Supporting Information for. Diesel Particle Filter and Fuel Effects on Heavy-Duty Diesel Engine Emissions

Transcription

1 Supporting Information for Diesel Particle Filter and Fuel Effects on Heavy-Duty Diesel Engine Emissions Matthew Ratcliff, A. John Dane, Aaron Williams, John Ireland, Jon Luecke, Robert L. McCormick, and Kent J. Voorhees The attached supporting information is eleven pages in length and includes details of exhaust gas sampling and analysis for volatile unburned hydrocarbons, carbonyls and PM associated PAH and nitro-pah. Nine tables summarize engine specifications (S1), test cycle (S2), internal standard recoveries from PM extraction (S3) and measured emissions (S4-S9). There is a reference to an NREL report which provides full PM extraction procedure and analysis details.

2 Supporting Information Table S1. Test engine specifications. Cummins ISB Serial Number Displacement, L 5.9 Cylinders 6 Rated Power, kw Rated Torque Bore x Stroke 224 at 2500 rpm 895 N-m at 1600 rpm 10.2x12 cm Compression Ratio 16.5:1 Fuel System Common Rail S1

3 Table S2. Modified AVL 8-mode engine test cycle and turbo-outlet exhaust temperatures. Mode Speed Torque Sampling Temp Temp Temp (rpm) (ft-lb) Time ( o C) ( o C) ( o C) (min) ULSD B20 B Volatile and Semi-volatile Unburned Hydrocarbon Sampling and Analysis. Exhaust gas samples were collected from the raw exhaust in evacuated 3.2 L fused silica coated canisters. Pneumatically actuated high-temperature valves located before and after the DPF were programmed to open to canisters for the time periods listed in Table 1. Exhaust gas flow was regulated into the canisters by a Restek Veriflo flow controller. This system provided repeatable, time-weighted composite raw exhaust samples. The samples were analyzed for products of incomplete combustion with an Agilent 6890 gas chromatograph/5975 mass selective detector (GCMS) equipped with an Entech 7100 preconcentrator. A cold-trap dehydration (CTD) method was used to concentrate 200 ml of the collected exhaust sample. The CTD method used one empty trap and one Tenax S2

4 bead trap as the first two concentration stages. This was followed by trapping in a cryofocuser prior to desorption and transport to the GCMS column inlet. The GCMS was equipped with a Restek Rxi-1ms column (60 m x 0.32 mm with a 1 µm film). The temperature program started at -10 o C for 13 minutes, followed by a 20 o C/min ramp to 100 o C, which was held for 8 minutes, followed by a 30 o C/min ramp to 300 o C. A gas standard known as CRC Mix No. 4 from Scott Gas was used for calibration and quantitation of the C 3 to C 11 hydrocarbons in the samples. Carbonyl Sampling and Analysis. Carbonyl compounds were captured in Waters Sep- Pak dinitrophenylhydrazine cartridges. Exhaust was sampled from the dilution tunnel through the cartridges at a rate of 0.5 L/min for the eight time periods shown in Table S2, producing time-weighted composite samples. The carbonyls, trapped as dinitrophenylhydrazones, were eluted from the cartridges with acetonitrile and diluted to a measured mass of g. The solutions were analyzed by an Agilent 1200 highpressure liquid chromatography equipped with a Thermal Electron Delta Bond AK-Fast column (100 mm x 4.6 mm with 5 µm particle size) and gradient of acetonitrile/water solvent buffered to a ph of 2.5. Compound detection was by ultraviolet (UV) absorption. Particulate Matter Sampling and Extraction. The PM samples were collected from the dilution tunnel using a conventional PM sampling system comprising a sample probe, a stainless steel filter holder loaded with a pre-weighed Tissuquartz filter (Pall Corp. No. 7202, 47 mm diameter), a sample pump, and mass flow controller. The PM samples were collected at a flow rate of 60 L/min. The PM samples were obtained with either the DPF S3

5 removed (engine-out) or installed (DPF-out). Engine-out test cycles were run in triplicate using a new, pre-weighed Tissuquartz filter for each test. However, because the PM emissions were expected to be very low with the DPF installed, a single Tissuquartz filter was used to collect all the PM from the triplicate runs for the DPF-out samples. In this case, the final PAH and nitro-pah levels were calculated by dividing the total value by 3 to determine the per-test value. Particle filter handling and weighing were conducted in an environmental chamber/clean room with constant humidity and barometric pressure and temperature control. UV filtering acrylic covered the fluorescent lamps in the clean room to protect the PM samples from UV radiation; additional protection was provided by filter storage in amber petri dishes. Filter weighing was conducted on a Sartorius SC2 microbalance with a readability of 0.1 µg. After sample collection, the filters were allowed to equilibrate for a minimum of 12 hours before weighing. The final filter weights were then compared with the clean filter weights to compute the PM mass data. Subsequently, the samples were stored in a refrigerator until extraction and analysis was performed. Improved procedures for extracting the PAH and nitro-pah from the PM samples were developed during this project; details are reported elsewhere (1). Briefly, the filters were spiked with solutions of deuterated PAH and deuterated nitro-pah internal standards, then Soxhlet extracted with dichloromethane (DCM) a minimum of 200 cycles. The DCM solutions were placed in an ice bath and evaporated under flowing nitrogen to 10 ml. The DCM concentrates were diluted with pentane and purified by solid-phase extraction. Toluene was added to the extracted solutions and the solutions were again placed in an ice bath and evaporated S4

6 to a final volume of 200 µl (173 mg). Internal standard recoveries are shown in Table S3; the low recovery of naphthalene is related to its relatively high vapor pressure during evaporation steps. Table S3. Recovery rates of PAH and nitro-pah internal standards from PM filters. % Recoveries ULSD B20 B100 naphthalene-d 8 44 ± ± 2 12 ± 3 phenanthrene-d ± ± 5 84 ± 11 chrysene-d ± 7 62 ± ± 22 1-nitronaphthalene-d 7 69 ± ± 8 60 ± 9 1,5-dinitronaphthalene-d 6 87 ± ± 2 67 ± 16 9-nitroanthracene-d 9 87 ± ± 4 83 ± 18 1-nitropyrene-d 9 85 ± ± ± 19 PAH and nitro-pah Analysis. A four-sector MStation JMS-700T gas chromatographmass spectrometer (JEOL USA, Inc., Peabody, Massachusetts) equipped with an experimental JEOL trochoidal electron monochromator ionization source was used for the PAH and nitro-pah analyses. One µl of each sample was injected onto a Hewlett Packard 6890 Series GC system, equipped with an on-column injection port (Agilent Technologies, Palo Alto, California) maintained at the GC oven temperature. A 0.25 mm i.d. x 30 m RTX-5Sil MS column (Restek Corp, Bellefonte, Pennsylvania), with an intermediate polarity guard column (Supelco Inc., Bellefonte, Pennsylvania) connected S5

7 on the front end, was used to separate the various analytes. The GC was temperatureprogrammed using the following parameters: 100 C, hold 1 min; 10 C/min to 190 C, hold 2 min; 5 C/min to 240 C, hold 1 min; 10 C/min to 310 C, hold 3 min. The NO 2 - containing compounds were detected using selected ion monitoring (SIM) at 3.5 ev for m/z 46, allowing identification of each compound by retention time. SIM monitoring of the molecular ions at near 0 ev allowed confirmation of the molecular weights and quantification of the individual compounds. Nitrobenzene was used as the calibration standard to set the electron energies, while nitrobenzene, hexafluorobenzene, and perfluorokerosene standards (Sigma-Aldrich) were used to calibrate the masses monitored in SIM mode. Analyses for PAH were performed using positive ion electron ionization (EI+, ~25 ev), and the molecular ion chromatograms were used to quantify the amounts of each PAH. The MS conditions were as follows: two-sector operation, 6.0 kv accelerating voltage, 1000 resolution, and 100 ms switching rate (magnetic field control). The source temperature was 280 C. S6

8 Table S4. Engine-out nitro-pah mass emissions per mg PM. ULSD B20 B100 Ratio Ratio (ng/mg) (ng/mg) (ng/mg) ULSD/B20 ULSD/B100 1-nitronaphthalene 0.32 ± ± ± nitronaphthalene 1.44 ± ± ± nitroanthracene 9.96 ± ± ± nitrophenanthrene 3.89 ± ± ± nitrophenanthrene 0.80 ± ± ± nitrofluoranthene 0.80 ± ± ± nitropyrene 40.8 ± ± ± Total nitro-pah Table S5. Engine-out regulated emissions, weighted according to the AVL 8-mode cycle (2). ULSD g/bhp-hr B20 g/bhp-hr B100 g/bhp-hr Average Std. Dev. Average Std. Dev. Average Std. Dev. NO x THC CO S7

9 Table S6. Engine-out PAH quantities for each fuel per mg PM. ULSD (ng/mg) B20 (ng/mg) B100 (ng/mg) Ratio ULSD/B20 Ratio ULSD/B100 naphthalene 29.8 ± ± ± biphenyl 6.7 ± ± ± acenaphthylene 0.8 ± ± ± fluorene 4.7 ± ± ± phenanthrene ± ± ± anthracene 18.0 ± ± ± fluoranthene ± ± ± pyrene ± ± ± benzo[c]- phenanthrene 5.0 ± ± ± chrysene 6.8 ± ± ± Total PAH S8

10 Table S7. Summary of quantified nitro-pahs extracted from PM samples. Reported error is the standard deviation of triplicate samples analyzed. No error reported for singular DPF-out samples comprising PM from triplicate engine tests. DPF-out (ng/sample) Engine-out (ng/sample) Ratio of Engine-out / DPF-out shown in parentheses ULSD B20 B100 ULSD B20 B100 1-nitronaphthalene 0.95 ± ± ± 0.04 ND ND ND 2-nitronaphthalene 4.33 ± ± ± 0.18 ND ND ND 9-nitroanthracene 30.1 ± ± ± 1.89 ND ND ND 9-nitrophenanthrene 11.8 ± ± ± 0.37 ND ND ND 3-nitrophenanthrene 2.41 ± ± ± 0.37 ND ND ND 3-nitrofluoranthene 2.44 ± ± ± (35) 0.06 (19) 0.05 (13) 1-nitropyrene ± ± ± (1.5) 53.5 (1.1) 15.0 (1.8) Total nitro-pah (2.2) 53.6 (1.8) 15.1 (2.8) S9

11 Table S8. Summary of quantified PAHs extracted from PM samples. Reported error is the standard deviation of triplicate samples analyzed. No error reported for singular DPFout samples comprising PM from triplicate engine tests. DPF-out (ng/sample) Engine-out (ng/sample) Ratio of Engine-out / DPF-out shown in parentheses ULSD B20 B100 ULSD B20 B100 naphthalene 82.7 ± ± ± (2.9) 58.4 (1.1) 7.7 (22) biphenyl 16.4 ± ± ± (4.8) 9.6 (1.7) 1.9 (12) acenaphthylene 2.0 ± ± ± (25) 0.09 (21) 0.10 (18) fluorene 12.9 ± ± ± 0.5 ND ND ND phenanthrene ± ± ± (557) 1.7 (312) 11.5 (7.8) anthracene 57.7 ± ± ± (525) 0.07 (507) 0.12 (17) fluoranthene ± ± ± (391) 1.8 (331) 25.1 (7.5) pyrene ± ± ± (548) 0.25 (1328) 15.1 (8.7) benzo[c]- phenanthrene 13.2 ± ± ± 1.2 ND ND ND chrysene 19.2 ± ± ± 3.3 ND ND ND Total PAH (67) 71.9 (22) 61.5 (10) S10

12 Table S9. Summary of quantified carbonyls. Reported error is standard deviation of quadruplicate samples analyzed (triplicate for acrolein). DPF-out (ng/l) Engine-out (ng/l) Ratio of Engine-out / DPF-out shown in parentheses ULSD B20 B100 ULSD B20 B100 formaldehyde 147 ± ± ± ± 25 (0.63) 171 ± 26 (0.77) 100 ± 28 (1.31) acetaldehyde 51 ± 2 47 ± 3 43 ± 5 ND ND ND o-tolualdehyde 33 ± 5 26 ± 3 30 ± 10 ND ND ND acrolein 24 ± 10 9 ± 1 7 ± 1 ND ND ND acetone 13 ± 1 14 ± 1 14 ± 1 ND ND ND 2,5-dimethylbenzaldehyde 12 ± 7 12 ± 5 9 ± 4 ND ND ND propionaldehyde 10 ± 1 10 ± 2 11 ± 1 ND ND ND valeraldehyde 9 ± 4 8 ± 1 8 ± 3 ND ND ND butyraldehyde 8 ± 2 7 ± 3 9 ± 2 ND ND ND benzaldehyde 7 ± 2 5 ± 2 2 ± 3 ND ND ND Total Carbonyls 314 ± ± ± ± 25 (1.3) 171 ± 26 (1.6) 100 ± 28 (2.6) (1) Dane, J.; Voorhees, K.J. Investigation of Nitro-Organic Compounds in Diesel Engine Exhaust Final Report. NREL/SR , May (2) AVL 8-Mode Heavy-Duty Cycle, 2008; S11