Loading effect correction for real-time aethalometer measurements of fresh diesel soot.

ABSTRACT

In this study, a correction was developed for the aethalometer to measure real-time black carbon (BC) concentrations in an environment dominated by fresh diesel soot. The relationship between the actual mass-specific absorption coefficient for BC and the BC-dependent attenuation coefficients was determined from experiments conducted in a diesel exposure chamber that provided constant concentrations of fine particulate matter (PM; [PM.sub.2.5]; PM <2.5 [micro]m in aerodynamic diameter) from diesel exhaust. The aethalometer reported BC concentrations decreasing with time from 48.1 to 31.5 [micro]g [m.sup.-3] when exposed to constant [PM.sub.2.5] concentrations of 55 [+ or -] 1 [micro]g [m.sup.-3] and [b.sub.scat] = 95 [+ or -] 3 [Mm.sup.-1] from diesel exhaust. This apparent decrease in reported light-absorbing PM concentration was used to derive a correction K(ATN) for loading of strong light-absorbing particles onto or into the aethalometer filter tape, which was a function of attenuation of light at 880 nm by the embedded particles.

INTRODUCTION

Carbonaceous aerosols are composed mainly of organic carbon (OC) and elemental carbon (EC), the latter a measure of black carbon (BC) aerosol. Carbonaceous aerosols may have a detrimental impact on human health from both acute and chronic exposures because of their toxicological effect and because they are able to penetrate into the lower respiratory system. (1,2) Susceptible populations include those suffering from chronic obstructive pulmonary disease, cardiovascular patients, and children with asthma. (3-6) In addition, BC aerosol influences climate directly and indirectly through light extinction in the atmosphere (7-9) and can lead to low visibility. (10,11)

It is known that the absorption efficiency of BC aerosol varies depending on the source and chemical composition, (12-16) and the assumption that all light-absorbing material is because of the presence of BC aerosol is not always accurate. (17-21) Light-absorption efficiencies of aerosols embedded in a reflective filter matrix are known to be enhanced compared with the same aerosols in the atmosphere. (22-25) For an early version of the aethalometer operating with an incandescent lamp, a value of [[sigma].sub.ATN] = 19 [m.sup.2] [g.sup.-1] was used to translate changes in light attenuation through a quartz filter into BC mass concentration. (26) This value was compared with the absorption efficiency of similar aerosols in the air ([[sigma].sub.abs] ~10 [m.sup.2] [g.sup.-1]; Horvath (8)) to account for absorption enhancements because of multiple scattering within the filter matrix. (23,24)

In addition, an optical effect because of the accumulation of particles in the filter has been reported for the aethalometer (25,27-29) and similar applications of the light transmission method. (13,30) As the filter becomes loaded with particles, the enhancement of the light absorbed per unit mass of added BC decreases, which results in lower reported BC concentrations for loaded filters compared with lightly loaded filters. This optical effect is referred to as a "shadowing effect." Weingartner et al. (28) reported that this effect is more pronounced for freshly emitted soot than for aged atmospheric aerosol. The current aethalometer algorithm used to translate filter light attenuation into BC mass concentration does not correct for this loading effect. In this study, we examine this optical effect using a constant source of diesel soot, and a correction for the current aethalometer algorithm was developed to measure real-time concentrations of BC from fresh diesel exhaust.

EXPERIMENTAL WORK

Two portable versions of the dual-wavelength aethalometers (AE41 and AE42, Magee Scientific Company) were simultaneously exposed to controlled concentrations of ultrafine diesel-generated particles (soot) in a specially designed diesel chamber. We used two exposures at variable concentrations of fine particulate matter (PM; [PM.sub.2.5]) and one at a constant (55 [+ or -] 1 [micro]g [m.sup.-3]) concentration of [PM.sub.2.5] from diesel soot, which was used to derive the correction to the aethalometer algorithm. The conditions of approximately 50 [micro]g [m.sup.-3] of [PM.sub.2.5] from diesel exhaust generated for the experiments were comparable to similar pollution observed near a high traffic road in Paris, France, (31) and lower than the BC concentrations observed when chasing a transit bus in Los Angeles, CA. (32)

Diesel Chamber

The diesel chamber was located in Seattle and was operated by the Department of Environmental and Occupational Health Sciences at the University of Washington. The chamber dimensions were 8.8 x 5.5 x 2.4 m with a total volume of 116 [m.sup.3]. The volumetric flow rate through the chamber was 28.3 [m.sup.3] [min.sup.-1], and the incoming air was filtered so that the PM concentration inside the chamber was unaffected by background PM. Diesel soot was generated from a turbocharged direct-injection 5.9-L Cummins B-series diesel engine (6BT5.9G6, Cummins, Inc), which was comparable to engines used in delivery trucks and school buses. The engine drove a 100-kW generator connected to an electric load bank (Simplex), and the load applied to the running engine was set to 75 kW. The engine fuel was highway grade diesel No. 2 undyed, which is commonly used in delivery vehicles.

The [PM.sub.2.5] concentration inside the chamber was established by a two-stage dilution process that mixed air with diesel exhaust. The degree of dilution and resulting concentration was adjusted with a variable speed fan that was electronically controlled by a system that used two light-scattering nephelometers (one sensing upstream of the chamber and the other inside the chamber). This provided feedback to the system to adjust the amount of diverted diesel exhaust to achieve and maintain a target [PM.sub.2.5] concentration. Under these conditions of controlled constant [PM.sub.2.5] concentration from diesel exhaust, the chamber exhibited a linear relationship between EC and [PM.sub.2.5] mass concentration (intercept = -6.1; slope = 0.85; [R.sup.2] = 0.97).

Measurements

To verify constant conditions and to compare chamber-generated diesel soot to other experiments reported in the literature, several supporting PM parameters were measured in the chamber, including continuous [PM.sub.2.5] concentration from a tapered element oscillating microbalance (TEOM) monitor with a [PM.sub.2.5] inlet (1-min; Series 1400a, Thermo Electron Co.) and light scattering coefficient, [b.sub.scat], from a nephelometer (1 min; M903, Radiance Research). The nephelometer was calibrated using zero air and carbon dioxide for span setting. In addition, integrated [PM.sub.2.5] samples were taken from collocated single-stage 5 L [min.sup.-1] low-volume samplers (Airmetrics Inc.). Two-stage filter cassettes (47-mm) with Teflon filter (2-[micro]m pores; part 7592-104, Whatman Inc.) and quartz filter after Teflon, as well as single-stage filter cassettes with quartz (part 1851047, Whatman Inc.), were used with the samplers to estimate OC and EC fractions of the [PM.sub.2.5]. In addition, two EcoChem samplers (EcoChem Analytics) were deployed to measure particle active surface area (PASA) and total particle-bound polycyclic aromatic hydrocarbons (PPAHs). These instruments have been used to characterize sources and type of aerosols based on the relationship between PPAHs and particle active surface area. (33,34)

Sample and Data Analysis

[PM.sub.2.5] collected on the Teflon filters was analyzed gravimetrically using a microbalance (model UMT2, Mettler-Toledo, Inc.) at a constant temperature (22.2 [+ or -] 1.8 [degrees]C) and relative humidity (34.8 [+ or -] 2.5%), after the samples were equilibrated for [greater than or equal to]24 hr before weighing. Sections of the quartz filters (1.5 [cm.sup.2]) were analyzed for OC and EC via thermal optical evolved gas analysis (thermal optical transmittance [TOT], Sunset Laboratory Inc.) using a modified version of the National Institute for Occupational Safety and Health 5040 method. (33) The TOT carbon analyzer was calibrated using a standard solution of sucrose (20 [micro]L of 4.5 g C [L.sup.-1] solution = 90 [micro]g C). More details about the temperature steps, carrier gases, and standards used in this thermal optical analysis are reported in Pang et al. (35)

BC concentrations from the aethalometer were compared with the EC concentrations measured by thermal optical analysis of the quartz filters. In addition, aethalometer measurements were compared with other PM measurements, including [PM.sub.2.5] concentration, [b.sub.scat], PASA, and PPAHs. This was done to evaluate the temporal behavior of the aethalometer measurements during the chamber experiment at constant PM concentrations. Finally, a correction was proposed for the aethalometer to account for loading effect of the filter-particle matrix.

Model Framework

This study developed a correction for the aethalometer algorithm that could account for the loading effect. This work is specific for air with fresh diesel exhaust dominated aerosols.

The optical attenuation (ATN) of light by particles deposited in the quartz filter is given by the following relationship:

ATN = -100 X ln(I/[I.sub.0]) = -100 x ln(T) (1)