Complementary use of simulations and molecular-thermodynamic theory to model micellization

Molecular-thermodynamic descriptions of micellization in aqueous media can be utilized to model the self-assembly of surfactants possessing relatively simple chemical structures, where it is possible to identify a priori what equilibrium position they will adopt in the resulting micellar aggregate. For such chemical structures, the portion of the surfactant molecule that is expected to be exposed to water upon aggregate self-assembly can be identified and used as an input to the molecular-thermodynamic description. Unfortunately, for many surfactants possessing more complex chemical structures, it is not clear a priori how they will orient themselves within a micellar aggregate. In this paper, we present a computational approach to identify what portions of a surfactant molecule are hydrated in a micellar environment through the use of molecular dynamics simulations of such molecules at an oil/water interface (modeling the micelle core/water interface). The local environment of each surfactant segment is determined by counting the number of contacts of each segment with the water and oil molecules. After identifying the hydrated and the unhydrated segments of the surfactant molecule, molecular-thermodynamic modeling can be performed to predict: (i) the free-energy change associated with forming a micellar aggregate, (ii) the critical micelle concentration (CMC), and (iii) the optimal shape and size of the micellar aggregate. The computer simulation results were found to be sensitive to the atomic charge parameters utilized during the simulation runs. Two different methods of assigning atomic charges were tested, and the computer simulation and molecular-thermodynamic modeling results obtained using both sets of atomic charges are presented and compared. The combined computer simulation/molecular-thermodynamic modeling approach presented here is validated first by implementing it in the case of anionic (sodium dodecyl sulfate, SDS), cationic (cetyltrimethylammonium bromide, CTAB), zwitterionic (dodecylphosphocholine, DPC), and nonionic (dodecyl poly(ethylene oxide), C12E8) surfactants possessing relatively simple chemical structures and verifying that good predictions of CMCs and micelle aggregation numbers are obtained. In the case of C12E8, the challenges and limitations associated with simulating a single, polymeric E8 moiety at the oil/water interface to model its behavior at the micelle/water interface are discussed. Subsequently, the combined modeling approach is implemented in the case of the anionic surfactant 3-hydroxy sulfonate (AOS) and of the nonionic surfactant decanoyl-n-methylglucamide (MEGA-10), which possess significantly more complex chemical structures. The good predictions obtained for these two surfactants indicate that the combined computer simulation/molecular-thermodynamic modeling approach presented here extends the range of applicability of molecular-thermodynamic theory to allow modeling of the micellization behavior of surfactants possessing more complex chemical structures.