Amanda Hughes - Identification of Structural Properties Associated with Polychlorinated Biphenyl Dechlorination Processes

Version 1

      Publication Details (including relevant citation   information):

      Hughes, A. S., J. M. Vanbriesen and M. J. Small (2010).   "Identification of Structural Properties Associated with   Polychlorinated Biphenyl Dechlorination Processes." Environmental   Science & Technology 44(8): 2842-2848.

      Abstract:

      Polychlorinated biphenyl molecules can be biologically   dechlorinated through sequential losses of a chlorine atom,   following 840 pathways from higher chlorinated to lesser   chlorinated congeners and biphenyl. Previously, eight recurring   sets of pathways, herein referred to as explicitly reported

      pathways in dechlorination processes, have been identified   through qualitative analysis of shifts in congener masses in   field and laboratory studies. Dechlorination process   generalizations were qualitatively extrapolated based on limited   attributes of the congeners dechlorinated in the explicitly   reported pathways. They are valuable because they allow   comparisons of dechlorination patterns across laboratory   experiments and contaminated sites. However, due to analytical   limitations and a paucity of studies, the explicitly reported   pathways in dechlorination processes likely do not represent all   of the pathways that could occur at contaminated sites. This work   presents an alternative, quantitative, and replicable approach to   the identification of candidate pathways for inclusion in   dechlorination process generalizations through use of   classification trees. This method considers 46 structural and   property attributes of dechlorination pathways. Trees fit for   pathway inclusion in each of the eight dechlorination processes   with alternative assumptions are compared in terms of critical   congener attributes. The classification trees correctly classify   explicitly reported pathways into dechlorination processes at   rates of 0.90 to 0.99. While many of the attributes used in the   original generalizations were also selected as predictors by the   classification trees, the extra attributes allow identification   of additional dechlorination pathways that can be considered as   candidates for monitoring in future studies.

      Address (URL): http://pubs.acs.org/doi/abs/10.1021/es902109w