Next Article in Journal
Runoff Response to Climate in Two River Basins Supplied by Small Glacier Meltwater in Southern and Northern Tibetan Plateau
Next Article in Special Issue
Emission Characteristics of Tyre Wear Particles from Light-Duty Vehicles
Previous Article in Journal
Influences of the Mid-Level Vortex on the Formation of Tropical Cyclone Toraji (2013)
Previous Article in Special Issue
Evaluating the Measurement Uncertainty of On-Road NOx Using a Portable Emission Measurement System (PEMS) Based on Real Testing Data in China
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Physicochemical Analysis of Particle Matter from a Gasoline Direct Injection Engine Based on the China Light-Duty Vehicle Test Cycle

1
School of Mechanics and Transportation, Southwest Forestry University, Kunming 650224, China
2
Key Laboratory of Vehicle Environmental Protection and Safety in Plateau Mountain Area of Yunnan University, Kunming 650224, China
*
Authors to whom correspondence should be addressed.
Atmosphere 2023, 14(4), 710; https://doi.org/10.3390/atmos14040710
Submission received: 9 March 2023 / Revised: 28 March 2023 / Accepted: 11 April 2023 / Published: 13 April 2023
(This article belongs to the Special Issue Traffic Related Emission)

Abstract

:
This paper investigated the physical and chemical properties of gasoline direct injection (GDI) engine particulate matter (PM). The physical properties mainly included the particulate aggregate morphology, primary particle size, and internal nanostructure. High-resolution transmission electron microscopy (HRTEM) and scanning electron microscopy (SEM) were used to obtain particle morphology information and to conduct image processing and analysis. The chemical characterization tests included X-ray photoelectron spectroscopy (XPS), energy dispersive scanning (EDS), Raman spectroscopy, Fourier-transform infrared spectroscopy (FTIR), and thermal gravimetric analysis (TGA). XPS can be used to observe the content of carbon and oxygen components and the surface carbon chemistry status, EDS can be used to obtain the elemental composition of particles, and TGA is used to analyze the oxidative kinetics of particles. Samples were collected from the exhaust emissions of a passenger vehicle compliant with China’s VI emission standards under China Light-Duty Vehicle Test Cycle (CLTC) test conditions. The study found that the particle morphology mainly comprised primary particles stacked on top of each other to form agglomerate structures, and the primary particles exhibited a core–shell structure. Analysis showed that carbon and oxygen were the predominant components of the particles, with other metallic elements also present. The XPS observations agreed with the FTIR results, indicating a small amount of oxygen was present on the particle surface and that the carbon components consisted mainly of sp2 hybridized graphite and sp3 hybridized organic carbon. The TGA results indicated high characteristic temperatures and low oxidation activity.

1. Introduction

Gasoline direct injection (GDI) engines offer better fuel economy and lower CO2 emissions than port fuel injection (PFI) engines. However, due to fuel impingement and uneven mixing, incomplete combustion occurs in GDI engines, resulting in the formation of fine particles and the number of particles generated is much higher than that from PFI engines [1]. These fine particles can penetrate the lungs and cause health problems, and long-term exposure to particulate matter can increase the risk of stroke, heart disease, and other health issues [2,3]. Particulate matter also has adverse environmental effects, reduces visibility, and contributes to acid rain [4]. Therefore, China and the European Union have established stricter automotive emission standards, with CHINA 6 and EU VI setting the particulate number limitation standard for light vehicles at 6 × 1011#/km and limiting the total PM mass to 0.005 g/km [5,6].
The mainstream technology for controlling particulate emissions mainly involves using a gasoline particle filter (GPF) to filter out particles, which gets rid of particles through oxidation regeneration [7]. The filtering efficiency of the GPF depends mainly on particle geometry, surface area, and other factors. The oxidation rate of particles is influenced by their agglomeration and nanoscale structure. Consequently, it is imperative to understand the physicochemical properties of these particles [8,9]. Observation of the agglomeration morphology and primary particle nanoscale structure of the particles using transmission electron microscopy (TEM) is a mature method [10]. Fractal theory can be employed to examine the agglomeration structure.
Furthermore, nanostructure parameters such as length, tortuosity, and spacing of carbon crystals can provide valuable insight. These parameters can be quantified using MATLAB image processing methods. The resulting data can then be utilized in the design of post-treatment systems such as particulate filters [11]. Some scholars have also studied the nature of particulate matter using various analytical techniques. For example, Hay Mon Oo et al. used TEM-EDS, X-ray diffraction (XRD), and TGA to characterize the physicochemical properties of particulate matter, finding that the morphology of GDI particulate matter was similar to that of diesel and the primary particles consisted of curved carbon crystals with an average particle size of 24 nm and a high proportion of amorphous carbon [12]. Sharma used Raman spectroscopy, FTIR, and high-resolution transmission electron microscopy (HRTEM) techniques to compare GDI particulate matter at three different ethanol ratios [13]. Liu et al. employed TGA, FTIR, and XPS to analyze the surface functional groups, carbon chemical states, and graphitization of diesel engine particles [14]. Jasiński et al. used FTIR, SEM-EDS, and EGA techniques to examine the particulate matter from gasoline engines operating at varying loads. Their findings indicated that an increase in load resulted in an increase in both particle concentration and diameter (with an average diameter up to 90 nm) [15]. Xuyuang Zhang et al. used HRTEM, XPS, EDS, and TGA to analyze the particulate matter produced by GDI gasoline engines under Worldwide Harmonized Light Vehicles Test Cycle (WLTC) operating conditions, including primary particle morphology, nanostructure, chemical fraction, carbon chemical state, and soot reactivity [16]. However, those study only examined particulate matter generated under a single speed or load condition, which was not a good representation of the physical and chemical characteristics of particulate matter from GDI vehicles under actual driving conditions.
In addition to the physicochemical properties of particulate matter, the oxidation capacity of GDI engines has also garnered significant attention. The oxidation rate parameters can be obtained using thermogravimetric analysis (TGA) [9,17]. Understanding carbon black’s oxidation process and kinetic parameters can aid in optimizing filter regeneration strategies [18]. By studying the physicochemical properties of particulate matter emitted by GDI vehicles, potential health and environmental risks can be determined. This information can also serve as a reference for emission control technology.

2. Materials and Methods

2.1. Particulate Sampling

The engine of the test vehicle was a direct injection gasoline engine that met the China 6 standard, and the test fuel was the China 6 test base fuel. Using the particle collection device, 47 mm of quartz fiber filter paper was used to collect the particles in the exhaust gas for physicochemical analysis; the dilution temperature was 200 °C and the sampling flow rate was 15 L/min−1. The analysis samples were the particulate matter emissions from three cold starts and three hot starts of the engine under CLTC cycle test conditions.
The CLTC cycle test was conducted with the chassis dynamometer. The test conditions included three speed ranges: low speed, medium speed, and high speed, where the maximum speed was 114 km/h and the average speed was 28.96 km/h, which was lower than the maximum speed of 120 km/h in NEDC and 134 km/h in WLTC. CLTC is mainly applicable to high-speed road conditions in China, so a more comprehensive range of the test cycle time was added: 1800 s. Compared with WLTC, the high-speed part required the least amount of time and the maximum speed and average speed were lower, which is more in line with the actual situation of urban driving in China. In order to ensure the adequacy and accuracy of the analysis samples, the same vehicle was tested several times under CLTC conditions.

2.2. Characterization Method

2.2.1. High-Resolution Transmission Electron Microscopy (HRTEM)

HRTEM was used to obtain images of particle aggregation morphology, primary particle microscopic morphology, and different classes of nanostructures. The HRTEM used was the FEI Talos F200X G2. The collected samples were pretreated by ultrasonic oscillation. Then, absolute ethanol containing the particulate matter was dropped onto the carbon film copper mesh, dried, and observed by HRTEM to observe the micromorphological characteristics of the particulate matter.

2.2.2. Scanning Electron Microscopy with Energy Dispersive Spectroscopy (SEM+EDS)

The chemical composition of the GDI particles was characterized by EDS combined with SEM. A small sample was taken and directly glued to the conductive adhesive, and the Oxford Quorum SC7620 sputter coater was used to spray gold. Then, the ZEISS Gemini SEM 300 scanning electron microscope was used to capture images of the sample morphology, perform energy spectrum mapping, and other tests. The accelerating voltage was 3 kV to obtain topography information, and the accelerating voltage was 15 kV for energy spectrum mapping.

2.2.3. X-ray Photoelectron Spectroscopy (XPS)

XPS can be used to identify the composition and valence states of the elements on the surface of the sample, and the sample can be analyzed qualitatively or quantitatively according to the binding energy. The quantitative X-ray microscopic analysis of the material was carried out using the Thermo Scientific K-Alpha XPS, and the C1s peak at the location of sp2 hybrid carbon was benchmarked against the binding energy of 284.8 eV. The full-spectrum scanning general energy was 150 eV, and the step size was 1 eV. The narrow-spectrum sweep had a general energy of 50 eV in steps of 0.1 eV.

2.2.4. Fourier-Transform Infrared Spectrometer (FTIR)

The particulate matter was analyzed using the Thermo Scientific Nicolet iS50 Fourier transform infrared spectrometer. The preparation of the particulate samples for this test involved tableting potassium bromide (KBr). During the test, the infrared spectrum with a wavenumber of 400~4000 cm−1 was acquired, and the acquired infrared spectrum was corrected and smoothed by OMNIC software for subsequent particle surface functional group analysis.

2.2.5. Raman Spectroscopy

The instrument used in the experiment was the HORIBA Scientific LabRAM HR Evolution micro confocal laser Raman spectrometer. The light source of the Raman spectrometer was a laser with a wavelength of 532 nm, each group of samples was scanned 30 times, and the scanning range was 50~3400 cm−1. The experiment used Origin software to perform peak fitting for Raman spectroscopy. The four peaks’ position, intensity, half-height width, and other parameters were obtained through peak fitting. The structure of the exhaust particulate matter and changes in gasoline engine activity were analyzed using these parameters.

2.2.6. Thermo Gravimetric Analysis (TGA)

The oxidation reaction capacity of the GDI particles was verified using a NETZSCH STA2500 thermogravimetric analyzer and an alumina crucible. Thermogravimetric analysis (TGA) refers to using specific heating procedures to obtain the relationship between the mass of the sample and the change in temperature in order to study changes in particulate composition and thermal stability. The particle weight loss (TG) curve was measured by TGA, and the thermal gravimetric (DTG) curve was obtained using Origin software.

3. Results and Discussion

3.1. Aggregate Morphology

Figure 1 shows images of the morphology of the GDI vehicle particulate agglomerates under CLTC conditions at scales of 1000, 500, 200, 100, 50, and 20 nanometers. The groups of particles can be seen to be made up of dozens to hundreds of roughly spherical primary particles. These primary particles formed agglomerates after adsorbing part of the SOF component. The agglomerates of particles were clustered or in chain-like structures, and the whole was in a disordered state. At the same time, it could be seen that some areas of the agglomerates were darker, which was caused by multiple primary particles stacked on top each other and also caused the outline of some primary particles to be blurred. The amount of particulate matter generated at the edge of the agglomerates was reduced. Fewer particles will reduce the chance of collision between primary particles, thus reducing the number of particles required to form a chain. It is not easy for smaller particles to combine to form large-sized particles [19]. At the same time, smaller particles are easily oxidized, which is conducive to regenerating the particulate matter filter.
The primary particles in Figure 1d were measured using Image J software. Due to the partial overlap of the boundaries, the particles with partial boundary overlap were ignored when measuring. According to the data, the particle size and number were determined, and the histogram was drawn to perform normal fitting. The data were usually fitted, and the average value of the distribution was the average diameter of the primary particles of the particulate matter. Figure 2 shows the primary particle size distribution from the GDI engine under the CLTC test, and the primary particle size diameter range was mainly 10–50 nm. Compared with the results of other researchers, the particle size distribution range of the primary particles was the same, and the average diameter of the primary particles was 23.8 nm.
The accumulation of numerous primary particles created irregular structures known as particle agglomerates. An agglomerate’s fractal dimension (Df) represents the degree of compactness of the particles. The higher the fractal dimension, the more compact the structure of the agglomerate and the greater the overlap between its primary particles [20,21]. ImageJ software measures the average diameter (DP) of easily identifiable primary particles to quantify these agglomerate morphologies. Equation (1) includes the fractal dimension (Df):
N p = k f ( D g D p ) D f
where kf is the fractal factor; Dg is the rotational diameter of the particle agglomerate; Dp is the average diameter of the particle; and Np is the total number of primary particles determined by Equation (2):
N p = ( A a A p ) a
where Aa is the projection area of the particulate agglomerates, Ap is the projection surface area of the primary particles; and a represents the empirical index of the projected area. The value of a is 1.09 according to the literature [22]. The relationship between the rotational diameter of the particulate matter and the maximum projection length L is given by:
L D g = 1.50 ± 0.05
When calculating the rotational diameter of the particle Dg using Equation (3), it was found that it was not easy to accurately measure the center of mass of the particle, and the phenomena of accumulation and overlap occurred between the primary particles. It was difficult to directly measure the radius of rotation of the particle. Therefore, the rotational diameter of particles was calculated using the indirect estimation method.
Particulate matter mainly occurs in chains, branches, or rings. At the same time, it was observed that some of the exhaust particulate matter was darker as a result of multiple elementary carbon particles piling up on top of one another, obfuscating the edges of the individual elementary carbon particles. The size of the fractal dimension of the particulate matter can reflect the degree of density among the particulate matter. When the fractal dimension of the particulate matter is large, the overlap rate between the particulate matter is also more significant, and the particulate matter is more closely arranged. As depicted in Figure 3, the analytical dimension of the particulate agglomerates under CLTC conditions was 1.84. As can be seen from Table 1, this value was greater than that observed for GDI particle generation under WLTC conditions. This discrepancy may have been due to the relatively small proportion of high-speed sections present under CLTC conditions compared to under WLTC conditions. Smaller particles undergo oxidation during ultra-high-speed cases under WLTC conditions, thereby reducing particulate matter generation. This decreases the probability of collision and combination between elementary particles and makes it more difficult for them to connect, merge, or agglomerate to form larger particles. Under CLTC conditions, the degree of particulate agglomeration was enhanced, and the structure was more compact.

3.2. Carbon Nanostructure

The microscopic morphology of the primary carbon particles was like an onion, stacked from the inside to the outside, and the different primary carbon particles were arranged in concentric spheres. As shown in Figure 4, the microcrystalline arrangement in the central area of these particles was relatively disordered. In contrast, the peripheral microcrystals were arranged more regularly and wrapped around the “core” to form a “shell”. This “shell” had a circular layer or core–shell structure, making the internal carbon crystals even more disordered. Gaddam et al. [24] believed that the interior of this primary particle was a separate core composed of carbon crystals while the outer shell was composed of several parallel carbon crystals, and a stripe structure could even be observed. Despite its complexity and many influencing factors, the overall morphology of primary carbon particles consistently presented as a “shell + core” structural form.
As shown in Figure 5a, under CLTC conditions, the overall length of the primary particle carbon crystals from the GDI engine ranged from 0.6 to 1.8 nm. The peak value was between 0.28 and 0.42 nm, and the average fringe length of the carbon crystals was 0.841 nm. This result was larger than that reported by Gaddam [24] (0.8 nm) and Xuyuang Zhang [16] (0.836 nm). When the microcrystal size of primary carbon particles is relatively large, the disorder of the carbon layer structure is low. Compared with particles with regular graphite structure, particles with amorphous morphology are more easily oxidized, resulting in lower activation energy. Compared with carbon atoms at the base of the carbon layer, carbon atoms at the edge have more vigorous oxidation activity. Longer microcrystal size means fewer carbon atoms at the edge, making oxidation more complex [25].
Figure 5b shows that the basic particle carbon crystal tortuosity of the GDI engine particulate matter was 1.0~1.4. From Table 2, we can see that this result was slightly higher than the research results of Gaddam et al. [24] (1.18~1.21), which was mainly related to the engine’s operating conditions. The average carbon crystal tortuosity was about 1.282, exceeding Xuyuang Zhang’s [16] test result (1.175). The tortuosity is caused by the presence of 5- or 7-membered carbon rings in the carbon layer structure and represents the degree of missing carbon atoms within the carbon layer. Tortuosity is also an important parameter reflecting the disorder of the carbon layer structure. When the tortuosity of the carbon layer crystal plane is large, the proportion of straight and graphitic crystal planes significantly decreases. Conversely, amorphous carbon structures become more prominent, making it more difficult for carbon layers to oxidize.
Figure 5c shows that the distance between primary particle carbon crystals of the GDI engine particulate matter presented a multi-peak distribution. The maximum separation distance was about 0.38–0.40 nm, and the average interlayer spacing was about 0.358 nm, which was slightly smaller than the test result of Xuyuang Zhang et al. [16] (0.362 nm). Layer spacing is one of the critical parameters that can reflect the oxidation activity of the primary carbon particles. With a decrease in layer spacing, the structure inside the primary carbon particles is relatively compact and the number of independent carbon layers is large, which reduces the contact area between the carbon layer and oxygen, thus increasing the oxidation difficulty and decreasing the oxidation activity.

3.3. Elemental Composition Analysis

The elemental composition of the gasoline engine particles was analyzed by an X-ray spectrometer equipped with a field emission SEM, starting with selecting the target area from the obtained gasoline engine particulate matter samples for EDS analysis. The scanning electron microscopy results are shown in Figure 6, with the particles shown at the 100 nm–200 μm scale. At the same time, the agglomerates comprised many primary particles, nearly spherical particles, and complex compounds that were tightly bonded together. The scanning electron microscopy results indicated the morphology of the particles discharged by the engine was similar to rice grains, wand they were more closely distributed. Their contours became blurred due to particle agglomeration and bonding. It can be seen from the figure that the morphology of the particles by SEM showed more apparent multi-layer cluster accumulation shapes, more scattered branch shapes, and more dispersed chain structures.
Figure 7 shows the EDS element analysis of the sample at points selected on the target particle sample. It can be seen from the spectrum that the main component of the particulate matter was C, with the largest value in the energy spectrum analysis. Secondly, the contents of O and Si were also high. O was mainly derived from the incomplete oxidation of hydrocarbons during fuel combustion. Aluminum (Al) was mostly present as fine particles produced by the wear and tear of the piston and other parts of the gasoline engine during operation, which were absorbed by the particles in the exhaust process. The high content of Si was partly due to additives and partly due to the quartz fiber filter membrane used to collect the particles. Mg, Ca, Zn, Ba, and other elements mainly originated from lubricating oil and other additives [27]., The weight percentage of various elements of gasoline particulate can be observed in Table 3.

3.4. Surface Chemical Composition

XPS can be used to qualitatively analyze the particle’s elemental composition, functional group content, and hybridization of carbon atoms [18]. Figure 8 shows the XPS full-spectrum scan of the GDI gasoline engine particulate matter collected under CLTC conditions. According to the findings of the EDS analysis, carbon was the primary component of the exhaust particulate matter, followed by oxygen. The oxygen on the surface of the particulate matter was mainly derived from the oxygen-containing active groups adsorbed at the edge defects of its microcrystalline structure and oxygen in the circulating intake air during the generation and oxidation of the particulate matter. In the figure, C1s and O1s represent photoelectrons excited by ionizing the 1 s orbitals of C and O atoms, respectively. It can be seen from the figure that the surface of the particulate matter produced by the combustion of the GDI engine mainly contained the C and O elements, and with less content of other elements. At the same time, by comparing the peak areas of the C1s and O1s peaks in the XPS full spectrum, the molar percentages of the C and O elements on the surface of the particulate matter were obtained, and the ratio of C and O elements calculated by Avantage software was 12.91.
The signals obtained from XPS full-spectrum scans are relatively coarse. They can help to determine the types and locations of elements in particles. However, they cannot provide information on chemical valence states or molecular structures. To obtain detailed information about the carbon element on the surface of the particles, high-resolution scans of the C1s peak were required, which can be used to obtain the relative oxygen-containing functional group content and C atom hybridization forms on the particle surface by fitting the fine carbon spectrum to peaks. This study used Avantage software to fit the fine carbon spectra of four peaks. The C1s peak was fitted into four peaks, representing sp2, sp3, hydroxyl (C–OH), and carbonyl (C=O). The corresponding peak positions were 284.8, 285.4, 286.6, and 288.8 eV, respectively. Figure 9 shows the C1s peak fitting diagram for the particles. The ratio of each fitted peak area to the C1s peak area was used to determine the relative content of sp2, sp3, C–OH, and C=O.
Figure 9 reveals that exhaust particles mostly existed in sp2 and sp3 hybridization forms. The dominant form was sp2. An increase in sp2 content indicates stronger particle carbon crystal orderliness. An increase in sp3 content indicates stronger particle disorder. The larger the ratio of sp2/sp3, the stronger the particle order. Enhanced carbon structure order makes particle oxidation difficult. By analyzing the fitted carbon spectrum after peak fitting, we obtained the percentages of C–OH and C=O functional groups and the hybridization ratio of the C element. As particle order increases, its surface will have a longer microcrystal size and fewer irregular breaks. This reduces the number of chemical reaction activation sites. During combustion, it becomes more difficult for active oxygen-containing groups to bind to particles, thus reducing oxidation activity.

3.5. Surface Functional Groups

Figure 10 shows the FTIR spectrum of particle samples from GDI vehicles under CLTC conditions. The absorption peak at 3435 cm−1 corresponded to O-H’s stretching vibration absorption peak, which was due to the small amount of water molecules adsorbed on the surface of the gasoline particles. Near 2920 cm−1, we found a characteristic peak corresponding to the C–H functional group of aliphatic hydrocarbons. In addition, there was a strong absorption peak at the wavelength of 1590 cm−1 that corresponded to a telescopic vibration peak on the C–C bond in the graphite lattice. The infrared spectral wavenumber is 1000~1800 cm−1, including carbon and oxygen functional groups and aromatic rings. Among them, the functional groups of carbon and oxygen were mainly manifested as changes in absorption peaks at wavenumbers 1000~1300 cm−1. The hydrocarbon functional groups were mainly manifested as changes in absorption peaks at wavenumbers 1390, 1450, and 2920 cm−1, among which the change in absorption peak at 2920 cm−1 was more pronounced [28].
Due to the inability to precisely control the mass of particles required for infrared detection, it was impossible to accurately quantify the contents of functional groups. The asymmetric methylene peak near 2920 cm−1 in the infrared spectrum is generally regarded as a characteristic peak for carbon–hydrogen bonding. The aromatic ring peak near 1590 cm−1 is considered a characteristic peak for C–C bonding. The relative content of aliphatic C–H functional groups on the surface of the particulate matter was determined by two characteristic peak-equivalent peak-to-height ratios (IC-H/IC=C). The relative content of aliphatic C-H functional groups in particulate matter is positively correlated with the value of IC-H/IC=C; the larger the value, the higher the content. It was found that the oxidative activity of PM was positively correlated with the content of C-H functional groups.

3.6. Degree of Disorder

Raman spectroscopy can determine the molecular structure of particulate samples. There is no noticeable difference in the Raman spectra of particulate samples under different working conditions. The difference between particles can be accurately measured only by analyzing the carbon structure by peak fitting. There is a frequency doubling relationship between first-order and second-order Raman spectra, so only the first-order Raman spectrum can be considered when peak division is fitted. D and G peaks appeared in the carbon particles’ first-order Raman spectral curves, corresponding to Raman frequency shifts of 1360 and 1590 cm−1, respectively. Lattice defects and respiratory vibrations mainly caused the D peak, with the local structure of the crystal losing symmetry or shifting to lower symmetry. The spectrum was realized as the D peak generated by the symmetric vibration of the A1g graphite lattice, indicating the disorder degree of the carbon particles. The G peak was mainly caused by the tensile vibration of the hybridization of carbon atoms, sp2, indicating the degree of graphitization of the carbon particles. Both the degree of disorder and graphitization of the carbon particles were related to their oxidation activity. Origin software was used to perform four-peak fitting through Lorentz and to compare the area values of the D and G peaks in order to study the degree of graphitization of the soot particles.
Figure 11 shows a particulate sample’s first-order Raman spectral peak fitting with four peaks: one Gaussian peak and three Lorentz peaks. In addition to the two existing peaks at 1360 cm−1 (D peak) and 1580 cm−1 (G peak), a third peak was fitted at 1500 cm−1 (D3 peak). Furthermore, a peak at 1180 cm−1 (D4 peak) was obtained by fitting. Amorphous carbon components caused the D3 peak, including functional groups, organic components, and fragments within the exhaust particles. This peak was key to the overlap between the D1 and G peaks. Its intensity was related to the functional groups, organic components, and fragments within the emitted particles. Asymmetric graphite layers caused the D4 peak, which resulted from the stretching vibrations of C–C and C=C bonds or sp2–sp3 hybridization in polyene-like structures. According to Table 4, the value of ID1/IG was 0.97, which are higher than those of J.H. Chan (0.94) [29], indicating a high degree of graphitization. The carbon crystal arrangement was more orderly and the oxidation activity was weaker.

3.7. Soot Oxidation Reactivity

TGA was performed in two gas atmospheres. A 1 mg sample was analyzed using TGA. Figure 12 shows the relationship between temperature and time during TGA analysis. The heating range was set from 40 to 800 °C at 10 degrees Celsius per minute. The process was divided into two stages. The first stage was conducted under an inert gas (N2) atmosphere. The first stage used pure nitrogen to eliminate oxidation reactions caused by oxygen. This allowed for the analysis of only the volatilization and decomposition processes of the sample. As a result, the impact on the particles’ SOF adsorption capacity could be better explored. The temperature was raised at 10 °C per minute from 40 °C, and it was held for 10 min. Then, it was raised to 400 °C and held for another 10 min. This stage removed water and volatile organic compounds. After holding for 10 min, the temperature was lowered to 150 °C. In the second stage, under an oxygen atmosphere, the temperature was again raised at a rate of 10 °C/min to 800 °C and held for another 10 min. This stage involved solid carbon pyrolysis, where carbon particles and non-volatile organic compounds were burned off to leave behind metal salts.
Figure 13 shows the TG and DTG curves of the particulate matter. It can be seen that in the warming range of 200–500 °C, the weight loss rate of particles was low, the highest weight loss rate was reached at about 660 °C, and then the weight loss rate gradually decreased. Before 400 °C, the TG curve of the particulate matter did not change significantly, and as the temperature increased, the particulate matter began to produce a combustion pyrolysis reaction. When the temperature reached about 710 °C, the quality of the particles remained unchanged because some metal components contained in the particles adsorbed onto the surface of the particles, making oxidation difficult.
The ignition temperature (Ti) is the temperature at which 10% of the mass of the soot is lost. The temperature corresponding to when the weight loss rate of the soot particles reaches 50% is a crucial temperature parameter to measure the peak period of soot combustion. The burnout temperature (Tb) corresponds to the weight loss of soot particles reaching 90% [30]. Maximum oxidation rate temperature (Tp), the temperature at which the soot particles reach their maximum weight loss rate during oxidation, is the minimum point of the TG differential curve. Compared with the results of Xuyuan Zhang et al., the peak and burndown temperatures were higher. It can be seen from Table 5 that the temperature of each characteristic point of the particles under CLTC conditions was higher than that under other conditions, and the oxidation temperature of the particles was higher, which makes GPF reduction more difficult.

4. Conclusions

The aim of this paper was to analyze the physicochemical properties of exhaust particles collected from GDI vehicles under CLTC conditions. GDI vehicle particles include organic and soot particles generated by GDI technology, which are small in size but large in number and pose a threat to the environment and human health. This paper examined the morphology, nanostructure, and elemental composition of the particles, as well as the surface functional groups, chemical state of carbon, and oxidation capacity.
(1).
It was found that CLTC working conditions produced more compact particle agglomerates, with a primary particle diameter of 23.26 nm, an agglomerate fractal dimension of around 1.84, a larger particle microcrystal size, an average carbon crystal fringe length of 0.841 nm, an average tortuosity of 1.282, and an average carbon crystal separation distance of 0.358 nm.
(2).
The graphite structure on the surface of the particles was more orderly and arranged. The main elements on the particle surface were C and O, with a ratio of 12.91. The sp2/sp3 ratio was 5.75, indicating a high degree of orderliness in the particles. The proportion of carbon–hydrogen functional groups to oxygen-containing functional groups in the particles was relatively low. The degree of branching and disorderliness was also low. This resulted in strong chemical stability and weaker oxidation activity.
(3).
At the same time, the surface functional group content increased, the particle ignition temperature (Ti), burnout temperature (Tb), and peak oxidation rate temperature (Tp) increased, and Tb reached 698.6 °C. The oxidation activity decreased, causing the regeneration temperature of GPF to increase, which makes GPF reduction more difficult.

Author Contributions

Conceptualization, C.H. and H.W.; methodology, H.Y. and J.L.; software, H.W.; writing—original draft, H.W.; validation, C.H., H.Y., and J.L.; funding acquisition, C.H. and X.L.; formal analysis, H.Y. and J.L.; resources, X.L.; data curation, X.L. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by National Natural Science Foundation of China, grant number 51968065; Yunnan Provincial high level talent support project, grant number YNWR-QNBJ-2018-066, YNQR-CYRC-2019-001; The scientific research foundation of education bureau of Yunnan Province of China, grant number 2022J0500.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Not applicable.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Saliba, G.; Saleh, R.; Zhao, Y.; Presto, A.A.; Lambe, A.T.; Frodin, B.; Sardar, S.; Maldonado, H.; Maddox, C.; May, A.A.; et al. Comparison of Gasoline Direct-Injection (GDI) and Port Fuel Injection (PFI) Vehicle Emissions: Emission Certification Standards, Cold-Start, Secondary Organic Aerosol Formation Potential, and Potential Climate Impacts. Environ. Sci. Technol. 2017, 51, 6542–6552. [Google Scholar] [CrossRef] [PubMed]
  2. Tung, N.T.; Ho, K.-F.; Niu, X.; Sun, J.; Shen, Z.; Wu, F.; Cao, J.; Dung, H.B.; Thuy, T.P.C.; Hsiao, T.-C.; et al. Loss of E-Cadherin Due to Road Dust PM2.5 Activates the EGFR in Human Pharyngeal Epithelial Cells. Environ. Sci. Pollut. Res. 2021, 28, 53872–53887. [Google Scholar] [CrossRef] [PubMed]
  3. Xing, J.; Shao, L.; Zhang, W.; Peng, J.; Wang, W.; Hou, C.; Shuai, S.; Hu, M.; Zhang, D. Morphology and Composition of Particles Emitted from a Port Fuel Injection Gasoline Vehicle under Real-World Driving Test Cycles. J. Environ. Sci. 2019, 76, 339–348. [Google Scholar] [CrossRef] [PubMed]
  4. Michelsen, H.A. Probing Soot Formation, Chemical and Physical Evolution, and Oxidation: A Review of in Situ Diagnostic Techniques and Needs. Proc. Combust. Inst. 2017, 36, 717–735. [Google Scholar] [CrossRef] [Green Version]
  5. Ko, J.; Kim, K.; Chung, W.; Myung, C.-L.; Park, S. Characteristics of On-Road Particle Number (PN) Emissions from a GDI Vehicle Depending on a Catalytic Stripper (CS) and a Metal-Foam Gasoline Particulate Filter (GPF). Fuel 2019, 238, 363–374. [Google Scholar] [CrossRef]
  6. Zhu, R.; Hu, J.; Bao, X.; He, L.; Lai, Y.; Zu, L.; Li, Y.; Su, S. Tailpipe Emissions from Gasoline Direct Injection (GDI) and Port Fuel Injection (PFI) Vehicles at Both Low and High Ambient Temperatures. Environ. Pollut. 2016, 216, 223–234. [Google Scholar] [CrossRef]
  7. Dangsunthonchai, M.; Visuwan, P.; Komintarachat, C.; Theinnoi, K.; Chuepeng, S. Nanoparticle Components and Number–Size Distribution of Waste Cooking Oil-Based Biodiesel Exhaust Gas from a Diesel Particulate Filter-Equipped Engine. ACS Omega 2022, 7, 3384–3394. [Google Scholar] [CrossRef]
  8. Martínez-Munuera, J.C.; Zoccoli, M.; Giménez-Mañogil, J.; García-García, A. Lattice Oxygen Activity in Ceria-Praseodymia Mixed Oxides for Soot Oxidation in Catalysed Gasoline Particle Filters. Appl. Catal. B Environ. 2019, 245, 706–720. [Google Scholar] [CrossRef] [Green Version]
  9. Choi, S.; Seong, H. Oxidation Characteristics of Gasoline Direct-Injection (GDI) Engine Soot: Catalytic Effects of Ash and Modified Kinetic Correlation. Combust. Flame 2015, 162, 2371–2389. [Google Scholar] [CrossRef] [Green Version]
  10. Li, Z.; Qiu, L.; Cheng, X.; Li, Y.; Wu, H. The Evolution of Soot Morphology and Nanostructure in Laminar Diffusion Flame of Surrogate Fuels for Diesel. Fuel 2018, 211, 517–528. [Google Scholar] [CrossRef]
  11. Yehliu, K.; Vander Wal, R.; Boehman, A. Development of an HRTEM Image Analysis Method to Quantify Carbon Nanostructure. Combust. Flame—COMBUST FLAME 2011, 158, 1837–1851. [Google Scholar] [CrossRef]
  12. Oo, H.M.; Karin, P.; Charoenphonphanich, C.; Chollacoop, N.; Hanamura, K. Physicochemical Characterization of Direct Injection Engines’s Soot Using TEM, EDS, X-Ray Diffraction and TGA. J. Energy Inst. 2021, 96, 181–191. [Google Scholar] [CrossRef]
  13. Sharma, N.; Agarwal, R.A.; Agarwal, A.K. Particulate Bound Trace Metals and Soot Morphology of Gasohol Fueled Gasoline Direct Injection Engine. J. Energy Resour. Technol. 2019, 141, 022201. [Google Scholar] [CrossRef]
  14. Liu, Y. Variations in Surface Functional Groups, Carbon Chemical State and Graphitization Degree during Thermal Deactivation of Diesel Soot Particles. J. Environ. Sci. 2023, 124, 678–687. [Google Scholar] [CrossRef] [PubMed]
  15. Jasiński, R.; Strzemiecka, B.; Koltsov, I.; Mizeracki, J.; Kurzawska, P. Physicochemical Analysis of the Particulate Matter Emitted from Road Vehicle Engines. Engines 2021, 18, 8556. [Google Scholar] [CrossRef]
  16. Zhang, X.; Song, C.; Lyu, G.; Li, Y.; Qiao, Y.; Li, Z. Physicochemical Analysis of the Exhaust Soot from a Gasoline Direct Injection (GDI) Engine and the Carbon Black. Fuel 2022, 322, 124262. [Google Scholar] [CrossRef]
  17. Shi, Y.; Zhou, Y.; Li, Z.; Cai, Y.; Li, X.; He, Y.; Fang, J. Effect of Temperature Control Conditions on DPF Regeneration by Nonthermal Plasma. Chemosphere 2022, 302, 134787. [Google Scholar] [CrossRef]
  18. Bogarra, M.; Herreros, J.M.; Tsolakis, A.; Rodríguez-Fernández, J.; York, A.P.E.; Millington, P.J. Gasoline Direct Injection Engine Soot Oxidation: Fundamentals and Determination of Kinetic Parameters. Combust. Flame 2018, 190, 177–187. [Google Scholar] [CrossRef]
  19. Pfau, S.A. Nanoscale Characterisation of Soot Particulates from Gasoline Direct Injection Engines. Available online: https://www.researchgate.net/publication/356493097_Nanoscale_characterisation_of_soot_particulates_from_gasoline_direct_injection_engines (accessed on 8 March 2023).
  20. Köylü, Ü.Ö.; Faeth, G.M.; Farias, T.L.; Carvalho, M.G. Fractal and Projected Structure Properties of Soot Aggregates. Combust. Flame 1995, 100, 621–633. [Google Scholar] [CrossRef]
  21. Swapna, M.S.; Devi, H.V.S.; Raj, V.; Sankararaman, S. Fractal and Spectroscopic Analysis of Soot from Internal Combustion Engines. Eur. Phys. J. Plus 2018, 133, 106. [Google Scholar] [CrossRef]
  22. MEGARIDIS, C.M.; DOBBINS, R.A. Morphological Description of Flame-Generated Materials. Combust. Sci. Technol. 1990, 71, 95–109. [Google Scholar] [CrossRef]
  23. Saffaripour, M.; Chan, T.W.; Liu, F.; Thomson, K.A.; Smallwood, G.J.; Kubsh, J.; Brezny, R. Effect of Drive Cycle and Gasoline Particulate Filter on the Size and Morphology of Soot Particles Emitted from a Gasoline-Direct-Injection Vehicle. Environ. Sci. Technol. 2015, 49, 11950–11958. [Google Scholar] [CrossRef] [PubMed]
  24. Gaddam, C.K.; Vander Wal, R.L. Physical and Chemical Characterization of SIDI Engine Particulates. Combust. Flame 2013, 160, 2517–2528. [Google Scholar] [CrossRef]
  25. Liati, A.; Schreiber, D.; Dimopoulos Eggenschwiler, P.; Arroyo Rojas Dasilva, Y.; Spiteri, A.C. Electron Microscopic Characterization of Soot Particulate Matter Emitted by Modern Direct Injection Gasoline Engines. Combust. Flame 2016, 166, 307–315. [Google Scholar] [CrossRef]
  26. Pfau, S.A.; La Rocca, A.; Haffner-Staton, E.; Rance, G.A.; Fay, M.W.; Brough, R.J.; Malizia, S. Comparative Nanostructure Analysis of Gasoline Turbocharged Direct Injection and Diesel Soot-in-Oil with Carbon Black. Carbon 2018, 139, 342–352. [Google Scholar] [CrossRef]
  27. Yang, J.; Roth, P.; Ruehl, C.R.; Shafer, M.M.; Antkiewicz, D.S.; Durbin, T.D.; Cocker, D.; Asa-Awuku, A.; Karavalakis, G. Physical, Chemical, and Toxicological Characteristics of Particulate Emissions from Current Technology Gasoline Direct Injection Vehicles. Sci. Total Environ. 2019, 650, 1182–1194. [Google Scholar] [CrossRef] [Green Version]
  28. Yang, H.; Zhang, Y.; Li, C.; Fang, Y.; Li, X. Physicochemical Characteristics of Particulate Matter Emitted from the Oxygenated Fuel/Diesel Blend Engine. Aerosol. Air Qual. Res. 2021, 21, 210175. [Google Scholar] [CrossRef]
  29. Chan, J.H.; Tsolakis, A.; Herreros, J.M.; Kallis, K.X.; Hergueta, C.; Sittichompoo, S.; Bogarra, M. Combustion, Gaseous Emissions and PM Characteristics of Di-Methyl Carbonate (DMC)-Gasoline Blend on Gasoline Direct Injection (GDI) Engine. Fuel 2020, 263, 116742. [Google Scholar] [CrossRef]
  30. Shi, Y.; Lu, Y.; Cai, Y.; He, Y.; Zhou, Y.; Fang, J. Evolution of Particulate Matter Deposited in the DPF Channel during Low-Temperature Regeneration by Non-Thermal Plasma. Fuel 2022, 318, 123552. [Google Scholar] [CrossRef]
Figure 1. TEM images of the GDI PM. (a) 1000 nm; (b) 500 nm; (c) 200 nm; (d) 100 nm; (e) 50 nm; (f) 20 nm.
Figure 1. TEM images of the GDI PM. (a) 1000 nm; (b) 500 nm; (c) 200 nm; (d) 100 nm; (e) 50 nm; (f) 20 nm.
Atmosphere 14 00710 g001
Figure 2. The primary size distribution of GDI PM.
Figure 2. The primary size distribution of GDI PM.
Atmosphere 14 00710 g002
Figure 3. Statistical determination of soot aggregate fractal dimension (Df) for the GDI PM.
Figure 3. Statistical determination of soot aggregate fractal dimension (Df) for the GDI PM.
Atmosphere 14 00710 g003
Figure 4. Typical HRTEM images of the GDI PM. (a) 10 nm; (b) 5 nm.
Figure 4. Typical HRTEM images of the GDI PM. (a) 10 nm; (b) 5 nm.
Atmosphere 14 00710 g004
Figure 5. Histograms of the fringe length (Lf), tortuosity (Tf), and separation distance (Ds) for the GDI PM. (a) fringe length; (b) tortuosity; (c) separation distance.
Figure 5. Histograms of the fringe length (Lf), tortuosity (Tf), and separation distance (Ds) for the GDI PM. (a) fringe length; (b) tortuosity; (c) separation distance.
Atmosphere 14 00710 g005aAtmosphere 14 00710 g005b
Figure 6. SEM micrographs of the GDI PM based on CLTC. (a) 2 μm; (b) 500 nm; (c) 200 nm; (d) 100 nm.
Figure 6. SEM micrographs of the GDI PM based on CLTC. (a) 2 μm; (b) 500 nm; (c) 200 nm; (d) 100 nm.
Atmosphere 14 00710 g006
Figure 7. EDS quantitative elemental analysis of GDI PM.
Figure 7. EDS quantitative elemental analysis of GDI PM.
Atmosphere 14 00710 g007
Figure 8. XPS spectrum of GDI PM.
Figure 8. XPS spectrum of GDI PM.
Atmosphere 14 00710 g008
Figure 9. C1s peak for the GDI soot with curve fitting for component resolution.
Figure 9. C1s peak for the GDI soot with curve fitting for component resolution.
Atmosphere 14 00710 g009
Figure 10. FTIR spectrum of GDI PM.
Figure 10. FTIR spectrum of GDI PM.
Atmosphere 14 00710 g010
Figure 11. First-order Raman spectra for GDI PM.
Figure 11. First-order Raman spectra for GDI PM.
Atmosphere 14 00710 g011
Figure 12. TGA methodology.
Figure 12. TGA methodology.
Atmosphere 14 00710 g012
Figure 13. TG. and DTG profiles for the GDI spectrum of GDI PM (under N2).
Figure 13. TG. and DTG profiles for the GDI spectrum of GDI PM (under N2).
Atmosphere 14 00710 g013
Table 1. Summary of the morphological features of GDI particles in previous studies.
Table 1. Summary of the morphological features of GDI particles in previous studies.
ReferenceEngine TypeTestDP (nm)Df
Present Study GDI engineCLTC23.81.84
Xuyuang Zhang [16]GDI engineWLTC27.63 1.72
Meghdad Saffaripour [23]GDI engineUS0624.6–26.61.77
FTP75251.8
Table 2. Summary of the nanostructure of GDI soot particles in previous studies.
Table 2. Summary of the nanostructure of GDI soot particles in previous studies.
ReferenceEngine TypeLf (nm)TfDs (nm)
Present Study GDI engine0.7411.1720.391
Xuyuang Zhang [16]GDI engine0.8361.263 0.362
SB40.8221.2810.366
Chethan K. Gaddam [24]SIDI engine0.81.175_
S.A. Pfau [26]1.0L GTDI0.971.130.414
1.4L GTDI0.991.130.427
Table 3. The elemental content of gasoline particulates.
Table 3. The elemental content of gasoline particulates.
ElementalPercentage by WeightElementalPercentage by Weight
C49.43S0.15
O25.54K0.14
Mg1.87Ca5.53
Al0.17Fe7.22
Si4.41Zn0.82
P3.19Zr1.20
Table 4. ID1/IG, AD1/AG, ID3/IG, AD3/AG,ID4/IG, and AD4/A obtained from Raman spectra.
Table 4. ID1/IG, AD1/AG, ID3/IG, AD3/AG,ID4/IG, and AD4/A obtained from Raman spectra.
Title 1ID1/IGAD1/AGID3/IGAD3/AGID4/IGAD4/AG
Present study 0.972.370.280.420.120.24
J.H Chan [29]0.942.010.350.49--
Table 5. Initial combustion characteristics of PM.
Table 5. Initial combustion characteristics of PM.
ReferenceEngine TypeTi (°C)Tp (°C)Tb (°C)
Present Study GDI engine557.7658.2689.6
Xuyuang Zhang [16]GDI engine-612.5643.1
J.H. Chan [29]SIDI engine-510-
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Wang, H.; He, C.; Yu, H.; Li, J.; Liu, X. Physicochemical Analysis of Particle Matter from a Gasoline Direct Injection Engine Based on the China Light-Duty Vehicle Test Cycle. Atmosphere 2023, 14, 710. https://doi.org/10.3390/atmos14040710

AMA Style

Wang H, He C, Yu H, Li J, Liu X. Physicochemical Analysis of Particle Matter from a Gasoline Direct Injection Engine Based on the China Light-Duty Vehicle Test Cycle. Atmosphere. 2023; 14(4):710. https://doi.org/10.3390/atmos14040710

Chicago/Turabian Style

Wang, Hao, Chao He, Haisheng Yu, Jiaqiang Li, and Xueyuan Liu. 2023. "Physicochemical Analysis of Particle Matter from a Gasoline Direct Injection Engine Based on the China Light-Duty Vehicle Test Cycle" Atmosphere 14, no. 4: 710. https://doi.org/10.3390/atmos14040710

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop