Abstract
Canola (or rapeseed) oil and waste vegetable oil (WVO) are used commonly to make biodiesel fuels composed completely from these oils (B100) or as blends with petroleum diesel (B0). However, no studies have reported the mutagenic potencies of the particulate matter with diameter ≤2.5 μm (PM2.5) or the mutagenicity emission factors, such as revertants/MJthermal (rev/MJth) for these biodiesel emissions. Using strains TA98 and TA100 with the Salmonella (Ames) mutagenicity assay, we determined these metrics for organic extracts of PM2.5 of emissions from biodiesel containing 5% soy oil (soy B5); 5, 20, 50, and 100% canola (canola B5, B20, B50, B100), and 100% waste vegetable oil (WVO B100). The mutagenic potencies (rev/mg PM2.5) of the canola B100 and WVO B100 emissions were generally greater than those of B0, whereas the mutagenicity emission factors (rev/MJth, rev/kg fuel, and rev/m3) were less, reflecting the lower PM emissions of the biodiesels relative to B0. Nearly all the rev/mg PM2.5 and rev/MJth values were greater in TA98 with S9 than without S9, indicating a relatively greater role for polycyclic aromatic hydrocarbons, which require S9, than nitroarenes, which do not. In TA100 −S9, the rev/mg PM2.5 and rev/MJth for the biodiesels were generally ≥ to those of B0, indicating that most of these biodiesels produced more direct-acting, base-substitution mutagenic activity than did B0. For B100 biodiesels and petroleum diesel, the rev/MJth in TA98 +S9 ranked: petroleum diesel > canola > WVO > soy. The diesel emissions generally had rev/MJth values orders of magnitude higher than those of large utility-scale combustors (natural gas, coal, oil, or wood) but orders of magnitude lower than those of inefficient open burning (e.g., residential wood fireplaces). These comparative data of the potential health effects of a variety of biodiesel fuels will help inform the life-cycle assessment and use of biodiesel fuels.
Keywords: Diesel exhaust, biodiesel, canola oil, waste cooking oil, mutagenicity
Introduction
Although biodiesel made from peanut oil was the first fuel used in the late 1890s by Rudolf Diesel to power his invention, the diesel engine, petroleum diesel soon became the dominant diesel fuel [1]. Recently, however, biodiesel has seen a resurgence as an alternative to petroleum diesel, due largely to environmental and economic concerns. Extensive research has explored the most suitable types of starting materials and the best methods for converting those materials into biodiesel fuels [1,2].
Biodiesel is a mixture of fatty acid methyl esters derived from the trans-esterification of fatty acids (triglycerides) from plant oils or animal fats. Although fuels composed entirely of biodiesel (B100) can be used in most diesel engines, typical commercially available biodiesel fuels are a blend of petroleum diesel (B0) with various percentages of biodiesel. Biodiesel derived from soybean oil at 20% (B20) is the most common type of biodiesel used in the U.S., whereas biodiesel derived from rapeseed (or canola) oil at 30% (B30) or even 100% (B100) is most common in Canada and Europe [2]. Biodiesels made from waste vegetable oils (WVO) from commercial cooking operations are also used extensively [3,4].
Both rapeseed oil and canola oil are derived from the seeds of the rapeseed plant, which is a member of the genus Brassica [5]. Canola was developed in the 1970s by Canadian scientists who developed a cultivar of the rapeseed plant that was missing glucosinolates and erucic acid, which were considered inedible or toxic at high doses [5]. They named the newly developed cultivar canola, which was formed from the words “Canadian” and “Oil” (or ola); both rapeseed oil and canola oil are used to make biodiesel.
The health effects of diesel exhaust from engines fueled by petroleum diesel have been studied extensively, with the first comprehensive toxicological assessment published in 1982 [6]. The particulate matter (PM) from diesel exhaust contributes in varying degrees to air pollution, especially in urban areas. Chronic exposure to ambient PM, especially that associated with gasoline or diesel exhaust, is genotoxic to humans [7] and is associated with an increased risk for lung cancer, asthma, and cardiovascular disease [8,9].
An extensive review by Bűnger et al. [10] concluded that the emissions from biodiesel (mostly rapeseed) relative to those from petroleum diesel had generally lower concentrations of carbon monoxide, hydrocarbons, polycyclic aromatic hydrocarbons (PAHs), and PM but higher concentrations of nitrogen oxide and aldehydes. As reviewed by Mutlu et al. [11], 10 of 18 studies (55%) reported that the extractable organics from the PM of biodiesel emissions were less mutagenic in the Salmonella (Ames) mutagenicity assay than were those from petroleum diesel emissions. Most of these studies used rapeseed biodiesel, but of the five that evaluated soy biodiesel, three reported that the emissions from soy biodiesel were less mutagenic than those from petroleum diesel. Based on the extractable organics from PM, we found that the mutagenicity emission factors, e.g., revertants/megajoule of thermal energy (rev/MJth) or rev/kg fuel burned from B20, B50, and B100 of soy biodiesel were notably lower than those from petroleum diesel [11–13]. There are no reports on the mutagenicity of canola or WVO biodiesel emissions.
In the present study we have extended this literature by determining the mutagenic potency of the extractable organic matter (EOM) and the PM2.5 (PM with a diameter ≤2.5 μm), and we have calculated the mutagenicity emission factors of the emissions from soy B5 biodiesel, which we had not reported in our previous study of soy biodiesel [11–13]; canola B5, B20, B50, and B100 biodiesel, and WVO B100 biodiesel. Besides our previous study of soy biodiesel [11–13], only one study with rapeseed biodiesel reported data that permitted the calculation of a mutagenicity emission factor [14]. Thus, we compared the mutagenicity emission factors for soy, canola, rapeseed, and WVO to those of petroleum diesel, and we compared all of them to those of other combustion emissions to place the results in context to a wide range of emissions.
Materials and methods
Chemicals and reagents
We purchased dimethyl sulfoxide (DMSO) and the positive controls (2-nitrofluorene, sodium azide, and 2-aminoanthracene) from Sigma, St. Louis, MO and Aroclor-induced Sprague-Dawley rat liver S9 from Moltox, Boone, NC. We purchased ultralow-sulfur (<15 ppm) diesel fuel from Red Star Oil, Raleigh, NC; and soy, canola, and waste vegetable oil (WVO) biodiesel fuel (methyl esters) from Piedmont Biofuels, Pittsboro, NC. These biodiesels were selected based on availability and common usage.
Engine operation, parameters, PM collection, and organic extractions
The engine operation, monitors used, and the collection of PM2.5 were as described for our previous study with soy biodiesel [13]. Briefly, we combusted petroleum diesel (B0); 5% soy biodiesel (soy B5); 5, 20, 50, or 100% canola biodiesel (canola B5, B20, B50, or B100, respectively); and 100% waste vegetable oil biodiesel (WVO B100) in a Yanmar L70 diesel engine (Adairville, GA) coupled with a Pramac E3750 generator (Marietta, GA). This air-cooled, single-cylinder, direct-injection, 320-cm3 engine was rated for 5.8 hp (4.3 kW) continuous load at 3600 rpm. We maintained an operating load of 3 kW using two electric heaters. Further details and rationale for using this engine are described in Mutlu et al. [15].
We measured engine inlet air velocity, fuel consumption, gas-phase engine emissions, and PM2.5 mass, which was collected on filters at a rate of 20 L/min and diluted ~10:1 prior to capture on 90-mm Teflon® filters. In our previous diesel and biodiesel studies [11–13], we collected PM for mutagenicity evaluation in conjunction with animal-exposure experiments. These always included gravimetric determination of PM concentrations within the exposure chambers. The present study, however, did not include accompanying animal exposures and, thus, did not include gravimetric PM concentrations. Instead, we determined PM concentrations (mg PM/m3) using a tapered oscillating electronic microbalance (TEOM) and used the measurements to calculate PM concentrations. Although TEOM and filter PM measurements are closely related, TEOM-based values for diesel PM measurements are usually ~10–15% lower than filter-based values. This is likely because the TEOM detector operates at 55⁰C compared to room-temperature for filter measurements, resulting in the loss of some volatile organic compounds from the diesel PM at 55°C. Thus, to make direct comparisons to our previous study, we recalculated the soy data from Mutlu et al. [11,12] based on TEOM PM measurements that we had made at the time but had not published.
Because our PM measurements and samples were taken from exposure chambers after dilution with filtered air, we used measurements of nitric oxide (NO) concentrations in the chambers and in the engine exhaust to determine dilution factors, and we corrected chamber PM concentrations to reflect PM concentrations in the emissions. For the samples collected for extraction, we determined NO ratios of ~10 to 1 and used these to calculate PM exhaust concentrations (mg/m3).
In addition to emission factors based on exhaust volumes, it is also useful to calculate them based on mass of fuel consumed (kg fuel burned) and energy released (megajoulethermal or MJth). Emission factors calculated in this way allow direct comparison between different sources. To calculate these, we determined the engine volume exhaust rate (m3/h) and calculated a PM emission rate (mg/h). We did this by mounting a hot wire anemometer within a plastic cowl and attaching the cowl over the engine air inlet. The anemometer software was configured to automatically multiply the measured inlet air velocity (m/h) by the cowl cross sectional area (m2) to yield the engine inlet volumetric flow (m3/h).
Measured or calculated engine inlet flows can be corrected to reflect engine exhaust flows with a small stoichiometric correction. This calculated correction recognizes that reactant and product moles (and volumetric flows) are not conserved but results in minor 2–3% volume increases. The stoichiometric corrections were based on assumed compositions of B0 (cetane, C15H32) and B100 (methyl linoleate, C19H34O2), complete combustion, and the measured concentrations of O2 in the engine exhaust (15.5%). Minor differences in the compositions of the three biofuels were negligible, and we interpolated the values for the different blends.
In our previous study of soy biodiesels [11–13], we measured (and reported) engine volumetric flow values of ~70 m3/h, but in our current study we measured values of ~50 m3/h. However, because the experimental conditions were identical, we suspected an error. We approximated the engine volumetric flow based on the fixed displacement (320 cm3) and the constant (3600 rpm) operating speed. However, also required in the calculation was the understanding that for 4-stroke engines, air intake occurs every other downstroke, and the addition of an assumed volumetric efficiency, defined as the ratio of the actual air flow divided by the theoretical air flow. Volumetric efficiencies greater than 1.0% are possible for turbocharged engines and advanced variable-valve timing designs; however, for the small simple engine that we used, values between 0.75–0.85% are more typical due to flow restrictions that create reduced pressure and density of the incoming air. Based on an assumed volumetric efficiency of 0.8%, we calculated an approximate engine volumetric flow of 27.7 m3/h.
Upon investigation for this apparent discrepancy, we found that anemometer measurements from both our previous and current studies were affected by a software-conversion error in which the logged data were multiplied by a factor of 1.88. That is, the data collected and logged every 3 min were 1.88 times larger than the data displayed on the anemometer. Correcting the current data for this error produced engine inlet flows of 27.1 to 28.8 m3/h, in agreement with our calculated value of 27.7 m3/h noted above.
Based on this finding, we also corrected the engine inlet flow rate values from our previous study [11–13] but obtained values of 36.7 to 39.4 m3/h rather than 27.1 to 28.8 m3/h. Upon additional investigation, we found that the anemometer was highly susceptible to error if misaligned and not oriented precisely perpendicular to the incoming flow. We discovered that any misalignment may result in partial blocking of the air inlet, a higher perceived velocity (and subsequent higher volumetric flow). Thus, we concluded that such a potential misalignment in our previous study [11–13] likely produced erroneous flow rates. Consequently, we present in the Results the corrected mutagenic emission factors for our published data [11–13] based on our newly calculated inlet flow rate of 27.7 m3/h. A corrigendum has been published in Inhalation Toxicology that is linked to Mutlu et al. [11–13] that presents a correction factor (divisor) of 2.56 (70.9 m3/h/27.7 m3/h) and identifies the tables and column values affected for the B0 and soy B20, B50, and B100 results.
In addition to engine exhaust flows (m3/h), measurements of fuel consumption rates (kg/h) and fuel heating values (MJth/kg) allow PM emission factors to be calculated based on fuel consumption (mg/kg) and heat released (mg/MJth). Fuel consumption rates for the soy experiments were measured continuously, and higher heating values (HHVs) for the B0 and soy B100 were measured experimentally [13]. Measurements by Mehta and Anand [15] indicate only small differences in measured lower heating values (LHVs, MJth/kg) for soy B100 (37.68), canola B100 (37.70), peanut B100 (38.05), and corn B100 (38.73); thus, we assumed that the fuel consumption rates and heating values for canola and WVO fuels were identical to soy. As with the stoichiometric correction, we interpolated HHVs for the different blends. We present our results based on HHVs to compare directly the diesel engine emission factors to those from a variety of stationary combustors and other open-combustion sources.
We extracted organics from the PM2.5 with dichloromethane, determined the percentage of extractable organic matter (% EOM) gravimetrically, and solvent-exchanged the organics into DMSO. Additional operating details are described by Mutlu et al. [13].
Mutagenicity assays
We performed the Salmonella plate-incorporation mutagenicity assay as described [16] using the frameshift strain TA98 and the base-substitution strain TA100 with and without S9. The availability of staff and access to the combustion facility permitted us to collect enough sample to evaluate all of the biodiesel emissions for mutagenicity in strain TA98 with and without S9. This was essential for us to compare the results to various other combustion emissions that we have tested in TA98 +S9 [11–13]. In cases where additional sample still remained, we first tested the remaining sample in TA100 + S9, followed by TA100 −S9. Due to limited sample, canola B100 was not evaluated in TA100, and the remaining canola blends were not evaluated in TA100 +S9.
We typically evaluated the extracts of B0 and the B5 and B20 samples at 10, 25, 50, and 100 μg EOM/plate, and we typically added higher doses (125, 250, and 500 μg EOM/plate) for the B50 and B100 samples to assure that we would characterize adequately the dose-response curve. We selected doses based on our previous studies of biodiesel emissions [11–13]. To conserve sample, we used our best judgement to select the doses for various emissions in various strains. Based on availability of sample, we evaluated the samples at 1 plate/dose in 3 independent experiments in TA98 −S9, in 2 independent experiments in TA98 +S9, and in one experiment each in TA100 +/− S9.
The negative control was DMSO, and the positive controls were 2-nitrofluorene (3 μg/plate) for TA98 −S9, sodium azide (3 μg/plate) for TA100 −S9, and 2-aminoanthracene (0.5 μg/plate) for TA98 and TA100 +S9. We incubated the plates for 3 days at 37°C and then counted the mutant colonies (revertants, rev) on an AccuCount™ 1000 automatic colony counter (BioLogics, Manassas, VA).
Calculation of mutagenic potencies
We defined a positive mutagenic response as a reproducible, dose-related increase of twofold or greater of rev/plate relative to the DMSO control. We calculated the linear regression over the linear portion of the dose-response curve to determine the mutagenic potency of each extract, expressed as rev/μg EOM. The linear portion was defined by the line with the highest r2 value. We deleted doses that caused a down-turn in the curve and reduced the r2 value relative to that produced by inclusion of the lower doses [17]. We multiplied the mutagenic potencies by the % EOM to give rev/μg PM2.5, and then by 1000 to give rev/mg PM2.5, which is the same as rev/mg particle.
Calculation of mutagenicity emission factors
We calculated mutagenicity emission factors expressed as rev/m3 of exhaust, rev/kg fuel burned, and rev/MJth energy released using the standard and measured values reported in Table 1. The mass of the particles equals the mass of PM2.5; thus, we multiplied mg PM2.5/m3, mg PM2.5/kg fuel burned, and mg PM2.5/MJth (Table 1) by rev/mg PM2.5 to calculate the various mutagenicity emission factors as described previously [11]. All diesel exhaust particles extracted for evaluation in this study were ≤2.5 μm.
Table 1.
Engineering parameters of combustion of biodiesel fuels
| Sample | TEOM PM conc (mg/m3) | Dilution factor | Exhaust PM conc (mg/m3) | Engine air inlet rate (m3/h) | Engine exhaust ratea (m3/h) | PM emission rate (mg/h) | Fuel consumption rate (kg fuel/h)b | PM emission factor (mg/kg fuel) | Fuel higher heating value (MJth/kg fuel)b | Fuel heat release rate (MJth/h) | PM emission factor (mg/MJth) |
|---|---|---|---|---|---|---|---|---|---|---|---|
| B0 | 2.4 | 10.0 | 24.0 | 28.6 | 29.2 | 700 | 1.05 | 670 | 45.8 | 47.8 | 14.6 |
| Soy B5 | 1.9 | 12.0 | 22.8 | 27.5 | 28.0 | 639 | 1.05 | 609 | 45.5 | 47.8 | 13.4 |
| Canola B5 | 1.7 | 9.7 | 15.9 | 27.6 | 28.1 | 448 | 1.05 | 426 | 45.5 | 47.8 | 9.4 |
| Canola B20 | 2.0 | 10.0 | 20.0 | 28.8 | 29.4 | 587 | 1.08 | 544 | 44.6 | 48.2 | 12.2 |
| Canola B50 | 2.0 | 9.0 | 18.0 | 28.3 | 28.9 | 519 | 1.12 | 466 | 43.0 | 47.9 | 10.8 |
| Canola B100 | 1.5 | 10.0 | 15.0 | 27.2 | 27.7 | 416 | 1.17 | 355 | 40.2 | 47.0 | 8.9 |
| WVO B100 | 1.8 | 7.5 | 13.1 | 27.1 | 27.6 | 363 | 1.17 | 310 | 40.2 | 47.0 | 7.7 |
The engine exhaust flow rates were calculated from engine air inlet flow rates, and a stoichiometric correction assuming that B0 was cetane (C15H32), B100 was methyl linoleate (C19H34O2), there was complete combustion, and the measured engine exhaust O2 concentration was 15.5%. This correction resulted in a small, 2% increase that was weakly dependent on the type of fuel.
Based on negligible differences between the heating values for corn, soy, and canola biodiesels reported by Mehta and Anand [15], fuel consumption and fuel heating values for this study were assumed identical to those for the soy biodiesel studies reported by Mutlu et al. [13]. Fuel consumption values for soy B5 and canola B5 were interpolated here from the B0 and soy B100 values reported by Mutlu et al. [13].
Results
Engineering parameters
Table 1 presents emissions and process measurements used to calculate mutagenic emission factors for the new fuels and fuel blends examined. These included a B0 replicate sample; a soy B5 blend not studied previously; canola B5, B20, B50, B100; and WVO B100. The biodiesel exhausts had PM concentrations (mg/m3), PM emission rates (mg/h), and PM emission factors (mg/kg fuel and mg/MJth) that were lower than those of petroleum diesel (B0). In all cases, WVO B100 had the lowest such values.
Mutagenic potencies of EOM and particles
The primary mutagenicity data (rev/plate) of the organic extracts of the PM2.5 of the emissions of the various diesel fuels in strains TA98 and TA100 of Salmonella are shown in Table 2. All extracts were mutagenic in all strains in which they were tested, both with and without S9. Where replicate data were available in a strain/S9 condition, we combined the replicate data to generate linear regressions as described in the Materials and Methods. The slopes of these regressions (rev/μg EOM) were the mutagenic potencies of the EOM (Table 3).
Table 2.
Mutagenicity of organic extracts of emissions in Salmonella
| Rev/platea | ||||||||
|---|---|---|---|---|---|---|---|---|
| Dose (μg | TA98 −S9 | TA98 +S9 | TA100 | |||||
| Sample | (EOM/plate) | Exp 1 | Exp 2 | Exp 3 | Exp 1 | Exp 2 | −S9 | +S9 |
| B0 | 0 | 24 | 25 | 22 | 29 | 25 | 115 | NDc |
| 10 | 47 | NDc | NDc | NDc | NDc | NDc | NDc | |
| 25 | 54 | 65 | 50 | 130 | 74 | 154 | NDc | |
| 50 | 75 | 108 | 69 | 221 | 99 | 222 | NDc | |
| 100 | 120b | 204b | 112b | 316 | 153 | 376 | NDc | |
| Soy B5 | 0 | 24 | 25 | 22 | 29 | 25 | 115 | 131 |
| 10 | 48 | 60 | 37 | 90 | 52 | 158 | 217 | |
| 25 | 84 | 120 | 73 | 164 | 95 | 240 | 494 | |
| 50 | 142 | 194 | 104 | 305 | 109 | 360 | 675 | |
| 100 | 259 | NDc | NDc | NDc | NDc | NDc | NDc | |
| Canola B5 | 0 | 24 | 25 | 22 | 29 | 25 | 115 | NDc |
| 10 | 32 | NDc | NDc | NDc | NDc | |||
| 25 | 65 | 95 | 45 | 160 | 77 | 198 | NDc | |
| 50 | 94 | 163 | 85 | 249 | 128 | 343 | NDc | |
| 100 | 192 | 275 | 144 | 357b | 159b | 569 | NDc | |
| Canola B20 | 0 | 24 | 25 | 22 | 29 | 25 | 115 | NDc |
| 10 | 51 | NDc | NDc | NDc | NDc | NDc | ||
| 25 | 80 | 103 | 63 | 154 | 73 | 263 | NDc | |
| 50 | 131 | 163 | 94 | 252 | 131 | 455 | NDc | |
| 100 | 216 | 361 | 169 | 405b | 192b | 705 | NDc | |
| Canola B50 | 0 | 24 | 25 | 22 | 29 | 25 | 115 | NDc |
| 25 | NDc | 67 | 72 | 148 | 71 | 183 | NDc | |
| 50 | 88 | 128 | 102 | 241 | 106 | 263 | NDc | |
| 100 | NDc | 330 | NDc | 398 | NDc | NDc | NDc | |
| 125 | 219 | NDc | 187b | NDc | 198b | 551 | NDc | |
| 250 | 422 | NDc | NDc | NDc | NDc | NDc | NDc | |
| Canola B100 | 0 | 24 | 25 | 22 | 29 | 25 | NDc | NDc |
| 50 | NDc | 145 | 80 | 226 | 91 | NDc | NDc | |
| 100 | 185 | NDc | NDc | NDc | NDc | NDc | NDc | |
| 125 | NDc | 366 | 177 | 432 | 177 | NDc | NDc | |
| 250 | 471 | 678 | 435 | 722 | 342 | NDc | NDc | |
| 500 | 892 | NDc | NDc | NDc | NDc | NDc | NDc | |
| WVO B100 | 0 | 24 | 25 | 22 | 29 | 25 | 115 | 131 |
| 5 | 28 | NDc | NDc | NDc | NDc | NDc | NDc | |
| 10 | 47 | NDc | NDc | NDc | NDc | NDc | NDc | |
| 25 | 55 | NDc | NDc | NDc | NDc | NDc | NDc | |
| 50 | 65 | 125 | 64 | 196 | 75 | 157 | 418 | |
| 100 | 123 | NDc | NDc | NDc | NDc | NDc | NDc | |
| 125 | NDc | 225 | 145 | 392 | 133 | 386 | 862 | |
| 250 | NDc | 445 | 234 | 614 | 262 | 728 | 1138b | |
Data are the results from single plates/dose except for the DMSO control (0), where each value is the average of 3 plates. Positive controls were used with each experiment at 3 plates/dose; positive control data (average rev/plate, range) were 2-aminoanthracene (+S9) at 0.5 μg/plate for TA98 (650, 598–676) and TA100 (1254, 1219–1296); sodium azide (−S9) at 3 μg/plate for TA100 (626, 579–656); and 2-nitrofluorene (−S9) at 3 μg/plate for TA98 (574, 367–863).
These data were not used in the linear regressions because they were outside of the linear portion of the dose-response curves.
ND = not done.
Table 3.
Mutagenic potencies (rev/μg EOM) of organic extractsa
| Sample | TA98 −S9 | TA98 +S9 | TA100 −S9 | TA100 +S9 |
|---|---|---|---|---|
| B0 | 1.2 | 2.0 | 2.7 | Not done |
| Soy B5 | 2.4 | 3.6 | 5.0 | 11.3 |
| Canola B5 | 1.8 | 3.2 | 4.6 | Not done |
| Canola B20 | 2.2 | 2.3 | 6.0 | Not done |
| Canola B50 | 1.6 | 3.6 | 3.5 | Not done |
| Canola B100 | 1.8 | 2.0 | Not done | Not done |
| WVO B100 | 1.3 | 1.6 | 2.6 | 5.9 |
Data are the slopes of linear regressions calculated from the data in Table 2.
The mutagenic potencies of the EOMs were greater in TA98 +S9 than in TA98 −S9, indicating a relatively greater role for PAHs (+S9) than nitro-PAHs (nitroarenes) (−S9) in the mutagenicity of these organic extracts in TA98 (Table 3). Although such a comparison was possible for only two of the EOMs in TA100, we found that both EOMs were 2–3 times more mutagenic in the presence than the absence of S9 (Table 3). The mutagenic potencies of the EOMs of the canola blends were higher than those of B0 in all strain/S9 conditions. In contrast, those of WVO were either the same or lower than that of B0 (Table 3).
By multiplying the mutagenic potencies in Table 3 by the % EOM values (footnote of Table 4), we generated the mutagenic potencies of the PM2.5 (rev/mg PM2.5) (Table 4), which is the same as the mutagenic potencies of the particles (rev/mg particle). As with the EOM, the mutagenic potencies of the PM2.5 were generally greater in the presence of S9 than the absence of S9 in TA98, indicating a relatively greater role for PAHs than nitroarenes in the mutagenicity of the PM2.5. As with the EOM, the highest mutagenic potencies of the PM2.5 were in TA100 +S9, which is the strain/S9 combination most sensitive for the detection of PAHs. The mutagenic potencies of the PM2.5 of the biodiesels were either similar to or, more typically, greater than those of petroleum diesel (B0) (Table 4).
Table 4.
Mutagenic potencies (rev/mg PM) of organic extractsa
| Sample | TA98 −S9 | TA98 +S9 | TA100 −S9 | TA100 +S9 |
|---|---|---|---|---|
| B0 | 400 | 600 | 800 | Not done |
| Soy B5 | 600 | 900 | 1200 | 2700 |
| Canola B5 | 500 | 1000 | 1400 | Not done |
| Canola B20 | 500 | 500 | 1400 | Not done |
| Canola B50 | 500 | 1100 | 1100 | Not done |
| Canola B100 | 600 | 700 | Not done | Not done |
| WVO B100 | 500 | 600 | 1000 | 2300 |
These values were calculated by first multiplying the values in Table 3 by the % EOM of each sample and then multiplying by 1000 to convert μg to mg. The % EOM values were 31.2% for B0, 23.9% for B5 soy, 30.2% for B5 canola, 23.4% for B20 canola, 30.8% for B50 canola, 35.2% for B100 canola, and 39.6% for B100 WVO.
Mutagenicity emission factors
Using the formulas described in the Materials and Methods, we calculated the mutagenicity emission factors expressed as rev/m3 exhaust (Table 5), rev/kg fuel burned (Table 6), and rev/MJth energy released (Table 7). Dividing the values expressed as rev/m3 exhaust (Table 5) by 1000 gives rev/L of exhaust (data not shown). As with the mutagenic potencies, nearly all the mutagenicity emission factors in any one strain were greater in the presence than the absence of S9 (Tables 5–7), indicating a relatively greater role for PAHs, which require S9 to be mutagenic, than nitroarenes, which do not.
Table 5.
Mutagenic emission factors (rev × 104/m3) for the fuelsa
| Sample | TA98 −S9 | TA98 +S9 | TA100 −S9 | TA100 +S9 |
|---|---|---|---|---|
| B0 | 1.0 | 1.4 | 1.9 | Not done |
| Soy B5 | 1.4 | 2.1 | 2.7 | 6.2 |
| Canola B5 | 0.8 | 1.6 | 2.2 | Not done |
| Canola B20 | 1.0 | 1.0 | 2.8 | Not done |
| Canola B50 | 0.9 | 2.0 | 2.0 | Not done |
| Canola B100 | 0.9 | 1.1 | Not done | Not done |
| WVO B100 | 0.7 | 0.8 | 1.3 | 3.0 |
Table 6.
Mutagenic emission factors (rev × 105/kg fuel) for the fuelsa
| Sample | TA98 −S9 | TA98 +S9 | TA100 −S9 | TA100 +S9 |
|---|---|---|---|---|
| B0 | 2.7 | 4.0 | 5.4 | Not done |
| Soy B5 | 3.7 | 5.5 | 7.3 | 16.0 |
| Canola B5 | 2.1 | 4.3 | 6.0 | Not done |
| Canola B20 | 2.7 | 2.7 | 7.6 | Not done |
| Canola B50 | 2.3 | 5.1 | 5.1 | Not done |
| Canola B100 | 2.1 | 2.5 | Not done | Not done |
| WVO B100 | 1.5 | 1.9 | 3.1 | 7.1 |
Table 7.
Mutagenic emission factors (rev × 104/MJth) for the fuelsa
| Sample | TA98 −S9 | TA98 +S9 | TA100 −S9 | TA100 +S9 |
|---|---|---|---|---|
| B0 | 0.6 | 0.9 | 1.2 | Not done |
| Soy B5 | 0.8 | 1.2 | 1.6 | 3.6 |
| Canola B5 | 0.5 | 0.9 | 1.3 | Not done |
| Canola B20 | 0.6 | 0.6 | 1.7 | Not done |
| Canola B50 | 0.5 | 1.2 | 1.2 | Not done |
| Canola B100 | 0.5 | 0.6 | Not done | Not done |
| WVO B100 | 0.4 | 0.5 | 0.8 | 1.8 |
The mutagenicity emission factors due to PAHs (TA98 +S9) of WVO B100 and canola B20 and B100 were lower relative to those of petroleum diesel (B0) regardless of the units of expression (Tables 5–7). The mutagenicity emission factors due to nitroarenes (TA98 −S9) of the WVO and all of the canola biodiesels were also lower relative to those of petroleum diesel (B0) regardless of the units of expression (Tables 5–7). Thus, the mutagenicity emission factors of most of the biodiesels that were due presumptively to PAHs and nitroarenes were lower than those of petroleum diesel.
The mutagenicity emission factors due to compounds that induced direct-acting base-substitution mutations (TA100 −S9) were lower only for B100 WVO relative to those of B0 regardless of the units of expression (Tables 5–7). In contrast, the mutagenicity emission factors in TA100 −S9 for all the remaining biodiesels were the same or greater than those of B0 regardless of the units of expression (Tables 5–7). This category of mutagenicity was generally higher for the canola biodiesels relative to petroleum diesel, indicating that canola biodiesel emissions produced more direct-acting, base-substitution mutagenicity than did petroleum diesel.
Discussion
Comparison of mutagenicity emission factors based on PM mass determined by gravimetric method using filters versus the TEOM method
As discussed in the Materials & Methods, there are at least two methods to determine the PM mass: gravimetric assessment of filters and TEOM. We are unaware of a previous comparison of mutagenicity emission factors determined by these two methods; however, we made such a comparison for a range of soy biodiesel blends. We determined the mutagenicity emission factors for several soy biodiesel blends in [11,12] but published only the filter data at the time. Here we show those filter data, corrected as described in the footnote to Table 8, along with the TEOM data, which we have calculated from data collected in Mutlu et al. [11,13] (Table 8; Figure 1). Both methods produced similar results, with the rev/MJth values either the same or slightly lower based on TEOM data relative to filter data; however, the trend across the blends was similar.
Table 8.
Corrected mutagenicity emission factors (rev × 104/MJth) of soy biodiesel using PM mass data based on filters and TEOM measurements from Mutlu et al. (2015)
| Source of PM mass | Sample | TA98 −S9 | TA98 +S9 | TA100 −S9 | TA100 +S9 |
|---|---|---|---|---|---|
| Filtersa | B0 | 1.2 | 1.7 | 3.0 | 12.0 |
| B20 | 0.7 | 0.6 | 1.8 | 4.1 | |
| B50 | 0.4 | 0.3 | 1.3 | 1.5 | |
| B100 | 0.4 | 0.2 | 0.8 | 1.5 | |
| TEOM | B0 | 1.0 | 1.4 | 2.5 | 9.9 |
| B20 | 0.7 | 0.5 | 1.8 | 3.9 | |
| B50 | 0.4 | 0.3 | 1.1 | 1.2 | |
| B100 | 0.3 | 0.2 | 0.7 | 1.3 |
As described in the Materials & Methods, we discovered, while preparing this manuscript, an error in a computer program that we had used in Mutlu et al. [11–13] to calculate the mutagenicity emission factors based on the mass of PM collected on filters for these soy biodiesels. Thus, incorporating the correction factors and using the data in Table 6 of Mutlu et al. [11], we show here the corrected filter values for TA98 and TA100. A corrigendum has been published in Inhalation Toxicology that is linked to Mutlu et al. [11–13] that shows the corrected values for all the other Salmonella strains expressed as rev/MJth as well as rev/kg fuel for B0 and soy B20, B50, and B100.
Figure 1.
Mutagenicity emission factors (rev/MJth) for soy biodiesel fuels of various blends relative to petroleum diesel (B0) in TA98 +S9, with the mass of the PM2.5 determined either by filter weights or TEOM measurements as described in the Materials & Methods. Data for all but B5 are from Table 8; data for B5 are from Table 7.
These results show that determination of PM mass by TEOM resulted in mutagenicity emission factors comparable to those determined gravimetrically from filters. This provides support for our present study in which we had only TEOM data to determine the PM mass. Figure 2 shows graphically the mutagenicity emission factors (rev/MJth) for the canola and WVO biodiesels using the data in Table 7. Although canola B50 was an outlier, canola B20 and B100 had mutagenicity emission factors lower than that of B0 (Figure 2).
Figure 2.
Mutagenicity emission factors (rev/MJth) for canola biodiesels and B100 WVO relative to petroleum diesel (B0) in TA98 +S9, with the mass of the PM2.5 determined by TEOM measurements as described in the Materials & Methods. All data are from Table 7.
Comparison of mutagenicity emission factors for 3 biodiesels and petroleum diesel
Based on TEOM data, Figure 3 shows a comparison of the mutagenicity emission factors (rev/MJth) for various blends of the biodiesels and petroleum diesel combusted under the same conditions in the same engine and evaluated for mutagenicity in three strain/S9 conditions based on data in Tables 7 and 8. Figure 4 is a sub-set of Figure 3 showing the results for all three B100 biodiesels and petroleum diesel in TA98 +S9. Regardless of blend or strain/S9 conditions, all three biodiesels except canola B50 in TA98 +S9 had lower mutagenicity emission factors than did petroleum diesel. The B20 blends of canola and soy had similar rev/MJth values, but these were still as much as 50% less than those of B0 in TA98 +S9 (Figure 3A). Except for canola B50 in TA98 +S9, the B50 and B100 biodiesels ranked in TA98 +/− S9 as petroleum > canola > WVO > soy (Figure 3). As with B20, the B50 and B100 biodiesels had mutagenicity emission factors in TA100 −S9 that were lower than those of petroleum diesel, but their values were similar to each other.
Figure 3.
Mutagenicity emission factors (rev/MJth) of biodiesels and petroleum diesel in three strain/S9 conditions. Values are from Tables 7 and 8, and all are based on TEOM data. B0 values are the average of those in Table 7, which are all TEOM values, and the TEOM values in Table 8.
Figure 4.
Mutagenicity emission factors (rev/MJth) of three biodiesels and petroleum diesel in TA98 +S9. Values are from Tables 7 and 8, and all are based on TEOM data.
This analysis is the first comparative assessment of the mutagenicity emission factors for a variety of different biodiesel fuels. Recently, Bűnger et al. [18] found a high correlation between the mutagenicity emission rate (rev/L) and the number of unsaturated fatty acids based on data for biodiesels made from rapeseed oil, palm tree oil, linseed oil, and coconut oil. Thus, the higher the number of double bonds in the fatty acids, the higher the mutagenicity emission rate. Figure 4 compares the rev/MJth for canola, soy, and WVO B100 relative to petroleum diesel. However, we were unable to correlate these results to the number of unsaturated fatty acids in each fuel because the saturated fatty acid values were similar for soy and canola and unknown for WVO.
The distinction between mutagenic potency and mutagenicity emission factor explains the disparity in the literature on the mutagenicity of biodiesels
As described in the Results, the mutagenic potencies of the EOM and PM2.5 for the canola and WVO biodiesel emissions were generally greater than those of B0 in nearly all strain/S9 conditions, whereas the mutagenicity emission factors were generally less for the biodiesel emissions than for those of B0. This was not the case for soy biodiesel, where both expressions of mutagenic potency as well as the mutagenicity emission factors were generally less for the biodiesel blends than for B0 [11]. A comparison of the various fuels burned under the same conditions in our laboratory showed that three different biodiesels had mutagenicity emission factors that were generally lower than that of petroleum diesel (Figure 3). We consider the mutagenicity emission factors to be the most meaningful expression of the mutagenicity of the emissions, and those comparative data clearly indicate that the emissions from a variety of biodiesel fuels are generally less mutagenic than those from petroleum diesel.
Other than our study of soy biodiesel [11–13] and our present study of canola and WVO biodiesel, only one other study (of rapeseed biodiesel) permitted the calculation of a mutagenicity emission factor [14]. We suggest that much of the disparity in the literature as to whether biodiesel emissions are less mutagenic than those of petroleum diesel is because essentially all studies have reported only the mutagenic potency of the EOM or PM, which can be greater or smaller for biodiesels relative to petroleum diesel. In contrast, the mutagenicity emission factors, which we have determined under the same combustion conditions for soy, canola, and WVO, are generally lower for biodiesels relative to petroleum diesel.
Comparison of biodiesel and diesel emissions with other combustion emissions
We have determined the mutagenicity emission factors for a wide variety of combustion emissions other than for biodiesel emissions [17,19,20–22], and Figure 5 compares the rev/MJth in strain TA98 +S9 of the biodiesel emissions reported here to some of these other combustion emissions. The petroleum diesel and biodiesel mutagenicity emission factors are generally greater, sometimes by several orders of magnitude, than those associated with highly controlled combustion, such as utility-sized combustors using natural gas, oil, coal, or wood. However, they are several orders of magnitude lower than those associated with highly uncontrolled, open burning, such as residential wood fireplaces or the open burning of tires.
Figure 5.
Comparison of the mutagenicity-emission factors for a variety of combustion emissions. The three sets of biodiesel studies are identified by bars with different fillings; studies with the same filling refer to the petroleum diesel (B0) in that set. Data for canola, waste vegetable oil (WVO), and its associated B0 are TEOM data from Table 7; data for soy and its associated B0 are TEOM data from Table 8. Data for the 150-kW engine are from Turrio-Baldassarri et al. [14] calculated as described in Mutlu et al. [11]. The remaining data (black bars) are from DeMarini et al. [19].
These comparative data illustrate the potential health effects associated with various types of diesel exhaust relative to other combustion emissions that contribute to air pollution. The potential health effects of diesel/biodiesel emissions should be considered within the context of these other emissions, given the genotoxicity of traffic exhaust to humans [7] and the risk for lung cancer from petroleum diesel [8] and air pollution [9], which is frequently contaminated with diesel emissions.
Other toxicological effects of biodiesel emissions
As reviewed [11], two studies found that the emissions from rapeseed biodiesel induced less DNA damage in mammalian cells in vitro (comet assay) than did those from petroleum diesel. Likewise, organics from the PM of emissions of rapeseed B100 biodiesel [23] and soy B20 biodiesel [24] induced lower levels of DNA adducts in calf thymus DNA in vitro and in Salmonella [24] exposed to the organics from the PM of soy biodiesel emissions. Exposure of cells at the air-liquid interface to biodiesel emissions from an engine equipped with emission controls did not induce DNA adducts [25]. Studies in human lung cell lines (A549 and BEAS-2B) found little difference in the levels of single-strand DNA breaks, micronuclei, cytotoxicity, oxidative stress, or gene expression between the emissions from rapeseed biodiesels or petroleum diesel [26,27]. A recent study of rapeseed biodiesel found little difference between the mutagenicity in Salmonella and cytotoxicity in human embryonic kidney 293T cells of organic extracts of PM from rapeseed biodiesel versus that of petroleum diesel [28].
As reviewed by Bűnger et al. [10] and Madden [29], studies in rodents exposed to the emissions from various types of biodiesel (primarily rapeseed) have given mixed results relative to exposures to emissions from petroleum diesel. However, recent studies with soy biodiesel found that B20 induced lower levels of inflammatory effects in mice [30] and was less arrhythmogenic in rats [31] relative to petroleum diesel. The emissions from soy B100 biodiesel (but not B20) had generally lower cardiovascular impacts on rats than did those from petroleum diesel [32].
Implications for public health
As with soy biodiesel [11–13], most of the reduction in the mutagenicity emission factor for canola and WVO was due to reduced concentrations of chemical classes that included PAHs, nitroarenes, and possibly aromatic amines and oxy-PAHs. Based on the Salmonella strains used here, we can infer that the lower mutagenicity emission factors for canola and WVO biodiesel relative to petroleum diesel were also due to a reduction in PAHs and nitroarenes.
The health effects of the emissions of biodiesel fuels are only one aspect of an assessment of biodiesel fuels. Other considerations include ethical concerns associated with the use of food crops for fuels [33]; the impact on greenhouse-gas emissions [34]; and the overall life-cycle assessment of the energy needed to grow, harvest, and process plants for biofuel use versus the energy obtained from the fuel [35]. Recent studies have shown that WVO may have the lowest environmental impact as well as the highest economic benefit compared to pristine plant-derived oils [3,4]. Our results show that although the mutagenicity emission factors for WVO were not the lowest (soy was the lowest) among the biodiesels we studied, they were as much as 50% lower than those of petroleum diesel. These data can contribute to a more complete life-cycle assessment of biodiesel fuels and provide additional information to support policy decisions regarding the promotion and use of biodiesel fuels.
Acknowledgements
This research was funded by the intramural research program of the National Health and Environmental Effects Research Laboratory, Office of Research and Development, U.S. Environmental Protection Agency, Research Triangle Park, NC. We thank Brian Chorley and Michael Lewandowski for their helpful comments on this manuscript. We thank Q. Todd Krantz for his help in conducting the engine experiments. This article was reviewed by the National Health and Environmental Effects Research Laboratory, U.S. Environmental Protection Agency and approved for publication. Approval does not signify that the contents reflect the views of the agency nor does mention of trade names or commercial products constitute endorsement or recommendation for use.
Footnotes
* Dedicated to Dr. Bruce N. Ames, whose Salmonella (Ames) mutagenicity assay was the first assay to permit the routine evaluation of complex environmental mixtures for mutagenicity.
Conflict of interest
None.
References
- [1].Knothe G, History of vegetable oil-based diesel fuels, in: Knothe G, Krahl J, Van Gerpen J (Eds.), The Biodiesel Handbook, second ed, AOCS Press, Urbana, IL, 2010, pp. 5–19. [Google Scholar]
- [2].Mujeeb MA, Vedamurthy AB, Shivasharana CT, Current strategies and prospects of biodiesel production: a review, Adv. App. Sci. Res 7 (2016) 120–133. [Google Scholar]
- [3].Liang S, Xu M, Zhang T, Life cycle assessment of biodiesel production in China, Bioresource Technol. 129 (2013) 72–77. [DOI] [PubMed] [Google Scholar]
- [4].Fawaz EG, Salam DA, Preliminary economic assessment of the use of waste frying oils for biodiesel production in Beirut, Lebanon, Sci. Total Environ 637–638 (2018) 1230–1240. [DOI] [PubMed] [Google Scholar]
- [5].Canola Council of Canada. 2017. What is Canola? https://www.canolacouncil.org/oil-and-meal/what-is-canola/ [accessed 15 December 2018].
- [6].Lewtas J, Toxicological Effects of Emissions from Diesel Engines, Developments in Toxicology and Environmental Science, Vol 10, Elsevier Biomedical, New York, 1982. [Google Scholar]
- [7].DeMarini DM, Genotoxicity biomarkers associated with exposure to traffic and near-road atmospheres: a review, Mutagenesis 28 (2013) 485–505. [DOI] [PubMed] [Google Scholar]
- [8].International Agency for Research on Cancer, IARC Monographs on the Evaluation of Carcinogenic Risks to Humans. Diesel and gasoline engine exhausts and some nitroarenes. IARC Mongr. Eval. Carcinog. Risks Hum 105 (2014) 1–704. [Google Scholar]
- [9].International Agency for Research on Cancer, IARC Monographs on the Evaluation of Carcinogenic Risks to Humans. Outdoor air pollution. IARC Mongr. Eval. Carcinog. Risks Hum 109 (2016) 1–448. [Google Scholar]
- [10].Bűnger J, Krahl J, Munack A, Schröder O, Schmidt L, Westphal GA, Potential hazards associated with combustion of bio-derived versus petroleum-derived diesel fuel, Crit. Rev. Toxicol 42 (2012) 732–750. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [11].Mutlu E, Warren SH, Matthews PP, King C, Walsh L, Kligerman AD, Schmid JE, Janek D, Kooter IM, Linak WP, Gilmour MI, DeMarini DM, Health effects of soy-biodiesel emissions: mutagenicity-emission factors, Inhal. Toxicol 27 (2015) 585–596. [DOI] [PubMed] [Google Scholar]
- [12].Mutlu E, Warren SH, Matthews PP, Schmid JE, Kooter IM, Linak WP, Gilmour MI, DeMarini DM, Health effects of soy-biodiesel emissions: bioassay-directed fractionation for mutagenicity, Inhal. Toxicol 27 (2015) 597–612. [DOI] [PubMed] [Google Scholar]
- [13].Multu E, Nash DG, King C, Krantz TQ, Preston WT, Kooter IM, Higuchi MM, DeMarini DM, Linak WP, Gilmour MI, Generation, characterization, and comparison of diesel engine emissions from petroleum and soybean biodiesel fuels and the application of an inhalation exposure system for studying their relative toxicity, Inhal. Toxicol 27 (2015) 515–532. [DOI] [PubMed] [Google Scholar]
- [14].Turrio-Baldassarri L, Battistelli CL, Conti L, Crebelli R, De Berardis B, Iamiceli AL, Gambino M, Iannaccone S, Emission comparison of urban bus engine fuel with diesel oil and “biodiesel” blend, Sci. Total Environ 327 (2004) 147–162. [DOI] [PubMed] [Google Scholar]
- [15].Mehta PS, Anand K, Estimation of a lower heating value of vegetable oil and biodiesel fuel, Energy & Fuels 23 (2009) 3893–3898. [Google Scholar]
- [16].Maron DM, Ames BN, Revised methods of the Salmonella mutagenicity test, Mutation Res. 113 (1983) 173–215. [DOI] [PubMed] [Google Scholar]
- [17].Mutlu E, Warren SH, Ebersviller SM, Kooter IM, Schmid JE, Dye JA, Linak WP, Gilmour MI, Jetter JJ, Higuchi M, DeMarini DM, Mutagenicity and pollutant emission factors of solid-fuel cookstoves: comparison with other combustion sources, Environ. Health Perspect 124 (2016) 974–982. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [18].Bűnger J, Bűnger JF, Krahl J, Munack A, Schrȍder O, Brűning T, Hallier E, Westphal GA, Combusting vegetable oils in diesel engines: the impact of unsaturated fatty acids on particle emissions and mutagenic effects of the exhaust, Arch. Toxicol 90 (2016) 1471–1479. [DOI] [PubMed] [Google Scholar]
- [19].DeMarini DM, Lemieux PM, Ryan JV, Brooks LR, Williams RW, Mutagenicity and chemical analysis of emissions from the open burning of scrap rubber tires, Environ. Sci. Technol 28 (1994) 136–141. [DOI] [PubMed] [Google Scholar]
- [20].DeMarini DM, Warren SH, Lavrich K, Flen A, Aurell J, Mitchell W, Greenwell D, Preston W, Schmid JE, Linak WP, Hays MD, Samet JM, Gullett BK, Mutagenicity and oxidative damage induced by an organic extract of the particulate emissions from a simulation of the Deepwater Horizon surface oil burns, Environ. Mol. Mutagen 58 (2017) 162–171. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [21].Kim YH, warren SH, Krantz QT, King C, Jaskot R, Preston WT, Hays MD, Landis MS, Higuchi M, DeMarini DM, Gilmour MI, Mutagenicity and lung toxicity of smoldering versus flaming emissions from various biomass fuels: implications for health effects from wildland fire, Environ. Health Perspect 126 (2018) 017011. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [22].Mutlu E, Warren SH, Matthews PP, King C, Linak WP, Kooter IM, Schmid JE, Ross JA, Gilmour MI, DeMarini DM, Bioassay-directed fractionation and sub-fractionation for mutagenicity and chemical analysis of diesel exhaust particles, Environ. Mol. Mutagen 54 (2013) 719–736. [DOI] [PubMed] [Google Scholar]
- [23].Topinka J, Milcova A, Schmuczerova J, Mazac M, Pechout M, Voitisek-Lom M, Genotoxic potential of organic extracts from particle emissions of diesel and rapeseed oil powered engines. Toxicol. Lett 212 (2012) 11–17. [DOI] [PubMed] [Google Scholar]
- [24].Ross JA, Nelson GB, Mutlu E, Warren SH, Gilmour MI, DeMarini DM, DNA adducts induced by in vitro activation of extracts of diesel and biodiesel exhaust particles, Inhal. Toxicol 27 (2015) 576–584. [DOI] [PubMed] [Google Scholar]
- [25].André V, Barraud C, Capron D, Preterre D, Keravec V, Vendeville C, Cazier F, Pottier D, Morin JP, Sichel F, Comparative mutagenicity and genotoxicity of particles and aerosols emitted by the combustion of standard vs. rapeseed methyl ester supplemented bio-diesel fuels: Impact of after treatment devices: oxidation catalyst and particulate filter. Mutation Res 777 (2015) 33–42. [DOI] [PubMed] [Google Scholar]
- [26].Kowalska M, Wegierek-Ciuk A, Brzoska K, Joewodzka M, Meczynska-Wielgosz S, Gromadzka-Ostrowska J, Mruk R, Øvrevik J, Kruszewski M, Lankoff A, Genotoxoic potential of diesel exhaust particles from the combustion of first- and second-generation biodiesel fuels—the FuelHealth project, Environ. Sci. Pollut. Res 24 (2017) 24223–24234. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [27].Barraud C, Corbière C, Pottier I, Estace E, Blanchard K, Logie C, Lagadu S, Kéravec V, Pottier D, Dionnet F, Morin JP, Préterre D, André V, Monteil C, Sichel F, Impact of after-treatment devices and biofuels on diesel exhausts genotoxicity in A549 cells exposed at air-liquid interface, Toxicol. In Vitro 45 (2017) 426–433. [DOI] [PubMed] [Google Scholar]
- [28].Agarwal AK, Singh AP, Gupta T, Agarwal RA, Sharma N, Rajput P, Pandey SK, Ateeq B, Mutagenicity and cytotoxicity of particulate matter emitted from biodiesel-fueled engines, Environ. Sci. Technol., Environ. Sci. Technol 52 (2018) 14496–14507. [DOI] [PubMed] [Google Scholar]
- [29].Madden MC, A paler shade of green? The toxicology of biodiesel emissions: Recent findings from studies with this alternative fuel, Biochim. Biophys. Acta 1860 (2016) 2856–28862. [DOI] [PubMed] [Google Scholar]
- [30].Gavett SH, Wood CE, Williams MA, Cyphert JM, Boykin EH, Daniels MJ, Copeland LB, King C, Krantz TQ, Richards JH, Andrews DL, Jaskot RH, Gilmour MI, Soy biodiesel emissions have reduced inflammatory effects compared to diesel emissions in healthy and allergic mice, Inhal. Toxicol 27 (2015) 533–544. [DOI] [PubMed] [Google Scholar]
- [31].Hazari MS, Haykal-Coates N, Winsett DW, King C, Krantz QT, Gilmour MI, Farraj AK, The effects of B0, B20, and B100 soy biodiesel exhaust on aconitine-induced cardiac arrhythmia in spontaneously hypertensive rats, Inhal. Toxicol 27 (2015) 557–563. [DOI] [PubMed] [Google Scholar]
- [32].Farraj AK, Haykal-Coates N, Winsett DW, Gilmour MI, King C, Krantz QT, Richards J, Hazari MS, Comparative electrocardiographic, autonomic and systemic inflammatory responses to soy biodiesel and petroleum diesel emissions in rats, Inhal. Toxicol 27 (2015) 564–575. [DOI] [PubMed] [Google Scholar]
- [33].Buyx A, Tait J, Ethical framework for biofuels, Science 332 (2011) 540–541. [DOI] [PubMed] [Google Scholar]
- [34].Mullins KA, Griffin WM, Matthews HS, Policy implications of uncertainty in modeled life-cycle greenhouse gas emissions of biofuels, Environ. Sci. Technol 45 (2011) 132–138. [DOI] [PubMed] [Google Scholar]
- [35].McKone TE, Nazaroff WW, Berck P, Auffhammer M, Lipman T, Torn MS, Masanet E, Lobscheid A, Santero N, Mishra U, Barrett A, Bomberg M, Fingerman K, Scown C, Strogen B, Horvath A, Grand challenges for life-cycle assessment of biofuels, Environ. Sci. Technol 45 (2011) 17541–17556. [DOI] [PubMed] [Google Scholar]