Study on the Evolution of Physicochemical Properties of Carbon Black at Different Regeneration Stages of Diesel Particulate Filters Regenerated by Non-Thermal Plasma

Abstract

As a new type of aftertreatment technology, non-thermal plasma (NTP) can effectively decompose the particulate matter (PM) deposited in diesel particulate filters (DPFs). In this paper, a regeneration test of a DPF loaded with carbon black was carried out using an NTP injection system, and the changes of oxidative activity, elemental content, and occurrence state, microstructure and graphitization degree of carbon black were analyzed to reveal the evolution of the physicochemical properties of carbon black at different regeneration stages of the DPF regenerated by NTP. As the regeneration stage of the DPF advanced, Ti, Tmax, and Te of the carbon black at the bottom of the DPF decreased, which were higher than those at the regeneration interface. After the NTP reaction, the proportion of C element decreased to less than 80%, while the proportion of O element increased to more than 20%; C-O was converted to C=O and the relative content of C=O increased. The average microcrystalline length and average spacing decreased, while the average microcrystalline curvature increased. The I_D1/I_G (relative peak intensities) of carbon black samples decreased from 3.31 to 3.10, and the R3 (relative peak intensities, R3 = I_D3/(I_G + I_D2 + I_D3)) increased from 0.41 to 0.58. The content of carbon clusters had a great influence on the disorder of the microcrystalline structure, so the graphitization degree of carbon black decreased and the oxidation activity increased.

1. Introduction

Diesel engines have the advantages of high torque, good economic performance, low speed, and high safety, and are widely used in industrial fields such as construction machinery, large cargo transportation, and electric power [1,2]. However, particulate matter (PM) emitted from diesel engines can cause serious harm to the environment and human health [3]. PM can cause a variety of health problems such as eye irritation, nasal allergy, coughing, and breathing difficulties in humans. Prolonged exposure to PM may also lead to serious respiratory diseases and heart diseases [4]. Diesel particulate filters (DPFs) can effectively capture PM and prevent it from entering the atmosphere, and are currently the most effective solution [5]. However, over time, a DPF’s channels will be blocked by PM and the exhaust back pressure increases. When the back pressure of the diesel engine exceeds a certain limit, its working performance decreases significantly and the DPF needs to be regenerated [6].

The strong oxidizing substances in a non-thermal plasma (NTP) have been shown to be effective in decomposing PM in diesel engines [7,8]. Ma et al. [9] introduced exhaust gas with PM into an NTP generator for reaction. It was shown that the total number and mass of PM were reduced by 40% and 76.9%, respectively. And the purification rate of PAHs in the particulate phase was 58.4% after NTP treatment. Liu et al. [10] conducted removal tests of PM at lower temperature conditions via synergistic action of NTP and MnO_x/CeO₂ catalysts. It was shown that the removal efficiencies of PM reached 85.2% and 94.3% at 20 °C and 200 °C, respectively. Fushimi et al. [11] carried out experiments on the oxidative decomposition of PM in a dielectric barrier NTP reactor with different structures, and pointed out that O₃ and NO₂ were the main active substances for the oxidation of PM, and the decomposition products of PM were CO and CO₂. Based on this research, scholars have applied NTP technology to the field of DPF regeneration. Hongsuk et al. [12] tested regeneration of a DOC + DPF system by entering compressed air into the generator in a rotating manner to produce rotating NTP. It was shown that the rotating plasma generator had good durability under vehicle driving conditions, and the regeneration efficiency of a DPF containing 5.5 g/L carbon was 93%. Shi et al. [13,14,15] investigated the effect of initial regeneration temperature on a DPF regenerated by NTP, and found that the higher the temperature the lower the amount of PM decomposition, and that the peak temperature and temperature gradient inside the DPF were much lower than the failure limit value. The effect of ambient temperature on the regeneration performance of DPFs has been investigated, and NTP can effectively regenerate a DPF without a catalyst, demonstrating superiority compared to the conventional regeneration.

Carbon black, as a cheap and readily available material with similar physicochemical properties to PM, is gradually replacing diesel PM for DPF regeneration studies [16,17,18]. Naseri et al. [19] used a thermal gravimetric analyzer (TGA) and high resolution transmission electron microscope (HRTEM) to characterize carbon black samples untreated, and heat-treated to 1400 °C and 1600 °C. The oxidation of carbon black was found to follow the steps of surface oxidation, nanochannel formation, and microcrystalline oxidation. Tighe et al. [20] performed oxidation tests on diesel PM and carbon black at the temperature range of 450 °C to 550 °C. It was found that the internal channel structures of the spherical particles of both were consistent. During the combustion process, the channels of the primary carbon particles of both were gradually enlarged, while O₂ entered the interior of the particles to oxidize the carbon microcrystals. Meng et al. [21,22] conducted thermogravimetric analysis of carbon black and regeneration test studies on a DPF loaded with carbon black. The study showed that the oxidation process of carbon black went through the process of organic matter volatilization, amorphous carbon oxidation, graphitization, and combustion of carbon black particles, and the gas emission was mainly concentrated in the rapid regeneration period of the DPF. Fang et al. [23] loaded a mixture of carbon black and different masses of ash into the DPF for the regeneration test. It was shown that a larger ash loading was conducive to improving the regeneration efficiency of the DPF, and the regeneration efficiency increased by 47% when the ash loading was increased from 0 g/L to 10 g/L. Chen et al. [24] constructed an experimental system to regenerate a DPF loaded with carbon black by oxidizing NTP, and investigated the influence of NTP injection flow rate on the removal rate of carbon black and the regeneration of the DPF. It was shown that the DPF regeneration efficiency increased and then decreased with the increase of NTP injection flow rate. The highest DPF regeneration efficiency of 85.1% was achieved when the NTP injection flow rate was 30 L/min. Most previous research mainly focuses on improving the regeneration efficiency of DPFs. However, few research studies focus on the decomposition mechanism and evolution of carbon black during the DPF regeneration process.

In this study, a DPF loaded with carbon black was regenerated using an NTP generator and the carbon black was sampled at different stages of regeneration. The oxidation characteristics, elemental ratio and occurrence state, micro-structure, and graphitization degree of the carbon black were analyzed using TGA, X-ray photoelectron spectroscopy (XPS), HRTEM, and Raman spectra (RNS) to study the changes in physical and chemical properties of carbon black during the regeneration process of the DPF. The changes in physical and chemical properties were investigated to provide a theoretical basis and experimental basis for the evolution of carbon black by NTP oxidative decomposition.

2. Experimental System and Research Methodology

2.1. DPF Loaded Carbon Black Test System

Figure 1 illustrates the schematic diagram of the DPF loading carbon black device, which mainly includes filters, air compressor, vacuum pump, a DPF, and piping. The DPF (NGK Environmental Ceramics Co., Suzhou, China) used in the test is made of cordierite ceramic, and its structural parameters are shown in

Table 1. DPF structural parameters. Reproduced with permission from Xulong Chen, Yunxi Shi et al., Carbon Letters; published by MDPI, 2023 [24].

DPF	Parameter
Diameter × length/mm	144 × 152
Wall thickness/mm	0.46
Number of holes/cpsi	100
Volume/L	2.48
Material	cordierite
Maximum operating temperature/°C	1200

. The carbon black (Orion Inc., Espoo, Finland) is Printex-U, and its parameters are shown in

Table 2. Physical properties of carbon black. Reproduced with permission from Xulong Chen, Yunxi Shi et al., Carbon Letters; published by MDPI, 2023 [24].

Printex-U	Parameter
Particle size/nm	25
Ash/%	0.02
Volatile matter (950 °C)/%	5
BET/m2·g−1	92

.

2.2. NTP-Regenerative DPF Test System

Figure 2 illustrates the schematic diagram of the experimental NTP-regenerated DPF system, which includes an NTP generator injection system and a DPF regeneration system. The NTP generator had a coaxial cylindrical structure of dielectric barrier discharge type, the barrier dielectric layer was comprised of a quartz glass tube, and the stainless steel tube and the steel mesh were used as the low-pressure electrode and high-pressure electrode, respectively. The generator utilized a combination of internal water cooling and external air cooling, with an infrared thermometer monitoring the surface temperature in the NTP generator discharge region. The NTP generator was powered by a CTP-2000K intelligent electronic shock machine. The TDS3034B digital oscilloscope monitored the working voltage and frequency of the NTP generator. A voltage divider capacitor was used in the circuit to assist in the measurement of the discharge voltage, and the capacitance ratio of the voltage divider capacitor was C₁:C₂ = 1:1000. An oxygen generator controlled the flow of oxygen injected into the NTP reactor. Oxygen was ionized by the NTP generator and converted into strong oxidizing reactive particles (O₃, O, etc.). The DPF regeneration system consisted of a DPF regeneration device alongside testing and analysis instruments. A gas analyzer was used to monitor the volume fraction of CO_x, the decomposition product of carbon black decomposition product at the back end of the DPF. An O₃ analyzer was used to monitor the mass concentration of O₃ at the inlet and outlet ends of the DPF. The temperature inspector monitored the internal temperature of the DPF. Figure 3 illustrates the distribution of thermocouples inside the DPF.

2.3. DPF Regeneration Test Method

The oxidative decomposition of carbon black is exothermic, and the temperature of each measurement point inside the DPF illustrates a tendency of increasing and then decreasing with the regeneration of the DPF. When the temperature at the measurement point decreases, it can be assumed that the carbon black in the channel of the DPF has been completely oxidized and decomposed by the NTP active substance [25]. The flow rate of NTP was 30 L/min, and the initial regeneration temperature was 20 °C, and the initial loading of carbon black was 6 g/L. The DPFs were reacted separately at different regeneration stage, and the DPFs were dissected along the median axis, as shown in Figure 4. The carbon black samples were taken from the regeneration interfaces and the bottom of the DPF. Figure 4a illustrates the profile of DPF1, regenerated to 1/4; Figure 4b illustrates the profile of DPF2 regenerated to 1/2; Figure 4c illustrates the profile of DPF3 regenerated to 3/4; Figure 4d illustrates the profile of DPF4, which was basically regenerated completely and the retained portion was used for sampling. The red and orange boxes in Figure 4 are the carbon black sampling positions of the DPF regeneration interface and the bottom, identified by 1~6. The regeneration interfaces of DPF3 and DPF4 were close to the bottom, so the samples were taken from the same position.

Table 3. Maximum temperatures of temperature measurement points at the regeneration of carbon black samples.

Sample Name	Temperature Measurement Point Number	Temperature/°C
0 (Original machine carbon black)	\	\
1 (DPF1 bottom)	10	148.7
2 (DPF2 bottom)	10	183.0
3 (DPF3 bottom)	10	231.1
4 (DPF4 bottom)	10	239.3
5 (DPF1 interface)	1	241.3
6 (DPF2 interface)	4	272.8

illustrates the maximum temperature of regeneration at the carbon black sampling position, and sample 0 is the initial carbon black before the DPF regeneration. The maximum temperature at the sampling position at the bottom of the DPF increased as the regeneration stage of the DPF advanced, and the maximum temperature of the sampling position at the regeneration interface of the DPF was higher than that of the bottom at the same regeneration stage.

The thermogravimetric properties of the carbon black samples were analyzed using a TGA8000 thermogravimetric analyzer, and the thermogravimetric test procedure is shown in

Table 4. Thermogravimetric test procedure. Reproduced with permission from Xulong Chen, Yunxi Shi et al., Carbon Letters; published by MDPI, 2023 [24].

Step	Program Temperature	Atmosphere and Flow
1	50 °C for 1 min	Nitrogen 60 mL/min
2	Temperature increase to 450 °C at a rate of 10 °C/min	Nitrogen 60 mL/min
3	450 °C for 1 min	Nitrogen 60 mL/min
4	Cool down to 50 °C at a rate of 10 °C/min	Nitrogen 60 mL/min
5	50 °C for 1 min	Oxygen 20 mL/min + Nitrogen 40 mL/min
6	Increase the temperature to 750 °C at a rate of 10 °C/min.	Oxygen 20 mL/min + Nitrogen 40 mL/min
7	750 °C for 1 min	Oxygen 20 mL/min + Nitrogen 40 mL/min

. The program setting of N₂ atmosphere ensures the complete volatilization of the volatile fraction (VF) and the effective separation of the VF from the elemental carbon (EC) [26]. EscaLab 250Xi XPS (Thermo Fisher Scientific, Waltham, MA, USA) was used to determine and analyze the elemental species, relative content, and the fugitive states of the carbon black samples [27]. The morphology and structure of primary carbon particles of the carbon black samples were measured using a JEOL-2100 F transmission electron microscope (Thermo Fisher Scientific, Waltham, MA, USA). The magnifications of carbon black samples were 50,000 and 2 million times, respectively. The degree of graphitization of the carbon black samples was analyzed using a DXR-type laser Raman spectrometer (Thermo Fisher Scientific, Waltham, MA, USA). The testing process was performed for 30 consecutive scans with a spectral scanning range of 50 cm⁻¹ to 3400 cm⁻¹. Different regions of the same sample were tested to ensure the accuracy of the results of the test samples [28].

3. Test Results and Analysis

3.1. Oxidation Characteristics of Carbon Black Samples

Figure 5 illustrates the TG and DTG curves of the carbon black samples. The carbon black samples have less VF content, and three specific temperatures in the EC oxidation process can be chosen as the characteristic parameters for evaluating the oxidizing activity of the carbon black: 1. the ignition temperature Ti (the temperature corresponding to the initial mass loss of 10% during the oxidation reaction); 2. the maximum oxygen oxidation rate temperature Tmax (the temperature corresponding to the maximum mass loss rate is the lowest point of the DTG curve.); 3. the final combustion temperature Te (the temperature appearing in a mass loss of 90%).

The Ti, Tmax, and Te of carbon black sample 0 were 543.6 °C, 604.5 °C, and 640.3 °C, respectively. After the NTP reaction, Ti of the carbon black sample decreased and the Tmax and Te increased. As the DPF regeneration stage advances, the Ti, Tmax, and Te of carbon black samples 1~4 at the bottom position are decreasing. The Ti of carbon black samples 1~4 decreased from 513.5 °C to 492.8 °C, Tmax from 636.3 °C to 627.7 °C, and Te from 652.4 °C to 641.4 °C. The Ti, Tmax, and Te at the regeneration interface in the same regeneration stage of the DPF were less than those at the bottom sampling position. The lower the apparent activation energy of the carbon black samples, the easier the oxidation reaction occurs and the higher its oxidation activity. The apparent activation energy of carbon black was calculated by the Arrhenius equation [29,30], as shown in Equation (1).

$$\ln \left[-\frac{dm}{m_{0}dt}\right]=\ln \left[A\cdot P_{O_2}\right]-\frac{E_a}{RT}$$

(1)

where m₀ is the initial mass of the sample, mg; t is the length of time the sample is oxidized, s; A is the finger forward factor, s⁻¹; P_O2 is the partial pressure of O₂ in the heated atmosphere, Pa; E_a is the apparent activation energy, kJ/mol; R is the molar gas constant, 8.314 J/mol·k; and T is the oxidation temperature, K.

The activation energy of carbon black sample 0 before DPF regeneration was 179.89 kJ/mol. The activation energy of carbon black samples 1~4 at the bottom sampling position decreased from 156.5 kJ/mol to 125.7 kJ/mol, and the activation energies of carbon black samples 5 and 6 at the regeneration interface were 117.91 kJ/mol and 109.62 kJ/mol as the regeneration stage of the DPF advanced. The activation energy of sample 0 before DPF regeneration was the highest, and a higher ignition temperature was needed to achieve rapid oxidation of carbon black. The activation energies of carbon black samples 1~6 were all reduced after the NTP reaction. The activation energy of carbon black samples 1~4 at the bottom position of the DPF decreased continuously and the oxidation activity increased as the regeneration stage advanced. The activation energy of carbon black samples 5~6 at the regeneration interface was lower than those at the bottom sampling position. The NTP active substance was easy to decompose by heat, and in addition, the concentration of NTP active substances gradually decreased as they react with carbon black during the process of passing from the top to the bottom of the DPF. Under the combined effect of high NTP active substance and the reaction temperature, the oxidative decomposition of the regeneration interface between NTP and carbon black was more intense compared with that of the bottom, and so was the oxidation activity of carbon black. The oxidative decomposition reaction between NTP and carbon black at the regeneration interface was more intense than that at the bottom, and the oxidation process of carbon black was faster.

3.2. Elemental Content and State of Carbon Black Samples

Each carbon black sample was fully dried and the sample powder was ground, and then the sample was spread on double-sided adhesive and placed at the center of clean aluminum foil (wiped with acetone). Another piece of aluminum foil covering the sample was placed on the platform of the tableting machine to pull the pressure bar to 10 MPa, and the sample to be tested was made into a 1 cm × 1 cm tablet for a period of time. In the XPS narrow scan analysis, the carbon C1s with a binding energy of 284.6 eV was used to correct the measurement results by deducting the charge effect and performing multi-peak fitting.

Figure 6 illustrates the full spectrum of the XPS scan of the carbon black samples. The C and O elements in the carbon black samples were scanned narrowly, and the peaks were fitted using Origin2021 software (OriginLab Corp., Northampton, MA, USA). The occurrence state and content of C and O elements were analyzed according to the area of photoelectron peaks in the energy spectrum [31]. The significant elemental peaks were observed near 248 eV and 532 eV, corresponding to C1s and O1s spectra. The scanned full spectrum of each carbon black sample can be used to calculate its relative content ratios of C and O elements, as shown in

Table 5. Relative content of major elements in carbon black samples at different regeneration stages of DPF.

Sample Name	C/%	O/%
0	88.4	11.6
1	79.1	20.9
2	78.5	21.5
3	77.8	22.2
4	77.3	22.8
5	71.9	28.1
6	71.9	28.1

. The relative contents of C and O in sample 0 were 88.4% and 11.6% before regeneration. The relative content of C element in carbon black decreased to below 80% and the content of O element increased to more than 20% after the NTP reaction. As the reaction proceeded, the relative content of O element in carbon black samples 1~4 gradually increased from 20.9% to 22.8%. The relative content of O element at the regeneration interface at the same stage was higher than that at the bottom. The elevation of the relative content of elemental O in the samples indicates that part of C element combined with O element and existed in the carbon black in an incompletely oxidized state, so the oxidizing activity of carbon black was elevated [32], which was consistent with the analytical results of TG.

In order to analyze the oxidation state of carbon black more intuitively, the C and O elements of each sample were narrowly scanned and split-peak fitted to analyze the state of elemental assignment and its relative content. Figure 7 illustrates an example of the peak-fitting curve of the C1s elemental spectrum. The common assignment patterns of C elements are as follows: carbon–carbon and carbon–hydrogen bonds (C-C, C=C, C-H) in hydrocarbons, corresponding to electron binding energy peaks of 284.6 eV~285.0 eV; carbon–oxygen single bonds (C-O, C-O-H) in ethers and hydroxyls, corresponding to electron binding energy peaks of 285.3 eV~286.8 eV; carbon–oxygen double bonds in carbonyls (C=O), carboxyl carbon (O-C=O), corresponding to the peak electron binding energy of 286.3 eV~287.8 eV.

Table 6. Relative content of different fugacity states of C1s spectra in carbon black samples.

Sample Name	C-C/%	C-O/%	O-C=O/%
0	77.2	14.6	8.1
1	75.7	16.4	8.0
2	75.6	16.5	7.8
3	75.0	18.0	7.1
4	73.8	20.0	6.2
5	72.3	21.6	6.1
6	71.3	23.1	5.6

illustrates the relative content of C1s spectrum in the carbon black samples in different occurrence states. Most of the C atoms in carbon black were not bound to O atoms and this type of C atom in Sample 0 accounted for 77.2% of the total C atoms. In sample 0, there are C atoms bound to two O atoms, which accounted for 8.13% of the total C atoms. The C atoms content of samples 1–4 in the form of C-C and C atoms bound to two O atoms in the form of carboxyl groups and lipid bonds decreased and the content of C atoms bound to one O atom in the form of ethers, hydroxyl groups, and carbonyl groups increased by 3.56% after NTP oxidation reaction. The content of C atoms in the form of C-C at the regeneration interface was lower than at the bottom at the same stage. Due to the interaction of high-energy ions and free radicals in the NTP active substance, C atoms underwent oxidation reactions to produce oxygen-containing functional groups. These oxygen-containing functional groups could increase the oxidizing activity of carbon black and make it easier to react with NTP [33,34]. It was shown that after the reaction with NTP-reactive substances, C atoms bound with two O atoms were oxidized to CO₂, while some C-C bonds were oxidized to C-O bonds. The oxygen-containing functional groups of carbon black increased significantly, and the oxidative activity of carbon black was enhanced [35].

Figure 8 shows an example of the split-peak fitting curve of the O1s elemental spectrum. The O element existed in the form of carbon–oxygen single bond (C-O) and carbon–oxygen double bond (C=O), corresponding to the peak positions of the electronic binding energies of 531.0 eV~531.8 eV and 532.3 eV~533.0 eV, respectively [36].

Table 7. Relative content of different fugacity states of O1s spectra in carbon black samples.

Sample Name	C=O/%	C-O/%
0	62.7	37.3
1	60.7	39.3
2	59.4	40.6
3	59.4	40.6
4	58.7	41.3
5	55.2	44.8
6	53.1	47.0

illustrates the relative contents of different occurrence states of O1s spectra in carbon black samples. After the NTP oxidation reaction, the content of O atoms in the form of C-O in carbon black samples 1~4 gradually decreased, and the content of O atoms in the form of C=O gradually increased. The content of C atoms in the form of C-O at the regeneration interface was lower than that at the bottom at the same stage. This phenomenon indicates that the oxygen-containing functional groups are mainly present in carbon black in the form of C-O. As the regeneration stage progresses, C-O was converted to C=O, and the relative content of C=O in the carbon black sample increased.

3.3. Micro-Nano Structure of Carbon Black Samples

The carbon black samples at the regeneration interface and the bottom sampling position were immersed in anhydrous ethanol, and the suspension was prepared by centrifugal extraction for 20 min using an ultrasonic oscillator. A sampler was used to drop the suspension on a support membrane (copper material) for sufficient drying before shooting. The morphology and structure of primary carbon particles of the carbon black and PM samples were measured using JEOL-2100 F transmission electron microscopy.

Figure 9 illustrates the aggregated morphology and core-shell structure of primary carbon particles of the carbon black samples obtained from TEM. Carbon black consists of dozens to tens of spherical primary particles forming a chain or cluster flocculent structure, which is due to van der Waals force and electrostatic gravity between particles [37,38]. The stacking of primary carbon particles in carbon black samples 1 to 4 gradually decreased and evolved from a cluster floc structure to a chain structure as the DPF regeneration stage advanced. As the NTP regeneration of the DPF advanced, NTP active substance entered into the internal cluster flocculated structure of particles through free diffusion and forced convection. The weaker part of the cluster flocculated structure was the first one to break under the oxidizing effect of the NTP active substance, and the cluster flocculated structure decomposed into the chain-like structure [39]. The aggregated morphology of primary carbon particles of carbon black samples 5~6 showed a clear chain-like structure. The surface charge of some carbon particles increased after being oxidized by NTP, leading to an increase in the electrostatic attraction between the particles. And this interaction caused the carbon particles to be more tightly packed together, thus reducing the particle size of the aggregated state [40].

The primary carbon black particles are spherical core-shell structures with diameters of about 20 nm~35 nm, with the outer shell consisting of longer microcrystals stacked layer by layer to form a graphite layer-like structure, and the inner core consisting of disordered microcrystals stacked haphazardly with shorter lengths. The primary carbon particles in Sample 0 had more amorphous carbon components in the core, and the boundary of the outer shell was obvious and clearly distinguishable. The diameter of primary carbon particles in carbon black samples 1~4 decreased as the DPF regeneration stage advanced, and the diameter of primary carbon particles at the regeneration interface was smaller than that at the bottom at the same regeneration stage, and the boundary between the kernel and the shell was gradually blurred. The presence of unpaired sp₂ electrons at the edges of carbon microcrystals made it easier for them to combine with C atoms in the NTP actives, eroding the edges of the carbon particles, and the C is removed as CO or CO₂, while some of the long microcrystals were cleaved into short microcrystals [41,42]. Figure 10 shows the schematic diagram of extracting characteristic parameters of primary carbon particle microcrystals.

Figure 11 illustrates the distribution of microcrystal length (La) in primary carbon particles in the carbon black samples. The average value of La in sample 0 was 0.48 nm, and 80% of the microcrystalline size was concentrated in the range of 0.19 nm~0.97 nm. The microcrystalline lengths of carbon black samples 1~4 gradually decreased and the average value of La in samples 1~4 after oxidation gradually decreased from 0.33 nm to 0.27 nm as the regeneration stage advanced. The range of microcrystal length distribution in the primary carbon particles also significantly reduced, with a significantly higher percentage of microcrystals with lengths ranging from 0.12 nm to 0.80 nm. The presence of unpaired sp₂ electrons at the microcrystalline edges of the primary carbon particles made it easier for the carbon particle components to combine with the O atoms in the NTP active substances and be removed from the surface of the primary carbon particles in the form of CO or CO₂. NTP preferentially reacted with the physically defective parts of the longer microcrystals to form shorter microcrystals. Therefore, the average microcrystal length was significantly reduced after NTP treatment [43]. The microcrystal length at the regeneration interface at the same stage was smaller than that at the bottom, and the average values of La in samples 5 and 6 were 0.26 nm and 0.22 nm. The oxidative decomposition reaction between NTP and carbon black at the regeneration interface was more intense compared with that at the bottom, and more long microcrystals were cleaved to form shorter microcrystals, so that the length of the microcrystals was shorter than that at the bottom.

Figure 12 illustrates the distribution of the dimension spacing (d) of microcrystals in primary carbon particles in the carbon black samples. The dimension spacing of microcrystals in primary carbon particles in sample 0 before DPF regeneration showed a single-peak distribution, with the peaks concentrated in the range 0.50 nm~0.55 nm, the average dimension spacing was 0.47 nm, and the maximum dimension spacing was not more than 0.60 nm. As the regeneration stage of the DPF advanced, the average dimension spacing of samples 1~4 gradually decreased from 0.46 nm to 0.42 nm. At the same stage, the layer spacing at the regeneration interface was smaller than at the bottom, and the average values of d in samples 5~6 were 0.41 nm and 0.40 nm. After the NTP reaction, the active substances entered into the interior of the primary carbon particles and oxidative reactions occurred, and the long microcrystals in the nucleus were cleaved into fine microcrystals, and the spacing of the cleaved fine microcrystals was shorter and the arrangement of the fine microcrystals was more tight, so the average microcrystalline spacing in the process was decreased dramatically [44].

Figure 13 illustrates the distribution of microcrystalline curvature (C) of primary carbon particles in the carbon black samples. The average curvature of microcrystals in sample 0 was 1.21. As the DPF regeneration stage advanced, the average microcrystalline curvature of samples 1 to 4 increased from 1.23 to 1.28, and microcrystals with curvatures of 1.50 to 1.60 increased significantly. The microcrystalline curvature at the regeneration interface at the same stage was higher than that at the bottom, and the average values of C in samples 5 and 6 were 1.34 and 1.35, respectively. The higher the microcrystalline curvature of the primary carbon particles, the higher the number of active sites on their surfaces, and the stronger the ability to adsorb reactive oxygen species, so that all the oxidizing activity of carbon black was enhanced [45].

3.4. Degree of Graphitization of Carbon Black Samples

The carbon black samples were placed on glass slides wiped with anhydrous ethanol, and then the samples to be tested were placed on the X-Y automatic platform. Different determination areas of the same sample were selected for repeated determination by microscope to ensure the accuracy of the test sample results.

The Raman test results of carbon black showed strong peaks near 1360 cm⁻¹ and 1580 cm⁻¹, which are D peak (Disorder peak) and G peak (Graphite peak), respectively. According to the five-peak fitting method proposed by Sadezky,

Table 8. Raman peak positions and corresponding structures. Reproduced with permission from Xulong Chen, Yunxi Shi et al., Carbon Letters; published by MDPI, 2023 [24].

Location (cm−1)	Peak	Corresponding Structures
~1580 cm−1	G	for the E2g symmetric mode, representing the C-C stretching vibration of the sp2 hybridized atoms located within the carbon layer
~1360 cm−1	D1	A1g mode of symmetry, representing the vibration of structural defects located at the edge of the carbon layer and the impurity elements contained therein
~1620 cm−1	D2	corresponding to the E2g mode of symmetry, representing the vibrations of structural defects located in the graphite layer on the surface of graphite crystals
~1500 cm−1	D3	carbon clusters in the amorphous carbon and functional groups of organic components
~1180 cm−1	D4	corresponding to the A1g mode of symmetry, representing sp2–sp3 hybridization or C-C, C=C stretching vibrations and impurity ions in polyene structures

illustrates the Raman peak positions and corresponding structures [46,47]. The first-order Raman spectra of the carbon black samples were fitted to the peaks using Origin2021 software (OriginLab Corp., Northampton, MA, USA), and the fitting results are shown in Figure 14.

The relative peak intensity of D1 peak (I_D1/I_G) was used to characterize the structural defects at the edge of the graphite microcrystalline carbon layer in the carbon black samples, and the higher the I_D1/I_G, the more structural defects in the microcrystals and the higher the disorder. Since I_D3/I_G fluctuates greatly, it is difficult to accurately reflect the relative content of amorphous carbon, so the relative peak intensity of the D3 peak, R₃ = I_D3/ (I_G + I_D2 + I_D3), is defined to characterize the content of carbon clusters in the carbon black sample. A high percentage of carbon clusters is usually considered to enhance the oxidizing activity [48]. I_D1/I_G and R3 of the carbon black samples are shown in Figure 15. As seen in Figure 14, I_D1/I_G of sample 0 is 4.27, and the most structural defects were found at the edge of the microcrystalline carbon layer of the carbon black before DPF regeneration. After the NTP reaction, I_D1/I_G of carbon black samples 1~4 decreased from 3.31 to 3.10; I_D1/I_G of Samples 5~6 located at the regeneration interface were decreased dramatically to 2.98 and 2.54, indicating that the structural defects at the edges of the microcrystalline carbon layer were reduced. The R3 values of the carbon black samples were significantly increased after the NTP reaction. R3 of samples 1~4 increased from 0.41 to 0.47, and that of samples 5~6 were 0.52 and 0.58, respectively, which indicated that the content of carbon clusters in the amorphous carbon increased. The R3 of samples 5~6 were 0.52 and 0.58, respectively, and the content of carbon clusters in amorphous carbon increased. The ordering of graphite structure in PM and the oxidation activity of PM were closely related. The stronger the ordering of graphite structure in the carbon material, the relatively weaker its oxidation activity was. From the TG analysis, the oxidizing activity of carbon black was elevated after NTP action. Although the reduction of structural defects at the edge of the microcrystalline carbon layer made the microcrystalline structure more ordered to some extent, the increase in the content of carbon clusters in the amorphous carbon contributed more to the disorder of the microcrystalline structure, which made the oxidizing activity of the final carbon black enhanced.

4. Conclusions

A test system for regeneration of DPFs by NTP oxidation was set up to carry out regeneration tests at different stages on a DPF loaded with carbon black. The carbon black was sampled at the regeneration interface and at the bottom position, and the changes in the physicochemical properties of the carbon black samples during the regeneration of the DPF were analyzed by TG, XPS, HRTEM, and RNS. The main conclusions are as follows:

(1)

As the DPF regeneration stage advanced, the Ts, Tmax, and Te of carbon black samples 1–4 decreased. After regeneration, the activation energy of carbon black at the bottom of the DPF gradually decreases. The activation energy of the carbon black sample at the regeneration interface was lower than that at the bottom, and its oxidation activity was stronger.

(2)

With the advancement of the DPF regeneration process, the average microcrystalline length and the average microcrystalline curvature of the primary carbon particles in the carbon black sample increased. The decrease in microcrystalline length and interlayer spacing of carbon particles leads to a denser structure, which leads to an increase in the curvature of carbon particles and an increase in the number of surface active sites. The number of sites increased and the oxidation activity increased.

(3)

The proportion of C element in carbon black before regeneration was 88.4% and O element was 11.6%. After the NTP reaction, the proportion of C element decreased and the proportion of O element increased. The increase in oxygen-containing functional groups in carbon black indicated that the oxidizing activity of carbon black was enhanced, and the increase in amorphous carbon content played a major role in the decrease of graphitization degree of carbon black.

(4)

The DPF was continuously consumed by the reaction with carbon black during the transfer from the top to the bottom, resulting in a gradual decrease in the concentration of NTP active substances from the NTP flow direction. At the same regeneration stage, the highest reaction temperature at the regeneration interface was higher than that at the bottom, so the oxidative decomposition reaction between NTP and carbon black at the regeneration interface was more intense than that at the bottom, and the oxidation process of carbon black was faster.