0% found this document useful (0 votes)
73 views19 pages

MD Lecture 1 - Intro To MD

This document provides an introduction to molecular dynamics simulations. It discusses why simulations are performed, as experiments can sometimes be impossible, dangerous, expensive or blind. It describes different types of simulations including quantum mechanics, molecular mechanics, and mixed quantum/classical methods. The document focuses on molecular dynamics, explaining that it calculates atom motion using Newtonian dynamics to determine net forces and accelerations. It also discusses energy functions, pair potentials, embedded atom methods, bond order potentials, and forcefields used in molecular dynamics. Finally, it briefly introduces numerical integration methods like the Verlet algorithm for solving the equations of motion.

Uploaded by

ADITYA VERMA
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
73 views19 pages

MD Lecture 1 - Intro To MD

This document provides an introduction to molecular dynamics simulations. It discusses why simulations are performed, as experiments can sometimes be impossible, dangerous, expensive or blind. It describes different types of simulations including quantum mechanics, molecular mechanics, and mixed quantum/classical methods. The document focuses on molecular dynamics, explaining that it calculates atom motion using Newtonian dynamics to determine net forces and accelerations. It also discusses energy functions, pair potentials, embedded atom methods, bond order potentials, and forcefields used in molecular dynamics. Finally, it briefly introduces numerical integration methods like the Verlet algorithm for solving the equations of motion.

Uploaded by

ADITYA VERMA
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 19

08-10-2022

ME F315 Advanced Manufacturing Process


Introduction to Molecular Dynamics

BITS Pilani Dr. Radha Raman Mishra


Pilani Campus
Mr. Ayush Khaitan

Why do we do Simulation?

▪ In some cases, experiment is


• Impossible
• Too dangerous
• Expensive
• Blind

▪ It’s a very useful complement as


it can
• Replace Experiment
• Provoke Experiment
• Explain Experiment
BITS Pilani, Pilani Campus

1
08-10-2022

Kinds of Simulation

BITS Pilani, Pilani Campus

Kind of Simulation
▪ Quantum Mechanics (QM)/Ab-
initio/DFT
• Electronic Structure (Schrodinger)
• Accurate, but extremely expensive
• For smaller systems

▪ Molecular Mechanics /Molecular


Dynamics
• Empirical forces (Newton)
• Less accurate, but practically fast
• Can be used for macromolecular
systems

▪ Mixed Quantum/Classical (QM/MM)


BITS Pilani, Pilani Campus

2
08-10-2022

Molecular Dynamics
▪ Molecular dynamics simulations calculate the motion of the
atoms in a molecular assembly using Newtonian dynamics to
determine the net force and acceleration experienced by each
atom. Each atom i at position ri, is treated as a point with a mass
mi and a fixed charge qi.

▪ Quantitative and/or Qualitative information about macroscopic


behavior of macromolecules can be obtained from simulation of a
system at atomistic level.
▪ Macroscopic properties are often determined by Molecule-level
behavior

BITS Pilani, Pilani Campus

Molecular Dynamics

▪ Molecular Dynamics Process:


• Particles and atoms have potential energies from interaction and
bonding with other atoms, this energy can be descried by a
mathematical function called energy function:

• In Classical Mechanics, the Forces on atoms can be calculated if


we knew the potential energies of these atoms

• From Newton's Law

BITS Pilani, Pilani Campus

3
08-10-2022

Molecular Dynamics
▪ Molecular Dynamics Process:
• So we determine the trajectory and motion of the atoms by using
the following expression.

BITS Pilani, Pilani Campus

Energy Functions: Interatomic


Potentials
▪ Empirical Potential Functions:
Assume a functional form for the potential function and then choose the
parameters to reproduce a set of experimental data. (Leonard-Jones,
Morse, Born-Mayer, Bond Order Potentials by Tersoff and Brenner)

▪ Semi-Empirical Potential Functions:


Derived from Quantum-Mechanical arguments, One can calculate the
electronic wavefunction for fixed atomic positions. This is difficult for a
system of many atoms. (Embedded Atom Method (EAM))

▪ Ab-initio:
One can perform direct electronic-structure (quantum-mechanics-based)
calculations of forces during so-called Ab-Initio Molecular Dynamics
(AIMD) simulation.

BITS Pilani, Pilani Campus

4
08-10-2022

Energy Functions: Pair


Potentials
▪ Pair potentials are the simplest form of a potential, that considers
only the two-atom interactions, and ignores the rest

▪ There are (N – 1) interactions per atom, and thus the number of


pairs, N pair, is on the order of N2:

BITS Pilani, Pilani Campus

Pair Potentials: Leonard –


Jones Potential
▪ Leonard – Jones (1924): are the simplest form of a potential, that
considers only the two-atom interactions, and ignores the rest

BITS Pilani, Pilani Campus

10

5
08-10-2022

Pair Potentials: Leonard –


Jones Potential
▪ This simple pair potential can express the atomic interactions of
noble gases (Ne, Ar, Kr, etc.), spherical molecules, and secondary
bonds very well

BITS Pilani, Pilani Campus

11

Energy Functions: Embedded


Atom Method (EAM)
▪ In metals and transition metals, atoms (or ions) are in the middle of
the electron sea and mostly coordinated with 8–12 other atoms.
▪ The coulomb interactions between them are long-ranged ones,
covering easily tens of atoms.
▪ The embedded atom method (EAM) considers the effective
electron density at a given atomic site as one of the parameters,
thereby capturing some electronic effects while maintaining the
simplicity of a potential.

BITS Pilani, Pilani Campus

12

6
08-10-2022

Energy Functions: Embedded


Atom Method (EAM)
▪ The EAM considers both contributions from the pair potential and
the embedding energy that approximates the missing N-atom
effect in a pair potential

BITS Pilani, Pilani Campus

13

Energy Functions: Embedded


Atom Method (EAM)
▪ This type of potential works very well for most metals and transition
metals, especially for FCC metals.
▪ Does not work very well for Alloys

BITS Pilani, Pilani Campus

14

7
08-10-2022

Energy Functions: Bond


Order Potentials
▪ In covalent solids, deviations from the scheme of pairwise additives
are even larger since these solids are less closely-packed (often
with a coordination number of only 4) and have strong directional
bonds of equivalent strength.
▪ Bond angle and Bond order are the prime characteristics of these
materials, and the strength of a bond between two atoms depends
on the local environment such as coordination number and bond
angle.

BITS Pilani, Pilani Campus

15

Energy Functions: Tersoff


Potential
▪ Tersoff Potential is one of the potentials which includes extra
geometrical and bond order related terms

BITS Pilani, Pilani Campus

16

8
08-10-2022

Energy Functions: Forcefields


▪ Force field potentials are designed for organics, polymers, and
bioorganic systems (for example, proteins)
▪ The forcefield is a collection of equations and associated
constants designed to reproduce molecular geometry and selected
properties of tested structures

▪ It used too describe the time evolution of


• Bond lengths
• Bond angles and torsions
• Van der Waals Interactions
• Electrostatic interactions

BITS Pilani, Pilani Campus

17

Energy Terms in Forcefields

BITS Pilani, Pilani Campus

18

9
08-10-2022

Forcefields
▪ Examples on different
forcefields:
• AMBER
• CHARMM
• GROMOS
• OPLS
• MMFF

▪ All vary in:


• Functional Form
• Empirical Parameters
• Simulations Target

BITS Pilani, Pilani Campus

19

Numerical Integration: Finding


Trajectories
▪ Newton’s equation represents a set of N second order differential
equations
▪ It’s impossible to solve this set of equations exactly to obtain the
trajectories of atoms
▪ We solve these equations Numerically at discrete time steps to
determine the trajectory of each atom.

BITS Pilani, Pilani Campus

20

10
08-10-2022

Numerical Integration: Verlet


Algorithm
▪ There are several algorithms available for the numerical integration
of Newton’s equations of motion and calculation of atomic
trajectories in practical MD runs:
• Verlet Algorithm (1967): Taylor expansion for forward and
backward positions in time

• Summing these two equations

BITS Pilani, Pilani Campus

21

Numerical Integration:
VelocityVerlet Algorithm
▪ The velocity Verlet algorithm (Verlet 1968) is one of the most
frequently implemented methods in MD

BITS Pilani, Pilani Campus

22

11
08-10-2022

Timesteps
▪ We need to determine the timestep of the numerical integration of
the simulations trajectories, in which every unit of this time step we
do the integration
▪ The Timestep of MD simulations is limited by the event which has
the smallest time (Why!!)

BITS Pilani, Pilani Campus

23

Timesteps: 1 fs barrier
▪ Atoms in a solid lattice vibrate in the range of 10−14

▪ Ideally the timestep should be 1/10 highest frequency (1/10×10−14s


= 1 femtosecond)

BITS Pilani, Pilani Campus

24

12
08-10-2022

Molecular Dynamics
Ensembles
▪ In a simulation box of an MD run, each atom moves and behaves
differently, and a new microstate of the system is generated in
every timestep
▪ An ensemble is the collection of possible systems that belongs to
the different microscopic states but with an identical macroscopic
or thermodynamic state.

BITS Pilani, Pilani Campus

25

Molecular Dynamics
Ensembles
▪ As with actual experiments in a lab, we impose these external
constraints on ensembles to have particular properties out after a
run.
▪ Thermodynamic properties such as entropy S, Helmholtz free
energy F, Gibbs free energy G, and the chemical potential μ can
be derived from data obtained from these prearranged ensembles

BITS Pilani, Pilani Campus

26

13
08-10-2022

Periodic Boundary Conditions


▪ Periodic boundary condition (PBCs) are a set of boundary conditions
which are often chosen for approximating a large (infinite) system by
using a small part called a unit cell. PBCs are often used in computer
simulations and mathematical models.
▪ A “real” system might contain about NA = 6 × 1023 molecules. Simulation
may contain about 100-1,000,000 molecules. A small cube of 10x10x10
molecules has about half of the molecules on the Surface, this is Not
appropriate to study “bulk” properties
▪ PBCs have the following properties:
• Small box replicated in all directions
• A particle that leaves the box on one side is
replaced by an image particle that enters from the
other side
• There are no walls and no surface particles
• Actual simulation takes place only for the atoms
inside the primary
box at the center, and all image boxes just copy the
primary box.

BITS Pilani, Pilani Campus

27

Minimization
▪ Minimization is a procedure that attempts to minimize the potential
energy of the system to the lowest possible point.

BITS Pilani, Pilani Campus

28

14
08-10-2022

Minimization: Steepest
Descent
▪ Steepest Descent is a method for obtaining local stationary points of a
multi-dimensional function
• X0 initial structure
• Compute U(R) and ∇𝑈(𝑅) Energy and Gradient
• Update Xi+1= Xi - γi ∇𝑈(𝑅) Displacement
• Check for Convergence
if Yes: Done Xf = Xi
if No: Return to step 2

▪ Convergence Check:

• | ∇𝑈 𝑅 | ≈ 0 Gradient
• | 𝑈 𝑥𝑖 − 𝑈(𝑥𝑖−1) | ≈ 0 Energy
• |𝑥𝑖 − 𝑥𝑖−1| ≈ 0 Displacement
BITS Pilani, Pilani Campus

29

Minimization: Conjugate
Gradient
▪ Conjugate Gradient is a more advanced energy minimization
algorithm than steepest descent. It uses step history to accelerate
convergence

BITS Pilani, Pilani Campus

30

15
08-10-2022

MD Algorithm
1. Assign initial

2. Compute

3. Update

4. Update

5. Repeat until

6. Compute Desired
Properties

BITS Pilani, Pilani Campus

31

Limitations of MD
▪ Timescales: Require shorter timesteps for numerical stability
• Usually 1 fs (1 × 10-15s)

▪ Timescales: Structural changes in protein can take up to


nanoseconds, microseconds, milliseconds or even seconds
• Requires millions, trillions or even more time steps, and thus
calculations

▪ Energy Functions accuracy: inherently, they are approximations

▪ No Electricity or Magnetism or Optical properties in Classical


MD

BITS Pilani, Pilani Campus

32

16
08-10-2022

Aims of MD simulation studies


▪ Understanding the properties of assemblies of molecules in terms
of their structure and the microscopic interactions between them
• Serves as a complement to conventional experiments, enabling us
to learn something new, something that cannot be found out in
other ways, …
▪ Bridge between microscopic length and time scales and the
macroscopic world of the laboratory
▪ Provides a guess at the interactions and “exact” predictions of bulk
properties
▪ Revealing hidden details of the bulk measurements
• E.g. Link between diffusion coefficient and velocity autocorrelation
function

BITS Pilani, Pilani Campus

33

Software Tools for MDS


▪ Multiple Software packages are available for MD Simulations
• AMBER, NAMD, GROMACS, Desmond, Desmond-GPU
(Shrodinger), ACEMD (HTMD), OpenMM, CHARMM, LAMMPS
• Core functionality is same, force fields are compatible
• Differ in speed, scalability, cost and feasibility
▪ Visualization Tool: VMD , OVITO, Avogadro, and VESTA

BITS Pilani, Pilani Campus

34

17
08-10-2022

What can we study with MD


Simulations
▪ Since the time the MD method was introduced, it has been used to
investigate a wide range of problems in different research fields:
• Chemistry and Biochemistry: molecular structures, reactions,
drug design, vibrational relaxation and energy transfer, structure of
membranes, dynamics of large biomolecules, protein folding, …
• Statistical Mechanics and Physics: theory of liquids, correlated
many-body motion, properties of statistical ensembles, structure
and properties of small clusters, phase transitions..
• Materials Science: point, linear, and planar defects in crystals and
their interactions, microscopic mechanisms of fracture, surface
reconstruction, melting and faceting, film growth, friction, …

BITS Pilani, Pilani Campus

35

What can we study with MD


Simulations

BITS Pilani, Pilani Campus

36

18
08-10-2022

THANK YOU
BITS Pilani
Pilani Campus

37

19

You might also like