The EPA considers PM_2.5 (particulate matter with particle sizes of 2.5 microns or less) to be the most important air pollutant because it can cause significant environmental and health problems. On 17 October 2006, the EPA issued the final amendments to the National Ambient Air Quality Standards (NAAQS) that reduce the PM_2.5 24-hour standard from 65 μg/m³ to 35 μg/m³ (Federal Register: Oct. 17, 2006, 71, 200). This regulation could pose a challenge for Department of Defense (DoD) air bases because almost all of the particles produced in gas turbine engines are PM_2.5 emissions. They include solid particulates (soot) and volatile aerosols formed from condensed gases and chemi-ions. To date, attention has been directed primarily at the carbonaceous particulate matter. However most of the particulate emissions in aircraft exhaust are in the form of condensed volatile aerosols. These particulates can grow via continuing condensation and agglomeration. They consist mostly of acids, such as H₂SO₄/H₂O, HNO₃/H₂O and H₂O/H₂SO₄/HNO₃, as well as neat H₂O. These aerosols can have a significant impact on both the local and global environment by providing surface areas for heterogeneous chemistry and sinks for already present atmospheric condensable gases. These effects are not understood nor have they been adequately characterized.

The objective of this proof-of-concept project was to evaluate extreme light, femtosecond (fs)-laser induced breakdown spectroscopy (LIBS) and conventional nanosecond (ns)-LIBS as techniques for accurately making time and spatially resolved in-situ measurements of total mass, composition, number density, and size distribution of solid and volatile aerosol particulates in controlled, simulated turbine-engine plume environments.

The approach was to explore the use of nanosecond (ns, 10-9 s) and femtosecond (fs, 10-15 s) LIBS in ways that would take advantage of the difference in LIBS response to species in gaseous and particulate states. Carbon was selected for study because it is relatively straight forward to generate gaseous and particulate sample streams and because there are published results documenting the different in-LIBS response for carbon in multi-phase samples. The conditions studied were not those expected in a gas turbine engine exhaust but those that were conducive to exploring new concepts for in situ measurements of total particulates. If a concept proved feasible, then it would be evaluated in more realistic environments. This program investigated a combined single- and dual-pulse ns-LIBS technique for overcoming the multi-phase problem. The idea of using fs-LIBS resulted from literature studies that suggest fs lasers, can dissociate a sample more efficiently than an ns laser.

This program demonstrated that single ns-LIBS and dual ns/ns-LIBS measurements could be used, in a unique way, to estimate the percentage of gaseous and total particulate carbon in a mixture of air, CO₂, and carbon particulates. The idea followed from the observation that the dual ns/ns-LIBS signal increased with increasing particulate carbon concentration at a different rate than the traditional single ns-LIBS signal. Calibration curves were obtained that quantitatively related the dual-pulse ns/ns-LIBS and single-pulse ns-LIBS signals with the percentage of gaseous and particulate carbon. Using these calibration curves, it was demonstrated that the ratio of the dual- to single-pulse LIBS signals could be used to estimate the percentage of gaseous and total particulate carbon in a multiphase mixture.

Although the sensitivity of the technique needs to be improved, an approach that shows promise for distinguishing between gas and particle species was demonstrated. In an attempt to improve the sensitivity of this approach, the replacement of the dual ns/ns-LIBS with fs/ns-LIBS was investigated. An fs/ns-LIBS capability was established and demonstrated. Single-pulse ns-LIBS and dual-pulse fs/ns LIBS were then used to obtain calibration curves for: (1) carbon dioxide, (2) oxalic acid, and (3) mixtures of air and these species. The results showed that both the single and dual LIBS techniques gave a larger response to carbon from particulates than that from gases.

However, the slopes of the linear calibration curves were nearly the same for the single and dual LIBS techniques. The impact was that the ratio of the fs/ns-LIBS to the ns-LIBS signals was almost independent of percent gaseous carbon. Thus, the use of single and dual LIBS to determine the percentage of gaseous and total particulate carbon did not seem to apply to the dual fs/ns- LIBS in place of the ns/ns-LIBS for the conditions studied. The cause of the difference in sensitivity of the dual ns/ns-LIBS and the fs/ns LIBS techniques is not clearly understood at this time but it is likely related to the significantly reduced interaction volume and plasma dynamics associated with the fs laser plasma.

This project explored the feasibility of developing a LIBS based technique for measuring total particulates in a sample containing carbon in gas and particulates species. The program was partly successful in that a LIBS technique was demonstrated for measuring the percentage of gaseous carbon and total particulate carbon in a multi-phase sample. The technique involves establishing a calibration curve that relates the ratio of dual to single-pulse ns-LIBS signals to the percentage of gaseous or total particulate carbon in a multi-phase sample. The program is considered only partly successful because additional research is required to improve the sensitivity so it is a viable technique for other species and for broader ranges of solid and gas ratios.

While this project was not entirely successful, the information obtained will provide the basis for future studies aimed at the development of in situ measurement techniques for volatile particulate matter in aircraft engine exhaust, an extremely difficult and challenging measurement problem.  Recommendations are provided for additional research to investigate ways to improve the sensitivity of using the ratio of dual and single pulse LIBS signals to determine the total percentage of particulates in a multi-phase sample.  Recommendations are also presented on other approaches for measuring particulates that resulted during the course of this project.