1: %\documentclass[preprint,aps,showpacs]{revtex4}
2: \documentclass[twocolumn,pre,aps,showpacs]{revtex4}
3: \usepackage{graphicx}
4:
5: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
6: %%% Abbreviations %%%
7: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
8: \def\be{\begin{equation}}
9: \def\ee{\end{equation}}
10: \def\bea{\begin{eqnarray}}
11: \def\eea{\end{eqnarray}}
12:
13: \newcommand{\zbar}{\bar{Z}}
14: \newcommand{\bfr}{{\bf r}}
15:
16: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
17:
18:
19: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
20: \begin{document}
21: \title{Temperature Relaxation in Hot Dense Hydrogen}
22: \author{Michael S. Murillo$^1$ and M. W. C. Dharma-wardana$^2$}
23: %\email{murillo@lanl.gov}
24: \affiliation{$^1$Physics Division, Los Alamos National Laboratory, Los Alamos, New Mexico 87545\\
25: $^2$National Research Council, Ottawa, Canada K1A 0R6}
26: \email{chandre@argos.phy.nrc.ca}
27: \date{\today}
28:
29: \begin{abstract}
30: Temperature equilibration of hydrogen is studied for conditions relevant to inertial confinement
31: fusion. New molecular-dynamics simulations and results from quantum many-body theory are compared with
32: Landau-Spitzer (LS) predictions for temperatures $T$ from 50 eV to 5000 eV, and densities with Wigner-Seitz radii
33: $r_s = 1.0$ and $0.5$. The relaxation is slower than
34: the LS result, even for temperatures in the keV range, but converges to agreement in the
35: high-$T$ limit.
36: \end{abstract}
37:
38: \pacs{52.25.Kn,71.10.-w,52.27.Gr}
39: % 52.25.Kn Thermodynamics of plasmas
40: % 52.25.Gj Fluctuation and chaos phenomena
41: % 71.10.-w Theories and models of many-electron systems
42: % 52.27.Gr Strongly coupled plasmas
43:
44: \maketitle
45:
46:
47: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
48: {\it Introduction --}
49: While a first-principles description of the equilibrium properties of strongly coupled Coulomb systems
50: is a formidable task \cite{EOS}, nonequilibrium systems pose an even greater challenge. Short-pulse
51: lasers and shock waves create nonequilibrium states. Thus, Coulomb systems as diverse as warm dense
52: matter \cite{Riley}, ultracold plasmas \cite{Killian}, shocked semiconductors \cite{Ng}, and dense
53: deuterium \cite{Knudsen} can now be readily created in the laboratory, but initially under
54: nonequilibrium conditions. Similarly, energy relaxation (ER) in astrophysical plasmas is important
55: to the physics of fusion of H, C, etc., in determining stellar evolution\cite{deWitt}.
56: Here, we consider the ER of nonequilibrium dense hydrogen due to
57: its importance in inertial confinement fusion (ICF)\cite{ICF}.
58:
59: The earliest theories of ER in plasmas were formulated by Landau
60: \cite{Land} and Spitzer \cite{Spit} (denoted LS). The LS approach is applicable to
61: dilute, hot, fully ionized plasms where the collisions are weak, binary, and involve negligible
62: quantum effects; essentially, LS is Rutherford's Coulomb scattering formula applied to a Maxwellian
63: distribution. A characteristic feature of the LS approach is the
64: use of a Coulomb logarithm (CL), i.e.,
65: \be
66: {\cal L} \equiv \ln\Lambda \sim \int_{b_{min}}^{b_{max}}\!db/b \sim
67: \int_{k_{max}}^{k_{min}}\!dk/k
68: \ee
69: where $b_{min}$ (or $1/k_{max}$) and $b_{max}$ (or $1/k_{min}$) are suitable,
70: but {\it ad hoc}, impact parameter (or momentum) cutoffs for the Coulomb
71: collision. The full quantum mechanical method, based on calculating a transition rate,
72: does not suffer from this problem. Calculations at the Fermi golden-rule
73: level and beyond have been made by Dharma-wardana et al. \cite{DWP,mwcd, Hazak}.
74: Such methods automatically include degeneracy effects, effects of collective modes, and strong coupling.
75: Other approaches employ convergent kinetic equations
76: \cite{Gould_Dewitt,GMS}. Hansen and McDonald \cite{HM} (denoted HM) and Reimann and Toepffer have directly
77: obtained ER via molecular
78: dynamics (MD) simulations.
79:
80:
81: In a hydrogen plasma the particle charges $z_i,z_j$ are $\pm1$,
82: in atomic units, where the electronic charge $|e|$=$\hbar$=$m_e$=1.
83: The mean electron and proton densities $n$ and $\rho$ are identical.
84: The ratio of a typical Coulomb energy to the kinetic energy becomes, in the classical
85: regime, $\Gamma=1/(r_sT)$, where $T$ is the temperature in energy units, and $\Gamma=r_s$
86: in the quantum regime.
87: $r_s=\left[3/(4\pi n)\right]^{1/3}$ is the radius of the Wigner-Seitz sphere of an
88: electron or a proton.
89: The properties of partially degenerate fully-ionized plasmas
90: require two independent parameters, e.g., both $r_s$ and $\theta=T/E_F$,
91: where $E_F=\left(3\pi^2 n\right)^{2/3}/2$ is the
92: Fermi energy \cite{Murillo_tut}.
93: The LS analysis of ER is in terms of e-p
94: collisions in a Maxwellian gas.
95: Relative to the LS approach, some theoretical approaches relax faster \cite{GMS},
96: or slower \cite{DWP,Hazak}. HM concluded that many-body effects are negligible
97: and that the the LS result holds\cite{HM}.
98: There is currently no
99: direct experimental data for ER rates, although such experiments
100: are under way\cite{Taccetti,GaretaRiley}.
101:
102:
103: Here, we present larger HM-like molecular-dynamics (MD) simulations and
104: quantum many-body calculations to narrow the gap in our predictions of ER rates for hot, dense hydrogen
105: relevant to ICF targets composed of dense cryogenic fuel rapidly laser compressed to keV temperatures.
106: The modeling of these dense plasmas covers physical processes over many orders of magnitude in density
107: and temperature. Nonequilibrium quantum simulations require theoretical
108: breakthroughs that are not yet fully established; however, the MD techniques
109: that employs quantum-corrected effective potentials can be usefully applied to these problems.
110: Since MD simulations attempt to solve the many-body equations of motion exactly, they are
111: likely to provide accurate ER rates for hot, dense hydrogen.
112:
113: {\it Molecular dynamics.--}
114: Several issues arise in simulations of plasmas with temperatures in the
115: $10^2$-10$^3$ eV range. Because the screening length and mean-free-path
116: are larger at higher temperatures,
117: we have varied the number of particles widely (maximum of several thousand), with $N$=500 used for
118: the results presented here. Also, because the ER time varies roughly as $T_e^{3/2}$, these simulations
119: required millions of time steps for the higher-temperature cases. Compounding the longer runs was
120: the need for much smaller time steps (as small as $0.005\omega_{pe}^{-1}$, where $\omega_{pe}$ is the
121: usual electron plasma frequency), as a result of high velocity collisions at high temperature. Such
122: issues are not unexpected, but have rarely been dealt with, since MD is typically applied to strongly
123: coupled (i.e., cooler) classical systems. We carried out direct simulations of electrons
124: and protons since the Born-Oppenheimer approximation is not applicable to this problem. The
125: mass ratio ($1:1836$) was used due to our interest in mass-dependent collective mode effects.
126: Finally, integrations were carried out with the velocity-Verlet algorithm using an ${\cal O}(N^{3/2})$
127: Ewald method.
128:
129: The Coulomb interaction $1/r$ of classical physics is replaced by the mean-value of the
130: operator $1/\hat{r}$ in quantum systems. This feature modifies the short-ranged
131: behaviour of the electron-electron and electron-proton interactions, since the
132: de Broglie wavelength
133: of the electron is not negligible for small $r$.
134: Also, unlike the classical electron-proton
135: interaction which always leads to a bound state, the delocalized electron does not bind to
136: the proton for regimes studied here. These effects are included in the
137: e-e and e-p potentials by solving the relevant Schrodinger equations for the two-body scattering
138: processes. These lead to diffraction-corrected Coulomb potentials
139: $v_{ij}^{dfr}(r_{ij})$
140: which are Coulomb-like for
141: distances larger than the respective de Broglie lengths,
142: $\lambda_{ij}=1/\sqrt{2\pi\mu_{ij}T_e}$.
143: Here $\mu_{ij}$ are the effective masses of the colliding pair.
144: The choice of $b_{min}=\lambda_{ij}$ is usal in
145: the LS approaches.
146:
147: Then interaction $\phi_{ij}(r)$ between the pair $i,j$ is given by
148: \bea
149: \label{phys_mod_dfr}
150: \phi_{ij}&=&v_{ij}^{dfr}(r_{ij})+v_{ee}^{Pau}(r)\\
151: v_{ij}^{dfr}(r_{ij})& =& \frac{z_iz_j}{r_{ij}}\left[1-e^{\left(-\frac{f_{ij}r_{ij}}{
152: \lambda_{ij}} \right)} \right] + v^{Ewl}_{ij}(r_{ij})
153: \eea
154: where $v^{Ewl}_{ij}$ is the Ewald potential.
155: The potential, $v_{ee}^{Pau}$,
156: accounts for the spin-averaged Pauli exclusion between two electrons.
157: This ensures that the ``non-interacting'' electron pair distribution function (PDF)
158: calculated from a classical simulation
159: is just the non-interacting quantum PDF \cite{Lado, CHNC}.
160: The explicit form\cite{JonesMurillo} used by HM is adequate
161: for most of the range studied
162: in this paper.
163: %\be
164: %\label{phys_mod_Pauli}
165: %v_{ee}^{Pau}(r_{ee}) =
166: % -T_e\ln\left[1-\frac{1}{2}\exp\left(-\frac{r_{ee}^2}{\pi\lambda^2_{ee}}
167: % \right) \right].
168: %\ee
169: %This potential of Eq.~\ref{phys_mod_dfr}, (Model I), with $f_{ij}=1$,
170: %is the diffraction-corrected potential
171: %\cite{JonesMurillo} (used by HM). Model II is
172: %the complete Pauli potential \cite{Lado}.
173: %Model II is essential for small $\theta<1$,
174: %and is the same as the Pauli potential implemented
175: %in the classical-map hypernetted-chain (CHNC)
176: %approach \cite{CHNC}.
177: The factor $f_{ij}$ is unity for all except the
178: cross-species case $f_{ep}$. This is chosen using CHNC.
179: The CHNC uses the above potentials and a classical fluid
180: temperature $T_{cf}=\surd{T_e^2+T_q^2}$
181: where the quantum temperature $T_q$ is defined in Ref.~\cite{CHNC}.
182: The $f_{ep}$ factor at a given $r_s,\, T_e,\, T_p$
183: is fixed by requiring that the CHNC $g_{ep}(r)$, at $r=0$
184: is the same as the Kohn-Sham value of $g_{ep}(r=0)$, as discussed
185: more fully in Ref.~\cite{DW-Murillo}.
186: The factors $f_{ep}$ are within 5\% of unity for the
187: conditions of our study.
188:
189:
190: To validate the potentials under two-temperature, dense-plasma conditions, we
191: carried out (using $\phi_{ij}$ and its simplified HM forms)
192: MD and CHNC for the conditions $r_s=1.0$,
193: $T_e=50$eV, and $T_i=10$eV. In the MD, electron and proton velocity-scaling thermostats
194: were used to
195: create the two-temperature system, whereas recent results for the cross-species
196: temperature $T_{ep}$
197: \cite{DW-Murillo} were used in the CHNC. The pair distribution
198: functions $g_{ij}(r)$ are shown in
199: Fig. (\ref{rdf_fig}).
200:
201:
202:
203: \begin{figure}[t]
204: %\includegraphics[angle=-90, width=3in]{Plots/RDF_Plots/rdf_1}
205: %\includegraphics[angle=0, width=3in]{Plots/RDF_Plots/t50}
206: \includegraphics[angle=0, width=3in]{t50.eps}
207: \caption{Pair distribution functions from MD (data points), CHNC (thick lines)
208: for dense hydrogen, $r_s=0.5$ and 1, $T/E_F=$0.25 and 1. The $g_{ee}$
209: from a simple HNC calculation (i.e., no quantum temperature, no Pauli potential)
210: with just $v_{ij}^{dfr}$ is also shown for $r_s=1$.}
211: \label{rdf_fig}
212: \end{figure}
213:
214:
215: Temperature equilibration rates were determined for the two densities $r_s=0.5$
216: ($n_e=1.5\times 10^{25}$cm$^{-3}$) and $r_s=1.0$ ($n_e=1.61\times 10^{24}$cm$^{-3}$) over
217: the temperature range $0.25<\theta<20$. In practice, an equilibration
218: stage with separate electron and proton velocity-scaling thermostats
219: was used to establish a
220: two-temperature system. This was followed by a microcanonical evolution
221: in which the temperatures
222: were relaxed, with
223: \be
224: T_j(t)=\frac{m_j}{3N_j}\sum_{i=1}^{N_j} v_j^2(t),
225: \ee
226: for each species $j$. Energy conservation was carefully monitored to assure stability
227: at the elevated temperatures.
228: Stability was quantified by $\Delta E = \frac{1}{N}\sum_{i=1}^N\left|\frac{E_i-E_0}{E_0}\right|$,
229: where $E_i$ is the energy at the $i$-th timestep. The timestep was chosen by first performing
230: several simulations with varying timesteps for temperatures $T=100, 250, 500, 750$eV and
231: noting the impact on $\Delta E$. As mentioned above, the timestep required for stability
232: decreases dramatically with increasing temperature. A fit to the slope of the temperature profiles
233: yielded the equilibration rate; in practice, the ion temperature was fixed for all runs at
234: $T_i=10$eV so that the reported results are $\partial T_e/\partial t$. Over the range of
235: temperatures considered, the relaxation rate should be very insensitive to the ion temperature.
236:
237: {\it Quantum transition rates.--}
238: The most transparent, strictly quantum approach to the calculation of the
239: ER rate is to treat it as a
240: transition rate where an electron in an initial momentum state
241: $\vec{k}_i$ transfers to a final state
242: $\vec{ k}_f$, while a proton in the initial momentum
243: eigenstate $\vec{p}_i$ absorbs energy and transfers to a final
244: eigenstate $\vec{p}_f$. The availability of such states depends on the products of Fermi
245: occupation factors $n_{k_i}(1-n_{k_f})$,
246: and similarly for the proton states. The strength of the transition depends
247: on the matrix element between
248: the initial and final states. This matrix element may be taken in
249: lowest-order theory (Born approximation)
250: or in higher order (i.e, a $T$-matrix evaluation).
251: These are the usual ingredients of the Fermi golden rule (FGR) for the
252: transition rate. The summation over all such pair processes
253: gives the total ER rate. But such summations
254: immediately convert the description of the plasma into a
255: description containing the full
256: spectrum of single-particle and collective modes. The spectrum of
257: all modes is given by the spectral
258: function $A_j(q,\omega,T_j)$ where the species index $j=e$ or $p$.
259: These spectral functions are given by the imaginary parts of the
260: corresponding dynamic response functions $\chi^{j}(\vec{k},\omega)$, e.g,
261: Eq.~(16) of Ref.~\cite{Hazak}.
262: The ER rate evaluated within the Fermi golden rule, $R_{fgr}$
263: can be expressed in terms of
264: the response functions of the plasma as follows, given in Eqs. (4)-(7)
265: of Ref.\cite{mwcd}, and Eq.~(15) of Ref.~\cite{Hazak}:
266: %\bea
267: %A_j(q,\omega,T_j)&=&-2\Im \chi_{jj}(q,\omega,T_j)\\
268: %\chi_{jj}(q,\omega,T_j)&=&\chi_{jj}^0/\{1-V_{jj}(q)[1-G_{jj}(q,\omega)\chi_{jj}^0]\}\\
269: %R_{fgr}&=&\Sigma_{q,\omega}|V_{ep}|^2\omega A^e(q,\omega)
270: %A^p(q,\omega)\Delta N_{ep}(\omega)\\
271: %\Delta N_{ep}(\omega)&=&N(\omega/T_e)-N(\omega/T_p)\\
272: %N(\omega/T_j)&=&1/[exp^{\omega/T_j}-1]
273: %\eea
274: \bea
275: \label{cothform}
276: R_{fgr}&=&\frac{dK}{dt}=\int \frac{d^{3}k}{\left( 2\pi \right) ^{3}}
277: \frac{\omega d\omega}
278: {2\pi}(\Delta B) F_{ep}\\
279: \Delta B&=&\coth(\omega/2T_e)-\coth(\omega/2T_p)\\
280: %&\left[ \coth \left( \frac{ \omega }{2T_{e}}\right) -
281: %\coth
282: %\left( \frac{ \omega }{2T_{p}}\right) \right]\\
283: F_{ep}&= &|( V_{ep}(k)
284: | ^{2}\Im\left[ \chi ^{p}(\vec{k},\omega )\right] \Im
285: \left[ \chi ^{e}\left( \vec{k},\omega \right) \right]
286: \eea
287: In the above we have used the spherical symmetry of the plasma
288: to write scalars $q,k$ instead of
289: $\vec{q},\, \vec{k} $ to simplify the notation.
290: %
291: \begin{figure}[t]
292: \includegraphics[angle=0, width=3.3in]{tau.eps}
293: \caption{The relaxation time $\tau/1000$,
294: in units of the inverse electon-plasma frequency,
295: for dense hydrogen, $r_s=0.5$ and 1, from the
296: degenerate to the classicle region. The Landau-Spitzer (LS) result
297: using the Hansen-McDonald(HM) prescription for the $k$-cutoffs,
298: the results from the Fermi Golden rule(FGR), and the coupled-mode(CM)
299: are shown, together with the MD results.
300: }
301: \label{taugraph}
302: \end{figure}
303: %
304:
305: The excess-energy density is denoted by $K=K_e-K_i$, and becomes $3(T_e-T_p)n/2$,
306: in the classical regime, i.e., where the chemical potential $\mu$ is negative.
307: The interaction $V_{ep}(k)$ in this equation is the full Coulomb matrix element
308: and {\em not} the diffraction corrected form used in the CHNC and the
309: classical simulations. In the simplest form of the FGR, $V_{ep}(k)=4\pi/k^2$ since
310: the momentum states are taken to be plane waves. A $T$-matrix evaluation would use
311: phase-shifted plane waves and the corresponding modified density of states, instead
312: of $d^{3}k/\left( 2\pi \right) ^{3}$.
313: With the onset of the classical regime where
314: $\mu<0$, which occurs for $\theta>1$, the $\Delta B$ factor
315: becomes $2\Delta/\omega$ where $\Delta =(T_e-T_p)$. It was shown in
316: Hazak et al.\cite{Hazak} how we may do the $\omega$-integration by
317: exploiting the $f$-sum rule and the fact that the ion-spectral function,
318: peaking near the ion-plasma frequency, resides way below the electron
319: spectral function. Then Eq.~\ref{cothform} can be written, to a good approximation
320: as:
321: \bea
322: \frac{1}{\Delta}\frac{d\Delta}{dt}&=&\frac{2}{3n}\omega _{ion}^{2}\int_0^\infty\frac{2}{\pi}
323: \left[ \frac{\partial }{\partial \omega }\Im\chi ^{ee}\left(
324: k,\omega \right) \right] _{\omega =0}dk
325: \label{fgrBA}
326: \eea
327: where $\omega _{ion}$ is the proton-plasma frequency. If we keep the proton
328: temperature $T_p$ fixed, we see that Eq.~\ref{fgrBA} leads to a
329: relaxation time $\tau$ for the electron temperature
330: $T_e$, involving the inverse of the r.h.s. of Eq.~\ref{fgrBA}.
331:
332: The above analysis treats the plasma as two independent subsystems. In reality,
333: the ion-density fluctuations are screened by the electron subsystem, and the
334: ion-plasma mode becomes an ion-acoustic mode. The excitations in the
335: coupled-mode system are described by the zeros of Eq.~(45) of Ref.~\cite{DWP}.
336: In the static, $k\to0$ limit this denominator converts the electron screening
337: parameter $k^e_{DH}$ to $\surd\{(k^e_{DH})^2+(k^p_{DH})^2\}$. However, the
338: proton-density modes act dynamically in the relaxation process.
339: Thus the use of static ion screening
340: is incorrect. The coupled-mode(CM) approach is fully dynamical and
341: includes another denominator,
342: \be
343: \label{dcm}
344: d_{cm}=|1-V_{ep}^2(k)\chi ^{ee}\left(k,\omega \right)\chi ^{pp}\left(k,\omega \right)|^2
345: \ee
346: into the integrand in Eq.~\ref{cothform}. That is, $F_{ep}$ in Eq.~\ref{cothform} is
347: replaced by $F_{ep}/d_{cm}$.
348:
349: The simple Landau-Spitzer form can also be be written in the same form as Eq.~\ref{fgrBA},
350: as shown in Ref.~\cite{Hazak}.
351: The quantum approaches in CM and FGR automatically contain the diffraction and screening
352: effects. Thus, while Eq.~\ref{fgrBA} use the full integration $0\to\infty$,
353: LS needs the cutoffs $k_{min}$ and $k_{max}$ to obtain a convergent result.
354: The calculated LS-values of $\tau$ does depend somewhat
355: on the choice of $k_{min}$ and $k_{max}$. Hence
356: different realizations
357: of the LS-form need not reduce to the same result at finite $T/E_F$.
358: In fact Lee and More\cite{leemore} use cutoffs based on the full
359: static screening length which includes the ions as well and
360: differ significantly from LS.
361: However, the CM analysis clearly
362: shows that the ion response in ER is dynamic.
363:
364: The non-interacting response function $\chi^0(q,\omega,T)$ at arbitrary
365: degeneracies was given by Khanna and Glyde\cite{KH}. We use a generalized RPA form
366: where local filed corrections $G_{ee}(k)$ may be included\cite{DWP}. However, these
367: are quite small for the conditions of this study.
368: Both the FGR and CM calculations assume that linear response can be used to
369: discuss the interaction of a proton with the electrons. The resulting
370: calculations are shown in Fig.~\ref{taugraph}.
371: %
372: The coupled-mode (CM) calculation is
373: quite close to the Fermi golden rule (FGR)
374: f-sum result. This is expected since
375: the H-plasmas considered here are relatively
376: weakly coupled, with $\Gamma<1$. Nevertheless, the inclusion of CM leads to
377: better agreement with the MD simulation.
378: Also, we have assumed the bare $4\pi/k^2$ form for the $V_{ep}$ in the
379: FGR and CM formulae,
380: without the moderating effects of a pseudo-potential. Such effects
381: would tend to make the $\tau$ larger than that from the present FGR or CM calculation.
382: This linear-response assumption is more satisfactory for the $r_s=0.5$ plasma.
383: Thus the MD results at $r_s=0.5$ are very close to the CM results.
384: %
385:
386: {\it Conclusion.--} We have evaluated the temperature relaxation time in hot,
387: dense hydrogen using
388: state-of-the-art molecular dynamics simulations and quantum
389: many-body theory. We find that the relaxation is slower than the LS value even for
390: temperatures in the kilovolt range, which suggests that burning plasmas are slightly more out
391: of equilibrium that might have been expected. Unlike in the calculations presented in, say,
392: Ref.~\cite{DWP,mwcd}, where strongly coupled Al-plasmas were considered, the present calculations
393: are for systems with $\Gamma \sim 1$ or less.
394: The temperatures have been pushed to $T/E_F\simeq20$. Thus we see
395: that the CM, FGR and the LS forms converge for sufficiently large $T/E_F$.
396: The MD results are slightly higher than from the analytical models which
397: use linear response. It is also clear that
398: the LS form is inadequate for highly compressed low-$T$,
399: partially degenerate plasmas.
400:
401: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
402: \begin{thebibliography}{99}
403: \bibitem{EOS} P. Umari, A. J. Williamson, G. Galli, and N. Marzari, {\it Phys. Rev. Lett.}
404: {\bf 95}, 207602 (2005); C. Pierleoni, D. M. Ceperley, and M. Holzmann, {\it Phys. Rev. Lett.}
405: {\bf 93}, 146402 (2004); C. Dharma-wardana and F. Perrot, Phys. Rev. B {\bf 66}, 014110 (2002) .
406: \bibitem{Riley} D. Riley, N. C. Woolsey, D. McSherry, I. Weaver, A. Djaoui, and E. Nardi,
407: {\it Phys. Rev. Lett.} {\bf 84}, 1704 (2000).
408: \bibitem{Killian} Y. C. Chen, C. E. Simien, S. Laha, P. Gupta, Y. N. Martinez, P. G. Mickelson,
409: S. B. Nagel, T. C. Killian, {\it Phys. Rev. Lett.} {\bf 93}, 265003 (2004).
410: \bibitem{Ng} P. Celliers, A. Ng, G. Xu, and A. Forsman, {\it Phys. Rev. Lett.} {\bf 68}, 2305 (1992).
411: \bibitem{Knudsen} M. D. Knudson, D. L. Hanson, J. E. Bailey, C. A. Hall, and J. R. Asay,
412: {\it Phys. Rev. Lett.} {\bf 90}, 035505 (2003).
413: \bibitem{deWitt}
414: A.I. Chugunov, H. E. DeWitt, and D. G. Yakovlev, Phys. Rev. D {\bf 76}, 025028 (2007);
415: F. A. Agronyan and R. A. Syunyaev,
416: Astrophysics {\bf 27}, 413-422; Translated from Astrofizika; 27: No. 1, 131-145 ( 1987)
417: \bibitem{ICF} R. Linford, R. Betti, J. Dahlburg, J. Asay, M. Campbell, Ph. Colella, J. Freidberg, J.
418: Goodman, D. Hammer, J. Hoagland, S. Jardin, J. Lindl, G. Logan, K. Matzen, G. Navratil, A. Nobile,
419: J. Sethian, J. Sheffield, M. Tillack and J. Weisheit, {\it J. Fusion Energy} {\bf 22}, 93 (2003).
420: \bibitem{Land} L. D. Landau, JETP {\bf 7}, 203 (137); E. M. Lifshitz and L. P. Pitaevskii, {\sf Physical
421: Kinetics} (Pergamon, Oxford, 1981).
422: \bibitem{Spit} L.\ Spitzer, {\sf Physics of Fully Ionized Gases} (Interscience, New York, 1967).
423: \bibitem{DWP} M. W. C. Dharma-wardana and F. Perrot, {\it Phys. Rev. E} {\bf 58}, 3705 (1998);
424: {\it Phys. Rev. E} {\bf 63}, 069901 (2001).
425: \bibitem{mwcd}
426: M. W. C. Dharma-wardana, Phys. Rev. E {\bf 64} 035401 (2001)
427: \bibitem{Hazak} G. Hazak, Z. Zinamon, Y. Rosenfeld, and M. W. C. Dharma-wardana, {\it Phys.
428: Rev. E} {\bf 64}, 066411 (2001).
429: \bibitem{Gould_Dewitt} H. A. Gould and H. E. DeWitt, {\it Phys. Rev.} {\bf 155}, 68 (1967).
430: \bibitem{GMS} D. O. Gericke, M. S. Murillo, and M. Schlanges, {\it Phys. Rev. E} {\bf 65}, 036418
431: (2002).
432: \bibitem{HM} J. P. Hansen and I. R. McDonald, {\it Phys. Lett.} {\bf 97A}, 42 (1983).
433: \bibitem{RT} U. Reimann and C. Toepffer, {\it Laser and Part. Beams} {\bf 8}, 771 (1990).
434: \bibitem{Murillo_tut} M. S. Murillo, {\it Phys. Plasmas} {\bf 11}, (2004).
435: \bibitem{Taccetti} J. M. Taccetti et al., {\it J. Phys. A: Math Gen.} {\bf 39}, 4347 (2006).
436: \bibitem{GaretaRiley} J. J. Angulo Gareta and D. Riley, {\it High Energy Density Physics} {\bf 2}, 83
437: (2006).
438: \bibitem{Lado} F. Lado, {\it J. Chem. Phys.} {\bf 47}, 5369 (1967).
439: \bibitem{CHNC} M. W. C. Dharma-wardana and F. Perrot, {\it Phys. Rev. Lett.} {\bf 84}, 959 (2000).
440: \bibitem{JonesMurillo} C. S. Jones and M. S. Murillo, {\it High Energy Density Physics} {\bf 3},
441: 379 (2007).
442: \bibitem{DW-Murillo} M. W. C. Dharma-wardana and M. S. Murillo, arXiv:0710.2888v1
443: [cond-mat.stat-mech]
444: {\it Phys. Rev. E}, submitted.
445: \bibitem{leemore}
446: Y. T. Lee and R. M. More, Phys. Fluids {\bf 27}, 1273 (1984)
447: \bibitem{KH}
448: F. C. Khanna and H. R. Glyde, Can .J. Phys. {\bf 54}, 648 (1978);
449:
450: \end{thebibliography}
451:
452:
453: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
454: %%% FIGURE CAPTIONS %%%
455: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
456:
457: \end{document}
458:
459: