Nitrogen Monoxide and Soot Oxidation in Diesel Emissions with Platinum–Tungsten / Titanium Dioxide Catalysts: Tungsten Loading E ﬀ ect

: Compared with Pt / TiO 2 , tungsten-loaded Pt–W / TiO 2 catalysts exhibit improved activity for NO and soot oxidation. Using catalysts prepared by an incipient wetness method, the tungsten loading e ﬀ ect was investigated using Brunauer–Emmett–Teller surface areas, X-ray di ﬀ raction, transmission electron microscopy (TEM), CO pulse chemisorption, H 2 temperature-programmed reduction, NH 3 temperature-programmed desorption (NH 3 -TPD), and pyridine Fourier transform infrared (FT-IR) spectroscopy. Loading tungsten on the Pt / TiO 2 catalyst reduced the platinum particle size, as revealed in TEM images. CO pulse chemisorption showed that platinum was covered with tungsten and the dispersion of platinum decreased when 5 wt.% or more of tungsten was loaded. The NH 3 -TPD and pyridine-FT-IR results demonstrated that the number of strong acid sites and Brønsted acid sites in the catalyst were increased by the presence of tungsten. Therefore, a catalyst containing an appropriate amount of tungsten increased the dispersion of platinum, thereby increasing the number of active sites for NO and soot oxidation, and increased the acidity of the catalyst, thereby increasing the activity of soot oxidation by NO 2 .


Introduction
Globally, harmful exhaust emissions regulations have been tightened owing to the increasing use of diesel engines in various fields, such as coal-fired power plants, ships, aviation, and automobiles [1,2]. Exhaust emissions from diesel engines include hydrocarbons, soot, NOx, SOx, and carbon monoxide. Among these emissions, soot, which consists of unburned carbon particles, can cause air pollution, various respiratory diseases, and lung cancer. The World Health Organization (WHO) and the California Air Resources Board (CARB) classify soot as a first-level carcinogen [3]. The European Union (EU) introduced the EURO emission regulation in 1994 as a means to reduce the emission of pollutants in the exhaust gas of diesel cars. EURO-6c was introduced in September 2017, which applies the Worldwide Harmonized Light Vehicle Test Procedure (WLTP) and Real Driving Emissions (RDE) instead of the New European Driving Cycle (NEDC), an existing EURO-6 emission measurement method. As a result, it has become more difficult to meet emission limits [4].
Methods for reducing soot include pre-treatment technologies, which involve fuel reforming and improvement of combustion conditions, and post-treatment technologies, which involve the collection C_ soot + O 2 → CO 2 (1) The reaction to oxidize soot using O 2 , as shown in Equation (1), occurs at temperatures of 550 • C or higher; therefore, post-treatment technologies based on this reaction require a separate heat source to raise the DPF temperature. In addition, when a large amount of soot is combusted, the exothermic reaction may raise the temperature of the catalyst and the DPF to approximately 1000 • C, thereby causing degradation of the catalyst and structural damage to the filter. In contrast, NO, which accounts for 80-90% of NOx in diesel exhaust gas, can be oxidized using a DPF loaded with a catalyst and the generated NO 2 can then oxidize soot, as shown in Equations (2) and (3). As this process occurs at a lower temperature (250 • C), a separate heat source is not required to increase the temperature of the DPF, and the possibility of durability degradation owing to thermal aging of the catalyst and filter is low [6,7].
As platinum catalysts have high oxidation activities, many studies on soot oxidation have focused on such catalysts. Table 1 provides a summary of platinum catalysts that have been applied for soot oxidation. Oi-Uchisawa et al. [8] studied the characteristics of soot oxidation by loading 0.3 wt.% platinum on TiO 2 , ZrO 2 , SiO 2 , and Al 2 O 3 supports. When the reaction gas contained SO 2 , the catalyst on the TiO 2 support showed high activity for soot oxidation, whereas the catalysts on SiO 2 and Al 2 O 3 supports showed low activities. The specific surface area and the platinum dispersion did not significantly affect the catalytic activities of these catalysts. However, in sulfate accumulation experiments, accumulation of SO 4 2− was observed on platinum in the Pt/SiO 2 catalyst and on TiO 2 in the Pt/TiO 2 catalyst, indicating that the activity of platinum, which is an active metal, was maintained. Matarrese et al. [9] performed soot experiments using a Pt[1]-K[5.4]/Al 2 O 3 catalyst and found that the activity of soot oxidation was increased by the NOx adsorption and desorption characteristics of potassium. Krishna et al. [10] studied soot oxidation using a CeO 2 support with redox properties and an excellent oxygen storage capacity. They explained that the reactive oxygen species present on the CeO 2 support of a Pt[2.5]/CeO 2 catalyst oxidized NO adsorbed on platinum, thereby activating the soot oxidation process. In addition, Liu et al. [11] added MnOx-CeO 2 to Al 2 O 3 to enhance the thermal stability of the Pt/Al 2 O 3 catalyst and the activity of soot oxidation. This modification lowered T 50 ( • C) (temperature with a 50% soot conversion rate) from 474 to 442 • C for soot oxidation. However, the active sites of catalysts containing CeO 2 and alkaline metals such as potassium are poisoned during reactions involving SO 2 , resulting in catalyst deactivation [10,11]. Liu et al. [12] studied soot oxidation using a Pt/ZSM-5 catalyst to reduce poisoning of the catalyst by SO 2 . During NO oxidation in the presence of SO 2 , a higher amount of NO 2 was produced at a low temperature when ZSM-5 was used rather than Al 2 O 3 as a support for platinum. During soot oxidation in the presence of SO 2 , T 50 ( • C) decreased by 20 • C with the Pt/ZSM-5 catalyst. However, a small amount of platinum was located in the micropores (0.57 nm) of ZSM-5, and it was difficult for soot larger than the pore size of ZSM-5 to come into contact with platinum in the pores. Therefore, to prevent platinum from being present in the zeolite pores, Gao et al. [13] synthesized platinum nanoparticles with an average particle size of 7.5 nm and then loaded the particles on an H-ZSM-5 support. Using this catalyst, T 50 ( • C) of 433 • C was obtained, even though the amount of NO was reduced to 500 ppm in the soot oxidation experiment. Tungsten is known to increase the acidity and NOx adsorption properties of catalysts [14]. However, heretofore, there has been no exploration of tungsten-added Pt catalysts for their soot oxidation activity. The exhaust gas temperature required for passive regeneration is gradually being lowered owing to the development of exhaust gas recirculation (EGR) and diesel engine combustion technology. Therefore, to oxidize soot by passive regeneration, it is necessary to develop a catalyst with excellent activity at low temperatures. In this study, we investigated the effect of the tungsten content in a Pt/TiO 2 catalyst on the soot oxidation characteristics at low exhaust gas temperatures. To determine the oxidation characteristics of soot on the catalyst, we confirmed the formation of NO 2 , which acts as an oxidant of soot, and examined soot oxidation by NO 2 and NO reaction gases. Moreover, tungsten-loaded Pt/TiO 2 catalysts were characterized and, based on the results, their oxidation characteristics were investigated. Table 2 summarizes the results of the BET (Brunauer-Emmett-Telle) analysis. When the tungsten loading amount on the catalyst increased to 3 wt.%, the specific surface area increased by approximately 3.3 m 2 /g relative to that of the catalysts with 0 wt.% tungsten. However, when the tungsten loading amount was more than 5 wt.%, the specific surface area decreased. The specific pore volumes were similar for all the catalysts, but the average pore diameter decreased from 28.9 to 26.6 nm as the tungsten content increased. Table 2. Analysis of the structural characteristics of Pt-W[x]/TiO 2 catalysts (BET surface area (S BET ), external surface area (S ext ), pore volume (V P ), and average pore size (D P )).

Sample
S BET (m 2 /g) S ext (m 2 /g) V P (m 2 /g) D P (nm) The XRD (X-ray diffraction) patterns of the Pt-W[x]/TiO 2 catalysts are shown in Figure 1. Anatase and rutile peaks, which are characteristic of TiO 2 , were observed for all the synthesized catalysts, but no platinum or WOx peaks were detected. Platinum may not have been observed because the metal loading amount on the catalyst was as low as 1 wt.% [15]. In the case of WOx, the characteristic peaks (2θ = 23.3 • , 24.4 • , and 33.8 • ) were not observed even when the metal loading amount on the catalyst was 7 wt.%. This suggests that amorphous or nano-sized WOx was well dispersed on TiO 2 [16].    (Figure 2c) were 1.8 and 1.6 nm, respectively. The platinum particle size distributions show that the size of the platinum particles decreased and the width of the particle size distribution narrowed as the tungsten content of the catalyst increased. However, tungsten on the surface of TiO2 was not observed to form clusters (Figure 2b,c). EDS (energydispersive X-ray spectroscopy) mapping was performed to confirm the distribution of platinum and  X-ray spectroscopy) mapping was performed to confirm the distribution of platinum and tungsten on the catalyst surface, and the results are shown in Figure 3. Tungsten, which was difficult to observe using S-TEM, was found to be distributed throughout the catalyst.
Catalysts 2020, 10, x FOR PEER REVIEW 5 of 14 tungsten on the catalyst surface, and the results are shown in Figure 3. Tungsten, which was difficult to observe using S-TEM, was found to be distributed throughout the catalyst.    Table 3 shows the results of the CO pulse chemisorption analysis for the Pt-W[x]/TiO 2 catalysts. When the tungsten loading amount on the catalyst was less than 3 wt.%, the dispersion and particle size of platinum were not significantly affected. However, when the loading amount was more than 5 wt.%, the dispersion of platinum drastically decreased and the platinum particle size dramatically increased. However, the average particle sizes of platinum obtained from the CO pulse chemisorption and TEM analyses are considerably different. According to the results of the TEM analysis, the amount of CO chemically adsorbed on platinum during CO pulse chemisorption decreased owing to partial blocking of the platinum particles by WOx.    a Pt loadings were calculated from ICP-MS (Inductively Coupled Plasma Mass Spectrometer). b Pt dispersions and particle sizes were calculated from CO pulse chemisorption. c Pt particle sizes from S-TEM images. Figure 4 shows the H 2 -TPR (H 2 temperature-programmed reduction) results for the Pt-W[x]/TiO 2 catalysts. For the catalyst without tungsten, the reduction peaks produced by the strong metal-support interaction (SMSI) effects of PtOx and Pt-TiO 2 appeared at 60 and 170 • C, respectively, and the surface reduction peak of TiO 2 appeared at 320 • C [17]. In contrast, for the catalysts containing tungsten, a new peak was detected at approximately 230 • C, which was considered to be generated by the Pt-W interaction. The reduction peak resulting from the SMSI effect of Pt-TiO 2 shifted to higher temperatures when up to 3 wt.% tungsten was added, but then shifted back to lower temperatures when more than 3 wt.% tungsten was added. This change may have resulted from overloading of tungsten, thereby increasing the W-Ti interaction and decreasing the Pt-TiO 2 interaction [18]. Figure 5 shows the NH 3 -TPD (NH 3 temperature-programmed desorption) curves of the Pt-W[x]/TiO 2 catalysts. The NH 3 -TPD curves can be deconvoluted into three peaks corresponding to weak, medium, and strong acid sites. The acid site densities of the catalysts with increasing tungsten content were calculated from the peak areas (Table 4). As the tungsten content increased, the total number of acid sites decreased. However, the amount of strong acid sites increased with increasing tungsten content, and broad peaks were detected at 460 and 480 • C in catalysts containing more than by the Pt-W interaction. The reduction peak resulting from the SMSI effect of Pt-TiO2 shifted to higher temperatures when up to 3 wt.% tungsten was added, but then shifted back to lower temperatures when more than 3 wt.% tungsten was added. This change may have resulted from overloading of tungsten, thereby increasing the W-Ti interaction and decreasing the Pt-TiO2 interaction [18].  Catalysts 2020, 10, x FOR PEER REVIEW 7 of 14 Figure 5 shows the NH3-TPD (NH3 temperature-programmed desorption) curves of the Pt-W[x]/TiO2 catalysts. The NH3-TPD curves can be deconvoluted into three peaks corresponding to weak, medium, and strong acid sites. The acid site densities of the catalysts with increasing tungsten content were calculated from the peak areas (Table 4). As the tungsten content increased, the total number of acid sites decreased. However, the amount of strong acid sites increased with increasing tungsten content, and broad peaks were detected at 460 and 480 °C in catalysts containing more than 3 wt.% tungsten. Thus, as the tungsten loading amount on the catalyst increased, WOx covered TiO2, thereby decreasing the amount of TiO2-related weak and medium acid sites and increasing the amount of WOx-related strong acid sites [19].    To investigate the relative concentrations of Brønsted and Lewis acid sites on the surfaces of the Pt-W[x]/TiO 2 catalysts, FT-IR (Fourier transform infrared) analyses were performed by adsorbing pyridine at 200 • C. At Brønsted acid sites, pyridine is adsorbed as a pyridinium ion after accepting a proton, whereas it is adsorbed as covalently bonded pyridine at Lewis acid sites via electron-pair sharing [20,21]. Figure 6 shows the FT-IR spectra of the pyridine-adsorbed Pt-W[x]/TiO 2 catalysts. Generally, peaks corresponding to the Brønsted acid sites appear at 1545 and 1640 cm −1 , whereas those corresponding to the Lewis acid sites appear at 1435 and 1598 cm −1 [22,23]. Brønsted acid sites were not detected in the catalyst without tungsten. However, a peak assigned to the Brønsted acid sites (1640 cm −1 ) slightly increased with increasing tungsten content. Table 5 shows the IR-band area and ratio (B/L ratio) of the Brønsted and Lewis acid sites with increasing tungsten content in Pt-W[x]/TiO 2 . Assuming the adsorption sites of pyridine are identical to those of ammonia, the densities of the Brønsted and Lewis sites were calculated by multiplying total acid site density by B/L B/L+1 and 1 B/L+1 , respectively. As the tungsten content in the catalyst increased to 7 wt.%, the peak areas of the Brønsted and Lewis acid sites increased. In addition, the B/L ratio indicated a three-fold increase in the number of Brønsted acid sites. These results reveal that tungsten loading resulted in the formation of Brønsted acid sites and a decrease in the proportion of Lewis acid sites.
Catalysts 2020, 10, x FOR PEER REVIEW 8 of 14 those corresponding to the Lewis acid sites appear at 1435 and 1598 cm −1 [22,23]. Brønsted acid sites were not detected in the catalyst without tungsten. However, a peak assigned to the Brønsted acid sites (1640 cm −1 ) slightly increased with increasing tungsten content. Table 5 Figure 5. b B, L area and B/L ratios were obtained from the pyridine adsorption FTIR spectroscopy results. c Brønsted and Lewis acid density were calculated using total acid density and B/L ratio.

Catalytic Activities of Pt-W[x]/TiO 2 for NO and Soot Oxidation
The catalytic oxidation of NO to NO 2 is an important reaction for NO emitted from diesel engines, as the produced NO 2 can be used as an oxidant in soot oxidation. Figure 7a illustrates the NO oxidation results over the Pt-W[x]/TiO 2 catalysts. The amount of NO 2 produced by NO oxidation increased as the content of tungsten increased at temperatures below 350 • C. At 250 • C, the concentration of NO 2 generated increased approximately 3.5 times as the amount of tungsten in the catalysts increased to 7 wt.%. As revealed by the NH 3 -TPD analysis, this phenomenon may occur because the amount of strong acid sites on the surface of the catalyst increased as the tungsten loading amount increased. Thus, the adsorption of NO on the catalyst surface was favored and the adsorbed NO was oxidized to NO 2 on the platinum surface [14].
Next, soot oxidation by NO 2 was performed by supplying 300 ppm NO 2 , 3 ppm SO 2 , 10% O 2 , and balance N 2 . Figure 7b shows the experimental results over the Pt-W[x]/TiO 2 catalysts. When 1 wt.% tungsten was loaded on the catalyst, the activity was similar to that of the Pt/TiO 2 catalyst without tungsten. However, when 3 wt.% tungsten was loaded, the soot oxidation rate rapidly increased, and similar reaction results were also observed at tungsten loading amounts of more than 3 wt.%. Thus, as the acidity of the catalyst surface increased, NOx was more easily absorbed. However, the addition of more than 3 wt.% tungsten did not significantly affect soot oxidation by NO 2 . Figure 7c [13]. The dependence of the soot oxidation performance on NO 2 was quite high. Moreover, the experimental results for the oxidation of NO into NO 2 (Figure 7a) and the oxidation of soot using NO 2 (Figure 7b) showed that as the content of tungsten in the Pt/TiO 2 catalyst increased, the supply of NO 2 , the oxidant required for soot oxidation, increased. However, the soot oxidation results differed when NO was used (Figure 7c). These experimental results indicate that there was a significant variable other than NO 2 production involved in the soot oxidation reaction.
increased as the content of tungsten increased at temperatures below 350 °C. At 250 °C, the concentration of NO2 generated increased approximately 3.5 times as the amount of tungsten in the catalysts increased to 7 wt.%. As revealed by the NH3-TPD analysis, this phenomenon may occur because the amount of strong acid sites on the surface of the catalyst increased as the tungsten loading amount increased. Thus, the adsorption of NO on the catalyst surface was favored and the adsorbed NO was oxidized to NO2 on the platinum surface [14]. Next, soot oxidation by NO2 was performed by supplying 300 ppm NO2, 3 ppm SO2, 10% O2, and balance N2. Figure 7b shows the experimental results over the Pt-W[x]/TiO2 catalysts. When 1 wt.% tungsten was loaded on the catalyst, the activity was similar to that of the Pt/TiO2 catalyst without tungsten. However, when 3 wt.% tungsten was loaded, the soot oxidation rate rapidly increased, and similar reaction results were also observed at tungsten loading amounts of more than 3 wt.%. Thus, as the acidity of the catalyst surface increased, NOx was more easily absorbed. However, the addition of more than 3 wt.% tungsten did not significantly affect soot oxidation by NO2. Figure 7c [13]. The dependence of the soot oxidation performance on NO2 was quite high. Moreover, the experimental results for the oxidation of NO into NO2 (Figure 7a) and the oxidation of soot using NO2 (Figure 7b) showed that as the content of tungsten in the Pt/TiO2 catalyst increased, the supply of NO2, the oxidant required for soot oxidation, increased. However, the soot oxidation results differed when NO was used (Figure 7c). These experimental results indicate that there was a significant variable other than NO2 production involved in the soot oxidation reaction. Figure 8 shows illustrations of the soot oxidation process using NO based on the analysis of the catalyst properties. As shown in Figure 8a, NO2 produced on the platinum surface migrates to soot when the platinum and the soot are adjacent to each other, thereby leading to rapid soot oxidation. Thus, the Pt-W[3]/TiO2 catalyst, which is loaded with an appropriate amount of tungsten, is considered advantageous for soot oxidation using NO. In contrast, as shown in Figure 8b, when platinum was completely covered by WOx, gaseous NO could easily react with platinum to generate NO2, but contact between soot and platinum, the active site, is difficult. Therefore, soot oxidation is relatively unlikely to occur. Therefore, NO oxidation was dominant in the catalyst loaded with 7 wt.% tungsten, but the soot oxidation rate using NO was relatively low.  Figure 8 shows illustrations of the soot oxidation process using NO based on the analysis of the catalyst properties. As shown in Figure 8a, NO 2 produced on the platinum surface migrates to soot when the platinum and the soot are adjacent to each other, thereby leading to rapid soot oxidation. Thus, the Pt-W[3]/TiO 2 catalyst, which is loaded with an appropriate amount of tungsten, is considered advantageous for soot oxidation using NO. In contrast, as shown in Figure 8b, when platinum was completely covered by WOx, gaseous NO could easily react with platinum to generate NO 2 , but contact between soot and platinum, the active site, is difficult. Therefore, soot oxidation is relatively unlikely to occur. Therefore, NO oxidation was dominant in the catalyst loaded with 7 wt.% tungsten, but the soot oxidation rate using NO was relatively low.  Figure 9 displays the temperature at which soot conversion rates of 10% (T10(°C)) and 20% (T20(°C)) are achieved, as well as the respective regression curves according to the tungsten content of the catalyst. When the catalyst was loaded with 3 wt.% tungsten, the lowest T10(°C) and T20(°C) values of 284 and 301 °C, respectively, were obtained. The regression curves suggest that a tungsten loading of 3.5-4.0 wt.% will result in the lowest temperature for soot conversion.  Figure 9 displays the temperature at which soot conversion rates of 10% (T 10 ( • C)) and 20% (T 20 ( • C)) are achieved, as well as the respective regression curves according to the tungsten content of the catalyst. When the catalyst was loaded with 3 wt.% tungsten, the lowest T 10 ( • C) and T 20 ( • C) values of 284 and 301 • C, respectively, were obtained. The regression curves suggest that a tungsten loading of 3.5-4.0 wt.% will result in the lowest temperature for soot conversion.  Figure 9 displays the temperature at which soot conversion rates of 10% (T10(°C)) and 20% (T20(°C)) are achieved, as well as the respective regression curves according to the tungsten content of the catalyst. When the catalyst was loaded with 3 wt.% tungsten, the lowest T10(°C) and T20(°C) values of 284 and 301 °C, respectively, were obtained. The regression curves suggest that a tungsten loading of 3.5-4.0 wt.% will result in the lowest temperature for soot conversion.

Catalyst Preparation
Degussa P-25 was used as the TiO2 support, ammonium paratungstate (APT, [(NH4)10(H2W12O42)·4H2O]) was used as a precursor for tungsten, and platinum was obtained from a platinum solution (A type, 16.6%, SNS Co., Hwa-seong, Korea). In this study, the content of platinum

Catalyst Preparation
Degussa P-25 was used as the TiO 2 support, ammonium paratungstate (APT, [(NH 4 )10(H 2 W 12 O 42 )·4H 2 O]) was used as a precursor for tungsten, and platinum was obtained from a platinum solution (A type, 16.6%, SNS Co., Hwa-seong, Korea). In this study, the content of platinum on the support was fixed at 1 wt.%, and the content of tungsten was 0-7 wt.%. The Pt-W[x]/TiO 2 catalysts (where x is the content of tungsten in wt.%) were prepared as follows. Firstly, calculated APT was added to solution of oxalic acid dehydrate ([HO 2 CCO 2 H·2H 2 O], Sigma-Aldrich, St. Louis, MO, USA). The mixture was stirred at 300 rpm for 10 min until clear solution formed. An APT solution was loaded on 10 g of TiO 2 by an incipient wetness method and then dried at 100 • C for 5 h to prepare W[x]/TiO 2 powder. Next, the calculated platinum solution was loaded on the 10 g of dried W[x]/TiO 2 powder by the incipient wetness method. The Pt-W[x]/TiO 2 catalyst was dried at 100 • C for 24 h and then calcined in an air atmosphere at 500 • C for 4 h using a calcination furnace at a heating rate of 1 • C/min.
For NO oxidation, the Pt-W[x]/TiO 2 catalyst powder was mixed with distilled water (catalyst (g)/water (g) = 0.5) and milled for 30 min using a mortar to prepare a slurry. Next, the catalyst slurry was coated on a cordierite honeycomb (400 cells per square inch (cpsi) with a height of 0.6 cm) at a rate of 100 g/L. SEM images of the non-catalyst coating and catalyst coating samples are shown in Figure S1, and the properties of cordierite honeycomb are shown in the Table S1.

NO Oxidation
A tube-type quartz reactor with an inner diameter of 1.6 cm was used to confirm the NO oxidation performance of the Pt-W[x]/TiO 2 catalysts. The prepared honeycomb catalysts were fixed in the quartz reactor, and the NO and NO 2 concentrations after the reaction were analyzed using non-dispersive infrared spectroscopy (Fuji Co., Tokyo, Japan). The experimental conditions for NO oxidation were as follows: 300 ppm NO, 3 ppm SO 2 , 10% O 2 , 5% H 2 O, and N 2 balance. The total flow rate of the reactants was 1.0 L/min (GHSV = 50,000 h −1 ).

Soot Oxidation
A tube-type quartz reactor with an inner diameter of 0.4 cm was used to confirm the soot oxidation performance of the Pt-W[x]/TiO 2 catalysts. The Pt-W[x]/TiO 2 catalyst powder and soot were mixed (catalyst (g)/soot (g) = 4) using a mortar for 30 min to ensure good contact. The model soot used in the experiment was Printex U (Degussa, 20-30 nm). The experiment was carried out at 200-450 • C under the following reaction conditions: 300 ppm NO or NO 2 , 3 ppm SO 2 , 10% O 2 , 5% H 2 O, and N 2 balance. The total flow rate of the reactants was 100 mL/min (GHSV = 50,000 h −1 ). The concentration of CO 2 generated after soot oxidation was measured at the outlet using gas chromatography (7890A, Agilent Technology Co, Santa Clara, CA., USA) with a thermal conductivity detector and a capillary column (Carboxen 1010 PLOT, 30 m × 0.533 mm, SUPELCO Co., St. Louis, MO, USA).

Catalyst Characterization
The BET surface areas and pore characteristics of the Pt-W[x]/TiO 2 catalysts were measured using a Micromeritics ASAP 2020 instrument.
XRD (D/max-2200/PC, Rigaku Co., Tokyo, Japan) with CuKα radiation was conducted at 40 kV and 200 mA in the 2θ range of 10-100 • to confirm the crystal structure as the tungsten content of the catalyst increased.
To determine the platinum particle size, tungsten loading state, and dispersion of the catalyst, FE-TEM (JEM-2200FS, JEOL Co., Tokyo, Japan) was used. To analyze the size and distribution of platinum particles on TiO 2 accurately, S-TEM was used. The elemental composition of the catalyst surface was analyzed by EDS mapping. For these measurements, the sample was dispersed in ethanol and prepared on a copper grid (300 mesh).
CO pulse chemisorption, H 2 -TPR, and NH 3 -TPD were performed using a Micromeritics Autochem 2920 instrument. CO pulse chemisorption was performed to determine the platinum particle size and dispersion of the catalyst. The catalyst was subjected to reduction treatment by maintaining a temperature of 300 • C for 1 h while injecting a mixed gas of 10% H 2 /Ar. Then, He gas was injected for 30 min to remove adsorbed hydrogen. Next, the temperature was lowered to room temperature, and 10% CO/He was pulse injected at a flow rate of 50 mL/min to measure the particle size and dispersion of platinum based on the amount of adsorbed CO. H 2 -TPR analysis was performed to determine the reducing ability of the catalyst according to temperature. The calcined sample was heated to 800 • C at a heating rate of 5 • C /min while injecting a mixed gas of 10% H 2 /Ar. NH 3 -TPD was conducted to analyze the acid sites in the catalyst. The calcined sample was activated by maintaining a temperature of 300 • C for 1 h under a mixed gas of 10% H 2 /Ar. After exposure to a mixed gas of 15% NH 3 /He at room temperature for 1 h, the temperature of the sample was raised to 100 • C to desorb any NH 3 physically adsorbed on the surface of the catalyst, and then He was injected for 30 min. Subsequently, the analysis was conducted by raising the sample temperature to 600 • C at 10 • C/min while flowing He at 50 mL/min.
To determine the distribution of Brønsted and Lewis acid sites on the catalyst, pyridine was adsorbed, and FT-IR spectra (FTLA2000-104, ABB Co., Zürich, Switzerland) were obtained. The sample was mounted on a diffuse reflection infrared Fourier transform (DRIFT) cell, and a background IR spectrum was obtained after degassing in a 300 • C vacuum atmosphere for 2 h. Next, the pyridine solution was injected into the DRIFT cell using a syringe, and pyridine molecules were adsorbed on the catalyst. The adsorption process was allowed to proceed at 200 • C for 30 min to saturate the acid sites on the catalyst surface with pyridine, and IR spectra were obtained.

Conclusions
In this study, a Pt-W[x]/TiO 2 catalyst showing excellent soot oxidation was developed. As the content of tungsten in the Pt-W[x]/TiO 2 catalyst increased, the amount of strong acid sites and Brønsted acid sites increased, and the catalytic activity for the oxidation of NO to NO 2 increased. The catalysts with 3 wt.% or more tungsten showed similar activities for soot oxidation by NO 2 . In contrast, the catalyst loaded with 3 wt.% tungsten showed the highest activity for soot oxidation using NO. Although increased oxidation of NO to NO 2 occurred at higher tungsten loadings, the surface of the active metal, platinum, was covered by WOx, making it difficult to deliver the produced NO 2 to soot and thus lowering the reaction activity. We confirmed that an appropriate surface acidity and coverage of platinum by tungsten were important in the Pt-W[x]/TiO 2 catalysts for effective soot oxidation. In this study, we found that the Pt-W[3]/TiO 2 catalyst showed the highest rate of soot oxidation.