Regeneration of the Waste Selective Catalytic Reduction Denitrification Catalyst by Nitric Acid Washing.

In this study, the waste V2O5-WO3/TiO2 denitrification catalysts from the coal-fired power plant were washed with water or nitric acid, followed by impregnating different contents of V2O5. The effects of the HNO3 concentration and the additional amount of vanadium on the low-temperature selective catalytic reduction denitrification activity were investigated under the condition of high concentration of SO2 and H2O. The catalysts were characterized by inductively coupled plasma optical emission spectrometry, X-ray powder diffraction , N2 adsorption/desorption, H2-temperature-programmed reduction, NH3-temperature-programmed desorption , Fourier transform infrared spectroscopy , and Raman spectroscopy. The evaluation results revealed that optimum activity was achieved by using 0.8 mol/L HNO3 solution and loading 1.60 wt % V2O5 to make the total V2O5 reach 2.3 wt %. The characterization results showed that nitric washing can remove most of the ammonium salts deposited on the surface of the waste catalyst and produce crystalline WO3, which can effectively inhibit the agglomeration of vanadium species in the process of impregnation. Furthermore, it can also increase the amount of oligomeric VO x , which can improve the denitration activity.

Introduction

In the past decades, more and more attention has been paid to the emission of NO (NO, NO2) because of its serious harm to human health and environment. More than 90% of NO in the atmosphere comes from the combustion of coal, oil, and natural gas; 70% of which comes from the direct combustion of coal. According to the statistics in 2013,[1] the power industry as a large coal user accounted for 46% of coal consumption in China. The coal-fired power plants are one of the main sources of NO emissions. Among all NO elimination technologies, the selective catalytic reduction (SCR) by ammonia is regarded as one of the most effective technologies. It is widely used in power plants and other industrial sites.[2] Moreover, it is necessary to continue to expand the application of the SCR method in flue gas denitrification of coal-fired power plants in order to meet increasingly stringent environmental requirements. The core technology of the SCR method is the V2O5–WO3(MoO3)/TiO2 catalyst which accounts for 30–50% of the overall cost[3] and works in the temperature range of 300–450 °C.[4]

The catalytic ability of the SCR catalysts will gradually decrease with the increase of running time because of the deactivation of the catalyst. The SCR catalyst deactivation is usually caused by the following factors: blockage of fly ash and ammonium salts,[5] poisoning of alkali metals[6] and alkaline earth metals,[7] arsenic poisoning,[8] plumbum poisoning,[9] catalyst sintering,[10] mechanical wear, and so on. Therefore, it needs to be replaced regularly. Moreover, the waste SCR catalyst (vanadium–titanium series) has been incorporated into hazardous solid waste in China because of its own potential environmental hazards and various heavy metals adsorbed in the process of operation.[3]

Since 2013, a large number of SCR denitrification projects have been carried out in thermal power units. Eighty percent of the SCR catalysts is imported and their service life is 5 years. Therefore, China will usher in the first peak of catalyst scrapping in 2019. The waste SCR catalysts are renewable and recyclable. Regeneration is mainly to improve the denitrification activity of the waste catalyst, which has a low cost. Recycling is mainly to extract TiO2, V2O5, and WO3 from the waste catalyst, which has a high cost. Seventy to eighty percent of these catalysts is renewable,[11] and it is necessary to strengthen the research on waste catalyst regeneration to replace the common landfill treatment.[12]

Taking the honeycomb type as an example, the regeneration of the waste catalyst is generally divided into two parts: the preliminary washing treatment and the later active component supplementation. The later active component supplementation is usually carried out by immersing the block catalyst in a vanadium-containing solution; hence, the main difference lies in the preliminary washing mode. Common washing methods include the acid method, alkali method, acid–alkali combined method, and ammonium salt method. The acid method includes sulfuric acid and oxalic acid cleaning which can remove alkali metals and other impurities on the surface of the waste catalyst.[13,14] The alkali method mainly uses sodium hydroxide cleaning which can remove aluminum sulfate deposited on the surface of the waste catalyst and reduce its blockage.[15] The acid–alkali combined method uses alkali washing followed by acid washing,[16] which aims at removing a small amount of alkali metals introduced in the process of alkali washing. The ammonium salt method mainly contains ammonium sulfate and ammonium chloride cleaning,[17] and it is found that this method can also improve the denitrification activity of the waste catalyst.

At present, catalyst regeneration is usually carried out by washing a whole piece of the waste catalyst. The washing solution can effectively remove the alkali metal, ammonium salt, and other impurities on the surface of the waste catalyst, but the impurities in the pore passage are difficult to clean and in some local areas, deactivation still remained, which will also lead to misdistribution of the active components on the carrier. Currently, sulfuric acid is commonly used during the process of acid washing and most of the regeneration of the catalyst focused on its recovery to the original activity as far as possible. Therefore, in this study, the waste catalyst was first crushed and then washed with dilute nitric acid. Finally, certain active components were added and reshaped, which was focused on the improvement of the overall SCR denitrification activity including the low-temperature activity under the condition of high concentration of SO2 and H2O.

Experimental Section

Source of Catalysts

The waste commercial honeycomb monolith catalysts were obtained from a coal-fired power plant in Shanxi Province provided by a catalyst recovery plant. The catalysts had been in operation for about 40 000 h. The size of the whole piece of the waste catalyst is 90 cm × 14 cm × 14 cm, and the wall thickness is 1.5 mm. In addition, the surface and the pore passage of catalysts were repeatedly rinsed using high-pressure water in the catalyst recovery plant, so it was considered that the original sample had been treated with water. The composition of the waste catalyst after water washing is shown in Table . The waste catalyst is the V2O5–WO3/TiO2 catalyst. It can be found that the waste catalyst contains a high concentration of the poisoning elements such as Ca and S.

Table 1

Composition of the Waste Catalyst after Water Washing

component	Ti	W	V	Al	Si	Ca	Fe	S
content (wt %)	48.16	3.88	0.47	0.49	4.20	1.00	0.093	0.54

Catalyst Regeneration

Before regeneration, the waste catalysts were crushed to a particle size of less than 75 μm, and the regeneration solution was a mixture of oxalic acid and ammonium metavanadate. In these experiments, the following two methods were used to regenerate the catalysts.

One method was to directly support active vanadium on the waste catalysts which had been washed with water. Incipient wet impregnation was used and the mass fractions of V2O5 were selected as 1.3, 1.8, 2.3, and 3.3%, which were based on the total vanadium content of the catalysts. Then, the catalysts were dried at 100 °C for 10 h and calcined at 500 °C for 3 h to obtain the regenerated catalysts.

The other method was to wash the waste catalysts with 0.1, 0.3, 0.5, 0.8, and 1.0 mol/L HNO3 at 30 °C for 3 h, L/S = 5, and the catalysts filtered with a lot of deionized water were dried at 110 °C for 2 h. Then, each sample was loaded with 1.60% V2O5 to make the total V2O5 reach 2.3%. Finally, the catalysts were dried at 100 °C for 10 h and calcined at 500 °C for 3 h to obtain the regenerated catalysts. The catalyst samples were labeled as shown in Table .

Table 2

Labeling of the Catalysts

catalysts	descriptions
WW	the waste catalysts obtained from coal-fired power plant and which had been washed by water.
WW-x V (y %)	the WW catalysts were loaded with x % V2O5, the total V2O5 reach to y %. The value of x was 0.47, 0.97, 1.47 and 2.47, the value of y was 1.3, 1.8, 2.3 and 3.3, respectively.
zNW	The WW catalysts were washed by z mol/L HNO3.
zNW-1.60 V	The zNW catalysts were loaded with 1.60% V2O5, the value of z was 0.1, 0.3, 0.5, 0.8 and 1.0, respectively.

Catalytic Activity Evaluation

All the catalytic evaluations of the regenerated catalysts were tested using a fixed-bed quartz reactor, and the inner diameter of the quartz tube was 10 mm. The device schematic diagram of NH3-SCR denitrification is shown in Figure S1, and the related introduction of the experiment device was described in our published article.[18] During a denitration experiment, the particle size of the catalyst was 0.25–0.42 mm and 200 mL/min simulated flue gas balanced with He was injected into the system. The composition of the flue gas was NO (500 ppm), NH3 (500 ppm), O2 (5 vol %), SO2 (1000 ppm), and H2O (10 vol %). Each flow rate was monitored by mass flow meters and the gas hourly space velocity of the experiment was set as 120 000 h–1. The concentration of NO and N2O was measured by Fourier transform infrared spectroscopy (FT-IR). The NO conversion and selectivity of N2 were calculated according to eqs and 2.Where ηNO stands for the NO conversion, CNO,in and CNO,out represent the import and export concentrations of NO, respectively. SN is the N2 selectivity, CNO is the outlet concentration of NO2, CN is the outlet concentration of N2O, and CNH is the inlet concentration of NH3.

Catalyst Characterization

The inductively coupled plasma optical emission spectrometry (ICP–OES) analysis was carried out on an ICP emission spectrometer (SPECTRO, Kleve, Germany). The sample digestion method was as follows: 0.06 g of sample was placed in the reaction kettle, and then 7 mL of HNO3, 2 mL of H2O2 and 1 mL of HF were added to the kettle for 30 min at 120 °C. Finally, the solution was introduced from the kettle into a volumetric flask, and the volume of the solution was found to be 100 mL. X-ray diffraction (XRD) analysis was carried out on a Rigaku MiniFlex 600 type X-ray diffractometer equipped with Cu Kα radiation source (λ = 0.154056 nm). The XRD pattern was scanned in the 2θ range from 5° to 90° at the scanning rate of 10°/min. The specific surface area and pore structure parameters of the catalysts were carried out with a nitrogen physical adsorption instrument (Micromeritics, Norcross, GA, USA). H2-temperature-programmed reduction (H2-TPR) and NH3-temperature-programmed desorption (NH3-TPD) analyses were carried out on the AutoChem II type high-performance chemical adsorption instrument (Micromeritics, Norcross, GA, USA). For H2-TPR, the 100 mg sample was pretreated in a He flow (20 mL/min) for 1 h at 300 °C and the reactor was cooled down to 50 °C. Then, the sample was heated up to 1000 °C at a rate of 10 °C/min in a 10% H2/Ar flow (20 mL/min), and the concentration of H2 was monitored by the signal of the thermal conductivity detector (TCD). For NH3-TPD, pretreatment is the same as that for H2-TPR and then exposed to 10% NH3/He at 50 °C for 30 min, followed by a temperature ramp to 800 °C at a rate of 10 °C/min, and the concentration of the desorbed NH3 was also monitored by the TCD. The FT-IR spectra were collected on a Bruker TENSOR 27 (Bruker Corporation, Karlsruhe, Germany). The Raman spectra were recorded with a Renishaw in Via Raman spectrometer (RENISHAW LTD, UK). Scanning electron microscopy (SEM) (Jsm-6700F, JEOL, Japan) images of the samples were obtained using an accelerating voltage of 10 kV.

Results and Discussion

Chemical Compositions of Catalysts

The main chemical compositions of zNW catalysts are presented in Table . With the increase of nitric acid concentration, the mass fraction of V2O5 (calculated based on the mass fraction of V) generally presents a downward trend, while it is basically maintained at about 0.7%. The content of Na decreases gradually, and finally about 83% of Na is removed, which greatly reduces the alkali metal poisoning of the catalysts. A part of Fe can be effectively removed by nitric acid washing. Besides, the content of S is greatly reduced, while the content of Ca and Al is basically unchanged. The main forms of sulfur may be sulfuric acid, ammonium sulfate salt, aluminum sulfate, and calcium sulfate. However, the surface of the waste catalyst is substantially free of aluminum sulfate and calcium sulfate. The significant reduction of S is caused by the dissolution of a large amount of ammonium sulfate salts or sulfuric acid on the surface of the catalyst. In addition, the content of heavy metals such as Pb, As, and Hg is at an extremely low level, which is basically outside the detection range of ICP–OES. This indicates that the catalyst is not poisoned by heavy metals.

Table 3

Main Chemical Compositions of the zNW Catalysts

	content (wt %)
sample	Ti	W	V	Al	Si	Ca	Fe	S	Na
WW	48.16	3.88	0.47	0.49	4.20	1.00	0.093	0.54	0.060
0.1NW	47.86	3.90	0.41	0.48	3.80	0.99	0.080	0.14	0.031
0.3NW	46.40	3.94	0.40	0.46	3.74	0.94	0.070	0.08	0.025
0.5NW	48.20	3.89	0.39	0.48	3.86	0.99	0.071	0.06	0.022
0.8NW	48.22	3.84	0.37	0.45	3.77	0.92	0.067	0.06	0.010
1.0NW	48.43	4.05	0.39	0.49	4.00	1.02	0.062	0.07	0.010

Pore Structure Parameters of Catalysts

The surface area, pore volume, and average pore diameter of WW and zNW catalysts are presented in Table . With the increase of the concentration of nitric acid, the specific surface area of the waste catalyst increases gradually and the pore size decreases generally. It can be concluded that the appropriate increase of the nitric acid concentration can reopen the blocked pores and increase the amount of pores with smaller diameter, which is beneficial to the uniform distribution of vanadium added to the surface of waste catalysts and the increase of new denitration active sites.

Table 4

Surface area and average Pore Size of WW and zNW Catalysts

sample	surface area (m2/g)	pore volume (cm3/g)	average pore diameter (nm)
WW	55.97	0.271	19.34
0.1NW	57.50	0.270	18.78
0.3NW	64.12	0.278	17.39
0.5NW	69.13	0.234	13.52
0.8NW	69.52	0.259	14.90
1.0NW	69.87	0.257	14.72

The pore size distribution of catalysts is shown in Figure . Compared with the WW catalyst, the pore volume between 2 and 16 nm of the zNW catalysts is greatly increased with increasing concentration of nitric acid. It is speculated that the catalyst produces many pores with pore sizes between 2 and 16 nm. Moreover, nitric acid plays a main role in the pore enlargement of the waste catalyst during the washing process.

Figure 1

Pore size distribution of catalysts.

Catalytic Activity Results

Figure a shows the denitration activity profiles of the WW-x V (y %) catalysts. It can be seen that its low-temperature activity in the range of 180–250 °C increases with increasing vanadium content under the condition that the total V2O5 does not exceed 2.3% especially when the total V2O5 amount is 2.3%, the low-temperature activity is greatly improved. The denitration rate increases from 30 to 80% at 225 °C. Besides, its high-temperature activity in the range of 250–400 °C also increased with increasing vanadium content but obviously lower than those of WW catalysts. This is due to the masking effect on the original active components caused by the reloaded active components. When the total amount of V2O5 is 3.3%, the high-temperature activity is unchanged compared with that of the former, but the low-temperature activity is significantly decreased. Therefore, the following experiments are based on the total amount of V2O5 being 2.3%.

Figure 2

Denitration activity of the (a) WW-x V (y %) and (b) zNW-1.60 V, 0.8NW catalysts.

According to the results of ICP, the total amount of V2O5 can reach 2.3% when the waste catalysts after different concentrations of HNO3 washing were all loaded with 1.60% V2O5. Figure b shows the denitration activity profiles of the zNW-1.60 V and 0.8NW catalysts. When the temperature is lower than 225 °C, the denitration activities of zNW-1.60 V and WW-1.47 V (2.3%) are similar. However, in the temperature range of 225–400 °C, the concentration of the acid has important effects on the activity of catalysts. In summary, the overall activity of the 0.8NW-1.60 V catalyst reaches the highest level. The high-temperature activity increases to more than 85%, which is higher than that of WW. This is consistent with the activities of 0.8NW. Because the concentration of sulfur dioxide is different in actual working conditions, the denitration performance under different SO2 concentrations is also investigated in order to test its regeneration performance, which is shown in Figure S2. It can be seen that good denitration activity can still be maintained with the change of sulfur dioxide concentration.

According to the denitration activity of all the regenerated catalysts in the two schemes, the WW-1.47 V (2.3%) and 0.8NW-1.60 V catalysts are selected to detect the N2O formation and N2 selectivity which is presented in Figure . Clearly, the N2O formation of the 0.8NW-1.60 V catalyst is less than that of WW-1.47 V (2.3%) and the 0.8NW-1.60 V catalyst has a better N2 selectivity which can be above 99% in the range of 100–360 °C. It indicates that nitric acid washing can improve the N2 selectivity of the regenerated catalyst.

Figure 3

N2O formation (a) and N2 selectivity (b) of the WW-1.47 V (2.3%) and 0.8NW-1.60 V catalysts.

Crystal Structures of Catalysts

The XRD patterns of WW, WW-1.47 V (2.3%), 0.8NW, and 0.8NW-1.60 V are presented in Figure . All these catalysts exhibit the typical anatase phase of TiO2, and no crystalline V2O5 species are observed. It indicates that the crystal structure of TiO2 is not destroyed by nitric acid washing and it is still anatase. The vanadium added to the surface of the waste catalyst is highly dispersible.

Figure 4

XRD patterns of catalysts.

H2-TPR Analysis

The H2-TPR profiles of the WW, WW-1.47 V (2.3%), 0.8NW, and 0.8NW-1.60 V catalysts are shown in Figure a. The WW catalyst shows two peaks at 525 and 850 °C. The first reduction peak is attributed to the reduction process of V5+ → V3+, W6+ → W4+, and SO42– → SO2, and the second reduction peak is the reduction process of W4+ → W0.[19] It can be seen that the first reduction peak of both WW-1.47 V (2.3%) and 0.8NW-1.60 V are reduced from 525 to 459 °C, indicating that the reduction ability of these two regenerated catalysts are significantly enhanced and close to each other, mainly because of the addition of vanadium in the waste catalysts. Correspondingly, the low-temperature activity of these two catalysts is greatly enhanced and also close to each other. For 0.8NW, the first reduction peak is reduced from 525 to 475 °C and its reduction ability has also been improved to some extent. But the peak intensity is significantly weakened because a small amount of the vanadium is washed away by nitric acid. At the same time, 0.8NW and 0.8NW-1.60 V show a new reduction peak located at around 580 °C, which may be the reduction peak of crystalline WO3 according to literature studies.[20]

Figure 5

H2-TPR results of (a) WW, WW-1.47 V (2.3%), 0.8NW, and 0.8NW-1.60 V and (b) zNW-1.60 V catalysts.

According to the previous research, for the V2O5–WO3/TiO2 catalyst, WO3 is an auxiliary agent, which is introduced to compete with the interaction between VO and TiO2 through the interaction between WO and TiO2 so as to influence the distribution and morphology of VO on the carrier.[21] Because VO is easily distributed on TiO2 adjacent to WO, the content and morphology of WO directly affects the state of VO on the carrier. WO on the carrier can be divided into amorphous WO and crystalline WO3, and the appropriate amount of amorphous WO can promote the dispersion of moderately distorted vanadium species and prevent the self-island process of vanadium species.[19] However, excessive amorphous WO will squeeze its neighboring vanadium species and lead to VO aggregation or the formation of crystalline V2O5.[20] In the case of only a certain amount of amorphous WO, when the amount of vanadium loading is low, it is beneficial to the formation and uniform dispersion of moderately distorted vanadium species. When the amount of vanadium loading is high, the compression effect of WO on single-layer VO will occur. Therefore, the existence of a suitable amount of crystalline WO3 can alleviate the competition between VO and WO by reducing the richness of WO and releasing a certain surface space on the carrier.

As shown in Figure b, the denitration activity of the 0.8NW-1.60 V catalyst is better than that of the WW-1.47 V (2.3%) catalyst in the temperature range of 225–400 °C. It is possible that the amorphous WO on the surface of the WW catalyst has a strong compression effect on the large amount of vanadium added which leads to a certain degree of agglomeration of VO, resulting in the decrease of activity at high temperature. For the 0.8NW-1.60 V catalyst, because of the oxidation of nitric acid, a certain amount of crystalline WO3 is formed on the surface of the waste catalyst after nitric acid washing, which can reduce the compression effect of WO on VO and inhibit VO agglomeration even if the amount of vanadium is high. In this way, more effective active sites are provided and the high-temperature activity is also enhanced. As shown in Figure b, with the washing treatment at a higher nitric acid concentration, the reduction peak strength of crystalline WO3 increases gradually, indicating that the amount of crystalline WO3 on the surface increases gradually.

NH3-TPD Analysis

Figure shows the NH3-TPD results of the catalysts. There are four desorption peaks at about 90, 165, 330, and 575 °C which represent the weak acid, medium strong acid, strong acid[22] and decomposition of ammonium sulfate on the surface of the catalysts, respectively.[23] The peak of the weak acid and medium strong acid could be related to NH3 desorbed from Brønsted acid sites and the peak of the strong acid could belong to NH3 desorbed from Lewis acid sites.[24] It can be concluded that ammonium sulfate deposits on the surface of the waste catalyst. For WW-1.47 V (2.3%) catalysts, the peak of ammonium sulfate has almost disappeared which may be because of the decomposition of ammonium salts caused by calcination in the sample preparation process, and a part of the blocked channel is opened and the reloaded vanadium increases its low-temperature activity. However, it can be seen from the figure that because of a certain degree of vanadium agglomeration on the surface of the carrier, the number of strong acid sites is greatly reduced, resulting in a decrease in the high-temperature activity. Compared to the WW-1.47 V (2.3%) catalyst, the number of strong acid sites of 0.8NW increases except for the disappearance of the ammonium sulfate peak, which raises its high-temperature activity. However, the number of weak acid sites decreases significantly, which decreases its low-temperature activity. For 0.8NW-1.60 V catalysts, in addition to the further increase in the number of strong acid sites, the number of weak acid sites has also been greatly improved. Therefore, its overall denitration activity reaches the optimum level. This shows that most of the ammonium sulfate is washed away, and more strong acid sites are produced under the condition of high vanadium content after being washed by 0.8 mol/L HNO3.

Figure 6

NH3-TPD profiles of the WW, WW-1.47 V (2.3%), 0.8NW, and 0.8NW-1.60 V catalysts.

FT-IR Analysis

Figure shows the FT-IR spectra of the WW, WW-1.47 V (2.3%), and 0.8NW-1.60 V catalysts. The absorption bands at 850–1075 cm–1 are enlarged in Figure a. The weak band at 879 cm–1 can be assigned to the V–O stretching mode of the metavanadate polymeric species.[25] Generally, the intensity of this peak will gradually decrease with the increase of vanadium, and the reason is that the vanadium ions interacted with hydroxyl groups on the surface of the carrier will gradually transform into vanadium oxide clusters and weaken the interaction with the carrier. The 0.8NW-1.60 V catalyst has a distinct peak at this position compared with that of WW-1.47 V (2.3%), which indicates that there are a certain number of vanadium ions on the surface, and the agglomeration is at a relatively low level. The band at 987 cm–1 can be assigned to the characteristic peak of W=O and V=O, and the peak strength increases slightly with the increase of the vanadium content. The band at 1050 cm–1 is related to the typical band caused by V=O,[26] in which the peak intensity increases with increasing the content of vanadium oxides, generally. However, the peak strength of WW-1.47 V (2.3%) is less than that of WW, which indicates that the addition of vanadium to the water-washed catalyst may give rise to the agglomeration of vanadium oxide on the surface of the carrier, and the amount of effective V=O is reduced, that is, the number of Lewis acid sites is decreased, resulting in the decrease of the high-temperature activity of the catalyst. For the 0.8NW-1.60 V catalyst, the peak strength at 1050 cm–1 is clearly stronger than that of the other two catalysts, which indicates that the amount of effective V=O on its surface has increased significantly, and its high-temperature activity is higher than that of the other two catalysts.

Figure 7

FT-IR spectra at (a) 850–1075 and (b) 1050–1500 cm–1 of the WW, WW-1.47 V (2.3%), and 0.8NW-1.60 V catalysts.

The absorption bands at 1050–1500 cm–1 are enlarged in Figure b. The bands at 1211 and 1134 cm–1 are recognized as the characteristic peaks of SO42−,[27] and the band at 1400 cm–1 is associated with NH4+.[28] It can be seen that the WW catalyst has obvious peaks of SO42– and NH4+, indicating that a certain amount of ammonium salt or other sulfates are deposited on the surface of the WW catalyst, causing the blockage of pores. These peak intensities of WW-1.47 V (2.3%) are greatly reduced, which may be because of the decomposition of the ammonium salt caused by calcination in the sample preparation process or the masking of vanadium but some sulfates still exist. These peak intensities of 0.8NW-1.60 V are almost completely disappeared, indicating that most sulfates can be removed by 0.8 mol/L nitric acid washing, and it is more conducive to the distribution of reloaded vanadium. The bands at 1450 and 1378 cm–1 are considered as the symmetric and asymmetric bending vibrations, respectively, of NH4+ coordinated to Brønsted acid sites,[29] while the bands at 1280 and 1090 cm–1 are considered as the asymmetric and symmetric bending vibrations of NH3 coordinated to Lewis acid sites.[30] Compared with the WW catalyst, the absorption peak strength of the WW-1.47 V (2.3%) catalyst at 1378 cm–1 increases slightly, and the Brønsted acid sites increase, which increases its low-temperature activity. For the 0.8NW-1.60 V catalyst, it produces two new peaks at 1450 and 1280 cm–1. The peak strength at 1378 cm–1 is further enhanced, and the peak strength at 1090 cm–1 is increased significantly compared with the other two catalysts, which shows that more Lewis acid sites are produced, increasing its high-temperature activity.

Raman Analysis

There are four micrographs with the same multiple taken by using a Raman spectrometer as shown in Figure S3. The surface of the WW catalyst is rough with a large particle size and incompact structure. It indicates that the catalyst has a certain sintering and agglomeration phenomenon. The yellow part on the surface of WW-1.47 V (2.3%) increases obviously and is more aggregated, which indicates that the dispersity of vanadium is poor. The surface of the 0.8NW catalyst is very smooth and flat, and most of the sulfate is removed. Besides, the yellow part of the 0.8NW-1.60 V catalyst has good dispersion and no phenomenon of high agglomeration. This is consistent with the results of SEM and element mapping in Figure S4.

Figure a shows the results of Raman spectroscopy of three samples. The bands at 396, 515, and 638 cm–1 are attributed to the anatase TiO2 structure,[31] and all three catalysts have clear anatase characteristic peaks. According to the previous study, the band at 1026 cm–1 can be assigned to the V=O vibration related to the surface of oligomeric VO,[32] and the band at 990 cm–1 is the characteristic peak of the V–O vibration of crystalline V2O5 when the vanadium loaded is high.[33,34] The band at 800 cm–1 can be assigned to the W–O vibration of crystalline WO3.[35,36] It can be seen that the peak strength of WW-1.47 V (2.3%) at 990 cm–1 is obviously enhanced compared with that of WW, which indicates that there are more crystalline V2O5 on the surface with distinct agglomeration. This may be the reason for the decrease in high-temperature activity. However, the peak of 0.8NW-1.60 V disappears at 990 cm–1 and a new peak is generated at 1026 cm–1, showing that the vanadium species on the surface of the sample has no agglomeration. It produces more oligomeric VO, which increases its high-temperature activity. Meanwhile, the peak strength at 800 cm–1 is significantly enhanced, indicating that nitric acid treatment increases the amount of crystalline WO3 on the surface, which reduces the richness of WO on the surface of the carrier and relieves the compression effect on VO. This effectively prevents vanadium species from aggregating in the case of high vanadium content, resulting in the formation of more oligomeric VO, which is consistent with the results of H2-TPR.

Figure 8

Raman spectra of the (a) WW, WW-1.47 V (2.3%), and 0.8NW-1.60 V and (b) zNW-1.60 V catalysts.

The Raman bands of several zNW-1.60 V catalysts at 850–1200 cm–1 are enlarged in Figure b. Compared with WW-1.47 V (2.3%), the peak intensity of zNW-1.60 V at 990 cm–1 decreases until it disappears with the increase of nitric acid concentration. It also means that the amount of crystalline V2O5 is gradually decreased, even if the total amount of vanadium is equal to each other. Finally, more and more oligomeric VO is produced, which improves the high-temperature activity of the catalysts.

Conclusions

In this article, the mixed solution of ammonium metavanadate and oxalic acid was used as a precursor solution, and the regenerated catalyst was prepared by separately loading vanadium onto the waste catalyst after water or nitric acid washing. The conclusions can be drawn as follows:

In WW-x V (y %) type catalysts, the WW-1.47 V (2.3%) catalyst exhibits an excellent low-temperature activity, especially the denitration rate increases from 30 to 80% at 225 °C, but its high-temperature activity is poor, only about 60%. In zNW-1.60 V type catalysts, the 0.8NW-1.60 V catalyst not only has the same low-temperature activity as the former but also has a high-temperature activity of over 85%.

The H2-TPR and Raman results show that nitric acid washing could generate a certain amount of crystalline WO3 on the surface of the waste catalyst, which reduces the compression effect of WO on VO when the amount of vanadium is high and inhibits the aggregation of vanadium effectively. Meanwhile, more oligomeric VO are formed on the surface, improving its overall denitration activity.

The 0.8NW-1.60 V catalyst can not only fully achieve the denitrification effect of commercial catalysts in coal-fired power plants but also can be applied to the environment with lower flue gas temperature, such as glass plants and coking plants, because of its excellent low-temperature activity and sulfur resistance.