Automated aerosol Raman spectrometer for semi-continuous sampling of atmospheric aerosol. (February 2017)

• Record Type: Journal Article
• Authors: Doughty, David C.; Hill, Steven C.
• Is Part Of: Journal of quantitative spectroscopy & radiative transfer. Volume 188(2017)
• Page Start: 103
• Page End: 117
• Publication Date: 2017-02
• Subjects: Raman Spectroscopy -- Aerosol -- HULIS/Soot; Spectrum analysis -- Periodicals; Radiation -- Periodicals; Analyse spectrale -- Périodiques; Rayonnement -- Périodiques; Radiation; Spectrum analysis; Periodicals; 543.0858
• DOI: 10.1016/j.jqsrt.2016.06.042
• Languages: English
• ISSNs: 0022-4073

Abstract

Abstract: Raman spectroscopy (RS) is useful in characterizing atmospheric aerosol. It is not commonly used in studying ambient particles partly because automated instrumentation for aerosol RS has not been available. Battelle (Columbus, Ohio, USA) has developed the Resource Effective Bioidentification System (REBS) for automated detection of airborne bioagents based on RS. We use a version of the REBS that measures Raman spectra of one set of particles while the next set of particles is collected from air, then moves the newly collected particles to the analysis region and repeats. Here we investigate the use of the REBS as the core of a general-purpose automated Aerosol Raman Spectrometer (ARS) for atmospheric applications. This REBS-based ARS can be operated as a line-scanning Raman imaging spectrometer. Spectra measured by this ARS for single particles made of polystyrene, black carbon, and several other materials are clearly distinguishable. Raman spectra from a 15 min ambient sample (approximately 35–50 particles, 158 spectra) were analyzed using a hierarchical clustering method to find that the cluster spectra are consistent with soot, inorganic aerosol, and other organic compounds. The ARS ran unattended, collecting atmospheric aerosol and measuring spectra for a 7 hr period at 15-min intervals. A total of 32, 718 spectra were measured; 5892 exceeded a threshold and were clustered during this time. The number of particles exhibiting the D-G bands of amorphous carbon plotted vs time (at 15-min intervals) increases during the morning commute, then decreases. This data illustrates the potential of the ARS to measure thousands of time resolved aerosol Raman spectra in the ambient atmosphere over the course of several hours. The capability of this ARS for automated measurements of Raman spectra should lead to more extensive RS-based studies of atmospheric aerosols. Highlights: Automated collection and measurement of single-particle Raman spectra is illustrated. Spectra of atmospheric aerosol measured in 15 min increments for 7 hr are shown. 5000 spectra were clustered into inorganic, HULIS/Soot, organic and fluorescent types. Some ambient supermicron aerosols were seen to not be homogeneously mixed. The number of soot-like spectra increased during the morning rush hour then decreased. The instrument should be useful characterizing and studying atmospheric aerosols.