Representation and Validation of Liquid Densities for Pure Compounds and Mixtures Vladimir Diky,* ,† John P. O’Connell, ‡ Jens Abildskov, § Kenneth Kroenlein, † and Michael Frenkel † † Applied Chemicals and Materials Division, National Institute of Standards and Technology, Boulder, Colorado 80305-3337, United States ‡ Department of Chemical Engineering, University of Virginia, 102 Engineers’ Way, Charlottesville, Virginia 22904-4741, United States § CAPEC-PROCESS, Department of Chemical and Biochemical Engineering, Building 229, Technical University of Denmark, 2800 Kgs. Lyngby, Denmark ABSTRACT: Reliable correlation and prediction of liquid densities are important for designing chemical processes at normal and ele- vated pressures. A corresponding-states model from molecular theory was extended to yield a robust method for quality testing of experi- mental data that also provides predicted values at unmeasured conditions. The model has been shown to successfully represent and validate the pressure and temperature dependence of liquid densities greater than 1.5 of the critical density for pure compounds, binary mixtures, and ternary mixtures from the triple to critical temperatures at pressures up to 10 6 kPa. The systems include the full range of organic compounds, including complex solutions, and ionic liquids. Minimal data are required for making predictions. 1. INTRODUCTION The doubling every 10 years 1 of published experimental ther- mophysical property data provides significant societal benefits, but the associated rate of published erroneous data is alarming. It is estimated that about 30 % of the articles reporting new thermophysical-property data obtained from direct measure- ments contain erroneous information in numerical values, metadata associated with these values, or their uncertainties. 2 As this information propagates through the community’s data infrastructure, it can lead (and has led) to significant damage 1 as it is fed into data-driven analyses and is used as the basis for new correlations and prediction methodologies. To avoid such effects, data validation methods designed to identify questionable data and prevent the use of erroneous data in modeling and engineering applications must be developed. Such data validation should be considered a major component of critical data evaluation to increase the “value” of the data first to the level of information (generating combined expanded uncertainties 3 for every data point to be processed), then to the level of knowledge where thermodynamic consis- tency between all related properties is enforced, and finally to the level of wisdom where critically evaluated thermophysical- property data are combined with other relevant data to support major engineering applications such as chemical product and process design. 4,5 A broad variety of data validation methods have been developed, tested, and incorporated into the ThermoData Engine software (TDE) 6 for thermophysical properties of pure compounds, binary mixtures, and ternary mixtures 7-9 as a part of the implementation of the dynamic data evaluation concept 10 within a framework of a Global Information System in Thermo- dynamics. 1,11 Volumetric properties are a major category of thermophysical- property information within the context of any dynamic data evaluation such as the TDE evaluation for thermophysical and thermochemical properties, along with phase boundary property information and energy-related property informa- tion. 10 Each category can be evaluated systematically in order to accumulate understanding about the behavior of a chemical system in thermodynamic space. Historically, there seems to be no rigorous density validation method available, and analysis of data quality for volumetric properties has been limited to the elimination of outliers identified via direct comparison of independent data sets. In TDE, density data are represented based on their association with phase and composition (for mixtures) information as a function of temperature and, for state points away from phase boundaries, pressure. Prior to this work, saturation-state densities for pure compounds have been represented by specialized polynomial expressions 10 and the pressure dependence of single-phase liquid densities given by the Tait equation. 12 The Span-Wagner equation of state 13-15 based on the Helmholtz energy and covering all fluid phases (including the supercritical region) could be fitted at a user’s Special Issue: Memorial Issue in Honor of Anthony R. H. Goodwin Received: June 9, 2015 Accepted: July 27, 2015 Article pubs.acs.org/jced © XXXX American Chemical Society A DOI: 10.1021/acs.jced.5b00477 J. Chem. Eng. Data XXXX, XXX, XXX-XXX