Diesel engines account for a significant share of industrial and public transportation, owing to their excellent fuel economy. However, the emissions from these engines remain a challenging issue. The formation of soot during unsteady combustion is governed by both kinetics and thermodynamics, whereas NO_x are produced under oxygen-rich, high temperature conditions [1].Soot and NO_x are both detrimental to human health and the environment [2], and soot is the primary source of PM2.5 air pollution [3]. The trade-off tendency between reducing these two pollutants has greatly increased the requirements for post-combustion processing system. As an example, diesel particulate filters (DPFs) can trap particulate matter (PM), but need time to regenerate within a specific temperature window to maintain working efficiency. Catalytic diesel particulate filters (CDPFs) are a potential means of mitigating this problem, and are both inexpensive and simple.

The materials included in a CDPF system are carefully selected. Considering that most transitional metal oxides (such as Co₃O₄) and precious metals (including Pt and Ag) are sensitive to sulfur, it is essential to include a sulfur trap upstream of the catalytic material so as to prolong the lifetime of the catalyst [4]. The use of higher quality oil provides more options for developing novel catalysts, including the two above-mentioned materials. In addition to the catalyst design, the cost of the catalyst also has to be evaluated. Fortunately, although silver is classified as a noble metal, its cost is relatively reasonable. Thus, as the result of economic considerations, silver-based catalysts have been widely applied in catalytic oxidation processes such as the removal of CO, formaldehyde and volatile organic compounds (VOCs) [5-8]. In recent years, there has also been interest in the application of silver and its oxides to catalytic soot gasification. Among the metal oxides, Ag₂O exhibits the best performance during the catalysis of soot combustion, although its crystal structure degrades during combustion processes [9]. For this reason, bulk Ag₂O cannot be used in catalytic systems operating at higher temperature (above 300 ℃). In comparison, metallic silver is relatively stable, and its strong interactions with metal oxide supports during catalytic soot removal are also helpful. As an example, the thermal stability of the catalyst is further enhanced by interactions between silver and SnO₂ [10]. The Ag/CeO₂ system is currently the focus of much research, primarily owing to the beneficial properties of CeO₂, such as thermal stability and oxygen storage capacity [11], as well as the ease with which platinum group metals can be dispersed on its surface. At present, however, our understanding of the interactions between silver and CeO₂ is incomplete. A 'rice ball' shaped Ag/CeO₂ catalyst has been synthesized, and its catalytic performance has been found to be superior to that of material prepared by the impregnation method [12]. Silver also accelerates oxygen migration from the silver surface to the CeO₂ via the interface between the two materials, accompanied by active oxygen generation [13]. Contrary to commonly accepted catalytic mechanisms, the combination of CeO₂ and silver forms Ag-O-Ce bonds that stabilize surface oxygen and hinder oxygen transfer. This complex interaction eventually results in the inferior performance of Ag/CeO₂ compared with that of Ag/ZrO₂ or Ag/Al₂O₃ [14]. In addition to pure metal oxide supports such as CeO₂, Al₂O₃, ZrO₂ and SnO₂, mixed-oxide supports have been explored in an attempt to enhance the silver-support interaction. Mixed metal oxides such as perovskite-type catalysts have been adopted as a potential candidate. In Ag/ABO₃ systems, metallic silver is believed to be the primary phase and provides most of the active oxygen species [15, 16]. Consequently, it appears that silver ions are partially incorporated into A sites in ABO₃ perovskite-type materials, generating extra oxygen vacancies [17]. Although silver is widely regarded as a promoter of the catalytic oxidation process, there is not yet a consensus regarding its role. On the basis of previous work regarding silver, a few studies have focused on the diverse behavior of this metal during soot removal in various atmospheres, especially in NO_x. Hence, it would be helpful to have a better understanding of the function of silver species in the NO_x atmosphere, since this would help to explain the reaction mechanisms and to elucidate the role of silver in O₂ atmospheres.

In our previous studies, Co-Ce composite oxides were investigated, and the synergistic effects between Co₃O₄ and CeO₂ were found to be responsible for the enhanced performance of these catalysts [18]. However, the light-off temperatures observed during this prior work were not low enough to meet the requirements for soot emissions under certain conditions, such as when idling. In the present research, the most active of the Co-Ce composite oxides (Co_0.93Ce_0.07) was employed as a silver support, with the aim of improving the catalytic performance and gaining a better understanding of the effects of silver on catalytic activities as well as the reaction mechanisms under various atmospheres.

Co (Ac)₂·4H₂O, Ce (NO₃)₃·6H₂O, AgNO₃ and C₆H₈O₇·H₂O, applied as precursors, were all analytical reagent (AR). The method used to prepare the catalysts has been reported in a previous paper [18]. The catalysts are denoted herein as xAg/Co_0.93Ce_0.07, where x (=0, 0.05, 0.1, 0.2 or 0.3) represents the molar ratio of Ag to the other metals (that is, x=Ag/(Co+Ce)). Co₃O₄ and 0.2Ag/Co₃O₄ were synthesized by the same method for comparison purposes.

X-ray diffraction (XRD) patterns were acquired using a Rigaku D/max-2200/PC over the 2θ range of 20°-80°. Raman spectra were obtained with a SENTERRA R200-L (Bruker Optics), employing a diode laser operating at 532 nm and 0.2 mW. Electron paramagnetic resonance (EPR) spectroscopy was performed at 25 ℃ using an EMX BRUKER spectrometer with a cavity frequency of approximately 9.83 GHz, a magnetic field modulation of 100 kHz and a power level of 0.5 mW. X-ray photoelectron spectroscopy (XPS) was used to determine the surface states of selected elements, employing a Kratos Analytical-A (Shimadzu) with a monochromatic Al K_α source at 1486.6 eV. Fourier transform infrared (FTIR) transmission spectra were obtained with a Nicolet 6700 (Thermofisher). Samples were prepared in the form of pressed wafers (ca. 1% sample in KBr). All spectra involved the accumulation of 80 scans at 4 cm^-1 resolution.

NO_x-temperature programmed desorption (TPD) was performed using a fixed bed quartz tube system in conjunction with an NO_x analyzer (42i-LS, Thermofisher Electron Corporation) to detect NO, NO₂ and NO_x. Samples were pretreated under a 180 mL/min N₂ flow at 300 ℃ for 1 h and then allowed to cool naturally to 50 ℃ under the same conditions. The apparatus was subsequently purged with a flow of 300 ppm NO in N₂ for 60 min, followed by exposure to N₂ at 180 mL/min until no NO_x was observed. After the pretreatment, NO-TPD measurements were carried out under a N₂ flow at 180 mL/min from 50 to 500 ℃ at a ramp of 10 ℃/min.

Soot-TPR was conducted using a thermogravimetric analysis (TGA) instrument (STA449F3). The test sample, consisting of a highly compact mixture of catalyst and soot, was heated at 100 ℃ under Ar at a flow rate of 50 mL/min to provide a clean surface. Subsequently, the sample was heated from 100 to 950 ℃ at a ramp of 10 ℃/min. H₂-TPR was performed using a Micromeritics Chemisorb 2720. In each trial, a 0.02 g quantity of the catalyst was pretreated under a 25 mL/min N₂ flow at 300 ℃ for 30 min and then cooled to 50 ℃ under the same conditions. The sample was subsequently exposed to 5% H₂ in N₂ at a flow rate of 25 mL/min until a stable thermal conductivity detector signal was obtained. Finally, the sample was assessed under the same conditions upon heating from 50 to 600 ℃ at a ramp of 10 ℃/min.

The catalytic performance was evaluated by a temperature-programmed oxidation (TPO) procedure with a fixed bed quartz tube system. During each run, the temperature was increased at 1.6 ℃/min from 100 to 400 ℃. Well mixed specimens were prepared by first combining a typical model soot (Printex U) with the catalyst at a fixed mass ratio (1:19) followed by grinding for 15 min. Prior to each trial, a 0.33 g quantity of the soot-catalyst mixture was placed in the quartz tube and silica wool was inserted on either side of the sample. This was followed by pretreatment at 200 ℃ for 1 h under a He flow to remove adsorbed water and hydrocarbons. The reactor was then cooled to 100 ℃ and a reactant gas containing either 5% O₂ or 2000 ppm NO + 5% O₂ in He was fed into the reactor at a flow rate of 25 mL/min. The outlet gas from the reactor was analyzed at 15 min intervals using a gas chromatograph (Shimazu GC-14B) with a thermal conductivity detector, in conjunction with a Porapak Q column to separate CO₂ and N₂O and a 5A molecular sieve column to analyze N₂, O₂, NO and CO. The catalytic activity was evaluated based on the T₁₀ values of each sample, defined as the temperature at which 10% of the soot was consumed during the reaction.

The catalytic performance of xAg/Co_0.93Ce_0.07 samples in either 5% O₂ or NO_x atmospheres (5% O₂+ 2000 ppm NO) is displayed in Fig. 1. It is evident that the silver-free samples ignited the soot at higher temperatures in O₂ compared with the NO_x environment, while the silver-based samples exhibited the opposite tendency. The light-off temperatures also decreased with increasing silver content under both O₂ and NO_x, with the sample having the maximum silver loading (x=0.3) showing a light-off temperature of 197 ℃. Notably, both the 0.2Ag/Co_0.93Ce_0.07 and 0.3Ag/Co_0.93Ce_0.07 initiated soot combustion at 226 ℃ in combination with NO_x. These results indicate that silver plays an unconventional role during soot combustion over these materials. The 0.2Ag/Co_0.93Ce_0.07 ignited the soot at 257 ℃ and complete combustion of the soot was achieved below 350 ℃ even under real-world loose contact conditions (data not shown).

The XRD patterns of the catalysts demonstrated the presence of metallic silver, cubic Co₃O₄ and fluorite-type CeO₂ (data not shown). Together with the results of Raman spectroscopy (data not shown), these data indicate that the addition of silver did not modify the lattice structures of the Co-Ce mixed oxides (not shown). From the H₂-TPR data, it is evident that the silver did not enhance the redox ability of the Co-Ce mixed oxides, and that the activity of the lattice oxygen was unchanged under Ar (data not shown). The silver particle sizes calculated from XRD patterns (data not shown) ranged from 51.5 to 55.0 nm, and this extremely small variation in particle sizes eliminates the normal relationship between size and catalytic performance. The BET surface areas of the specimens were found to be quite low (Table 1), indicating that these materials were sintered during calcination. However, it is unlikely that this affected the catalytic performance, because the reactions being monitored proceeded at the three-phase boundary between the soot, catalyst and gas phase [19, 20].

NO_x-TPD was performed to study the surface adsorption of NO species on the catalysts. Previous research has shown that the thermal stability of adsorbed NO_x species follows the sequence: nitrite ions > nitrate ions > surface nitrosyl (50 ℃). The TPD profiles contained four peaks corresponding to different temperatures: low (150 ℃), middle (230 ℃) and high (330 and 410 ℃). The low and middle temperature peaks correspond to the decomposition or desorption of nitrosyl and nitrite species, respectively, while the double peaks located at higher temperatures are ascribed to nitrate bonded to cobalt and cerium, in sequence [21]. As shown in Fig. 2 and Table 1, a lower loading of silver (x < 0.3) increased the total amount of NO_x desorbed, although this increment in the desorption peaks occurred at a lower temperature rather than at a higher temperature. The desorption amount decreased as the temperature rose above 300 ℃, likely due to the partial oxidation of nitrite species via active oxygen provided by the silver. The significant promotion of NO₂ desorption below 150 ℃ over the silver-based catalysts can possibly be attributed to the capacity of these materials for the physical adsorption of reactants, and to their ability to stabilize and dissociate oxygen. With increases in the silver loading (x=0.3), the NO_x desorption exhibited a decreasing trend, presumably because the silver particles occupied the Co₃O₄ and CeO₂ adsorption sites. This occupation weakened the synergistic effect between Co₃O₄ and CeO₂ and adversely affected both NO_x adsorption and desorption. Excessive silver coverage lowers NO_x adsorption, accounting for the lack of any further increases in desorption in the case of samples with higher loadings (x=0.2 and 0.3).

The XPS data for these catalysts are displayed in Table 2 and Fig. 3. The O 1s spectra are composed of three overlapping peaks. The O_Ⅰ peak at 529.6 eV is ascribed to surface lattice oxygen, while the O_Ⅱ shoulder signal at 531.7 eV results from O^- and/or-OH [22, 23]. From these data, we can conclude that the O_Ⅱ peak is associated with chemisorbed oxygen. The O_Ⅲ peak at approximately 532.7 eV is due to weakly bonded surface oxygen [24].Oxygen adsorption was evidently enhanced over the silver-based samples, and the adsorption oxygen percentage (A_o) increased by 6.60% and 8.98% over the 0.2Ag/Co₃O₄ and 0.2Ag/Co_0.93Ce_0.07, respectively. Moreover, the presence of CeO₂ also increased the A_o value, as a result of the higher oxygen ratio in this material [18]. The soot functioned as a reducing agent and significantly altered the distribution of surface oxygen species. The reduction in the lattice oxygen ratio indicates the transformation of surface lattice oxygen to adsorbed oxygen. The Ag 3d profiles are composed of Ag 3d_3/2 and Ag 3d_5/2 spectra, and the latter was adopted to analyze the state of surface silver due to its stronger intensity. The asymmetric peak demonstrates the presence of silver ions under lean combustion conditions, and consists of two peaks due to metallic silver at 368 eV and oxidized silver (Ag^δ+) at 367 eV [25, 26]. The Ag^δ+ concentration was higher on the 0.2Ag/Co₃O₄ than the 0.2Ag/Co_0.93Ce_0.07, likely because the silver oxides were partially covered by metallic silver [14]. The amount of Ag^δ+was not the only factor affecting soot combustion; the activity of the support also played an important role in this process. The synergistic effect between the CeO₂ and Co₃O₄ accelerates soot combustion, and this effect still functioned in the case of the silver-based samples, leading to better performance over the 0.2Ag/Co_0.93Ce_0.07 than over 0.2Ag/Co₃O₄. The soot also appeared to stimulate surface reduction at room temperature. The shift in the Co 2p_3/2 peak to a higher value also illustrates the reduction of cobalt atoms having higher oxidation states. The Ce 3d spectrum shows a similar effect; a characteristic Ce³⁺ peak at 904 eV was generated by the 0.2Ag/Co_0.93Ce_0.07 and soot mixture [27, 28].

The durability tests were performed under NO_x (Fig. 4). Almost no loss of catalytic activity was observed after seven replicate trials, demonstrating excellent reproducibility. In these trials, varying the pre-treatment conditions provided a better understanding of the catalysts. The catalytic performance was found to decrease considerably after purging with He at 700 ℃ for 1 h, while the ignition temperature rose to 260 ℃. Interestingly, cooling the pre-treated sample to room temperature followed by exposure to the reactant gas for 3 h seemed to regenerate the material. That is, the catalytic performance recovered to the level of the fresh material. FTIR and XPS provided information regarding this regeneration (Figs. 5 and 6). A peak at 1384 cm^-1, corresponding to nitrate species, was generated by each sample except that pre-treated at 700 ℃ in He [29-32], and no evidence of nitrite species was observed at any time [33, 34]. We believe that nitrate species were crucial to soot abatement under NO_x. The XPS survey was used to determine the oxidation state of silver in the materials. High temperature pre-treatment in He produced a symmetric Ag 3d peak, demonstrating the disappearance of Ag^δ+. A high binding energy peak located at 369 eV (at approximately 10%) was also generated by the NO_x-treated sample, and is attributed to the formation of AgNO₃[35, 36]. It is noteworthy that no Ag^δ+ was identified in this sample, presumably because purging with NO_x converted any silver oxides to AgNO₃. On the basis of the catalytic performance data and the FTIR and XPS results, AgNO₃ species are thought to be the key to soot gasification under NO_x.

On the basis of literature reports and the present experimental results, the effects of silver can be categorized as follows.

Among the metal oxides, bulk Ag₂O has been shown to be the most active, although its poor stability limits its potential applications. Theoretical calculations have demonstrated that silver-based oxygen species exist below 200 ℃ on the Ag (111) crystal plane [37]. Li and co-workers [38] also found various oxidized silver species such as Ag₂O, Ag₂O₃ and Ag₃O on the Ag (100) and Ag (111) crystal faces. These mixed silver oxides are highly active and have been recognized as an important factor in various catalytic oxidation reactions [39-41]. The effect of Ag^δ+ during soot removal has been elucidated by Haneda et al. [42]. In the present study, silver ions were detected even at low O₂ partial pressures during XPS analyses (Fig. 3 and Table 2), indicating the formation of surface silver oxides in the presence of O₂. The data also revealed the continued presence of silver oxides between room temperature and lower reaction temperatures. In summary, it is apparent that the surface adsorbed oxygen directly participates in the reaction over silver-based samples.

Zero-valence silver is able to dissociate molecular oxygen into active oxygen species such as peroxides and superoxides [43, 44]. The combination of silver and various supports further promotes the formation of this active oxygen, and superoxide species have been observed over Ag/activated carbon catalysts even under fuel-rich combustion conditions [45]. CeO₂ combined with silver has also been shown to produce an intense superoxide signal [46]. In the present study, the process required a relatively "clean" metallic silver surface, thought to be acquired at higher temperatures following the destruction of silver oxides.

Alkali metals such as potassium have low melting points and sufficient mobility to improve the degree of soot-catalyst contact. In most cases, alkali metals exist in the form of ions rather than elemental atoms. Silver nitrate is also a low melting point compound, melting at temperatures as low as 210 ℃ and decomposing at 440 ℃ (Table 3) [47]. This melting point is close to the T₁₀ value of the 0.2Ag/Co_0.93Ce_0.07, implying a connection between the two values. Prior to the consumption of the reducing agent, the AgNO₃ in the mixture is in a molten state, such that Ag^δ+ migration facilitates soot-catalyst contact. This behavior is entirely different from that exhibited by various silver oxides. Ag₂O is decomposed after a single reaction trial, while AgO degrades above 100 ℃. Thus, silver oxides are unlikely to improve the contact conditions, although the mobility of silver has to be taken into account. Gardini et al. [48] studied the movements of silver particles via ETEM, to determine why silver-based catalysts maintain relatively high catalytic activity under loose contact conditions. In fact, while the mobility of silver is inferior to that of potassium salts [49], this mobility does accelerate soot combustion. In the present work, an EPR study was carried out to clarify the migration of silver (Fig. 7). The resulting plot is composed of two overlapping signals: S1 and S2. S1 can be ascribed to soot, and its intensity depends on the amount of radicals on the soot surface. S2 reflects the contact points between the soot and the catalyst. In most cases, S2 is very weak and so the contact conditions were assessed based on the intensity of the superimposed signals. Tuning of the pre-treatment temperature was found to modify the amount of volatiles on the soot surface, and higher temperatures led to the release of fewer radicals. The signal generated by the Co₃O₄/soot mixture gradually decayed with increasing temperature in an inert gas. In contrast, the signals of the silver-based samples suddenly increased at 200 ℃, demonstrating the contribution of the S2 peak. Above 200 ℃, the reduced signal intensity indicates that the increase in S2 was unable to compensate for the loss of S1 [50]. At higher temperatures, agglomerated silver particles tend to migrate, thus improving the mixture and generating more contact points. These metal particles have also been shown to exhibit different mobilities over various substrates [51].

The data obtained in the present study offer some clues to understanding the effects of silver on the reaction mechanism in the presence of O₂. The silver loading exerted a significant influence on soot ignition under O₂ without modifying the structure of the Co-Ce mixed oxides, demonstrating that the reaction occurred over the silver phase at low temperatures. The loss of Ag^δ+ in the presence of soot was confirmed by XPS, showing that oxygen species were transferred from the mixed silver oxides to the soot surface and subsequently assisted in soot removal. Ag^δ+ was evidently stabilized at the catalyst surface following durability tests, a result that can be attributed to the mutual transformation between mixed silver oxides and metallic silver during soot removal. It can be inferred from these data that Ag^δ+ is self-regenerated. Many researchers have reported that the formation of a relatively stable AgO_y layer is followed by the adsorption and dissociation of molecular oxygen [25, 52]. On the basis of the aforementioned analyses, a mixed silver oxides layer (mainly Ag₂O) evidently plays a crucial role as an oxidizing agent during soot combustion under O₂. The properties of this layer are similar to those of bulk Ag₂O, in which nucleophilic oxygen acts as an excellent electron donor. In addition, the silver oxides promote the adsorption of reactants to form complex π bonds that are essential for the formation of peroxide and superoxide species [53].

The proposed reaction mechanisms and pathways can be summarized by two stages distinguished by the decomposition of silver oxides (AgO_y). In the first stage, AgO_y not only reacts with soot but also favors the formation of active oxygen species before its decomposition. AgO_y initially decomposes and its released oxygen migrates to soot surfaces to form carbon-oxygen intermediates (R1). Meanwhile, active oxygen species attack soot to form carbon-oxygen intermediates (R2 and R3), and these intermediates subsequently undergo further oxidation (R4). Finally, the adsorbed oxygen on the silver, whose properties are identical to those of bulk Ag₂O [53], promotes the regeneration process (R5). The active oxygen species and the redox reaction of AgO_y dominate the reaction at low temperatures. A "clean" silver surface is obtained at a higher temperature because AgO_y is unlikely to persist at such temperatures. The dissociated active oxygen species are the key to these reactions (R6), as these substances accelerate soot combustion throughout the heating process (R7 and R8). In addition, higher amounts of Ag^δ+ explain the superior performance of the materials with more silver. Thus, silver oxides situated on the silver surface in the presence of O₂ are critical oxidizing agents and can be regenerated at low temperatures. Both oxygen spillover and the MK mechanism (that is, the redox mechanism) are involved in this reaction.

Low-temperature pathways:

(R1)

(R2)

(R3)

(R4)

(R5)

High-temperature pathways:

(R6)

(R7)

(R8)

Nitrate species, observed at low temperatures under NO_x, are likely to play a significant role in soot abatement. Unfortunately, it is difficult to quantify the amounts of nitrate species, although their effects can be categorized into the following three aspects [54, 55].

(Ⅰ) A low melting point salt (such as AgNO₃) promotes mobility and thus increases the extent of soot-catalyst contact. The mobility of AgNO₃ should therefore be considered based on its melting point. The minor gap between the ignition temperature and the melting point of AgNO₃indicates a possible close correlation. As silver loading reaches a certain value, the three-phase contact points (between AgNO₃, NO_x and soot) become saturated. Further introduction of silver shows no effect owing to the limited amount of AgNO₃ that can be formed and/or the weakened synergistic effect between the silver and metal oxides. It is believed that the mobility of the AgNO₃ is likely a crucial factor during soot gasification in the presence of NO_x.

(Ⅱ) Nitrate species are reduced to nitrites and subsequently assist in soot combustion. The reaction occurs at the boundary between soot, nitrate and NO_x. In the first step, nitrate is reduced to nitrite, accompanied by the formation of intermediate compounds in reactions between soot and NO_x. In the next step, the nitrite is oxidized to nitrate and the intermediates are eventually gasified to CO₂[56]. AgNO₃ acts as an oxidizing agent in this case, and the oxidation properties of AgNO₃ are discussed in the following paragraphs.

(Ⅲ) The decomposition product (NO₂) catalyzes soot removal. Previous reports have pointed out that the decomposition process forms NO₂, which is active for soot gasification. In the present study, the thermal stabilities of Co (NO₃)₂ and Co (NO₃)₃ were less than that of AgNO₃. Thus, cobalt-based nitrate salts might make significant contributions. Considering that silver significantly lowers the light-off temperature, this decomposition effect is unlikely to be the dominant factor during soot ignition. However, released NO₂ promotes soot abatement well above T₁₀.

Unlike most metal oxides, these xAg/Co_0.93Ce_0.07 catalysts perform even better in O₂ than in NO_x. The transitional oxidizing agents (silver oxides and silver nitrates) are the main factor producing this result. The AgO_y stabilized in the presence of the O₂ atmosphere is quickly converted to AgNO₃ and/or other nitrate salts with the assistance of NO_x. The disappearance of AgO_y subsequently decreases the catalytic performance. It can be inferred that certain rapid reactions that normally make up the catalytic process do not proceed in the absence of AgO_y, or rather that AgNO₃ is not as active as AgO_y. This effect can be seen by examining the results shown in Fig. 8. These differential thermogravimetry (DTG) data demonstrate that Ag₂O, the primary phase of AgO_y, attacks soot at a lower temperature than does AgNO₃ in both O₂-rich and O₂-free atmospheres. Ag₂O thus appears to be a stronger oxidant than AgNO₃. Hence, the unusual catalytic behavior of this material largely results from the activity of silver-based intermediates.

Highly active xAg/Co_0.93Ce_0.07 composite catalysts that generate ignition temperatures below 200 ℃ were explored for the purpose of catalytic soot combustion. Silver significantly enhanced the catalytic behavior in various atmospheres. In the presence of O₂, a low temperature redox cycle evidently occurs between the soot and the silver oxides situated on the surface of metallic silver. In the NO_x-rich atmosphere, these active silver oxides transform into AgNO₃, which is less active than silver oxides. The difference in activity between the two transition state oxidization agents, Ag₂O and AgNO₃, accounts for the unconventional catalytic performance of these materials in the presence of O₂ or NO_x. Overall, these highly active Ag/Co-Ce catalysts show promise as potential candidates for CDPF coating layers. We therefore anticipate that the unique properties of silver could be used in the future development of catalytic materials for soot abatement or even VOC oxidation, resulting in low light-off temperatures.