Basic Research Needs for Geosciences - Energetics Meetings and ...

GRAND CHALLENGE: COMPUTATIONAL THERMODYNAMICS OF COMPLEX FLUIDS AND SOLIDSslightly complex systems (Gubbins 1985). It is surprising that even for common systems (e.g.,CO 2 -H 2 O-CH 4 , Al 3+ -H + -H 2 O-salt), sufficient data do not exist to develop an accurate EOS. Forexample, for systems containing aluminum, the most common metal in the earth’s crust and anubiquitous component of natural waters, there is insufficient information to even uniquely definethe speciation in solution (e.g., the formation of polynuclear hydroxide species, etc.) (Baes andMesmer 1986). This is also the case for solutions containing other charged ions important toDOE applications, such as Fe 3+ , UO 2 2+ , etc. Without this knowledge, the development of reliablethermodynamic models with good interpolation and extrapolation properties is not possible. Forthese systems the likely contribution of simulation methods in the next generation will be tounderstand the relative stability, composition and structure of species (e.g., ion pairs, metal ionpolymers) in the solution. This is an essential step in the development of an EOS that reproducesthe behavior of the system with near-experimental accuracy and can be parameterized from thelimited data likely to be available.For closed shell uncharged molecules (e.g., CO 2 , H 2 O, O 2 , N 2 , CH 4 ), it should be possible todevelop molecular dynamics models that can calculate phase equilibrium and thermophysicalproperties from first principles. For these systems it would be worthwhile to invest in thedevelopment of very high accuracy interaction models at the molecular level (i.e., for CO 2 -H 2 O,CO 2 -CH 4 interactions; see Cao et al. 2001). These could then be used to simulatethermodynamics for these systems via molecular dynamics and Monte Carlo methods. While thisis not a difficult simulation, the magnitude of the problem grows quickly when the desired resultis thermodynamic data, especially free energy data. Even for these systems the scaling of presentday simulation algorithms is not good (less than 200 processors) (Fitch et al. 2006). It isimportant to develop highly scalable simulation algorithms, including those that directly simulatefree energy for future implementation on next-generation computers.78 Basic Research Needs for Geosciences: Facilitating 21 st Century Energy Systems