Study on oxidation activity of Ce–Mn–K composite oxides on diesel soot

As an effective method, diesel particulate filter (DPF) technology has a great contribution in reducing soot emissions from diesel engines. To achieve passive regeneration of DPF at low temperatures, K-doped Ce0.5Mn0.5O2 catalysts were synthesized using sol–gel method. The effect of K-doped catalysts-Kz–Ce0.5Mn0.5O2-on the oxidation of soot had been studied by thermogravimetric analysis, and the corresponding catalytic properties were evaluated based on X-ray diffraction (XRD), hydrogen temperature programmed reduction (H2-TPR), O2 temperature programmed desorption (O2-TPD) Raman spectroscopy (Raman), Brunauer–Emmett–Teller (BET) and Fourier-Transform-Infrared (FTIR).The results showed that K doping facilitated the oxidation of diesel particulate matter, which was indicated by the entire mass loss curve shifting to lower temperatures. K0.2–Ce0.5Mn0.5O2 showed the best performance among the series of K-doped catalysts. Compared with the findings for Ce0.5Mn0.5O2, the ignition temperature of soot oxidation (Ti) had been lowered by 28 ℃, and the maximum peak combustion temperature (Tm) of the dry soot decreased by 61 °C. Furthermore, compared with the Ce0.5Mn0.5O2-catalyzed reaction, K doping led to a lower activation energy and significantly improved pre-exponential factor. The minimum reaction activation energy of 27.46 kJ/mol was exhibited by K0.2–Ce0.5Mn0.5O2.

www.nature.com/scientificreports/ with Fe, Zr, La, Pr, and Nd show stronger catalytic activity than CeO 2 alone 24,25 . Kohn et al. 26 doped CeO 2 with Sm and La via coprecipitation, and the doped CeO 2 material was found to display a smaller subgrain size and larger specific surface area. By doping CeO 2 with Mn and Cu, March et al. 27 found that Mn ions entered the CeO 2 lattice to form a solid solution, which exhibited a larger number of oxygen vacancies and a significantly increased concentration of surface adsorbed oxygen. The Cu ions were found to disperse on the surface of the CeO 2 fluorite structure, and the interaction between Cu and Ce enabled rapid release of lattice oxygen in a reducing atmosphere. The formation of the Ce-Mn solid solution promotes the mobility of oxygen species, and leads to a significantly increased number of oxygen vacancies. Studies have shown that the catalyst maintains the basic cubic fluorite structure of CeO 2 when the Ce/Mn ratio is greater than 1 28 . To address the high-temperature sintering of CeO 2 , Hemeryck et al. 29 doped Ce-Mn catalyst with Ba, which prevented the separation of the Ce-Mn phases and was found to be effective for suppressing the sintering of oxides.
Our previous studies show that Mn enters the crystal lattice of CeO 2 to form a Ce-Mn solid solution 30 . The introduction of Mn induces valence state variation of Ce 4+ /Ce 3+ , and increases the number of surface oxygen vacancies. The synergistic effect of Ce and Mn enhances the catalyst selectivity for soot oxidation, leading to significantly enhanced catalytic activity. Ce 0.5 Mn 0.5 O 2 was selected as the base material as it displayed the highest activity in our previous research. To further lower the soot oxidation temperature, K was doped into Ce 0.5 Mn 0.5 O 2 , which was found to exhibit the highest activity in our previous research. The sol-gel method was used to synthesize a series of K z -Ce 0.5 Mn 0.5 O 2 catalysts. To determine the optimal K doping ratio, the K-doped catalysts were characterized and their catalytic activities were evaluated. The findings revealed the influence of K doping on the crystal structure of Ce-Mn catalyst and the oxygen species mobility, as well as the consequent valence variation of the Ce and Mn ions. Furthermore, the effect of K doping on diesel particulate matter oxidation was analyzed, providing reference for future studies on the catalyzed reaction of diesel particulate matter and catalyst coating of DPF.

Experimental
Catalyst preparation. The group of K z -Ce 0.5 Mn 0.5 O 2 catalysts was synthesized using the sol-gel method.
For a typical synthesis, Mn(NO 3 ) 2 and Ce(NO 3 ) 3 ·6H 2 O with a particular stoichiometric ratio were thoroughly mixed in deionized water and then KNO 3 was introduced into the mixture. The amount of KNO 3 required was calculated according to the ratio of K in the final product. After the solution was magnetically stirred at a constant temperature for 5 min, citric acid of the equivalent mole amount to that of metal cations in the solution was added. The mixed solution was then ultrasonicated for 10 min, before being placed in a water bath at a constant temperature of 80 °C. The mixed solution was magnetically stirred in the water bath until it formed a gel. The newly-formed gel was dried overnight at 120 °C in a blast dryer and a muffle furnace was utilized to calcined the dried product for 4 h. The catalysts synthesized were denoted K z -Ce 0.5 Mn 0.5 O 2 (z = 0.1, 0.2, 0.3), with the specific formulas being K 0.1 -Ce 0.5 Mn 0.5 O 2 , K 0.2 -Ce 0.5 Mn 0.5 O 2 , and K 0.3 -Ce 0.5 Mn 0.5 O 2 , separately.
Catalyst characterization method. The crystal structure of the catalyst samples was analyzed by the Burker D8 Avdance X-ray diffraction (XRD). The instrumental test conditions were as follows: running at specific condition (40 kV, 30 mA), utilizing Cu Kα radiation (λ = 0.15418 nm) filtered with nickel. The scanning range of small angle diffraction was 2 = 0.8°-8° and the scanning step length was 0.002°. The scanning range of wide angle diffraction was 2 = 10°-80° and the scanning step length was 0.02°.The scanning step is 4°/min. H 2 -temperature programming reduction (H 2 -TPR) was performed on a Chemisorb 2720 pulse chemisorption system equipped with a TPx (temperature-programmed controller and software) system and a TCD detector (Micromeritics). 10 mg of the sample was heated up to 700 ℃ from room temperature with a heating addition rate of 10 ℃/min. The reducing atmosphere gas (the mixture of 10 vol% H 2 and N 2 ) was supplied at a flow rate of 25 mL/min 31 .
The TPDRO 1100 instrument was used for the O 2 temperature-programmed desorption (O 2 -TPD) O 2 -TPD experiment. The specific experimental steps are as follows: 200 mg of catalyst was weighed into the sample cell; then, the temperature was raised to 600 °C at a heating rate of 15 °C/min in an O 2 atmosphere of 20 mL/min, and the temperature was maintained for 30 min. Finally, after the temperature drops to room temperature, oxygen desorption rises 900 °C at 10 °C/min in He atmosphere from room temperature. Same conditions were applied in these O 2 -TPD experiments to control variable and maintain accuracy. A TCD detector was used to characterize and observe the concentration signals of the desorbed O 2 .
The Raman spectrometer used in the experiment was a Renishaw micro Raman spectrometer produced by Renishaw, Germany. It uses a CCD multi-channel detector, an excitation light source is a 633 nm He-Ne ion laser, and the resolution of the Raman dynamic line is 0.5/cm. The N 2 adsorption-desorption isotherm was determined on the ASAP2020 physical adsorption apparatus. Before the test, the catalyst was fully ground in a mortar, about 0.1 g of the sample was weighed, and a predegassing treatment was performed under a vacuum condition at 200 °C for 10 h. Then, the liquid nitrogen was used as the adsorption medium, the adsorption and desorption volumes of the catalyst for N 2 at different pressures (relative pressure range 0.0-1.0) were tested at 77 K to obtain the adsorption-desorption isotherms of the samples.
The Fourier-Transform-Infrared (FTIR) spectrum of the sample was recorded on a Bruker Tensor 27 spectrometer. The instrument resolution was 4/cm and 32 scans were accumulated. The in situ pool without sample was scanned in He atmosphere to obtain the background spectrum. The composite catalyst with a load of K adsorbed carbon smoke was pressed into self-supporting sheets with a thickness of 7.5 mg/cm 2 , which were fixed in the in-situ infrared transmission pool, which was connected to the air path system and heated to 500 ℃. After the sample was pretreated at 200 ℃ for 1 h and reduced to room temperature, 100 mL/min of high purity He gas Scientific RepoRtS | (2020) 10:10025 | https://doi.org/10.1038/s41598-020-67335-5 www.nature.com/scientificreports/ was injected, and the temperature was raised to 500 ℃ at 5 ℃/min, and the temperature was kept at 500 ℃ for 3 h, during which the FTIR spectrum was recorded.
Catalytic testing. T SOF , T pre , T m , and T i were used to evaluate the catalyst activity in this paper.They indicate the temperatures corresponding to the mass loss peaks of the soluble organic fraction (SOF), soot precursor, and dry soot, and the soot ignition temperature (the temperature corresponding to 10% soot mass loss), respectively. In Table 1, the main technical parameters of the test diesel engine were listed. TGA/DSC1 thermogravimetric analyzer from Swiss METTLER company was used for the thermogravimetric analysis (TGA) of particle samples, with a high precision microgram electronic balance and temperature sensor. The PM samples data was collected by AVL SPC472 partial flow point collection system. An oxygen atmosphere with a mass fraction of 12% was selected for thermogravimetric test, which was close to the oxygen content in diesel exhaust. N 2 in high-purity had been performed as the protective gas, and its flow velocity was at a constant rate of 100 mL/ min under the setting temperature from 40 to 800 ℃, and the heating addition rate was maintained at 15 ℃/ min. The sample weight was 3 mg. Mixed at the ratio of 4:1 evenly, the tested diesel soot particles and catalyst became an uniform mixture.

Results and discussion
Activity tests. The thermogravimetric (TG) and differential thermogravimetric (DTG) curves of diesel particulate matter with the catalysts are presented in Fig. 1. It is obvious that the mass loss rate peaks of the soluble organic fraction (SOF), soot precursor, and dry soot, shifted significantly to lower temperatures with the doping of K. The introduction of K significantly lowered the oxidation temperature of diesel exhaust particulates compared with the catalytic activity of Ce 0.5 Mn 0.5 O 2 . The catalyst K 0.2 -Ce 0.5 Mn 0.5 O 2 (z = 0.2) displayed the best catalytic effect on diesel soot oxidation.
The weight loss characteristics of diesel particulate matter in the presence of various catalysts are shown in Table 2. It is noticeable in the table that K doping reduced the oxidation temperature of SOF. Specifically, with increasing K content, the SOF oxidation temperature decreased by 11, 19, and 18 °C, respectively. This indicates that the ability of the Ce-Mn solid solution to oxidize diesel particulate matter at low temperatures was improved   Table 3. Doping with K led to a decreasing trend in reaction activation energy, implying that the energy required for catalytic oxidation of soot was lower with K-doped catalyst, which resulted in easier soot oxidation. The minimum observed activation energy of 27.46 kJ/mol was achieved for the reaction catalyzed by K 0.2 -Ce 0.5 Mn 0.5 O 2 , which is about 20 kJ/ mol lower than that reported in relevant literature 33,34 . Moreover, the pre-exponential factor of soot oxidation was found to increase significantly with K-doped Ce 0.5 Mn 0.5 O 2 as the catalyst. A larger pre-exponential factor represents more effective collisions between the catalyst and soot during the reaction process, which would facilitate soot oxidation.  In addition, Ce 4+ was partially substituted by K + during the doping process, which was accompanied by the transition of electrons between ions, generating oxygen vacancies. Figure 3 also indicates that upon K doping, the characteristic peaks of KNO 3 appeared at 2θ = 23.5° and 41.8°. With the continuous increase of K content, a peak at 2θ = 33.8°, which is also attributed to KNO 3 , was observed for K 0.3 -Ce 0.5 Mn 0.5 O 2 . Meanwhile, the characteristic diffraction peak at 2θ = 15.5° of a new compound K 2 Mn 4 O 8 also appeared for catalysts with higher K content. The K 2 Mn 4 O 8 phase was absent and the peak positions of the Mn oxide phase were shifted to lower angles for the catalyst with low K content (K 0.1 -Ce 0.5 Mn 0.5 O 2 ). The formation of K 2 Mn 4 O 8 at elevated K content is attributed to the fact that in addition to forming the tetrahedral structure with Ce, K also combines with Mn oxides and dissolves in the solid solution. Therefore, at higher K concentrations, in addition to the K species covering the surface of the solid solution, residual K combined with the Mn oxide to form a new K 2 Mn 4 O 8 phase.
The redox capacity is an important indicator of the catalytic performance of a catalyst, particularly in cases where the catalyst is applied for catalytic oxidation of diesel particulate matter. H 2 -TPR is an effective technique that reflects the reducing ability of a catalyst. Comparing the H 2 -TPR profiles of Ce 0.5 Mn 0.5 O 2 and K z -Ce 0.5 Mn 0.5 O 2 (Fig. 4) revealed that the catalytic peak positions varied depending on the amount of K doped.  www.nature.com/scientificreports/ The reactive oxygen species present can be characterized using the temperatures corresponding to the reduction peaks, and the amount of reactive oxygen species represents the catalytic ability of the catalyst in the oxidation of particulate matter. In our previous work, Ce 0.5 Mn 0.5 O 2 displayed relatively good reduction peak positions in the range of 100-500 °C, indicating high activity, which was a result of the conversion and electron transition between Mn 4+ /Mn 3+ and Ce 4+ /Ce 3+ ion pairs 32 . The peak positions of the H 2 -TPR profile varied with increasing K content (Fig. 4). Compared with the peak positions of Ce 0.5 Mn 0.5 O 2 at 240 and 381 °C, the corresponding peak temperatures of K 0.1 -Ce 0.5 Mn 0.5 O 2 dropped to 226 and 366 °C, respectively. Therefore, K doping resulted in the shift of the reduction peaks (< 400 °C) to lower temperatures. The K + ions doped into the material substituted some of the Ce 4+ ions, which caused greater conversion of Mn 3+ to the higher valence state (Mn 4+ ), leading to a gradually increasing Mn 4+ /Mn 3+ ratio with higher K + content. Consequently, the surface charge of the catalyst became unbalanced, and the mobility of lattice oxygen was enhanced. The generation of oxygen vacancies then enhanced the adsorption of reactive oxygen species by the catalyst 35 . The H 2 reduction peaks indicated that the K + ions remained relatively stable without going through valence state conversion. Higher degrees of K doping led to a mounting in the amount of oxygen species adsorbed by the catalyst and significantly improved oxygen mobility. It is speculated that for the K-doped catalysts the active oxygen species on K sites could spill over to the soot surface, and react with the free carbon sites to form ketene species with C=C=O structure; the active oxygen species at K sites were then continuously supplied with gaseous oxygen via mobile lattice oxygen until the particulate matter was fully oxidized. This is thought to be caused by the improved stability among atoms induced by the small amount of K during the calcination of the catalyst, which inhibited the desorption capability of the adsorbed oxygen species. The oxygen desorption intensity was significantly enhanced with further increase of the K content. When z = 0.2, the adsorbed atomic oxygen (O − ) actively desorbed, displaying a distinct oxygen desorption peak at ~ 480 °C. In addition, the lattice oxygen desorption at 700 °C was also intensified.
The Raman spectra of Ce 0.5 Mn 0.5 O 2 and K z -Ce 0.5 Mn 0.5 O 2 catalysts are shown in Fig. 6. Ce 0.5 Mn 0.5 O 2 exhibited a typical cubic fluorite structure, indicated by a vibration peak (447/cm), which was attributed to the typical F 2g vibration of the CeO 2 cubic fluorite structure. The peak shifted slightly upon K doping, however, the overall cubic fluorite structure of CeO 2 was not altered, which is accordant with the XRD results. In our previous work we reported that the vibration peak at 641/cm was caused by Mn entering the CeO 2 lattice. Characteristic peaks at the same position were also observed for K z -Ce 0.5 Mn 0.5 O 2 catalysts, which were attributed to the vibration of Mn-O. However, the peaks were broader and showed a slight red shift. This observation suggests the existence of a small amount of MnO x in the K z -Ce 0.5 Mn 0.5 O 2 catalysts, and that K doping led to the variation of the structural www.nature.com/scientificreports/ valence between Mn and Ce, which subsequently resulted in a change in the amount of oxygen vacancies, facilitating the migration of oxygen species on the catalyst surface and promoting the catalytic combustion of soot. Table 4 exhibits the information of the three CeO 2 samples, which are surface area, pore volume and average pore diameter. The specific surface area of the sample was calculated using BET (Brunauer-Emmett-Teller) method, and the pore volume and aperture were calculated by the isotherm adsorption branch using BJH (Barrett-Joyner-Halenda) model, where the pore volume was calculated using the adsorption volume at the relative pressure p/p0 = 0.99. It can be seen from Table 4 that the surface area of the prepared K z -Ce 0.5 Mn 0.5 O 2 is significantly greater than that of Ce 0.5 Mn 0.5 O 2 . The larger surface area indicates more surface active sites per unit mass of K z -Ce 0.5 Mn 0.5 O 2 . Furthermore, there are more opportunities to achieve a closer contact between the catalysts and the reactants.
In situ FTIR results are shown in Fig. 7. It is found that ketene species may exist in this reaction (1388/cm). At first, the mixture of soot and K 0.2 -Ce 0.5 Mn 0.5 O 2 was heated to 430 °C in O 2 + He followed by cooling down to 200 °C with purging with He. In this step, some soot was depleted and thus a clear FTIR signal and lots of free carbon sites were obtained. The corresponding spectra are shown in Fig. 7 (20   www.nature.com/scientificreports/ a large amount of NO, the real exhaust atmosphere is simulated. NO was introduced and switched off when the spectrum did not change significantly. As expected, the band of the ionic nitrite was observed in Fig. 7 (40 min). At this time, free carbon sites and ionic nitrite were abundant on the surfaces of soot and K 0.2 -Ce 0.5 Mn 0.5 O 2 , respectively. The mixture was progressively heated up to higher temperatures in He (60 and 80 min). During this period, the band of the ionic nitrite gradually decreases in intensity, simultaneously with the formation of the ketene group. These facts suggest that the ionic nitrite may be consumed with the production of the ketene group. In other words, the ionic nitrite on K 0.2 -Ce 0.5 Mn 0.5 O 2 interacts with the free carbon sites on the soot to form the ketene group, which can be described as 36 : The ketene group has been identified as the intermediate of soot oxidation with O 2 or NO 2 , which is a surface oxygen complex formed on the surface of soot with graphite structures. Likewise, the ketene group can serve as the intermediate of soot oxidation with NO. During this process, chelating bidentate carbonate (1,257/cm) and ionic carbonate (1,091/cm) are formed, which have been observed in soot oxidation with O 2 . These carbonates originate from the adsorption of the produced CO 2 on potassium sites. Catalytic characteristics. The catalytic oxidation of soot using K-doped Ce 0.5 Mn 0.5 O 2 as the catalyst was analyzed by relating the characterization results indicating variations in structure and surface ions, to the redox capacity of K-doped Ce 0.5 Mn 0.5 O 2 , as shown in Fig. 8. For the K-doped Ce 0.5 Mn 0.5 O 2 catalysts, the active oxygen species on the K sites spill over to the soot surface and react with the free carbon sites to form ketene species with C=C=O structure. Meanwhile, the K sites enable the supplementation of active oxygen species in the catalyst by activating surrounding gaseous oxygen and enhancing the mobility of lattice oxygen until the soot is fully catalytically oxidized to CO 2 37 . The activated oxygen on the surface of K site may overflow to the free carbon site on the soot, forming a carbon-oxygen complex, that is reaction intermediate, ketene group. The K effect is used to supplement the consumed surface oxygen by chemical adsorption and dissociation of gas-phase oxygen or surface lattice oxygen. In the absence of transient reactions, carbothermal reduction and gas phase oxygen, surface lattice oxygen participates in soot combustion. The ketene group is further oxidized to carbon dioxide by other active oxygen, which increases the number of exposed free carbon sites. The selectivity of soot combustion is due to the fact that these free carbon can be directly oxidized into CO by gas phase oxygen. The number of active sites increases with the increasing of K, which will occupy more free carbon sites and avoid combining with gas phase O 2 to form CO, resulting in a small increase in CO 2 selectivity. What's more, K can promote the escape of oxygen to soot by forming ketene group.

conclusions
In the current study a series of K-doped Ce 0.5 Mn 0.5 O 2 catalysts were synthesized using the citrate sol-gel method. Their ability to catalyze diesel particulate matter oxidation was investigated by thermogravimetric analysis. The activation energy and pre-exponential factor of each reaction were determined by further analysis of particle oxidation kinetics. The catalytic mechanism was explored on the microstructure level by means of XRD, H 2 -TPR, O 2 -TPD, and Raman. The following conclusions were reached: (1) K doping promoted the oxidation of diesel particulate matter, which manifested as the entire mass loss curve shifting to lower temperatures. Compared with the Ce 0.5 Mn 0.5 O 2 -catalyzed reaction, the ignition temperature of soot decreased by 17, 28, and 26 °C, with K z -Ce 0.5 Mn 0.5 O 2 (z = 0.1, 0.2, 0.3, respectively) as the catalyst. While, the maximum peak combustion temperature of dry soot was also lowered by 45, 61, and 57 °C, upon K doping (z = 0.1, 0.2, and 0.3, respectively). The prepared K z -Ce 0.5 Mn 0.5 O 2 has a larger specific surface area and the catalytic activity of K z -Ce 0.5 Mn 0.5 O 2 increases with its specific surface area increasing. Doping with K significantly enhanced the catalytic activity of the Ce-Mn solid solution for special matter oxidation and reduced the oxidation temperature of soot.
(1) C = C * + K + − NO − 2 → C = C = O + NO + K + − www.nature.com/scientificreports/ (2) Analysis of the oxidation kinetics of diesel particulates indicated lower activation energies and increased pre-exponential factors for reactions catalyzed by K-doped Ce 0.5 Mn 0.5 O 2 . The minimum observed activation energy of 27.46 kJ/mol was achieved using K 0.2 -Ce 0.5 Mn 0.5 O 2. The oxygen species on K sites played an important role in soot oxidation, and were continuously supplemented by activating surrounding oxygen molecules to complete the oxidation process. (3) The K z -Ce 0.5 Mn 0.5 O 2 catalysts synthesized using the citrate sol-gel method displayed the original cubic fluorite structure of CeO 2 . Therefore, material structures remained stable upon K doping. However, the valence states of Ce and Mn ions were altered by K doping. The amount of oxygen vacancies on the catalyst surface increased when the flow mode of the oxygen species diversified. (4) K doping also resulted in more adsorbed oxygen species and significantly improved oxygen mobility. In addition, the structural balance between Mn and Ce was also altered by K doping, which resulted in changes of the amount of oxygen vacancies, facilitating the migration of oxygen species on the catalyst surface and promoting the catalytic combustion of soot.