Issues associated with solid particle measurement

Transcription

1 Issues associated with solid particle measurement David B. Kittelson, Jacob Swanson*, Heejung Jung** Department of Mechanical Engineering University of Minnesota *Engineering Department University of Cambridge, UK **Department of Mechanical Engineering University of California, Riverside PMP Meeting 6th December 2011 DG JRC, Ispra, Italy

2 Outline Background UCR on-road and laboratory tests UMN laboratory tests Conclusions

3 At very low emission levels number measurements have much better resolution than mass but.. Current PMP method regulates solid particles larger than 23 nm For engines equipped with particle filters regulating to 23 nm effectively regulates all sizes Under extreme conditions false counts of nucleated semi-volatile have been observed For engines without filters (advanced fuels, combustion modes, gasoline) there may be large concentrations of solid particles below 23 nm that are not counted by current method. The next generation of high efficiency direct injection gasoline engines are challenged by the current standard even with the 23 nm limit An international solid number standard for aircraft emissions is being considered and is likely use a lower counting limit to 10 nm Extending solid PM measurements to 10 nm Significant semi-volatile particles downstream of PMP VPR often observed No significant semi-volatile formation downstream of catalytic stripper in this size range Extending solid PM measurements to below 10 nm problematic Particles as small as sub 3 nm formed in large concentrations downstream of PMP VPR Some evidence of solid particle formation by thermal denuder Sub 10 nm particle formation observed downstream of CS under some conditions Downstream nucleation and solid particle losses greatly reduced in new mini CS Removal of sulfate or other low vapor pressure species critical

4 Recent papers raise issues about solid particle measurements, especially when applied to particles smaller than 23 nm Work done at University of California, Riverside, CE-CERT Johnson, Kent C., Thomas D. Durbin, Heejung Jung, Ajay Chaudhary, David R. Cocker III, Jorn D. Herner, William H. Robertson, Tao Huai, Alberto Ayala, and David Kittelson, Evaluation of the European PMP Methodologies during On-Road and Chassis Dynamometer Testing for DPF Equipped Heavy Duty Diesel Vehicles, Aerosol Science and Technology, 43: , Zheng, Zhongqing, Kent C. Johnson, Zhihua Liu, Thomas D. Durbin, Shaohua Hu, Tao Huai, David B. Kittelson, and Heejung Jung, Investigation of solid particle number measurement: Existence and nature of sub-23nm particles under PMP methodology, Journal of Aerosol Science 42 (2011) Work done at the University of Minnesota, CDR Swanson, Jacob and David Kittelson, Evaluation of thermal denuder and catalytic stripper methods for solid particle measurements, Journal of Aerosol Science, Volume 41, Issue 12, Pages Swanson, Jacob and David Kittelson, Evaluation of PMP APC and catalytic stripper methods for solid particle measurements, in preparation

5 PMP number measurement system A specially designed CPC with a lower size cutoff of 23 nm is used

6 Why solid, why only larger than 23 nm? 1.00E+10 Number Concentrations (Part./cm³) 7.00E E E E E E+08 DN/DLog Dp (Part./cm³) 1.00E E E+07 T dilution = 32 C, Primary DR ~ rpm, 50% load 1.00E E E Dilution Temperature ( C) Residence Time (ms) 1.00E Dp (nm) Residence time = 1000 ms 100 ms 230 ms Filtration Efficiency, % min 2.25 min 3.25 min 4.25 min 5.25 min 6.25 min 7.25 min 8.25 min 9.25 min min Dp, nm The concentration of volatile nucleation mode particles is very dependent on sampling conditions Most of these particles are smaller than 23 nm If the engine is fitted with a particle filter particles below 50 nm or so are very effectively removed Thus regulating solid particles above 23 nm is really regulating soot particles is effectively regulating all particles for a trap equipped Without a trap the story is different

7 Engine out, light-load, low soot conditions: Most of the number emissions are solid with Dp < 23 nm 6.0E+07 Carbonaceous soot Mode 1 solid 5.0E+07 Mode 2 solid dn/dlogd p (part/cm 3 ) 4.0E E E+07 CPC cutoff Ash Mode 5 solid 1.0E E D p (nm) Cummins 2004 ISM engine, BP 50 fuel, AVL modes

8 Spark ignition engines can also produce tiny solid nanoparticles, especially with metal additives 4.0E+07 CPC cutoff Transient EUDC Manganese solid 3.0E+07 Idle Manganese solid 50 kph Manganese solid dn/dlogdp, part/cm 3 2.0E E E Dp, nm Gidney, Jeremy T., Martyn V. Twigg, and David B. Kittelson, Effect of Organometallic Fuel Additives on Nanoparticle Emissions from a Gasoline Passenger Car, Environmental Science and Technology, v 44, n 7, p

9 Recently published results from SWRI show solid particles between 10 and 23 nm for a modern GDI car From: Khalek, Imad A., Thomas Bougher, and Jeff J. Jetter, Particle Emissions from a 2009 Gasoline Direct Injection Engine Using Different Commercially Available Fuels, SAE paper number

10 Gasoline engine running in pure HCCI mode shows no solid particles above 10 nm 9.0E E+07 dn/dlogdp (particles/cm 3 ) 8.0E E E E E E rpm Tin100C 1200rpm Tin100C CS 1500rpm Tin90C 1500rpm Tin90C CS 2000rpm Tin rpm Tin70C CS dn/dlogdp (particles/cm 3 ) 8.0E E E E E E rpm Tin100C CS 1500rpm Tin90C CS 2000rpm Tin70C CS 2.0E E E E E Particle diameter (nm) 0.0E Particle diameter (nm) Emissions depend upon speed, load, temperature in-cylinder thermal processing Solid PN measured with catalytic stripper (CS) total PN without Right plot shows solid fraction on 10x expanded scale Most of the particles emitted are volatile but the solid ones are very small From: Bika, Anil, Wei Fang, Luke Franklin, Bin Huang, and David Kittelson, Particle Emissions from a Soot Free Engine, 15th ETH-Conference on Combustion Generated Nanoparticles, June 2011.

11 Using a 23 nm cut with aircraft plumes will lead to undercounting a significant fraction of the particles 78% below 23 nm 41% below 23 nm Typical fast-scan size distributions measured from a J85-GE turbo-jet engine at various engine power settings: Han, H.-S.; Chen, D.-R.; Pui, D.Y.H.; and Anderson, B.E. A nanometer aerosol size analyzer (nasa) for rapid measurement of highconcentration size distributions, Journal of Nanoparticle Research, v 2, n 1, p 43-52, March 2000

12 Outline Background UCR on-road and laboratory tests UMN laboratory tests Conclusions

13 UCR on road tests using PMP protocol show unexpected solid particles and many particles below 23 nm A heavy-duty truck equipped with a CRT was tested on road and on a chassis dynamometer It showed large concentrations of solid particles below 23 nm at high load conditions These conditions favor sulfate particle formation. Filtration efficiency for particles below 23 nm should be very high. 1e+7 Primary dilution Secondary dilution MEL CVS MEL secondary diluter Filter sampling train (gravimetric) PMP diluter (Clone system) PMP diluter (Alternative system) Sampling & Measurement TSI 3790 CPC (23 nm) TSI 3760A CPC (11 nm) TSI 3025A CPC (3 nm) TSI 3760A CPC (11 nm) TSI 3025A CPC (3 nm) TSI 3022 CPC (7 nm) Particle concentration (#/cc) 1e+6 1e+5 1e+4 1e+3 1e+2 1e * 1e+0 ETC UrbanETC CruiseARB Creep UDDS Route 1 Route 2 Johnson, Kent C., Thomas D. Durbin, Heejung Jung, Ajay Chaudhary, David R. Cocker III, Jorn D. Herner, William H. Robertson, Tao Huai, Alberto Ayala, and David Kittelson, Evaluation of the European PMP Methodologies during On-Road and Chassis Dynamometer Testing for DPF Equipped Heavy Duty Diesel Vehicles, Aerosol Science and Technology, 43: , 2009.

14 UCR chassis dynamometer tests comparing APC and catalytic stripper (CS) Comparisons of fully compliant PMP system with measurement system using catalytic stripper for volatile particle removal Use a variety of counting instruments with different lower size cutoffs TSI nm TSI EEPS 6 nm TSI nm TSI nm TSI 3025A 3 nm TSI nm Tests with exhaust aerosols from heavy-duty vehicle operating on chassis dynamometer Freightliner class 8 truck with 14.6 liter, 2000 Caterpillar C-15 engine, equipped with Johnson Matthey Continuously Regenerating Trap (CRT TM ) Two steady state cruise conditions, constant speed 56 mph at 26% and 74% of full load Tests with laboratory challenge aerosols Zhongqing Zheng, Kent C. Johnson, Zhihua Liu, Thomas D. Durbin, Shaohua Hu, Tao Huai, David B. Kittelson, Heejung S. Jung, Investigation of solid particle number measurement: existence and nature of sub 23 nm particles under PMP methodology, Journal of Aerosol Science 42 (2011)

15 Comparisons of a Catalytic Stripper and APC Volatile Particle Remover with exhaust aerosols Exhaust from heavy-duty vehicle with CRT Zhongqing Zheng, Kent C. Johnson, Zhihua Liu, Thomas D. Durbin, Shaohua Hu, Tao Huai, David B. Kittelson, Heejung S. Jung, Investigation of solid particle number measurement: existence and nature of sub 23 nm particles under PMP methodology, Journal of Aerosol Science 42 (2011)

16 Steady state total particle number measurements 26% load 74% load Zhongqing Zheng, Kent C. Johnson, Zhihua Liu, Thomas D. Durbin, Shaohua Hu, Tao Huai, David B. Kittelson, Heejung S. Jung, Investigation of solid particle number measurement: existence and nature of sub 23 nm particles under PMP methodology, Journal of Aerosol Science 42 (2011)

17 Steady state 74% load strongly bimodal Accumulation mode Nuclei mode Zhongqing Zheng, Kent C. Johnson, Zhihua Liu, Thomas D. Durbin, Shaohua Hu, Tao Huai, David B. Kittelson, Heejung S. Jung, Investigation of solid particle number measurement: existence and nature of sub 23 nm particles under PMP methodology, Journal of Aerosol Science 42 (2011)

18 At 74% load PMP compliant system closely tracks the accumulation mode Zhongqing Zheng, Kent C. Johnson, Zhihua Liu, Thomas D. Durbin, Shaohua Hu, Tao Huai, David B. Kittelson, Heejung S. Jung, Investigation of solid particle number measurement: existence and nature of sub 23 nm particles under PMP methodology, Journal of Aerosol Science 42 (2011)

19 74% load 3790_APC (23 nm) and 3772_CS (10nm) The 3790 is always connected thru the PMP system and the 3772_CS is always connected thru the CS. The lower reading with the CS is due to thermophoretic losses ~40% APC500 PMP system, 500 dilution ratio APC100 PMP system, 100 dilution ratio CS Catalytic Stripper CVS sampling directly from main dilution tunnel, no removal of volatile particles Changing APC dilution ratio doesn t change results it shouldn t Zhongqing Zheng, Kent C. Johnson, Zhihua Liu, Thomas D. Durbin, Shaohua Hu, Tao Huai, David B. Kittelson, Heejung S. Jung, Investigation of solid particle number measurement: existence and nature of sub 23 nm particles under PMP methodology, Journal of Aerosol Science 42 (2011)

20 74% load 3790_APC (23 nm) on APC, 3772_CS (10 nm) on CS 3772 (10 nm) switched between APC and CS Switching from 100 to 500 overall dilution ratio has no impact on the APC results the desired result When connected thru the PMP system the 3790 and 3772 agree no particles between 10 and 23 nm Zhongqing Zheng, Kent C. Johnson, Zhihua Liu, Thomas D. Durbin, Shaohua Hu, Tao Huai, David B. Kittelson, Heejung S. Jung, Investigation of solid particle number measurement: existence and nature of sub 23 nm particles under PMP methodology, Journal of Aerosol Science 42 (2011)

21 74% load 3790_APC (23 nm) on APC, 3772_CS (10 nm) on CS 3776 (2.5 nm) switched between APC and CS When connected thru the PMP system the 3776 reads much higher than the 3790 due to particle formation below 10 nm, thru the stripper it also reads progressively higher than the 3772 again indicating sub 10 nm particles Zhongqing Zheng, Kent C. Johnson, Zhihua Liu, Thomas D. Durbin, Shaohua Hu, Tao Huai, David B. Kittelson, Heejung S. Jung, Investigation of solid particle number measurement: existence and nature of sub 23 nm particles under PMP methodology, Journal of Aerosol Science 42 (2011)

22 Summary of results at 74% load cruise Downstream of PMP system 3790 and 3772 agree no particles between 10 and 23 nn 3025A and 3776 agree and read progressively higher than 3772 and 3790 as time goes on particles forming between 3 and 10 nm Same trend at 100 and 500 dilution ratio Downstream of CS In first time window all instruments agree no particle below 23 nm In second and third time windows 3776 and 3025A read higher than 3772 particle formation between 3 and 10 nm Zhongqing Zheng, Kent C. Johnson, Zhihua Liu, Thomas D. Durbin, Shaohua Hu, Tao Huai, David B. Kittelson, Heejung S. Jung, Investigation of solid particle number measurement: existence and nature of sub 23 nm particles under PMP methodology, Journal of Aerosol Science 42 (2011)

23 26% load 3790_APC (23 nm) and 3772_CS (10 nm) Dependence on dilution ratio suggests particle formation or incomplete removal APC500 PMP system, 500 dilution ratio APC100 PMP system, 100 dilution ratio CS Catalytic Stripper CVS sampling directly from main dilution tunnel, no removal of volatile particles Changing APC dilution ratio changes results it shouldn t Zhongqing Zheng, Kent C. Johnson, Zhihua Liu, Thomas D. Durbin, Shaohua Hu, Tao Huai, David B. Kittelson, Heejung S. Jung, Investigation of solid particle number measurement: existence and nature of sub 23 nm particles under PMP methodology, Journal of Aerosol Science 42 (2011)

24 26% load 3790_APC (23 nm) on APC, 3772_CS (10 nm) on CS 3772 (10 nm) switched between APC and CS When connected thru the PMP system the 3790 and 3772 agree no particles between 10 and 23 nm but some additional particles above 23 nm at the lower APC dilution ratio Zhongqing Zheng, Kent C. Johnson, Zhihua Liu, Thomas D. Durbin, Shaohua Hu, Tao Huai, David B. Kittelson, Heejung S. Jung, Investigation of solid particle number measurement: existence and nature of sub 23 nm particles under PMP methodology, Journal of Aerosol Science 42 (2011)

25 26% load 3790_APC (23 nm) on APC, 3772_CS (10 nm) on CS 3776 (2.5 nm) switched between APC and CS When connected thru the PMP system the 3776 shows higher concentration than the 3790 indicating particles below 10 nm, especially at higher dilution ratio, little evidence of this with CS Zhongqing Zheng, Kent C. Johnson, Zhihua Liu, Thomas D. Durbin, Shaohua Hu, Tao Huai, David B. Kittelson, Heejung S. Jung, Investigation of solid particle number measurement: existence and nature of sub 23 nm particles under PMP methodology, Journal of Aerosol Science 42 (2011)

26 26% load 3025A (3 nm) and 3776 (2.5 nm) How can total be less than solid? The 3025A and 3776 disagree at the higher dilution ratio downstream of the PMP system but agree for 50 nm calibration aerosols suggesting that the particles are near the lower detection sizes of the instruments, < 3 nm.. When connected directly to CVS, no removal of volatiles, CPCs agree and show lower concentration than when volatiles are removed at 500 DR Zhongqing Zheng, Kent C. Johnson, Zhihua Liu, Thomas D. Durbin, Shaohua Hu, Tao Huai, David B. Kittelson, Heejung S. Jung, Investigation of solid particle number measurement: existence and nature of sub 23 nm particles under PMP methodology, Journal of Aerosol Science 42 (2011)

27 Summary of results at 26% load cruise Much lower concentrations than at 74% Downstream of PMP system In first time window, DR = and 3772 agree no particles between 10 and 23 nm 3776 and 3025A read much higher and disagree many particles below lower cutoff size of these instruments, 2.5 to 3 nm In second time window, DR = and 3772 read higher but agree no particles between 10 and 23 nm but formation above 23 nm 3776 and 3025A agree but read only slightly higher than 3790 and 3772 nearly all particles have grown to above 23 nm Downstream of CS Consistently lower reading and agreement between instruments In last time window instruments bypass volatile particle removal systems and are directly connect to CVS measure total solid and volatile particles fewer particles than DR = 500 APC, clear evidence of particle formation by APC Zhongqing Zheng, Kent C. Johnson, Zhihua Liu, Thomas D. Durbin, Shaohua Hu, Tao Huai, David B. Kittelson, Heejung S. Jung, Investigation of solid particle number measurement: existence and nature of sub 23 nm particles under PMP methodology, Journal of Aerosol Science 42 (2011)

28 Laboratory and engine tests both show concentration swings tracking VPR temperature Laboratory test aerosol Chassis dyno 74% load Zhongqing Zheng, Kent C. Johnson, Zhihua Liu, Thomas D. Durbin, Shaohua Hu, Tao Huai, David B. Kittelson, Heejung S. Jung, Investigation of solid particle number measurement: existence and nature of sub 23 nm particles under PMP methodology, Journal of Aerosol Science 42 (2011)

29 Summary UCR tests High load on-road tests of CRT equipped heavy-duty truck shows evidence of overloading PMP system Large increase in particle counts with all CPCs even PMP 3790 May be associated with storage and release of sulfate Chassis dynamometer tests at 74% and 26% load highway cruise At 74% load APC and CS agree above 10 nm and track accumulation mode APC and to a lesser extent CS show significant particle formation below 10 nm APC sub 10 nm concentration swings associated with evaporation tube temperature At 26% load APC shows dilution ratio dependence evidence of particle formation above 23 nm APC leads to formation of extremely small particles ~ 3 nm Laboratory generated aerosols show formation or incomplete removal of particles larger than 23 nm with evaporation tube temperature dependence.

30 Outline Background UCR on-road and laboratory tests UMN laboratory tests Conclusions

31 Performance of thermal denuder and catalytic stripper with ambient aerosol 1.E+07 1.E+06 Downstream TD dn/dlogdp, part/cm 3 1.E+05 1.E+04 1.E+03 Atmospheric aerosol Downstream CS 1.E Dp, nm Swanson, Jacob and David Kittelson, Evaluation of thermal denuder and catalytic stripper methods for solid particle measurements, Journal of Aerosol Science, Volume 41, Issue 12, Pages

32 Evaluating thermal denuder and catalytic stripper with pure semi-volatile challenge aerosols The test compounds Sulfuric acid (reagent grade 99.9% pure, 18 M) Solid tetracosane (C 24 H 50 ) flakes (reagent grade 99.9% pure) Dioctyl sebacate liquid (C 26 H 50 O 4 ) (technical grade 90% pure) An evaporation and condensation technique was used to generate nanometer sized particles. Test compounds were heated in an alumina combustion boat to various temperatures ranging from 120 C to 250 C. Temperatures were chosen such that the saturation vapor pressure above each compound was roughly the same, so that the evaporation rates would be similar and they would be similar mole fractions of each in the carrier gas stream at the time they were mixed. The vapor(s) were entrained with hot dry air, mixed and heated to 250 C, and then cooled by dilution with a dilution ratio of ~20:1. The cooling causes the vapors to nucleate, forming a high concentration of nanoparticles composed of the evaporated compounds. Swanson, Jacob and David Kittelson, Evaluation of thermal denuder and catalytic stripper methods for solid particle measurements, Journal of Aerosol Science, Volume 41, Issue 12, Pages

33 Test of TD and CS with laboratory generated semivolatile particles dn/dlogdp, part/cm 3 1.0E E E E E E E+02 Increasing challenge concentration Challenge aerosol 28 µg/m Dp, nm Downstream CS <0.01 Downstream TD + CS: <0.01 Challenge aerosol 14 µg/m 3 Challenge aerosol 380 µg/m 3 Pure tetracosane challenge aerosol Wide range of concentrations Particles form downstream of TD No particle downstream of CS No particles downstream of TD + CS particles formed by TD volatile 1.0E E+07 Downstream CS <0.01 part/cm 3 Downstream TD + CS: <0.01 part/cm 3 Tetracosane plus sulfuric acid challenge aerosol dn/dlogdp, part/cm 3 1.0E E E E+03 Downstream TD Challenge aerosol 32 µg/m 3 Significant particle formation downstream of TD No particle downstream of CS No particles downstream of TD + CS particles formed by TD volatile Swanson, Jacob and David Kittelson, Evaluation of thermal denuder and catalytic stripper methods for solid particle measurements, Journal of Aerosol Science, Volume 41, Issue 12, Pages E Dp, nm

34 Test of TD and CS with laboratory generated semivolatile particles 1.0E+08 dn/dlogdp, part/cm 3 dn/dlogdp, part/cm 3 1.0E E E E E E E E E E E E+03 Downstream TD + CS Downstream TD Downstream TD + CS Downstream CS Dp, nm Challenge aerosol 270 µg/m 3 Downstream CS: <0.01 Downstream TD 0.16 mg/m 3 Challenge aerosol 3.8 mg/m Dp, nm Tetracosane plus sulfuric acid challenge aerosols Higher challenge aerosol concentrations Significant particle formation downstream of TD A few particles downstream of CS at highest challenge aerosol concentrations Particles downstream of TD + CS particles suggesting solid particle formation by TD Swanson, Jacob and David Kittelson, Evaluation of thermal denuder and catalytic stripper methods for solid particle measurements, Journal of Aerosol Science, Volume 41, Issue 12, Pages

35 New miniature CS for use in APC Core length m Core width m Cell density 600 cells/in Wall thickness 2.5 mil Flowrate 1.5 L/min C 40 H 82 penetration 0.2 % SO 2-4 penetration % 50% cut off size 7 nm Penetration AVL APC - experimental Penetration curves for the CS and AVL APC. For the CS, the data are fitted to diffusion and thermophoresis loss model. For the APC, the data are fitted to a diffusion model only. 0.2 AVL APC - fitted to diffusion curve CS - experimental 0.0 CS - fitted to diffusion + thermophoresis curve Dp, nm

36 H 2 SO 4 particle generation method 1 L/min HEPA filter Cleaned compressed air H 2 O reservoir Mixing orifice To CS or SMPS H 2 SO 4 reservoir 4 L/min Hot plate Humidified dilution air RH measured = 70% H 2 O reservoir Initial acid concentration ~ 98%

37 H 2 SO 4 particle distributions 1.2E E mg/m mg/m mg/m 3 dn/dlogdp, part/cm 3 8.0E E E mg/m mg/m 3 28 mg/m 3 2.0E E Dp, nm

38 Some calculations of theoretical threshold for sulfuric acid nucleation and experimentally observed ~ 14 mg/m 3 Extremely low penetration necessary to prevent nucleation Experimental data indicates penetration ~ or ~99.99% removal efficiency Penetration Theorethical calculation of penetration required to give J = 1 part/m 3 - s Threshold concentration for nucleation in perfect system Predicted penetration based on diffusion theory and CS conditions H 2 SO 4 concentration, mg/m 3

39 Generation of hydrocarbon particles Cleaned compressed air 2 L/min HEPA filter Mixing orifice 1.E+06 To DMA for size selection 8.E+05 Dp = 60 nm C 40 H 82 reservoir σ g = 1.05 Hot plate Concentration = 5.6 ± 0.5 E4 part/cm 3 dn/dlogdp, part/cm 3 6.E+05 4.E+05 2.E+05 0.E Challenge particle Dp, nm

40 Hydrocarbon particle removal 1.E+05 10,000 part/cm 3 Upstream Downstream 1.E+04 Total concentration, part/cm 3 1.E+03 1.E+02 1.E+01 Corresponds to 99.98% removal 1.E Challenge particle Dp, nm

41 Conclusions UMN lab tests Thermal denuders and related devices have significant potential for downstream nucleation and at high concentrations may create solid residue from pure volatile materials CS system only shows downstream nucleation of sulfuric acid at very high upstream concentrations ~ 14 mg/m 3 APC VPR system shows downstream sulfuric acid nucleation at ~ 100µg/m 3 Renucleation of sulfuric acid or other low vapor pressure species critical need performance standard?

42 Outline Background UCR on-road and laboratory tests UMN laboratory tests Conclusions

43 Conclusions Current PMP method regulates solid particles larger than 23 nm For engines equipped with particle filters regulating to 23 nm effectively regulates all sizes Under extreme conditions false counts of nucleated semi-volatile have been observed For engines without filters (advanced fuels, combustion modes, gasoline) there may be large concentrations of solid particles below 23 nm that are not counted by current method The next generation of high efficiency direct injection gasoline engines are challenged by the current standard even with the 23 nm limit An international solid number standard for aircraft emissions is being considered and is likely use a lower counting limit to 10 nm Extending solid PM measurements to 10 nm Significant semi-volatile particles downstream of PMP VPR often observed No significant semi-volatile formation downstream of catalytic stripper in this size range Extending solid PM measurements to below 10 nm problematic Particles as small as sub 3 nm formed in large concentrations downstream of PMP VPR Some evidence of solid particle formation by thermal denuder Sub 10 nm particle formation observed downstream of CS under some conditions Downstream nucleation and solid particle losses greatly reduced in new mini CS Removal of sulfate or other low vapor pressure species critical

44 Acknowledgements All catalytic strippers used in this work were supplied by Martyn Twigg of Johnson-Matthey

45 Questions?

46 Backup slides

47 Outline H 2 SO 4 particle generation method Results for challenge concentrations 0.03 to 28 mg/m 3 Brief conclusion Previous test schematic and experimental results Calculations Hydrocarbon removal

48 H 2 SO 4 particle distributions for challenge concentrations <10 mg/m 3 1.0E E mg/m mg/m3 3.4 mg/m 3 dn/dlogdp, part/cm 3 1.0E E E E E mg/m mg/m 3 1.0E E+01 Downstream measurements are noise Dp, nm

49 Results show renucleation of H 2 SO 4 vapor for particle challenge concentrations ~28 mg/m 3 1.0E E mg/m mg/m mg/m3 9.3 mg/m 3 1.0E+07 Downstream CS at 28 mg/m 3 28 mg/m 3 dn/dlogdp, part/cm 3 1.0E E E E mg/m 3 1.0E E Dp, nm

50 Quick summary Current results suggest re-nucleation occurs between 10 and 28 mg/m 3 upstream

51 Past results H 2 SO 4 particle generation method 19 L/min 20: 1 dilution Cleaned compressed air To CS or SMPS Oxidation catalyst 1 L/min Permeation tube H 2 O reservoir HEPA filter SO 2 bottle

52 APC challenge with 12 mg/m 3 dn/dlogdp, part/cm 3 1.E+10 1.E+08 1.E+06 1.E+04 Downstream APC - particles are mostly the result of the re-nucleation of sulfuric acid Upstream mass concentation = 12 mg/m 3 1.E+02 Downstream APC + CS for mass concentation of 12 mg/m 3 - particles are solid residues Dp, nm

53 APC challenge with 1.0 mg/m 3 Highest inlet concentration possible without re nucleation downstream. Inlet concentration is really 100 µg/m 3 because aerosol is first diluted 10x. 1.0E E+08 Upstream mass concentration = 1.0 mg/m 3 dn/dlogdp, part/cm 3 1.0E E E+05 No re-nucleation downstream - these particles are solid residuals 1.0E E Dp, nm

54 CS challenge 12 mg/m 3 No difference between 1 or 2 CS units, corrected for size dependant penetration 1.E+09 1.E+08 Uptream mass concentration = 12 mg/m 3 dn/dlogdp, part/cm 3 1.E+07 1.E+06 1.E+05 1.E+04 1.E+03 Downstream CS Downstream CS + CS 1.E Dp, nm

55 CS challenge 16 mg/m 3 1.0E E+08 Use of 2 CS units shows sulfuric acid re nucleation Re-nucleation of H 2 SO 4 Uptream mass concentration = 16 mg/m 3 1.0E+07 dn/dlogdp, part/cm 3 1.0E E E E+03 Downstream CS Downstream CS + CS 1.0E Particle Diameter, nm

56 Conclusions New particle generation method using glass beakers eliminates solid particle residues Method is not completely stable due to constant uptake of water by the acid but it works Current results suggest renucleation occurs between 10 and 28 mg/m 3 upstream Consistent with past result where renucleation was observed at 12 mg/m 3 Compared to renucleation in APC at ~ 100µg/m 3