Grantee Research Project Results
Final Report: Continuous Measurement of PM2.5 and Associated Semi-Volatile Particulate Species
EPA Grant Number: R825367Title: Continuous Measurement of PM2.5 and Associated Semi-Volatile Particulate Species
Investigators: Eatough, Delbert J.
Institution: Brigham Young University
EPA Project Officer: Chung, Serena
Project Period: October 1, 1996 through September 30, 1999
Project Amount: $353,103
RFA: Analytical and Monitoring Methods (1996) RFA Text | Recipients Lists
Research Category: Air Quality and Air Toxics , Environmental Statistics , Water , Land and Waste Management , Air , Ecological Indicators/Assessment/Restoration
Objective:
The hypothesis of this research project is that fine particulate mass is significantly underdetermined in urban environments using current standard reference methods because of the loss of volatile material from the particles during sampling. This project involves the development of new instrumentation for the continuous monitoring of PM2.5, including the determination of fine particulate semi-volatile species which can be lost from fine particles during sampling with a filter.The specific objectives of the program were to: (1) develop a "Real-time Ambient Mass Sampler" (RAMS) for the real-time determination of the mass of both particles and semi-volatile species lost from the particles during sampling; (2) evaluate the collection after the impactor-concentrator-denuder-dryer inlet of fine particles and volatile particulate species lost from particles by a variety of "sandwich" filters in both controlled chamber and ambient studies for the sampling system developed in 1; (3) determine the effect of changing relative humidity on the collection of water by the "sandwich" filters and on the retention of water by collected particles in both controlled chamber and ambient studies for the sampling system(s) developed in 1; (4) validate the determination of fine particle associated water using the analytical techniques developed in 3 with results obtained from humidity dependent differential mobility analyzer and particle size distribution data in ambient studies; (5) based on the results obtained in 3 and 4, determine if paradigms can be developed for the estimation of water associated with fine particles using the new continuous mass monitor; (6) if particle-bound water can be measured by the RAMS, evaluate the determination of water soluble gases (e.g., HNO3, SO2) associated with hygroscopic fine particles using the developed sampling system(s) in both controlled chamber and ambient studies; and (7) characterize the precision and accuracy of the new continuous monitor for the determination of PM2.5 and associated volatile species at several ambient locations which represent a matrix of temperature, relative humidity, fine particle acidity, and fine particle chemical composition.
Summary/Accomplishments (Outputs/Outcomes):
The research program has developed and validated new instrumentation for the continuous monitoring of PM2.5, including the determination of particulate semi-volatile species which can be lost from fine particles during sampling with a filter. As presently developed, the RAMS (Figure 1) has both a monitor and an active blank sampler. The monitor has eight parts: (1) a combined impactor-particle concentrator inlet to separate the fine particles from coarse particles and from 70 percent of the sampled gas stream; (2) two triethanolamine coated annular denuders to remove NO2; (3) a Nafion dryer to remove gas phase water from the sampled air stream; (4) a charcoal impregnated multi-parallel filter diffusion denuder (BOSS denuder) to remove gas phase semi-volatile organic material, nitric acid, ammonia, and oxidants (e.g., ozone, hydrogen peroxide) which can interfere with the measurement of fine particulate mass. This denuder is heated (about 30 C) to minimize the absorption of water by the charcoal; (5) one triethanolamine coated annular denuder to remove any NO2 formed from NO on the charcoal of the BOSS denuder; (6) a final Nafion dryer to remove gas phase water from the sampled air stream; (7) a "sandwich" filter for the collection of fine particles and semi-volatile species which can be lost from the particles during sampling; and (8) a TEOM? monitor to measure the rate of collection of fine particulate material by the "sandwich" filter. The first three components of the sampler are identical for both the active blank sampler and the monitor. After the first Nafion dryer, the flow stream is split, half entering the remainder of the monitor system. The second half passes through a 47 mm quartz fiber filter to remove particles and then through components identical to components 4 through 8 of the monitor to provide a particle free sample steam for the active blank monitor, Figure 1.The instrument developed and validated in the EPA Science to Achieve Results (STAR) program will assist in the improved assessment of environmental problems associated with exposure to fine particulate matter (PM2.5) as the new standard for fine particulate matter is promulgated. The research has developed, validated, and tested new sampling procedures for both the routine monitoring of total PM2.5, and the determination of the fraction of the total PM2.5, which is associated with various volatile species. Specific volatile fine particulate species which the new instrument will monitor are: (1) semi-volatile nitrate compounds (ammonium nitrate); and (2) semi-volatile organic compounds. The RAMS will not monitor for particle bound water and associated dissolved gases.
The design of the instrument in its final stage of development in the BYU EPA STAR program is shown schematically in Figure 1. The RAMS contains both a monitor and an active blank sampler. The first component of each of the two systems is a cyclone or impactor inlet with a particle cut at 2.5 m, the cut point for the new fine particle standard. The next three components of both systems is a particle concentrator to remove 70 percent of the gas phase species in equilibria with the fine particles with minimal change in the concentrations of the gas phase species with which the particles are in equilibrium in the atmosphere, two triethanolamine coated annular denuders to remove NO2, and a Nafion dryer to remove water. The flow is then split between the monitor and blank samplers. The next component of the blank sampler is a quartz filter to remove particles. Thus, the blank sampler monitors for any gas phase compounds not removed by the denuders and dryers, and measured by the TEOM monitor. The RAMS data are corrected for this active blank. The next component in both systems is a BOSS diffusion denuder to remove the gas phase species in equilibrium with the semi-volatile components of the particles. Species which are quantitatively removed by the diffusion denuder include nitric acid, sulfur dioxide and semi-volatile organic compounds. The denuder also removes most of the ammonia and oxidants (e.g., O3, NO2 and H2O2) which may react with collected fine particulate compounds. Since the charcoal surface of this denuder will absorb water at high relative humidity, the denuders are heated to 30?C. The annular denuder after the BOSS denuder of each system removes any additional NO2 which may be formed in the BOSS denuder. The next component of both systems is a diffusion dryer to minimize the humidity of the sampled air. The last component of each system in the instrument is a "sandwich" filter to collect particles and the nitrate and semi-volatile organic material which may be lost from the particles during sampling. Mass collected on the "sandwich" filter is determined on a real-time basis using a R&P Tapered Element Oscillating Microbalance, TEOM?, with the "sandwich" filter replacing the conventional TEOM filter. The "sandwich" filter is heated to 30-40?C to release particle bound water.
Both controlled chamber experiments and ambient studies have been conducted to characterize the collection and measurement by the RAMS of each of the target semi-volatile components of fine particulate matter; ammonium nitrate, semi-volatile organic material, particulate bound water and water associated small molecular weight gases. The results of the continuous monitoring of fine particulate material, including the determination of the various volatile particulate components has been compared to results obtained using conventional mass monitoring techniques and using integrated samples collected with diffusion denuder samplers. This has included the use of the PC-BOSS developed as part of the program. In these experiments, the RAMS was evaluated in environments which represent a matrix of temperature, relative humidity, fine particle acidity, and fine particle chemical composition. These combined experiments have tested the precision and accuracy of the RAMS for the continuous determination of PM2.5, including associated volatile species. The research has provided an instrument which can be used both for the routine monitoring of total fine particulate mass and for the specific determination of various classes of fine particulate volatile species in special studies of ambient fine particulate chemistry and associated epidemiology.
The research related to each of the project objectives was completed as follows:
Year 1. Research was conducted related to objectives 1 and 2, development and validation of the RAMS sampler. Research was initiated related to objectives 3-5, the development of paradigms for the determination of aerosol water.
Year 2. Research was continued relative to objectives 3, 4, and 5. In addition, two field studies were conducted relative to objective 7 in Riverside and Bakersfield, CA. Data from these field studies indicated that, under some field conditions, the blank for the RAMS data was substantial. Therefore the instrument was redesigned to give the configuration shown in Figure 1 and experiments were conducted to minimize the observed blank. Chamber experiments were also conducted to validate the RAMS for the collection of well defined particles at low humidity. These chamber studies included the study of ammonium sulfate, ammonium nitrate, glycerol, organic acids and combinations of these various compounds. The results of the various chamber and field studies indicated that the RAMS quantitatively determines the mass of any nitric acid or semi-volatile organic material which would be lost from a single filter sampler. Some of the ammonia lost from particulate ammonium sulfate (but not the nitrate) is not collected by the RAMS if the "sandwich" filter temperature is above 30?C.
Year 3. Field studies were conducted in both Provo, UT and Philadelphia, PA to validate the current RAMS configuration, Figure 1, for the measurement of fine particulate material, including ammonium nitrate and semi-volatile organic material (objective 7). In addition, the RAMS was used in the August 1999 Atlanta EPA Supersite program using both the EPA STAR and BYU EPA EMPACT funds. Research was completed relative to objectives 3-5 studying generated ammonium sulfate aerosols at varying relative humidity. These chamber experiments conclusively showed that the RAMS will not monitor for particle bound water. As a result, the research related to objective 6, the measurement of water soluble gases associated with hygroscopic fine particles, was not done.
Specific accomplishments for the research effort completed during the 3 years of the EPA STAR program are:
- The efficiency of the combined particle-concentrator/denuder portions of the RAMS for the removal of gas phase semi-volatile organic material, nitric acid and ammonia has been evaluated in both chamber and ambient sampling experiments. Ambient studies have been conducted on the removal of water and oxidants by the RAMS. These results have shown that the current design used in the RAMS minimizes these gas phase constituents prior to monitoring the mass of the collected fine particles. The early results of ambient and chamber studies of the RAMS and the papers describing the development of the PC-BOSS have been published or are submitted (Ding 2000a, 2000b, 1997, Eatough 1999, Obedi 2000a, 1999, 1997, Pang 1999, 1998a, 1998b) and papers describing the final design of the RAMS (Obeidi 2000b, 1999, Patterson 1988) the more recent field studies (Eatough 2000) and chamber work describing the collection of model aerosols an minimization of the blank (Obeidi 2000b, Patterson 2000) have been submitted or are being prepared for publication.
- The efficiency of the Nafion dryer for the removal of gas phase water prior to the determination of fine particulate mass has been evaluated in both chamber and ambient experiments (Obeidi 2000a, 2000b, 1999, Eatough 2000). The current instrument design has been shown to minimize the effect of ambient relative humidity on the determination of fine particulate mass. Manuscripts describing the ambient results in the two completed field studies using the final instrument design have been submitted (Obeidi 2000b) or are being prepared (Eatough 2000). A more detailed manuscript on the results obtained on the minimization of the blank associated with water and oxidants is being prepared (Patterson 2000).
- The retention of water by the particles collected on the "sandwich" filter of the RAMS has been evaluated in chamber studies using ammonium sulfate aerosol at both high (95%) and low (30%) relative humidity. This experiments showed that greater than 95% of the ammonium sulfate associated water is lost from the particles collected on the "sandwich" filter of the RAMS. The results obtained in these experiments have been published in an M.S. Thesis (Patterson 1998) and a Ph.D. dissertation (Obeidi 1999).
- The efficiency and lower cut point of the particle concentrator has been evaluated in a series of ambient and chamber experiments. Results obtained to date indicate that the lower cut point of the concentrator can be made as low as 0.05 m. Thus the RAMS should determine essentially all fine particulate material (Obeidi 1999, 1997, Eatough 1999).
- Field experiments to validate the results obtained with the initial RAMS configuration by comparison with the results obtained with a conventional filter pack and with diffusion denuder samplers for organic material and ammonium nitrate have been completed in a joint program between BYU and Harvard University at both Riverside (August - September 1997) and Bakersfield (February - March 1998), CA. Results of these studies have been published (Pang 1999) or submitted (Ding 2000a, Obeidi 2000a) for publication. The blank corrected RAMS data were in agreement with both PC-BOSS and conventional diffusion denuder results. However, single filter sampling procedures (e.g., the PM2.5 FRM) lost an average of 40 percent and 30 percent of the fine particle mass in the Riverside and Bakersfield studies, respectively.
- Field studies to validate the configuration of the RAMS shown in Figure 1 have been completed in Provo, UT (November-December 1998) (Obeidi 2000b) and Philadelphia, PA (June-July 1999) (Eatough 2000). The results of these studies have shown that continuous, accurate data can be obtained with the modified RAMS. The results of these studies will be presented at the PM2000 specialty conference in Charleston SC in January 2000, and an overview (Eatough 2000) and more detailed papers describing these field studies have been submitted (Obeidi 2000) or are being prepared for publication (Pang 2000).
- The experiments completed in 6. have shown that the precision of the RAMS data is limited by the absorption of water by the cellulose portion of the charcoal impregnated filter used in the RAMS "sandwich" filter. In cooperation with Schleicher and Schuell (Germany), a new charcoal impregnated filter based on the use of glass fibers has been developed. The absorption of water by this filter is much reduced over that for the cellulose fiber product and should allow the reduction of the precision of the RAMS data to about ? 2 ug/m2 for 30 minute continuous data. Data related to this development are included in a paper now being prepared for publication (Eatough 2000).
Outline of Results for the Ambient Sampling Programs to Validate the RAMS
AMS Blank. The various dryers and denuders of the RAMS are designed to remove gas phase organic compounds, nitric acid, sulfur dioxide, nitrogen dioxide, ammonia, ozone and water, each of which can be absorbed by the charcoal portion of the "sandwich filter." The concentrations of these species are reduced to acceptable, but not zero concentrations prior to the collection of particles on the TEOM probe. Therefore, a second, parallel system, preceded by a filter, is used to obtain an active blank for the correction of the RAMS monitor data. The true and measured blank differ slightly because the presence of the filter in the blank alters the residence time in, and hence the efficiency of the various denuders and dryers. The precision of the RAMS data is limited by the variance in this blank correction. The variability in the blank correction is dominated by the effects of water and major gas phase interferents (e.g. NO2 and O3) (Obeidi 2000b, 1999). Hence, the blank correction differs for each sampling site, Figure 1. As illustrated in Figure 2, the variability and precision of the blank corrections for samples collected in the summer is small for a small urban area with a dry climate (Provo, UT in the winter) and much larger for major urban areas with a humid climate (Philadelphia and Atlanta in the summer). The blank corrected data are illustrated in Figure 4 for Philadelphia and Atlanta. The data given in Figure 4 are presented as 2.5 h running averages to minimize the effect of the blank correction uncertainty.
Results for Bakersfield and Riverside. The results obtained with the first version of the RAMS in Bakersfield and Riverside, California (Eatough 1999, Obeidi 2000a) did not yield continuous data because of the high uncertainty in the blank during morning hours when water collected by the BOSS denuder of the RAMS during the night, at high relative humidity, was released as the relative humidity decreased. However, for those time periods when valid RAMS data were obtained, the PM2.5 mass measured with the RAMS and constructed mass obtained with the PC-BOSS were in good agreement, Figure 3. The PM2.5 mass measured at these two sites included an average of 40 percent and 30 percent of the fine particle mass present as semi-volatile ammonium nitrate and organic material at Riverside and Bakersfield, respectively. This semi-volatile material was not measured by single filter samplers such as the PM2.5 FRM.
Results for Provo and Philadelphia. The results obtained for the continuous determination of PM2.5 with the latest version of the RAMS, Figure 1 have been validated by comparison with results obtained from diffusion denuder integrated samples to determine the mass of fine particulate material retained on a filter and the semi-volatile organic material and ammonium nitrate lost from the filter during sampling with a PC-BOSS. Results obtained with RAMS and PC-BOSS denuder sampler show that PM2.5 mass, including semi-volatile fine particulate nitrate and organic species can be continuously and accurately monitored with the RAMS. The linear regression of RAMS vs PC-BOSS results give r2 = 0.74 (n=13) with a slope of 0.98 ?0.03 for the Provo UT data2, a slope of 1.02?06 (n=7) for the Philadelphia data, and comparable results for Atlanta. The precision, but not the accuracy, of RAMS data are affected by the variability in the blank data, Figures 2 and 4.
The variability of the blank data is dominated by the effect of the reversible adsorption of water by the cellulose fibers of the charcoal impregnated filter in the "sandwich" filter. In cooperation with Schleicher and Schuell, we have evaluated the use of a charcoal impregnated glass fiber filter in the "sandwich filter." The blank variability in the RAMS is markedly reduced with this "sandwich" filter, Figure 5, and half-hour averaged RAMS data with good precision can be obtained with this modification, Figure 6.
The RAMS with an impregnated charcoal glass fibre filter is now being used for the routine, continuous determination of PM2.5 mass, including semi-volatile species in the EPA Salt Lake City EMPACT program.
Journal Articles on this Report : 13 Displayed | Download in RIS Format
Other project views: | All 19 publications | 13 publications in selected types | All 13 journal articles |
---|
Type | Citation | ||
---|---|---|---|
|
Ding Y, Pang Y, Eatough DJ. High-volume diffusion denuder sampler for the routine monitoring of fine particulate matter: I. Design and optimization of the PC-BOSS. Aerosol Science and Technology 2002;36(4):369-382. |
R825367 (Final) |
Exit |
|
Ding Y, Pang Y, Eatough DJ, Eatough NL, Tanner RL. High-volume diffusion denuder sampler for the routine monitoring of fine particulate matter: II. Field evaluation of the PC BOSS. Aerosol Science and Technology 2002;36(4):383-396. |
R825367 (Final) |
Exit |
|
Eatough DJ, Obeidi F, Pang YB, Ding Y, Eatough NL, Wilson WE. Integrated and real-time diffusion denuder sampler for PM2.5. Atmospheric Environment 1999;33(17):2835-2844. |
R825367 (Final) R826373 (1998) R826373 (2000) R826373 (2002) |
Exit Exit |
|
Eatough DJ, Pang Y, Eatough NL. Determination of PM2.5 sulfate and nitrate with a PC-BOSS designed for routine sampling for semi-volatile particulate matter. Journal of the Air & Waste Management Association 1999;49:69-75. |
R825367 (Final) |
Exit |
|
Lewtas J, Pang P, Booth D, Reimer S, Eatough DJ, Gundel LA. Comparison of sampling methods for semi-volatile organic carbon associated with PM2.5. Aerosol Science and Technology 2001;34(1):9-22. |
R825367 (Final) R827993 (2001) |
Exit Exit |
|
Obeidi F, Eatough DJ. Continuous measurement of semivolatile fine particulate mass in Provo, Utah. Aerosol Science and Technology 2002;36(2):191-203. |
R825367 (Final) R827993 (2002) R827993 (2003) R827993 (Final) |
Exit Exit |
|
Obeidi F, Eatough NL, Eatough DJ. Use of the RAMS to measure semivolatile fine particulate matter at Riverside and Bakersfield, California. Aerosol Science and Technology 2002;36(2):204-216. |
R825367 (Final) |
Exit |
|
Pang Y, Ren Y, Obeidi F, Hastings R, Eatough DJ, Wilson WE. Semi-volatile species in PM2.5:Comparison of integrated and continuous samplers for PM2.5 research or monitoring. Journal of the Air & Waste Management Association 2001;51(1):25-36. |
R825367 (Final) R827993 (2002) R827993 (2003) R827993 (Final) |
Exit Exit |
|
Pang Y, Eatough NL, Wilson J, Eatough DJ. Effect of semivolatile material on PM2.5 measurement by the PM2.5 federal reference method sampler at Bakersfield, California. Aerosol Science and Technology 2002;36(3):289-299. |
R825367 (Final) |
Exit |
|
Pang Y, Eatough NL, Eatough DJ. Evaluation of the performance of annular denuder samplers. Aerosol Science and Technology 2002;36(6):790-798. |
R825367 (Final) |
Exit |
|
Pang Y, Eatough NL, Eatough DJ. PM2.5 semivolatile organic material at Riverside, California: Implications for the PM2.5 federal reference method sampler. Aerosol Science and Technology 2002;36(3):277-288. |
R825367 (Final) |
Exit |
|
Patterson E, Eatough DJ. Indoor/outdoor relationships for ambient PM2.5 and associated pollutants: epidemiological implications in Lindon, Utah. Journal of the Air & Waste Management Association 2000;50(1):103-110. |
R825367 (Final) |
|
|
Eatough DJ, Eatough NL, Obeidi F, Pang Y, Modey W, Long R. Continuous determination of PM2.5 mass, including semi-volatile species. Aerosol Science and Technology 2001;34(1):1-8. |
R825367 (Final) R827993 (2000) R827993 (2001) R827993 (Final) |
Exit Exit |
Supplemental Keywords:
continuous measurement, qater, volatile, ambient, monitoring, articulate, sampling., RFA, Scientific Discipline, Air, Ecosystem Protection/Environmental Exposure & Risk, particulate matter, Physics, Environmental Chemistry, Chemistry, Monitoring/Modeling, environmental monitoring, particle size, ambient particle properties, environmental measurement, fine particles, particulate, continuous measurement, PM 2.5, semi-volatile organic material, PM2.5, sampling, semi-volatile particulate species, urban environment, aerosolsProgress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.