J.D. Kessler and L.A. Currie

Objective: To expand the capabilities of conventional accelerator mass spectrometry (AMS) and scanning electron microscopy (SEM) measurements to accommodate the need for synergistic measurement techniques capable of carbon quantification at the sub-microgram level.

Problem: The lack of sufficient and automated techniques for preparation, measurement, and evaluation of low carbon concentration (ca. 1-10ng/g) aerosol samples, accompanied by the increasing realization that black carbon (BC) contributes to global warming, has fueled the investigation to solve these problems. Characterization of these aerosols must be conducted on three levels: (1) mineralogical characterization on a particle-by-particle basis with SEM to fingerprint the BC origin(s); (2) separation of the BC fraction of the sample; (3) establishment of submicromole AMS techniques for measuring 14C to distinguish between fossil and biomass burning sources of BC. Due to the vast quantities of geographical areas requiring these techniques, from polar to remote to rural, automated techniques are of interest.

Approach: All techniques and materials to achieve this objective were established at NIST and the National Ocean Sciences Accelerator Mass Spectrometry (NOSAMS) Facility with the use of NIST SRM 1649a, 1515, and 2975 and various Greenland aerosol, snow, firn, and ice core samples. Automated data collection methods and a reusable boron substrate created by Windsor at NIST have allowed rapid elemental quantification with 6=Z=30 in the SEM using energy and wavelength-dispersive spectroscopy. Investigation of several BC separation methods proved BC as an entity defined by its method of separation. The method of Thermal Optical Kinetics (TOK), developed by Currie, proved most automated, time efficient, and produced the lowest blank (<0.5µg C) in comparison to the other BC separation methods. The implementation of the microsample combustion-dilution facility at NIST in conjunction with the NOSAMS small sample (25µg) AMS target preparation facility have demonstrated successfully the modern carbon quantification limit (10% rsd) of ca. 0.8µg carbon.

Figure 1. Thermal character of organic (OC) and black carbon (BC) from a filtered, 6.1 kg surface snow sample collected on 19 December 1997.

Figure 2. EDS spectrum and SEM image (1000X) of a carbonaceous particle on a boron substrate extracted from 1996 Greenland snow pit.

Results and Future Plans: Three techniques are now established at NIST capable of preparing and measuring the elemental, with particular interest in carbon, content of remote polar samples. Figure 1 illustrates an example of the ability to quantitatively measure the organic and black carbon species in a sample of ca. 5 µg carbon by thermal character measurement. Figure 2 illustrates an image and subsequent X-ray spectrum of a carbonaceous particle extracted from a 1996 Greenland snow pit. The initial results resoundingly prove that quantification of carbonaceous species in remote air, snow, and ice samples is possible. The automated techniques to date encompass data collection, but are particularly lacking in linking X-ray spectroscopy data to their appropriate aerosols and minerals. This type of automated data analysis would not only be useful in the study of ice core, but in any geological field where mineral identification is necessary. Following data collection and evaluation optimization, these techniques can be applied to any measurement field where the study of atmospheric aerosols is of interest.

Web Contact micro@nist.gov