| 1 |
|
|
| 2 |
< |
\section{The Emperical Energy Functions} |
| 2 |
> |
\section{\label{sec:empiricalEnergy}The Empirical Energy Functions} |
| 3 |
|
|
| 4 |
< |
\subsection{Atoms and Molecules} |
| 4 |
> |
\subsection{\label{sec:atomsMolecules}Atoms, Molecules and Rigid Bodies} |
| 5 |
|
|
| 6 |
< |
The basic unit of an {\sc oopse} simulation is the atom. The parameters |
| 7 |
< |
describing the atom are generalized to make the atom as flexible a |
| 8 |
< |
representation as possible. They may represent specific atoms of an |
| 9 |
< |
element, or be used for collections of atoms such as a methyl |
| 10 |
< |
group. The atoms are also capable of having a directional component |
| 11 |
< |
associated with them, often in the form of a dipole. Charges on atoms |
| 12 |
< |
are not currently suporrted by {\sc oopse}. |
| 6 |
> |
The basic unit of an {\sc oopse} simulation is the atom. The |
| 7 |
> |
parameters describing the atom are generalized to make the atom as |
| 8 |
> |
flexible a representation as possible. They may represent specific |
| 9 |
> |
atoms of an element, or be used for collections of atoms such as |
| 10 |
> |
methyl and carbonyl groups. The atoms are also capable of having |
| 11 |
> |
directional components associated with them (\emph{e.g.}~permanent |
| 12 |
> |
dipoles). Charges on atoms are not currently supported by {\sc oopse}. |
| 13 |
|
|
| 14 |
< |
The second most basic building block of a simulation is the |
| 15 |
< |
molecule. The molecule is a way for {\sc oopse} to keep track of the atoms |
| 16 |
< |
in a simulation in logical manner. This particular unit will store the |
| 17 |
< |
identities of all the atoms associated with itself and is responsible |
| 18 |
< |
for the evaluation of its own bonded interaction (i.e.~bonds, bends, |
| 19 |
< |
and torsions). |
| 14 |
> |
\begin{lstlisting}[caption={[Specifier for molecules and atoms] A sample specification of the simple Ar molecule},label=sch:AtmMole] |
| 15 |
> |
molecule{ |
| 16 |
> |
name = "Ar"; |
| 17 |
> |
nAtoms = 1; |
| 18 |
> |
atom[0]{ |
| 19 |
> |
type="Ar"; |
| 20 |
> |
position( 0.0, 0.0, 0.0 ); |
| 21 |
> |
} |
| 22 |
> |
} |
| 23 |
> |
\end{lstlisting} |
| 24 |
> |
|
| 25 |
> |
Atoms can be collected into secondary srtructures such as rigid bodies |
| 26 |
> |
or molecules. The molecule is a way for {\sc oopse} to keep track of |
| 27 |
> |
the atoms in a simulation in logical manner. Molecular units store the |
| 28 |
> |
identities of all the atoms associated with themselves, and are |
| 29 |
> |
responsible for the evaluation of their own internal interactions |
| 30 |
> |
(\emph{i.e.}~bonds, bends, and torsions). Scheme \ref{sch:AtmMole} |
| 31 |
> |
shws how one creates a molecule in the \texttt{.mdl} files. The |
| 32 |
> |
position of the atoms given in the declaration are relative to the |
| 33 |
> |
origin of the molecule, and is used when creating a system containing |
| 34 |
> |
the molecule. |
| 35 |
> |
|
| 36 |
> |
As stated previously, one of the features that sets {\sc oopse} apart |
| 37 |
> |
from most of the current molecular simulation packages is the ability |
| 38 |
> |
to handle rigid body dynamics. Rigid bodies are non-spherical |
| 39 |
> |
particles or collections of particles that have a constant internal |
| 40 |
> |
potential and move collectively.\cite{Goldstein01} They are not |
| 41 |
> |
included in most simulation packages because of the requirement to |
| 42 |
> |
propagate the orientational degrees of freedom. Until recently, |
| 43 |
> |
integrators which propagate orientational motion have been lacking. |
| 44 |
> |
|
| 45 |
> |
Moving a rigid body involves determination of both the force and |
| 46 |
> |
torque applied by the surroundings, which directly affect the |
| 47 |
> |
translational and rotational motion in turn. In order to accumulate |
| 48 |
> |
the total force on a rigid body, the external forces and torques must |
| 49 |
> |
first be calculated for all the internal particles. The total force on |
| 50 |
> |
the rigid body is simply the sum of these external forces. |
| 51 |
> |
Accumulation of the total torque on the rigid body is more complex |
| 52 |
> |
than the force in that it is the torque applied on the center of mass |
| 53 |
> |
that dictates rotational motion. The torque on rigid body {\it i} is |
| 54 |
> |
\begin{equation} |
| 55 |
> |
\boldsymbol{\tau}_i= |
| 56 |
> |
\sum_{a}(\mathbf{r}_{ia}-\mathbf{r}_i)\times \mathbf{f}_{ia} |
| 57 |
> |
+ \boldsymbol{\tau}_{ia}, |
| 58 |
> |
\label{eq:torqueAccumulate} |
| 59 |
> |
\end{equation} |
| 60 |
> |
where $\boldsymbol{\tau}_i$ and $\mathbf{r}_i$ are the torque on and |
| 61 |
> |
position of the center of mass respectively, while $\mathbf{f}_{ia}$, |
| 62 |
> |
$\mathbf{r}_{ia}$, and $\boldsymbol{\tau}_{ia}$ are the force on, |
| 63 |
> |
position of, and torque on the component particles of the rigid body. |
| 64 |
> |
|
| 65 |
> |
The summation of the total torque is done in the body fixed axis of |
| 66 |
> |
the rigid body. In order to move between the space fixed and body |
| 67 |
> |
fixed coordinate axes, parameters describing the orientation must be |
| 68 |
> |
maintained for each rigid body. At a minimum, the rotation matrix |
| 69 |
> |
(\textbf{A}) can be described by the three Euler angles ($\phi, |
| 70 |
> |
\theta,$ and $\psi$), where the elements of \textbf{A} are composed of |
| 71 |
> |
trigonometric operations involving $\phi, \theta,$ and |
| 72 |
> |
$\psi$.\cite{Goldstein01} In order to avoid numerical instabilities |
| 73 |
> |
inherent in using the Euler angles, the four parameter ``quaternion'' |
| 74 |
> |
scheme is often used. The elements of \textbf{A} can be expressed as |
| 75 |
> |
arithmetic operations involving the four quaternions ($q_0, q_1, q_2,$ |
| 76 |
> |
and $q_3$).\cite{allen87:csl} Use of quaternions also leads to |
| 77 |
> |
performance enhancements, particularly for very small |
| 78 |
> |
systems.\cite{Evans77} |
| 79 |
> |
|
| 80 |
> |
{\sc oopse} utilizes a relatively new scheme that propagates the |
| 81 |
> |
entire nine parameter rotation matrix internally. Further discussion |
| 82 |
> |
on this choice can be found in Sec.~\ref{sec:integrate}. An example |
| 83 |
> |
definition of a riged body can be seen in Scheme |
| 84 |
> |
\ref{sch:rigidBody}. The positions in the atom definitions are the |
| 85 |
> |
placements of the atoms relative to the origin of the rigid body, |
| 86 |
> |
which itself has a position relative to the origin of the molecule. |
| 87 |
> |
|
| 88 |
> |
\begin{lstlisting}[caption={[Defining rigid bodies]A sample definition of a rigid body},label={sch:rigidBody}] |
| 89 |
> |
molecule{ |
| 90 |
> |
name = "TIP3P_water"; |
| 91 |
> |
nRigidBodies = 1; |
| 92 |
> |
rigidBody[0]{ |
| 93 |
> |
nAtoms = 3; |
| 94 |
> |
atom[0]{ |
| 95 |
> |
type = "O_TIP3P"; |
| 96 |
> |
position( 0.0, 0.0, -0.06556 ); |
| 97 |
> |
} |
| 98 |
> |
atom[1]{ |
| 99 |
> |
type = "H_TIP3P"; |
| 100 |
> |
position( 0.0, 0.75695, 0.52032 ); |
| 101 |
> |
} |
| 102 |
> |
atom[2]{ |
| 103 |
> |
type = "H_TIP3P"; |
| 104 |
> |
position( 0.0, -0.75695, 0.52032 ); |
| 105 |
> |
} |
| 106 |
> |
position( 0.0, 0.0, 0.0 ); |
| 107 |
> |
orientation( 0.0, 0.0, 1.0 ); |
| 108 |
> |
} |
| 109 |
> |
} |
| 110 |
> |
\end{lstlisting} |
| 111 |
> |
|
| 112 |
> |
\subsection{\label{sec:LJPot}The Lennard Jones Potential} |
| 113 |
> |
|
| 114 |
> |
The most basic force field implemented in {\sc oopse} is the |
| 115 |
> |
Lennard-Jones potential, which mimics the van der Waals interaction at |
| 116 |
> |
long distances, and uses an empirical repulsion at short |
| 117 |
> |
distances. The Lennard-Jones potential is given by: |
| 118 |
> |
\begin{equation} |
| 119 |
> |
V_{\text{LJ}}(r_{ij}) = |
| 120 |
> |
4\epsilon_{ij} \biggl[ |
| 121 |
> |
\biggl(\frac{\sigma_{ij}}{r_{ij}}\biggr)^{12} |
| 122 |
> |
- \biggl(\frac{\sigma_{ij}}{r_{ij}}\biggr)^{6} |
| 123 |
> |
\biggr] |
| 124 |
> |
\label{eq:lennardJonesPot} |
| 125 |
> |
\end{equation} |
| 126 |
> |
Where $r_{ij}$ is the distance between particles $i$ and $j$, |
| 127 |
> |
$\sigma_{ij}$ scales the length of the interaction, and |
| 128 |
> |
$\epsilon_{ij}$ scales the well depth of the potential. Scheme |
| 129 |
> |
\ref{sch:LJFF} gives and example partial \texttt{.bass} file that |
| 130 |
> |
shows a system of 108 Ar particles simulated with the Lennard-Jones |
| 131 |
> |
force field. |
| 132 |
> |
|
| 133 |
> |
\begin{lstlisting}[caption={[Invocation of the Lennard-Jones force field] A sample system using the Lennard-Jones force field.},label={sch:LJFF}] |
| 134 |
> |
|
| 135 |
> |
/* |
| 136 |
> |
* The Ar molecule is specified |
| 137 |
> |
* external to the.bass file |
| 138 |
> |
*/ |
| 139 |
> |
|
| 140 |
> |
#include "argon.mdl" |
| 141 |
> |
|
| 142 |
> |
nComponents = 1; |
| 143 |
> |
component{ |
| 144 |
> |
type = "Ar"; |
| 145 |
> |
nMol = 108; |
| 146 |
> |
} |
| 147 |
> |
|
| 148 |
> |
/* |
| 149 |
> |
* The initial configuration is generated |
| 150 |
> |
* before the simulation is invoked. |
| 151 |
> |
*/ |
| 152 |
> |
|
| 153 |
> |
initialConfig = "./argon.init"; |
| 154 |
> |
|
| 155 |
> |
forceField = "LJ"; |
| 156 |
> |
\end{lstlisting} |
| 157 |
> |
|
| 158 |
> |
Because this potential is calculated between all pairs, the force |
| 159 |
> |
evaluation can become computationally expensive for large systems. To |
| 160 |
> |
keep the pair evaluations to a manageable number, {\sc oopse} employs |
| 161 |
> |
a cut-off radius.\cite{allen87:csl} The cutoff radius is set to be |
| 162 |
> |
$2.5\sigma_{ii}$, where $\sigma_{ii}$ is the largest Lennard-Jones |
| 163 |
> |
length parameter present in the simulation. Truncating the calculation |
| 164 |
> |
at $r_{\text{cut}}$ introduces a discontinuity into the potential |
| 165 |
> |
energy. To offset this discontinuity, the energy value at |
| 166 |
> |
$r_{\text{cut}}$ is subtracted from the potential. This causes the |
| 167 |
> |
potential to go to zero smoothly at the cut-off radius. |
| 168 |
> |
|
| 169 |
> |
Interactions between dissimilar particles requires the generation of |
| 170 |
> |
cross term parameters for $\sigma$ and $\epsilon$. These are |
| 171 |
> |
calculated through the Lorentz-Berthelot mixing |
| 172 |
> |
rules:\cite{allen87:csl} |
| 173 |
> |
\begin{equation} |
| 174 |
> |
\sigma_{ij} = \frac{1}{2}[\sigma_{ii} + \sigma_{jj}] |
| 175 |
> |
\label{eq:sigmaMix} |
| 176 |
> |
\end{equation} |
| 177 |
> |
and |
| 178 |
> |
\begin{equation} |
| 179 |
> |
\epsilon_{ij} = \sqrt{\epsilon_{ii} \epsilon_{jj}} |
| 180 |
> |
\label{eq:epsilonMix} |
| 181 |
> |
\end{equation} |
| 182 |
> |
|
| 183 |
> |
|
| 184 |
> |
|
| 185 |
> |
\subsection{\label{sec:DUFF}Dipolar Unified-Atom Force Field} |
| 186 |
> |
|
| 187 |
> |
The dipolar unified-atom force field ({\sc duff}) was developed to |
| 188 |
> |
simulate lipid bilayers. The simulations require a model capable of |
| 189 |
> |
forming bilayers, while still being sufficiently computationally |
| 190 |
> |
efficient to allow large systems ($\approx$100's of phospholipids, |
| 191 |
> |
$\approx$1000's of waters) to be simulated for long times |
| 192 |
> |
($\approx$10's of nanoseconds). |
| 193 |
> |
|
| 194 |
> |
With this goal in mind, {\sc duff} has no point |
| 195 |
> |
charges. Charge-neutral distributions were replaced with dipoles, |
| 196 |
> |
while most atoms and groups of atoms were reduced to Lennard-Jones |
| 197 |
> |
interaction sites. This simplification cuts the length scale of long |
| 198 |
> |
range interactions from $\frac{1}{r}$ to $\frac{1}{r^3}$, allowing us |
| 199 |
> |
to avoid the computationally expensive Ewald sum. Instead, we can use |
| 200 |
> |
neighbor-lists, reaction field, and cutoff radii for the dipolar |
| 201 |
> |
interactions. |
| 202 |
> |
|
| 203 |
> |
As an example, lipid head-groups in {\sc duff} are represented as |
| 204 |
> |
point dipole interaction sites. By placing a dipole of 20.6~Debye at |
| 205 |
> |
the head group center of mass, our model mimics the head group of |
| 206 |
> |
phosphatidylcholine.\cite{Cevc87} Additionally, a large Lennard-Jones |
| 207 |
> |
site is located at the pseudoatom's center of mass. The model is |
| 208 |
> |
illustrated by the dark grey atom in Fig.~\ref{fig:lipidModel}. The |
| 209 |
> |
water model we use to complement the dipoles of the lipids is our |
| 210 |
> |
reparameterization of the soft sticky dipole (SSD) model of Ichiye |
| 211 |
> |
\emph{et al.}\cite{liu96:new_model} |
| 212 |
> |
|
| 213 |
> |
\begin{figure} |
| 214 |
> |
\epsfxsize=\linewidth |
| 215 |
> |
\epsfbox{lipidModel.eps} |
| 216 |
> |
\caption{A representation of the lipid model. $\phi$ is the torsion angle, $\theta$ % |
| 217 |
> |
is the bend angle, $\mu$ is the dipole moment of the head group, and n |
| 218 |
> |
is the chain length.} |
| 219 |
> |
\label{fig:lipidModel} |
| 220 |
> |
\end{figure} |
| 221 |
> |
|
| 222 |
> |
We have used a set of scalable parameters to model the alkyl groups |
| 223 |
> |
with Lennard-Jones sites. For this, we have borrowed parameters from |
| 224 |
> |
the TraPPE force field of Siepmann |
| 225 |
> |
\emph{et al}.\cite{Siepmann1998} TraPPE is a unified-atom |
| 226 |
> |
representation of n-alkanes, which is parametrized against phase |
| 227 |
> |
equilibria using Gibbs ensemble Monte Carlo simulation |
| 228 |
> |
techniques.\cite{Siepmann1998} One of the advantages of TraPPE is that |
| 229 |
> |
it generalizes the types of atoms in an alkyl chain to keep the number |
| 230 |
> |
of pseudoatoms to a minimum; the parameters for an atom such as |
| 231 |
> |
$\text{CH}_2$ do not change depending on what species are bonded to |
| 232 |
> |
it. |
| 233 |
> |
|
| 234 |
> |
TraPPE also constrains all bonds to be of fixed length. Typically, |
| 235 |
> |
bond vibrations are the fastest motions in a molecular dynamic |
| 236 |
> |
simulation. Small time steps between force evaluations must be used to |
| 237 |
> |
ensure adequate sampling of the bond potential to ensure conservation |
| 238 |
> |
of energy. By constraining the bond lengths, larger time steps may be |
| 239 |
> |
used when integrating the equations of motion. A simulation using {\sc |
| 240 |
> |
duff} is illustrated in Scheme \ref{sch:DUFF}. |
| 241 |
> |
|
| 242 |
> |
\begin{lstlisting}[caption={[Invocation of {\sc duff}]Sample \texttt{.bass} file showing a simulation utilizing {\sc duff}},label={sch:DUFF}] |
| 243 |
> |
|
| 244 |
> |
#include "water.mdl" |
| 245 |
> |
#include "lipid.mdl" |
| 246 |
> |
|
| 247 |
> |
nComponents = 2; |
| 248 |
> |
component{ |
| 249 |
> |
type = "simpleLipid_16"; |
| 250 |
> |
nMol = 60; |
| 251 |
> |
} |
| 252 |
> |
|
| 253 |
> |
component{ |
| 254 |
> |
type = "SSD_water"; |
| 255 |
> |
nMol = 1936; |
| 256 |
> |
} |
| 257 |
> |
|
| 258 |
> |
initialConfig = "bilayer.init"; |
| 259 |
> |
|
| 260 |
> |
forceField = "DUFF"; |
| 261 |
> |
|
| 262 |
> |
\end{lstlisting} |
| 263 |
> |
|
| 264 |
> |
\subsubsection{\label{subSec:energyFunctions}{\sc duff} Energy Functions} |
| 265 |
> |
|
| 266 |
> |
The total potential energy function in {\sc duff} is |
| 267 |
> |
\begin{equation} |
| 268 |
> |
V = \sum^{N}_{I=1} V^{I}_{\text{Internal}} |
| 269 |
> |
+ \sum^{N}_{I=1} \sum_{J>I} V^{IJ}_{\text{Cross}} |
| 270 |
> |
\label{eq:totalPotential} |
| 271 |
> |
\end{equation} |
| 272 |
> |
Where $V^{I}_{\text{Internal}}$ is the internal potential of molecule $I$: |
| 273 |
> |
\begin{equation} |
| 274 |
> |
V^{I}_{\text{Internal}} = |
| 275 |
> |
\sum_{\theta_{ijk} \in I} V_{\text{bend}}(\theta_{ijk}) |
| 276 |
> |
+ \sum_{\phi_{ijkl} \in I} V_{\text{torsion}}(\phi_{ijkl}) |
| 277 |
> |
+ \sum_{i \in I} \sum_{(j>i+4) \in I} |
| 278 |
> |
\biggl[ V_{\text{LJ}}(r_{ij}) + V_{\text{dipole}} |
| 279 |
> |
(\mathbf{r}_{ij},\boldsymbol{\Omega}_{i},\boldsymbol{\Omega}_{j}) |
| 280 |
> |
\biggr] |
| 281 |
> |
\label{eq:internalPotential} |
| 282 |
> |
\end{equation} |
| 283 |
> |
Here $V_{\text{bend}}$ is the bend potential for all 1, 3 bonded pairs |
| 284 |
> |
within the molecule $I$, and $V_{\text{torsion}}$ is the torsion potential |
| 285 |
> |
for all 1, 4 bonded pairs. The pairwise portions of the internal |
| 286 |
> |
potential are excluded for pairs that are closer than three bonds, |
| 287 |
> |
i.e.~atom pairs farther away than a torsion are included in the |
| 288 |
> |
pair-wise loop. |
| 289 |
> |
|
| 290 |
> |
|
| 291 |
> |
The bend potential of a molecule is represented by the following function: |
| 292 |
> |
\begin{equation} |
| 293 |
> |
V_{\text{bend}}(\theta_{ijk}) = k_{\theta}( \theta_{ijk} - \theta_0 )^2 \label{eq:bendPot} |
| 294 |
> |
\end{equation} |
| 295 |
> |
Where $\theta_{ijk}$ is the angle defined by atoms $i$, $j$, and $k$ |
| 296 |
> |
(see Fig.~\ref{fig:lipidModel}), $\theta_0$ is the equilibrium |
| 297 |
> |
bond angle, and $k_{\theta}$ is the force constant which determines the |
| 298 |
> |
strength of the harmonic bend. The parameters for $k_{\theta}$ and |
| 299 |
> |
$\theta_0$ are borrowed from those in TraPPE.\cite{Siepmann1998} |
| 300 |
> |
|
| 301 |
> |
The torsion potential and parameters are also borrowed from TraPPE. It is |
| 302 |
> |
of the form: |
| 303 |
> |
\begin{equation} |
| 304 |
> |
V_{\text{torsion}}(\phi) = c_1[1 + \cos \phi] |
| 305 |
> |
+ c_2[1 + \cos(2\phi)] |
| 306 |
> |
+ c_3[1 + \cos(3\phi)] |
| 307 |
> |
\label{eq:origTorsionPot} |
| 308 |
> |
\end{equation} |
| 309 |
> |
Here $\phi$ is the angle defined by four bonded neighbors $i$, |
| 310 |
> |
$j$, $k$, and $l$ (again, see Fig.~\ref{fig:lipidModel}). For |
| 311 |
> |
computational efficiency, the torsion potential has been recast after |
| 312 |
> |
the method of CHARMM,\cite{charmm1983} in which the angle series is |
| 313 |
> |
converted to a power series of the form: |
| 314 |
> |
\begin{equation} |
| 315 |
> |
V_{\text{torsion}}(\phi) = |
| 316 |
> |
k_3 \cos^3 \phi + k_2 \cos^2 \phi + k_1 \cos \phi + k_0 |
| 317 |
> |
\label{eq:torsionPot} |
| 318 |
> |
\end{equation} |
| 319 |
> |
Where: |
| 320 |
> |
\begin{align*} |
| 321 |
> |
k_0 &= c_1 + c_3 \\ |
| 322 |
> |
k_1 &= c_1 - 3c_3 \\ |
| 323 |
> |
k_2 &= 2 c_2 \\ |
| 324 |
> |
k_3 &= 4c_3 |
| 325 |
> |
\end{align*} |
| 326 |
> |
By recasting the potential as a power series, repeated trigonometric |
| 327 |
> |
evaluations are avoided during the calculation of the potential energy. |
| 328 |
> |
|
| 329 |
> |
|
| 330 |
> |
The cross potential between molecules $I$ and $J$, $V^{IJ}_{\text{Cross}}$, is |
| 331 |
> |
as follows: |
| 332 |
> |
\begin{equation} |
| 333 |
> |
V^{IJ}_{\text{Cross}} = |
| 334 |
> |
\sum_{i \in I} \sum_{j \in J} |
| 335 |
> |
\biggl[ V_{\text{LJ}}(r_{ij}) + V_{\text{dipole}} |
| 336 |
> |
(\mathbf{r}_{ij},\boldsymbol{\Omega}_{i},\boldsymbol{\Omega}_{j}) |
| 337 |
> |
+ V_{\text{sticky}} |
| 338 |
> |
(\mathbf{r}_{ij},\boldsymbol{\Omega}_{i},\boldsymbol{\Omega}_{j}) |
| 339 |
> |
\biggr] |
| 340 |
> |
\label{eq:crossPotentail} |
| 341 |
> |
\end{equation} |
| 342 |
> |
Where $V_{\text{LJ}}$ is the Lennard Jones potential, |
| 343 |
> |
$V_{\text{dipole}}$ is the dipole dipole potential, and |
| 344 |
> |
$V_{\text{sticky}}$ is the sticky potential defined by the SSD model |
| 345 |
> |
(Sec.~\ref{sec:SSD}). Note that not all atom types include all |
| 346 |
> |
interactions. |
| 347 |
> |
|
| 348 |
> |
The dipole-dipole potential has the following form: |
| 349 |
> |
\begin{equation} |
| 350 |
> |
V_{\text{dipole}}(\mathbf{r}_{ij},\boldsymbol{\Omega}_{i}, |
| 351 |
> |
\boldsymbol{\Omega}_{j}) = \frac{|\mu_i||\mu_j|}{4\pi\epsilon_{0}r_{ij}^{3}} \biggl[ |
| 352 |
> |
\boldsymbol{\hat{u}}_{i} \cdot \boldsymbol{\hat{u}}_{j} |
| 353 |
> |
- |
| 354 |
> |
\frac{3(\boldsymbol{\hat{u}}_i \cdot \mathbf{r}_{ij}) % |
| 355 |
> |
(\boldsymbol{\hat{u}}_j \cdot \mathbf{r}_{ij}) } |
| 356 |
> |
{r^{2}_{ij}} \biggr] |
| 357 |
> |
\label{eq:dipolePot} |
| 358 |
> |
\end{equation} |
| 359 |
> |
Here $\mathbf{r}_{ij}$ is the vector starting at atom $i$ pointing |
| 360 |
> |
towards $j$, and $\boldsymbol{\Omega}_i$ and $\boldsymbol{\Omega}_j$ |
| 361 |
> |
are the orientational degrees of freedom for atoms $i$ and $j$ |
| 362 |
> |
respectively. $|\mu_i|$ is the magnitude of the dipole moment of atom |
| 363 |
> |
$i$, $\boldsymbol{\hat{u}}_i$ is the standard unit orientation |
| 364 |
> |
vector of $\boldsymbol{\Omega}_i$, and $\boldsymbol{\hat{r}}_{ij}$ is |
| 365 |
> |
the unit vector pointing along $\mathbf{r}_{ij}$. |
| 366 |
> |
|
| 367 |
> |
|
| 368 |
> |
\subsubsection{\label{sec:SSD}The {\sc duff} Water Models: SSD/E and SSD/RF} |
| 369 |
> |
|
| 370 |
> |
In the interest of computational efficiency, the default solvent used |
| 371 |
> |
by {\sc oopse} is the extended Soft Sticky Dipole (SSD/E) water |
| 372 |
> |
model.\cite{Gezelter04} The original SSD was developed by Ichiye |
| 373 |
> |
\emph{et al.}\cite{liu96:new_model} as a modified form of the hard-sphere |
| 374 |
> |
water model proposed by Bratko, Blum, and |
| 375 |
> |
Luzar.\cite{Bratko85,Bratko95} It consists of a single point dipole |
| 376 |
> |
with a Lennard-Jones core and a sticky potential that directs the |
| 377 |
> |
particles to assume the proper hydrogen bond orientation in the first |
| 378 |
> |
solvation shell. Thus, the interaction between two SSD water molecules |
| 379 |
> |
\emph{i} and \emph{j} is given by the potential |
| 380 |
> |
\begin{equation} |
| 381 |
> |
V_{ij} = |
| 382 |
> |
V_{ij}^{LJ} (r_{ij})\ + V_{ij}^{dp} |
| 383 |
> |
(\mathbf{r}_{ij},\boldsymbol{\Omega}_i,\boldsymbol{\Omega}_j)\ + |
| 384 |
> |
V_{ij}^{sp} |
| 385 |
> |
(\mathbf{r}_{ij},\boldsymbol{\Omega}_i,\boldsymbol{\Omega}_j), |
| 386 |
> |
\label{eq:ssdPot} |
| 387 |
> |
\end{equation} |
| 388 |
> |
where the $\mathbf{r}_{ij}$ is the position vector between molecules |
| 389 |
> |
\emph{i} and \emph{j} with magnitude equal to the distance $r_{ij}$, and |
| 390 |
> |
$\boldsymbol{\Omega}_i$ and $\boldsymbol{\Omega}_j$ represent the |
| 391 |
> |
orientations of the respective molecules. The Lennard-Jones and dipole |
| 392 |
> |
parts of the potential are given by equations \ref{eq:lennardJonesPot} |
| 393 |
> |
and \ref{eq:dipolePot} respectively. The sticky part is described by |
| 394 |
> |
the following, |
| 395 |
> |
\begin{equation} |
| 396 |
> |
u_{ij}^{sp}(\mathbf{r}_{ij},\boldsymbol{\Omega}_i,\boldsymbol{\Omega}_j)= |
| 397 |
> |
\frac{\nu_0}{2}[s(r_{ij})w(\mathbf{r}_{ij}, |
| 398 |
> |
\boldsymbol{\Omega}_i,\boldsymbol{\Omega}_j) + |
| 399 |
> |
s^\prime(r_{ij})w^\prime(\mathbf{r}_{ij}, |
| 400 |
> |
\boldsymbol{\Omega}_i,\boldsymbol{\Omega}_j)]\ , |
| 401 |
> |
\label{eq:stickyPot} |
| 402 |
> |
\end{equation} |
| 403 |
> |
where $\nu_0$ is a strength parameter for the sticky potential, and |
| 404 |
> |
$s$ and $s^\prime$ are cubic switching functions which turn off the |
| 405 |
> |
sticky interaction beyond the first solvation shell. The $w$ function |
| 406 |
> |
can be thought of as an attractive potential with tetrahedral |
| 407 |
> |
geometry: |
| 408 |
> |
\begin{equation} |
| 409 |
> |
w({\bf r}_{ij},{\bf \Omega}_i,{\bf \Omega}_j)= |
| 410 |
> |
\sin\theta_{ij}\sin2\theta_{ij}\cos2\phi_{ij}, |
| 411 |
> |
\label{eq:stickyW} |
| 412 |
> |
\end{equation} |
| 413 |
> |
while the $w^\prime$ function counters the normal aligned and |
| 414 |
> |
anti-aligned structures favored by point dipoles: |
| 415 |
> |
\begin{equation} |
| 416 |
> |
w^\prime({\bf r}_{ij},{\bf \Omega}_i,{\bf \Omega}_j)= |
| 417 |
> |
(\cos\theta_{ij}-0.6)^2(\cos\theta_{ij}+0.8)^2-w^0, |
| 418 |
> |
\label{eq:stickyWprime} |
| 419 |
> |
\end{equation} |
| 420 |
> |
It should be noted that $w$ is proportional to the sum of the $Y_3^2$ |
| 421 |
> |
and $Y_3^{-2}$ spherical harmonics (a linear combination which |
| 422 |
> |
enhances the tetrahedral geometry for hydrogen bonded structures), |
| 423 |
> |
while $w^\prime$ is a purely empirical function. A more detailed |
| 424 |
> |
description of the functional parts and variables in this potential |
| 425 |
> |
can be found in the original SSD |
| 426 |
> |
articles.\cite{liu96:new_model,liu96:monte_carlo,chandra99:ssd_md,Ichiye03} |
| 427 |
> |
|
| 428 |
> |
Since SSD is a single-point {\it dipolar} model, the force |
| 429 |
> |
calculations are simplified significantly relative to the standard |
| 430 |
> |
{\it charged} multi-point models. In the original Monte Carlo |
| 431 |
> |
simulations using this model, Ichiye {\it et al.} reported that using |
| 432 |
> |
SSD decreased computer time by a factor of 6-7 compared to other |
| 433 |
> |
models.\cite{liu96:new_model} What is most impressive is that these savings |
| 434 |
> |
did not come at the expense of accurate depiction of the liquid state |
| 435 |
> |
properties. Indeed, SSD maintains reasonable agreement with the Soper |
| 436 |
> |
diffraction data for the structural features of liquid |
| 437 |
> |
water.\cite{Soper86,liu96:new_model} Additionally, the dynamical properties |
| 438 |
> |
exhibited by SSD agree with experiment better than those of more |
| 439 |
> |
computationally expensive models (like TIP3P and |
| 440 |
> |
SPC/E).\cite{chandra99:ssd_md} The combination of speed and accurate depiction |
| 441 |
> |
of solvent properties makes SSD a very attractive model for the |
| 442 |
> |
simulation of large scale biochemical simulations. |
| 443 |
> |
|
| 444 |
> |
Recent constant pressure simulations revealed issues in the original |
| 445 |
> |
SSD model that led to lower than expected densities at all target |
| 446 |
> |
pressures.\cite{Ichiye03,Gezelter04} The default model in {\sc oopse} |
| 447 |
> |
is therefore SSD/E, a density corrected derivative of SSD that |
| 448 |
> |
exhibits improved liquid structure and transport behavior. If the use |
| 449 |
> |
of a reaction field long-range interaction correction is desired, it |
| 450 |
> |
is recommended that the parameters be modified to those of the SSD/RF |
| 451 |
> |
model. Solvent parameters can be easily modified in an accompanying |
| 452 |
> |
{\sc BASS} file as illustrated in the scheme below. A table of the |
| 453 |
> |
parameter values and the drawbacks and benefits of the different |
| 454 |
> |
density corrected SSD models can be found in reference |
| 455 |
> |
\ref{Gezelter04}. |
| 456 |
> |
|
| 457 |
> |
\begin{lstlisting}[caption={[A simulation of {\sc ssd} water]An example file showing a simulation including {\sc ssd} water.},label={sch:ssd}] |
| 458 |
> |
|
| 459 |
> |
#include "water.mdl" |
| 460 |
> |
|
| 461 |
> |
nComponents = 1; |
| 462 |
> |
component{ |
| 463 |
> |
type = "SSD_water"; |
| 464 |
> |
nMol = 864; |
| 465 |
> |
} |
| 466 |
> |
|
| 467 |
> |
initialConfig = "liquidWater.init"; |
| 468 |
> |
|
| 469 |
> |
forceField = "DUFF"; |
| 470 |
> |
|
| 471 |
> |
/* |
| 472 |
> |
* The reactionField flag toggles reaction |
| 473 |
> |
* field corrections. |
| 474 |
> |
*/ |
| 475 |
> |
|
| 476 |
> |
reactionField = false; // defaults to false |
| 477 |
> |
dielectric = 80.0; // dielectric for reaction field |
| 478 |
> |
|
| 479 |
> |
/* |
| 480 |
> |
* The following two flags set the cutoff |
| 481 |
> |
* radius for the electrostatic forces |
| 482 |
> |
* as well as the skin thickness of the switching |
| 483 |
> |
* function. |
| 484 |
> |
*/ |
| 485 |
> |
|
| 486 |
> |
electrostaticCutoffRadius = 9.2; |
| 487 |
> |
electrostaticSkinThickness = 1.38; |
| 488 |
> |
|
| 489 |
> |
\end{lstlisting} |
| 490 |
> |
|
| 491 |
> |
|
| 492 |
> |
\subsection{\label{sec:eam}Embedded Atom Method} |
| 493 |
> |
|
| 494 |
> |
Several other molecular dynamics packages\cite{dynamo86} exist which have the |
| 495 |
> |
capacity to simulate metallic systems, including some that have |
| 496 |
> |
parallel computational abilities\cite{plimpton93}. Potentials that |
| 497 |
> |
describe bonding transition metal |
| 498 |
> |
systems\cite{Finnis84,Ercolessi88,Chen90,Qi99,Ercolessi02} have a |
| 499 |
> |
attractive interaction which models ``Embedding'' |
| 500 |
> |
a positively charged metal ion in the electron density due to the |
| 501 |
> |
free valance ``sea'' of electrons created by the surrounding atoms in |
| 502 |
> |
the system. A mostly repulsive pairwise part of the potential |
| 503 |
> |
describes the interaction of the positively charged metal core ions |
| 504 |
> |
with one another. A particular potential description called the |
| 505 |
> |
Embedded Atom Method\cite{Daw84,FBD86,johnson89,Lu97}({\sc eam}) that has |
| 506 |
> |
particularly wide adoption has been selected for inclusion in {\sc oopse}. A |
| 507 |
> |
good review of {\sc eam} and other metallic potential formulations was done |
| 508 |
> |
by Voter.\cite{voter} |
| 509 |
> |
|
| 510 |
> |
The {\sc eam} potential has the form: |
| 511 |
> |
\begin{eqnarray} |
| 512 |
> |
V & = & \sum_{i} F_{i}\left[\rho_{i}\right] + \sum_{i} \sum_{j \neq i} |
| 513 |
> |
\phi_{ij}({\bf r}_{ij}) \\ |
| 514 |
> |
\rho_{i} & = & \sum_{j \neq i} f_{j}({\bf r}_{ij}) |
| 515 |
> |
\end{eqnarray}S |
| 516 |
> |
|
| 517 |
> |
where $F_{i} $ is the embedding function that equates the energy required to embed a |
| 518 |
> |
positively-charged core ion $i$ into a linear superposition of |
| 519 |
> |
spherically averaged atomic electron densities given by |
| 520 |
> |
$\rho_{i}$. $\phi_{ij}$ is a primarily repulsive pairwise interaction |
| 521 |
> |
between atoms $i$ and $j$. In the original formulation of |
| 522 |
> |
{\sc eam} cite{Daw84}, $\phi_{ij}$ was an entirely repulsive term, however |
| 523 |
> |
in later refinements to EAM have shown that non-uniqueness between $F$ |
| 524 |
> |
and $\phi$ allow for more general forms for $\phi$.\cite{Daw89} |
| 525 |
> |
There is a cutoff distance, $r_{cut}$, which limits the |
| 526 |
> |
summations in the {\sc eam} equation to the few dozen atoms |
| 527 |
> |
surrounding atom $i$ for both the density $\rho$ and pairwise $\phi$ |
| 528 |
> |
interactions. Foiles et al. fit EAM potentials for fcc metals Cu, Ag, Au, Ni, Pd, Pt and alloys of these metals\cite{FDB86}. These potential fits are in the DYNAMO 86 format and are included with {\sc oopse}. |
| 529 |
> |
|
| 530 |
> |
|
| 531 |
> |
\subsection{\label{Sec:pbc}Periodic Boundary Conditions} |
| 532 |
> |
|
| 533 |
> |
\newcommand{\roundme}{\operatorname{round}} |
| 534 |
> |
|
| 535 |
> |
\textit{Periodic boundary conditions} are widely used to simulate truly |
| 536 |
> |
macroscopic systems with a relatively small number of particles. The |
| 537 |
> |
simulation box is replicated throughout space to form an infinite lattice. |
| 538 |
> |
During the simulation, when a particle moves in the primary cell, its image in |
| 539 |
> |
other boxes move in exactly the same direction with exactly the same |
| 540 |
> |
orientation.Thus, as a particle leaves the primary cell, one of its images |
| 541 |
> |
will enter through the opposite face.If the simulation box is large enough to |
| 542 |
> |
avoid \textquotedblleft feeling\textquotedblright\ the symmetries of the |
| 543 |
> |
periodic lattice, surface effects can be ignored. Cubic, orthorhombic and |
| 544 |
> |
parallelepiped are the available periodic cells In OOPSE. We use a matrix to |
| 545 |
> |
describe the property of the simulation box. Therefore, both the size and |
| 546 |
> |
shape of the simulation box can be changed during the simulation. The |
| 547 |
> |
transformation from box space vector $\mathbf{s}$ to its corresponding real |
| 548 |
> |
space vector $\mathbf{r}$ is defined by |
| 549 |
> |
\begin{equation} |
| 550 |
> |
\mathbf{r}=\underline{\mathbf{H}}\cdot\mathbf{s}% |
| 551 |
> |
\end{equation} |
| 552 |
> |
|
| 553 |
> |
|
| 554 |
> |
where $H=(h_{x},h_{y},h_{z})$ is a transformation matrix made up of the three |
| 555 |
> |
box axis vectors. $h_{x},h_{y}$ and $h_{z}$ represent the three sides of the |
| 556 |
> |
simulation box respectively. |
| 557 |
> |
|
| 558 |
> |
To find the minimum image of a vector $\mathbf{r}$, we convert the real vector |
| 559 |
> |
to its corresponding vector in box space first, \bigskip% |
| 560 |
> |
\begin{equation} |
| 561 |
> |
\mathbf{s}=\underline{\mathbf{H}}^{-1}\cdot\mathbf{r}% |
| 562 |
> |
\end{equation} |
| 563 |
> |
And then, each element of $\mathbf{s}$ is wrapped to lie between -0.5 to 0.5, |
| 564 |
> |
\begin{equation} |
| 565 |
> |
s_{i}^{\prime}=s_{i}-\roundme(s_{i}) |
| 566 |
> |
\end{equation} |
| 567 |
> |
where |
| 568 |
> |
|
| 569 |
> |
% |
| 570 |
> |
|
| 571 |
> |
\begin{equation} |
| 572 |
> |
\roundme(x)=\left\{ |
| 573 |
> |
\begin{array}{cc}% |
| 574 |
> |
\lfloor{x+0.5}\rfloor & \text{if \ }x\geqslant0\\ |
| 575 |
> |
\lceil{x-0.5}\rceil & \text{otherwise}% |
| 576 |
> |
\end{array} |
| 577 |
> |
\right. |
| 578 |
> |
\end{equation} |
| 579 |
> |
|
| 580 |
> |
|
| 581 |
> |
For example, $\roundme(3.6)=4$,$\roundme(3.1)=3$, $\roundme(-3.6)=-4$, $\roundme(-3.1)=-3$. |
| 582 |
> |
|
| 583 |
> |
Finally, we obtain the minimum image coordinates $\mathbf{r}^{\prime}$ by |
| 584 |
> |
transforming back to real space,% |
| 585 |
> |
|
| 586 |
> |
\begin{equation} |
| 587 |
> |
\mathbf{r}^{\prime}=\underline{\mathbf{H}}^{-1}\cdot\mathbf{s}^{\prime}% |
| 588 |
> |
\end{equation} |
| 589 |
> |
|
