The presence of a small ‘peak’ during the fourth temperature step is not necessarily due to carbonate. Other carbonaceous matter (e.g., char) is sometimes removed during the fourth step. Unlike carbonates, which produce a relatively sharp peak, other materials typically evolve as a small, broad peak. If determination of relatively low (e.g., CC < 10% of TC) amounts of CC is desired, a second portion of the sample should be acidified and analyzed to verify the presence of carbonate and reveal any underlying baseline features contributed by other materials. A more accurate estimate of CC can then be obtained by integrating the first (non-acidified) sample over the missing peak area (i.e., area removed through acidification). f. Sampling Artifacts: Organic Aerosol Face velocity. Quartz-fiber filters are routinely used to collect airborne particulate matter for carbon determination. Quartz filters have high collection efficiency for particulate carbon, but collection of organic aerosol is artifact prone [63–67]. Particulate carbon (EC and nonvolatile OC) results should not depend on filter face velocity. This is true for EC, but OC (and TC) results can have a face velocity dependence due to sampling artifacts. Namely, adsorption of organic vapor can positively bias the results, while evaporative losses have the opposite effect. Although both processes may occur, recent studies indicate vapor adsorption is the dominant artifact [65, 67]. In one study [65], the OC (TC) results for air samples collected at face velocities of 20, 40, and 80 cm/sec had a face velocity dependence before, but not after, correction for adsorbed vapor. Lack of dependence after correction supports adsorption as the dominant artifact. At the much lower face velocities typical of occupational monitoring (e.g., about 4 cm/sec with a 37-mm filter and 2 L/min flow rate), OC losses induced by pressure drop across the filter are expected to be minor relative to adsorption. OC sampling artifacts were acknowledged when the thermal-optical method was proposed [48, 49], but this issue was not investigated prior to publication of NMAM 5040 because an EC surrogate was recommended. In view of the MSHA-proposed air standard for TC, and the fact that both OC and EC are determined by the method, the issue has since been addressed. Adsorbed OC. Correction for adsorbed organic vapor through use of traditional blanks (media or field) may not be accurate because the amount of vapor adsorbed by them is variable [68]. More importantly, traditional blanks collect vapor passively, while the samples collect it actively (during sampling). A more representative correction for adsorbed organic vapor can be made through use of two filters in tandem [65, 67]. The air sample is collected with a sampler containing a Teflon or quartz upper filter and a bottom quartz filter. After sampling, the bottom filter is used to correct for adsorbed vapor. The vapor adsorbed on the bottom filter more closely represents that adsorbed on the sample filter (quartz) because both are collected actively. Use of two quartz filters is preferable to Teflon and quartz because the filter used to correct for adsorbed vapor is in the same sampler; however, the second (bottom) quartz filter may underestimate the amount of adsorbed OC relative to a quartz filter under Teflon [65, 67]. These results have been attributed to depletion of the vapor concentration by the top quartz filter, which presents a lower concentration to the bottom one. A shorter equilibration time (for partitioning between gas phase and adsorbed state) is expected with Teflon filters because they have less surface area and are more inert than quartz. Thus, less adsorption is expected on Teflon relative to quartz. Because differences in the amount of adsorbed OC have been reported for the two sampling configurations, both were examined.
3/15/03
237
NIOSH Manuual of Analytical Methods
