Configurational-bias Monte Carlo simulations in the isobaric-isothermal ensemble and in the canonical and isobaric-isothermal versions of the Gibbs ensemble were carried out to investigate the thermophysical properties of a 2-methoxy-1-(2-methoxypropoxy)propane/2-methoxy-1-((1-methoxypropan-2-yl)oxy)propane mixture and the mutual solubility of this dipropylene glycol dimethyl ether (DPGDME) mixture and water. The TraPPE-UA and TIP4P force fields were used for the DPGDME isomers and water, respectively. The TraPPE-UA force field yields satisfactory predictions for the specific density of the liquid phase and for the normal boiling point of the DPGDME mixture. Preliminary simulation data for the mutual solubility of the DPGDME mixture and water at T= 298 K pointed to an underprediction of the miscibility gap and, hence, simulations were carried out for variations of the TraPPE-UA DPGDME model with different partial charges for the ether moiety. Both the organic and the aqueous phase exhibit pronounced microheterogeneity and the persistence of very stable and large aggregates poses significant sampling challenges. The trajectories of independent simulations started from different initial compositions indicate either (i) that equilibrium has not yet been achieved, (ii) that composition fluctuations are sampled insufficiently over the current trajectory lengths, or (iii) that the system size is too small. Based on these preliminary simulation data, cautious estimates of the mutual solubilities at temperatures from 283 to 353. K are made using a model with modified charges. These predictions yield mean signed deviations from the experimental benchmark data for the mutual solubilities of +5% and -4% for the organic and aqueous phases, respectively.
Bibliographical noteFunding Information:
We thank Jake Rafferty for helpful discussions, Tom Hoye for help with the preparation of Fig. 1 , and Jonathan Moore for organizing the 6th Industrial Fluid Properties Simulation Challenge. Financial support from the National Science Foundation ( CBET-0756641 ) and a Graduate School Fellowship (P.B.) is gratefully acknowledged. Part of the computer resources were provided by the Minnesota Supercomputing Institute.
- Aqueous solution
- Gibbs ensemble
- Liquid-liquid equilibria
- Monte Carlo simulation