Conny J M, Norris G A, Gould T R
Surface and Microanalysis Science Division, National Institute of Standards and Technology, 100 Bureau Drive Stop 8372, Gaithersburg, MD 20899, United States.
Anal Chim Acta. 2009 Mar 9;635(2):144-56. doi: 10.1016/j.aca.2008.12.046. Epub 2009 Jan 8.
Thermal-optical transmission (TOT) analysis measures black carbon (BC) in atmospheric aerosol on a fibrous filter. The method pyrolyzes organic carbon (OC) and employs laser light absorption to distinguish BC from the pyrolyzed OC; however, the instrument does not necessarily separate the two physically. In addition, a comprehensive temperature protocol for the analysis based on the Beer-Lambert Law remains elusive. Here, empirical response-surface modeling was used to show how the temperature protocol in TOT analysis can be modified to distinguish pyrolyzed OC from BC based on the Beer-Lambert Law. We determined the apparent specific absorption cross sections for pyrolyzed OC (sigma(Char)) and BC (sigma(BC)), which accounted for individual absorption enhancement effects within the filter. Response-surface models of these cross sections were derived from a three-factor central-composite factorial experimental design: temperature and duration of the high-temperature step in the helium phase, and the heating increase in the helium-oxygen phase. The response surface for sigma(BC), which varied with instrument conditions, revealed a ridge indicating the correct conditions for OC pyrolysis in helium. The intersection of the sigma(BC) and sigma(Char) surfaces indicated the conditions where the cross sections were equivalent, satisfying an important assumption upon which the method relies. 95% confidence interval surfaces defined a confidence region for a range of pyrolysis conditions. Analyses of wintertime samples from Seattle, WA revealed a temperature between 830 degrees C and 850 degrees C as most suitable for the helium high-temperature step lasting 150s. However, a temperature as low as 750 degrees C could not be rejected statistically.
热光透射（TOT）分析用于测量纤维滤膜上大气气溶胶中的黑碳（BC）。该方法对有机碳（OC）进行热解，并利用激光吸收将BC与热解后的OC区分开来；然而，该仪器并不一定能在物理上分离这两者。此外，基于比尔-朗伯定律的用于分析的全面温度方案仍然难以捉摸。在此，使用经验响应面模型来展示如何修改TOT分析中的温度方案，以便基于比尔-朗伯定律将热解后的OC与BC区分开来。我们确定了热解后OC（σ(Char)）和BC（σ(BC)）的表观比吸收截面，这考虑了滤膜内的个体吸收增强效应。这些截面的响应面模型源自三因素中心复合因子实验设计：氦气阶段高温步骤的温度和持续时间，以及氦气-氧气阶段的加热增量。随仪器条件变化的σ(BC)的响应面显示出一个脊，表明氦气中OC热解的正确条件。σ(BC)和σ(Char)表面的交点表明了截面相等的条件，满足了该方法所依赖的一个重要假设。95%置信区间表面定义了一系列热解条件的置信区域。对华盛顿州西雅图冬季样本的分析表明，830℃至850℃之间的温度最适合持续150秒的氦气高温步骤。然而，低至750℃的温度在统计上也不能被排除。
