Lecture 1 - Introduction

Download as pdf or txt
Download as pdf or txt
You are on page 1of 38

Lecture Note for Molecular Modeling and Simulation

Lecture 1. Introduction

Jeong Won Kang


Department of Chemical and Biological Engineering
Korea University
Syllabus

◆ Text Book
0. Atkins “Physical Chemistry”
1. D. Frenkel and B. Schmit, "Understanding Molecular Simulation",
Academic Press, San Diego (1996)
2. M. P. Allen, D. J. Tildesley, "Computer Simulation of Liquids",
Clarendon, Oxford (1987)
◆ References
1. C. J. Cramer, "Essentials of Computational Chemistry",John Wiley &
Sons, New York (2002)
2. T. L. Hill, "An Introduction to Statistical Thermodynamics", Addison-
Wesley, London (1960)
Syllabus
◆ Computer Software
– Material Studio
◆ Programming Language
– C / C++ or FORTRAN, Python, Mathlab
◆ Exam
– End of the semester
◆ Small Project
– MD simulation of 2D/3D sphere
◆ Grade
– Exam (60 %), Project (20 %), QUIZ and Attendances (20%)
Overview of the Course
Lecture 1. Introduction
Lecture 2. Basic Statistics
Lecture 3. Introduction to Statistical Mechanics
Lecture 4. Model System – Force Field
Lecture 5. Molecular Monte Carlo Simulation
Lecture 6. Molecular Dynamics Simulations
Project: MD Simulation of 2D Sphere, 3D Sphere
Lecture 7,8 – Software Exercises
Lecture 9. Quantum Mechanics
Lecture 10. Molecular Modeling -1
Lecture 11. Molecular Modeling -2
Final Exam
Difficulties in Learning Molecular Modeling and Simulation

◆ Principles in various fields are required…


– Basic Science
• Statistics, Mathematics
• Statistical Mechanics, Statistical Thermodynamics
• Quantum Mechanics
– Applications
• Chemistry and Biology
• Thermodynamics
• Reaction Engineering
• Material Sciences
– Software and Hardware
• Gaussian, Hyperchem, Accerlys Products,…
• Programming Language : C/C++ or FORTRAN, Python
• Parallel Computing, Distributed Computing
Introduction – Motivation

◆ Why we should learn molecular modeling and


simulation ?

Understanding the nature of molecules are


ultimate goal in many science and
engineering fields !

Reaction
Biology
Engineering

Chemistry Thermodynamics Material Science


Introduction - Motivation

◆ Molecular Modeling and Molecular Design

Given structure Molecular Find properties


Modeling
and
Simulation
Toolbox

Molecular Design Given Properties


- Ultimate Goal for Design of Molecules
Find molecular structure
What is molecular modeling and simulation ?

◆ Molecular Modeling (Computational Chemistry)


– Computer modeling at molecular level
– Quantum Mechanics
– Modeling of single or fewer molecules
– Interested in electronic structure/energy surface/interaction/….
– Numerical solution to Schrödinger equation (QM)
– Computationally intensive (QM)

◆ Molecular Simulation
– Properties calculation based on many-body simulation
– Interested in bulk properties (Average of may molecules)
– Two methods are well-known
• Molecular Monte-Carlo Simulation (MC)
• Molecular Dynamics Simulation (MD)
– Two methods require force field as input
– The force field can be estimated using Molecular Modeling Technique
Molecular Modeling and Simulation

Molecular Specification Electronic Structure Properties


Pressure Plot

O 28

O 26

O 24

Pressure
22

20

18

N 16

14
0 200 400 600 800 1000 1200

Time

Shroedinger Equation Force Field (FF) Phase Equilibria

Molecular Modeling
(Computational Molecular Simulation
Quantum Chemistry) Monte Carlo
Ab-initio Molecular Dynamics
HF, DFT
Structure and Thermodynamic Properties
Energy of individual molecules Transport Properties
Molecular Modeling and Simulation
◆ Computer Modeling/Simulation become a general research tools
– Many major journals in science and engineering fields
– No. of papers are increasing
◆ Understanding the “Black Box” greatly increases the efficiency using it
◆ The technique can be applied to various field of science and engineering
– Chemistry and Physics
– Chemical Engineering
• Thermodynamics
• Transport Phenomena
• Reaction Kinetics
– Bimolecular Engineering
• Protein Structure Prediction
• Protein Binding – Pharmaceuticals design
• Enzyme Kinetics
– Material Science
• Functional Material Design
Motivation of electronic computers

◆ World War II
– US battleships can fire shells weighing as much as small car over distance up to 25 miles (~38 km)
– Physicists could write the equations that describes how atmospheric drag, wind, gravity, … would
determine the trajectory of the shell.
– US military was looking for females (with math majors) to hire the job of computing “firing tables”
– Not enough humans could be found to keep up with the need for new table
– The early success was the Harvard Mark I computer (Harvard and IBM, 1944)
– However, Mark I was not pure electronic computer (mechanical switches, relays, rotating shafts, …)
Origin of “Computer”

Human Computers
History of Using Computer

◆ MANIAC, 1952
– After World War II, scientists seek to use of computer
– N. C. Metropolis was interested in solving broad spectrum of problems on
this machine
– At Los Alamos, in the 1950, a group of researches led by Metropolis,
including John von Neumann and Stanislaw Ulam, developed Monte Carlo
Method.

Metropolis Monte-Carlo
Simulation Method
Molecular simulation B.C. (Before Computer)

◆ Approximate Theories
– Van der Waals equation for non-polar fluids
– Debye-Huckel for electrolytes

◆ Mechanical Simulation
– Plastic foam balls
– Metal bearings
• Tedious, laborious
• Quite realistic ▪ J.D.Bernal’s model of a simple liquid

“ …I took a number of rubber balls and stuck them


together with rods of a selection of different lengths
ranging from 2.75 to 4 inches. I tried to do this in
the first place as casually as possible, working in my
office, being interrupted every five minutes or so
and not remembering what I have done before the
interruption …”
Theory, Model and Simulation

Liquid State Lennard-Jones potential


Gas State Real Hard Sphere potential
Polymer Model Square-well potential
Nano space
Problem

Carry out Construct


Perform
Computer Approximate
Experiments
Simulation Theories

Experimental Exact Result Theoretical


Result for Model Predictions

Van der Waals


Comparison Comparison Debye-Huckel

Test of Test of
Model Theories
Procedure for Molecular Simulation

Statistical
Computer Simulation
Model Averaging
Implementation Result
Method

Interaction Statistical Random Number


Test of Model
Energy model Mechanics Generation
Structural Ensemble Random
Test of Theory
Model Average Walk
Statistical Property
Treatment Prediction
Method of
New Discovery
Integration
Examples of Research Topics
◆ Simple properties estimation
– Heat of reaction, heat of formation
– Properties for complex molecules
– Properties at nano space
◆ Reaction Mechanism Study
◆ Crystal Morphology Estimation
◆ Model parameters estimation
– Critical constants
– Parameters for a specific model
◆ Structure prediction
– Protein folding problem
◆ Computer-Aided Molecular Design (CAMD)
– Group contribution based design
– QM/MM application
Simple Properties Estimation
◆ Heat of reaction / Heat of formation (QM Application)
– G2 Method (Pople et al. J.Physics, 1989)
• A high level ab-initio method (from quantum mechanics)
– Case 1
• NH=CHNH2 + CH3-N=CH-CH3 →CH3-N=CH-NH2 + NH=CH-CH3
• Heat of reaction using Benson Method : 0 kJ/mol Prone to hydrolysis and
decomposition
• Heat of reaction using G2 Method : 1.8 kJ/mol
– Case 2
• Heat of formation for THP (1,4,5,6-tetrahydropropyrimidine)
– DIPPR Project 871 : 13.2 + 0.5 kcal /mol
– G2 Method : 18.9 + 0.1 kcal /mol
» 40 hours on CRAY C-90 computer and 10 GB storage space
• G2 Method is believed to be more accurate !

• Sir John Pople first developed G1 method (Novel Prize Winner in 1998)
• Gaussian-4 theory is now available.
Heat of Formation Tool in Accerlys Software
Properties of alternative molecules
◆ Properties Prediction for alternative refrigerant (HCFCs)
◆ Fermeglia et al. FPE, 2003
– Accurate force field (FF) calculation is most important factor
– FF is important due to complex nature of molecular associations in H-F
bonds

Properties
Molecular Structure Force Field
Phase Equilibria

Ab-initio calculation Mote-Carlo Gibbs Ensemble


Emprical parameterization Simulation

Software : Cerius v. 4.2 from Accelrys Inc.


Hardware : Silicon Graphics Origin 2000
Properties of HCFCs
Critical Property Estimation
◆ Critical constants are normally used for
parameters of cubic EOS.
◆ Panagiotopoulos (1987) proposed a new
method for the simulation of vapor-
liquid equilibrium ( Gibbs Ensemble
Method )
◆ Using Gibbs Ensemble Method, Smit
and coworkers (Siepman et al., 1993)
calculated coexistence diagram for
normal paraffins to Nc = 48.
Critical Property Estimation

Data of Steele (1993)


Data of Ansleme (1990)
Simlulation (Siepman, 1993)

Data of Anselme shows


maximum at Nc= 8 while
data of Steele does not.
Simulation calculations
have settled the
disagreement
Verification of Reaction Mechanism

◆Trans-esterification of Lipase (CALB), 권정훈 2007


◆ Hydrogen bond lengths of the CALB-(±)-1e complex

100
(-) Form
(+) Form
80

Conversion
Conversion Yield (%)

60

yield (%)
40

(-)Form; 66.7%
20

(+)Form; 35.4%
0
0 20 40 60 80

Reaction Time (hr)


Crystal Morphology
Recrystallization acetone + d-HBP ethanol+ d-HBP methanol+ d-HBP

◆ 정정영, 2008 ➢ 1 g of HBP


+
10 ml of solvent

100 ml
beaker

Experimental
Slow evaporation at room
temperature without acetone+ t-HBP ethanol+ t-HBP methanol+ t-HBP
mechanical vibration

Grown crystal was


sampled and taken the
pictures using SEM or a
camera

❖ t-HBP
Eatt,hkl E'att,hkl E'att,hkl E'att,hkl
face
in vacuum in acetone in ethanol in methanol
{1 0 1} -21.29 -33.72 -34.09 -31.41
{1 1 1} -19.16 -71.79 -53.49 -59.73
{1 1 0} -18.93 -26.56 -32.52 -19.94
{1 0 0} -20.83 -25.36 -17.29 -21.66
{0 1 0} -9.36 -12.35 -12.40 -10.64
Molecular Modeling {0 0 1} -9.41 -16.02 -14.49 -7.39

✓ t-HBP from acetone ✓ t-HBP from ethanol ✓ t-HBP from methanol


Protein Structure Prediction

▪ BPTI structure
• X-ray structure
▪ The lowest-energy conformation
• (b) MC simulation annealing in gas phase
• (c) In aqueous solution, by the distance-distance dielectric function
• (d) In aqueous solution, by the solvent-accessible surface area
Molecular Design/ QSAR / QSPR

◆ Two important methods used for product design


– QSAR (Qualitative Structure-Activity Relationship)
– QSPR (Quantitative Structure-Property Relationship)
◆ QSAR have been used in drug discovery
◆ QSPR is very similar to group-contribution approach
in chemical engineering field
◆ Design a candidate molecules those have desired
property / activity
Molecular Design

◆ Drug and Product Design have


been long studied in the field of
pharmaceutical industries and
chemical industries.

◆ Dendrimer as an alternative for


human blood ingredient
(Fermaglia et al. 2002,
Bioorganic and Medicinal
Chemistry)
Molecular Design
◆ Ionic liquids
– Promising candidate for clean solvent
– Almost infinite number of candidate ionic liquid exist due to
the variations in cation-anion pair combinations
– Property measurement / process design is not a primary issue
in this subject.
– The structure-property relationship is the essential problem for
the design of a ionic solution as a specific solvent
COSMOtherm
Conductor-like Screening Model for Realistic Solvation
Questions
◆ Q1 : Are MM/Simulation methods really useful in real
application ?

◆ Q2 : Do research institutes in advanced countries invest


money the development of theoretical MM methods ?

◆ Q3 : Is it really possible to do modeling for a very complex


molecules ? Is it economical ?

◆ Q4 : Why chemical engineers ? Will we just leave this subject


to pure chemist or biologist ?
Answer to Q1
Are MM/Simulation methods really useful in
real application ?

◆ Enthalpy of reaction and enthalpy of formation


– Now in “maintenance” mode !
◆ Drug discovery
– Used by pharmaceutical companies
◆ Reaction Mechanism Analysis
– Research is going on…
◆ In some fields, they are used extensively and for
some field research is going on…
Answer to Q2
Do research institutes in advanced countries invest
money the development of theoretical MM methods ?
Air Liquide Design zeolites for O2/N2 separation

Air Prod & Chem Adhesives, adsorption

Albemarle Flame retardancy

Amoco Catalysis-homo/heterogeneous, thermochemistry

BP CH4 adsorption evaluation

Chevron Gas hydrates, lubricants

Dow Thermochemistry, reaction networks, AHrxn (safety)

DuPont Thermo, kinetics, catalysis

Exxon R&E NOx kinetics, elementary and networked reactions, safety

Compiled by Phil Westmoreland, UMass


Answer to Q2
Air Liquide Design zeolites for O2/N2 separation
Air Prod & Chem Adhesives, adsorption
Albemarle Flame retardancy
Amoco Catalysis-homo/heterogeneous, thermochemistry
BP CH4 adsorption evaluation
Chevron Gas hydrates, lubricants
Dow Thermochemistry, reaction networks, AHrxn (safety)
DuPont Thermo, kinetics, catalysis
Exxon R&E NOx kinetics, elementary and networked reactions, safety

Hercules Polysaccharide rheology


Lubrizol Hydrocarbon-chlorine reactions, anti-corrosion additives
General Motors Monolithic catalysts
Procter & Gamble Designed detergent enzyme
Royal Dutch Shell Diffusion in porous media
Schlumberger Setting of cements
Shell USA Solvent separations, catalysis
Union Carbide Catalysis, materials
Xerox Modify and develop materials
Answer to Q2

◆ USA
– NSF(National Science Foundation), DOE (Department of Energy), NIST (National Institute for
Science and Technology) , DARPA (Defense Advanced Research Projects Agency) , AFOSR(Air
Force Office of Scientific Research), NIH (National Institute for Health), …
◆ UK
– CCP Project (Collaborative Computational Project)
– CCP1 - The electronic structure of molecules
CCP2 - Continuum states of atoms and molecules
CCP3 - Simulation of physical and electronic properties of surfaces and interfaces
CCP4 - Protein crystallography
CCP5 - Computer simulation of condensed phases
CCP6 - Quantum molecular dynamics
CCP9 - Computational studies of the electronic structure of solids
CCP11 - Biosequence and structure analysis
CCP12 - High Performance Computing in Engineering
CCP13 - Fibre and polymer diffraction
CCP14 - Powder and small molecule single crystal diffraction
Answer to Q3
Is it really possible to do modeling for a very complex
molecules ? Is it economical ?
◆ Moore’s Law
– Computing speed doubles every 18 months
– Order of magnitude in every 5 years
◆ Parallel computing
– Add 2-3 order of magnitude for computing
Answer to Q4

Why chemical engineers ? Will we just leave this subject to pure


chemist or biologist ?
◆ Previous MM methods development was primary at normal
conditions (especially for drug design)
– 25 – 35 degree C, below 5 bar

◆ Chemical engineering covers wider range of application and various


properties

◆ MM methods have to be implemented in process simulators using


statistical molecular thermodynamics for the replacement of physical
property database.
Perspective
◆ Many routine thermophysical and thermochemical properties
of low molecular weight systems will be predictable
computationally at better accuracy and higher precision than
experiment
◆ Increasing emphasis on prediction by simulation
◆ Theory and experiment will continue to be essential for
systems challenged by simulation
– Long relaxation times (polymers, glasses)
– Materials whose properties are determined at scales larger than
molecular (mesocale)
– Materials in which quantum physical and/or chemical processes are
important
◆ Progress in simulation is possible only with new theories for
bridging time scales

You might also like