Reply to comment by S. Decesari et al. on “On the use of anion exchange chromatography for the characterization of water soluble organic carbon”

[1] Chang et al. [2005] applied an anion exchange chromatography technique, first proposed by Decesari et al. [2000] to characterize water soluble organic carbon (WSOC) in fog samples and in aqueous extracts of aerosol samples. The chromatograms obtained were similar to those reported by Decesari et al. and the method was successful in classifying a large fraction of organic carbon in our samples. We also reported that challenges to the method, made by analysis of individual organic compounds important in atmospheric samples, revealed some problems with WSOC classification by the method as previously described in the peer-reviewed literature. In particular, several compounds eluted in fractions different than one would expect based on the names assigned by Decesari et al. [2000] to the eluted fractions. Because results obtained using the method of Decesari et al. [2000] have gained widespread attention as a way to help classify the complex mixture of organics present in atmospheric samples and because previous publications had not clearly demonstrated the performance of the published method when challenged by analysis of individual atmospherically relevant compounds, we felt it was important to bring attention to observed limitations of the published protocol. We also felt compelled to caution users of results obtained using the published protocol to exercise some caution in interpreting the results as an accurate description of the composition of complex, real atmospheric aerosol samples. It was not our intent to discredit the application of anion exchange chromatography as one of many useful tools for characterizing organic mixtures in fogs and aerosols. Rather, we hoped that our publication would provoke a more thorough evaluation of this approach that would ultimately lead to improvements in our understanding of fog and aerosol composition. [2] The response to our publication, provided by Decesari et al. [2005], indicates that our goals of prompting further discussion of method performance and limitations and of improving its performance through various analytical modifications are already being met. We commend these responders on taking up the challenge and look forward to ongoing improvements in their novel approach to the challenging problem of classifying the organic composition of atmospheric samples.