Gibbs Reactor: Solution Failed

[deliobs]

I'm trying to perform a simulation using Gibbs Reactor Unit Operation for the following reactions

1: (c6h5)3CCl(l) + CH3OH (l) ----> (c6h5)3COCH3(l) + HCl(l)
2: (c6h5)N(l)+ HCl(l) ----> (c6h6)NHCl(s)

For assurance a system in liquid state, I pressurized the feed and under 10 atm and the inlet temperature is 25 C. The main reaction is the number 1, but number 2 also happened due to system dilution and my troubles are laying on that reaction. After run the simulation I'm getting the following errors from the simulation report as follows

+++ specifications +++

Temperature: 317.417470643401 K
Pressure drop: 0 Pa
Tolerance: 0.0001
Maximum iterations: 1000

+++ no solution +++

Solution failed: Failed to calculate entropyF, enthalpyF and their derivatives: CalcSinglePhaseProp failed: Property evaluation using equation of state failed: Missing critical properties for one or more compounds

This problem belongs to Example 5-1. Scoot Fogler, Kinetic and Chemical Engineering Reactions (3rd edition)

[jasper]

The error message in your post seems to suggest you are running the reactor in the liquid or vapor phase, which is not an option with solid-only compounds in there.

The way the problem is attached, the Gibbs reactor is configured to solve for the solid phase, which is not an option because you have non-solid compounds in there.

TEA has only very limited support for solid-only compounds. They can only be used in a solid phase, and may not appear in a liquid or vapor phase. All other compounds on the other hand may not appear in the solid phase. Hence, you cannot mix these in a single phase. And the Gibbs reactor is a single phase reactor.

I suppose what you are after is minimize Gibbs energy at phase equilibrium? This would put solid only compounds in a solid phase and other compounds in vapor and/or liquid. The Gibbs reactor will not do that however - you'd need to make a custom unit operation for that (e.g. Matlab, Scilab, Excel, Python, C++, ...) and perform the minimization (note: you need formation terms in there so use enthalpyF-T*entropyF for Gibbs energy as GibbsEnergy may or may not contain formation terms, and at phase equilibrium you will need to sum this value, multiplied by phase fraction, over all phases present at phase equilibrium) but be aware that a multiphase Gibbs minimization problem is not per se that straight forward due to local minima.

[deliobs]

That what's really happening, so what if I assigned for those compounds' supposed to be solid anyother classification? Would it work?

For example, I know that there will be a solid precipitate from the product of the second reaction, but I won't set it up as solid-only compound instead of I will classify it in some other way else. I also know that its melt temperature is under the reactor's temperature.

[jasper]

If you mark these compounds as not-solid-only they will take part in the vapor liquid equilibrium and in the case of the Gibbs (single phase) reactor, you can calculate liquid phase properties that include these compounds. But then you must supply some details for these compounds that allow for calculation of these properties (for equations of state, typically the critical properties, or for activity models, typically a vapor pressure curve and interactions with other compounds).

"Will it work" is of course a different question to "is it realistic" - in particular if you use a Gibbs reactor that tries to evaluate chemical equilibrium exactly from how you define your chemical potentials (fugacity, activity, ...)