Grantee Research Project Results

Final Report: Real-Time Analysis of PAH Bound to Size-Resolved Atmospheric Particles by Tandem Time of Flight Mass Spectrometers

EPA Grant Number: R825391
Title: Real-Time Analysis of PAH Bound to Size-Resolved Atmospheric Particles by Tandem Time of Flight Mass Spectrometers
Investigators: Smith, Kenneth A. , Worsnop, Douglas R.
Institution: Massachusetts Institute of Technology
EPA Project Officer: Hahn, Intaek
Project Period: October 1, 1996 through September 30, 1999 (Extended to November 30, 2000)
Project Amount: $375,000
RFA: Exploratory Research - Air Engineering (1996) RFA Text |  Recipients Lists
Research Category: Air Quality and Air Toxics , Air , Safer Chemicals , Land and Waste Management

Objective:

The goal of this research was to develop and demonstrate an instrument capable of quantifying polycyclic aromatic hydrocarbons (PAHs) associated with individual size-segregated atmospheric particles in real time. Our approach to this task involves combining an aerosol mass spectrometer (AMS) developed at Aerodyne Research, Inc. (ARI) with a sensitive PAH detection scheme using photoionization with time-of-flight mass spectrometers. This program is a collaborative effort between the research groups of Professor Kenneth A. Smith at Massachusetts Institute of Technology (MIT), Department of Chemical Engineering, and Aerodyne Research, Inc. The experimental results reported here have been largely performed at Aerodyne Research by Dr. Xuefeng Zhang, an MIT postdoctoral research associate supported by this program.

Combustion exhaust particles and associated adsorbed organic compounds released into the atmosphere are of concern because of their possible health effects. Of particular concern are the PAHs, which are known to be mutagenic. PAH material formed by combustion processes is listed as a hazardous air pollutant in the Clean Air Act. In the atmosphere, PAH material partitions between the gas and particulate phases. Measurements of the amount of PAH associated with different aerosol size fractions are critical for a complete understanding of the environmental fate of and human exposure to fine particles containing PAH. Current methods to quantify PAH bound to size-segregated atmospheric particles require long sampling times (~100 hours) and laborious (~1000 man-hours) and costly post-collection chemical analysis.

This research program was designed to develop and demonstrate an instrument that is capable of quantifying PAH associated with individual size-segregated atmospheric particles in real time. Our approach to developing a compact and efficient field-portable aerosol mass spectrometer was built on our earlier efforts at performing single particle mass spectrometric detection using resistively heated surfaces as the particle vaporization source (Jayne, et al., 2000; Allen and Gould, 1981; and Sinha, et al., 1982). The AMS described here combines the high performance aerosol sampling inlet of the Liu, et al. (1995a,b) design with particle detection via efficient thermal vaporization. In this approach, particles are focused into a narrow beam while being sampled into vacuum and directed onto a resistively heated surface. The PAH components in/on the particle flash vaporize upon contact in the heater, and the gas phase molecular constituents are then photo-ionized by UV laser resonance-enhanced multi-photon ionization (REMPI). The ionized molecular PAH species are classified using time-of-flight (TOF) molecular mass spectrometric analysis techniques. Particle aerodynamic size is determined by measuring particle time-of-flight using a mechanical particle beam chopper. This is performed by firing the laser at specific time intervals synchronized to the phase of the particle beam chopper with subsequent detection of ion signals in the TOF spectrometer. This particle sizing scheme takes advantage of the size-dependent distribution of particle velocities generated by the expansion of the aerosol into the vacuum (Jayne, et al., 2000). The resulting instrument is capable of measuring ambient aerosol sizes from several tens of nanometers to several microns in diameter while simultaneously providing size-resolved composition information on single particles. In addition, the REMPI process provides a very selective and, hence, sensitive means of PAH ionization and classification when coupled with molecular mass spectrometric detection. This is a result of the relatively low ionization potentials typical of PAH material.

Summary/Accomplishments (Outputs/Outcomes):

During this project, we have developed and quantified the aerodynamic particle sizing technique (Jayne, et al., 2000). Size-dependent particle velocity, particle beam width, and particle transmission and collection efficiency have been measured as a function of particle sizes from 50 to ~1000 nm diameter. The measured performance of the aerodynamic sizing technique has been compared to its theoretical performance with very good agreement.

We also have coupled a TOF mass spectrometer (RM Jordan, Grass Valley, CA), a KrF pulse excimer laser (MPB Technologies, Montreal, Canada) operating at 248 nm, and a vacuum compatible cartridge heater (Heatwave, Watsonville, CA), which serves as the particle evaporator. We also have demonstrated that the proposed photoionization and TOF analysis scheme provides high sensitivity detection of PAH components bound to soot particles generated from laboratory flames and gasoline engines.

An instrument schematic is shown in Figure 1. It consists of three main sections: (1) an aerosol sampling inlet, (2) a particle sizing chamber, and (3) a particle composition detection chamber. Each chamber is separated by critical apertures and is differentially pumped.

The performance of the aerodynamic lens has been modeled using a commercially available fluid dynamics program (FLUENT, 1995). Calculations describing the particle focusing capability, the particle transmission efficiency, and the size-dependent particle velocities generated by the aerodynamic inlet have been performed and support the results of the measured lens performance. It was found that, by optimizing lens geometry, particles in the size range of 40-5000 nm can be efficiently collimated. The modeling results also show that, by increasing lens upstream pressure, one can preferentially transmit larger particles, or vice versa. Depending on the particular sampling application, the aerodynamic lens may be tuned to provide optimum sampling efficiency. Three papers which model the performance of the aerodynamic lens and the pinhole have been submitted to or currently being prepared for peer reviewed journals.

The results of the aerodynamic particle velocity measurements are plotted in Figure 2 as a function of aerodynamic diameter. As can be seen, the measurements indicate a very good correlation between particle velocity and particle aerodynamic size. For spherical particles (DOP and PSL), aerodynamic diameter is the product of the geometric diameter and the particle density. For non-spherical crystalline NH4NO3 particles, however, the effect of particle shape needs to be considered in defining an aerodynamic diameter. To obtain the best fit of the NH4NO3 particle velocity data shown Figure 2 a shape factor of 0.80 was applied. The dashed line in Figure 2 is the result from the FLUENT model calculation for spheres with a density of 1 g/cc. The agreement between the model and the data is very good. These experiments used a quadrupole mass spectrometer with electron impact ionization which is more convenient for particles of these chemical compositions.

Experiments also were performed to determine the AMS particle collection efficiency as a function of particle size. For these experiments, a differential mobility analyzer (model 3071, TSI, St. Paul, MN), and a condensation particle counter (CPC3022A, TSI) were used to produce a calibrated monodisperse aerosol source. The monodisperse aerosol flow from the output of the DMA was sampled by both the AMS and the CPC, and the particle count rates from both the AMS and the CPC were compared for a range of particle sizes. The AMS particle collection efficiency measurements are summarized in Figure 3. The two symbols in the figure (circles and squares) represent the two modes of signal processing offered by the AMS: (1) direct particle counting (circles) in which each incident particle gives rise to a discrete single particle signal and (2) total particle mass from integrating ion signal intensities for many single particle events (squares). The solid line in Figure 3 is the prediction based on the FLUENT calculations. The agreement between the model line and the measurements is good. The decrease in collection efficiency for the larger diameter particles is due to impaction losses at the entrance to the aerodynamic lens. The more dramatic decrease in collection efficiency for the smaller diameter particles (<60 nm) results from less efficient focusing in the aerodynamic lens. Although these particles are transmitted through the lens, their angular dispersion upon exiting the final lens is sufficiently large that they do not pass the skimmer and pinhole which constitute the entrance the detection chamber.

The ionization/detection scheme that has been employed for this program uses an excimer laser to photoionize the gas phase PAH constituents and a TOF mass spectrometer for molecular detection. This approach can provide a complete mass spectrum for a single particle vaporization event.

Resonance-enhanced multiphoton ionization (REMPI) provides the most sensitive detection technique available for many polyaromatic molecules, particularly for species such as PAH with strong UV absorption features. Since its original demonstration (Bernstein, 1982; Squire, et al., 1983), REMPI is now routinely used for extremely sensitive detection of PAH (e.g., Tanda, et al., 1993; Hepp, et al., 1995; Siegmann, et al., 1995; and Hankin, et al., 1997. Ionization efficiencies can reach 100 percent within the ionization volume (Boesl, et al., 1981, Lubman and Kronick, 1982). The ionized gas molecules, are sampled using the TOF molecular mass spectrometer. The TOF spectra are recorded by a 300 MHz Tektronic digital oscilloscope (model TDS3032). We have chosen UV light at 248 nm because it is suitable for a very wide range of combustion-generated organic compounds like PAH; in fact, Siegmann and Sattler (2000) have observed PAHs in the range of 78 to 788 AMU (generated from a methane flame) by operating a laser at 248 nm.

Operationally, the ionization laser is triggered at specific time intervals which are synchronized to the phase of the particle beam chopper. This is equivalent to selecting a specific particle time-of-flight or particle aerodynamic diameter for mass spectrometric analysis.

The performance of the instrument has been tested using atomized pure pyrene particles, freshly generated soot particles from a propane flame and soot particles exhausted from a Honda EG1400 gasoline engine. The flame or engine particle sources were passed through a diffusion dryer then into an 2.2 cm x 40 cm glass flow tube where particle concentration could be diluted by adding filtered air flow. A small fraction of particles is sampled near the exit of the flow tube by both the AMS (1.6 scc/sec) and a condensation particle counter (25 scc/sec, TSI model CPC3022A). The majority of the flow tube effluent was vented. Results are discussed below.
A typical spectrum of soot particles from a propane diffusion flame is shown in Figure 4. Figure 4A is an average ion current signal over 512 laser firing events versus time, where the solid line is for soot particles and dashed line is for atomized pure pyrene particles for molecular weight calibration. In the plot, the peak at t=0 is electronic "pickup" indicating the time of the laser firing. One important feature in the pyrene spectrum is that it shows only one ion peak at 17.5 s time of flight, which corresponds to the pyrene parent mass peak (202 AMU). The pyrene spectrum indicates that, with the current laser beam intensity (107 W/cm2), the REMPI process is "soft," i.e., production of ion fragments is minimal. The dominant ion peak in the upper plot indicates that pyrene is a major semi-volatile component of soot particles from propane flames.

The spectrum in Figure 4A is plotted in Figure 4B versus molecular weight. The plot shows that the lowest PAH peak is at 140 AMU corresponding to an 11 carbon molecule (C11H8). The plot also shows all PAH peaks from 11 carbon molecules (C11H8) to 42 carbon molecules (C42H16). It also is observed that the PAHs containing odd numbers of carbon atoms produce smaller ion intensity peaks. This is consistent with the observation of gas phase PAHs from methane diffusion flame by Siegmann and Sattler (2000), in which it was explained that the PAHs containing odd numbers of C atoms cannot be completely conjugated or benzenoided, and therefore, they are less stable than PAHs of comparable structure containing even numbers of C atoms.

Figure 5A is a series of PAH spectra (averaged over 512 laser firing events) from particles of different particle times-of-flight or diameters. The particle number density is about 5x106/cc. The figure shows a variation in ion intensity as a function of laser firing time relative to the chopper position. When laser is fired at zero particle TOF, a small ion signal is measured due to the background in the chamber. The intensity of the spectrum increases with particle TOF and reaches a maximum at about 2 ms, which corresponds to particles about 245 nm in diameter. The intensity of the spectrum goes down with further increases in particle TOF.

Figure 5B represents normalized ion signals and particle counting frequency versus particle diameter. In the plot, the open symbols represent areas of signal peaks obtained under high particle loadings, 5x106/cc, at three molecular weights (178, 202, and 252 AMU). The values are normalized by area under peak 202 AMU and 280 nm diameter. This is the diameter at which the area under 202 AMU is a maximum. The two filled symbols are normalized numbers of particles counted for particles sources of two different number densities (5x106/cc and 2x105/cc). The filled squares (particles diluted to 2x105/cc by adding a filtered particle free flow) show that particles are distributed around 300 nm in diameter. At 5x106/cc, the results show substantial saturation effects. On the other hand, all the open symbols (PAH signals from the same particle source, 5x106/cc) show roughly the same shape as the filled squares. This indicates that both PAH ion signal levels and particle counting frequency versus particle time-of-flight provide a means of measuring particle size distribution.

To demonstrate the ability of the present approach to measure size- and molecular weight-resolved PAH species carried by soot particles generated from realistic systems, we performed preliminary tests using the exhaust from a small gasoline engine (Honda EG 1400, about 7 horse power). Figure 6A is a typical spectrum averaged for 512 laser shots. The engine is operated in a fuel rich condition by closing the choke valve by 2/3 of full scale (fuel valve was at full scale). The plot shows that the spectrum of engine exhaust is similar to that of propane diffusion flame as shown Figure 5. The soot particles from both sources carry appreciable amounts of PAHs up to 500 AMU. Figure 6B shows that soot particles from the engine also have a size distribution that is similar to that for soot particles from a propane flame. We also conducted measurements on the engine that was operated at less fuel rich conditions. In the experiments, both fuel and air valves were opened at full scale. It is observed that the amount of major PAHs (AMU 202-276) in the less fuel rich condition were about 10 times less than those shown in Figure 6A; and the amount of those heavier ones (>276 AMU) reduced even more (~20 times). This suggests that the fuel rich conditions generate more heavy PAHs (generally more toxic) both in absolute and relative values.

The results in Figures 5 and 6 demonstrate that the present configuration offers a major advantage over other single particle-detecting based instruments in that the present design can make real-time measurements of size- and molecular weight-segregated particles containing epidemiologically important PAH or other organic species and without limits in particle number density. This is because the detection of the present instrument is not limited on single particles, instead, it can detect PAH mass loading of particle assemblies if there is coincidence of multiple particles.

We have compared the data in Figure 6 to PAH data in literature obtained through filter sampling and offline GC/MS analysis. In the data interpretation, total PAH over all size range was obtained by integrating the PAH signals (the circles, the squares, and the triangles) shown in Figure 6B. Data of four PAHs (202, 228, 252, 276 AMU) for which measurement data are available in the literature, also are presented in Figure 7 as C/Ctotal. The AMS data were corrected with absorbance of PAHs at 248 nm assuming that ionization efficiency is proportional to the absorbance. Of course, the assumption is valid only if the ionization process is not saturated, or, in other words, the laser flux is not too high. Data in the literature on light duty vehicles, catalyst equipped autos, and ambient measurements are included in the figure. In the figure, data at each molecular weight were added up for all PAH isomers available in the literature. The figure shows that the AMS data exhibit a similar trend as those GC/MS data, i.e, MW 276 seems a major PAH peak. However, on the other hand, the AMS data at 252 AMU appears much lower than the overall GC/MS data. This might indicate that the AMS laser ionization process for the PAH is saturated so that signal at this MW is over corrected, because this is the peak of the highest adsorbance (two to eight times as other PAHs in the figure).

To obtain absolute mass loading of PAHs carried by particles from AMS data, calibration on particles of known PAH mass is desired. Such calibrations are in progress as part of further development of the AMS.

Summary. An AMS has been developed that uses an aerodynamic inlet, a chopper, and a thermal particle evaporation source. The vaporized particle plume is ionized via a UV laser-induced REMPI process, and ions are analyzed by a molecular TOF mass spectrometer. Tests using soot particles from propane flames and a gasoline engine have demonstrated that the current design is able to detect quantitatively or semi-quantitatively the amount of size- and composition-resolved PAHs carried by particles, in realtime. The present design allows sampling in both highly concentrated particle sources where particle number density is well above 107/cc, and in clean air where particle number density is well below 100/cc.

The current AMS can be used for monitoring the level of PAHs in ambient particles for regulatory purposes or for providing data for epidemiological studies. The AMS also can be used to characterize size- and composition-resolved soot particles from combustion emission sources as a function of combustion conditions, which is important both for developing control strategies and for regulatory decision making.

To obtain absolute mass loading of PAHs carried by particles from AMS data, calibration on particles of known PAH mass is desired. Such calibrations are in progress as part of further development of the AMS.

Acknowledgment and Disclaimer

This material is based upon work supported by the U.S. Environmental Protection Agency (EPA) through an award to MIT, which is gratefully acknowledged. Any opinions, findings, and conclusions or recommendations expressed in this report or future publications are those of the authors and do not necessarily reflect the views of the EPA.

References:

Allen JO, Dookeran NM, Smith KA, Sarofim AF, Tahizadeh K, Lafleur AL. Environmental Science and Technology 1996;30:1023.

Allen J, Gould RK. Review of Scientific Instruments 1981;52:804-809.

Bernstein RB. Journal of Physical Chemistry 1982;86:1178.

Bosel U, Neusser HJ, Schlag EW. Journal of Chemical Physics 1981;55:193.

Fluent. Version 4.47, Fluent Inc., Lebanon, NH, 1995.

Hankin SM, John P, Smith GP. Analytical Chemistry 1997;69:2927.

Hepp H, Siegmann K, Sattle K, Chemical Physics Letters 1995;233:16.

Jayne JT, Leard DC, Zhang X, Davidovits P, Smith KA, Kolb CE, Worsnop DR Aerosol Science and Technology 2000;33:49-70.

Liu P, Ziemann PJ, Kittelson DB, McMurry PH. Aerosol Science and Technology 1995a;22:293.

Liu P, Ziemann PJ, Kittelson DB, McMurry PH. Aerosol Science and Technology 1995b;22:314.

Lubman DM, Kronick MN. Analytical Chemistry 1982;54:660.

Marr LC, Kirchstetter TW, Harley RA, Miguel AH, Hering SV, Hammond SK. Environmental Science and Technology 1999;33:3091.

Rogge WF, Hildemann LM, Mazurek MA., Cass GL. Environmental Science and Technology 1993;27:636.

Sinha MP, Ginffin CE, Norris DD, Estes TJ, Vilker VL, Friedlander SK. Journal of Colloidal and Interface Science 1982;87:140-153.

Squire DW, Barbalas MP, Bernstein RB. Journal of Physical Chemistry 1983;87:1701.

Siegmann K, Hepp S, Frank S, Malinowski N, Martin TP. Journal of Aerosol Science 1995;26:S661.

Tanda TN, Velazquez J, Hemmi N, Cool TA. Physical Chemistry 1993;97:1516.

Venkataraman C, Friedlander SK. Environmental Science and Technology 1994;28:563.

Ziemann PJ, Liu P, Rao N, Kittelson DB, McMurry PH. Journal of Aerosol Science 1995;26:745.

Journal Articles on this Report : 3 Displayed | Download in RIS Format

Publications Views

Other project views:	All 10 publications	3 publications in selected types	All 3 journal articles

Publications

Type	Citation	Project	Document Sources
Journal Article	Jayne JT, Leard DC, Zhang XF, Davidovits P, Smith KA, Kolb CE, Worsnop DR. Development of an aerosol mass spectrometer for size and composition analysis of submicron particles. Aerosol Science and Technology 2000;33(1-2):49-70.	R825391 (1999) R825391 (2000) R825391 (Final) R825253 (Final) R828172 (Final)	Full-text: Taylor & Francis-Full Text PDF Exit Abstract: Taylor & Francis-Abstract Exit
Journal Article	Zhang X, Smith KA, Worsnop DR, Jimenez J, Jayne JT, Kolb CE. A numerical characterization of particle beam collimation by an aerodynamic lens-nozzle system: Part I. An individual lens or nozzle. Aerosol Science and Technology 2002;36(5):617-631.	R825391 (2000) R825391 (Final)	Full-text: Taylor & Francis-Full Text PDF Exit Abstract: Taylor & Francis-Abstract Exit Other: University of Colorado-Full Text PDF Exit
Journal Article	Zhang X, Smith KA, Worsnop DR, Jimenez JL, Jayne JT, Kolb CE, Morris J, Davidovits P. Numerical characterization of particle beam collimation: Part II integrated aerodynamic-lens--nozzle system. Aerosol Science and Technology 2004;38(6):619-638.	R825391 (Final)	Full-text: Colorado University-Full Text PDF Exit Abstract: Taylor&Francis-Abstract Exit Other: Taylor&Francis-Full Text PDF Exit

Supplemental Keywords:

exposure, risk, risk assessment, health effects, ecological effects, carcinogen, environmental chemistry, monitoring, transportation, particles, polycyclic aromatic hydrocarbons, PAHs, emissions, atmosphere, ambient, aerosol, instrument, field, environmental fate, air pollution., RFA, Air, Scientific Discipline, Engineering, particulate matter, mobile sources, Environmental Chemistry, Engineering, Chemistry, & Physics, air toxics, aerosol particles, fine particulates, mutagenic pollutants, human exposure, fine particles, combustion, fate, ultraviolet excimer laser, vapor plume, monitoring, atmospheric particles, ambient emissions, particulates, air quality models, flight mass spectrometer, PAH

Progress and Final Reports:

Original Abstract

1997 Progress Report

1998

1999 Progress Report

2000 Progress Report