Insight into Platinum Poisoning Effect on Cu-SSZ-13 in Selective Catalytic Reduction of NO x with NH 3

: Platinum’s (Pt) poisoning effect on Cu-SSZ-13 and its regeneration were investigated. The Pt enhanced the parallel reactions, such as NH 3 oxidation and NO oxidation reactions, which decreased the deNOx activities. In the temperature range below 330 ◦ C, the deactivation of Cu-SSZ-13 by Pt poisoning was primarily caused by the overconsumption of NH 3 , due to the enhanced NH 3 -selective oxidation reaction, while the formation of NO x in NH 3 oxidation and NO oxidation into NO 2 further aggravated the degradation when the temperature was above 460 ◦ C. The non-selective NH 3 oxidation and non-selective NO x catalytic reduction reactions resulted in increased N 2 O formation over Pt-doped samples. The transformation of Pt 0 into PtO x after hydrothermal aging recovered the deNO x activities of the Pt-poisoned samples.


Introduction
Cu-exchanged SSZ-13 has recently been commercialized as a catalyst for the selective catalytic reduction of NO x with a NH 3 (NH 3 -SCR) section in a diesel engine exhaust aftertreatment system [1]. For automotive applications, there is a critical requirement for the application life of the SCR catalyst in order to meet the NOx emission standard [2].
Specific to hydrothermal treatment, the deactivation of Cu-SSZ-13 is typically attributed to the destruction and transformation of cupric sites from SCR-active to -inactive forms, which is irreversible [5,9]. Nevertheless, the effect of chemical poisoning on Cu-SSZ-13 is more complex than that of sulfur poisoning and hydrothermal treatment. In addition to the poisoning effect of the destruction and transformation of isolated Cu 2+ ions into inactive forms [6,[10][11][12], the additional chemical contaminants may act as active sites that catalyze side reactions in the NH 3 -SCR reaction system.
In general, the standard SCR reaction is the target pathway for the NO x elimination in the NH 3 -SCR system, as shown in Reaction 1 (R.1): However, some parallel reactions also occur along with the standard reaction, as shown in R. 2 Without the contaminants, the rate of the standard NH 3 -SCR reaction is much higher than that of the side reactions when using the Cu-SSZ-13 as catalyst, because the isolated Cu 2+ ions primarily catalyze the standard reaction [13,14]. Therefore, the Cu-SSZ-13 exhibited excellent deNOx activity. However, the presence of additional chemical contamination sites might help to catalyze the reactions in R.2-R.7, which increases the parallel reactions rate [6]. As such, the competition between the parallel reactions (R.2-R.7) and the standard reaction (R.1) would reduce the SCR activity of Cu-SSZ-13 (R.1) [6].
Pt is a typical chemical contamination species for Cu-SSZ-13 [6]. In an after-treatment system, the SCR catalyst is located downstream of the diesel oxidation catalyst (DOC) and diesel particulate filter (DPF) sections [15]. The Pt coating in DOC and DPF would be volatilized and deposited on the SCR catalyst due to the high exhaust temperature, which significantly decreases the deNOx activities of the SCR catalysts [16][17][18]. The previous work in the industry investigated the Pt poisoning effect on Cu-SSZ-13 with Pt loading in the weight range of 0-0.016 wt. % [19]. This work found that 0.008 wt. % was the threshold with little influence on the activity of Cu-SSZ-13, above which the activity decreased significantly. Lezcano-Gonzalez et al. [6] investigated the effect of series pollutants (Pt, Zn, Ca and P) on Cu/SSZ-13 catalysts during the NH 3 -SCR process, and Pt poisoning (with Pt loading of 1-2 wt. %) resulted in the NOx conversion reduction, which was predicted to be due to the Pt species' oxidation ability that promoted the NH 3 oxidation reaction (R.3-R.6). A similar poisoning effect of Pt was also observed on Cu-SAPO-34 [18]. It should be noted that the amount of Pt has a significant impact on the SCR performance of Cu-SSZ-13 [19]. Therefore, the effect of a Pt loading of 0.01-0.1 wt. % on Cu-SSZ-13 should be investigated. In addition, the hydrothermal treatment effect on the Pt poisoning influence on Pt/Cu-SSZ-13 has also not been reported.
In this work, a series of Pt/Cu-SSZ-13 samples with different Pt doping contents (0.01-0.1 wt. %) was prepared to elucidate the poisoning impact of Pt on the NH 3 -SCR performance of the catalysts. The Pt's influence on the NH 3 -SCR reaction network was investigated based on transient reaction experiments. The Pt poisoning effects on the structure and cupric sites of Cu-SSZ-13 were evaluated by XRD and H 2 -TPR measurements. The effect of hydrothermal treatment on the recovery of a Pt-contaminated sample was analyzed.

Results and Discussion
2.1. Influence of Pt Poisoning on the Catalytic Performance of Cu-SSZ-13 Figure 1 shows the SCR performance of the Cu-SSZ-13 and Pt poisoning samples with and without hydrothermal treatment. The Cu-SSZ-13 presented above 75% NOx conversion at 150-550 • C, and the N 2 O concentration was below 20 ppm, indicating that the NO and NH 3 were primarily converted via the standard NH 3 -SCR reaction (R.1) pathway over Cu-SSZ-13. For Cu-SSZ-13 doping with Pt (PtxCu), the NOx conversion decreased with increasing Pt loading, particularly in the temperature range above 300 • C. Notably, larger concentrations of N 2 O were formed in PtxCu samples than in Cu-SSZ-13 ( Figure 2). This indicates that the parallel reactions (e.g., R.2-R.7) should be accelerated by Pt sites [18,20]. In addition, the N 2 O formation shows a volcanic curve, with the peaks at around 330 • C in the Pt-doped samples. This demonstrates that the parallel reactions (such as R.2 and R.4) contributing to the N 2 O formation mainly affect the NO x conversion below 330 • C. In other words, some other parallel reactions resulted in a decrease in deNO x activity above 330 • C. This part will be further analyzed in Section 3.2, accompanied by the transient reaction results. In addition, in our work, the NO x conversion decreased with the increasing Pt loading from 0.01 to 0.1 wt. % ( Figure 1). However, in the work by Lezcano-Gonzalez et al. [6], the Pt-poisoned Cu-SSZ-13 showed a similar loss of NOx conversion in the Cu-SSZ-13 with Pt loadings of 1 and 2 wt. %. This indicates that the Pt poisoning effect might not change significantly when the Pt loading reaches a high content (around 1 wt.%).  The NO x conversion of Cu-SSZ-13 decreased after hydrothermal treatment. However, the hydrothermally aged PtxCu samples showed only slightly inferior NO x conversions as compared to the Cu-H sample, which was much higher than its fresh counterparts, as shown in Figures 1 and 2. In addition, the N 2 O formation over the PtxCu samples decreased after hydrothermal treatment. This demonstrates that the hydrothermal treatment partially recovered the activity of Pt-poisoned Cu-SSZ-13.

Transient Reactions
To investigate the effect of Pt on the standard NH 3 -SCR reaction and parallel reactions with Cu-SSZ-13 below and above 330 • C, the transient reactions were conducted, and the results are given in Table 1. The procedure of the transient reactions is shown in Figure 3. The NH 3 oxidation results in Figure S1 reveal that the Pt poisoning enhanced the NH 3 oxidation capacity of PtxCu samples. However, small amounts of NO and NO 2 were formed on the Cu-SSZ-13 and Pt-poisoned samples during the NH 3 oxidation period at 330 • C, as shown in Table 1. This indicates that the Pt sites mainly enhanced the NH 3 selectivity of the oxidation reaction to N 2 (R.3), which led to insufficient NH 3 taking part in the standard SCR reaction, and decreased the NO x conversion of PtxCu samples ( Figure 1) below 330 • C as a result. Both the non-selective NH 3 oxidation reaction (NSNO, R.4) and the non-selective NO x catalytic reduction reaction (NSCR, R.2) contributed to the N 2 O formation [21]. The larger content of N 2 O in the PtxCu sample than Cu-SSZ-13 ( Figure 2) must be attributed to the enhancement of the NSNO and NSCR reaction rates. However, the N 2 O formation cannot be attributed to the enhancement of the non-selective NH 3 oxidation reaction (NSNO, R.4), because only a low concentration of N 2 O was formed in the NH 3 oxidation period at 330 • C, as shown in Table 1. After the introduction of NO, the N 2 O concentration was significantly increased (Table 1). This indicates that the N 2 O formation was primarily attributed to the non-selective NO x catalytic reduction reaction (NSCR, R.2), which was catalyzed by Pt sites. More NO was formed during the NH 3 oxidation at 460 • C (R.6) than at 330 • C. This indicates that, in addition to the decrease in reductant involved in the standard SCR reaction due to the enhancement of the NH 3 oxidation reaction by Pt, the oxidation of NH 3 to NO (R.6) also contributed to the decrease in NO x conversion at high temperatures (460 • C).
Besides this, at 460 • C, a large amount of NO 2 was also produced during the NH 3 oxidation of Pt doping samples, while none was observed in Cu-SSZ-13 (Table 1), indicating that a part of the NH 3 was oxidized to NO 2 by Pt at 460 • C. Meanwhile, NO oxidation into NO 2 (R.7) was also enhanced in the PtxCu samples, as confirmed by the higher concentration of NO 2 formed over PtxCu than Cu-SSZ-13 during the NO oxidation period (Table 1). Furthermore, the total concentration of NO 2 formed in the NH 3 oxidation and NO oxidation reactions was similar to that formed in the NH 3 -SCR period. This demonstrates that Pt loading accelerated the reactions in R.6 and R.7, resulting in the increased NO 2 formation in the NH 3 -SCR process. Those observations above imply that, beside the overconsumption of NH 3 following R.1, the additional NO and NO 2 production via NH 3 and NO oxidation (R.5-R.7) also led to the decline in NO x conversion in the PtxCu samples at high temperatures (around 460 • C) [22,23].

Pt Poisoning Effect on the Texture and Structure
The ICP results of the samples are given in Table 2. The fresh and aged samples showed similar Cu loading contents, and the Pt amount increased with the increment in the Pt loading. The hydrothermal treatment did not change the Pt content. The surface area and pore volume decreased with the increasing Pt loading (Table 2). No significant framework dealumination took place as a result of Pt impregnation, as only a slight decline of the peak at 58 ppm, attributed to the tetrahedral Al species, was observed in the Pt-doped samples compared to Cu-SSZ-13 ( Figure S4) [14,24]. In addition, no significant formation of octahedrally coordinated Al species (showing a resonance peak at 0 ppm) could be detected [14,24]. These observations above indicate that the dealumination should not be the primary reason for the decline in BET surface area. Actually, SSZ-13 possesses small windows of about 3.8 Å, so mass transfer limitations might be induced due to the deposition of impurities, even at low quantities. Moreover, taking into account the diameter of Pt 0 and PtO x (around 1 nm, shown in the TEM results in Figure S6), it is reasonable to assume that some of the Pt might be located in the zeolite micropores, causing a decrease in the micropore volume and the BET surface area. However, those values were largely maintained after hydrothermal treatment, suggesting that the structure of the hydrothermally aged sample remains intact, consistent with the XRD results in Figure S2.

H 2 -TPR
The isolated Cu 2+ ions are identified as the active sites that catalyze the standard NH 3 -SCR reaction [25][26][27]. H 2 -TPR was carried out on the catalysts to detect the state of cupric species, and the results are shown in Figure 4a. For the fresh samples, two peaks in 220-380 • C and in 600-750 • C were observed, representing the reduction of isolated Cu 2+ to Cu + (R.8) and the reduction of Cu + to Cu 0 (R.9), respectively [28,29].
Catalysts 2021, 11, 796 7 of 11 380 °C and in 600-750 °C were observed, representing the reduction of isolated Cu 2+ to Cu + (R.8) and the reduction of Cu + to Cu 0 (R.9), respectively [28,29].  After deconvolution, the peak in the 220-380 °C range can be divided into two peaks with tops at 250 °C and 325 °C, assigned to the reduction of isolated Cu 2+ located in eightand six-member rings, respectively [30]. After the hydrothermal treatment, the reduction peak below 500 °C was widened, and an additional peak at 190 °C appeared. This can be attributed to the transformation of a part of the isolated Cu 2+ ions to CuO during the hydrothermal treatment [31].
The isolated Cu 2+ ions content of each sample was normalized using the area of the reduction peaks of isolated Cu 2+ ions in Cu-SSZ-13 as a reference, shown in Figure 4b. The Cu-SSZ-13 and PtxCu samples contained similar isolated Cu 2+ ions contents. This suggests that the Pt poisoning has no influence on the cupric sites, which is consistent with previ- After deconvolution, the peak in the 220-380 • C range can be divided into two peaks with tops at 250 • C and 325 • C, assigned to the reduction of isolated Cu 2+ located in eightand six-member rings, respectively [30]. After the hydrothermal treatment, the reduction peak below 500 • C was widened, and an additional peak at 190 • C appeared. This can be attributed to the transformation of a part of the isolated Cu 2+ ions to CuO during the hydrothermal treatment [31].
The isolated Cu 2+ ions content of each sample was normalized using the area of the reduction peaks of isolated Cu 2+ ions in Cu-SSZ-13 as a reference, shown in Figure 4b. The Cu-SSZ-13 and PtxCu samples contained similar isolated Cu 2+ ions contents. This suggests that the Pt poisoning has no influence on the cupric sites, which is consistent with previous work [6]. Therefore, the parallel reactions should exclusively be catalyzed by the Pt species, which reduced the SCR activities. Furthermore, the SCR active sites, i.e., isolated Cu 2+ ions, decreased, and the Pt content remained constant (Figure 4b), while the SCR activity was increased for the Pt x Cu samples after hydrothermal treatment. This indicates that the hydrothermal conditions might change the state of Pt, which plays a critical role in determining the SCR's performance.

X-ray Photoelectron Spectroscopy
To detect the changes of state in Pt, the XPS technique was utilized, and the Pt 4f spectra of the fresh and aged samples are given in Figure 5a,b, respectively. The Cu-SSZ-13 sample displayed no Pt signals, and a broad peak at 70-80 eV was observed in the PtxCu samples. The signals for the Pt 4f 7/2 peaks with a binding energy value at 71.32 eV confirmed the presence of platinum in the zero oxidation state (Pt 0 ) [32]. The Pt 4f 7/2 signals at around 72.96 and 74.01 are attributed to PtO and PtO 2 species, respectively, which were formed due to the Pt 0 oxidation during the catalyst's calcination in preparation [33]. In comparison, only reduced Pt was observed in the previous work on Pt-poisoned Cu-SSZ-13 [6]. This might be attributed to the different Pt precursors used in the work by Lezcano-Gonzalez et al. ([Pt(NH 3 ) 4 ](NO 3 ) 2 ) and in our work (Pt(NO 3 ) 2 ), as the different precursors have an influence on the state of the Pt [6,34]. After the hydrothermal treatment, the XPS spectra of the hydrothermally aged samples changed. Only the PtO and PtO 2 species remained in the aged samples. This indicates that all the Pt 0 was oxidized during the aging process, which is consistent with the pervious works that reported that the Pt 0 would be easily oxidized at high temperatures [16,18].
Catalysts 2021, 11,796 8 o spectra of the hydrothermally aged samples changed. Only the PtO and PtO2 species mained in the aged samples. This indicates that all the Pt 0 was oxidized during the ag process, which is consistent with the pervious works that reported that the Pt 0 would easily oxidized at high temperatures [16,18]. The Pt 0 content of each sample was calculated, as shown in Figure 5c (details sho in Figure S3). The Pt 0 content increased in the order of Pt3Cu > Pt2Cu > Pt1Cu > Cu-SSZ consistent with the extent of the deactivation of the Pt-loaded samples ( Figure 1). C bined with the transient reaction results in Table 1, it is suggested that the Pt 0 cataly the R.2-7, decreasing the deNOx activities of the PtxCu samples ( Figure 1). In addit the aged Cu-SSZ-13 and PtxCu exhibited similar SCR activities after hydrothermal tr ment, although a large content of PtOx still existed in the PtxCu-H samples (Figures 1  5b). This demonstrates that the PtOx is ineffective in catalyzing the parallel reactions other words, the transformation of Pt 0 into PtOx results in the regeneration of PtxCu s ples after hydrothermal treatment.

Catalyst Preparation
The Cu-SSZ-13 was synthesized via one-pot hydrothermal synthesis using cop tetraethylenepentamine (Cu-TEPA, Guangfu, Tanjin) as a structure-directing agent. In The Pt 0 content of each sample was calculated, as shown in Figure 5c (details shown in Figure S3). The Pt 0 content increased in the order of Pt 3 Cu > Pt 2 Cu > Pt 1 Cu > Cu-SSZ-13, consistent with the extent of the deactivation of the Pt-loaded samples (Figure 1). Combined with the transient reaction results in Table 1, it is suggested that the Pt 0 catalyzed the R.2-7, decreasing the deNO x activities of the PtxCu samples ( Figure 1). In addition, the aged Cu-SSZ-13 and PtxCu exhibited similar SCR activities after hydrothermal treatment, although a large content of PtO x still existed in the PtxCu-H samples (Figures 1 and 5b). This demonstrates that the PtO x is ineffective in catalyzing the parallel reactions. In other words, the transformation of Pt 0 into PtO x results in the regeneration of PtxCu samples after hydrothermal treatment.

Catalyst Preparation
The Cu-SSZ-13 was synthesized via one-pot hydrothermal synthesis using coppertetraethylenepentamine (Cu-TEPA, Guangfu, Tanjin) as a structure-directing agent. In total, 180 mL of gel was prepared with a molar ratio of 2.5 Na 2 O:1 Al 2 O 3 :10 SiO 2 :150 H 2 O:2 Cu-TEPA, which was then transferred into a 250 mL autoclave with a Teflon liner and crystallized at 140 • C for 4 days. The solid product was filtrated, washed with water, and then dried at 120 • C for 6 h. Then, 15 g of as-synthesized sample was ion-exchanged with 300 mL NH 4 NO 3 solution (1 mol/L) at 80 • C and kept for 9 h to remove excess Cu. After filtration, washing, and drying at 100 • C, the sample was calcinated in synthetic air at 600 • C for 6 h to obtain Cu-SSZ-13. The Si/Al ratio of Cu-SSZ-13 is 4, as determined by ICP-OES. The details of the preparation can be found in our previous work [14].
All catalysts were treated at 750 • C in air with 10% H 2 O for 12 h. The flow rate was kept at 500 mL/min. The hydrothermally aged samples with Pt loading were distinguished with the suffix of (-H), and Cu-H was used as the abbreviation for the aged Cu-SSZ-13.

Catalytic Performance Test
The NH 3 -SCR activities of the catalysts were tested in a fixed-bed quartz reactor (10 mm in diameter). Typically, 100 mg of the catalyst (40-60 mesh) was used, which was pre-activated in 5% O 2 /N 2 flow at 550 • C for 1 h. The reactant gases contain 500 ppm NO, 500 ppm NH 3 , 5 vol.% O 2 , and N 2 as balance. The gas flow was kept at 500 mL/min, and the corresponding GHSV was 150,000 h −1 . The concentrations of the products (NO, N 2 O and NO 2 ) were recorded on an FTIR gases analyzer (Thermo Nicolet iS10, Madison, WI, USA) with a 2 m gas cell when the reaction reached a steady state (less than 1 ppm change in ten minutes). The NO x (NO x = NO + NO 2 ) conversion was calculated according to R.10:

Transient Reaction
Transient reactions were carried out at 330 and 460 • C. First, a similar reaction condition as in the standard SCR measurement mentioned above was adopted for the NH 3 oxidation reaction, in which NO was not introduced into the gas mixture. After the reaction reached a steady state, 500 ppm NO was switched into the reaction system. The concentrations of the gases (NO, NO 2 , N 2 O and NH 3 ) were measured with the FTIR gas analyzer (Thermo Nicolet iS10).
The surface area and pore volume were determined by N 2 physisorption measurement at −196 • C using a Quantachrome Autosorb-1, with pre-degassing at 250 • C for 6 h. The chemical composition of the sample was determined by ICP-OES (VISTA-MPX, Varian, Palo Alto, CA, USA). The H 2 temperature-programmed reduction (H 2 -TPR) was conducted on an AutoChem II 2920 (Micromeritics, Norcross, GA, USA) instrument. The samples were treated at 550 • C in 10%O 2 /N 2 for 1 h before testing. X-ray photoelectron spectroscopy (XPS) was performed with a PHI-1600 ESCA system spectrometer(Perkin-Elmer Co., USA) using Mg Kα as the X-ray source (1253.6 eV) under a residual pressure of 5 × 10 −6 Pa. The error of the binding energy was calibrated using C1s at 284.6 eV as the standard.

Conclusions
The Cu-SSZ-13 suffered severe deactivation after Pt impregnation, resulting in a decrease in NO x conversion. The Pt 0 species accelerated the NH 3 selective oxidation reaction to N 2 in the temperature range below 330 • C, reducing the deNO x activity of PtxCu samples due to the insufficient amount of reductant taking part in the standard NH 3 -SCR reaction. In addition to the over-consumption of NH 3 due to NH 3 oxidation, the formation of NO and NO 2 via NH 3 and NO oxidation (R.3-R.7) further increased the loss of deNOx activity in Pt-poisoned samples at high temperatures (around 460 • C). The NSCR (R.2) reaction contributed to the increased N 2 O formation in Pt-poisoned samples at around 330 • C. After hydrothermal treatment, the activity was mostly recovered due to the transformation of Pt 0 , which catalyzed the parallel reactions, into PtO x species.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/catal11070796/s1, Figure S1: NH 3 conversion as a function of temperature during NH 3 oxidation experiment for the Cu-SSZ-13 and PtxCu samples. Reactant feed contains 500 ppm NH 3 , 5% O 2 5% H 2 O, balanced with N 2 ; flow rate at 500 mL/min, GHSV: 150,000 h −1 , Figure S2: NO 2 formation as a function of temperature during NH 3 -SCR reaction. The reactant feed contains 500 ppm NH 3 , 5% O 2 , 5% H 2 O, balanced with N 2 ; flow rate at 500 mL/min, GHSV: 150,000 h −1 , Figure S3: XRD patterns of the (a) fresh and (b) aged catalysts, Figure S4: Solid state 27 Al MAS NMR spectra of the Cu-SSZ-13 and Pt-poisoned samples, Figure S5: Ratio of different Pt species in the Pt loading samples quantified by the XPS spectra in Figure 5, Figure

Conflicts of Interest:
The authors declare no conflict of interest.