Adsorption and Decomposition of NO x on Heteropolyacids: An Evaluation of the Adsorption Performance

In order to develop an efficient method by which to eliminate NO x pollution, several new catalyst systems including different heteropolyacids (HPAs) and supported phosphotungstic acids (HPWs) (HPW/SiO 2 , HPW/SnO 2 , HPW/USY and HPW/ZSM-5) for adsorption-decomposition of NO x were prepared and studied. The obtained catalysts were characterized using Brunauer-Emmett-Teller (BET) measurement and Fourier transform infrared (FT-IR) analysis. The results showed that W-containing HPAs were superior to Mo-containing HPAs. Among the selected catalysts, HPW/SnO 2 with a 50% HPW loading had the highest NO x adsorption rate of 77.3%, for which the amount of saturated NO x adsorption (85.4 mg g –1 ) was much higher than that of HPW (50.5 mg g –1 ). The NO x adsorption performance of the catalyst was mainly determined by the interaction between the support and the HPW, which was also affected by the specific surface area of the catalyst. FTIR characterization revealed that the adsorbed NO x mainly existed in the HPW bulk phase in the form of NOH + . A gas chromatograph-mass spectrometer (GC-MS) was used to confirm the effectiveness of NO x decomposition into N 2 .


INTRODUCTION
) are major air pollutants derived from stationary power sources and mobile sources that not only are extremely toxic to human body, but also can lead to photochemical smog, acid rain, ozone depletion, and the greenhouse effect (Liu et al., 2014;Jiang et al., 2016;Huang et al., 2019). Alternative well-developed technologies have been employed for eliminating NO x emissions in response to environmental protection requirements. Among them, selective catalytic reduction of NO x with NH 3 (NH 3 -SCR) has been considered to be one of the most efficient techniques for NO x removal due to its efficiency, economy, and selectivity (Forzatti et al., 2009;Imanaka and Masui, 2012). However, the leakage of ammonia can cause secondary pollution. The catalytic decomposition of NO x has attracted much attention because it can directly decompose NO x into N 2 and O 2 without consuming any reducing agent, and the operation process is simple, economical, and without secondary pollutants. In recent years, various catalysts, including precious metals, metal oxides, ion exchange molecular sieves, perovskite-type composite oxides and carbonitrides (Konsolakis and Michalis, 2014;Sun et al., 2014), have been used in direct decomposition of NO x under high temperature conditions (> 500°C).
Polyoxometalates (POMs) have been used in a variety of catalytic fields owing to their strong Brønsted acidity, remarkable redox properties, and pseudo-liquid phase behavior (Wang and Yang, 2015;Ren et al., 2017a, b). In particular, HPAs have shown attractive NO x adsorption performance and can decompose the adsorbed NO x into N 2 through rapid heating. Yang and Chen (1994; reported that NO x was adsorbed on HPW in the form of protonated NO (NOH + ), and N-O cleavage was generated in the rapid heating process, thus resulting in N 2 yield. However, the small specific surface area and low mechanical strength of HPA and its inferior thermal stability makes it difficult to reuse in practical applications. Therefore, many supported HPA materials have been developed to improve its adsorption performance, catalytic decomposition activity, and stability. In representative lean-gas mixture conditions, the NO x storage and reduction performance of H 3 PW 12 O 40 ·6H 2 O loaded on Ti-Zr and Ce-Zr mixed oxides were studied and shown to have high N 2 selectivity (Gómez-García et al., 2005a, b). Platinum (Pt) and HPW doped mesoporous MSU-type silica mixtures were used for NO x adsorption (Hamad et al., 2007). HPW/CNTs catalysts were prepared, and the results showed that the yield of N 2 after microwave heating was higher than that obtained with rapid heating using a resistance furnace (Zhang et al., 2012). Zhang et al. (2013) (Wu et al., 2021). The NO x adsorption performance of the catalyst determines the amount of NO x decomposed. The development of catalysts with high activity and stability have consistently been a research hotspot and have been proven to be quite challenging. The factors affecting the adsorption performance of catalysts also require further exploration. The objective of this work is to develop a novel, economical, and environmentally friendly NO x conversion catalyst and to further understand the constraints that affect catalytic performance. In this work, various HPAs and supported HPW (HPW/SiO 2 , HPW/SnO 2 , HPW/USY and HPW/ZSM-5) were prepared and used for the adsorption and decomposition of NO x . The physical and chemical properties of the catalyst were characterized using BET and FTIR spectra. The relationship between their NO x adsorption performance and their physicochemical properties was analyzed through comparative studies.  (HSiMo) and SnO 2 , were obtained from Sinopharm Chemical Reagent Co., Ltd (China). SiO 2 was purchased from Alfa Aesar. USY and ZSM-5 molecular sieves were purchased from the Zhoucun Catalyst Factory and Jiangsu Aoke Petrochemical Science and Technology Company. All reagents were analytical grade and were used directly without further purification.

Synthesis of HPW-Loaded Materials
HPW-loaded materials were prepared using the impregnation method. The mass ratios of HPW/(HPW + support ) were 0, 16.7%, 28.6%, 37.5%, 44.4%, and 50.0%. A specific mass ratio of the support was immersed in a quantitative HPW solution for 24 h and then dried in a 60°C water bath for 1 h to obtain the samples. The samples were denoted as HPW/SiO 2 , HPW/USY, HPW/SiO 2 , and HPW/ZSM-5.

Catalyst Characterization
FTIR spectra of the samples were recorded using a Nicolet Avatar370, and the spectral domain ranged between 400 and 4000 cm -1 at a 4 cm -1 resolution. The BET surface areas of the catalysts were measured based on the N 2 adsorption at -196°C using a Micromeritics ASAP 2010 apparatus.

Standard test procedure: Adsorption stage
The experiments were carried out in a fixed-bed quartz tube reactor (inner diameter = 8 mm). 0.3 g of the sample was placed in the middle of the reactor between two quartz wool plugs and pretreated with highly purified N 2 at 150°C for 1 h. During NO x adsorption, the feed gas containing 800 ppm NO, 8% O 2 , 4.2% H 2 O, and balance He was passed through a bench blending reactor and partially converted to NO 2 (2NO + O 2 = 2NO 2 ), yielding an NO x mixture. NO and NO 2 analyzers (TH-990S) were used to continuously record the concentration at the NO and NO 2 outlet.
The NO x adsorption amount in terms of NO 2 was calculated by integrating the curve below the baseline (1000 ppm) and expressed in mg g -1 . The calculation formula is shown below: where m represents the mass of the catalyst (g); Q is the total gas flow (m 3 min -1 ); C is the NO x removal concentration (mg m -3 ), and t is the adsorption time (min).

Decomposition stage
After NO x adsorption saturation, the reactor containing the catalyst was purged with He flow (5 mL min -1 ) for 1 h. Then, the reactor was placed in a tube furnace and rapidly heated from 30°C to 450°C at a heating rate of 150°C min -1 . The decomposition product was detected with a GC-MS analyzer.
NO conversion was then calculated using the following formula: where NO x (adsorbed) represents the amount of HPW the NO x adsorbed, and N 2 formation represents the amount of N 2 generated during NO conversion.

FTIR Analysis
FTIR spectra are the most commonly used research and measurement method applied to polyanions. The FTIR spectra of the supports before and after loading with HPW are shown in Fig. 1(a-d), and the FTIR spectrum of the HPW used in this work is shown in Fig. 1(e). Four characteristic peaks of the Keggin structure can be observed at 1080 cm -1 , 983 cm -1 , 892 cm -1 , and 801 cm -1 , corresponding to P-O a , W = O d , W-O b -W, W-O c -W, respectively, which is consistent with the literature (Yang and Chen, 1994). As shown in Fig. 1(a), for HPW/SiO 2 , the P-O a absorption peak of HPW at 1080 cm -1 became broader due to being masked by a wide SiO 2 absorption peak at 1086 cm -1 . The other three absorption peak positions of HPW shifted slightly, and their peak intensity decreased significantly. These findings implied that the HPW Keggin structure still existed after loading HPW on SiO 2 . The change in the characteristic peaks indicated that an interaction between HPW and SiO 2 occurred. The negative charge of the heteropolyanion increased after the HPW was loaded on SiO 2 , where the increased negative charge could fill its antibonding molecular orbital, leading to a reduction in the chemical bond strength and a decrease in the vibration frequency.
In Fig. 1(b), a broad peak can be observed at 400-750 cm -1 (650 cm -1 ), which was mainly attributed to the stretching vibration of the Sn-O band (Abello et al., 1998;Singh et al., 2014). When comparing the IR spectra of SnO 2 and HPW/SnO 2 , it can be seen that the four characteristic peaks of the Keggin structure on HPW/SnO 2 were stable, indicating that HPW was definitely loaded on the SnO 2 , and there was no strong interaction between the SnO 2 and the HPW, thereby preserving the heteropolyanion phase. As shown in Fig. 1(c), the USY framework structure can be identified by the peaks at 1700-400 cm -1 (Zhang et al., 2001;Munir and Usman, 2018). The absorption band at 1636 cm -1 was assigned to the bending vibration of the -OH group. The broad band from 1000 to 1250 cm -1 represented the asymmetric stretching vibration of Si-O-Si. The absorption bands at 815 cm -1 and 455 cm -1 were attributed to the bending vibration of Si-O-Si. In addition, a band at 590 cm -1 was observed, indicating the presence of a zeolite framework with tetrahedral units of SiO 4 and AlO 4 . The characteristic peak of the of USY continued to exist on HPW/USY, indicating that the cage-type micropore structure of USY was preserved, and the strong interaction between USY and HPW caused the characteristic peak of the zeolite to shift. The broader absorption peak of USY at 1054 cm -1 covered the P-O a band and the HPW W = O d band. The W-O b -W band and the W-O c -W band shifted slightly, and their peak intensity was greatly reduced or even disappeared. This indicated that the Keggin structure in HPW/USY basically existed, but the strong interaction between HPW and USY may have deformed the PO 4 tetrahedral structure in the heteropolyanion and weakened W-O bond outside the heteropolyanion structure. Fig. 1(d) shows the IR spectrum of HPW/ZSM-5. The peaks appearing at 1223 cm -1 , 1090 cm -1 , 794 cm -1 , 544 cm -1 , and 448 cm -1 can be ascribed to the typical characteristic peaks of a ZSM-5 molecular sieve framework Shirazi et al., 2010). In the case of HPW/ZSM-5, the characteristic peak of the ZSM-5 framework structure and the HPW Keggin structure can be found with a weak deviation. Therefore, HPW/ZSM-5 not only maintained the porous zeolite structure of ZSM-5, but also the HPW Keggin structure on ZSM-5 was not destroyed, indicating that the interaction between HPW and ZSM-5 was relatively small, which was similar to HPW/SnO2. The combination of the characteristics of HPW and the support was more conducive to the improvement in the deNO x performance. Table 1 lists the specific surface area of the four catalyst systems with the preferred HPW loading. The specific surface area of HPW was only 3.2 m 2 g -1 . For the supports, an obvious difference in specific surface area was observed. The specific surface areas of SiO 2 , SnO 2 , USY, and ZSM-5 were 334.2 m 2 g -1 , 10.4 m 2 g -1 , 559.8 m 2 g -1 , and 305.4 m 2 g -1 , respectively. In general, for adsorption reactions, a larger specific surface area is beneficial in terms of providing more adsorption sites, thereby improving the adsorption performance of the catalyst. Actually, it can be found that for all supports loaded with HPW, as the HPW loading increased, the specific surface area of the sample gradually decreased to varying degrees. This indicated that the HPW was effectively loaded on the supports, and the interaction between the highly dispersed HPW and the supports had an inhibiting effect in specific surface areas.

Adsorption Performance of Heteropoly Acids for NO x
The NO x adsorption performance of four heteropoly acids (HPW, HSiW, HPMo, and HSiMo) was investigated. As shown in the Fig. 2, during the experimental period, relatively stable adsorption rates were observed on the four heteropoly acids. The NO x adsorption rates after a 40 min reaction time with HPW, HSiW, HSiMo, and HPMo were 66.5%, 62.9%, 30.7%, and 24.4%, respectively. Among the four heteropolyacids, HPW showed the best NO x adsorption performance. From the experimental results, the NO x adsorption rates of the W-containing heteropolyacids (HPW and HSiW) were much higher than that of Mo-containing heteropolyacids (HSiMo and HPMo), indicating that the NO x adsorption performance of heteropolyacid could be closely related to the composition of its anions, especially the type of coordination atoms in the anions. This result was consistent with the conclusion of Belanger and Moffat (1995). Indeed, the acid strength of the W-based heteropolyacids was higher than that of the Mo-based heteropolyacids (Jozefowicz et al., 1993). It can be inferred that the NO x adsorption rates of the four heteropolyacids were correlated with their acid strengths.

Effect of HPW Loading on NO x Adsorption for Various Supports
NO x adsorption experiments were carried out on various supports with different HPW loading amounts. The NO x adsorption rate was plotted as a function of the adsorption time reached by each catalyst in the NO-TPO experiment (Fig. 3). These results were derived as shown in Fig. 4 in order to better compare the effect of different HPW loading amounts on the NO x adsorption performance of the catalyst. As shown in Fig. 4(a), the loading amounts of HPW were 0, 16.7%, 28.6%, 37.5%, 44.4%, and 50.0%, for which the corresponding NO x adsorption rates of HPW/SiO 2 were 35.2%, 39.6%, 62.3%, 56.6%, 49.3%, and 50.9%, respectively. It was found that the NO x adsorption performance of pure SiO 2 was very low. After loading with HPW, the NO x adsorption efficiency of the catalytic system increased immediately. When the HPW loading exceeded 30%, the NO x adsorption efficiency of HPW/SiO 2 decreased steadily with increases in the loading, but it was still higher than the performance of the pure SiO 2 . The best deNO x performance was obtained when the HPW loading was 28.6% for the HPW/SiO 2 system. The NO x adsorption rate for HPW/SnO 2 is shown in Fig. 4(b). Pure SnO 2 exhibited an extremely low NO x adsorption rate (13.7%), where after a small amount of HPW was loaded, the NO x adsorption performance increased sharply. The deNO x performance of HPW/SnO 2 basically increased with increases in the amount of loaded HPW, but the performance decreased slightly at 37.5% and then rebounded again. It should be noted that the NO x adsorption rate of HPW/SnO 2 with a 50% HPW loading amount reached 77.3%. Fig. 4(c) shows the NO x adsorption rate of HPW/USY. HPW/USY exhibited a similar trend to that of the HPW/SiO 2 system; that is, the NO x adsorption rate increased first and then decreased with increases in the HPW loading amounts. For HPW/USY, the optimal HPW load was 16.7%. Fig. 4(d) shows the NO x adsorption rates of HPW/ZSM-5 corresponding to HPW loadings of 0, 28.6%, 37.5%, 44.4% and 50% were 27.7%, 57.6%, 62.6%, 65.2% and 70.5%, respectively. The results showed that as the HPW loading amount increased, the NO x adsorption performance of HPW/ZSM-5 increased steadily. This may have been due to the fact that the ZSM-5 is a molecular sieve material with a large specific surface area and a rich pore structure. The heteropoly acid loading can cause the specific surface area and pores to decrease, but it does not lead to the destruction of the structure. Thus, a synergistic effect on the NO x adsorption performance was observed.
Comparing the four sets of data, it was found that the adsorption curve showed two different changes, which was a meaningful for determining the nature of the NO x adsorption. When unloaded supports were used for NO x adsorption, their adsorption rate sequences were USY > SiO 2 > ZSM-5 > SnO 2 , which was consistent with their specific surface area sequence, indicating that a specific surface area plays a certain role in NO x adsorption. When SiO 2 and USY were used as the HPW supports for NO x adsorption, the basic trend in the adsorption performance indicated similar volcanic changes. It can be seen that the highest NO x adsorption rates of HPW/SiO 2 and HPW/USY were 62.4% and 65.6%, respectively, both still slightly lower than that of pure HPW (66.5%), which meant that HPW was the main active component leading to NO x adsorption. According to the BET data, the increased HPW loading on SiO 2 and USY caused a significant decrease in their specific surface area. For HPW/SiO 2 and HPW/USY with the 50% HPW loading, the specific surface area decreased by 59.3% and 56.1%, respectively, compared to the pure supports. The high loading amount made it possible for a significant amount of HPW to enter the support pores, which blocked its porous passages or collapsed the microporous structure, leading to a reduction in the specific surface area. After the highly dispersed HPW was wrapped in the pores of the support, it may have instead reduced the probability of contact between the HPW body phase and reactive molecules, thus limiting the HPW-specific pseudo liquid phase behavior. This may be one reason for the initial increase in the deNO x performance of HPW/SiO 2 and HPW/USY and the subsequent decrease with HPW loading. In addition, the strong interaction between HPW and the surface properties (hydroxyl groups and acid sites) of the support was an important factor affecting the deNO x performance. After loading, the HPW acid amount and acid type remain unchanged, but the acid strength decreased, and the degree to which the acid strength was decreased was also related to the strength of the surface interaction, where a stronger surface interaction led to the lower acid strength. Rocchiccioli-Deltcheff et al. (2010) believed that the effect of HPW and hydroxyl groups on the surface of SiO 2 would cause the acidity and redox of the heteropoly acid to decrease at the same time, thereby leading to reduced catalytic performance. Mccormick et al. (1998) also reported that HPW interacts strongly with the hydroxyl groups on the surface of the SiO 2 support, resulting in a completely different secondary structure from HPW.
For HPW/USY, there was a strong interaction and a weak interaction between the HPW and the silicon hydroxyl group (Si-OH) on the surface of USY, respectively, where the strong interaction consumed H + to produce water, and the weak interaction caused a H + transfer (Liu et al., 2003). Indeed, there were various hydroxyl groups on the surface of the USY channel, including surface Si-OH, skeleton surface Al-OH, and non-framework Al-OH, so the source of acidity of HPW/USY was more complicated. HPW may also interact weakly with the abundant Si-OH in the hydroxyl socket of the USY secondary pore, resulting in a decrease in the acid strength. In addition, the acid-base interaction between the Al species in the USY and HPW may have decreased the acid strength. The decrease in the total acid content on HPW/USY with a high HPW loading may have been due to the decreased surface area of the catalyst, which led to a reduction in the available surface acid centers, thereby reducing the adsorption performance of the catalyst. ZSM-5 and USY are both zeolite molecular sieves with a Si-Al framework structure. However, they have different frameworks, pore structures, and surface properties. It is clear that the specific surface area of HPW/ZSM-5 from pure ZSM-5 to the 50% loading was only decreased by 44.1%, which indicated that the negative effect of increasing the HPW loading amount and the clogging of the carrier pores was less than that in the case of USY. Furthermore, the amount of non-framework Al remaining in the pores of ZSM-5 may have been much less than that in USY (Shen et al., 2007), which reduced the probability of non-framework Al damaging the acidic center of the HPW with a Keggin structure. Thus, the NOx adsorption rate and the trend of the changes in the deNO x performance with the HPW loading amount were obviously different. It is worth noting that SnO 2 was completely different from the other conventional supports in terms of interacting with HPW to participate in NO x adsorption. Conventional porous supports generally rely on their own excellent pore structure and high specific surface area to efficiently disperse HPW active centers to facilitate the entry of NO x molecules into the HPW body. However, the specific surface area of SnO 2 is very small and clearly does not have the properties characteristic of porous materials. In addition, SnO 2 is a semiconductor material with a specific concentration of oxygen vacancies. It has thus been speculated that SnO 2 can be used as a high-efficiency catalytic promoter in the place of ordinary supports. Zhang et al. (2002) studied the desorption and decomposition of NO by HPW/TiO 2 and speculated the trapping effect of oxygen vacancies on the surface of TiO 2 . O 2 is essential in the process of HPW adsorption of NO x . The doping of SnO 2 can increase the O 2 adsorption capacity of the catalyst surface in the NO x adsorption process. Therefore, this combination can demonstrate significantly increased NO x adsorption capacity through synergistic effects that occur between SnO 2 and HPW, where HPW/SnO 2 with the 50% loading had the best deNO x performance.

Adsorption Capacity
Adsorption capacity is an important indicator to evaluate the performance of an adsorbent. The NO x saturation adsorption curves for HPW and HPW/SnO 2 with a 50% HPW loading amount at 150°C were investigated, as shown in Fig. 5. The calculated saturation adsorption capacity of HPW was 50.5 mg g -1 . The NO x saturation adsorption capacity of HPW/SnO 2 was equivalent to 85.4 mg g -1 , which was higher than that of Ti-Zr (28 mg g -1 ) and TiO 2 (46 mg g -1 ) as supports reported in the previous literature (Hodjati et al., 2001;Gómez-García et al., 2005b). It was thus found that the amount of HPW used was reduced, and the stability of the catalyst increased after the HPW was supported on SnO 2 . Furthermore, this combination still maintained excellent NO x adsorption performance.

Adsorption Mechanism
In this experiment, the HPW and HPW/SnO 2 samples after adsorption saturation were characterized using FTIR, as shown in Fig. 6. It can be observed that four characteristic peaks of the Keggin structure still existed in the range of 700-1100 cm -1 , indicating that the two samples maintained a stable Keggin structure after being saturated with NO x . In addition, a new characteristic peak appeared at 2266 cm -1 for the two samples. Chen (1994, 1995) carried out IR characterization of pure HPW saturated with NO x and found an absorption peak at 2270 cm -1 . According to TGA and the calculation of the mass conservation principle of nitrogen, the absorption peak at 2270 cm -1 was regarded as ionic protonated NO (NOH + ). Thus, a classic NO adsorption mechanism was proposed where the adsorbed NO was retained in the bulk HPW phase in the form of NOH + by replacing the crystal water in HPW. Herring et al. (1998Herring et al. ( , 2010 improved the NO adsorption mechanism and confirmed that the protons existed in the secondary structure of heteropoly acid in the form of H 3 O + or H 5 O 2 + . Particularly, protonated water was a prerequisite for NO to enter the HPA secondary structure and exist as NOH + . According to the above analyses, the vibration peak at 2266 cm -1 detected in this experiment could be ascribed to NO + . Furthermore, the characteristic bending vibration peaks of NO 2 + (570 cm -1 ) were not detected. The HPA used in this work was H 3 PW 12 O 40 ⋅6H 2 O. At most, 6 crystal water molecules could be replaced in one HPW molecule, and there was competitive adsorption between H 2 O and NO x . Consequently, it could be inferred that for the HPW/SnO 2 system, the characteristic peak at 2266 cm -1 should also be created by the NOH + vibration entering the HPW secondary structure. The interaction between SnO 2 and HPW was correspondingly small and did not cause changes in the HPW structure. HPW with a pseudo liquid phase acted as the main component for NO x adsorption, and SnO 2 acted as a catalyst assistant. In the presence of O 2 and H 2 O, NO x could be adsorbed on the surface of the HPW or further into its bulk phase to replace part of the crystal water in the secondary HPW structure, where, finally, adsorbed NO x existed in the catalyst in the form of NOH + .

Catalytic Decomposition of Adsorbed NO x
Catalytic decomposition products of adsorbed NO x on HPW and HPW/SnO 2 was detected using GC-MS. N 2 was found in the products, for which the obtained results are shown in Fig. 7. For HPW, N 2 formation was detected 2 min after the start of the increase in the temperature, and a N 2 peak was detected around 8 min later. The decomposition basically ended after 16 min. For HPW/SnO 2 , the detected N 2 reached a peak in about 14 min, and the decomposition was basically completed  in about 24 min. The experimental results showed that the decomposition time of HPW/SnO 2 with the 50% loading was longer than that for HPW, so it can be reasonably concluded that the addition of the SnO 2 support was not only beneficial to the NO x adsorption performance, but the combination also contributed to improving the efficiency of NO conversion, thereby leading to more N 2 conversion. The used catalyst could be regenerated by placing it in humid air (5% water vapor) below 100°C, where HPW can replenish the lattice water replaced by NO x by adsorbing water molecules. Therefore, the new NO x adsorption-decomposition system screened in this work can achieve NO x to N 2 without the use of reducing agents.

CONCLUSIONS
In summary, several new catalyst systems, different HPAs, and supported HPWs (HPW/SiO 2 , HPW/SnO 2 , HPW/USY and HPW/ZSM-5) were prepared and used for the adsorption and decomposition of NO x . The results showed that W-containing HPAs were superior to Mo-containing HPAs. Among the selected catalysts, HPW/SnO 2 with a 50% HPW loading had the highest NO x adsorption rate of 77.3%, and its saturated NO x adsorption amount (85.4 mg g -1 ) was much higher than that of HPW (50.5 mg g -1 ). With HPW as the main active component, the interaction between HPW and the support affected the HPW properties, in turn affecting its NO x adsorption performance. When the interaction was stronger, the adsorption rate of the catalyst tended to become lower. The adsorbed NO x mainly existed in the HPW bulk phase in the form of NOH + . The GC-MS detection confirmed that the adsorbed NO x can be decomposed to N 2 using a temperature programming process. This study provides a new experimental basis for which to find a new multi-functional NO x catalytic decomposition system.