293f Carbon Replica from Silica Template Using Grand Canonical Monte Carlo Simulations

Surendra K. Jain1, Keith E. Gubbins1, and Roland J.-M. Pellenq2. (1) Department of Chemical Engineering, North Carolina State University, North Carolina State University, Raleigh, NC 27695, (2) CNRS, Campus de Luminy, Centre de Recherche en Matière Condensée et Nanosciences, Marseilles, 13288, France

We present a Grand Canonical Monte Carlo study to obtain replica mesoporous carbons from a silica template. Many experimental studies have been reported on obtaining such replica carbons using different templating materials, such as SBA-15, MCM-41, zeolites etc. The formation of these templated carbons is a two-fold process. A carbon rich precursor in vapor or liquid phase is first deposited inside the pores of the matrix template. The adsorbed precursor is then carbonized by heating at a high temperature in an oxygen free environment to produce a carbon-matrix composite. Finally, the matrix is removed by treating with an acid and the carbon replica is obtained.

Apart from the experimental work, very few simulation studies has been reported on modeling these carbon replicas. In a recent work Roussel et al [1] used GCMC simulations to obtain carbon replica from zeolites, by adsorbing carbon vapor in the zeolite pores. They used a Tight binding formalism to model the carbon-carbon interaction. In this work we follow a similar method to develop molecular models of the carbon replica of nanoporous oxides like MCM-48 and SBA-15 by adsorbing carbon vapor in the pores of the template matrix using Grand Canonical Monte Carlo simulations. The carbon-carbon interaction is modeled using the bond order potential developed by Brenner [2]. The matrix-carbon interaction is modeled using the PN-Traz potential [3]. We further study the stability of the resultant carbon replica by removing the matrix template and relaxing the final carbon replica using molecular dynamics simulations. The structure of the resultant carbon replica is investigated by calculating the carbon-carbon pair correlation functions, neighbor distribution, bond angle distribution and ring statistics.

[1] Roussel T., Bichara C., Pellenq, Adsorption 2005 ; 11: 709. [2] D. W. Brenner, Phys. Rev. B 1990; 42(15):9458. [3] R. J.-M. Pellenq and D. Nicholson, J. Phys. Chem. 1994; 98:13339.