1: \documentclass{article}%
2: \usepackage{amssymb}
3: \usepackage{graphics}%
4: \usepackage{amsmath}%
5: \setcounter{MaxMatrixCols}{30}%
6: \usepackage{amsfonts}%
7: \usepackage{graphicx}
8: %TCIDATA{OutputFilter=latex2.dll}
9: %TCIDATA{Version=5.00.0.2606}
10: %TCIDATA{LastRevised=Tuesday, May 10, 2005 13:38:54}
11: %TCIDATA{<META NAME="GraphicsSave" CONTENT="32">}
12: %TCIDATA{<META NAME="SaveForMode" CONTENT="1">}
13: %TCIDATA{BibliographyScheme=Manual}
14: %TCIDATA{Language=American English}
15: \begin{document}
16:
17: \title{Continuum percolation and the liquid--solid coexistence line of simple fluids.}
18: \author{Luis A. Pugnaloni\thanks{E-mail: luis@iflysib.unlp.edu.ar}, Marcos G. Valluzzi
19: \and and Fernando Vericat\\Instituto de F\'{\i}sica de L\'{\i}quidos y Sistemas Biol\'{\i}gicos,\\UNLP-CONICET - cc. 565, 1900 La Plata, Argentina}
20: \maketitle
21:
22: \begin{abstract}
23: We compare the percolation loci for chemical clusters with the liquid--solid
24: transition in the temperature--density phase diagram. Chemical clusters are
25: defined as sets of particles connected through particle-particle bonds that
26: last for a given time $\tau$. By using molecular dynamics simulations of a
27: Lennard--Jones system we obtain the percolation loci at different values of
28: $\tau$ as the lines in the temperature--density plane at which the system
29: presents a spanning cluster in 50 percent of the configurations. We find that
30: the percolation loci for chemical clusters shifts rapidly towards high
31: densities as $\tau$ is increased. For moderate values of $\tau$ this line
32: coincides with the low-density branch of the liquid--solid coexistence curve.
33: This implies that no stable chemical clusters can be found in the fluid phase.
34: In contrast, the percolation loci for physical clusters ---sets of particles
35: that remain close together at every instant for a given period $\tau$--- tends
36: to a limiting line, as $\tau$ tends to infinity, which is far from the
37: liquid--solid transition line.
38:
39: \end{abstract}
40:
41: \section{Introduction}
42:
43: Clustering and percolation in continuum systems are concepts of great interest
44: in many areas of physics. See for example ref.
45: \cite{Stauffer,Senger1,Starr1,Chen1,Butler1,Pugnaloni0,Pugnaloni0b}. In a
46: many-body disordered system, the central problem in continuum percolation
47: consists in determining the critical density (for a given temperature) at
48: which the system presents a cluster that spans the system volume. This, in
49: turn, leads to the crucial question of what is a cluster. The answer to this
50: question is many-sided. A cluster is a set of particles that can be considered
51: as a unit separated from the rest of the system and that can manifest itself
52: in an experiment. What we define as a cluster in a given system depends on the
53: properties of interest. Therefore, there is a degree of arbitrariness in the
54: cluster definition. The main premise here is that any proposed definition
55: allows a separation of the system into a number of disjoint sets of particles
56: which is appropriate to describe the phenomenon under study. For example, if
57: one intends to explain an insulator--conductor transition, and the particles
58: of the system are the conducting objects, we should call cluster to any set of
59: contacting particles that is not in contact with any other particle in the
60: system \cite{Chen1}. A good example is also found in deposited granular matter
61: where clusters correspond to mutually stabilizing sets of particles called
62: arches \cite{Pugnaloni0,Pugnaloni0b}.
63:
64: The first physically sound cluster definition for a simple fluid was given by
65: Hill \cite{Hill1} by contrast with the mathematical clusters introduced by
66: Mayer \cite{Mayer1}. In Hill's theory, the concept of cluster is directly
67: related to the idea of \textit{bonded pairs}. A bonded pair is a set of two
68: particles that are linked by some direct mechanism. Then, a cluster is defined
69: as a set of particles such that any pair of particles in the set is connected
70: through a path of bonded pairs. The definition of bonded pair becomes then the
71: essential part of the theory and a number of definitions have been used. Among
72: others, we can mention the energetic criterion and the geometric criterion for
73: bonded pairs proposed by Hill himself \cite{Hill1}. In the first case, two
74: particles are considered bonded at a given time if the negative of the pair
75: potential energy is higher than the relative kinetic energy of the pair (see
76: for example refs. \cite{Campi1} and \cite{Pugnaloni4}). In the second case,
77: two particles are bonded at a given time if they are within a distance $d$.
78: These later geometrical clusters are referred to as Stillinger's clusters (see
79: for example refs. \cite{Stillinger1} and \cite{Heyes1}).
80:
81: In a previous paper \cite{Pugnaloni1}, we have introduced a new definition of
82: cluster based on Hill's theory where the bonded pairs are defined as particles
83: that remain within a certain distance $d$ during a time interval $\tau$. We
84: call these clusters \textit{chemical clusters}\ since the stability
85: requirement on each connection remind us of a covalent bond. Note that a very
86: similar dynamic definition was advanced previously by Bahadur and McClurg
87: \cite{Bahadur1}. In the same paper, we have also presented a different
88: approach to cluster identification where the concept of bonded pairs is
89: meaningless. In this case, we identify as a cluster any set of particles that
90: remain interconnected (in the Stillinger's sense) at every instant over a time
91: interval $\tau$. Each particle within a given cluster may move around without
92: forming a stable bond with any other particle whereas it stays connected to
93: the group. We have termed these later clusters \textit{physical clusters} for
94: they exist without the presence of specific particle--particle bonds.
95:
96: Some preliminary studies of the properties of chemical and physical clusters
97: in Lennard--Jones systems where presented in ref. \cite{Pugnaloni1}. Moreover,
98: an integral equation theory was presented there ---and later on successfully
99: used \cite{Zarragoicoechea1}--- to calculate the cluster pair correlation
100: function for chemical clusters. Also, a first theoretical attempt to treat the
101: case of physical clusters was advanced in ref. \cite{Pugnaloni1}. In this
102: paper, we present the results of an exhaustive investigation of the
103: percolation of chemical clusters for a variety of values of $\tau$. We have
104: encountered that chemical clusters percolate at progressively higher densities
105: as $\tau$ grows from zero. Beyond a moderate value of $\tau$, the percolation
106: line coincides with the low-density branch of the liquid--solid coexistence
107: curve. This represents, to our knowledge, the first application of continuum
108: percolation where the percolation line coincides to such degree with a
109: coexistence line of a simple fluid. We also show that the percolation line for
110: physical clusters, although it moves towards higher densities as $\tau$ grows,
111: reaches a maximum density within the fluid phase beyond which the system
112: percolates irrespective of how large is $\tau$. From these findings, we
113: conclude that stable chemical clusters cannot exist in a simple fluid although
114: physical cluster do exist at relatively high densities. We also point out that
115: transient chemical clusters (by contrast with stable clusters) may help to
116: explain differences in experiments carried out on the same system when probing
117: different timescales.
118:
119: \section{Stillinger's clusters}
120:
121: In a many-body system, two particles at time $t$ are considered to form a
122: \textit{bonded pair} (in the Stillinger's sense) of proximity $d$, if they are
123: within a distance $d$ from each other at time $t$. Then, a set of particles at
124: time $t$ is called a \textit{Stillinger's cluster} of proximity $d$, if every
125: particle in the set is connected through a path of pair-bonded particles and
126: no other particle in the system is bonded to any of them. This cluster
127: definition is a basic standpoint for the following more realistic definitions.
128:
129: \section{Chemical clusters}
130:
131: Two particles at time $t$ are considered to form a \textit{bonded pair} of
132: lifespan $\tau$ and proximity $d$, if they were within a distance $d$ from
133: each other over the entire time interval $[t-\tau,t]$. Then, a set of
134: particles at time $t$ is called a \textit{chemical cluster} of lifespan $\tau$
135: and proximity $d$, if every particle in the set is connected through a path of
136: pair-bonded particles and no other particle in the system is bonded to any of
137: them. A chemical cluster with lifespan $\tau\gg\tau_{0}$ is called a
138: \textit{stable} chemical cluster. Here, $\tau_{0}$ means any time long enough
139: for the cluster to be detected by an experimental setup. A chemical cluster of
140: lifespan $\tau=0$ and proximity $d$ corresponds to a Stillinger's cluster of
141: proximity $d$.
142:
143: We would not normally expect to find stable chemical clusters in a simple
144: fluid since particles do not tend to form stable particle--particle bonds.
145: However, if we use an experimental technique able to resolve short times, we
146: might be able to detect the presence of ephemeral (small $\tau$) chemical
147: clusters. Moreover, if interactions are significantly stronger and dynamics
148: significantly slower, as is the case in colloidal systems, most experiments
149: are able to detect the existence of chemical clusters (see for example the
150: relation between percolation of chemical clusters and the sol-gel transition
151: \cite{Pugnaloni3}). Of course, chemical clusters are easily detected in true
152: chemical bonding (\textit{e.g.} polymerisation) and in strongly interacting
153: molecules (\textit{e.g.} hydrogen bonding \cite{Sciortino1}).
154:
155: \section{Physical clusters}
156:
157: Suppose that we identify a given set of particles at time $(t-\tau)$ as a
158: Stillinger's cluster of proximity $d$. We then focus on this set of particles
159: alone. As the system evolves in time, the particles of this Stillinger's
160: cluster approach other particles in the system, which we will disregard in our
161: analysis. Eventually, the original Stillinger's cluster fragments into two or
162: more disconnected subclusters --- in the Stillinger's sense. Each of these
163: subclusters can be now considered individually, disregarding the remaining
164: particles in the system --- even those that were part of the parent cluster
165: and now belong to a different subcluster. Again, during the evolution, each
166: subcluster may fragment. If we continue this analysis systematically until
167: time $t$ is reached (\textit{i.e.} over a period of time $\tau$), each final
168: subcluster, which is a product of numerous fragmentation events, is called a
169: \textit{physical cluster} of lifespan $\tau$ and proximity $d$. A physical
170: cluster with lifespan $\tau\gg\tau_{0}$ is called a \textit{stable} physical
171: cluster. Again, $\tau_{0}$ represents a time long enough for the cluster to be
172: detected by an experimental setup. A physical cluster of lifespan $\tau=0$ and
173: proximity $d $ corresponds to a Stillinger's cluster of proximity $d$.
174:
175: We expect to find stable physical clusters in a simple fluid. In fact, any
176: sample of liquid should correspond to a single physical cluster.
177:
178: \section{Molecular dynamics}
179:
180: The molecular dynamics (MD) simulation technique used to identify both
181: chemical and physical clusters was introduced and described in ref.
182: \cite{Pugnaloni1}. This technique is a slight modification of a standard $NVT$
183: MD simulation \cite{Allen1} where bonded pairs (for chemical cluster) and
184: entire Stillinger's clusters (for physical clusters) are tracked down along a
185: time interval of length $\tau$. We report results for a Lennard--Jones system,
186: \textit{i.e.} particles that interact through%
187: \begin{equation}
188: v(r)=4\varepsilon\left[ \left( \frac{\sigma}{r}\right) ^{12}-\left(
189: \frac{\sigma}{r}\right) ^{6}\right] .\label{e1}%
190: \end{equation}
191: Quantities are then expressed in Lennard--Jones reduced units in the rest of
192: the paper: $\sigma$ for length, $\sigma\sqrt{m/\varepsilon}$ for time, and
193: $\varepsilon$ for energy. Quantities in reduced units are indicated with an asterisk
194:
195: The results that we present in the following section are based on a leapfrog
196: MD simulation of $500$ particles in a cubic box with periodic boundary
197: conditions. A cut off distance equal to $2.7\sigma$ was used in the
198: Lennard--Jones pair potential. We have set the proximity value to $d^{\ast
199: }=1.5$. This value was chosen because it corresponds roughly to the maximum
200: distance at which two Lennard--Jones particles attract most strongly. Besides,
201: at $d^{\ast}=1.5$ lies the boundary between the first and the second neighbour
202: shell in the supercooled liquid state \cite{Keyes1}. This can be considered
203: the maximum distance for which a particle should be considered directly
204: connected from a geometrical point of view.
205:
206: In most cases, quantities are averaged over $10^{3}$ configurations generated
207: after stabilization. A few simulations performed with 4000 particles yielded
208: results indistinguishable from those obtained with the 500-particle-system. A
209: system is said to be in a percolated state if a cluster that spans the
210: periodic replicas is present 50 percent of the time \cite{Seaton1}. Then, a
211: percolation transition curve (the percolation loci), which separates the
212: percolated and the non-percolated states of the system, can be drawn above the
213: gas--liquid coexistence curve in the $T-\rho$ phase diagram.
214:
215: \begin{figure}[ptb]
216: \includegraphics{fig1.EPS}\caption{Temperature--density
217: phase diagram for the Lennard--Jones fluid and percolation loci for chemical
218: clusters. Solid squares (with trend line) correspond to the coexistence curve
219: for gas--liquid and liquid--solid transition. Open symbols (with dotted trend
220: line) correspond to the percolation loci for chemical clusters with proximity
221: $d^{\ast}=1.5$ and lifespan $\tau^{\ast}=$ 0 (circles), 0.5 (pentagons), 1.0
222: (squares), 5.0 (diamonds), 8.0 (down triangles), and 10.0 (up triangles).}%
223: \label{figure1}%
224: \end{figure}
225:
226: \section{Results}
227:
228: In fig. \ref{figure1} we show the phase diagram of the Lennard--Jones fluid in
229: the $T-\rho$ plane. The gas--liquid coexistence curve corresponds to the Gibbs
230: ensemble Monte Carlo (MC) simulation results from Panagiotopoulos
231: \cite{Panagio1} and the liquid--solid coexistence line corresponds to the MC
232: calculations of Hansen and Verlet \cite{Hansen1}. The percolation line for
233: chemical clusters of proximity $d^{\ast}=1.5$ and lifespan $\tau^{\ast
234: }=0,0.5,1,5,8$ and $10$ are also shown in fig. \ref{figure1}. The special case
235: with $\tau^{\ast}=0$ corresponds to \textit{instantaneous} Stillinger's
236: clusters. As we can see, the percolation line shifts towards high values of
237: $\rho$ as $\tau$ is increased. This is to be expected in a simple liquid
238: because, for a given temperature, the density needs to be increased in order
239: to promote caging and so extend the lifespan of particle--particle bonds.
240:
241: \begin{figure}[ptb]
242: \includegraphics{fig2.EPS}\caption{Temperature-density
243: phase diagram for the Lennard--Jones fluid and percolation loci for physical
244: clusters. Solid squares (with trend line) correspond to the coexistence curve
245: for gas--liquid and liquid--solid transition. Open symbols (with dotted trend
246: line) correspond to the percolation loci for physical clusters with proximity
247: $d^{\ast}=1.5$ and lifespan $\tau^{\ast}=$ 0 (circles), 1 (pentagons), 5
248: (squares), 10 (diamonds), 100 (down triangles), 1000 (up triangles). The solid
249: vertical line corresponds to the critical density $\rho_{f}$\ (see text).}%
250: \label{figure2}%
251: \end{figure}
252:
253: Interestingly, for $\tau^{\ast}\gtrsim8.0$ we find that the percolation line
254: coincides with the liquid branch of the liquid--solid transition curve.
255: Rigorously speaking, if $\tau^{\ast}\gtrsim8$, any state of the system to the
256: left of the liquid--solid coexistence line is non-percolated whereas any state
257: in the solid side of the phase diagram is percolated. The reason for this
258: resides in the fact that simple liquids have to make a transition to a
259: crystalline state in order to generate long-lasting particle--particle bonds.
260: In a crystalline state the neighbours of each particle remain the same for
261: extremely long periods of time. Of course, a similar effect can be achieved in
262: a glassy state, but we are only concerned with equilibrium states here. From
263: this result we can say that stable ($\tau\gg\tau_{0}$) chemical cluster do not
264: exist in the fluid phase of a simple liquid. It is worth mentioning that using
265: different values of $d$ (at least in the range $1.25<d^{\ast}<$ $2.0$) shifts
266: all the percolation curves but they still converge to the liquid--solid
267: transition line for large $\tau$. Larger values of $d$ requires larger values
268: of $\tau$ to achieve convergence.
269:
270: Our findings on the properties of the chemical cluster percolation line for
271: large $\tau$ suggests that we could use this analysis to predict the
272: liquid--solid transition line using a relatively simple simulation technique
273: and the appropriate cluster identification criterion. Previously, much effort
274: have been made to find a match between percolation and the gas--liquid
275: transition, especially in the two-state magnetic lattice model of a fluid
276: \cite{Stauffer2}. This studies have been unsuccessful so far, which can be
277: appreciated in the type of device (ghost spin) needed to prevent
278: inconsistencies in the theory \cite{Stauffer2}. As we can see, chemical
279: clusters are a more realistic definition of clusters and shows that the
280: percolation line should describe the liquid--solid transition rather than the
281: gas--liquid transition. Whether there exists a suitable definition of cluster
282: that yields a match between percolation and gas--liquid transition is still unresolved.
283:
284: For the sake of completeness, the percolation behaviour of physical clusters
285: is shown in fig. \ref{figure2}. In contrast with chemical clusters, physical
286: clusters percolate at moderate densities (far from the liquid--solid
287: transition) even at very large values of $\tau$ ($\tau^{\ast}\simeq10^{3} $).
288: In fact, the percolation curve seems to converge to a limiting line beyond
289: which the system is always percolated no matter how large is $\tau$. This
290: suggest that there exist a critical density $\rho_{f}$ ($\rho_{f}^{\ast}%
291: \simeq0.573$) in the supercritical phase above which physical clusters
292: percolates irrespective of the temperature and the required lifespan
293: \cite{Pugnaloni5}. This line is sketched in fig. \ref{figure2}, and it can be
294: thought of as a dividing line between a gas-like phase (to the left of
295: $\rho_{f}$) and a liquid-like phase (to the right of $\rho_{f}$). Within the
296: gas-like phase, physical clusters percolate only at finite lifespan. In the
297: liquid-like phase physical clusters always percolate the system. Within the
298: liquid-like phase ---except in a small neighbourhood around $\rho_{f}$---
299: nearly all particles in the system remain interconnected in a single large
300: physical cluster, \textit{i.e.}, no fragmentation takes places within any
301: finite time interval.
302:
303: The main objection we have to this separation into liquid-like and gas-like
304: regions of the supercritical phase is that states that we would normally
305: assign to a liquid reside in the gas-like side of the phase diagram. The
306: density $\rho_{f}$ is certainly higher than the thermodynamic critical density
307: $\rho_{c}$. All the states comprised between the gas--liquid coexistence
308: curve, the vertical line that passes through $\rho_{f}$ and the horizontal
309: line that passes through $T_{c}$ (the thermodynamic critical temperature) are
310: in a liquid state. According to our physical cluster properties, however, they
311: should be in a gas-like state. Interestingly, a different separation of the
312: supercritical phase proposed by Fisher and Widom \cite{Fisher1} does also meet
313: the gas--liquid coexistence line at a density above $\rho_{c}$ \cite{Vega1}.
314: However, the Fisher--Widom line does not coincide with the dividing line we
315: present here.
316:
317: \section{Conclusions}
318:
319: In this work, we have shown that chemical clusters and physical clusters
320: defined in ref. \cite{Pugnaloni1} have very different behaviour at large
321: lifespan. In particular, chemical clusters present a percolation curve that
322: coincides with the low-density branch of the liquid--solid transition curve
323: for any $\tau$ beyond a moderate value. Physical clusters, in contrast,
324: present a percolation curve that converges to a vertical line in the $T-\rho$
325: plane located at moderate densities when $\tau$ tends to infinity.
326:
327: The percolation properties of chemical clusters suggest that we can predict
328: the liquid--solid transition in a simple fluid by calculating the percolation
329: line at moderate lifespan values. An integral equation theory for chemical
330: clusters has been developed previously \cite{Pugnaloni1,Pugnaloni2}. Moreover,
331: a numerical technique has been used recently \cite{Zarragoicoechea1}\ to solve
332: the equations, but only at small values of $\tau$. This technique may allow us
333: to predict the position of the liquid--solid coexistence line without the need
334: of free energy calculations by setting $\tau$ to larger values ($\tau^{\ast
335: }\gtrsim8.0$).
336:
337: It is important to notice that is not possible to find stable (long lifespan)
338: chemical clusters in the fluid phase of a simple fluid. As soon as we exceed a
339: moderate value of $\tau$ we require the system to crystallise in order to find
340: chemical clusters of lifespan $\tau$. This finding seems rather intuitive;
341: however, until now, percolation models where not consistent with this picture.
342:
343: For small values of $\tau$, we find a family of percolation curves for
344: chemical clusters ranging from the classical instantaneous (Stillinger's)
345: cluster percolation up to the liquid--solid transition. Each of these
346: percolations curves may be observable as viscoelastic sol--gel-like
347: transitions measured at different shear frequencies \cite{Pugnaloni3}. Indeed,
348: working at very high shear frequencies will result in a sol--gel transition
349: line at relatively low densities. High frequencies probe short timescales and
350: any ephemeral (small $\tau$) stress-bearing percolating cluster will develop a
351: dominating elastic response in the system. Conversely, working at very low
352: frequencies will result in a sol--gel transition line at relatively high
353: densities. Low frequencies probe long timescales and the particle--particle
354: bonds have to last long enough (large $\tau$) for any percolating cluster to
355: develop a dominating elastic response.
356:
357: \textbf{Acknowledgments.}\ Support of this work by UNLP, CONICET and ANPCYT of
358: Argentina is very much appreciated. M.G.V. is a fellow of CONICET. L.A.P. and
359: F.V. are members of CONICET.
360:
361: \begin{thebibliography}{99} %
362: \bibitem {Stauffer}Stauffer D. and Aharony A., \textit{Introduction to
363: percolation theory}, Taylor and Francis, London, 1991, 2nd Ed.
364:
365: \bibitem {Senger1}Senger B., Schaaf P., Corti D. S., Bowles R., Voegel J. C.
366: and Reiss H., J. Chem. Phys. \textbf{110} (1999) 6421.
367:
368: \bibitem {Starr1}Starr F. W., Nielsen J. K. and Stanley H. E., Phys. Rev.
369: Lett. \textbf{82} (1999) 2294.
370:
371: \bibitem {Chen1}Chen S. H., Rouch J., Sciortino F. and Tartaglia P., J. Phys.:
372: Condens. Matter \textbf{6 (}1994) 10855.
373:
374: \bibitem {Butler1}Butler B. D., Hanley H. J. M., Hansen D. and Evans D. J.,
375: Phys. Rev. Lett. \textbf{74} (1995) 4468.
376:
377: \bibitem {Pugnaloni0}Pugnaloni L. A., Barker G. C. and Mehta A., Adv. Complex
378: Systems \textbf{4 (}2001) 289.
379:
380: \bibitem {Pugnaloni0b}Pugnaloni L. A. and Barker G. C., Physica A \textbf{337
381: (}2004) 428.
382:
383: \bibitem {Hill1}T. L. Hill, J. Chem. Phys. \textbf{23 (}1955) 617.
384:
385: \bibitem {Mayer1}J. E. Mayer, J. Chem. Phys. \textbf{5} (1937) 67.
386:
387: \bibitem {Campi1}Campi X., Krivine H. and Sator N., Physica A \textbf{296
388: (}2001) 24.
389:
390: \bibitem {Pugnaloni4}Pugnaloni L. A., M\'{a}rquez I. F. and Vericat F.,
391: Physica A \textbf{321 (}2003) 398.
392:
393: \bibitem {Stillinger1}Stillinger F. H., J. Chem. Phys. \textbf{38} (1963) 1486.
394:
395: \bibitem {Heyes1}Heyes D. M. and Melrose J. R., Mol. Phys. \textbf{66} (1989) 1057.
396:
397: \bibitem {Bahadur1}Bahadur R. and McClurg R. B., J. Phys. Chem. B \textbf{105}
398: (2001) 11893.
399:
400: \bibitem {Pugnaloni1}Pugnaloni L. A. and Vericat F., J. Chem. Phys.
401: \textbf{116} (2002) 1097.
402:
403: \bibitem {Zarragoicoechea1}Zarragoicoechea G. J., Pugnaloni L. A., Lado F.,
404: Lomba E. and Vericat F., Phys. Rev. E \textbf{71 (}2005) 0312021.
405:
406: \bibitem {Pugnaloni3}Pugnaloni L. A., cond-mat/0406713.
407:
408: \bibitem {Sciortino1}Sciortino F., Poole P. H., Stanley H. E. and Halvlin S.,
409: Phys. Rev. Lett. \textbf{64} (1990) 1686.
410:
411: \bibitem {Allen1}Allen M. P. and Tildesley D. J., \textit{Computer Simulation
412: of Liquids}, Clarendon Press, Oxford, 1987.
413:
414: \bibitem {Keyes1}Keyes T., J. Chem. Phys. \textbf{110 (}1999) 1097.
415:
416: \bibitem {Seaton1}Seaton N.A. and Glandt E.D., J. Chem. Phys. \textbf{86
417: (}1987) 4668.
418:
419: \bibitem {Panagio1}Panagiotopoulos A. Z., Mol. Phys. \textbf{61 (}1987) 813.
420:
421: \bibitem {Hansen1}Hansen J. P. and Verlet L., Phys. Rev. \textbf{184 (}1969) 151.
422:
423: \bibitem {Stauffer2}See Ref. \cite{Stauffer}, Chap. 7.
424:
425: \bibitem {Pugnaloni5}Pugnaloni L. A., [unpublished].
426:
427: \bibitem {Fisher1}Fisher M. E. and Widom B., J. Chem. Phys. \textbf{50} (1969) 3756.
428:
429: \bibitem {Vega1}Vega C., Rull L. F. and Lago S., Phys. Rev. E \textbf{51}
430: (1995) 3146.
431:
432: \bibitem {Pugnaloni2}Pugnaloni L. A. and Vericat F., Phys. Rev. E \textbf{61
433: (}2000) R6067.
434: \end{thebibliography}
435:
436:
437: \end{document}