yeongtong-Gu, Suwon 443-749, Republic of Koreac Functional Materials Laboratory, Department of Chemistry and Sungkyunkwan Advanced Institute of Nano Technology,

Sungkyunkwan University, Cheoncheon-Dong, Jangan-Gu, Suwon 440-746, Republic of Koread Catalyst and Surface Laboratory, Department of Applied Chemistry, Ajou University, Wonchun-Dong,

Yeongtong-Gu, Suwon 443-749, Republic of Korea

Received 20 December 2006; received in revised form 15 May 2007; accepted 17 May 2007

Available online 21 May 2007

Abstract

Manganese oxide catalysts prepared by a precipitation method with various precipitants were investigated for the low temperature selective

catalytic reduction (SCR) of NOx with NH3 in the presence of excess O2. Various characterization methods such as N2 adsorption, X-ray diffraction

(XRD), X-ray photoelectron spectroscopy (XPS), thermal gravimetric analysis (TGA) and X-ray absorption near edge structure (XANES) were

conducted to probe the physical and chemical properties of MnOx catalysts. The active MnOx catalysts, precipitated with sodium carbonate and

calcined in air at moderate temperatures such as 523 K and 623 K, have the high surface area, the abundant Mn4+ species, and the high

concentration of surface oxygen on the surface. The amorphous Mn3O4 and Mn2O3 were mainly present in this active catalyst. The carbonate

species appeared to help adsorb NH3 on the catalyst surface, which resulted in the high catalytic activity at low temperatures.

# 2007 Elsevier B.V. All rights reserved.

Keywords: NO reduction; Manganese oxides; NH3-SCR; Precipitation method; Calcinations

1. Introduction

The emission control of nitrogen oxides (NO, NO2 and N2O)

from various combustion processes has been a major

environmental concern related to the air quality because

nitrogen oxides have been reported as one of the most serious

pollutants causing acid rain along with sulfur dioxide and play a

major role in the photochemical chain reaction leading to the

formation of photochemical smog. The primary nitrogen oxides

(NOx) produced from combustion processes with fossil fuel are

nitric oxide (NO) and nitrogen dioxide (NO2), which are 90–

95% of the total NOx emission from automotive exhausts and

stationary sources such as thermoelectric power plant and

incinerator [1].

* Corresponding author. Tel.: +82 31 219 2513; fax: +82 31 219 2394.

E-mail address: yie@ajou.ac.kr (J.E. Yie).

0926-860X/$ – see front matter # 2007 Elsevier B.V. All rights reserved.

doi:10.1016/j.apcata.2007.05.024

The removal of nitrogen oxides from stationary or mobile

sources has become an important issue and a variety of

reduction methods to minimize the emission of NOx such as a

combustion control and a post-combustion control technology

have been developed. Among the proposed post-combustion

methods for NOx removal, the technologies using the catalysts

are known as one of most efficient methods in terms of

relatively low cost and high efficiency. Especially, the selective

catalytic reduction (SCR) of nitrogen oxide has been generally

recognized as the most effective and widely commercialized

removal technology of NOx emitted from the stationary sources

[1–3].

In order to convert NO contained in the flue gas into N2, the

reducing agent must be employed. NH3, CO, H2 and a variety of

hydrocarbons such as methane, propylene, and ethane have

been employed as reductants for NO removal reaction [4].

Although a number of reducing agents can be utilized in SCR,

ammonia is the most effective and widely commercialized,

which is called NH3-SCR, for stationary sources such as power

M. Kang et al. / Applied Catalysis A: General 327 (2007) 261–269262

plants and nitric acid plants [5]. The selective catalytic

reduction (SCR) is a process in which a reducing agent, e.g.

NH3, reacts selectively with NOx to produce N2 without

consuming an excess O2.

In the past few decades, the backbone of SCR technology is

the development of SCR catalysts such as noble metals [6],

supported metal oxides [7], zeolites [8] and others [1,9,10].

Among them, vanadia supported on titania is known to be the

most effective and widely used commercial SCR catalyst due to

its high activity and durability to sulfur compounds. Because

this catalyst exhibits high conversions only in the temperature

range of 573–673 K, the SCR should be applied before units for

particle removal and desulphurization where the gas tempera-

ture decreases [11]. However, when the flue gas has high

concentrations of particles and other contaminants which are

deleterious for the catalyst, proper units should be located at the

upstream of the catalyst bed to resolve above problems, which

causes the decrease of the exit gas temperature. Therefore, there

is a great interest in the development of SCR catalysts active at

low temperatures (<573 K).

A number of catalysts consisted of various transition metal

(V, Cr, Mn, Fe, Co, Ni and Cu) oxides on different commercial

supports such as silica and alumina have been studied. Among

these catalysts, manganese oxides such as MnOx/Al2O3 [12],

MnOx/NaY [13], MnOx/USY [14] and MnOx/TiO2 [15,16]

have attracted much interests due to their high catalytic

activities. These catalysts were prepared by the solution

impregnation method on supports using manganese nitrate or

acetate. In case of unsupported metal oxides, only limited

works have been reported because unsupported MnOx catalysts

suffer from very low surface areas [17]. Some additives such as

citric acid were added in the preparation step to increase the

surface area as well as the catalytic activity [18].

Recently, we found that MnOx catalysts prepared by a

simple precipitation method with sodium carbonate showed the

high catalytic activity for low temperature NH3-SCR [19]. This

catalyst also appeared to be stable in the presence of excess

water vapor. In the present study, we investigated the effect of

preparation methods including kinds of precipitants and

calcinations temperatures of the MnOx catalysts on their

structural features and catalytic performance for selective

catalytic reduction of NOx with ammonia.

2. Experimental

2.1. The preparation of catalysts

The various kinds of manganese oxides were prepared by a

precipitation method with different precipitants such as

ammonium carbonate (AC), potassium carbonate (PC), sodium

carbonate (SC), ammonium hydroxide (AH), potassium

hydroxide (PH) and sodium hydroxide (SH). Each catalyst is

denoted as MnOx-AC, MnOx-PC, MnOx-SC, MnOx-AH,

MnOx-PH and MnOx-SH, respectively. 0.5 M ammonium

carbonate ((NH4)2CO3, DAEJUNG, 99.5%) aqueous solution,

0.5 M potassium carbonate (K2CO3, DAEJUNG, 99.5%)

aqueous solution, 0.5 M sodium carbonate (Na2CO3, SHINYO,

99.5%) aqueous solution, ammonium hydroxide (NH4OH,

DAEJUNG, 25.0–28.0%) solution, potassium hydroxide

(KOH, DAEJUNG, 99%) or sodium hydroxide (NaOH,

SAMCHUN, 99%) was continuously added to 500 ml of

0.5 M manganese nitrate (Mn(NO3)2�xH2O, Aldrich, 98.0+%)

aqueous solution until the pH of the solution reached 8. The

resulting precipitate was aged at 298 K for 1 h, filtered, and

washed several times with distilled water. The cake was dried in

air at 393 K for 12 h and calcined in static air at different

temperatures such as 523 K, 623 K, 723 K and 823 K.

2.2. Catalyst characterization

X-ray diffraction (XRD) patterns were recorded on a Rigaku

D/MAC-III using Cu Ka radiation (l = 0.15406 nm), operated

at 50 kV and 30 mA (1.5 kW). BET surface areas were

calculated from N2 adsorption data that were obtained using

Autosorb-1 apparatus (Quantachrome) at liquid N2 tempera-

ture. Before the measurement, the sample was degassed for 12 h

at 150 8C. The amount of adsorbed NH3 and NO was measured

at 300 K by a pulse adsorption method using helium as a carrier

gas. X-ray photoelectron spectroscopy (XPS) data were

obtained with an Mg Ka (1253.6 eV) X-ray source using

ESCA2000 (VG Microtech) instrument. The binding energy of

C 1s (284.7 eV) was used as an internal standard. The thermal

gravimetric analysis (TGA) was carried out using a Perkin-

Elmer Series 7 thermal analysis system under a flow of dry air.

The temperature was raised from room temperature to 1173 K

using a linear programmer at a heating rate of 40 K min�1. The

X-ray absorption near edge structure (XANES) spectra were

taken in a transmission mode for the K-edge of Mn at beamline

3C1 of the pohang light source (PAL) operating at 2.5 GeV with

ca. 100–150 mA of stored current. The detector gas was N2 for

the incident beam and the transmitted beam. In addition to

catalyst samples, XAFS data were also obtained for Mn foil,

MnCO3, MnO, Mn2O3, Mn3O4, and MnO2 as references. They

were analyzed by using ATHENA [20].

2.3. Activity measurements

Catalytic activities were measured over a fixed bed of

catalysts in a tubular flow reactor of 8 mm i.d. Reactant gases

were fed to the reactor by means of electronic mass flow

controller (MKS).

The reactant gas typically consisted of 500 ppm NO,

500 ppm NH3, 5 vol.% O2, and N2 or He. The NOx

concentration in the inlet and outlet gas was analyzed by

means of a NO/NO2 combustion gas analyzer (Euroton). The

N2 and N2O in the effluent were separated at 353 K with

HeySep D column and their concentrations were analyzed with

a thermal conductive detector (TCD) in a gas chromatography

(Hewlett-Packard, HP 5890). From the concentration of the

gases at steady state both the conversion and the selectivity are

calculated according to the following formula [12,17]:

NO conversionð%Þ ¼ ½NO�in � ½NO�out

½NO� in� 100 (1)

Table 1

N2 adsorption data of the MnOx catalysts precipitated with a different pre-

cipitant such as (NH4)2CO3 (AC), K2CO3 (PC), Na2CO3 (SC), NH4OH (AH),

KOH (PH), and NaOH (SH)

Catalystsa SBET (m2/g) VP (cm3/g)

MnOx-AC (623) 132.7 0.32

MnOx-PC (623) 154.3 0.38

MnOx-SC (623) 173.3 0.35

MnOx-AH (623) 25.9 0.26

MnOx-PH (623) 29.1 0.29

MnOx-SH (623) 31.0 0.27

a Catalysts were calcined in air at 623 K before the measurement.

is A: General 327 (2007) 261–269 263

N2 selectivityð%Þ ¼ ½N2�out

½N2�out þ ½N2O�out

� 100 (2)

The subscripts in and out indicated the inlet concentration and

outlet concentration at steady state, respectively.

3. Results and discussion

3.1. The effect of kinds of precipitant

Fig. 1 compared the catalytic activities of the MnOx

catalysts precipitated with different kinds of precipitant in

terms of NOx conversion and N2 selectivity as a function of

reaction temperature. The NOx conversion was affected

noticeably by kinds of its precipitant especially at low

temperatures. The catalysts precipitated with precipitants

containing carbonate as a anion showed the higher NOx

conversion than those prepared with alkali or ammonium

hydroxides. The former also exhibited the better N2 selectivity

M. Kang et al. / Applied Catalys

Fig. 1. NOx conversions and N2 selectivities at different reaction temperatures

over various MnOx catalysts prepared by different precipitants. All catalysts

were calcined in air at 623 K. Reactants: 500 ppm NO, 500 ppm NH3 and

5 vol.% O2 in N2. The gas hourly space velocity (GHSV) was 50,000 h�1.

than the latter. For all catalysts, the N2 selectivity decreased

with increasing reaction temperature. The NOx conversion and

the N2 selectivity were also affected by kinds of cations such as

sodium ion, potassium ion, and ammonium ion and decreased

in this order. Therefore, the highest NOx conversion and N2

selectivity was observed over MnOx catalysts precipitated with

sodium carbonate.

Table 1 gives the BET surface area and pore volume of the

various MnOx catalysts prepared with different precipitants.

The surface areas and pore volumes of MnOx catalysts

precipitated with alkali or ammonium hydroxides are much

smaller than those of the MnOx catalysts prepared with

precipitants containing carbonate. The kinds of cations in the

precipitant also appeared to affect the surface area. The BET

surface area decreased in the order when Na+, K+, and NH4+

was utilized as a cation in the precipitant, respectively.

Therefore, the manganese oxide catalyst precipitated with

sodium carbonate had the highest surface area. It is worth

noting that the SCR activity closely correlates with the surface

area. Based on the previous work [19], the additional factors

besides the surface area must affect the SCR activity because

MnOx-SC showed the higher specific activity per surface area

than MnOx-AH.

Fig. 2 shows the XRD patterns of the various MnOx catalysts

just dried at 373 K and calcined in air at 623 K. All the as-

prepared MnOx-AH, MnOx-PH and MnOx-SH sample give

sharp XRD peaks representing Mn3O4 phase. This phase was

transformed into Mn5O8 phases after heat treatment at 623 K.

For MnOx-AC, MnOx-PC and MnOx-SC catalysts, MnCO3

phase was observed when they were dried at 373 K. However,

when they were calcined in air at 623 K, no noticeable

crystalline phase was observed. This can be interpreted that as-

prepared MnCO3 was decomposed into the amorphous

manganese oxide phase during calcinations. This partially

decomposed and amorphous structure of the MnOx catalysts

may also be a reason for the excellent catalytic activity at low

temperatures, in addition to the high surface area.

To find out the surface chemical states of MnOx catalysts

prepared with different precipitants, XPS spectra of Mn 2p, O

1s, and C 1s were obtained as shown in Fig. 3. For all samples,

two main peaks due to Mn 2p3/2 and Mn 2p1/2 were observed

from 639 eV to 660 eV. It was reported that 2p3/2 binding

energy of Mn(0) and Mn(IV)O2 was 639.0 eV and 642.1 eV,

respectively [21,22]. Therefore, it is not easy to discern the

Fig. 2. XRD patterns of MnOx catalysts just dried (A) and calcined in air at 623 K (B) after precipitated with different precipitants: (a) MnOx-AC, (b) MnOx-PC, (c)

MnOx-SC, (d) MnOx-AH, (e) MnOx-PH and (f) MnOx-SH (* MnCO3, ^ Mn3O4 and & Mn5O8).

M. Kang et al. / Applied Catalysis A: General 327 (2007) 261–269264

oxidation states of manganese species without peak deconvolu-

tion because there is only 3.1 eV energy difference between

Mn(0) and Mn(IV). The peak position of Mn 2p3/2 for MnOx

catalysts can exclude the possibility that there exist manganese

species whose oxidation state is 0 or +2. Therefore, the

deconvolution of Mn 2p3/2 was conducted based on the

assumption that there are only Mn(III) and Mn(IV) species.

Fig. 3. XPS spectra for (A) Mn 2p, (B) O 1s, and (C) C 1s of the MnOx catalysts: (a) M

(e) MnOx-PH (623) and (f) MnOx-SH (623).

The presence of satellite peak was also considered in the peak

deconvolution. This satellite structure, noticeable on the higher

binding energy side of the Mn 2p core-line in the XPS spectrum

of the manganese compounds, originates from the charge

transfer between outer electron shell of ligand and an unfilled

3d shell of Mn during creation of core-hole in the photoelectron

process [23]. As shown in Fig. 3, the asymmetric peak was

nOx-AC (623), (b) MnOx-PC (623), (c) MnOx-SC (623), (d) MnOx-AH (623),

Table 2

Mn 2p and O 1s binding energies on MnOx catalysts precipitated with a different precipitant such as (NH4)2CO3 (AC), K2CO3 (PC), Na2CO3 (SC), NH4OH (AH),

KOH (PH), and NaOH (SH)

Catalystsa Mn 2p (eV) Mn4+/Mn3+ O 1s (eV) Oa/(Oa + Ob)

Mn4+ Mn3+ Oa Ob

MnOx-AC (623) 643.7 641.9 0.97 531.3 529.6 42.6

MnOx-PC (623) 643.7 641.9 1.00 531.3 529.6 41.3

MnOx-SC (623) 643.7 641.9 1.10 531.4 529.7 52.2

MnOx-AH (623) 643.8 641.9 0.71 531.7 530.1 30.4

MnOx-PH (623) 643.8 641.9 0.63 531.5 530.0 31.0

MnOx-SH (623) 643.7 642.0 0.69 531.5 529.9 28.1

a Catalysts were calcined in air at 623 K before the measurement.

Fig. 4. NOx conversions and N2 selectivities at different reaction temperatures

over various MnOx-SC catalysts calcined at different temperatures. Reactants:

500 ppm NO, 500 ppm NH3 and 5 vol% O2 in N2. The gas hourly space velocity

(GHSV) was 50,000 h�1.

M. Kang et al. / Applied Catalysis A: General 327 (2007) 261–269 265

observed in XPS spectra of O 1s for all samples. The peak at

529.6–530.0 eV corresponds to lattice oxygen (O2�) (hereafter

denote as Ob) whereas the one at 531.3–531.7 eV corresponds

to several O 1s states assigned to the surface-adsorbed oxygen

such as O22� or O�, in the form of hydroxyl, OH�, and

carbonate, CO32� (hereafter denoted as Oa) [24,25]. XPS

spectra of C 1s were also obtained as shown in Fig. 3. For all

samples, two adjacent peaks were observed from 282 eV to

292 eV. The peak at 285 eV, 286.5 eV, and 289 eV corresponds

to contaminated carbon, CO bond, and surface carbonate

species, respectively. The peak intensity at a higher binding

energy in C 1s XPS spectra for MnOx prepared with

precipitants containing carbonate anions was stronger than

those for others.

Table 2 lists the surface compositions from the XPS spectra

of the MnOx catalysts prepared with different precipitants. The

value of the Mn4+/Mn3+ ratio in the surface layer of MnOx

catalyst was quite high (0.97–1.10) in catalysts precipitated

with carbonate-containing precipitants. Especially, this value

appeared to be the largest for MnOx-SC which showed the

highest catalytic activity for the selective NOx reduction with

NH3 at low temperatures. The relative concentration ratio of

Oa/(Oa+Ob) was also higher in MnOx-SC, MnOx-PC, and

MnOx-AC than that in MnOx-AH, MnOx-PH, and MnOx-SH.

The manganese oxide prepared with sodium carbonate which

was the best in NH3-SCR at low temperatures showed the

highest relative concentration ratio of Oa/(Oa+Ob) among all

catalysts. Therefore, the effect of calcinations temperature on

the catalytic activity and structure and electronic state of

catalysts were examined further for MnOx catalysts precipi-

tated with sodium carbonate.

3.2. The effect of calcination temperatures

Fig. 4 compares the NOx conversions and N2 selectivity over

MnOx-SC catalysts calcined at different temperatures. The

MnOx-SC catalyst dried at 373 K showed less than 20% NOx

conversion when the reaction temperature was below 400 K.

This catalyst exhibited more than 80% NOx conversion only in

the temperature range from 448 K to 498 K. However, 100%

NOx conversion was obtained from 348 K to 448 K when

MnOx-SC catalyst was calcined at 523 or 623 K. Further

increase in calcinations temperature above 623 K caused

decreasing NOx conversion especially at low reaction

temperatures. Therefore, the MnOx-SC catalysts calcined at

moderate temperature such as 523 or 623 K have the highest

catalytic activity for NOx reduction with NH3. The N2

selectivity gradually decreased with increasing calcination

temperatures. Table 3 shows the BET surface areas, pore

volumes, and average pore diameters of the MnOx-SC catalysts

calcined at different temperatures. The BET surface areas and

Fig. 5. XRD patterns of MnOx-SC catalysts calcined at different temperatures.

Table 3

Surface area, pore volume and average pore diameter of the MnOx catalysts

precipitated with sodium carbonate (SC) and calcined in air at different

temperatures

Catalysts SBET (m2/g) VP (cm3/g) DP (nm)

MnOx-SC (523) 178.5 0.35 8.0

MnOx-SC (623) 173.3 0.35 10.9

MnOx-SC (723) 80.4 0.31 15.5

MnOx-SC (823) 22.2 0.07 21.3

M. Kang et al. / Applied Catalysis A: General 327 (2007) 261–269266

the pore volumes of the MnOx-SC catalysts monotonically

decreased with increasing the calcination temperature from

523 K to 823 K. The corresponding average pore diameter

gradually increased from 8.0 nm to 21.3 nm as expected. This

can be due to the sintering to some extent at high calcinations

temperature.

Fig. 5 illustrates the XRD patterns of MnOx catalysts

precipitated with sodium carbonate and calcined at different

temperatures. For the MnOx-SC dried at 373 K, the strong

peaks representing crystalline MnCO3 phase were observed.

However, these strong peaks disappeared completely and very

weak peaks which cannot be assigned to any manganese oxides

were found when MnOx-SC catalysts were calcined at

temperatures from 523 K to 723 K. This can be interpreted

that MnCO3 was transformed into an amorphous phase during

calcinations at this temperature. This partially decomposed and

amorphous structure of the MnOx-SC catalyst may also be a

reason for the high NOx conversion and N2 selectivity at low

temperatures. This was also consistent with the previous

observation that nitrous oxide formation occurred preferen-

tially on crystalline phases [12]. When the MnOx-SC catalyst

was calcined at 823 K, the formation of Mn2O3 crystalline

phase was observed in XRD pattern.

XPS measurements were performed to examine the effect of

calcinations temperatures on the surface electronic state of

MnOx-SC catalysts. Fig. 6 shows the XPS spectra of Mn 2p, O

1s, and C 1s of MnOx-SC catalysts with different calcination

temperatures. For all samples, two main peaks due to Mn 2p1/2

and Mn 2p3/2 were observed from 639 eV to 660 eV, which can

exclude the possibility of the presence of Mn(0) and Mn(II)

species. Therefore, the deconvolution of Mn 2p3/2 was

conducted based on the assumption that there are only Mn(III)

and Mn(IV) species. The presence of satellite peak was also

considered in the peak deconvolution. As shown in Fig. 6, the

asymmetric peak was observed in XPS spectra of O 1s for all

samples. As stated previously, the peak at 530.0 eV and

531.5 eV can be assigned to the lattice oxygen (Ob) and the

Table 4

Mn 2p and O 1s binding energies on MnOx-SC catalysts precipitated with sodium

Catalysts Mn 2p (eV) Mn4+/M

Mn4+ Mn3+

MnOx-SC (523) 643.6 641.9 1.03

MnOx-SC (623) 643.7 641.9 1.10

MnOx-SC (723) 643.5 641.9 0.92

MnOx-SC (823) 643.5 641.8 0.78

surface-adsorbed oxygen (Oa). XPS spectra of C 1s were also

obtained as shown in Fig. 6. For all samples, two adjacent peaks

were observed from 282 eV to 292 eV. The peak at 285 eV,

286.5 eV, and 289 eV corresponds to contaminated carbon, CO

bond, and surface carbonate species, respectively. For MnOx-

SC catalysts, the peak intensity at a higher binding energy in C

1s XPS spectra decreased with increasing calcinations

temperature.

Table 4 lists the surface compositions from the XPS spectra

of the MnOx catalysts calcined at different temperatures. The

value of the Mn4+/Mn3+ ratio in the surface layer of MnOx

catalyst was quite high in catalysts calcined at 523 or 623 K

which showed the high catalytic activity for the selective NOx

reduction with NH3 at low temperatures. The relative

concentration ratio of Oa/(Oa+Ob) decreased with increasing

calcinations temperature for manganese oxides catalysts

precipitated with sodium carbonate. The most active manga-

nese oxide prepared with sodium carbonate and calcined at

523 K showed the highest concentration of surface oxygen.

Therefore, the active catalyst has the high Mn4+/Mn3+ ratio and

surface oxygen concentration in the surface layer of MnOx

catalyst.

carbonate (SC) and calcined in air at different temperatures

n3+ O 1s (eV) Oa/(Oa + Ob)

Oa Ob

531.6 529.8 68.7

531.4 529.7 52.2

531.5 529.7 42.5

531.5 529.9 40.2

Fig. 6. XPS spectra for (A) Mn 2p, (B) O 1s, and (C) C 1s of the MnOx-SC catalysts calcined in air at different temperatures.

Fig. 7. Thermal gravimetric analysis of manganese oxides precipitated with

sodium carbonate and dried at 373 K.

M. Kang et al. / Applied Catalysis A: General 327 (2007) 261–269 267

The value of the Mn4+/Mn3+ ratio was previously considered

as a significant parameter characterizing the intrinsic properties

of other metal oxide catalysts. Machocki et al. [22] have

investigated the catalytic combustion of methane over silver

modified manganese–lanthanum oxides and found that the rate

of methane oxidation is a linear function of the Mn4+/Mn3+

surface ratio. Tang et al. [26] also reached the same conclusion

that MnOx–CeO2 catalysts showed higher catalytic activity for

HCHO oxidation with increasing the Mn4+/Mn3+ surface ratio.

Similar result was observed in the present study.

In the general NH3-SCR process, NO is easily converted to

N2 and H2O by NH3:

4NO þ 4NH3þO2 ! 4N2þ 6H2O (3)

When both NO and NO2 are present, the following reaction

proceeds much more rapidly than the above reaction (3):

4NH3þ 2NO þ 2NO2 ! 4N2þ 6H2O (4)

Wallin et al. [27] reported that the presence of NO2 in the gas

mixture enhanced the SCR performance of a catalyst.

Therefore, it can be inferred that the difference in the maximum

NOx conversion between the catalysts is probably due to the

difference in the amount of NO2 produced from the oxidation of

NO by high Mn4+/Mn3+ surface ratio.

The value of Oa/(Oa + Ob) in the surface layer of MnOx

catalyst prepared with precipitants containing carbonate anions

was higher than that of others. The high relative concentration

ratio of Oa/(Oa+Ob) was previously found to be preferable for

SCR reactions over the manganese-containing catalysts.

Kapteijn et al. [17] indicated that the product distribution is

determined by the concentration of reactive oxygen and nitric

oxide, since both affect the relative distribution of the surface

species. A high surface oxygen concentration facilitates the (–

NH2) formation. Kijlstra et al. [10] studied the mechanism of

SCR of NO with ammonia at low temperatures on MnOx/Al2O3

catalyst. They proposed that the reaction starts with the

adsorption of NH3 on a Lewis acid and subsequently transforms

into NH2; the NH2 would then react with gas-phase NO (an ER

mechanism) and nitrite intermediates on the surface (a LH

mechanism). This is supported by our results, which indicated

that the N2 selectivity increased with the relative ratio between

concentration of surface oxygen and that of lattice oxygen.

The thermal gravimetric analysis was conducted to observe

the change in the weight of manganese oxides precipitated with

sodium carbonate with increasing temperature as shown in

Fig. 7. The gradual weight loss occurred from room

temperature to 850 K. This decreased weight (�35 wt.%) is

almost same to the calculated value assuming that MnCO3 was

transformed into Mn2O3. Therefore, MnOx-SC catalyst

calcined below 850 K must contain lots of carbon oxide

species (COx). There is no significant weigh loss in the MnOx-

AH catalyst. It is reasonable that the presence of COx species in

the MnOx-SC catalyst also affects the de-NOx activity at low

temperatures because the residual COx species may act as

acidic sites on the catalyst surface. These acidic sites can help

Fig. 8. Mn K-edge XANES spectra of Mn reference samples and MnOx-SC catalysts calcined at different temperatures. The fitted spectra were plotted in dotted line.

Table 5

The linear combination fitting result for MnOx catalysts precipitated with

sodium carbonate (SC) and calcined in air at different temperatures

Catalysts E0/(eV) Standard

sample

Weight

fraction

R-factora

MnOx-SC (373) 6548.733 MnCO3 0.966 0.15801

Mn2O3 0.034

MnOx-SC (523) 6548.719 MnCO3 0.424 0.02717

Mn3O4 0.279

Mn2O3 0.286

MnO2 0.011

MnOx-SC (623) 6548.719 MnCO3 0.135 0.52299

Mn3O4 0.280

Mn2O3 0.510

MnO2 0.075

MnOx-SC (723) 6548.719 MnCO3 0.128 0.12677

Mn3O4 0.314

Mn2O3 0.487

MnO2 0.071

MnOx-SC (823) 6553.719 MnCO3 0.012 0.15782

Mn3O4 0.106

Mn2O3 0.882

a R-factor is defined as follows. R = sum((data-fit)^2)/sum(data^2)).

M. Kang et al. / Applied Catalysis A: General 327 (2007) 261–269268

the basic reductant, NH3, adsorb on the surface at low

temperatures, and therefore the reductant may be enriched

compared with the surface of MnOx-AH catalyst. This was also

supported by the fact that 0.207 mmol NH3/gcat and

0.011 mmol NH3/gcat could be adsorbed on MnOx-SC and

MnOx-AH, respectively. This difference is much more than

expected even if MnOx-SC has seven times larger surface area

than that of MnOx-AH. The amount of chemisorbed NO on

MnOx-SC and MnOx-AH at room temperature was 0.50 mmol

NO/gcat and 0.12 mmol NO/gcat, respectively. This can be

mainly due to difference in surface area between two catalysts.

To determine the structural and electronic information of

MnOx catalysts which were determined to be amorphous from

XRD data, Mn K-edge XANES spectra were obtained. Fig. 8

presents Mn K-edge XANES spectra of MnOx-SC catalysts

calcined at different temperatures with Mn reference com-

pounds such as Mn foil, MnCO3, MnO, Mn3O4, Mn2O3, and

MnO2. Easily discernible features of XANES spectra were

observed for Mn reference samples and the edge energy shifted

toward a higher energy with an increase in the oxidation state of

Mn from metallic Mn foil to Mn(IV)O2. Mn K-edge XANES

spectra of as-prepared MnOx-SC catalyst was similar to that of

MnCO3, which is consistent with the fact that crystalline

MnCO3 phase was observed from XRD. MnOx-SC catalysts

calcined at 523 K, 623 K and 723 K appeared to be different

compared with any XANES spectra of Mn reference samples

measured. However, Mn K-edge XANES spectra of MnOx-SC

calcined at 823 K was similar to that of Mn2O3, which is

consistent with the fact that crystalline Mn2O3 phase was

observed from XRD.

To determine the quantitative amount of each manganese

oxides, a linear XANES fitting was conducted for MnOx-SC

catalysts calcined at different temperatures as shown in Table 5.

Different weight fractions of manganese oxides appeared to be

present in these catalysts. The manganese carbonate was

determined to be mainly present with a small amount of

Mn2O3 in MnOx-SC just dried at 373 K. When MnOx-SC was

calcined at 523 K, the weight fraction of MnCO3 decreased

noticeably and those of Mn3O4 and Mn2O3 increased. When this

catalyst was calcined at 623 or 723 K, most of MnCO3 must be

transformed into Mn2O3. Because Mn2O3 was determined to be

mainly present with a small amount of Mn3O4 in MnOx-SC

calcined at 823 K, some of Mn3O4 appeared to be further

transformed into Mn2O3. Although no noticeable difference in

the composition of manganese oxides for MnOx-SC calcined at

623 K compared with that of MnOx-SC calcined at 723 K from

XANES fitting results, the former was superior to the latter in the

catalytic activity at low temperatures. This indicates that the kind

of manganese oxide as well as its crystallinity should be related to

the catalytic activity in NH3-SCR at low temperatures.

M. Kang et al. / Applied Catalysis A: General 327 (2007) 261–269 269

The formation of ammonium nitrate on the catalyst surface

during the course of the reaction has been claimed to be a main

problem for low temperature NH3-SCR performance [28]. The

temperature programmed desorption (TPD) was conducted

from 373 K to 1073 K over the catalyst after a reaction and its

effluent gas was analyzed with a mass spectrometer (PFEIF-

FER Vacuum Quadstar). During TPD experiment, no NH3

species was detected. This can exclude the formation of

ammonium salts during a reaction.

4. Conclusions

The manganese oxides catalyst prepared by a precipitation

method using sodium carbonate and calcined at moderate

temperatures such as 523 K and 623 K showed the high NOx

conversion and N2 selectivity for low temperature selective

catalytic reduction of NOx with NH3. This active MnOx

catalyst has the high surface area, the abundant Mn4+ species,

and the rich concentration of surface oxygen on the surface.

The amorphous Mn3O4 and Mn2O3 were determined to be

mainly present. The residual carbonate species on this catalyst

also help NH3 adsorb on the surface, which resulted in the high

catalytic activity at low temperatures.

Acknowledgements

The financial support by the Research Initiation Program at

Ajou University (20041340) was appreciated by one of authors,

Eun Duck Park. Experiments at PLS were supported in part by

MOST and POSTECH.

References

[1] H. Bosch, F.J.I.G. Janssen, Catal. Today 2 (1988) 369–379.

[2] J.N. Armor, Appl. Catal. B 1 (1992) 221–256.

[3] H.S. Rosenberg, L.M. Curran, A.U. Slack, J. Ando, J.H. Oxley, EPRIRP-

925, October.1978.

[4] H. Hamada, Catal. Today 22 (1994) 21–40.

[5] J.M. Garcia-Cortes, Appl. Catal. B: Environ. 30 (2001) 399–408.

[6] G.A. Papapolymerou, L.D. Schmidt, Langmuir 1 (1985) 488–495.

[7] Y. Murakami, M. Inomata, K. Mori, K. Suzuki, A. Miyamoto, T. Hattori,

in: G. Poncelet, P. Grange, P.A. Jacobs (Eds.), Preparation of Catalysis III,

Elsevier, Amsterdam, 1983, pp. 531–534.

[8] J.R. Kiovsky, P.B. Koradia, C.T. Lim, Ind. Eng. Chem. Prod. Res. Dev. 19

(1980) 218–225.

[9] G. Busca, L. Lietti, G. Ramis, F. Berti, Appl. Catal. B: Environ. 18 (1998)

1–36.

[10] W.S. Kijlstra, J.C.M.L. Daamen, J.M. van de Graff, B. van der Linden,

E.K. Poels, A. Bliek, Appl. Catal. B: Environ. 7 (1996) 337–357.

[11] J. Muniz, G. Marban, A.B. Fuertes, Appl. Catal. B: Environ. 27 (2000) 27–

36.

[12] L. Singoredjo, R. Korver, F. Kapteijn, J. Moulijn, Appl. Catal. B: Environ.

1 (1992) 297–316.

[13] U. Bentrup, A. Bruckner, M. Richter, R. Fricke, Appl. Catal. B: Environ.

32 (2001) 229–241.

[14] G. Qi, R.T. Yang, R. Chang, Catal. Lett. 87 (2003) 67–71.

[15] P.G. Smirniotis, D.A. P na, B.S. Uphade, Angew. Chem. Int. Ed. 40 (2001)

2479–2482.

[16] G. Qi, R.T. Yang, Appl. Catal. B: Environ. 44 (2003) 217–225.

[17] F. Kapteijn, L. Singoredjo, A. Andreini, J.A. Moulijn, Appl. Catal. B:

Environ. 3 (1994) 173–189.

[18] G. Qi, R.T. Yang, Chem. Commun. (2003) 848–849.

[19] M. Kang, T.H. Yeon, E.D. Park, J.E. Yie, J.M. Kim, Catal. Lett. 106 (2006)

77–80.

[20] B. Ravel, M. Newville, J. Synchrotron Radiat. 12 (2005) 537–541.

[21] S. Ponce, M.A. Pena, J.L.G. Fierro, Appl. Catal. B 24 (2000) 193–

205.

[22] A. Machocki, T. Ioannides, B. Stasinska, W. Gac, G. Avgouropoulos, D.

Delimaris, W. Grzegorczyk, S. Pasieczna, J. Catal. 227 (2004) 282–

296.

[23] R.J. Iwanowski, M.H. Heinonen, W. Paszkowicz, R. Minikaev, T. Story, B.

Witkowska, Appl. Surf. Sci. 252 (2006) 3632–3641.

[24] J.P. Holgado, G. Munuera, J.P. Espinos, A.R. Gonzalez-Elipe, Appl. Surf.

Sci. 158 (2000) 164–171.

[25] R. Larachi, J. Pierre, A. Adnot, A. Bernis, Appl. Surf. Sci. 195 (2002) 236–

250.

[26] X. Tang, Y. Li, X. Huang, Y. Xu, H. Zhu, J. Wang, W. Shen, Appl. Catal. B:

Environ. 62 (2006) 265–273.

[27] M. Wallin, C. Karlsson, M. Skoglundh, A. Palmqvist, J. Catal. 218 (2003)

354–364.

[28] M. Koebel, M. Elsener, G. Madia, Ind. Eng. Chem. Res. 40 (2001) 52–59.