1: \documentclass[aps,pre,groupedaddress,showpacs,twocolumn]{revtex4}
2: \usepackage{graphicx}
3:
4: \newcommand{\fsolvas}{f_\mathrm{solv}^{\mathrm{AS}}}
5: \newcommand{\fsolvs}{f_\mathrm{solv}^{\mathrm{S}}}
6: \newcommand{\fsolv}{f_\mathrm{solv}}
7: \newcommand{\fb}{f_\mathrm{b}}
8: \newcommand{\fint}{f_\mathrm{int}}
9: \newcommand{\Tast}{T^\ast}
10: \newcommand{\Tw}{T_\mathrm{w}}
11: \newcommand{\Tc}{T_\mathrm{c}}
12: \newcommand{\kb}{k_\mathrm{B}}
13: \newcommand{\rme}{\mathrm{e}}
14: \newcommand{\abs}[1]{\left|#1\right|}
15: \newcommand{\yast}{y^\ast}
16:
17: \begin{document}
18: \title{Scaling of solvation force in 2D Ising strip}
19: \author{Piotr Nowakowski}
20: \email{pionow@fuw.edu.pl}
21: \author{Marek \surname{Napi\'orkowski}}
22: \affiliation{Faculty of Physics, University of Warsaw, ul. Ho\.za 69,
23: 00-681 Warsaw, Poland}
24: \date{\today}
25: \begin{abstract}
26: The solvation force for the 2D Ising strip is calculated via
27: exact diagonalization of the transfer matrix in two cases: the symmetric case corresponds to
28: identical surface fields, and the antisymmetric case to exactly opposite surface fields.
29: In the symmetric case the solvation force is
30: always negative (attractive) while in the antisymmetric case the solvation force is positive (repulsive)
31: at high temperatures and negative at low temperatures. It changes sign close to the
32: critical wetting temperature characterizing the semi--infinite system.
33: The properties of the solvation force are discussed and the
34: scaling function describing its dependence on temperature, surface field, and strip's width is proposed.
35: \end{abstract}
36: \pacs{05.50.+q, 68.08.Bc}
37: \maketitle
38: \section{Introduction}
39: Properties of the solvation force in various condensed matter systems have been the subject
40: of very intensive research during recent years
41: \cite{EvansMarconi, Christ1, Evans1, H1, EvansStecki, HSED, Evans2, KPH, SHD, Nature}.
42: Lattice models play special role among considered systems. Although many important results were
43: obtained in this field
44: by different methods of taking into account fluctuations which determine the properties of
45: the analyzed systems, the
46: approach based on the exact evaluation of the partition function via the transfer matrix method still
47: plays a distinguished role. Below we report our results on the properties of the solvation force for
48: 2D Ising strip. They are obtained via exact diagonalization of the transfer matrix which is then followed by
49: numerical solutions of equations for eigenvalues.
50:
51: We consider Ising model on a two--dimensional square lattice consisting of
52: $M$ rows and $N$ columns. There is no bulk magnetic field acting on the system
53: and there are two surface fields: $h_1$ acts on spins in the first row and $h_2$ acts on the $M$-th
54: row. The Hamiltonian of this model takes the standard form
55: \begin{eqnarray}
56: \nonumber\mathcal{H}\mathbf{(}\left\{s_i\right\}\mathbf{)} &=& -J
57: \sum_{n=1}^{N}\sum_{m=1}^{M-1}s_{n,m} s_{n,m+1}\\
58: \nonumber& &-J\sum_{n=1}^N\sum_{m=1}^M s_{n,m}s_{n+1,m} \\
59: \label{hamiltonian}& &-h_1\sum_{n=1}^N s_{n,1}-h_2\sum_{n=1}^N s_{n,M},
60: \end{eqnarray}
61: where $s_{n,m}=\pm1$ with $n=1,\ldots, N,\ m=1,\ldots,M$ denotes the spin located in the
62: $n$-th column and $m$-th row, and $J$ is the coupling constant. Periodic
63: boundary conditions in the horizontal direction are imposed:
64: $s_{N+1,m}\equiv s_{1,m}$.
65:
66: Our purpose is to determine the properties of the solvation force experienced by the system boundaries.
67: In the following we consider two cases: the symmetric case (S) corresponds to $h_1=h_2$, and
68: for the antisymmetric case (AS) one has $h_1=-h_2$.
69:
70: Although the strip of finite width experiences no phase transition, we shall often refer to two
71: critical temperatures: the bulk critical temperature
72: $ \kb \Tc = 2J/\ln\left(1+\sqrt{2}\right)$,
73: and the wetting temperature
74: $\Tw$ which characterizes the critical wetting transition in a semi--infinite system with surface field $h_{1}$.
75: The wetting temperature fulfills the equation
76: \begin{equation}
77: \cosh 2K_1=\cosh 2K-\sinh 2K \rme^{-2K},
78: \end{equation}
79: where $K=J/\left(\kb\Tw\right)$, $K_1=h_1/\left(\kb\Tw\right)$, and
80: $\kb$ is the Boltzmann constant.
81: For a small surface field $h_1$ this equation leads to
82: \begin{equation}
83: \label{wettemp}\frac{\Tc-\Tw}{\Tc}=\frac{1}{4}\left(1+\sqrt{2}\right)\,\ln\left(1+\sqrt{2} \right)
84: \left(\frac{h_1}{J}\right)^2, \quad \frac{h_1}{J} \to 0.
85: \end{equation}
86: For $h_1\geq J$ one has $\Tw=0$ and there is no wetting transition.
87:
88: \section{Solvation force}
89: The dimensionless free energy per one column is defined as
90: \begin{equation}
91: f\left(T,h_1,M\right)=\lim_{N\to\infty} \frac{1}{N} \,
92: \frac{\mathcal{F}}{\kb T} = -\lim_{N\to\infty}\frac{\ln \mathcal{Z}}{N},
93: \end{equation}
94: where $\mathcal{Z}$ is the canonical partition function for the
95: Hamiltonian (\ref{hamiltonian}) which we evaluate using the exact transfer
96: matrix method \cite{Kaufman,Abraham}. The free energy may be separated into three types of
97: contributions
98: \begin{equation}
99: f\left(T,h_1,M\right)=M
100: \fb\left(T\right)+f_\mathrm{s}\left(T,h_1\right)+\fint\left(T,h_1,M\right),
101: \end{equation}
102: where $\fb$ is the bulk free energy per one spin \cite{Onsager} and
103: $f_\mathrm{s}$ is the surface free
104: energy; both $\fb$ and $f_\mathrm{s}$ are $M$--independent. The remaining term $\fint$
105: describes the interaction between the system boundaries. It tends to 0 as $M$ goes to infinity
106: and from this term the solvation force originates.
107:
108: The solvation force is, in general, defined as the minus derivative of $\fint$
109: wrt $M$. In our case, because $M$ takes only integer values, we use the definition
110: \begin{equation}
111: \fsolv\left(T,h_1,M\right)\!=\!-\left[\fint\!\left(T,h_1,M+1\right)-
112: \fint\!\left(T,h_1,M\right)\right],
113: \end{equation}
114: which leads to the following expression
115: \begin{equation}
116: \fsolv\left(T,h_1,M\right)=f\left(T,h_1,M\right)-f\left(T,h_1,M+1\right)-
117: \fb.
118: \end{equation}
119:
120: The solvation force for the 2D strip has been already analyzed in the $\Tw=0$ case
121: \cite{EvansStecki} corresponding to $h_{1}=J$. Our analysis covers the whole spectrum
122: $\Tw \ge 0$, i.e. $h_1 \le J$. To calculate the solvation force we used the
123: methods described in \cite{Kaufman, Abraham, Stecki, Maciolek}. The complete
124: analysis (including also the inhomogeneous boundary fields) will be published elsewhere
125: \cite{NN}; here we discuss only the main results.
126:
127: First we briefly discuss the symmetric case (S). The corresponding solvation force $\fsolvs$
128: is plotted in Fig.\ref{fig1} for two cases:
129: $\Tw > 0$ (solid curve) and $\Tw=0$ (broken curve). The difference
130: between these two functions is only of quantitative nature ---
131: decreasing the surface field $h_{1}$, i.e. increasing the wetting temperature $\Tw$
132: results in decreasing the absolute value of the solvation force. The minimum of
133: the solvation force $\fsolvs$ is located at $T_\mathrm{min}^\mathrm{S} > \Tc$.
134:
135: \begin{figure}[t]
136: \includegraphics{fig1.eps}
137: \caption{\label{fig1} Comparison of the solvation force $\fsolvs$ for $\Tw= 0.8 \Tc$
138: ($h_1 \approx
139: 0.60 J$) (solid curve) and $\Tw=0$ ($h_1=J$) (broken curve) for symmetric case $h_{1}=h_{2}$.
140: Both curves correspond to the value $M=50$.}
141: \end{figure}
142:
143: For opposite surface fields (AS) the solvation force $\fsolvas$ in the $\Tw>0$ case, i.e., $h_1<J$,
144: differs substantially from the $\Tw=0$, i.e., $h_1=J$ case. Fig.\ref{fig2} presents a
145: typical plot of $\fsolvas$ as a function of $T$. For
146: low temperatures this force is negative (attractive) and has a minimum at $T_\mathrm{min}^\mathrm{AS}<\Tw$. The solvation force is negative at wetting temperature and
147: has a zero at $\Tast>\Tw.$ Above $\Tast$ the solvation force is positive
148: (repulsive) and has a maximum at $T_\mathrm{max}^\mathrm{AS}<\Tc$ (for $M$ large enough). This remains
149: in contrast with the $\Tw=0$ case in which the solvation force is positive for all
150: temperatures.
151:
152: Exactly at the bulk critical temperature $\Tc$ the dependence of the
153: solvation force on $M$ has been found using conformal invariance \cite{Nightingale,Cardy}
154: \begin{eqnarray}
155: \fsolvs \left(\Tc,h_1,M\right)=&-\frac{\pi}{48M^2}+\mathcal{O}\left(1/M^3\right),\\
156: \label{casimir}\fsolvas \left(\Tc,h_1,M\right)=&\frac{23\pi}{48
157: M^2}+\mathcal{O}\left(1/M^3\right).
158: \end{eqnarray}
159: This result, in particular the universal values of the amplitudes, has been
160: reproduced by our analysis.
161:
162: In the rest of this paper we exclusively discuss the antisymmetric case (AS).
163: \begin{figure}[t]
164: \includegraphics{fig2.eps}
165: \caption{\label{fig2}The solvation force $\fsolvas$ as a function of temperature for
166: $\Tw=0.8 \Tc$ (i.e. $h_1\approx 0.60 J$) and $M=50$ in the
167: antisymmetric case $h_1=-h_2$. Note that the difference between the zero
168: of the solvation force $\Tast$ and the wetting temperature $\Tw$ is not visible in this
169: scale.}
170: \end{figure}
171:
172: First we concentrate on the temperature $\Tast$ at which the solvation
173: force becomes zero. We have found that for fixed $h_1> 0$ and for
174: $M\to\infty$ this temperature approaches the wetting temperature $\Tw$ exponentially
175: \begin{equation}
176: \label{tast}\frac{\Tast-\Tw}{\Tc}=A\left(h_1\right) \rme^{-B\left(h_1\right)M},\quad M\to\infty,
177: \end{equation}
178: where $A\left(h_1\right)$ and $B\left(h_1\right)$ are positive functions of the surface field $h_1$
179: \footnote{In principle, these functions can be found by fitting Eq.(\ref{tast}) to our
180: numerical results. However, for $\Tw>0.5 \Tc$ and for $M\geq 10$ the
181: difference between $\Tast$ and $\Tw$ is of order of numerical errors
182: and we were unable to precisely determine the functions $A\left(h_1\right)$ and $B\left(h_1\right)$.}.
183: We have found that for $h_1\to 0$, i.e., $\Tw \to \Tc$ one has
184: \begin{equation}
185: \lim_{h_1\to 0} A\left( h_1\right)=0.
186: \end{equation}
187:
188: We note that this result is different from the corresponding result obtained within the
189: mean field theory where $\Tast$ is exponentially shifted
190: \textit{below} $\Tw$. It is also different form the corresponding result obtained for the
191: restricted
192: solid--on--solid (RSOS) model, where $\Tast$ is \textit{equal}
193: to $\Tw$ \cite{ParryEvans}.
194:
195: \section{Scaling function}
196:
197: In this section we discuss the scaling function that describes the
198: behavior of the solvation force $\fsolvas(T,h_1,M)$ for large $M$ and subcritical
199: temperatures. The relevant scaling function $\mathcal{X}(x)$ has already been
200: proposed to describe $\fsolvas$ in the $\Tw=0$ case \cite{EvansStecki}
201: \begin{equation}
202: \label{fscaleX}\fsolvas\left(T,h_1=J,M\right)=\frac{1}{M^2}
203: \mathcal{X}\left(\frac{M}{\xi^-\left(T\right)}\right).
204: \end{equation}
205: The correlation length for 2D Ising model close to $\Tc$ is $\xi^-\left(T\right)=\xi_0^- t^{-1}$, where $t=\left(\Tc-T\right)/\Tc>0$ and $\xi^-_0=\left[4
206: \ln\left(1+\sqrt{2}\right)\right]^{-1}$. The scaling function $\mathcal{X}(x)$ can
207: be obtained numerically from the transfer matrix spectrum and some of its properties can
208: be proved analytically (see Eq.(\ref{casimir}) and \cite{Danchev}), namely
209: \begin{eqnarray}
210: \mathcal{X}\left(0\right)&=&23\pi/48,\\
211: \mathcal{X}\left(x\right)&=&2\pi^2/x, \quad \text{for }x\to\infty.
212: \end{eqnarray}
213: Here we would like to extend this result to $h_1 < J$, i.e. $\Tw>0$ case.
214:
215: For 2D Ising model the gap exponent $\Delta_1=1/2$ and for $T<\Tc$ the following
216: scaling behavior
217: \begin{equation}
218: \label{fscale}\fsolvas\left(T,h_1,M\right)=\frac{1}{M^2} \mathcal{Y}\left(x,y\right),
219: \end{equation}
220: where
221: \begin{equation}
222: \label{A0}x=\frac{M}{\xi^-_0 t^{-1}},\quad y=\frac{A_0}{\kb\Tc} \frac{h_1}{t^{1/2}}
223: \end{equation}
224: comes into play in the limit $M\to \infty$ with $x$ and $y$ fixed. This implies additionally the
225: $t \to 0$, $h_1 \to 0$ and $\Tw \to \Tc$ limits.
226: The coefficient $A_0=\left[\left(1+\sqrt{2}\right)/\ln\left(1+\sqrt{2}\right)\right]^{1/2}$ in Eq.(\ref{A0})
227: has been introduced such that the value $y=1$ corresponds to $T=\Tw$ and then
228: Eq.(\ref{wettemp}) is satisfied. For $y < 1$ Eq.(\ref{fscale}) gives the solvation force $\fsolvas$
229: below the wetting temperature and for $y>1$ above $\Tw$.
230:
231: \begin{figure}[t]
232: \includegraphics{fig3.eps}
233: \caption{\label{fig3}Convergence of the scaling function. The function
234: $\mathcal{Y}_M$ is evaluated for $x=1$ and for several values of $y$ and $M$ presented on the plot.}
235: \end{figure}
236: \begin{figure}[t]
237: \includegraphics{fig4.eps}
238: \caption{\label{fig4}Plots of the scaling function $\mathcal{Y}$. All plots were obtained from
239: $\mathcal{Y}_M$ for $M=200$.}
240: \end{figure}
241: \begin{figure}[t]
242: \includegraphics{fig5.eps}
243: \caption{\label{fig5}The scaling function $\mathcal{Y}$ plotted for $y=0.5, 1.5,
244: 2.5, 3.5$ and the scaling function $\mathcal{X}$. All points were calculated
245: for $M=200$.}
246: \end{figure}
247:
248: Eq.(\ref{fscale}) may be rewritten in the form leading to the scaling
249: function
250: \begin{equation}
251: \label{limit}\mathcal{Y}\left(x,y\right)=\lim_{M\to\infty}
252: \mathcal{Y}_M\left(x,y\right),
253: \end{equation}
254: where
255: \begin{eqnarray}
256: \lefteqn{\mathcal{Y}_M\left(x,y\right) = }\\
257: \nonumber & M^2 \fsolv\left(\Tc\left(1-\frac{x
258: \xi_0^-}{M}\right), \kb \Tc y \left(\frac{M}{x \xi_0^-
259: A_0^2}\right)^{1/2} , M\right).
260: \end{eqnarray}
261:
262: Fig.\ref{fig3} presents plots of $\mathcal{Y}_M$ for fixed value of $x=1$ and selected values of
263: $y \in [0, 1.2]$ to exhibit the convergence of the series $\mathcal{Y}_M$. Typically,
264: for $M$ large enough one has
265: \begin{eqnarray}
266: \label{yyy}
267: \mathcal{Y}_M\left(
268: x,y\right)=\mathcal{Y}\left(x,y\right)+\frac{C\left(x,y\right)}{M}+O\left(M^{-2}\right).
269: \end{eqnarray}
270: To estimate the values of the function $\mathcal{Y}$ we evaluated $\mathcal{Y}_M$ for
271: $M=200$. We note that a different way of obtaining the function $\mathcal{Y}$, based on the least
272: squares method which allows to calculate the functions $\mathcal{Y}$ and $C$ in Eq.(\ref{yyy}),
273: leads to similar results (the differences are not visible on the scale
274: of our figures). The plots of function $\mathcal{Y}$ are shown on Fig.\ref{fig4}.
275:
276: Next we investigate the relation between the scaling functions
277: $\mathcal{Y}\left(x,y\right)$ and $\mathcal{X}\left(x\right)$. The function
278: $\mathcal{X}\left(x\right)$ is calculated in the
279: limit $M\to\infty$, $t\to 0$ with $h_1$ and $M t$ fixed. By applying
280: this limit to Eqs (\ref{fscale}) and (\ref{fscaleX}) one~gets
281: \begin{eqnarray}
282: \label{y1y1}
283: \mathcal{X}\left(x\right)=\lim_{y\to\infty}\mathcal{Y}\left(x,y\right).
284: \end{eqnarray}
285: The functions $\mathcal{X}\left(x\right)$ and $\mathcal{Y}\left(x,y\right)$ plotted
286: for selected values of $y$ are presented on Fig.\ref{fig5}.
287: Additionally we have found that $\mathcal{X}\left(x\right)-\mathcal{Y}\left(x,y\right)\propto y^{-2}$
288: for large $y$.
289:
290: The scaling function $\mathcal{Y}\left(x,y\right)$ changes its sign, see Figs \ref{fig4}, \ref{fig5}.
291: The zeros of the scaling function are denoted by $\yast\left(x\right)$, i.e.,
292: $\mathcal{Y}\left(x,\yast(x)\right)=0$.
293: We have found that for large $x$ the function
294: $\yast\left(x\right)$ approaches $1$ exponentially form above. This
295: allows us to show that in the scaling limit $M\to\infty$, $h_1\to 0$, and
296: $M h_1^2$ fixed one has
297: \begin{eqnarray}
298: \nonumber\frac{\Tast-\Tw}{\Tc}&=&\frac{\Tc-\Tw}{\Tc}-\frac{1}{M}
299: f\left(\frac{M h_1^2}{\kb^2
300: \Tc^2}\right)+\mathcal{O}\left(1/M^2\right),\\
301: \label{func_f}&=& \frac{1}{M} g\left(\frac{M h_1^2}{\kb^2
302: \Tc^2}\right)+\mathcal{O}\left(1/M^2\right),
303: \end{eqnarray}
304: where $f\left(\zeta\right)$ and $g\left(\zeta\right)$ are positive functions which can be determined via an implicit formula
305: \begin{eqnarray}
306: A_0 \zeta^{1/2}&=&\left[f\left(\zeta\right)\right]^{1/2} \yast\left(
307: f\left(\zeta\right)/\xi_0^- \right),\\
308: g\left(\zeta\right) &=& A_0^2-f\left(\zeta\right).
309: \end{eqnarray}
310: We note that Eq.(\ref{func_f}) is different form Eq.(\ref{tast}) because different limiting
311: procedures were applied in these two cases.
312: \section{Conclusions}
313:
314: In the symmetric case of identical surface fields the solvation force is negative (attractive) and
315: has minimum at supercritical temperature. The solvation force
316: calculated for system parameters such that $\Tw>0$ differs only quantitatively from the one in case
317: $\Tw=0$, see Fig.\ref{fig1}.
318:
319: In the antisymmetric case of opposite surface fields the solvation
320: force is positive at high temperatures and negative at low temperatures;
321: it changes its sign at temperature $T^\ast > \Tw$, see Fig.\ref{fig2}. In the case $h_1$
322: fixed and $M \to \infty$ the difference $\Tast-\Tw$ approaches $0$
323: exponentially quickly in $M$.
324:
325: The scaling function $\mathcal{Y}(x,y)$ was proposed to describe the behavior of the
326: solvation force for $T<\Tc$ in the limit $h_1\to 0$ and $M\to
327: \infty$, see Figs \ref{fig4}, \ref{fig5}. We checked that in the limit of high
328: surface field ($y\to\infty$ in Eq.(\ref{fscale})) this scaling function
329: approached the scaling function describing the scaling behavior of the for solvation force
330: in the $\Tw=0$ case. In addition, the zeros of the scaling function were investigated to find
331: the formula for $\Tast-\Tw$ in the $h_1\to 0$ limit.
332:
333: We thank A. Macio{\l}ek, D. Danchev, and S. Dietrich for many helpful discussions. This
334: work has been financed from the funds provided by Polish Ministry of Science and Higher Education
335: for scientific research for the years 2006--2008 under the research project N202 076 31/0108.
336:
337:
338: \begin{thebibliography}{99}
339: \bibitem{EvansMarconi} R. Evans and U. Marini Bettolo Marconi, J. Chem. Phys. \textbf{86}, 7138 (1987).
340: \bibitem{Christ1} H. K. Christenson and C. E. Blom, J. Chem. Phys. \textbf{86}, 419 (1987).
341: \bibitem{Evans1} R. Evans, J. Phys.: Condens. Matter \textbf{2}, 8989 (1990).
342: \bibitem{H1} R. Evans, J. R. Henderson, D. C. Hoyle, A. O. Parry, and Z. A. Sabeur, Mol. Phys.
343: \textbf{80}, 755 (1993).
344: \bibitem{EvansStecki} R. Evans and J. Stecki, Phys. Rev. B \textbf{49}, 8842 (1994).
345: \bibitem{HSED} A. Hanke, F. Schlesener, E. Eisenriegler, and S. Dietrich,
346: Phys. Rev. Lett. \textbf{81}, 1885 (1998).
347: \bibitem{Evans2} A. Drzewi\'nski, A. Macio{\l}ek, and R. Evans, Phys. Rev. Lett. \textbf{85}, 3079 (2000).
348: \bibitem{KPH} S. Kondrat, A. Poniewierski, and L Harnau, Eur. Phys. J. E \textbf{10}, 163 (2003).
349: \bibitem{SHD} F. Schlesener, A. Hanke, and S. Dietrich, J. Stat. Phys. \textbf{110}, 981 (2003).
350: \bibitem{Nature} C. Hertlein, L. Helden, A. Gambassi, S. Dietrich, and C. Bechinger, Nature \textbf{451}, 172 (2008).
351: \bibitem{Kaufman} B. Kaufman, Phys. Rev. \textbf{76}, 1232 (1949).
352: \bibitem{Abraham} D. B. Abraham and A. Martin-L\"of, Commun. Math. Phys. \textbf{32}, 245 (1973).
353: \bibitem{Onsager} L. Onsager, Phys. Rev. \textbf{62}, 559 (1942).
354: \bibitem{Stecki} J. Stecki, A. Macio{\l}ek, and K. Olaussen, Phys. Rev. B \textbf{49}, 1092 (1994).
355: \bibitem{Maciolek} A. Macio\l{}ek and J. Stecki, Phys. Rev. B \textbf{54}, 1128 (1996).
356: \bibitem{NN} P. Nowakowski and M. Napi\'orkowski, in preparation.
357: \bibitem{Nightingale} M. P. Nightingale and J. O. Indekeu,
358: Phys. Rev. Lett. \textbf{54}, 1824 (1985); H. W. J. Bl\"ote, J. L. Cardy, and M. P. Nightingale, \textit{ibid.} \textbf{56}, 742 (1986).
359: \bibitem{Cardy} J. L. Cardy, Nucl. Phys. B \textbf{275}, 200 (1986).
360: \bibitem{ParryEvans} A. O. Parry and R. Evans, Phys. Rev. Lett. \textbf{64}, 439 (1990); A. O. Parry and R. Evans, Physica A \textbf{181}, 250 (1992).
361: \bibitem{Danchev} D. M. Danchev (private communication), see also J. G. Brankov, D. M. Danchev and
362: N. S. Tonchev, \textit{Theory of Critical Phenomena in Finite-Size Systems, Scaling and Quantum Effects} (World Scientific, Singapore, 2000).
363: \end{thebibliography}
364: \end{document}