SPEADMD, September 2011

J. Richard Elliott

Video Tutorials

SPEADMD is a molecular dynamics simulation methodology for physical property estimation. The acronym stands for Step Potentials for Equilibria And Discontinuous Molecular Dynamics. This section of the help menu provides links to video tutorials that discuss the background of the methodology and show you which buttons to push to get results. The first video (27 min) discusses the physical basis of the molecular simulation model and how it can be used to derive physical properties. The second video (11 min) shows how to draw a picture of a molecule and use it to estimate properties.

Brief Introduction and Bibliography

SPEADMD is a molecular dynamics simulation methodology for physical property estimation. A more detailed description is given in SpeadIntro. This page is intended to provide a very brief linkage to the available information. The success of this molecular model hinges on three fundamental premises.
(1) The influence of repulsive forces dominates the physical properties. For example, the intermolecular distributions and their fluctuations are primarily influenced by how closely the atoms can approach each other. As another example, the entanglements that strongly influence transport properties occur because molecules cannot pass through each other, but must find a viable path for wriggling past each other.
(2) Repulsive effects are specific to the 3D structure of a molecule, necessitating molecular simulation of that specific molecule. Branching, rings, steric hindrance, and intramolecular proximity are all important molecular features with significant macroscopic implications for phase equilibria and transport properties.
(3) Thermodynamic Perturbation Theory (TPT) is sufficiently accurate that a quantitative treatment of the attractive details of the potential can be derived from theory. Since the TPT contributions are directly related to the atomistic distributions and their fluctuations, and these are accurately determined by the repulsive forces, this means there is no need to repeat the simulation for every possible specification of the attractive part of the potential. The parameterization of the attractive part of the potential can therefore be pursued in the manner of an engineering equation of state.

We refer to this methodology as "SPEADMD," abbreviating Step Potentials for Equilibria And Discontinuous Molecular Dynamics. The developments to date can be best understood by a brief demonstration through the website (~10 minutes). The demonstration is divided into three parts: (1) the graphical user interface for defining the ".mol" file, (2) the initialization interface specifying and submitting the ".m3d" file (3) the "fluid properties" interface for translating the simulation results into a customized equation of state and analyzing phase equilibria and transport properties. Instructions for conducting the demonstration are given in the help menus. Just select the “Critical Properties” link to get started.

Detailed descriptions of the methodology are described in the literature. An overview is given in SpeadIntro. SpeadIntro documents the fundamentals of perturbation theory and discretization of the potential, including comparison of the prediction of TPT in relation to simulating the full potential. Briefly, if the TPT prediction agrees quantitatively with the full potential simulation, and it is roughly 100 times faster, then why simulate the full potential? SpeadIntro also reviews the publications and typical results that have been obtained with figures and tables in more detail than the list of titles below. If you are especially interested in the fundamentals of Newton's laws of motion and discontinuous molecular dynamics, follow the "DMD" link at the top of the home page.

A brief bibliography is given below. These articles review the coverage of molecular types including: n-alkanes, branched alkanes, alkenes, alkynes, aromatics, naphthenics, alcohols, amines, amides, esters, nitrates, phosphates, halocarbons, perfluorocarbons, acids, sulfides, and thiophenes.

We have performed many simulations to characterize molecular interactions. The current database covers about 500 compounds. These are listed on the "Fluid Properties" link. Our results for vapor pressure (~10% average absolute deviation, %AAD) are quite good compared to other methods that are based solely on molecular structure.(cf. 4,5,7,9,14,15,16) The temperature range over which our results are applicable is important. In general, DMD/TPT is accurate to reduced temperatures of 0.45 whereas other united atom models are inaccurate below reduced temperatures of 0.6. Typical engineering equations of state like the Peng-Robinson equation are accurate to reduced temperatures of 0.45, so it is important to extend to lower reduced temperatures.  Most group contribution methods that are focused on the low pressure ranges (10-760 mmHg) and their accuracy at higher pressures is less known. (cf. 5,7) Vapor pressure accuracy is very important for phase equilibrium predictions, especially VLE (cf. 4,6,9,11).  For liquid density, our accuracy is roughly 3%AAD.  This compares to ~1.5%AAD for other molecular simulation models.  We are working to improve our accuracy for density by applying a non-transferable volume shift when any single density data have been measured and by implementing White's method for the critical region.  For diffusivity, our accuracy at present is ~25%AAD.  We are at an early stage of correlating transport properties for engineering applications. (cf. 3). The methodology can also be extended to polymers by inferring the asymptotic trends in the TPT terms in the long chain limit. These trends are characteristic of individual polymer architectures. (cf. 1, 13).  

Reference:

1.       Evaluating Perturbation Contributions in SAFT Models by Comparing to Molecular Simulation of N-Alkanes, Ahmad F. Ghobadi, J. Richard Elliott, Fluid Phase Equilibria, 305:57-66 (2011).

 

2.      'Historical Perspective and Current Outlook for Molecular Dynamics As a Chemical Engineering Tool,' Edward J. Maginn and J. Richard Elliott, Ind. Eng. Chem. Res., 49:3059-3078 (2010).

 

3.      Self-Diffusivity Estimation By Molecular Dynamics,’ Z. Nevin Gerek and J. Richard Elliott, Ind. Eng. Chem. Res., 49:3411-3423 (2010).

 

4.      Transferable Intermolecular Potentials and Equations of State for Carboxylic Acids and Their Phase Behavior,’Amir Vahid, J. Richard Elliott, AIChE J., 56:485-505 (2010).

 

5.      ‘Finitely Limited Group Contribution Correlations for Boiling Temperatures,’ Fateme Sadat Emami, Amir Vahid, J. Richard Elliott, Farzaneh Feyzi, J. Chem. Thermo., 31:530-537 (2009).

 

6.      ‘Correlation Of Mixture Vapor-Liquid Equilıbria With The SPEADMD Model,’Amir Vahid, Amanda D. Sans, J. Richard Elliott, Ind. Eng. Chem. Res., 47:7955-7964 (2008).
http://pubs.acs.org/cgi-bin/download.pl?ie800374h/L6Za

 

7.      ‘Group Contribution Prediction of Vapor Pressure with SAFT, PC-SAFT and ESD Equations of State,’ Fateme Sadat Emami, Amir Vahid, J. Richard Elliott, Farzaneh Feyzi, Ind. Eng. Chem. Res., 47:8401–8411 (2008). http://pubs.acs.org/cgi-bin/download.pl?ie800329r/A7on

 

8.     ‘Inferring Transferable Potential Models,’ Sinan Ucyigitler, Mehmet C. Camurdan, Metin Turkay, J. Richard Elliott, Molecular Simulation., 34:147-154 (2008).

 

9.      ‘Transferable Potentials for Perfluorinated Molecules,’ Amanda D. Sans, J. Richard Elliott, Fluid Phase Equilibria, 263:182-189 (2008).

 

10.  Combining Molecular Dynamics and Chemical Process Simulation: The SPEAD Model’ AsiaPacific J. Chem. Eng., 2:257-271 (2007).

 

11.   ‘Transferable Potentials for Alcohol-Amine Interactions,’ J. Richard Elliott, Amir Vahid, Amanda D. Sans, Fluid Phase Equilibria, 256:4-13 (2007).  

 

12.  ‘Molecular dynamic simulations and global equation of state of square-well fluids with well-widths from 1.1 to 2.1,’ Sergei B. Kiselev, James F. Ely, J. Richard Elliott, Mol. Phys, 104:2545-2559 (2006).

 

13.   Asymptotic Trends in Thermodynamic Perturbation Theory,’ J. Richard Elliott and Neil H. Gray, J. Chem. Phys, 123:184902 (2005).

 

14.   Transferable Step Potentials for Amines, Amides, Acetates, and Ketones,’ Suhan Baskaya, Neil Gray, Z. Nevin Gerek, and J. Richard Elliott, Fluid Phase Equilibria, 236:42-52 (2005).

 

15.    Molecular Modeling of Isomer Effects in Naphthenic and Aromatic Hydrocarbons,’ Neil Gray, Z. Nevin Gerek, and J. Richard Elliott, Fluid Phase Equilibria, 228-229C, 147-153 (2005).

 

16.  ‘Transferable Step Potentials for the Straight Chain Alkanes, Alkenes, Alkynes, Ethers, and Alcohols,’ Ozlem Unlu, Neil Gray, Z. Nevin Gerek, and J. Richard Elliott, Ind. Eng. Chem. Res., 43:1788-1793 (2004).

 

17.    ‘Phase Diagrams for Multi-step Potential Models of n-Alkanes by DMD/TPT,' J. Cui and J.R. Elliott, Jr., J. Chem. Phys., 116:8625 (2002).

 

18.  ‘Optimized Step Potential Models for n-Alkanes and Benzene,' J.R. Elliott, Jr., Fluid Phase Equilibria, 194:161 (2002).

 

19.   ‘Phase Envelopes For Variable Width Square Well Chain Fluids,' J. Cui and J.R. Elliott, Jr., J. Chem. Phys., 114:7283 (2001).

 

20. ‘Vapor Liquid Equilibria of Square-Well Chains,' L. Hu, H. Rangwalla, J. Cui, J.R. Elliott, Jr., J. Chem. Phys., 111:1293 (1999).

 

21.  ‘Vapor Liquid Equilibria of Square-Well Spheres,' J.R. Elliott, Jr. and L. Hu, J. Chem. Phys., 110:3043 (1999).