cond-mat0106334/ram.tex
1: %Lviv, 02.11.2000
2: %\documentclass{ws-p8-50x6-00}
3: \documentclass{ws-p8-50}
4: \usepackage{emlines}
5: \begin{document}
6: 
7: \title{Phase Transition in the Random Anisotropy Model}
8: 
9: \author{M. Dudka}
10: 
11: \address{Institute for Condensed Matter Physics, Ukrainian Acad.
12: Sci., UA-79011 Lviv, Ukraine \\E-mail: maxdudka@icmp.lviv.ua}
13: 
14: \author{R. Folk}
15: 
16: \address{Institut f\"ur theoretische Physik,
17: Johannes Kepler Universit\"{a}t Linz, A-4040 Linz,
18: Austria\\E-mail: folk@tphys.uni-linz.ac.at}
19: 
20: \author{Yu. Holovatch}
21: 
22: \address{Institute for Condensed Matter Physics, Ukrainian Acad.
23: Sci., UA-79011 Lviv and
24: Ivan Franko National University of Lviv,
25: UA-79005 Lviv, Ukraine\\E-mail: hol@icmp.lviv.ua}
26: 
27: \maketitle
28: 
29: \abstracts{The influence of a local anisotropy of random
30: orientation on a ferromagnetic phase transition is studied for 
31: two cases of anisotropy axis distribution. To this end a model of a random 
32: anisotropy magnet is analyzed by means of the field theoretical 
33: renormalization group approach in two loop
34: approximation refined by a resummation of the asymptotic
35: series. The one-loop result of Aharony
36: indicating the absence of a second-order phase
37: transition for an isotropic
38: distribution of random anisotropy axis at space dimension $d<4$ is
39: corroborated. For a cubic distribution the accessible
40: stable fixed point leads to disordered
41: Ising--like critical exponents.}
42: %============================================================================
43: %                          SECTION  I
44: %============================================================================
45: \section{Introduction}\label{I}
46: 
47: Modern understanding of universal properties of matter in the
48: vicinity of critical points is mainly due to the
49: application of the renormalization group
50:  (RG) ideas.\cite{books} Applied to the problems of
51: condensed matter physics in the early 1970s, the
52: RG technique proved to be a powerful
53: tool to study critical phenomena. For example, expressions for
54: critical exponents governing the magnetic phase transition
55:  in regular systems are known by now with record
56: accuracy both for isotropic\cite{Kleinerta} ($O(m)$ symmetrical)
57: and cubic\cite{Kleinertb} magnets. The RG
58:  approach also sheds light on the influence of structural
59: disorder on ferromagnetism.  In the present article we will apply
60: the field theoretical RG approach to
61: study peculiarities of magnetic behaviour influenced by  disorder
62: in a form of random anisotropy axis.\cite{Harris73}  It is special
63: pleasure for us to dedicate this paper to Prof. Hagen Kleinert on
64: the occasion of his 60th anniversary. His contribution to the
65: field is hard to be overestimated.
66: 
67: Although an influence of a weak quenched structural disorder on
68: universal properties of a ferromagnetic phase
69: transition has already been a problem of 
70: intensive 
71: study for several decades, there remains a number of unsettled
72: questions. Here, one should distinguish between random site,
73: random-field and random anisotropy
74: magnets. A weak quenched disorder
75: preserves second-order phase transition in
76: three-dimensional ($d=3$) random site magnets\cite{note1} but can
77: destroy this transition in random field systems\cite{note2} for
78: $d<4$. The situation for the random-anisotropy magnets is not so 
79: clear.
80: 
81: Typical examples of random-anisotropy
82: magnets are amorphous rare-earth -- transition metal alloys.  Some
83: of these systems order magnetically and for the description
84: of the ordered structure it has been proposed\cite{Harris73} to
85: consider a regular lattice of magnetic ions, each of them being 
86: subjected to a local anisotropy
87:  of random orientation. The Hamiltonian of this random
88: anisotropy model (RAM) reads:\cite{Harris73}
89: \begin{equation}
90: {\cal H} =  - \sum_{{\bf R},{\bf R'}} J_{{\bf R},{\bf R'}}
91: \vec{S}_{\bf R} \vec{S}_{\bf R'} -D_0\sum_{{\bf R}} (\hat {x}_{\bf
92: R}\vec{S}_{\bf R})^{2}, \label{1}
93: \end{equation}
94: where $\vec{S}_{\bf R}$ is an $m$-component vector on a lattice
95: site ${\bf R}$, $J_{{\bf R},{\bf R'}}$ is an exchange interaction,
96: $D_0$ is an anisotropy strength, and $\hat{x}_{\bf R}$ is a unit vector
97: pointing in the local (quenched) random direction of an uniaxial
98: anisotropy.
99: 
100: The model has been investigated by a variety of techniques including
101: mean--field theory,\cite{mfa} computer simulations,\cite{comp}
102: $1/m$--ex\-pan\-si\-on,\cite{1overm} renormalization
103: group
104: $\varepsilon$--expansion.
105: \cite{Aharony75,Chen77,Pelcovits78} The limit case of an
106: infinite anisotropy has been subject of a detailed study as
107: well.\cite{infram,Fischer85}  However the nature of
108: low--temperature phase  in RAM is not completely clear up to now,
109: although several low--temperature phases were discussed
110: like ferromagnetic ordering,\cite{mfa,comp} spin--glass
111: phase,\cite{comp,1overm} and quasi long--range
112: ordering.\cite{qlro}
113: 
114: The nature of ordering is connected with the distribution of
115: the random variables $\hat{x}_{\bf R}$ in (\ref{1}). For
116: an isotropic distribution arguments similar to those applied
117: by Imry and Ma\cite{Imry75} for a random-field Ising model bring
118: about the absence of ferromagnetic order for space
119: dimensions $d<4$,\cite{Pelcovits78,Ma78} whereas anisotropic
120: distributions may lead to a ferromagnetic order.\cite{cubic}
121: 
122: Application of Wilson RG technique to
123: RAM with the isotropic distribution of a local anisotropy axis
124: suggests\cite{Aharony75}  the possibility of ``runaway"
125: solutions of the recursion equations. Such a behaviour was
126: interpreted as a smeared transition. However this result was
127: obtained in first order of 
128: the $\varepsilon$--expansion and
129: remains to be confirmed also in higher orders.
130: 
131: Here, we will report results obtained by means of the field
132: theoretical RG technique in two loop
133: approximation refined by a resummation of
134: the resulting asymptotic series. We will
135: consider two cases of distribution of the random anisotropy
136: axis and show that a ferromagnetic second-order phase
137: transition takes place only when the
138: distribution is non-isotropic. Moreover we will show that the RAM
139: provides another example of a disordered model, where the
140: only possible new critical behaviour is of ``random Ising" type,
141: similar to the site--diluted magnets.\cite{note1}  More
142: detailed results can be found in Refs.\cite{Dudka00a,Dudka00b}
143: 
144: %============================================================================
145: %                          SECTION  II
146: %============================================================================
147: \section{Isotropic case}\label{II}
148: 
149: In order to deal with  quenched disorder, one  way to obtain the 
150: effective Hamiltonian of a RAM is to make use of the replica trick. For a 
151: given configuration of quenched random variables $\hat{x}_{\bf R}$ in
152: (\ref{1}) the partition function may then be written in a form of
153: functional integral of a Gibbs distribution depending on
154: $\hat{x}_{\bf R}$. To average over
155: configurations one should complete the model by choosing certain
156: distribution of $\hat{x}_{\bf R}$. We will analyze two cases: 
157: first the isotropic case, where 
158: the random vector $\hat x$ points with equal probability in any
159: direction of the $m$-dimensional hyperspace, 
160: and second the cubic case, where $\hat x$ lies along the edges of 
161: $m$-dimensional hypercube (cubic case). Other distributions may be
162: considered as well. In the first case the distribution function
163: reads:
164: \begin{equation}
165: p(\hat x)\equiv\left(\int d^m\hat x\right)^{-1}
166: \!=\frac{\Gamma(m/2)}{2\pi^{m/2}}. \label{2}
167: \end{equation}
168: Following the above described program, one ends up with the
169: replica $n\rightarrow 0$ limit of the effective
170: Hamiltonian\cite{Aharony75}:
171: \begin{eqnarray}
172: \label{3} {\cal H}_{eff} &{=}& {-}\int d^dR \left
173: \{\frac{1}{2}\!\left[{\mu_0}^2|\vec{\varphi}|^2{+}
174: |\vec{\nabla}\vec{\varphi}|^2\right]{+}u_0 |\vec{\varphi}|^4{+}
175: v_0\!\sum_{\alpha=1}^n|\vec{\phi}^\alpha|^4{+}\right. \nonumber\\
176: && \left. w_0\!\sum_{\alpha,\beta=1}^n\sum_{i,j=1}^{m}\!
177: \phi_i^\alpha\phi_j^\alpha\phi_i^\beta\phi_j^\beta \right\},
178: \end{eqnarray}
179: where ${\mu_0}^2$ and $u_0$, $v_0$, $w_0$ are defined by $D_0$ and
180: familiar bare couplings of an $m$-vector model, and
181: $\vec{\phi}^{\alpha}\equiv\vec{\phi}^{\alpha}_{\bf R}$ is a
182: $m$-dimensional
183: vector,  $|\vec{\varphi}|^2=\sum_{\alpha=1}^{n} |\vec{\phi}^\alpha|^2$.
184: The bare couplings are restricted by $u_0>0$, $v_0>0$, $w_0<0$. 
185: Furthermore, values of $u_0$ and $w_0$ are related to appropriate 
186: cumulants of the distribution function (\ref{2}) and their ratio 
187: equals $w_0/u_0=-m$. Note that the symmetry of $u_0$ and $v_0$ terms
188: corresponds to the random site $m$-vector model.\cite{Grinstein76}
189: However the $u_0$-term has an opposite sign.
190: 
191: In order to study long-distance properties of the Hamiltonian
192: (\ref{3}), we use the field-theoretical RG approach.\cite{books} Here 
193: the critical point of a system corresponds to a stable fixed 
194: point (FP) of the RG 
195: transformation. We apply the massive field theory renormalization
196: scheme\cite{Parisi80} performing renormalization at fixed space
197: dimension $d$ and zero external momenta.
198: In  two-loop approximation we get\cite{Dudka00a} expressions
199: for the RG functions
200: in form of asymtotic series in renormalized couplings
201: $u,v,w$.
202: \begin{figure} [htbp]
203: \begin{center}
204: {\input {isotr.pic}}
205: \end{center}
206: \vspace{-2.5cm} \caption{ \label{fig1} Fixed points of the RAM
207: with isotropic distribution of a local anisotropy axis. The 
208: f\/ixed points located in the octant $u>0,v>0,w<0$ are shown.  The
209: filled box shows the stable fixed point, the cross denotes typical
210: initial values of couplings.}
211: \end{figure}
212: 
213: As it was mentioned in the introduction, the only known
214: RG results for RAM with isotropic
215: distribution of the local anisotropy axis so far are those
216: obtained in first order in $\varepsilon$.\cite{Aharony75} In
217: total one gets eight fixed points.
218: All FPs with $u>0,v>0,w<0$ appear to be
219: unstable for $\varepsilon>0$ except of the ``polymer" $O(n=0)$
220: FP III which is stable for all $m$ (see Fig.
221: \ref{fig1}). However the presence of a stable FP is not a sufficient 
222: condition for a second-order phase transition. In 
223: order to be physically relevant, the FP should be 
224: accessible from the initial values of couplings.  This is not the case for 
225: the location of FPs shown in Fig. \ref{fig1}. Indeed starting
226: from the region of physical initial conditions (denoted
227: by the cross  in the Fig. \ref{fig1}) in the plane of
228: $v=0$ one would have to cross  the separatrix joining the
229: unstable FPs I and VI. This is not possible and so one never
230: reaches the stable FP III. As far as both FPs
231: I and VI are strongly unstable with respect to $v$, FP III is not 
232: accessible for arbitrary positive $v$ either. Finally, runaway solutions of 
233: the RG equations show that the second-order 
234: phase transition is absent in the model. 
235: \begin{table}[htp]
236: \begin{center}
237: \caption {\label{tab1} Resummed values of the fixed points and
238: critical exponents for isotropic case in
239: two-loop approximation for $d=3$. We absorb the value of a one-loop 
240: integral into  the normalization of couplings.
241: }
242: \begin{tabular}{|c|c|c|c|c|c|c|}
243:    \hline
244: FP &$m$ & $u^*$ & $v^*$ & $w^*$ &
245:   $\nu$ &$\eta$\\
246:    \hline
247: I&$\forall m$ &0 &0 &0& &\\ \hline II&2 &0 &0.9107 &0 &0.663
248: &0.027\\ &3 &0 &0.8102 &0 &0.693 &0.027\\ &4 &0 &0.7275 &0 &
249: 0.724&0.027\\
250:  \hline
251: III&$\forall m$ &1.1857 &0 &0 &0.590 &0.023\\ \hline IV&2 &-0.0322
252: &0.9454 &0& 0.668&0.027\\ &3 &0.1733 &0.6460& 0&0.659 &0.027 \\ &4
253: &0.2867&0.4851&0&0.653 &0.028\\ \hline VI&2 &1.4650 &0 &-1.6278
254: &0.449 &-0.028\\ \hline VIII&2 &0.7517 &0.7072 &-0.3984  &0.626
255: &0.031\\ &3 &0.8031 &0.5463 & -0.3305 & 0.620&0.029\\ &4 &0.8349
256: &0.4545 & -0.2888 &0.617 &0.029\\ \hline
257: \end{tabular}
258: \end{center}
259: \end{table}
260: The main question of interest here is whether the above described
261: picture of runaway solutions is not an artifact of the
262: $\varepsilon$--expansion.
263: To check this, we used a more refined analysis of the FPs
264: and their stability, considering the series for
265: RG functions directly at
266: $d=3$.\cite{Parisi80} It is known that series of this type are 
267: at best asymptotic  and a resummation procedure
268: has to be applied to obtain reliable data on their
269: basis. We make use of Pad\'e--Borel resummation
270: techniques,\cite{Baker78} first writing the
271: RG functions as resovent
272: series\cite{Watson74} in one auxiliary variable and then
273: performing resummation. Numerical values of the
274: FPs are given in  Table \ref{tab1}. Resummed
275: two-loop results qualitatively confirm the picture obtained in first-order  
276: $\varepsilon$--expansion: stability of the FPs does not change after 
277: the resummation. This supports the conjecture of 
278: Aharony\cite{Aharony75} that accessible stable FP for the RAM with 
279: isotropic distribution of the local anisotropy axis 
280:  is absent. 
281: In the table, we list values of the 
282:  correlation length
283: and pair correlation function critical exponents $\nu$ and $\eta$ 
284: which are resummed in a similar way. As they are calculated in unstable FPs 
285: they have rather to be considered as effective ones.
286: 
287: %============================================================================
288: %                          SECTION  III
289: %============================================================================
290: \section{Cubic case}\label{III}
291: Let us now consider the second example of anisotropy axis
292: distribution, when the vector $\hat x_{\bf R}$ (\ref{1}) points
293: only along one of the $2m$ directions of axes $\hat k_i$ of a
294: cubic lattice :
295: \begin{equation}
296: p(\hat x)=\frac{1}{2m}\sum_{i=1}^m[\delta^{(m)}(\hat x-\hat k_i)+
297: \delta^{(m)}(\hat x+\hat k_i)]. \label{4}
298: \end{equation}
299: The {\em rationale} for such a choice  is to mimic the situation
300: when an amorphous magnet still ``remembers" the initial (cubic) lattice
301: structure. Repeating the procedure described in the previous
302: section, one ends up with the following effective Hamiltonian
303:  which is of interest in the limit $n\rightarrow
304: 0$\cite{Aharony75}:
305: \begin{eqnarray}
306: {\cal H}_{eff}&{=}&-\int d^dR
307: \left\{\frac{1}{2}\left[{\mu_0}^2|\vec{\varphi}|^2{+}
308: |\vec{\nabla}\vec{\varphi}|^2\right]{+}u_0|\vec{\varphi}|^4{+}
309: v_0\sum_{\alpha=1}^n|\vec{\phi}^\alpha|^4{+}\right. \nonumber\\ &
310: & \left.w_0\sum_{i=1}^m\sum_{\alpha,\beta=1}^n
311: {\phi_i^\alpha}^2{\phi_i^\beta}^2{+}
312: y_0\sum_{i=1}^m\sum_{\alpha=1}^n{\phi_i^\alpha}^4\right\}.
313: \label{5}
314: \end{eqnarray}
315: Here, the bare couplings are $u_0>0$, $v_0>0$,
316: $w_0<0$. The $y_0$ term is generated  when the
317: RG transformation is applied
318: and may be of either sign. The symmetry of
319: $w_0$ terms differs in (\ref{3}) and (\ref{5}).  Furthermore,
320: values of $w_0$ and $u_0$ differ for Hamiltonians (\ref{3}) and
321: (\ref{5}) but their ratio equals $-m$ again.
322: 
323: We apply the massive field theory renormalization
324: scheme\cite{Parisi80} and get the RG
325: functions in two-loop approximation.\cite{Dudka00b} As  in
326: the previous case we reproduce the first-order
327: $\varepsilon$-results.\cite{Aharony75} Now one gets 14 FPs.
328: However, in first order of the $\varepsilon$--expansion all FPs
329: with $u>0,v>0,w<0$ appear to be unstable for $\varepsilon>0$
330: except of the ``polymer" $O(n=0)$ FP III which
331: is stable for all $m$ but not accessible (see Fig. \ref{fig2}). Now
332: the account of the $\varepsilon^2$-terms qualitatively changes the
333: picture. Indeed, the system of equations for the FPs appears
334: to be degenerated  at the one loop level. As known from
335: other cases in two loop order this leads to the appearance of a
336: new FP which is stable and is expressed by a
337: $\sqrt{\varepsilon}$ series.\cite{Grinstein76} The
338: possibility of such scenario was predicted already in
339: Refs.\cite{cubic} However it remained unclear whether there exist
340: any other accessible stable FPs.
341: \begin{figure} [htbp]
342: \begin{center}
343: {\input {cubic.pic}}
344: \end{center}
345: \vspace{-2.5cm}
346: \caption{ \label{fig2} Fixed points of the RAM with
347: distribution of a local anisotropy axis along hypercube axes for
348: $v=0$. The only f\/ixed points located in the region $u>0,w<0$ are
349: shown. Filled boxes show the stable fixed points, the cross denotes
350: typical initial values of couplings.} \end{figure}
351: 
352: Applying  Pad\'e--Borel resummation we get
353: 16 FPs. Values of the FPs with $u^*>0,v^*>0,w^*<0$ are
354: listed in the Table \ref{tab2}. The last FP XV in Table 
355: \ref{tab2} corresponds to the stable FP of 
356: $\sqrt\varepsilon$-expansion. It has coordinates with $u^*=v^*=0$, $w^*<0$ 
357: and $y^*>0$ and is accessible from the typical initial values of couplings 
358: (shown by a cross in the Fig. \ref{fig2}). Applying the 
359: resummation procedure, we have not found any other stable 
360: FPs in the region of interest. The effective Hamiltonian (\ref{5}) at 
361: $u=v=0$ in the replica limit $n\rightarrow0$ reduces to a product of
362: $m$ effective Hamiltonians of a weakly diluted quenched random
363: site Ising model. This means that for any value of $m>1$ the system
364: is characterized by the same set of critical
365: exponents as those of a weakly
366: diluted quenched random site Ising model. In the Table, we give 
367: \ref{tab2} values of critical exponents in
368: the other FPs as well: if the flows from initial values of couplings
369: pass near the these FPs, one may observe an effective critical
370: behaviour governed by these critical exponents.
371: 
372: \begin{table}[htbp]
373: \begin{center}
374: \caption
375: {\label{tab2} Resummed values of the FPs and critical
376: exponents for cubic distribution
377: in two-loop approximation for $d=3$. We absorb the value of a one-loop 
378: integral into  the normalization of couplings.
379: }
380: %\tabcolsep1.4mm
381:    \begin{tabular}{|c|c|c|c|c|c|c|c|}
382:    \hline
383: FP &$m$ & $u^*$ & $v^*$ & $w^*$ & $y^*$ &  $\nu$ &$\eta$\\
384:    \hline
385:    I&$\forall m$ &0 &0 &0&0 &1/2 &0 \\
386:    \hline
387:  &2 &0 &0.9107 &0& 0 &0.663 &0.027\\
388: II&3 &0 &0.8102 &0 & 0&0.693 &0.027\\
389: &4 &0 &0.7275 &0 & 0&0.720&0.026\\ \hline
390:  III&$\forall m$ &1.1857 &0 &0& 0 &0.590 &0.023 \\
391:  \hline
392: V&$\forall m$ &0 &0 &0 &1.0339&0.628 &0.026\\ \hline
393:  VI&3 &0.1733
394: &0.6460 &0 & 0&  0.659&0.027\\
395: &4 &0.2867 &0.4851 & 0& 0&0.653 &0.027\\
396: \hline
397:  VII&$\forall m$ &2.1112 &0 &-2.1112 &0&1/2 &0\\ \hline
398:  &2 &0 &1.5508 &0 &-1.0339& 0.628 &0.026\\
399:  VIII&3 &0 &0.8393 &0 &-0.0485& 0.693 &0.027\\
400:  &4 &0 &0.5259 &0 &0.3624& 0.709 &0.026\\
401: \hline
402:  IX&3 &0.1695 &0.7096 & 0& -0.1022&0.659& 0.027\\
403:  &4 &0.2751 &0.4190 & 0& 0.1432&0.653 &0.027\\
404: \hline
405: X&$\forall m$ &0.6678 &0 &-0.6678 &1.0339&0.628 &0.026\\
406: \hline
407: XV&$\forall m$ &0&0&-0.4401&1.5933&0.676&0.031\\
408: \hline
409: \end{tabular}
410: \end{center}
411: \end{table}
412: 
413: %============================================================================
414: %                          SECTION IV
415: %============================================================================
416: \section{Conclusions}\label{IV}
417: We applied the field theoretical
418: RG approach to  analyze critical
419: behaviour of a model of random anisotropy
420: magnets with isotropic and cubic distributions of a local
421: anisotropy axis. The origin of a low--temperature phase in this
422: model is not completely clear. General arguments based on an
423: estimate of the energy for formation of magnetic
424: domains\cite{Imry75} lead to the conclusion that for $d<4$ a
425: ferromagnetic order is absent.\cite{Pelcovits78,Ma78} However,
426: these arguments do not take into account entropy which may be
427: important for disordered systems.\cite{Fischer85} Furthermore,
428: they do not apply  for anisotropic distributions
429: of the random axis.\cite{cubic}
430: 
431: In the RG analysis the absence of a
432: ferromagnetic second-order phase transition corresponds to  the 
433: lack of a stable FP of the RG
434: transformation. However in the case of RAM with isotropic
435: distribution of a local anisotropy axis the scenario differs. Our
436: two-loop calculation leads to a $O(n=0)$
437: symmetric FP which is stable for any value of
438: $m$ for both isotropic and cubic distributions of a random
439: anisotropy axis. Note that this
440: FP is not accessible from the initial values of
441: the couplings. We checked the location of the FPs up to
442: second order in the 
443: $\varepsilon$--expansion and by
444: means of a fixed $d=3$ technique refined by Pad\'e--Borel
445: resummation.
446: 
447: In the case of isotropic distribution of a random
448: anisotropy axis our analysis supports
449: the conjecture of Aharony\cite{Aharony75} based on results linear in
450: $\varepsilon$ about runaway solutions of the
451: RG equations. For the cubic
452: distribution we get two stable FPs. One of them (FP III in Fig. 
453: \ref{fig2}) is not accessible as in the isotropic case.  But the disordered 
454: Ising-like FP (FP XV in Fig.  \ref{fig2}) may be reached 
455: from the initial values of couplings. Applying the resummation procedure we 
456: have not found any other stable FPs in the region of interest. This
457: means that RAM with cubic distributions of a random
458: anisotropy axis is governed by a  set of
459: critical exponents of a weakly site
460: diluted quenched Ising model.\cite{Grinstein76,note4}
461: 
462: To conclude we want to attract attention to a certain similarity in
463: the critical behaviour of both random-site\cite{Grinstein76} and
464: random-anisotropy\cite{Harris73} quenched magnets:
465: if there appears a {\em new critical behaviour} it {\em always is
466: governed by critical exponents of
467: site-diluted Ising type}. The above calculations of a ``phase diagram"
468: of RAM are based on two-loop expansions improved by a resummation
469: technique. Once the qualitative picture
470:  becomes clear there is no need
471: to go into higher orders of a perturbation theory as far as the
472: critical exponents of site-diluted Ising model are known by now with
473: high accuracy.\cite{note4}
474: 
475: %============================================================================
476: %                          Acknowledgement
477: %============================================================================
478: \section*{Acknowledgments}
479: Yu. H. acknowledges helpful discussions with  Mykola Shpot. This work
480: has been supported in part by "\"{O}ster\-rei\-chi\-sche Nationalbank
481: Ju\-bi\-l\"a\-ums\-fonds" through the grant No 7694.
482: 
483: \begin{thebibliography}{99}
484: \bibitem{books}
485: D.J. Amit  {\em Field Theory, the Renormalization
486: Group, and Critical Phenomena.} 1st ed. (World Scientific, Singapore, 1984);
487: J.  Zinn-Justin, {\em Quantum Field Theory and
488: Critical Phenomena} (Oxford University Press, Oxford, 1989); H.
489: Kleinert and V.
490: Schulte-Frohlinde, {\em Critical
491: Properties of $\phi^4$-Theories.}(World Scientific , Singapore, 2001).
492: 
493: \bibitem{Kleinerta}
494: H. Kleinert {\em et al},
495: \Journal{\PLB}{272}{39}{1991},
496: {\em ibid.} {319}{545}{1993};
497: R. Guida and J.
498: Zinn-Justin, \Journal{{\em J. Phys.}
499: A}{31}{8103}{1998}.
500: 
501: \bibitem{Kleinertb}
502: H. Kleinert and
503: V. Schulte-Frohlinde, \Journal{\PLB}
504: {342}{284}{1995};
505: J.M. Carmona, A.
506:  Pelissetto, and E. Vicari,
507:  \Journal{{\em Phys. Rev.} B}{61}{15136}{2000}.
508: 
509: \bibitem{Harris73}
510: R. Harris, M. Plischke, and M.J.
511: Zuckermann, \Journal{\PRL}{31}{160}{1973}.
512: 
513: \bibitem{note1}  A weak quenched random site dilution is
514: irrelevant, if the heat capacity critical exponent $\alpha_{pure}$
515: of the undiluted system is negative (A.B. Harris,
516: \Journal{{\em J. Phys.} C}{7}{1671}{1974}). For positive
517: $\alpha_{pure}$ dilution causes change in the universality
518: class (for a recent
519: review see e.g.  R. Folk, Yu.
520: Holovatch, and T. Yavors'kii, \Journal{{\em Phys. Rev.} 
521: B}{61}{15114}{2000}).
522: 
523: \bibitem{note2} See e.g. reviews by D.P. Belanger
524: and A.P. Young (\Journal{\em
525: JMMM}{100}{272}{1991}); D.P.  Belanger, eprint:
526: cond-mat/0009029.
527: 
528: \bibitem{mfa}
529: B. Derrida and J. Vannimenus,
530:  \Journal{{\em J. Phys.} C}{13}{3261}{1980}; R. Harris and D. 
531:  Zobin, \Journal{{\em J. Phys.}
532: F}{7}{337}{1977}; E. Callen, Y.J. Liu, and J.R.  Cullen, \Journal{{\em 
533: Phys. Rev.}B}{16}{263}{1977}; J.D. Patterson, G.R. 
534: Gruzalski, and D.J. Sellmyer, 
535: \Journal{{\em Phys. Rev.} B}{18}{1377}{1978};R. Alben, J.J. 
536: Becker, and M.C. Chi, \Journal{{\em J.  
537: Appl. Phys.}}{49}{1653}{1978}.
538: 
539: \bibitem{comp}
540: M.C. Chi and R. Alben,
541: \Journal{\em J. Appl. Phys.}{48}{2987}{1977};
542: M.C. Chi and T. Egami,
543: \Journal{\em J. Appl.
544: Phys.}{50}{1651}{1979};R. Harris and S.H.
545: Sung, \Journal{{\em J. Phys.} F}{8}{L299}{1978};C.
546: Jayaprakash and S.
547: Kirkpatrick, \Journal{{\em Phys.
548: Rev.}B}{21}{4072}{1980}.
549: 
550: \bibitem{1overm}
551: 
552: Y.Y. Goldschmidt,
553:  \Journal{\NPB}{225}{123}{1983};
554: Y.Y. Goldschmidt,
555: \Journal{{\em Phys. Rev.} B}{30}{1632}{1984}; A.
556: Khurana, A. Jagannathan, and
557: J.M. Kosterlitz, \Journal{\NPB}{40}{1}{1984};
558: A. Jagannathan, B. Shaub, and
559: J.M. Kosterlitz,
560: \Journal{\NPB}{265}{324}{1986};Y.Y. Goldschmidt
561:  and A.  Aharony, \Journal{{\em Phys. Rev.}
562: B}{32}{264}{1985}.
563: 
564: \bibitem{Aharony75}
565: A. Aharony,
566: \Journal{{\em Phys. Rev.} B}{12}{1038}{1975}.
567: 
568: \bibitem{Chen77}
569: J.H. Chen and T.C. Lubensky,
570: \Journal{{\em Phys. Rev.}B}{16}{2106}{1977}.
571: 
572: \bibitem{Pelcovits78}
573: R.A. Pelcovits, E. Pytte, and
574:  J. Rudnick, \Journal{\PRL}{40}{476}{1978}; R.A.
575: Pelcovits, \Journal{{\em Phys. Rev.}
576: B}{19}{465}{1979}; E. Pytte, \Journal{{\em Phys.
577: Rev.} B}{18}{5046}{1978}.
578: 
579: \bibitem{infram}
580: A.J. Bray and M.A. Moore, \Journal{{\em
581: J. Phys.} C}{18}{L139}{1985}; A.B. Harris, R.G.
582: Caflisch and J.R. Banavar,
583: \Journal{{\em Phys.  Rev.}B}{35}{4929}{1987}.
584: 
585: \bibitem{Fischer85}
586: K.M. Fischer and A. Zippelius, \Journal{{\em J. Phys.} 
587: C}{18}{L1139}{1985}.
588: 
589: \bibitem{qlro}
590: A. Aharony and E. Pytte, \Journal{\PRL}{45}{1583}{1980};
591: R. Fisch, \Journal{{\em Phys. Rev.} B}{57}{269}{1998};
592: D.E. Feldman, \Journal{\em JETP
593: Lett}{70}{135}{1999}; J. Chakrabati,
594: \Journal{\PRL}{81}{385}{1998}.
595: 
596: \bibitem{Imry75}
597: Y. Imry and S.-k. Ma,
598: \Journal{\PRL}{35}{1399}{1975}.
599: 
600: \bibitem{Ma78}
601: S.-k. Ma and J. Rudnick,
602: \Journal{\PRL}{40}{589}{1978}.
603: 
604: \bibitem{cubic}
605: D. Mukamel and G. Grinstein,
606: \Journal{{\em Phys. Rev.} B}{25}{381}{1982}; A.L.
607: Korzhenevskii and A.A.
608: Luzhkov, \Journal{\em Sov.
609: Phys.JETP}{67}{1229}{1988}.
610: 
611: \bibitem{Dudka00a} M. Dudka, R. Folk,
612: and Yu. Holovatch,
613: \Journal{\em Cond. Matt. Phys.}{4}{77}{2001} 
614: 
615: \bibitem{Dudka00b} M. Dudka, R. Folk,
616: and Yu. Holovatch, {\em Cond. Matt. Phys.} (submitted). 
617: 
618: \bibitem{Grinstein76}
619: G. Grinstein and A. Luther,
620: \Journal{{\em Phys. Rev.} B}{13}{1329}{1976}.
621: 
622: \bibitem{Parisi80}
623: G. Parisi (1973) (unpublished); \Journal{\em J. Stat.
624: Phys.}{23}{49}{1980}.
625: 
626: \bibitem{Baker78}
627: G.A. Jr. Baker, B.G. Nickel, and D.I. Meiron, 
628: \Journal{{\em Phys. Rev.} B} {17}{1365}{1978}.
629: 
630: \bibitem{Watson74}
631: P.J.S. Watson, \Journal{{\em J. Phys.}
632: A.}{7}{L167}{1974}.
633: 
634: 
635: \bibitem{note4} Up to  now the critical exponents of a weakly 
636: diluted quenched Ising model are calculated in RG technique with 
637: record six-loop accuracy in massive theory scheme: A. Pelissetto and E.  
638: Vicari, \Journal{{\em Phys. Rev.} B} {62}{6393}{2000}. In minimal 
639: subtraction scheme five loop results are available:  B.N. Shalaev, S.A. 
640: Antonenko, and A.I. Sokolov, \Journal{{\em Phys. Lett.} 
641: A}{230}{105}{1997}; R. Folk, Yu. Holovatch, and T. 
642: Yavors'kii, \Journal{\em J. Phys. Stud.}{2}{213}{1998}; 
643: \Journal{\em JETP Lett.}{69}{747}{1999}.
644: \end{thebibliography}
645: \end{document}
646: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
647: %% The End
648: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
649: 
650: 
651: