1: %&latex209
2:
3:
4: \documentstyle[preprint,aps,epsfig]{revtex}
5: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
6: %TCIDATA{OutputFilter=Latex.dll}
7: %TCIDATA{LastRevised=Tue Mar 14 13:18:27 2000}
8: %TCIDATA{<META NAME="GraphicsSave" CONTENT="32">}
9: %TCIDATA{Language=American English}
10: %TCIDATA{CSTFile=revtex.cst}
11:
12:
13:
14: \begin{document}
15: \title{Evolution of the universality class in slightly diluted ($1>p>0.8$) Ising systems.}
16: \author{Manuel I. Marqu\'es and Julio A. Gonzalo}
17: \address{Departamento de F\'isica de Materiales, C-IV
18: Universidad Aut\'onoma de Madrid, 28049, Madrid, Spain
19: email: julio.gonzalo@uam.es}
20: \date{14-03-2000}
21: \maketitle
22:
23: \begin{abstract}
24: The crossover of a pure (undiluted) Ising system (spin per site probability $p=1$) to a diluted Ising
25: system (spin per site probability $p<0.8$) is studied by means of Monte Carlo calculations
26: with $p$ ranging between 1 and 0.8 at intervals of 0.025. The evolution of the self averaging is
27: analyzed by direct determination of the normalized square widths $R_{M}$ and $R_{\chi}$ as a function of $p$. We find a monotonous and smooth evolution
28: from the pure to the randomly diluted universality class. The $p$-dependent transition
29: is found to be indepent of size ($L$). This property is very convenient for extrapolation towards the
30: randomly diluted universality class avoiding complications resulting from finite size effects.
31: \end{abstract}
32:
33: Systems with quenched randomness have been studied intensively for several
34: decades \cite{Domb}. One of the first results was establishing the so called
35: Harris criterion \cite{Harris}, which predicts that a weak dilution does not
36: change the critical behavior's character near second order phase transitions
37: for systems of dimension $d$ with specific heat exponent lower than zero
38: (the so called P systems), $\alpha _{pure}<0\Longleftrightarrow \nu
39: _{pure}>2/d$ due to the hyperscaling relationship, in the undiluted case.
40: This criterion has been confirmed by several renormalization group (RG)
41: analyses \cite{Lubensky,LubenskyII,Grinstein}, and by scaling analysis \cite
42: {Aharony}. It was shown to hold also in strongly diluted systems by Chayes
43: et al. \cite{Chayes}. For $\alpha _{pure}>0$ (the so called R systems), for
44: example the Ising 3D case, the system fixed point flows from that of a pure
45: (undiluted) fixed point towards a new stable fixed point at which $\alpha
46: _{random}<0$ \cite{Lubensky,LubenskyII,Grinstein,Aharony,Chayes,Kinzel} for
47: diluted systems.
48:
49: Recently Ballesteros et al. have used the Monte Carlo approach to study
50: diluted Ising systems in two\cite{BallesterosI}, three \cite{BallesterosII}
51: and four dimensions \cite{BallesterosIII}. The existence of a new
52: universality class for the randomly diluted Ising system (RDIS) (different
53: from that of the pure Ising model, and $p$-independent being $p$ the spin
54: per site probability) is proved using an infinite volume extrapolation
55: technique\cite{BallesterosII} based upon the leading correction to scaling.
56: The critical exponents obtained this way agree with the experimental
57: critical exponents for a random disposition of vacancies in diluted magnetic
58: systems \cite{Folk}.
59:
60: The crossover from the pure Ising system $p=1$ to the randomly diluted
61: system may occur for very large values of the average density of occupied
62: sites ($1>p>0.8$), i.e. systems with a very small amount of vacancies. In
63: this region, the specific heat critical exponent must flow from a value
64: grater than zero, $\alpha _{pure}>0$ for $p=1$, to a value smaller than
65: zero, $\alpha _{random}<0$ for $p=0.8$. It means that in principle is
66: possible to expect the existence of a critical density $p_{c}$ at which its
67: value is equal to zero. The $p_{c}$ value has been found to be around $0.9$
68: \cite{BallesterosII}.
69:
70: In principle it is not clear whether this crossover should occur smoothly or
71: whether the crossover should take place sharply at a critical value $p_{c},$
72: separating the two distinct universality classes. This is a crucial question
73: to establish whether the slightly diluted systems should be considered as
74: pure (basically undiluted), as randomly diluted systems, or, on the
75: contrary, as intermediate states between both extreme classes. There is an
76: intrinsic difficulty in detecting the evolution of critical exponents from
77: pure to diluted random Ising systems due to the fact that they are very
78: similar (see Table I). Following Ballesteros et al. we find $\alpha
79: _{random}=-0.051,\beta _{random}=0.3546,\gamma _{random}=1.342$ in
80: comparison with the pure undiluted values: $\alpha _{pure}=0.11,\beta
81: _{pure}=0.3267,\gamma _{{pure}}=1.237$ \cite{Blote} (incidentally, this does
82: not happen if the disorder is long range correlated \cite
83: {Weinrib,BallesterosIV,Marques}). That is why it is useful to study some
84: other universal quantity which clearly indicates the difference between the
85: pure and the random universality classes.
86:
87: For a random hypercubic sample of linear dimension $L$ and number of sites $%
88: N=L^{d}$, any observable singular property $X$ presents different values for
89: different random realizations corresponding to the same average dilution.
90: This means that X behaves as a stochastic variable with average $[X]$,
91: variance $(\Delta X)^{2}$ and a normalized square width $R_{X}=(\Delta
92: X)^{2}/[X]^{2}$. This quantity allows us to determine properly the evolution
93: from the pure to the randomly diluted system by an investigation of its self
94: averaging behavior. A system is said to exhibit self averaging (SA) if $%
95: R_{X}\rightarrow 0$ as $L\rightarrow \infty $. If the system is away from
96: criticality, $L>>\xi $ (being $\xi $ the correlation length) the central
97: limit theorem indicates that strong SA must be expected. However, the
98: self-averaging behavior of a ferromagnet at criticality (where $\xi >>L)$ is
99: not so obvious. This point has been studied recently. Wiseman and Domany
100: (WD) have investigated the self-averaging of diluted ferromagnets at
101: criticality by means of finite-size scaling calculations\cite{Wiseman},
102: concluding weak SA for both the P and R cases. In contrast Aharony and
103: Harris (AH), using a renormalization group analysis in $d=4-\varepsilon $
104: dimensions, proved the expectation of a rigorous absence of self-averaging
105: in critically random ferromagnets \cite{AharonyII}. More recently, Monte
106: Carlo simulations where used to check this lack of self-averaging in
107: critically disordered magnetic systems \cite
108: {BallesterosII,WisemanII,AharonyIII}. The absence of self-averaging was
109: confirmed. The source of the discrepancy with previous scaling analysis by
110: WD was attributed to the particular size ($L$) dependence of the
111: distribution of pseudocritical temperatures used in their work.
112:
113: In the present work we study the evolution of the normalized square width
114: from $p=1$ to $p=0.8$ at small steps $\Delta p=0.025,$ in an effort to
115: characterize in detail the evolution of the normalized square width from $%
116: R_{M}=R\chi =0$ to the zone where lack of self-averaging $(R_{M}\neq 0$ and $%
117: R\chi \neq 0)$ appears. We will determine whether there is some sharp
118: critical value $p_{c}$ separating both universality classes, or there is a
119: smooth evolution. We have performed Monte Carlo calculations using the Wolff
120: single cluster algorithm \cite{Wolff} in diluted three dimensional Ising
121: systems at criticality for different values of the site occupation spin
122: probability $1>p>0.8.$ In order to obtain good enough statistics in our
123: determination of the normalized square width for magnetization and
124: susceptibility, we have used 500 samples for the sizes $L=20,40,60.$ The
125: magnetization and susceptibility of each sample was determined using 50.000
126: MCS leaving the previous 100.000 MCS for thermalization. The critical
127: temperature for each dilution was taken by interpolation between the data
128: reported by Heuer et al. \cite{Heuer} and Ballesteros et al. \cite
129: {BallesterosII}. We may note that there are no much data in the literature
130: about the critical temperature in this region of slightly diluted systems. A
131: ''nearly-linear'' extrapolation of the data for $T_{c}$ vs. $p$ from $p=1$
132: to $p=0.8$ seems clear from Fig.1. To check this we have calculated the
133: critical temperature for several values of $\ p$ by means of statistics on
134: the Binder Cumulant, and we have found that the data lie over the
135: interpolation functions previously considered.
136:
137: We can build histograms with the values obtained for susceptibility or
138: magnetization at criticality. In the case of very high $p$ values,
139: corresponding to very-low dilution, the width of these histograms is very
140: small, indicating the existence of proper self-averaging. However, for
141: somewhat lower values of $p$, the system starts its crossover to the
142: randomly diluted behavior and the width of the histograms begins to
143: increase, indicating that lack of self-averaging is taken place. Fig.2a and
144: b show the evolution of these histograms for the case of the susceptibility
145: and for $L=60$. Note that the width of the histograms increases monotonously
146: as $p$ decreases, indicating a smooth flow towards the random diluted
147: universality class. We will see this point more clearly studying the value
148: of the normalized square width.
149:
150: The results obtained for the normalized square width of the magnetization
151: and the susceptibility are presented in Fig3 and Fig4 respectively. Note how
152: in both cases we find a smooth evolution indicating that the crossover from
153: the pure fixed point to the randomly diluted fixed point takes place
154: smoothly and continuously and that there is no apparent critical value of $%
155: p_{c}$ acting as a boundary between the two regimens.
156:
157: The value of the normalized square width for a given $p$, can be strongly
158: affected by finite size effects. In order to obtain a value of $R_{M}$ or $%
159: R_{\chi }$ independent of $p$ for $p<0.8$ it is necessary to consider very
160: high values of $L$ \cite{AharonyIII} or to use the so called infinite volume
161: extrapolation \cite{BallesterosII}. However, for small dilution $(1>p>0.8)$
162: the values of the normalized square width are nearly unaffected by finite
163: size effects \cite{BallesterosII}, {\bf but} they are dependent of $p$. That
164: is the reason why in Fig3 and Fig4 all the data seem to collapse over the
165: same curve. This does not happen for $p<0.8$ where finite size effects
166: clearly appear. To show this, we present in Fig5 data for the susceptibility
167: together with data by Ballesteros et al. \cite{BallesterosII} for different
168: values of $L$ and for values of $p<0.8.$ Note that the tendency of the data
169: for $L\longrightarrow \infty $, seems to be towards $R_{\chi }(p=0.8)$. It
170: means that the effect of the finite size is to introduce a apparent increase
171: in the value of the normalized square width which should not exist for the
172: sample with $L=\infty $. If we consider the data in the $p$-dependent zone,
173: that is $1>p>0.8$, where there is small $L$ dependence, we can make and
174: extrapolation to $p\longrightarrow p_{p}$ (being $p_{p}$ the probability for
175: which the system suffers percolation: $p_{p}\approx 0.31$\cite{Stauffer})
176: that is not going to be affected by finite size effects. In our case we have
177: used a hyperbolic tangent to fit our data, $R_{\chi }(L,1-p)=R_{\chi
178: }(\infty )tanh[const(1-p)]$, leaving free the universal value of the
179: normalized square width $R_{\chi }(\infty )$ and the slope constant ($const)$%
180: . Results are shown in Fig5. For $L=40,60$ we find a p-independent universal
181: value $R_{\chi }(\infty )\approx 0.155$, very close to previously reported
182: values : $0.150(7)$ \cite{BallesterosII}. For $L=20,$ on the other hand,
183: finite size effects are important even in the region $(1>p>0.8)$, and the
184: extrapolation gives a somewhat higher value $R_{\chi }(\infty )\approx 0.19$.
185:
186: In conclusion, we have presented Monte Carlo data for diluted Ising systems
187: in the region where crossover to the diluted random universality class takes
188: place $(1>p>0.8).$ The evolution of the normalized square width for the
189: magnetization $R_{M}$ and the susceptibility $R_{\chi }$ indicates a smooth
190: transition with no critical probability $p_{c}$ (corresponding to a well
191: defined boundary between the pure and the randomly diluted universality
192: classes). The transition zone studied is $p$-dependent but $L$-independent.
193: This result is very convenient for extrapolation to the universal value $%
194: R_{\chi }(\infty )$ which is independent of $L.$
195:
196: We acknowledge financial support from CGCyT through grant PB96-0037.
197:
198: \newpage
199:
200: \begin{references}
201: \bibitem{Domb} For a review, see R.B. Stinchcombe, in Phase Transitions and
202: Critical Phenomena, edited by C. Domb and J.L. Lebowitz (Academic, New York,
203: 1983), Vol. 7.
204:
205: \bibitem{Harris} A. B. Harris, J. Phys. C 7, 1671 (1974).
206:
207: \bibitem{Lubensky} T. C. Lubensky and A. B. Harris, in Magnetism and
208: Magnetic Materials, edited by C. D. Graham, G. H. Lander and J. J. Rhyne,
209: AIP Conf. Proc. No. 24 (AIP, New York, 1975), p.99; A. B. Harris and T. C.
210: Lubensky, Phys. Rev. Lett. 33, 1540 (1974).
211:
212: \bibitem{LubenskyII} T. C. Lubensky, Phys. Rev. B 11, 3573 (1975).
213:
214: \bibitem{Grinstein} G. Grinstein and A. Luther, Phys. Rev. B 13, 1329
215: (1976).
216:
217: \bibitem{Aharony} A. Aharony, in Phase Transitions and Critical Phenomena,
218: edited by C. Domb and M.S. Green (Academic, New York, 1976), Vol. 6, p. 357.
219:
220: \bibitem{Chayes} J. T. Chayes, L. Chayes, D. S. Fisher, and T. Spencer,
221: Phys. Rev. Lett. 57, 2999 (1986).
222:
223: \bibitem{Kinzel} W. Kinzel and E. Domany, Phys. Rev. B 23, 3421 (1981); D.
224: Andelman and A. N. Berker, Phys. Rev. B 29, 2630 (1984).
225:
226: \bibitem{BallesterosI} H. G. Ballesteros, L. A. Fern\'{a}ndez, V. Mart\'{i}%
227: n-Mayor, A. Mu\~{n}oz Sudupe, G. Parisi and J. J. Ruiz-Lorenzo, J. Phys. A
228: 30, 8379 (1997).
229:
230: \bibitem{BallesterosII} H. G. Ballesteros, L. A. Fern\'{a}ndez, V. Mart\'{i}%
231: n-Mayor, A. Mu\~{n}oz Sudupe, G. Parisi and J. J. Ruiz-Lorenzo, Phys. Rev. B
232: 58, 2740 (1998).
233:
234: \bibitem{BallesterosIII} H. G. Ballesteros, L. A. Fern\'{a}ndez, V. Mart%
235: \'{i}n-Mayor, A. Mu\~{n}oz Sudupe, G. Parisi and J. J. Ruiz-Lorenzo, Nucl.
236: Phys. B 512[FS], 681 (1998).
237:
238: \bibitem{Folk} R. Folk, Yu. Holovatch and T. Yavors'kii, cond-mat/9909121.
239:
240: \bibitem{Blote} H. W. J. Bl\"{o}te, E. Luijten and J. R. Heringa, J. Phys.
241: A: Math. Gen. 28, 6289 (1995).
242:
243: \bibitem{Weinrib} A. Weinrib and B. I. Halperin, Phys. Rev. B 27, 413
244: (1983).
245:
246: \bibitem{BallesterosIV} H. G. Ballesteros and G. Parisi, cond-mat/9903230.
247:
248: \bibitem{Marques} M. I. Marqu\'{e}s and J. A. Gonzalo, Phys. Rev. E 60,
249: 2394 (1999); cond-mat/0001388.
250:
251: \bibitem{Wiseman} S. Wiseman and E. Domany, Phys. Rev. E 52, 3469 (1995).
252:
253: \bibitem{AharonyII} A. Aharony and A. B. Harris, Phys. Rev. Lett. 77, 3700
254: (1996).
255:
256: \bibitem{WisemanII} S. Wiseman and E. Domany, Phys. Rev. Lett. 81, 22
257: (1998).
258:
259: \bibitem{AharonyIII} A. Aharony, A. B. Harris and S. Wiseman, Phys. Rev.
260: Lett 81, 252 (1998).
261:
262: \bibitem{Wolff} U. Wolff, Phys. Rev. Lett. 62, 361 (1989).
263:
264: \bibitem{Heuer} H. O. Heuer, J. Phys. A: Math. Gen. 26, L333 (1993).
265:
266: \bibitem{Stauffer} D. Stauffer and A. Aharony, {\it Introduction to the
267: Percolation Theory} (Taylor \& Francis, London, 1994).
268: \end{references}
269:
270: \newpage
271:
272: {\bf Figure Captions}
273:
274: Fig1: Critical temperature vs. spin per site probability in the slightly
275: diluted zone. Our data is compared with the extrapolation performed for data
276: by Heuer et al. \cite{Heuer} and Ballesteros et al. \cite{BallesterosII}.
277:
278: Fig2: Normalized histograms for the susceptibility values obtained at
279: criticality. The values of the spin per site probability considered are (a)
280: 0.8,0.825,0.85, 0.875 and (b) 0.9,0.925,0.95,0.975.
281:
282: Fig3: Normalized square width for the magnetization vs. $1-p$, for values of
283: $L=20,40,60$. Dotted line is just a guide for the eye.
284:
285: Fig4: Normalized square width for the susceptibility vs. $1-p$, for values
286: of $L=20,40,60$. Dotted line is just a guide for the eye.
287:
288: Fig5: Normalized square width for the susceptibility vs. $1-p$ for values of
289: $L=20,40,60$ (circles), together with the data reported in \cite
290: {BallesterosII}. The two continuous lines indicate the universal value of $%
291: R_{\chi }$ for the randomly diluted Ising system reported in \cite
292: {BallesterosII}. The hyperbolic extrapolation of the data is indicated by a
293: segmented line for $L=40,60$ and by a doted line for $L=20$. \newpage
294: \begin{table}[tbp]
295: \caption{Effective critical exponents and normalized square widths for the
296: pure and randomly diluted Ising system}
297: \label{puretable}
298: \begin{tabular}{cccccc}
299: x & $\alpha $ & $\beta $ & $\gamma $ & $R_{\chi} $ & \\ \hline
300: pure (undiluted) & 0.11 & 0.3269 & 1.237 & 0.000 & \\
301: diluted & -0.051 & 0.3546 & 1.342 & 0.155 & \\
302: $\delta x/x$ & -1.4636 & 0.0847 & 0.0848 & $\infty$ &
303: \end{tabular}
304: \end{table}
305:
306: \newpage
307:
308: \begin{figure}
309: \protect\epsfxsize=16cm\protect\epsfysize=20cm\protect\epsfbox{fig1.ps}
310: %\epsfig{width=0.95 \linewidth,figure=fig5.eps} %\vbox{\vspace{3.9truecm}}
311: %\caption{}
312: \label{fig1}
313: \end{figure}
314: \newpage
315:
316: \begin{figure}
317: \protect\epsfxsize=16cm\protect\epsfysize=20cm\protect\epsfbox{fig2a.ps}
318: %\epsfig{width=0.95 \linewidth,figure=fig5.eps} %\vbox{\vspace{3.9truecm}}
319: %\caption{}
320: \label{fig2a}
321: \end{figure}
322:
323: \newpage
324:
325: \begin{figure}
326: \protect\epsfxsize=16cm\protect\epsfysize=20cm\protect\epsfbox{fig2b.ps}
327: %\epsfig{width=0.95 \linewidth,figure=fig5.eps} %\vbox{\vspace{3.9truecm}}
328: %\caption{}
329: \label{fig2b}
330: \end{figure}
331:
332: \newpage
333:
334: \begin{figure}
335: \protect\epsfxsize=16cm\protect\epsfysize=20cm\protect\epsfbox{fig3.ps}
336: %\epsfig{width=0.95 \linewidth,figure=fig5.eps} %\vbox{\vspace{3.9truecm}}
337: %\caption{}
338: \label{fig3}
339: \end{figure}
340:
341: \newpage
342:
343: \begin{figure}
344: \protect\epsfxsize=16cm\protect\epsfysize=20cm\protect\epsfbox{fig4.ps}
345: %\epsfig{width=0.95 \linewidth,figure=fig5.eps} %\vbox{\vspace{3.9truecm}}
346: %\caption{}
347: \label{fig4}
348: \end{figure}
349:
350: \newpage
351:
352: \begin{figure}
353: \protect\epsfxsize=16cm\protect\epsfysize=20cm\protect\epsfbox{fig5.ps}
354: %\epsfig{width=0.95 \linewidth,figure=fig5.eps} %\vbox{\vspace{3.9truecm}}
355: %\caption{}
356: \label{fig5}
357: \end{figure}
358:
359: \end{document}
360: