SMF - Just Installed!

Main Menu

Calculation of adsorbed liquid mixture

Started by Fan Honglei, October 14, 2021, 04:11:33 AM

Previous topic - Next topic

Fan Honglei

Hello, Dr. Dubbledam,
     I learned that when calculating the adsorbate in the adsorbed gas phase, when it exceeds its saturated vapor pressure, the adsorption value will become negative.
     I refer to the literature and know that, for example, if I want to calculate the mixture of alcohol and water in the adsorbed liquid phase, I need to set the configuration biased Monte Carlo and calculate the pressure value of the gas-liquid equilibrium of alcohol and water.

My doubts are the following:
1. Is the pressure value when calculating the gas-liquid phase equilibrium (using Aspen software) the pressure value of the input file in Raspa.

2. Is the component ratio of alcohol and water in the input file the liquid phase component ratio in gas-liquid equilibrium? (there is also a gas phase component ratio of alcohol and water in gas-liquid equilibrium)

3. To enable configuration biased Monte Carlo, do I need to write CBMCprobability 1 under the components of adsorbate in the input file? Or do I need to write other parameters?

4. How to solve the problem of using Raspa to calculate the adsorbed liquid phase?

I consulted many literatures about calculating the adsorbed liquid phase, but I was confused about how to operate it in Raspa software.   

David Dubbeldam

Note sure I fully understand. There is no such thing as "adsorbed liquid phase", there is only an adsorbed phase. So, do you mean a Vapor-liquid simulation or a binary mixture adsorption simulation?

1) Note sure what Aspen does, but in VLE simulation pressure is an output property, and an input property for adsorption simulations. In the latter, you input the total pressure and the mol-fraction.

2) In adsorption simulations in the grand-canonical ensemble you input the pressure and the mol-fraction of the fluid phase. The imaginary reservoir and the adsorbed phase are in equilibrium and for these conditions you compute the loading of each component in the framework.

3) Moves like "SwapProbability" are already biased. You would use "CBMCProbability" (which is also "PartialReinsertionProbability") to specify partial-reinsertion moves.

4) Look at the binary mixture adsorption in the examples and adapt it for you system.