Effects of porous confinement on the structural properties of the Gaussian core model

Abstract

We have investigated the structural properties of a fluid in which particles, interacting via soft potentials, are imbibed into a disordered porous structure built up by soft particles. Using a Gaussian potential for all interactions involved, we determine via Monte Carlo simulations and integral-equation theory the fluid–fluid, fluid–matrix, connected, and blocked structure factors. Within the explored range of state parameters, the fluid–fluid structure factors display a distinct pre-peak, which we identify as a fingerprint of the structure of the matrix. We show that this feature at low wave-vectors arises from contributions of blocked correlations to the fluid–fluid structure factor. We argue that a similar feature may be found also in systems in which particles interact via harshly repulsive potentials. On the other hand, the variation of the main peak of the fluid–fluid structure factor resembles the behaviour of the equilibrium Gaussian core model fluid. In particular, the height of the main peak changes non-monotonically with increasing fluid density at fixed matrix density. Finally, we analyse the effect of a mismatch between the temperature of the fluid and the temperature of the matrix on the structural properties of the system.

Keywords:

• liquids
• porous media
• soft matter
• replica Ornstein–Zernike equations
• Monte Carlo simulation

Acknowledgements

Financial support by the Austrian Research Fund (FWF) under Proj. Nos. P17823-N08, P19890-N16 and W004 is gratefully acknowledged.

Notes

Notes

1. In the generalised percolation approximation, however, the density of the reference equilibrium fluid is , where is the Henry constant 5. For a QA hard-sphere system, H measures the fraction of volume available to the centre of a fluid particle adsorbed in a matrix of density ρ _m.

2. The presence of a slowly decaying ‘self-term’ in the blocked structure factors determined from simulations can be seen by inserting the expression of the one-particle density [see Equation (12)] into Equation (15). Contributions to the product in Equation (15) coming from equal particles at equal time steps give rise to a term equal to 1/M in , where M is the number of configurations used for averaging. Moreover, determined from simulations is affected by statistical uncertainty, which shows up as a positive background noise in . These features prevent [as defined in Equation (15)] from decaying fully to zero at very large k.

3. Remember that is the temperature of the equilibrium fluid from which we obtained the matrix configurations, as described in Section 2.