Short Course on Molecular
Dynamics Simulation
Lecture 5:
Boundary Conditions
Professor A. Martini
Purdue University
High Level Course Outline
1. MD Basics
2. Potential Energy Functions
3. Integration Algorithms
4. Temperature Control
5. Boundary Conditions
6. Neighbor Lists
7. Initialization and Equilibrium
8. Extracting Static Properties
9. Extracting Dynamic Properties
10. Non-Equilibrium MD
Molecular Dynamics Simulation A. Martini Last Updated 8/2009
Boundary Conditions & Minimum Images
Fixed simulation cell boundaries
– Repulsive boundary
– Atomistic rigid walls
– Atomistic semi-rigid walls
• Massive boundary atoms
• Spring potential
Periodic boundary conditions
Minimum image criteria
Molecular Dynamics Simulation A. Martini Last Updated 8/2009
Fixed Boundaries
Continuous barrier potential
– Simplest case, flat repulsive boundary
−12
⎛z⎞
z
u(z) = ε ⎜ ⎟
⎝σ ⎠
σ
– Virtual barrier texture introduced through more
complex expressions such as the (10-4-3) potential
Continuous function ⎡ 2 1 1 ⎤
u ( z ) = 2π ⎢ 10 − 4 − 3⎥
representing the (100)
face of an fcc lattice ⎢⎣ 5 z z (
3 / 2 z + 0.61 / 2 ⎥⎦ )
Molecular Dynamics Simulation A. Martini Last Updated 8/2009
Fixed Boundaries
Rigid atomistic walls
Δt
Molecular Dynamics Simulation A. Martini Last Updated 8/2009
Fixed Boundaries
Semi-rigid atomistic boundary Æ massive
boundary atoms
– Massive boundary atoms will be affected less by inter-
atomic interactions
– Consider the mb
interaction
between a F ab=F/mb
boundary and a
non-boundary
atom where F a=F/m
mb >> m
m ab << a
Molecular Dynamics Simulation A. Martini Last Updated 8/2009
Fixed Boundaries
Semi-rigid atomistic boundary Æ spring potential
– Boundaries both rigid enough to maintain structure and
flexible enough to interact with the other atoms
r
Lattice Wall
u(r) = k(r − rlattice)
1 2
Site Atom
2
Initial bcc configuration: During MD simulation:
Molecular Dynamics Simulation A. Martini Last Updated 8/2009
Periodic Boundary Conditions
PBCs enable macroscopic properties to be
calculated from fewer particles
Simulated interactions of “bulk” material
Out one side, in the other
Molecular Dynamics Simulation A. Martini Last Updated 8/2009
Periodic Boundary Conditions
The primary cell is replicated in all simulated
direction as image cells
Primary and image cells have the same
– Number, position, momentum of atoms
– Size
– Shape
Molecular Dynamics Simulation A. Martini Last Updated 8/2009
Periodic Boundary Conditions
The shape of a periodic simulation cell must fill
all space by translational operations of the
central box in 3D
Truncated Hexagonal Rhombic
Cube Octahedron Prism Dodecahedron
Molecular Dynamics Simulation A. Martini Last Updated 8/2009
Periodic Boundary Conditions
Forces between primary and image atoms
If i and j are in the primary cell:
L v
v ∂U (rij )
Fi = − v
α=(-1,1) α=(0,1) α=(1,1) ∂ri
If j is in an image cell:
α=(-1,0) α=(0,0) α=(1,0)
v v
v ∂U (rij − αL)
Fi = −∑ v
α=(-1,-1) α=(0,-1) α=(1,-1) v
α ∂ri
Molecular Dynamics Simulation A. Martini Last Updated 8/2009
Periodic Boundary Conditions
v v
v +1 +1 +1 ∂U (rij − αL)
Fi = − ∑ ∑ ∑ v
α x = −1 α y = −1 α z = −1 ∂ri
Minimum image distance
– Atom i can experience a force from atom j and its
26 images
– But only one image is a distance less than ½L
– So if the pair potential is truncated at rc ≤ 1/2 L,
either atom j or only one of its images can exert
force on atom i
– With PBCs, interactions are truncated at least to
this minimum image distance
Molecular Dynamics Simulation A. Martini Last Updated 8/2009