Grantee Research Project Results
Final Report: Particle Sampler for On-Line Chemical and Physical Characterization of Particulate Organics
EPA Grant Number: R831077Title: Particle Sampler for On-Line Chemical and Physical Characterization of Particulate Organics
Investigators: Smith, Kenneth A. , Worsnop, Douglas R.
Institution: Massachusetts Institute of Technology , Aerodyne Research Inc.
EPA Project Officer: Chung, Serena
Project Period: October 1, 2003 through August 1, 2006 (Extended to September 30, 2007)
Project Amount: $410,000
RFA: Measurement, Modeling, and Analysis Methods for Airborne Carbonaceous Fine Particulate Matter (PM2.5) (2003) RFA Text | Recipients Lists
Research Category: Air , Air Quality and Air Toxics , Particulate Matter
Objective:
The overall objective of this STAR project is to develop an innovative particle sampler that can be coupled to pre-existing commercial or research-grade analytical instruments for on-line, size-resolved, high time resolution (on the order of 1 hour) analysis of individual organic species in ambient aerosol particles. This coupling will enable:
- High efficiency separation of particles from the gas phase.
- Concentration of collected aerosol on a cryo-cooled surface under high vacuum (avoiding problems associated with filter sampling and solvent extractions).
- Direct injection of desorbed species from the particle sample into analytical instruments.
This program is a collaborative effort between the research groups led by Professor Kenneth A. Smith at the Massachusetts Institute of Technology (MIT), Department of Chemical Engineering, and Dr. Douglas R. Worsnop at Aerodyne Research, Inc. (ARI).
Summary/Accomplishments (Outputs/Outcomes):
The following achievements have been completed during the project period:
- The prototype particle sampler was designed and constructed in the first year. Polydisperse oleic acid samples were collected and detected with an attached GC/MS system. The particle collector was redesigned to minimize the dead volume. A schematic of the improved collector is shown in Figure 1. The collector consists of a copper part on which a stainless steel tube is brazed. The stainless steel tube is connected to the valve which connects the collector to the vacuum chamber. On the back end of the copper part of the collector, there is a hole for a cartridge heater for heating the collector. When the collector is cooled, the heater is switched off and liquid nitrogen is blown to the collector.
- In the second year, two 4-port valves (Valco valves) were installed for the necessary flow cycles of particle sampling, cleaning of the collector, and analysis. Additionally a new kind of inlet to the GC/MS, the Volatile Interface, was installed to increase the reproducibility of the sample retention time.
- An interrupt-driven C program was written to control the parameters (temperatures and durations of the stages of the sampling cycles) of the particle sampler and to switch the valves automatically. The GC/MS is a slave in this program and it is possible to start the GC/MS program and the flow control of the volatile interface automatically at a preset time. With this program, autonomous data collection is possible.
- Experiments were performed with the particle sampler using a hydrocarbon standard (ASTM) with several components. The chromatograms of these experiments show only some of the components of the standard. The associated peaks were very well resolved and sharp. Which peaks appeared depended on the chosen desorption temperature. Other experiments were performed in which a liquid sample of the ASTM standard was injected directly into the GC instead of via the particle sampler to investigate the transfer characteristics of the particle sampler. The chromatograms from these experiments yielded more individual components than the experiments with the particle sampler. This shows that the smaller hydrocarbons (for example hexane) are too volatile to be detected with the particle sampler. Furthermore, the largest molecular weight components decompose in the particle sampler.
- The particle sampler was connected to a Proton Transfer Reaction/Mass Spectrometer (PTR/MS) which is owned by Pacific Northwest National Laboratory (PNNL). Experiments were performed with different parameters, such as, temperatures of valves and lines, and times of different stages of a sampling cycle. Figure 3 shows the results for motor oil for two different temperatures of the Valco valves and the transfer line. The higher temperature leads to a more rapid transfer of material to the detector. These experiments with faster sampling time suggested that some of the sample is sticking and desorbing in the transfer line. By coating the transfer line with Silcosteel from Entech Instruments it was possible to get more reproducible results.
- Experiments were also made with oleic acid (C18H34O2, MW=282 amu) because it is a good surrogate for ambient organic aerosol. More mass than expected was needed to give a measurable response because part of the oleic acid decomposes. This was shown by GC/MS mass spectra where the majority of the ion intensity is low molecular weight fragments (<90 amu). Also significant levels of CO2 formed by de-carboxylation of the oleic acid were observed. This shows that thermal decomposition may be a limiting factor for this technique. However, ambient aerosol contains hundreds of other compounds which can be detected with the ACM.
- Experiments with a candle flame as a combustion source were performed to show the potential of the ACM for multi-component samples. Figure 4 shows the chromatogram from one of these experiments in which many (>15) individual components can be resolved. From the NIST mass spectra library, most of these peaks are identified as polyaromatic and aliphatic hydrocarbons (tri-, tetra-, penta-, and hexadecane, benzene and naphthalene derivatives). A cryotrap at the head of the column is not necessary because the sample is collected at the head of the cold column during the desorption stage and is desorbed after heating the column in the particular GC method used here.
- Tests of the particle sampler were performed at the Helmholtz Research Center in Juelich, Germany in 2007. The prototype of the particle sampler was coupled to their smog chamber where different aerosol compositions and loadings can be systematically produced. In these experiments a GC owned by the Helmholtz Research Center was used as the detector. With the ACM it was possible to speciate the aerosol in the chamber. Results from one chamber experiment at two different times are shown in Figure 5.
- All these experiments show that the transfer of the sample to the detector is critical for successful detection of particle components. For this reason several improvements in the transfer were introduced:
- The transfer line was shortened to minimum length (40 cm) and is now coated with Siltek from Restek instead of Silonite from Entech Instruments. The Siltek coating is designed for GC applications for semivolatile compounds.
- A new pair of Valco valves with valve bodies which are coated with Siltek were used for the new experiments.
- The inlet to the GC used for former experiments was a Volatile Interface (VI) from Agilent. This VI is coated with Silcosteel from Restek. However, Restek recommends a Siltek coating for chromatography applications. Therefore, the column was moved into the transfer line to avoid contact of the sample to the Silcosteel coated VI. This led to a significant increase in sensitivity.
- A dual chopper was designed to provide aerosol size selection. The dual chopper was constructed and tested during Year4 (see Figure 8).
Figure 1. Schematic of the collector
Figure 2. Carrier gas flow pathway
The carrier gas flow pathway through the Valco valves is shown in Figure 2. Before an experiment, the collector is in standby (Figure 2A). Carrier gas passes through the collector to a vent. For particle sampling (Figure 2B), the collector volume is decoupled from the carrier gas flow and the valve to the vacuum chamber is open. In this stage the collector is cooled with liquid nitrogen. After sampling the particles, the valve to the vacuum chamber is closed and the Valco valves are switched to the injection stage with the carrier gas flow through the collector to the analyzer (Figure 2D). Directly after switching the valves the collector is heated up and the sample desorbs into the carrier gas. Alternatively it is possible to heat up the collector before switching the Valco valves to injection/transfer (Figure 2C). Experiments showed however, that, in this case, the sample adsorbs on the walls of the thermally isolating support which is not heated. After the transfer of the sample to the detector the Valco valves switch to backflush and the collector is kept hot to clean it.
Figure 3. Motor oil detected with the PTR/MS at two different temperatures of the valves and the transfer line (red graph 150° C, blue graph 200° C)
Figure 4. Chromatogram of soot from a candle flame
Figure 5. Two measurements with the ACM during one chamber experiment. The first measurement was taken during the maximum of mass in the chamber, the second one about 22 hours later with a smaller total mass
Comparisons with SMPS data showed that for the upper graph a mass of about 4.6 μg and for the lower graph of about 1.2 μg was collected. The large peak at 19.5 min was identified as nopinone. The change of the peak ratio of the peaks at 31 and 32 min (not yet identified) shows the change of the composition of the aerosol in the chamber.
Figure 6. Comparison of the total number of GC counts from all measurements during one chamber experiment measured with the ACM (cross) with the total SMPS mass concentration (dotted line) and the total AMS organics mass concentration (straight line).
Figure 6 shows a comparison of the total GC counts from all measurements during one chamber experiment with the ACM (crosses) with the total SMPS mass concentration (dotted line) and the total AMS organics mass concentration (solid line). The SMPS and AMS data were taken later during a repeat chamber experiment. The total SMPS mass concentration is smaller than the total AMS organics mass concentration because the SMPS missed larger particles (> 630 nm) because of the use of an impactor in front of the SMPS.
The analysis of these experiments is still in progress. These experiments show the successful application of the ACM for chamber studies as an environmentally relevant use. They also show the success of the concept of the ACM as a self-contained module that can be attached to any analytical instrumentation.
With these improvements, experiments with octadecane with different mass loadings were performed. The octadecane particles were produced in an atomizer and size- selected with a differential mobility analyzer (DMA, 300nm). The particle number concentration entering the ACM was measured using a condensation particle counter (CPC). The results of these experiments are shown in Figure 7. The numbers are the experiment number. For the first time in experiments with the ACM, the sensitivity did not decrease over time, and the measured mass was reproducible for a given injected mass. The slight deviation from linearity and the small non-zero intercept may be due to partial adsorption of the sample on the transfer line and the valves in spite of improvements in the design.
Figure 7. Corrected peak area of the octadecane peak from the chromatograms versus mass loading obtained from DMA and CPC
After each experiment a blank was performed. In the blank, the experimental protocol was repeated but air was sampled through a particle filter. When there was still a signal, another blank was performed until the blank was clean. Usually a small signal (about 10% of the signal in the experiment) was detected in the first blank and the second blank was clean. In experiments with high mass loadings there was also a small signal in the second blank.
After long breaks (for example over night or over the weekend) the first blank showed signal even if the blank before was clean. This behavior also occurred when no sampling took place and only the column of the GC was heated up. This indicates that some sample adsorbs on the wall of the transfer line and the valves and desorbs slowly with time and is collected at the head of the (relatively cool) GC column. When the GC column temperature is ramped up, this material elutes into the MS detector.
Figure 8. Drawing of the dual chopper size selector
This size selector consists of two chopper wheels with different diameters mounted on a shaft, with a known offset of the two slits and driven by a stepper engine. For a given rotational velocity, only particles of a given speed (depending on their size) will pass to the collector. The chopper wheels can be moved in and out of the particle beam in four modes: 1) The beam can be blocked or 2) be transferred completely. In the third mode, the particle beam passes the smaller wheel and hits the larger wheel. This is the chopped mode in which each group of particles passing through the chopper slit contains the whole size distribution. And finally the beam can be chopped at both wheels which have slits with a radial offset. In this mode (velocity select mode) only particles with a certain velocity and hence a certain size are transferred through the second slit.
Figure 9. Results with the dual chopper size selector for chopper wheels giving 1% and 2% duty cycle
This system was tested in an Aerodyne Aerosol Mass Spectrometer (AMS) system using polydisperse ammonium nitrate particles. Results for chopper wheels with 1% and 2% slits are shown in Figure 9. The signals are from the mass spectrometer for mass 46 (NO2+). On the x-axis is the time which can be assigned to a size by a velocity calibration, e.g., with PSL particles whose sizes are known. The black curve is in chopped mode and shows the size distribution of the sample. The other curves are in velocity-select mode with different velocities of the wheels, i.e., different transit times to pass through both slits. Figure 9 shows that for different velocities of the chopper wheels different sizes are transferred. A comparison of the top and bottom panels of Figure 9 shows that smaller slits achieve better size resolution. In the ACM the velocity selector will allow the collection of aerosol mass for different particle size bins by choosing the rotational speed of the chopper wheels.
Conclusions:
The new particle sampler was designed, constructed and tested extensively. The experiments with the candle flame and those at the smog chamber in Juelich demonstrate the use of the particle sampler for samples with many compounds. It is possible to resolve several components with the GC/MS. A cryotrap at the head of the column is not necessary because the sample is adsorbed at the head of the column and desorbed from there when the column is heated during the GC/MS cycle. In ambient experiments with hundreds of compounds which cannot be resolved with a GC, coupling to a GCxGC/MS detector may give better chemical identification.
The experiments with the PTR/MS show the possibility of using the ACM with different detectors which was one of the goals of this project. Also for the experiments at the smog chamber in Juelich they used their own detector. This shows the success of the concept of the self-contained module which limits the cost and was one major goal of this project.
The experiments in Juelich show the possibility of applying the particle sampler for environmentally relevant research. With the new particle sampler, speciation of the chamber aerosol was achieved. In these experiments, the long transfer line coated with Silonite from Entech Instrument was still in use. In future experiments, the redesigned transfer coupling with better sensitivity will allow lower aerosol mass loadings to be measured.
All the experiments indicate that the transfer of the sample to the detector is critical and coatings and temperatures of all parts of the transfer hardware have to be optimized. Improvements in the hardware included a shorter transfer line with a new coating, coated Valco valves, and moving the column into the transfer line (see point 9 in the accomplishments). With these improvements, the experiments were reproducible and the sensitivity was increased.
The sensitivity is not yet sufficient for ambient aerosol measurements. With the current sensitivity and an ambient aerosol loading of about 4 μg/m3, sampling of about 20 hours would be needed to successfully speciate the organic components. This means that a further improvement of a factor of 10 to 20 is necessary. Further improvements may be possible by optimizing coatings and temperatures of the transfer parts. Experiments will continue with in-house support at Aerodyne Research, Inc. and with support from an SBIR grant from DOE. Further support comes from the collaboration with the Helmholtz Research Center in Juelich where the smog chamber data are still being analyzed. They are still very interested in buying a particle sampler. More experiments (with the improved particle sampler) with a PTRMS will be performed at the beginning of 2008.
The results of this project show the power of this technique for on-line compound separation and identification without costly post-collection analysis procedures. These post- collection analysis procedures are subject to artifacts, e.g., by chemistry on filters. Experiments with the AMS also show an efficient separation of the gas phase from the particle phase.
Overall this project was successful in demonstrating the technology for a particle sampling module for different detectors which can sample and speciate on-line particles without post-collection analysis procedures. The ACM will provide an important tool for environmental studies as shown by the application of the particle sampler in the Juelich smog chamber. The goal of sampling ambient aerosol was not achieved because the problems with the transfer were underestimated. However, much progress has been made on increasing sensitivity and reproducibility.
Project Tasks
|
|
|
% Complete |
Year 1 |
Task 1 |
Design and build particle sampler |
100 |
|
Task 2 |
Couple particle sampler to vacuum chambers |
100 |
Year 2 |
Task 3 |
Laboratory testing of particle sampler with GC/MS |
80 |
Year 3 |
Task 4 |
Modify particle sampler based on lab testing. |
85 |
|
Task 3b |
Laboratory testing of particle sampler with PTR/MS |
50 |
Year 4 |
Task 5 |
On-line characterization of ambient aerosol at ARI. |
0 |
|
Task 6 |
On-line characterization of urban and rural aerosol. |
0 |
|
Task 7 |
Data analysis and publication of results. |
0 |
Quality Assurance
Quality assurance requirements are being met through repeated laboratory experiments in order to determine the reproducibility, quantification and linearity of the data generated by the particle sampler. Because this technology is still under development, the QA process is under development as well.
Environmental experiments have not yet been performed. When environmental experiments are performed, we will address quality assurance requirements through careful calibration of the instrument and comparisons with data produced by co-located, well-characterized instrumentation, such as an AMS.
Journal Articles on this Report : 2 Displayed | Download in RIS Format
Other project views: | All 6 publications | 2 publications in selected types | All 2 journal articles |
---|
Type | Citation | ||
---|---|---|---|
|
Hohaus T, Trimborn D, Kiendler-Scharr A, Gensch I, Laumer W, Kammer B, Andres S, Boudries H, Smith KA, Worsnop DR, Jayne JT. A new aerosol collector for quasi on-line analysis of particulate organic matter: the Aerosol Collection Module (ACM) and first applications with a GC/MS-FID. Atmospheric Measurement Techniques 2010;3(5):1423-1436. |
R831077 (Final) |
Exit Exit |
|
Liu PSK, Deng R, Smith KA, Williams LR, Jayne JT, Canagaratna MR, Moore K, Onasch TB, Worsnop DR, Deshler T. Transmission efficiency of an aerodynamic focusing lens system: comparison of model calculations and laboratory measurements for the Aerodyne Aerosol Mass Spectrometer. Aerosol Science & Technology 2007;41(8):721-733. |
R831077 (Final) |
Exit Exit |
Supplemental Keywords:
RFA, Scientific Discipline, Air, Ecosystem Protection/Environmental Exposure & Risk, RESEARCH, particulate matter, Environmental Chemistry, Monitoring/Modeling, Monitoring, Environmental Monitoring, Ecological Risk Assessment, particle size, atmospheric dispersion models, atmospheric measurements, analysis of organic particulate matter, chemical characteristics, human health effects, air quality models, monitoring stations, gas chromatography, air quality model, air sampling, modeling, analytical chemistry, particulate matter mass, particle sampler, modeling studies, aerosol analyzersProgress 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.