Effect of Exhaust Gas Recirculation Rate on the Emissions of Persistent Organic Pollutants from a Diesel Engine

This study investigates the emission characteristics of toxic organic pollutants (PAHs, PCDD/Fs, and PCBs) generated by a heavy-duty diesel engine operating at various exhaust gas recirculation (EGR) rates during steady-state cycles. Tests on the exhaust gas composition were conducted before and after changing the EGR ratio. The fuel used in the study (B2 diesel) was a mixture of 2% biodiesel and 98% diesel. The main focus was on the emission factors for the organic toxic pollutants in the exhaust gas after EGR ratios of 0% and 5% were applied. At an EGR ratio of 5%, the total mass emission factors for the PAHs and PCBs increased by 9.1 times and 14.4 times, respectively, while the toxicity equivalent factors increased by 4.0 times and 4.8 times, respectively. A significant increase in pollutants with a higher molecular weight, particularly for the PAHs, was observed after applying an EGR ratio of 5%, implying incomplete combustion. The emission factors of carbon dioxide (CO2) and nitric oxides (NOx) decreased by 2.5% and 54.4%, respectively, when the EGR ratio was increased from 0% to 5%, but those of PM and carbon monoxide (CO) increased by 60.5% and 66%, respectively. Therefore, a combination of control strategies is necessary in order to achieve a significant reduction in the emission of all pollutants.

European emissions regulations are forcing automobile diesel engine manufacturers to find more sophisticated ways to reduce emissions, especially nitrogen oxides (NO x ) and particulate matter (PM) (Maiboom et al., 2009).Exhaust gas recirculation (EGR) is a reputable practice for reducing the concentration of NO x in the emissions (Maiboom et al., 2009;Chang et al., 2014).The NO x reduction mechanisms applied in EGR involve complex and sometimes reverse processes that occur during combustion (Rajesh Kumar and Saravanan, 2015).Actual EGR systems often result in uneven distribution of emissions from cylinder to cylinder, with air and recirculated exhaust gas not being completely mixed, which has an effect on other pollutants (Wu et al., 2014;Subbarayan and Senthil Kumaar, 2017).Subbarayan and Senthil Kumaar (2017) tested a cottonseed biodiesel blend for cold and hot EGR rates based on injection pressure and injection time and observed a significant increase in emissions of HC, smoke opacity, and NO x .However, under varying load conditions, the high-temperature EGR rate was associated with lower emissions as compared to the cold EGR rate (Chang et al., 2013b;Redfern et al., 2017a).
Application of fuel blends is a viable technique in pollution control.Overall, biodiesel and bio-alcohols (such as ethanol and butanol) have higher brake thermal efficiency (BTE) and lower emissions for carbon monoxide (CO), particulate matter (PM), and hydrocarbons (HC) (Weber and Kuch, 2003;Wang et al., 2009).However, biodiesels produce more nitrogen oxide (NO x ) emissions than fossilbased diesel, which has the ability to reduce NO x emissions owing to its huge heat of vaporization (Chang et al., 2013a).PM emissions from diesel engines provide carbon sources for the formation of PAHs and a large surface area for adsorption of other toxins in the ambient air (Liu et al., 2013).
According to Wang et al. (2003), POPs pose huge risks to human health and despite their emission levels being at very low concentrations, their importance as toxic pollutants in the atmosphere should not be ignored especially in the case of PCDD/Fs, which have been referred to as the poison of the century (Liu et al., 2016).The lethality of dioxins (PCDD/Fs) has been proven through a large number of animal experiments, with estimations showing that their toxicity is 130 times that of cyanide and 900 times that of arsenic (Ebert and Bahadir, 2003;Redfern et al., 2017b).After entering and accumulating in the human body, dioxins are believed to adversely affect human health, for example, through destruction of the immune system and initiation of cancer (Xing et al., 2017;Zhao et al., 2018b).
Even though the research on toxic and carcinogenic pollutants emitted by diesel engines is still relatively scarce, investigations on the pollutants emitted by diesel engines have been ongoing for decades (Chang et al., 2014).The International Agency for Research on Cancer (IARC) classifies the toxic pollutants emitted by diesel engines under the first group of human carcinogens.Because air pollution control is among the main strategies in the plan for control and prevention of cancer, efforts are underway to cut the emissions from diesel engines (Wang et al., 2007;Chang et al., 2014).
This study is aimed toward investigating the effects of different EGR ratios on the emissions from a heavy-duty diesel engine at a steady-state.Studies on emissions from diesel engines have been carried out for many years.However, to the best of our knowledge, no such studies have focused on the effect of different EGR ratios on the emissions from a diesel engine running at a steady state.Efforts were thus made herein to discuss the exhaust emission characteristics of organic toxic pollutants, especially PAHs, PCDD/Fs, and PCBs, in the exhaust gases emitted by diesel engines after changes in the EGR rate.

Diesel Engine
For the experiment, a six-cylinder diesel engine was used with the following specifications: naturally aspirated, water-cooled, 6-liter direct-injection, and heavy-duty.Further specifications of the diesel engine and the boundary conditions during the operation are listed in Table 1.An engine dynamometer was used to control the engine's speed and torque.In this experiment, the diesel engine was monitored under four modes of operation out of the possible thirteen European Steady-State Cycle (ESC) modes: the 1 st Mode (750 rpm, 0 Nm), 2 nd Mode (1650 rpm, 360 Nm), 7 th Mode (1650 rpm, 90 Nm), and 11 th Mode (1950 rpm, 96.2 Nm).

Sampling Procedures
For collection of the test emissions, runs were made using B2 fuel (98% fossil diesel and 2% biodiesel); EGR rates were set at 0% and 5%, and the cycle was repeated.The engine was started and left to run for 30 minutes before each sample was taken and for at least 3 minutes between the selected test modes.After each EGR rate change, the engine was preconditioned in the 11 th mode for 30 minutes.During each entire test cycle, the exhaust gases of the diesel engine were sampled right away and at equal speeds using a sampling system consisting of a flow meter, a fiberglass filter, a condenser, a two-stage glass cylinder, Table 1.A summary of the important parameters of the diesel engine under investigation.

Item
Hino W06E and a pump.The emission particle phases were captured using a fiberglass filter.A condenser located in front of the two-stage glass cylinder was used to cool the exhaust gas to a temperature below 5°C, and the moisture was then extracted from the exhaust gas.The gas phase contaminants were then collected using a two-stage glass cylinder.A box with 2.5 cm polyurethane foam plugs at each end was filled with 5.0 cm of XAD-2 resin, which weighed about 20 g.The analytes were concentrated by combining the four samples collected in the ESC mode to ensure that the analyte concentration was above the detection limit.The total sampling time was approximately 80 minutes (approximately 20 minutes per ESC mode).Finally, the sampled flue gas volumes were consolidated under the following physical conditions: a pressure of 760 mmHg and a temperature of 273 K.

Pre-treatment of Samples and Quality Assurance
Before the sampling process, the glass fiber filter was exposed to an oven temperature of 450°C for a duration of 8 hours to get rid of any possible organic matter.The POPs in the blank samples did not exceed 0.5% of the total POPs in the actual exhaust samples, with the exception of PBDEs, which were less than 2%.The concentrations of the pollutants of interest in the blank samples were thus insignificant compared to the corresponding exhaust samples.
The sampling system was tested for leakage in the section lying between the filter holder's inlet and the flow meter's outlet prior to each fuel test.In the preceding sampling exercise, a breakthrough test was performed on PAHs using a three-stage glass filter.From the results, it was clear that 16 specific PAHs in the third stage only accounted for between 0.409% and 4.37% of all three stages.Consequently, the two-stage glass filter applied in this investigation ensured a collection efficiency of 95%.The XAD-2 resin was spiked with an isotopically categorized PCDD/F in order to replace the standard prior to sampling at the exhaust.The recovery rate of PCDD/F replacement standard ranged between 90% and 94%, which was in line with the recovery standard of U.S. EPA Method 23 (70-130%).This recovery indicates that the breakthrough of PCDD/F was negligible.It was not possible to purchase alternative standards for other analytes; therefore, the corresponding PCDD/F was used for the PBDD/F sampler collection efficiency check instead of the standard recovery rate.The corrections for the sampler collection efficiencies in the concentrations of other chemical species reported in this study were not done.
Prior to the extraction process, the 13 C 12 -labeled internal standard depicted in Table 2 was incorporated into the extract for the quantification of the sample and monitoring of its recovery in the analytical procedure.Addition of the recovery standard solution to the sample prior was necessary during the instrumental analysis to ensure recapture in the analysis.Relevant standards were applied for the precision and recovery (PAR) of the POPs, replacements, and internal and recycling practices.A signal-to-noise ratio (S/N) greater than 3 was applied as the limit of detection (LOD) in this study.Meanwhile, an S/N greater than 10 was applied as the limit of quantitation (LOQ).Another assumption used for this investigation was that the undetected homologues were 50% of the LOD for the PAHs and PBDEs, while for the PCDD/Fs, PCBs, PCDEs, PBDD/Fs, and PBBs the undetected homologues were equal to 0.

Instrumental Analysis
The concentrations of PAHs were measured using Agilent 5890A and 5975 series GC/MS, which were equipped with capillary columns (50 m × 0.32 mm × 0.17 µm, HP Ultra 2).Earlier operating conditions applied by Chang et al. (2014) were also used for this investigation as follows: a splitless 1 µL injection at a temperature of 300°C, with an ion source temperature of 310°C.A 45°C furnace temperature was maintained for 1 minute and then raised to a temperature of 100°C within 5 minutes.Afterwards, the temperature was increased from 100°C to 320°C at a rate of 8°C per minute, and then it was kept at 320°C for 15 minutes.For the primary and secondary PAH ions, a scan mode for the pure PAH standard was used to establish their masses.The selected ion monitoring (SIM) mode was used to qualify the PAHs.

Analytical Stage
The fiber filter and a two-stage cartridge for each exhaust sampling were examined to determine the concentration of the captured pollutants.A Soxhlet extractor was used to extract the exhaust gas samples using n-hexane and dichloromethane with a volume ratio of 1:1; each 250 mL was the solvent for 24 hours.To concentrate the extract, the extracts were gently purged with ultrapure nitrogen and later purified using a silica gel column.After obtaining a concentrate of 1 mL, gas chromatography/mass spectrometry (GC/MS) was used for the analysis, and sixteen PAH homologs were detected.
Subsequently, seventeen 2,3,7,8-substituted PCDD/Fs, twelve dioxin-like PCBs, six PCDEs, twelve 2,3,7,8substituted PBDD/Fs, five PBBs, and fourteen PBDE homologs were analyzed.This stage of the analysis involved treatment of the solution using concentrated sulfuric acid, after which it was subjected to a series of sample purification and fractionation steps including a silica multilayer column, an alumina column, and a column of activated carbon.In the section with the alumina column, elution of the non-planar PCBs and PBBs was carried out with 15 mL of hexane, and then further elution using 25 mL of a 1:24 mixture of DCM and hexane was completed before progressing to the activated carbon column.For the activated carbon column, elution was done using 5 mL of a toluene, methanol, ethyl acetate, and hexane mixture in a volume ratio of 1:1:2:16 to obtain PCDEs, PBDEs, planar PCBs, and PBBs, and then 40 mL of toluene was added.The analytical procedures are described in more detail in a previous work.

Conventional "Criteria" Air Pollutants (CO, CO 2 , PM, and NO x )
At an EGR ratio of 0%, the PM and NO x concentrations in the exhaust from the diesel engine fueled with B2 were 70.6 mg Nm -3 and 352 ppm, respectively.With the power output as the basis of comparison for the pollutants, the emission factors of CO, CO 2 , NO x , and PM were 3.71 g kW-h -1 , 872 g kW-h -1 , 6.53 g kW-h -1 , and 0.38 g kW-h -1 , respectively.When the EGR ratio was increased to 5%, the CO, CO 2 , NO x , and PM emission factors were 6.16 g kW-h -1 , 850 g kW-h -1 , 2.98 g kW-h -1 , and 0.61 g kW-h -1 , respectively.The results indicate that the NO x emissions from the engine decreased by 54.4% and the CO 2 emission factors dropped by 2.5% for the 5% increase in the EGR rate.However, for the other two pollutants, the rate of emissions increased with an increase in the EGR ratio, and the CO and PM produced by the engine increased by 66%, and 60.5%, respectively.The experimental data shows that increasing the EGR only significantly decreased NO x emissions, while CO and PM exhibited a significant increase.Therefore, the EGR ratio should not be increased blindly.As an alternative, caution should be exercised in order to strike a balance in the trade-off between PM and NO x emissions and hence minimize the overall pollution effects.Similar findings have been reported by Sunil Naik and Balakrishna (2018), who conducted experiments with B10 and B20 fuels at EGR rates of 0%, 10%, and 20%.The average NO x emission reduction rates of the B10 and B20 test fuels at EGR rates of 0%, 10%, and 20% were approximately 18.24%, 17.67%, and 17.14%, respectively.However, CO in the exhaust increased significantly as the proportion of the exhaust gas recirculation rate increased.

POPs Emitted from a Diesel Engine Running on B2 Biodiesel
The mass concentrations of pollutants in the exhaust gases using B2 fuel (98% fossil-based diesel and 2% biodiesel) in a diesel engine are in Table 3. Their toxic equivalency factors (TEFs), which are parameters used to calculate the toxicity of POPs, were applied in the calculation of their respective equivalent toxicities (Van den Berg et al., 1998).The TEFs for PBBs, PBDEs, and PCDEs are not widely accepted yet and as such are only evaluated on the basis of mass concentration (Tsai et al., 2018).The results of the analysis for each persistent organic pollutant using the two EGR ratios are discussed below.

PAH Emission Factors
Without EGR, the emission factor for PAHs based on mass concentration was 107 µg kW-h -1 , while on the basis of toxic equivalency was 0.488 µg BaPeq kW-h -1 .After applying an EGR ratio of 5%, the total mass emission factor for the PAHs increased by 9.07 times, while the toxic equivalency factor increased by 3.97 times, as shown in Table 4.It can be seen from the characteristic profile presented in Fig. 1 that the species of polyaromatic hydrocarbons with higher molecular weight increased significantly, indicating that there was incomplete combustion at the 5% EGR ratio (Zhang et al., 2013).Specifically, there was a decrease in naphthalene, acenaphthylene, and acenaphthene, while an increase was observed for fluorene, phenanthrene, anthracene, fluoranthene, and pyrene.The highest increase by about 10 µg BaPeq kW-h -1 was observed for phenanthrene.The percentage contribution of the remaining PAHs were negligible for both EGR ratios (Alriksson and Denbratt, 2006;Borillo et al., 2018).

Concentration of PCDD/Fs in the Emissions
According to previous literature, PCDFs can be formed by a condensation reaction of chlorophenol with chlorobenzene via PCDEs (Liu et al., 2008).In addition, PCDEs can generate PCDD/Fs through a pyrolysis reaction.Therefore, PCDEs may be important precursors for the formation of PCDD/Fs.Consequently, the concentration of PCDEs is significantly lower than the concentration of PCDD/Fs in diesel engine exhaust.A similar phenomenon has been reported in a study of simulated fly ash in a flow reactor system (Liu et al., 2011).One explanation for this phenomenon is the formation of a common precursor of PCDEs and PCDD/Fs such as chlorobenzene and chlorophenol, as well as surface competition.Another explanation is that PCDEs are mostly converted to PCDD/Fs as intermediates during the reaction.From Table 3, the mass concentration of PCDD/Fs was 39.6 pg Nm -3 , with a toxicity equivalence of 2.54 pg I-TEQ Nm -3 .
From Table 4, it can be seen that there is a decrease in PCDD/F emission factors after using the 5% EGR as compared to an EGR ratio of 0%.The possible reason is that when the EGR ratio was 0%, the higher oxygen concentrations in the supplied air for combustion led to the formation of PCDD/Fs through de novo synthesis reactions (Chang et al., 2000).
From the congener profile presented in Fig. 2, it can be observed that except for OCDD and OCDF, the rest of the congeners had a contribution of less than below 10%.The  contribution of OCDD and OCDF was 50% and 20%, at 0%-EGR and 40% and 18% at 5%-EGR respectively.

Concentration of PCBs in the Emissions
PCBs had a concentration of 10.7 pg Nm -3 and a toxic equivalency of 0.0339 pg WHO 2005 -TEQ Nm -3 , as shown in Table 3.After applying an EGR ratio of 5%, the mass emission factor of PCBs was 14.4 times higher than the original value at 0% EGR, while on the basis of toxicity, it increased by 5 times, as shown in Table 4.The increase in both mass concentration and toxic equivalency can be attributed to incomplete combustion as a result of increasing the EGR ratio (Hedman et al., 2006).

CONCLUSION
An increase in EGR from 0% to 5% is a viable NO x control technology since it reduced NO x emission factor by 54.4%.However, it also increased the PM emission factor by 60.5% and as such should be combined with additional aftertreatment, for instance, diesel particle filters (DPF) to achieve simultaneous reductions in both NO x and PM.Raising the EGR reduced the combustion efficiency hence increasing the emission factors of PAHs and PCBs by 9.07 times and 14.36 times on the basis of mass and 3.97 times and 4.81 times based on the toxicity.Further investigations are therefore necessary in the development of more sustainable ways to reduce organic toxic pollutants from diesel engines.

Table 2 .
The internal standards used for the study.

Table 3 .
Concentrations of particulate matter, NO x and POPs in the emissions from a diesel engine using B2 as the fuel.

Table 4 .
Emission factors of the toxic organic pollutants in EGR exhaust gas from diesel engine flue gas.