2010 Annual Report
1a.Objectives (from AD-416)
Overall project objectives include: Collecting the best available information on the structures within the cotton fiber; Constructing fundamental models of these structures at different size scales; Providing additional fundamental models that have partial surfaces of hydrophobic molecules; and Monitoring moisture movement through the model structure during molecular dynamics simulations. Specific aspects of this work to be carried out by the CCRC include development of improvements of atomistic models of cotton cellulose and their interaction with water.
1b.Approach (from AD-416)
The main feature of this effort will be to examine the abilities of an improved version of the water model (TIP5P) and a parameter set that utilizes explicit lone pairs of electrons. The models will be examined for reproduction of known properties such as unit cell dimensions and retention of the geometric features of the cellulose structure. Distinctions in results between molecular dynamics and energy minimization will be explored. Necessary adjustments will be made to the modeling parameters to develop a final set. A model involving several smaller crystals will be devised to mimic a portion of a cotton fiber, and the behavior of interacting water will be observed. An ultimate goal, possibly not practical within the scope of this project, will be to gain insight on the division between “bound water” and “bulk water” and the sensitivity of the results to the exact details of the simulation. Deviations from the above plan are not necessarily expected but can be accommodated. For example, water models other than TIP5P might be more helpful, or other force field modifications other than the addition of explicit lone pairs may be more useful.
The University of Georgia (UGA) effort is directed to solving a problem observed in modeling cellulose crystals, namely, a twist occurs when the GLYCAM modeling force field developed by the University of Georgia group (and used by the Tulane group) acts on the model crystal. (GLYCAM was created to deal with the special problems of carbohydrate modeling.) Normally, crystals are not twisted. The initial hypothesis was that the hydrogen bonding energy is modeled in a simple way, and that a more complex evaluation might improve the calculations to the point where the crystal did not twist. A prototype force field, GLYCAM-LP, which incorporates lone electron pairs (LP), was used, as was a more complex water model. However, twisting still happened. Therefore another possible cause, the lack of polarization in GLYCAM, has been investigated by “pre-polarizing” the models. This involves reassessment of the atomic charges based on their neighbors in the solid state rather than as parts of isolated molecules in vacuum. A graphics processor-based computer was implemented by Professor Wood’s group at the University of Georgia for the work, and the GROMACS molecular modeling software was implemented on it to speed the calculations. The UGA work was presented at the Anselme Payen Award Symposium at the American Chemical Society Meeting in San Francisco and sparked good discussion.
The methods used to monitor activities for this agreement were annual reports, technical visits/e-mails/interactions, presentations at scientific and industry meetings, and publications.