| 56 |
|
Decomposition of the molecular orientational time correlation function showed |
| 57 |
|
different behavior for the short- and longer-time decay components approaching |
| 58 |
|
normal to the interface. Lastly we show through calculation of the interfacial |
| 59 |
< |
friction coefficient that the basal and pyramidal facets are more |
| 59 |
> |
friction coefficient and dynamic water contact angle measurement |
| 60 |
> |
that the basal and pyramidal facets are more |
| 61 |
|
hydrophilic than the prismatic and secondary prismatic facets.} |
| 62 |
|
\end{abstract} |
| 63 |
|
|
| 64 |
|
\keywords{ice|water|interface|contact angle|molecular dynamics} |
| 65 |
|
|
| 66 |
< |
\abbreviations{QLL, quasi liquid layer; MD, molecular dynamics} |
| 66 |
> |
%\abbreviations{QLL, quasi liquid layer; MD, molecular dynamics} |
| 67 |
|
|
| 68 |
|
%\dropcap{I}n this article we study the evolution of ``almost-sharp'' fronts |
| 69 |
|
%for the surface quasi-geostrophic equation. This 2-D active scalar |
| 179 |
|
apparent independence of water model on interfacial width, we have chosen to use the rigid SPC/E |
| 180 |
|
water model in this study. |
| 181 |
|
|
| 182 |
< |
\subsection{Pyramidal and secondary prismatic system construction} |
| 182 |
> |
\subsection{Pyramidal and secondary prismatic ice/water interface construction} |
| 183 |
|
To construct the pyramidal and secondary prismatic ice/water systems, |
| 184 |
|
first a proton-ordered zero dipole crystal of ice-I$_\mathrm{h}$ with exposed strips |
| 185 |
|
of H-atoms and lone pairs was constructed from Structure 6 of Hirsch |
| 194 |
|
three times that of the ice block. Each of the ice slabs and water |
| 195 |
|
boxes were independently equilibrated to 50K, and the resulting |
| 196 |
|
systems were merged by carving out any liquid water molecules within 3 |
| 197 |
< |
\AA\ of any atoms in the ice slabs. For a more detailed explanation of |
| 197 |
> |
\AA\ of any atoms in the ice slabs. Each of the combined ice/water |
| 198 |
> |
systems were then equilibrated to 225K, which was found to be a stable |
| 199 |
> |
temperature for each of the interfaces over a 5 ns simulation. |
| 200 |
> |
For a more detailed explanation of |
| 201 |
|
the ice/water systems construction, please refer to a previous |
| 202 |
|
paper\cite{Louden13}. The resulting dimensions, number of ice, and liquid water molecules |
| 203 |
|
contained in each of these systems can be seen in Table \ref{tab:method}. |
| 205 |
|
% Do we need to justify the sims at 225K? |
| 206 |
|
% No crystal growth or shrinkage over 2 successive 1 ns NVT simulations for |
| 207 |
|
% either the pyramidal or sec. prismatic ice/water systems. |
| 208 |
+ |
To perform the shearing simulations, the velocity shearing and scaling |
| 209 |
+ |
varient of reverse nonequilibrium molecular dynamics (VSS-RNEMD) was |
| 210 |
+ |
conducted. This method performs a series of simultaneous velocity |
| 211 |
+ |
exchanges between two regions of the simulation cell, to |
| 212 |
+ |
simultaneously create a velocity and temperature gradient. The thermal |
| 213 |
+ |
gradient is necessary when performing shearing simulations as to |
| 214 |
+ |
prevent frictional heating from the shear from melting the |
| 215 |
+ |
interface. For more details on the VSS-RNEMD method please refer to a |
| 216 |
+ |
pervious paper\cite{Louden13}. |
| 217 |
|
|
| 218 |
|
The computational details performed here were equivalent to those reported |
| 219 |
< |
in our previous publication\cite{Louden13}. The only changes made to the |
| 220 |
< |
previously reported procedure were the following. VSS-RNEMD moves were |
| 208 |
< |
attempted every 2 fs instead of every 50 fs. This was done to minimize |
| 219 |
> |
in a previous publication\cite{Louden13}, with the following changes. |
| 220 |
> |
VSS-RNEMD moves were attempted every 2 fs instead of every 50 fs. This was done to minimize |
| 221 |
|
the magnitude of each individual VSS-RNEMD perturbation to the system. |
| 210 |
– |
|
| 222 |
|
All pyramidal simulations were performed under the canonical (NVT) ensamble |
| 223 |
< |
except those |
| 213 |
< |
during which statistics were accumulated for the orientational correlation |
| 223 |
> |
except those during which configurations were accumulated for the orientational correlation |
| 224 |
|
function, which were performed under the microcanonical (NVE) ensamble. All |
| 225 |
< |
secondary prismatic |
| 216 |
< |
simulations were performed under the NVE ensamble. |
| 225 |
> |
secondary prismatic simulations were performed under the NVE ensamble. |
| 226 |
|
|
| 227 |
|
\subsection{Droplet simulations} |
| 228 |
< |
Here, we will calculate the contact angle of a water droplet as it spreads |
| 229 |
< |
across each of the four ice I$_\mathrm{h}$ crystal facets in order to |
| 230 |
< |
determine the surface's relative hydrophilicites. The ice surfaces were |
| 231 |
< |
oriented so that the desired facet was exposed to the positive z dimension. |
| 232 |
< |
The sizes and number of molecules in each of the surfaces is given in Table |
| 228 |
> |
To construct ice surfaces to perform water contact angle calculations |
| 229 |
> |
on, ice crystals were created as described earlier (see Pyramidal and |
| 230 |
> |
secondary prismatic ice/water interface construction). The crystals |
| 231 |
> |
were then cut from the negative $z$ dimension, ensuring the remaining |
| 232 |
> |
ice crystal was thicker in $z$ than the potential cutoff. The crystals |
| 233 |
> |
were then replicated in $x$ and $y$ until a sufficiently large surface |
| 234 |
> |
had been created. The sizes and number of molecules in each of the surfaces is given in Table |
| 235 |
|
\ref{tab:ice_sheets}. Molecular restraints were applied to the center of mass |
| 236 |
|
of the rigid bodies to prevent surface melting, however the molecules were |
| 237 |
< |
allowed to reorient themselves freely. The water doplet to be placed on the |
| 238 |
< |
surface contained 2048 SPC/E molecules, which has been found to produce |
| 237 |
> |
allowed to reorient themselves freely. The water doplet contained 2048 |
| 238 |
> |
SPC/E molecules, which has been found to produce |
| 239 |
|
agreement for the Young contact angle extrapolated to an infinite drop |
| 240 |
< |
size\cite{Daub10}. The surfaces and droplet were equilibrated to 225 K, at |
| 241 |
< |
which time the droplet was placed 3-5 \AA\ above the surface at 5 unique |
| 242 |
< |
locations. Each simulation was 5 ns in length and conducted in the NVE |
| 243 |
< |
ensemble. |
| 240 |
> |
size\cite{Daub10}. The surfaces and droplet were independently |
| 241 |
> |
equilibrated to 225 K, at which time the droplet was placed 3-5 \AA\ |
| 242 |
> |
above the positive $z$ dimension of the surface at 5 unique |
| 243 |
> |
locations. Each simulation was 5 ns in length and conducted in the NVE ensemble. |
| 244 |
|
|
| 245 |
|
|
| 246 |
|
\section{Results and discussion} |
