Preparation of vanadium-based catalysts for selective catalytic reduction of nitrogen oxides using titania supports chemically modified with organosilanes

Preparation of vanadium-based catalysts for selective catalytic reduction of nitrogen oxides using titania supports chemically modified with organosilanes

Studiesin SurfaceScienceand Catalysis 143 E. Gaigneauxet al. (Editors) 9 2002 ElsevierScienceB.V. All rightsreserved. 1089 Preparation of vanadium-b...

450KB Sizes 2 Downloads 38 Views

Studiesin SurfaceScienceand Catalysis 143 E. Gaigneauxet al. (Editors) 9 2002 ElsevierScienceB.V. All rightsreserved.

1089

Preparation of vanadium-based catalysts for selective catalytic reduction of nitrogen oxides using titania supports chemically modified with organosilanes H. Kominami, a M. Itonaga, a A. Shinonaga, a K. Kagawa, b S. Konishi, a and Y. Kera a aDepartment of Applied Chemistry, Faculty of Science and Engineering, Kinki University, Kowakae, Higashiosaka, Osaka 577-8502, Japan bTechnical Research Center, The Kansai Electric Power Co. Inc., Nakoji, Amagasaki, Hyogo 661-0974, Japan A titania (TiO2) sample with a large surface area (300 mZg-1) was chemically modified with 3-aminopropyltrimethoxysilane (APS) under a reflux of toluene. The thermal stability of the modified TiOz (APS-TiO2) and the adsorptivity of metavanadate anion (VO3-) on APS-TiO2 from an aqueous solution were investigated. Modification with APS suppressed crystal growth and transformation of anatase crystallite to rutile upon calcination, and the anatase phase was preserved even after calcination at 1000~ while transformation to rutile in the unmodified TiOz samples was observed at around 800~ Since there was little crystal growth in APS-TiO2, it possessed a large surface area of 205 mZg-1 after calcination at 700~ The amount of VO3- adsorbed on APS-TiO2 was ca. 1.5-times larger than that on unmodified TiO2 due to a strong affinity caused by acid-base interaction between VO3-and the amino group of APS-TiO2. A supported V205 catalyst that was prepared by decomposition of a VO3--adsorbing APS-TiOz sample had a large surface area despite the large V205 loading and exhibited higher activity than that of a catalyst prepared from an unmodified TiO2 sample. 1. I N T R O D U C T I O N Vanadia/titania (VzOs/TiO2) catalysts are widely used as commercial environmental catalyst for the selective catalytic reduction (SCR) of nitrogen oxides (NOx) with ammonia [1-3]. Due to current strict regulations, a high level of NOx removal is now required, and one of the most effective approaches for the removal of a large amount of NOx is to use a TiO2 support that has a large surface area. Generally, a large surface area is required for a catalyst support to disperse a catalyst material effectively and increase the number of active sites in the catalyst. Since a catalyst is usually used at high temperature, a high degree of thermal stability is also important. We have shown that chemical modification of metal oxide supports, such as silica (SiO2) and TiO2, with silane-coupling agents (organosilanes) having amino groups was effective for dispersion of a catalytic material such as phosphododecatungstate and

1090 phosphotetradecavanadate [4-8] and that the thus-obtained catalysts exhibited higher activities than those of catalysts without chemical modification in some reactions such as partial oxidation of methanol to formaldehyde and oxidative dehydrogenation of isobutyric acid to methacrylic acid [5, 7]. We applied this technique to the preparation of VzOs/TiO2 catalysts and here report the effect of organosilane on the thermal stability of a TiO2 support with a large surface area, the adsorption properties of metavanadate anion (VO3-), and the SCR activities of supported V205 catalysts prepared from VO3--adsorbing TiO2 supports. 2. EXPERIMENTAL

2.1. Modification of TiO2 with organosilane TiO2 powder with a large surface area (ST01) was kindly donated from Ishihara Sangyo, Osaka, Japan. Organosilane, 3-aminopropyltrimethoxysilane (H2N-(CH2)3-Si(OCH3)3; APS), was supplied from Shin-Etsu Chemical Co., Ltd., Tokyo, Japan and was used without further purification. Chemical modification of the ST01 TiO2 with APS was carried out according to the procedure previously reported [4-8]. The TiO2 powder was dried in advance at room temperature under reduced pressure for 24 h. Two grams of the TiO2 powder (2 g) was then suspended in a toluene solution (20 cm 3) of APS (22.1 mmol) and heated at ll0~ for 2 h. After completion of the reaction, the powder was filtered, washed with toluene, diethylether and finally methanol, and then allowed to stand in a desiccator for 30 min. Finally, the sample was dried at l l0~ for 1 h. Hereafter, the TiO2 powder modified with APS is designated APS-TiO2. 2.2. Adsorption of VO3" on modified and unmodified TiO2 Ammonium metavanadate (NH4VO3) (Kanto Chemicals, Tokyo, Japan) was dissolved in H20 to give solutions (0.04-2.00 • 10.2 mol din-3). A hundred milligrams of unmodified TiO2 or APS-TiO2 was added to 30 cm 3 of NH4VO3 solution and stirred to make VO3- adsorb at room temperature for 24 h. The samples were then collected by filtration under suction and dried at l l0~ for 1 h. The amount of VO3- adsorbed was determined using a UV-vis spectrometer (UV-2400, Shimadzu, Kyoto, Japan) from the difference in the concentrations of the solution before and after addition of TiO2. 2.3. SCR of NOx with NH3 over a V205[TiO2 catalyst Unmodified TiO2 and APS-TiO2 samples adsorbing VO3- were decomposed at 550~ for 1 h in air to form supported V205 catalysts. Two V205 catalysts were used for deNOx reaction. The catalytic reaction (catalyst weight: 80 mg) was carried out in a fixed bed flow reactor at temperatures ranging from 300 to 400~ (flow rate: 1.2 dm3 min -1) using a model gas containing 200 ppm NO, 240 ppm NH3, 500 ppm SO2, 33 ppm CO, 3% 02 and 12% H20. The concentration of NOx in the outlet gas was determined by using a NOx meter. 2.4. Characterization Powder X-ray diffraction (XRD) (MultiFlex, Rigaku, Tokyo, Japan) was measured using CuKct radiation with a carbon monochromater. Crystallite size (d101) was calculated

1091 from the half-height width of the 101 diffraction peak of anatase using Scherrer's equation. The value of the shape factor, K, was taken to be 0.9. The specific surface area was calculated using the BET single-point method on the basis of nitrogen (Na) uptake measured at 78 K at the relative pressure of 0.3. Before the N2 adsorption, each sample was dried at 403 K for 30 min in a 30% Na-helium flow. Differential thermal analysis (DTA) (TG-8120, Rigaku) was carried out at a rate of 10~ min -1 in air flow. The morphology of the sample was observed under a JEOL JEM-3010 transmission electron microscope (TEM) operated at 300 kV. Chemical analysis of the APS-TiO2 was performed at Galbraith Laboratory Inc., TN USA.

(a)

IU

(b)

1000~C

b

10001:,C

9001~C

9001:,C ..~.

-..

8001=C

800~C

7001~C .....

-

~

.

_. - , , -

7001~C

5501~C

5501:,C 'before calcination ,

20

I

30

I

40 20 /

,

I

,

50

I

60

degree

I

,

70

20

I

30

I

I

t

I

I

I

40 50 60 20 / degree

t

70

Fig. 1. XRD patterns of samples obtained by calcination of (a) unmodified TiO2 and (b) APS-Ti02 at various temperatures. 3. RESULTS AND DISCUSSION 3.1. Effect of modification with APS on thermal stability of TiOz In the DTA curve of APS-TiO2, a large exothermic peak was observed at around 300~ due to combustion of the aminopropyl group of APS-TiO2. Chemical analysis

1092 revealed that the degree of modification in the APS-TiO2 sample, D(APS), was 2.21 mmol-APS.g-TiO2 -1 and that the Si/Ti ratio was 0.177. Since the initial amount of APS in the toluene solution was 22.1 mmol for 2 g of TiO2, the degree of deposition of APS onto TiO2 was calculated to be 20%. The specific surface area of unmodified TiO2 (bare ST01), S(TiO2), was determined by the BET method to be 300 m2g-1, which is in good agreement with the value reported by the supplier. From D(APS), S(TiO2) and Avogadro's number, the number of APS per nm 2 of TiO2 was calculated to be 4.4, which is almost equal to the numbers (4.5-4.9) of surface hydroxyl groups per nm 2 of anatase TiO2 [9]. The surface coverage of APS on TiO2, 0 (APS), can be estimated by the following formula: 0 (APS)

= D(APS)

9 or (APS) / S(TiO2),

where cr(APS) denotes the cross-sectional area (m 2 mo1-1) of APS. By applying the values for D and S given above and cr = 7.8 • 104 m 2 mo1-1 by estimation from the value given by the supplier (minimum covered area, 436 m 2 g-l), 0(APS) of the present APS-TiO2 sample was estimated to be 0.58.

60

300 "~

(a)

E

~,~ 2OO

4O

.

(b) 9

9

I-U.l rn

m

20

0

I

~

I

,

i

~

!

,

I

200 400 600 800 1000 Tcal /13C

100

0

~

w

200 400 600 800 1000 TcaI/13C

Fig. 2. Changes in (a) crystallite size and (b) BET surface area of TiO2 samples upon calcination at various temperatures. Circles and squares show data of unmodified TiO2 and APS-TiO2, respectively.

The XRD patterns of samples obtained by calcination of unmodified TiO2 and APS-TiO2 at various temperatures are shown in Fig. 1. Transformation of the anatase phase in the unmodified TiO2 sample to the rutile form occurred at 800~ and was almost completed at 900~ (Fig. 1(a)), whereas the anatase crystallite in the APS-TiO2 sample was stable even after calcination at 1000~ (Fig. l(b)). Changes in dl01 and BET surface area of the unmodified TiO2 and APS-TiO2 samples upon calcination at various temperatures are shown in Fig. 2. Unmodified TiO2 exhibited a small dl01 of 8 nm before calcination and, as

1093 mentioned above, possessed a large surface area of 300 mZg-1. However, calcination of the TiOz sample at 550~ drastically increased d101 to 17 nm and decreased the surface area to 91 mZg-1, indicating low thermal stability of the original TiO2 powder. Modification of the TiOz powder with APS inhibited the growth of the anatase crystallite as well as its transformation to the rutile crystallite, as shown in Fig. 2. For example, the sample obtained by calcination at 700~ showed the almost same dl01 as that before calcination and retained a large surface area (>200 mZg-1) due to the inhibition of crystal growth. As can be seen in Fig. 3, TEM observation clearly showed a difference between the behaviors of crystal growth of unmodified and modified TiO2 upon calcination. It had been reported that silica-modified TiO2 samples synthesized from a mixture of titanium alkoxide and silicate ester by glycothermal reaction and thermal decomposition in toluene showed a high degree of thermal stability [10, 11]. Most of the silicon (silica) was incorporated into the anatase structure of both products and suppressed crystal growth and transformation of the TiOz phase. On the other hand, in the present APS-TiOz, silicon originating from APS was distributed only on the surfaces of TiOz crystals (particles). It should be noted that surface modification of TiO2 with APS results in significant improvement of the thermal stability of the original TiOz if the original TiOz has a large surface area.

Fig. 3. TEM photographs of samples after calcination of (a) unmodified TiO2 and (b) APS-TiO2 at 550~ 3.2. Effect of modification with APS on adsorption properties of TiOz Fig. 4(a) shows the amounts of VO3- adsorbed on unmodified TiO2 and APS-TiO2 (Ca~) as a function of equilibrium concentration of VO3-(Ceq). Both TiO2 samples gave Langmuir-type isotherms. From linear plots (Ceq vs. Ceq/Ca~s) (Fig. 4(b)), the limiting

1094 amounts of Cads (maXCads) were estimated to be 0.91 and 1.4 mmol g-1 for unmodified and modified TiOz, respectively, indicating that adsorptivity of VO3- onto TiO2 was improved by modification of TiOz with APS. Since the methoxy group of APS is reactive to HzO or moisture, pKa of APS can not be determined directly in the presence of HzO. Here, pKa of propylamine (10.7) is regarded as that of the amino group of APS-TiOz. The large pKa indicates that the amino group of APS-TiOz is protonated in the NH4VO3 solution (pH 6.8). Therefore, the strong interaction between VO3- and the protonated amino group in APS-TiO2 is attributable to the larger amount of VO3- adsorbed. From the results of chemical analysis and maXCads, the ratio of VO3- adsorbed to APS was calculated to be 1.0, indicating that all of the amino groups of APS-TiOz interacted with VO3-. Therefore, despite the higher loading, greater dispersion of VO3- on the APS-TiOz support is expected because one VO3- anion interacts with the protonated amino group of APS. 1.5

(a)

=

"o,

20 l

(b)

~ 15 0~

0

E E

"o

co

0

0.5 5

0 1

=

0

5 10 15 20 0 5 10 15 20 Fig. 4. (a) Isotherms of adsorption of VO3- (Ceq-Cads plot) and (b) Ceq-Ceq/Cads plot. Circles and squares show data of unmodified TiO2 and APS-TiOz, respectively. 0

3.3. Properties and deNOx activities of supported V205 catalysts The unmodified and modified TiOz samples adsorbing VO3- in the most-concentrated NH4VO3 solution (2.00 • 10.2 mol dm -3) were decomposed at 550~ for 1 h in air to form two supported V205 catalysts. The VzO5 catalyst, prepared from the VO3--adsorbing unmodified TiO2, had a small surface area (81 mZg-1), as expected from the low thermal stability of the unmodified TiOz. It is well known that an excess loading of VzO5 on a TiO2 support generally induces sintering and pore-mouth plugging of the support, resulting in a decrease in surface area of the catalyst. Therefore, the amount of VO3- loading (1.0 mmol g-i) on the unmodified TiOz from the NH4VO3 solution exceeded the limited amount of the support for effective dispersion of V205 on it. On the other hand, the other V205 catalyst, prepared from the VO3--adsorbing APS-TiOz sample, retained a sufficient surface area (144 mZg-1) as expected from the improved thermal stability of the APS-TiO2 support. This indicates that VzO5 species were effectively dispersed on the modified TiOz support despite

1095 the large amount of V205 loading. Thus the modification of a TiO2 support with a large surface area with APS enables preparation of V205 catalyst with both a large surface area and a large amount of V205 loading.

60

~

40

> 0 E •

0

20

Z

I

I

I

300 350 400 Temperature/bO Fig. 5. Removal of NOx over V205 catalysts. Circles and squares show results of the catalysts prepared from VO3-adsorbing unmodified and APS-modified TiO2 samples, respectively. Reaction conditions are described in the Experimental section. These two V205 catalysts were used for SCR of NOx with ammonia in a temperature range from 300 to 400~ and the results are shown in Fig. 5. With elevation in the reaction temperature, the amount of NOx removed increased. Due to the large amount of highly dispersed VzO5 species, i.e., large number of active sites, V=O, the V205 catalyst prepared from APS-TiO2 exhibited higher activity than that of a catalyst prepared from unmodified TiO2. 4. CONCLUSIONS Chemical modification of a TiO2 support with a large surface area with organosilane having an amino group suppressed the growth and sintering of anatase crystallite, resulting in improved thermal stability of the support. The adsorption properties of VO3- were also improved by the strong interaction with VO3- and the protonated amino group in the modified TiO2. A supported V205 catalyst prepared from VOa--adsorbing modified TiO2 satisfies both the requirements of large surface area and large V205 loading and, in SCR of NOx with ammonia, exhibits higher activity than that of a catalyst prepared from unmodified TiO2.

1096 REFERENCES

1. S. Matsuda and A. Kato, Appl. Catal., 8 (1983) 149. 2. M. Inomata, A. Miyamoto and Y. Murakami, J. Chem. Soc., Chem. Commun., (1980) 233. 3. E Luck, Bull. Soc. Chim. Belg., 100 (1991) 781. 4. M. Kamada, H. Nishijima and Y. Kera, Bull. Chem. Soc. Jpn., 66 (1993) 3565. 5. Y. Hanada, M. Kamada, K. Umemoto, H. Kominami and Y. Kera, Catal. Lett., 37 (1996) 229. 6. M. Kamada, H. Kominami and Y. Kera, J. Colloid Interface Sci., 182 (1996) 297. 7. M. Kamada, H. Kominami and Y. Kera, Nippon Kagaku Kaishi, (1996) 300 [in Japanese]. 8. Y. Kera, M. Kamada, Y. Hanada and H. Kominami, Composite Interfaces, 8 (2001) 109. 9. H.P. Boehm, Disc. Faraday Soc., 52 (1971) 264. 10. S. Iwamoto, W. Tanakulrungsank, M. Inoue, K. Kagawa and E Praserthdam, J. Mater. Sci. Lett., 19 (2000) 1439. 11 H. Kominami, M. Kohno, Y. Matsunaga and Y. Kera, J. Am. Ceram. Soc., 84 (2001) 1178.