Study on volatilization characteristic of vanadium and tungsten of vanadium-based selective catalytic reduction

. Vanadium-based selective catalyst reduction (V-SCR) has made significant contributions to De-NOx of diesel vehicles. However, the two main catalyst components, V 2 O 5 and WO 3 , are volatile and toxic. Therefore, there are some restrictions for its application on heavy duty diesel vehicles (HDDV) in the CHINA VI stage. In view of this problem, the volatilization rates of two kinds of volatiles under different inlet temperature and test duration were measured on three V-SCR samples from different catalyst manufacturers. The effects of V-SCR inlet temperature and test duration on the volatilization characteristics of the two volatiles on different samples were systematically studied and compared. The emission concentrations of V 2 O 5 and WO 3 at different temperatures were compared with the occupational exposure limits of chemical harmful factors in the workplace which is specified in the GBZ 2.1-2019. The results show that the initial temperature of volatilization of V-SCR is directly related to the technical level of the catalyst itself; whether the volatilization characteristics of V-SCR will change with the prolongation of test duration depends on its inlet temperature; V 2 O 5 emission concentration of V-SCR at 550 ℃ can reach a quite high exposure level, and the exposure level of WO 3 volatiles at 650 ℃ may be equivalent to that of V 2 O 5 volatiles at 550 ℃ .


Introduction
Selective catalytic reduction (SCR) has become the most widely used commercial diesel vehicle De-NOx technology, especially in China and Europe [1,2].The SCR technology used in the IV and V emission stages of heavy-duty diesel vehicles (HDDV) in Europe and China is V-W-Ti based SCR (V-SCR) with V2O5 as catalyst active substance, WO3 as catalyst assistant and TiO2 as coating material.However, V2O5 and WO3 are volatile, and the volatilized gas is toxic, among which V2O5 is identified as suspicious carcinogenic to human by the national occupational health standard GBZ 2.1-2019 [3].
In China and Europe, in the IV and V emission stages of HDDV, V-SCR is the unique aftertreatment device with the usual inlet temperature lower than 400 ℃ and unlikely to produce volatiles, so the application of V-SCR is relatively safe.From July 1, 2019, China began to implement the CHINA VI emission standard of HDDV [4], in which the particulate matter emission limit was tightened.In order to meet the regulatory requirements, the current prevailing technology is to install a diesel particle filter (DPF) in front of SCR, and then burn and remove the accumulated PM through active or passive regeneration.If the DPF adopts the active regeneration mode, the exhaust temperature may rise to 750 ℃ instantaneously, while exceeding the melting point of V2O5 and causing volatilization.Therefore, the new standard [4] puts forward special requirements for V-SCR that the inlet temperature lower than 550 ℃.If V-SCR technology is applied, the upstream DPF must adopt mild passive regeneration mode and verify the volatilization characteristics of vanadium volatiles.As for WO3, although its toxicity level is relatively low, some researchers also showed interests and studied it with V2O5 synchronous as well, and found that its volatilization rate was higher than V2O5 [2,[5][6][7].
With regard to the influencing factors of volatilization characteristics of V-SCR, Yannopoulos et al. [8] found that the main way of volatilization of V2O5 and WO3 is not by heat, but by the formation of volatile hydroxyl compounds with water in the exhaust gas.The existence of water can promote the volatilization and decrease the volatilization temperature.Liu et al. [7] showed that the technical level and aging degree of the catalyst itself can also lead to different volatilization characteristics.However, at present, there are only a little of data from foreign laboratories, which mainly focused on the high-temperature effect, and the test conditions such as water vapor content and the evaluation device are different.Thus, it is not convenient for data comparison.On the other hand, with the upgrading of emission regulations, the technical level of V-SCR has being improved, and the samples studied in these public papers cannot represent the technical level of V-SCR newly in China, so it is worthy to study the V2O5 and WO3 volatilization characteristics of state of the art V-SCR products used in domestic market.In view of this, three V-SCR samples (the technical level differences between them are unknow) from different catalyst manufacturers were tested in a simulated operating environment with a fixed water vapor concentration of 10 vol.% (about the maximum concentration of water vapor in the exhaust of diesel vehicles), and the volatilization rate of V2O5 and WO3 at different inlet temperatures with different test duration were tested.The effects of V-SCR inlet temperature and test duration on the volatilization characteristics of the two volatiles on different samples were systematically studied and compared.The emission concentrations of V2O5 and WO3 at different temperatures were compared with the occupational exposure limits of chemical harmful factors in the workplace which is specified in the GBZ 2.1-2019 [3].

The capture of Volatiles
The three V-SCR samples were named as sample A, B and C separately.Take 6 small samples from each V-SCR sample for testing at different temperature points separately, the size of small sample was about Φ 20 mm × 25 mm.Each small sample was put in the constant temperature zone of the quartz tube of the reactor.Two quartz capture traps contained a specific amount of γ-Al2O3 (7.0 g, particle diameter 2 -4 mm) as the adsorbent of volatiles were put in the constant temperature zone and the low temperature zone of the quartz tube separately.Use thermocouple to measure the inlet temperature of the sample, and the measuring point was at the center line and 1 cm before the front end of the sample.The Schematic of volatiles capture test is shown in Figure 1.During all the experiments, a space velocity (SV) of 35 000 h -1 was used.To minimize the volatilization in the process of transient temperature ramps, 5 vol.%O2 and balance N2 were introduced into the quartz tube, and the furnace temperature was programmed heating until the inlet temperature of small sample reached the desired temperature within 1 h.When the inlet temperature reached the desired point, 10 vol.% water vapor, 250 ppm NO and 250 ppm NH3 were added immediately, meanwhile the test duration begun to calculate.It was not until reaching the target duration that the heating was stopped and the water supply with all the gases except N2 was cut down.N2 was used to cool the reactor rapidly to below 100 ℃.Then take the two quartz capture traps out to analyze, and replace the fresh adsorbent to proceed the next testing stage.

Analysis method of adsorbent
Take out the adsorbent and then grind them into powder (particle size ≤ 0.15 mm) separately.Three test portions of 0.2 g for each test sample were transferred respectively to the Teflon pressure vessels with a mixture of 12 ml of HCl, 2 ml of HNO3, 0.5 ml of HF.The samples were heated in a microwave digestion instrument at 200 ℃ for 25 min.After cooling, the test solutions were diluted to 50 mL with high purity water.The concentrations of V, W and Ti were determined by Inductively Coupled Plasma -Optical Emission Spectroscopy (ICP-OES) from Perkin Elmer.Ti served as an internal marker for catalyst dust spill over.It is required that the relative standard deviation of each element in three parallel test portions should not exceed 5 %.

Calculation of minimum quantitative detection limit
According to HJ 168-2010 [9], the method detection limit (MDL) of 4 times was used as the minimum quantitative detection limit.The MDL calculation method is according to the method of "the target substance is detected in the blank test", fresh γ-Al2O3 from the same batch as the tests was taken as the blank sample and was determined 10 times in parallel according to all the steps for adsorbent analysis.MDL was calculated using Equation 1.
MDL is the method detection limit, μg/g; t (n-1, 0.99) is the t distribution (one side) when the freedom degree is 10-1 and the confidence is 99%, and the reference value is 2.821; S is the standard deviation of parallel determination results for 10 times, μg/g.

Volatilization rate
V2O5 and WO3 volatilization rates were calculated respectively using Equations 2 and 3. for each V-SCR sample, V is the volume of V-SCR small sample, t is the duration hours at the specified temperature.

𝑄𝑄𝑄𝑄 𝑉𝑉𝑉𝑉2𝑂𝑂𝑂𝑂5 = (
If V, W and Ti elements are detected in the adsorbent of cold trap, the results should be calculated in the same way and included in the volatilization rate.The volatilization rate of V2O5 and WO3 obtained from cold trap adsorbent should not exceed 5 % of that of corresponding hot trap adsorbent respectively.

Emission concentration
V2O5 and WO3 emission concentrations was calculated using Equation 4.
Ri is the emission concentration of V2O5 or WO3, μg/m 3 , i represents V2O5 or WO3; Qi is volatilization rate of V2O5 or WO, μg/(L•h), i represents V2O5 or WO3; SV is the space velocity fixed at 35 000 h -1 .

Effect of inlet temperature on volatilization
The inlet temperature of V-SCR for CHINA VI should be lower than 550 ℃ [4].In order to verify the influence of inlet temperature on volatilization rate, the volatilization rates of the two volatiles of three samples were tested at six inlet temperature points between 400 ℃ and 700 ℃ for 18 h.The comparison results are shown in Figure 2. It can be seen from Figure 2 that the initial temperature of volatilization for the three samples were slightly different.Thereinto, a small amount of volatilization of V2O5 and WO3 occurred early at 500 ℃ for sample B and C, and then at 550 ℃ for sample A. For the same sample, the initial temperatures of volatilization of V2O5 and WO3 were the same.As for the trend of the curve, it can be seen that the V2O5 and WO3 volatilization rates of the three samples turned out an approximate exponential growth with the rise of temperature, which is different from the stable linear growth trend that observed by Liu et al. [7] in the temperature range of 500 ℃ ~ 700 ℃ on reactor with 10.5 vol.% water vapor during 22h.Compared with the volatilization rate difference between samples, it can be found that the volatilization rates of the two volatiles on sample B were higher than those of samples A and C at all the testing temperature points, indicating that sample B had stronger volatility.In addition, The HDDV emission standard of CHINA VI requires that there should be no leakage of vanadium containing compounds in the whole life cycle of HDDV installed with V-SCR.The test requirement is that the volatilization of vanadium in the adsorbent should be lower than the minimum quantitative detection limit when the inlet temperature of V-SCR is kept at 550 ℃ for 18 h, and the matching ammonia slip catalyst is permitted to be at the back end of V-SCR as a whole sample.However, it is observed that all three samples at 550 ℃ produced the vanadium volatiles.Therefore, it can be seen that in order to meet the requirements of the regulations, it is necessary to further improve the technical level of the products so as to inhibit volatilization, or simultaneously use the capture measures of volatiles to prevent their leakage.For example, Hoej et al. [2] found that ammonia slip catalyst had an excellent adsorption effect on volatiles.

Effect of test duration on volatilization
At the constant inlet temperature, prolonging the duration can accumulate the heat effect and aggravate the catalyst aging degree of V-SCR, which may promote the volatilization of volatiles.T/CAEPI 12.2-2017 [10], a standard on technical requirements for environmental protection products, puts forward that the durability test requirement of V-SCR is aging at 550 ℃ for 200 h.In order to verify the variation of volatilization rate of the two volatiles during the duration, the two volatiles volatilization rates of three samples at 550 ℃ in 0-18 h, 19-50 h, 51-100 h and 101-200 h (18 h, 50 h, 100 h, 200 h for short), four separate stages, were tested respectively.The results are shown in Figure 3.It can be seen from Figure 3 that the volatilization rates of V2O5 and WO3 of the three samples at 550 ℃ had no obvious variation in the four test duration stages, which was consistent with the phenomenon observed by Hoej et al. [2] from the volatilization performance of a V-SCR sample within 200 h at the same temperature.The results show that the samples still kept good stability even if the test duration was extended to 200 h at 550 ℃, and the heat accumulated during the test had no obvious influence on the volatilization rate of volatiles.In order to compare the effect of test duration on the volatilization rate at higher temperature, the volatilization rates of the two volatiles in 18 h and 50 h at 600 ℃ were tested.The results are shown in Figure 4.The phenomena in Figure 4 are slightly different from that in Figure 3.At 600 ℃, only the volatilization rates of two volatiles of sample A had no significant variation in the two duration stages, while the volatilization rates of V2O5 and WO3 of sample B and sample C in the second stage were significantly raised compared with those in the first stage, with the corresponding increases of 114 % and 117 % for sample B and 64 % and 60 % for sample C respectively.It can be inferred that the order of hydrothermal stability of the three samples from high to low was A > C > B at 600 ℃.At this temperature, sample B and sample C showed obvious deterioration, which promoted the volatilization of V2O5 and WO3.Furthermore, it is observed that the volatilization rates of V2O5 and WO3 of sample B and sample C were basically the same in different duration stages, and the volatilization rate ratio of the two volatiles did not change significantly, indicating that prolonging the test duration or aggravating the aging degree of the catalyst had no significant effect on the volatilization rate ratio of the two volatiles.

Comparison of volatilization characteristics between V2O5 and WO3
Generally, V2O5 loading is 1 wt.% ~ 3 wt.% in V-SCR catalyst coating, while WO3 loading range is wider, which can be 1 wt.% ~ 15 wt.%, however, its loading is higher than V2O5 for the most part.The loading ratios of WO3 to V2O5 corresponding to the three samples is in the order of A > B > C. In order to compare the real volatilization characteristics of the two volatiles and exclude the influence of loading's difference on volatilization rate, the volatilization rate ratio of WO3 and V2O5 corresponding to the same loading was calculated according to Equation 5, and the results were shown in Figure 5.

𝑌𝑌𝑌𝑌 = 𝑄𝑄𝑄𝑄 𝑊𝑊𝑊𝑊𝑂𝑂𝑂𝑂3
: • 25 (5) Y is the volatilization rate ratio of WO3 to V2O5 corresponding to the equal loading; PW:V is the load mass ratio of WO3 to V2O5 on V-SCR.It can be seen from Figure 5 that the volatilization rate ratio of WO3 to V2O5 corresponding to the equal loading of the three samples ranged from 0.4 to 5.5, and increased with the rise of temperature.The volatilization rate ratio of WO3 to V2O5 corresponding to the equal loading was less than 1.0 for sample A at 550 ℃ and 600 ℃, and samples B and C at 500 ℃ separately, which indicates that the actual volatility of WO3 was slightly weaker than that of V2O5 under the specific condition.With the increase of temperature, the volatilization rate ratio of WO3 to V2O5 raised, which indicates that the increase of temperature has a greater promotion effect on the volatility of WO3 than that of V2O5.

Comparison of volatiles' concentrations in exhaust with their occupational exposure limits (OELs) in workplace
In order to evaluate the harmful effects degree of V2O5 and WO3 volatiles from V-SCR if discharge them directly into the respiratory zone of human without being adsorbed, the OELs and exposure levels for chemical hazardous factors in the workplace, which were constituted on the basis of the research results of toxicology and epidemiology and regulated in the mandatory occupational health standard GBZ 2.1-2019 [3] issued by the National Health Commission of China was referred to.And the V2O5 and WO3 emission concentrations of V-SCR (assuming that the space is full of diesel exhaust) can be compared with the OELs of V2O5 and WO3 to evaluate the exposure level.GBZ 2.1-2019 [3] divides the exposure level into five grades (Table 1).adverse health effect is respiratory system damage; the PC-TWA of W and its insoluble compounds (WO3 belongs to this category) is 5 000 μg/m 3 according to W and equivalent to 6 304 μg/m 3 of WO3, and its critical adverse health effect is lower respiratory tract stimulation.
The comparison results of ratio of V2O5 and WO3 emission concentrations of the three V-SCR samples within 18 h at 500 ℃ -700 ℃ with the corresponding PC-TWA in the workplace (named as exposure ratio) are shown in Figure 6.
As shown in Figure 6, at the same temperature, the exposure ratio of V2O5 was about two orders of magnitude higher than that of WO3.According to the description of exposure level in Table 1, it can be seen that, for V2O5 at 500 ℃, the exposure level of volatiles reached level II, the level with contact but no obvious health effect; at 550 ℃, sample B reached level III of significant exposure; at 600 ℃, except for sample C which still kept the exposure level of grade III, all others exceeded PC-TWA and reached level IV.Compared with V2O5, the exposure level of WO3 is relatively low.At the temperature below 650 ℃, the exposure level of WO3 volatiles was about two orders of magnitude lower than that of V2O5.It's noteworthy that at 650 ℃, the WO3 exposure level of samples A and C reached level II while sample B even reached level III, and the exposure level of WO3 of three samples was equivalent to that of V2O5 at 550 ℃ actually.

Summary and conclusions
Based on the test data of the samples investigated in this paper, the following conclusions can be drawn: (1) The initial volatilization temperature of V-SCR is directly related to the technical level of the catalyst itself.V2O5 and WO3 begin to volatilize at the same temperature on one V-SCR.
(2) The V2O5 and WO3 volatilization rates both have an approximate exponential growth with the rise of inlet temperature of V-SCR.
(3) Whether the volatilization characteristics of V-SCR will change with the prolongation of test duration depends on its inlet temperature.When the temperature is below 550 ℃, the extension of test duration has no obvious effect on the volatilization characteristics of V-SCR.However, when the temperature is over 550 ℃, the extension of test duration will lead to the simultaneous increase of V2O5 and WO3 volatilization of V-SCR with poor hydrothermal stability.
(4) The results of comparison of the volatility of V2O5 and WO3 are related to the inlet temperature of V-SCR.At the temperature near the initial volatilization temperature, the
QV2O5 and QWO3 are the V2O5 and WO3 volatilization rate respectively, μg/(L•h), CV, CW and CTi are the concentration of V, W and Ti found in the absorbent respectively, μg/g, m is the mass of absorbent in the hot trap, PV:Ti and PW:Ti are the loading ratio of V2O5 and WO3

Fig. 2 .
Fig. 2. Volatilization rate of V2O5 and WO3 within 18 h as a function of inlet temperature.

Fig. 5 .
Fig. 5. Volatilization rate ratio of WO3 to V2O5 corresponding to the equal loading as a function of temperature.

Table 1 .
[3]cription of occupational exposure levels OELs is the permissible concentration-time weighted average (PC-TWA), which means the allowable average exposure concentration for 8 h working day and 40 h working week.GBZ 2.1-2019[3]specifies the PC-TWA of V2O5 is 50 μg/m 3 , and its critical