Axel Drefahl - Extraction and Application of Environmentally Relevant Chemical Information from the ThermoML Archive

Version 1

      Publication Details (including relevant citation   information):

      EnviroInfo 2007 21st International Conference on Informatics   for Environmental Protection September 12-14, 2007, Warsaw,   Poland. Pages 71 to 78 in Volume 1: Plenary session papers.   Shaker Verlag GmbH, D-52018 Aachen, 2007.


      The ThermoML Archive supports an   open-source approach to science. The XML-structured plain-text   files in this archive provide thermodynamic data for chemical   compounds and mixtures abstracted from recently published   articles. We introduce a Web-based Property Viewer that displays   all currently available property data for user-selected pure   compounds. Extraction and transformation of ThermoML data using   Python-based scripts and JavaScript is discussed. We demonstrate   the general use of the Property Viewer and in particular discuss   access to data on organic salts such as ionic liquids. The   Property Viewer is an easy-to-use tool assisting comparison of   “conventional” solvents and those labelled as “green”. The   ThermoML Archive provides a rich source of data on   multi-component systems. We demonstrate access to information on   environmentally important (water + chemical)   systems.

      This paper emphasises the   request-to-target precision achieved by applying the Document   Object Model (DOM)  guided matching while inspecting   ThermoML-structured files. Such a precision cannot be obtained by   text-based search-key matching since the chemical and   environmental literature in general does not confine to unique,   unambiguous terms, neither for chemical names nor for terms of   properties and their associated units. The DOM-based granularity   and annotation of the chemical information in the ThermoML   Archive efficiently supports the design of software for customer   specified applications such as screening and selective data   extraction. As an example, we evaluate a published atom   additivity method for the estimation of molecular polarisability,   a critical molcular descriptor in ecotoxicological modelling, by   extracting from the archive new experimental mass density and   refractive index data not used in the original development of   that method.

      We forecast that the ThermoML Archive   will not just serve as an outstanding source for chemical   properties and risk assessment but will support eco-integrated   chemical design and thus advance future strategies in    sustainable chemistry.

      Address (URL):