Critical constants by simulation.

April 12, 2011 6:02 PM

I am wondering if there are generally accepted ways of getting the critical constants (acentric factor, critical pressure and temperature) of a substance from a simulation of some sort (i.e. quantum chemistry)

I just finished taking a chemical engineering course that focused on various models and equations of state for gases and liquids and how to use them to solve chemical engineering problems. The equations of state all depend on the acentric factor, critical pressure, and critical temperature.

Previously I've taken an introduction to computational chemistry course that focused on using ab initio methods to determine properties of a substance.

I'm wondering if there is some straightforward way of running a few simulations and exploiting some property of the critical point to get the critical constants.

posted by selenized to science & nature (3 answers total)

I'm a fellow ChemE and recently stumbled across the Joback Method, which can be used to approximate critical temperature and pressure based on the functional groups in your compound.
posted by rancidchickn at 6:39 PM on April 12, 2011

I think "generally accepted" is the weak link here. I've had people balk at the notion of using Henderson Hasselbalch to figure out what to mix to make a buffer because it was "just theoretical".
posted by Kid Charlemagne at 6:52 AM on April 13, 2011

rancidchicken The Joback Method seems pretty close to what I was thinking of, though I was thinking more in terms of results from ab initio calculations as opposed to counting the functional groups.

Kid Charlemagne I don't think the existence of people that would balk at using anything theoretical is enough to sink any notions of a "generally accepted" method. For example: many process simulators use models like UNIFAC to calculate activity coefficients for mixtures. This is really a mix of hand-waving theoretical stuff and empirical correlations. Is this *universally* accepted as the best way? No. Is it *generally* acceptable? I would say that since it has become the norm for calculations of this sort, yes.

Probably most of the practical calculations of chemical engineering would make both a die-hard empiricist and a die-hard theoretician scream and tear out their hair.
posted by selenized at 11:34 AM on April 13, 2011