1: \documentclass[twocolumn,showpacs,preprintnumbers,amsmath,amssymb]{revtex4}
2:
3: %\documentclass[preprint,preprintnumbers,amsmath,amssymb]{revtex4}
4:
5: \usepackage{graphicx}% Include figure files
6: % \usepackage{psfig}% Include figure files
7: \usepackage{dcolumn}% Align table columns on decimal point
8: \usepackage{bm}% bold math
9:
10:
11: \begin{document}
12:
13: \title{
14: Dipolar coupling and multidomain states
15: in perpendicularly polarized nanostructures}
16: %
17:
18: \author{Alexei N.\ Bogdanov}
19: \altaffiliation[Permanent address: ]%
20: {Donetsk Institute for Physics and Technology,
21: 340114 Donetsk, Ukraine
22: }
23: \email{a.bogdanov@ifw-dresden.de}
24: %
25: %
26: \author{Ulrich K.\ R\"o\ss ler}
27: \thanks
28: {Corresponding author
29: % IFW Dresden,
30: % Postfach 270116, D--01171 Dresden, Germany.
31: % Tel.: +49-351-4659-542; Fax: +49-351-4659-537
32: }
33: \email{u.roessler@ifw-dresden.de}
34:
35: \affiliation{
36: IFW Dresden,\\
37: %Leibniz-Institut f{\"u}r Festk{\"o}rper- und Werkstoffforschung\\
38: P.O. Box 270116, D--01171 Dresden, Germany
39: }%
40:
41: \date{\today}
42:
43: \begin{abstract}
44: %
45: {
46: Stripe states in multilayer systems
47: with perpendicular polarization are investigated
48: by analytical calculations within a general
49: continuum approach, applicable to ferromagnetic,
50: ferroelectric, or ferroelastic nanoscale superlattices.
51: %
52: The competition between
53: the long-range depolarization effect and
54: short-range interlayer couplings
55: can stabilize monodomain states and unusual stripe phase
56: ground states with antiparallel polarization in adjacent layers.
57: %
58: Geometric parameters of stable stripe domain states
59: and the phase transitions lines
60: between single domain states, aligned
61: and antialigned stripe states have been derived.
62: %
63: The theory is applied to analyze multidomain states
64: and phase transitions in antiferromagnetically
65: coupled multilayers [CoPt]/Ru.
66: }
67: \end{abstract}
68:
69: \pacs{
70: 75.70.Cn,
71: % Magnetic properties of interfaces (multilayers, superlattices, heterostructures)
72: %
73: 75.70.Kw,
74: % Domain structure (including magnetic bubbles)
75: %
76: 77.80.-e,
77: % Ferroelectricity and antiferroelectricity
78: 77.80.Dj,
79: % Domain structure; hysteresis,
80: %
81: }
82: % %%% PACS numbers
83:
84: \keywords{ }%Use showkeys class option if keyword
85: %display desired
86:
87: \maketitle
88:
89: % \clearpage
90:
91: %
92: Multidomain states considerably influence
93: physical properties of condense matter
94: systems with spontaneous polarization.
95: %
96: Such spatially inhomogeneous patterns
97: form ground states of
98: ferromagnetic \cite{Kittel,Hubert98},
99: ferroelectric \cite{Scott06},
100: or ferroelastic \cite{Bratkovsky01} films.
101: %
102: Recently multidomain structures have been observed
103: in nanoscale magnetic films and multilayers
104: with strong perpendicular anisotropy
105: \cite{Gehanno97,Hamada02,Hellwig03,Itoh03,PRB04}
106: and in ferroelectric superlattices
107: \cite{Streiffer02,Scott06,Lee04}.
108: %
109: Similar spatially modulated states can also arise
110: in polar or magnetic liquid crystals
111: \cite{Lagerwall,Sayar03},
112: polar multiblock copolymer layers \cite{Ruzette05},
113: in superconducting films or
114: magnetic-superconductor hybrids \cite{Gillijns05},
115: and in shape memory alloy films \cite{Dong04}.
116: %
117: Multilayer systems with perpendicular polarization components
118: provide ideal experimental models to investigate fundamental
119: aspects of ordered structures and stable pattern formation
120: in confining geometries.
121: %
122: Control of such regular depolarization patterns is also
123: of practical interest.
124: %
125: E.g., patterns may provide templates,
126: which could be decorated by nanoparticles or macromolecules,
127: or they may be used to calibrate imaging techniques.
128: %
129: Particular interest in ferroelectric superlattices
130: is driven by the exciting possibility of
131: using ferroelectric nanostructures
132: in nonvolatile memory applications,
133: new microelectromechanical systems (MEMS),
134: and for nonlinear optics devices \cite{Scott06}.
135: %
136: In nanomagnetism, antiferromagnetically coupled
137: superlattices with strong perpendicular
138: anisotropy \cite{Hamada02,Hellwig03,Itoh03}
139: are considered as promising candidates
140: for nonvolatile magnetic recording media
141: and other applications \cite{Fullerton03}.
142: %
143: %This requires a control of the
144: %domain structures and the switching processes
145: %under fields in polarized layered systems.
146: %
147: According to recent experiments
148: \cite{Hamada02,Hellwig03,Scott06}
149: these nanoscale superlattices
150: are characterized by new multidomain
151: states, unusual depolarization processes,
152: and other specific effects which
153: have no counterpart in other classes
154: of media with perpendicular polarization.
155:
156: We present here a detailed analysis of
157: multidomain states
158: in magnetic nanoscale superlattices
159: with perpendicular polarization components.
160: %
161: We derive simplified micromagnetic equations
162: for equilibrium parameters of stripe domains
163: in nanoscale multilayer systems.
164: %
165: %This provides effective tools to calculate
166: %equilibrium parameters of the stripe states,
167: %
168: These mathematical tools provide
169: a clear description of the
170: multidomain processes and
171: reveal the physical mechanism underlying
172: their unconventional properties.
173: %
174: It was found that in contrast to
175: other bulk and nanomagnetic systems,
176: the magnetic states here are determined
177: by a close competition between interlayer
178: exchange and dipolar couplings.
179: %
180: The enhanced stray field couplings are responsible
181: for the unusual switching processes and
182: specific transformation of the domain patterns observed
183: in synthetic metamagnetic multilayer systems, as [CoPt]/Ru
184: and others \cite{Hamada02,Hellwig03,Itoh03}.
185: %
186: The depolarization effects
187: revealed in this paper
188: have a universal character.
189: They arise similarly in ferroelectric
190: superlattices and in other nanosized polarized media.
191:
192: %
193: For definiteness we consider stripe domains in
194: magnetic nanolayers.
195: %
196: We analyse superlattices
197: consisting of $N$ identical layers
198: with thickness $h$ separated by spacer layers
199: of thickness $s$ (Fig.~\ref{StripeGeometry}).
200: %
201: The perpendicular anisotropy fixes the easy
202: magnetization direction.
203: %
204: Within the single layers the magnetization
205: ${\mathbf{ M}_i}(\mathbf{r})$,
206: may be spatially inhomogeneous.
207: %
208: The energy can be written in a phenomenological approach as
209: %
210: \begin{eqnarray}
211: \label{energy0}
212: & & W_N = \sum_{i=1}^{N-1} \int \int j(\mathbf{r}_i,\mathbf{r}_j)
213: \, \mathbf{M}_{i} \cdot
214: \mathbf{M}_{i+1} dv_i\,dv_j +
215: \\
216: & & \sum_{i=1}^{N} \int \left[-\frac{K}{2}\left(\mathbf{M}_i
217: \cdot\mathbf{n} \right)^2
218: -\mathbf{H}^{(e)}\cdot \mathbf{M}_i
219: -\frac{1}{2}\mathbf{H}_d \cdot \mathbf{M}_i \right] d v_i\,
220: \nonumber
221: \end{eqnarray}
222: %
223: where integrals are over the volume $v_i$ of the single layers.
224: %
225: $\mathbf{H}^{(e)}$ and $\mathbf{H}_d (\mathbf{r})$
226: are the externally applied
227: and the depolarizing magnetic fields,
228: respectively.
229: %
230: The unity vector ${{\mathbf n}}$ designates
231: the normal to the film.
232: %
233: $K > 0$ is a perpendicular anisotropy.
234: %
235: The short-range interlayer coupling parameter
236: between (nearly) homogeneous layers
237: $J = \int\int j(\mathbf{r}_i,\mathbf{r}_j) dv_i\,dv_j > (<)$~0
238: favours (anti)parallel orientation of the neighbouring layers.
239: %
240: To investigate general effects of competing
241: stray field and exchange interlayer interactions,
242: we consider a simple model of a multidomain structure,
243: %
244: namely stripe domains with antiparallel magnetization
245: of magnitude $M \equiv |\mathbf{M}_i=\textrm{const}$
246: within all the magnetic layers.
247: %
248: The adjacent domains are
249: separated by \textit{thin} domain
250: walls with energy density $\sigma$.
251: %
252: The magnetic field $\mathbf{H}$ is applied
253: perpendicular to the layer surfaces.
254: %
255: The stripe structure is
256: described
257: by the widths of domains $d_{\pm}$
258: that are
259: polarized
260: in the directions parallel ($+$) and antiparallel ($-$)
261: to the field.
262: %
263: The geometry of stripes in a multilayer
264: structure is conveniently defined by
265: the stripe period $D = d_{+}+d_{-}$,
266: and a set of reduced parameters
267: %
268: \begin{eqnarray}
269: \label{pq}
270: q = \frac{d_{+}-d_{-}}{D}, \;
271: p = 2\pi\frac{h}{D}, \quad
272: \nu = \frac{s}{h}, \;
273: \tau = 1+\nu\,,
274: \end{eqnarray}
275: %
276: where $q$ is proportional to the average magnetization
277: in a stripe structure, $p$ is the reduced
278: thickness of single ferromagnetic layers,
279: $\nu$ is the thickness ratio between
280: magnetized layer and interlayer,
281: and $\tau$ fixes the superlattice period
282: (Fig. \ref{StripeGeometry}).
283:
284: %
285: The reduced energy $w_N = W_N/(2 \pi M^2 N$)
286: of a system with $N$ can be written
287: %
288: \begin{eqnarray}
289: \label{energy1}
290: w_N = \frac{2 \Lambda p}{\pi^2} - \frac{H q}{2 \pi M}
291: + \varkappa \left( 1 - \frac{1}{N} \right) + w_m (p,q)\,.
292: \end{eqnarray}
293: %
294: In Eq.~(\ref{energy1}), the term linear in $p$
295: is the energy of the domain walls.
296: %
297: The next two terms are Zeeman energy and
298: the short-range interlayer coupling, respectively.
299: %
300: The stray field energy $w_m (p,q)$
301: must be derived by solving
302: the corresponding magnetostatic problem
303: \cite{Suna85}.
304: %
305: The reduced energy (\ref{energy1})
306: depends on the two dimensionless materials parameters
307: %
308: \begin{eqnarray}
309: \label{kappa}
310: \varkappa = \frac{J}{2\pi M^2}, \quad
311: \Lambda = \pi \frac{l}{h}\,.
312: \end{eqnarray}
313: %
314: The strengths of the exchange coupling
315: is measured by $\varkappa $ given by
316: the ratio between the exchange and stray field energies.
317: %
318: The parameter $\Lambda$ characterizes
319: the balance between the domain wall energy and
320: stray field energies.
321: %
322: It is fixed by the ratio of
323: \textit{characteristic length}, $l=\sigma/(4\pi\, M^2)$,
324: which is a fundamental material parameter,
325: see Ref.~\onlinecite{Hubert98}) and the
326: thickness of the magnetic layers.
327:
328: %
329: Due to the mathematical identity of
330: electro- and magnetostatic equations
331: \cite{Hubert98} the multilayer
332: with stripes can be thought of
333: as a set of planes with ``charged''
334: stripes (Fig. \ref{StripeGeometry}).
335: %
336: For two such planes separated by an interlayer with
337: thickness $a=\omega h$,
338: the magnetostatic energy is
339: %
340: \begin{eqnarray}
341: \label{energyMS1}
342: f(\omega) = \frac{4}{p}
343: \sum_{n=1}^{\infty}
344: \frac{1 -(-1)^n \cos (nq)}{n^3} \exp(-n\,p\,\omega)\,.
345: \end{eqnarray}
346: %
347: The stray-field energy of the ``charges''
348: within the same plane is
349: $w_m^{(0)} = q^2 +f(0)$.
350: %
351: The dipolar coupling energy
352: between the ``poles'' on different
353: sides of the same layer is $f(1)$.
354: %
355: Hence, the magnetostatic energy of
356: individual layers is \cite{MalekKambersky,KooyEnz60}
357: %
358: \begin{eqnarray}
359: w_m^{(\textrm{self})} =q^2 +f(0)-f(1)
360: \label{energyMS0}
361: \end{eqnarray}
362: %
363: The stray field energy of the multilayer,
364: $w_m$
365: originally derived by Suna,\cite{Suna85}
366: can be written as a sum of
367: ``self'' energies of the layers and
368: interactions between them
369: %
370: \begin{eqnarray}
371: \label{energyMS}
372: w_m & = & w_m^{(\textrm{self})}+ w_m^{(\textrm{int})},
373: \\
374: w_m^{(\textrm{int})} & = &
375: \pm \frac{1}{N} \sum_{k=1}^{N-1} (N-k)F_k\,,
376: \nonumber\\
377: F_k & = & f(\tau k +1) + f(\tau k -1) -2 f(\tau k)\,.
378: \nonumber
379: \end{eqnarray}
380: %
381: The upper (lower) signs in $w_m^{(\textrm{int})}$
382: correspond to parallel (antiparallel) arrangement
383: of the polarization in adjacent layers
384: (Figs.~\ref{FMPeriods},~\ref{SolutionsD}).
385: %
386: We denote these modes
387: as \textit{ferro} (F) or \textit{antiferro}
388: (AF) stripes.
389: %
390: The factors $F_k$ equals
391: stray field coupling energies
392: between two layers separated by distance $\tau k$.
393: %
394: They are composed of the four contributions
395: of the magnetodipole coupling
396: between pairs of the planes bounding
397: the layers.
398: %
399: In particular, for two adjacent
400: layers, $k =1$ (Fig. \ref{StripeGeometry}),
401: the interactions
402: $1\leftrightarrow 3$
403: and $2\leftrightarrow 4$ yield
404: equal (positive) energy contributions $f(\tau)$,
405: while
406: $1\leftrightarrow 4$ and $2\leftrightarrow 3$
407: yield negative energy contributions,
408: $f(\tau+1)$ and $f(\tau-1)$, correspondingly.
409:
410: %
411: In common polarized systems
412: with characteristic sizes far
413: beyond the nanoscale range
414: the equilibrium domain sizes are usually
415: much smaller than the individual
416: layer thicknesses, $p \gg 1$.
417: %
418: Numerous observations indicate
419: that, as soon as domain sizes approaches
420: the layer thickness, coercitivity suppresses
421: the formation of regular multidomain patterns
422: \cite{Hubert98}).
423: %
424: This establishes a natural limit for domain
425: sizes in classical systems.
426: %
427: For $D \ll h$ there is no effective
428: dipole interaction between different surfaces.
429: %
430: Hence, in Eq.(\ref{energyMS})
431: for all nonzero $\omega$, one hase $ f(\omega) \ll f(0)$,
432: and the stray field energy of the multilayer
433: is reduced to $w_m^{(0)}$ \cite{Kittel}.
434: %
435: In such decoupled superlattices the multidomain
436: states should have similar properties as those
437: in isolated layers.
438: %
439: \begin{figure}
440: %
441: % \vspace{40mm}
442: \includegraphics[width=7.0cm]{Stripes_Fig_1.eps}
443: %
444: \caption{
445: \label{StripeGeometry}
446: A fragment of a superlattice:
447: two ferromagnetic
448: nanolayers (of thickness $h$)
449: with stripe domains are
450: separated by a nonmagnetic
451: spacer of thickness $s$.
452: %
453: The domains are coupled
454: by the exchange ($J$)
455: and magnetostatic forces.
456: %
457: }
458: \end{figure}%
459: %
460:
461: On the contrary, in perpendicular polarized
462: nanoscale films and multilayers
463: the periods of \textit{regular}
464: multidomain patterns is of the same
465: order as their thicknesses
466: or exceed these thicknesses
467: \cite{Gehanno97,Hellwig03,Streiffer02}.
468: %
469: In such systems dipole interactions
470: between different surfaces have
471: a sizable effect.
472: %
473: Mathematically, this is seen from slowly converging
474: sums of interaction terms between poles
475: far apart and on different internal surfaces.
476: %
477: For such structures, numerical evaluation becomes arduous,
478: and sharpened analytical methods are required.
479: %
480: To overcome the slow convergence in $w_m$,
481: we extend the method introduced in \cite{FTT80}.
482: %
483: With the help of the identity
484: $ \int_0^{\infty} t^{(m-1)}\exp(-nt) dt = m!/n^m$,
485: the infinite sums in Eq. (\ref{energyMS1})
486: can be transformed into integrals on the interval
487: $[0,1]$
488:
489: \begin{eqnarray}
490: f(\omega) - f(0)= \pi^2 (1-q^2) -2p \Omega(\omega)
491: \label{int1}
492: \end{eqnarray}
493:
494: where
495: %
496: \begin{eqnarray}
497: \label{Omega}
498: \Omega (\omega)= \omega^2 \int_0^1 (1-t)\ln
499: \left[1 + \frac{ \cos^2 \left(\pi q/2 \right)}
500: { \sinh^2 \left(\omega p t/2 \right)}\right]d t\,.
501: \end{eqnarray}
502:
503: Then
504: the dipolar stray field
505: energy (\ref{energyMS})
506: can be written as
507: %
508: \begin{eqnarray}
509: \label{energyStripeMC}
510: w_m = 1- \frac{2 p}{\pi^2} \Omega (1) \pm
511: \frac{2 p }{\pi^2 N} \sum_{k=1}^{N-1} (N-k)\Xi_k (\tau k),
512: \end{eqnarray}
513:
514: \begin{eqnarray}
515: \label{energyStripeMC2}
516: \Xi_k (\tau k) = 2 \Omega(\tau k)
517: -\Omega(\tau k +1)
518: -\Omega(\tau k -1)\,.
519: \end{eqnarray}
520: %
521: %
522: Minimization of $w_N$ (\ref{energy1}) with respect
523: to $p$ and $q$ yields the equilibrium parameters
524: of the stripes in the multilayers.
525: %
526: By the form of the multilayer energy Eq.~(\ref{energy1})
527: the widths of the stripes are independent of the
528: interlayer couplings $\varkappa$.
529: %
530: A complete analysis of these equations in
531: applied field will be published elsewhere
532: \cite{PRBstripes}.
533: %
534: Here we investigate the ground state of the
535: system in zero applied field, $\mathbf{H} = 0$.
536: %
537: In this case $q =0$ and the parameter
538: $p$ is derived from the equation
539: $d w_N /d p = 0$
540: %
541: \begin{eqnarray}
542: \Lambda = \Omega_p (1)
543: \mp \frac{1}{N}
544: \sum_{k=1}^{N-1} (N-k)\Xi_k^{(p)} (\tau k +1)
545: \label{equationP})
546: \end{eqnarray}
547: %
548: where $\Xi_k^{(p)} (\tau k +1) =
549: 2 \Omega_p (\tau k)
550: -\Omega_p (\tau k +1)
551: -\Omega_p (\tau k -1)$, and
552: $\Omega_p (\omega) = d \Omega /d p(\omega)$.
553: %
554: After elementary transformations
555: this function can be written
556: %
557: \begin{eqnarray}
558: \Omega_p (\omega)= -2 \omega^2 \int_0^1 t
559: \ln \left[ \tanh \left( \frac{p t}{2} \right) \right] d t\,.
560: \label{omegaP}
561: \end{eqnarray}
562: %
563: %
564: \begin{figure}
565: %
566: % \vspace{70mm}
567: \includegraphics[width=7.0cm]{Stripes_Fig_2.eps}
568: %
569: \caption{
570: \label{FMPeriods}
571: The equilibrium reduced period $D/l$ as a function
572: of the reduced layer thickness $h/l$ for ferro (F)
573: stripes in an $N=10$ multilayer for different values
574: of parameter $\nu$.
575: %
576: Inset shows the function $D/l$ ($h/l$)
577: for an isolated layer \cite{MalekKambersky}
578: and the parameters of the minimum point.
579: %
580: Dashed line indicates $D \propto \sqrt{h}$
581: fit corresponding to Kittel theory.
582: %%
583: }
584: \end{figure}%
585: %
586:
587:
588: \begin{figure}
589: %
590: % \vspace{70mm}
591: \includegraphics[width=7.5cm]{Stripes_Fig_3.eps}
592: %
593: \caption{
594: \label{SolutionsD}
595: The solutions $D/l$ ($h/l$) for ferro and
596: antiferro (AF) stripes in
597: systems with $\nu = 0.1$.
598: %
599: AF stripes exist for thickness
600: larger than the critical thickness $h_{cr}$.
601: %
602: For a two-layer systems ($N =2$) the
603: stripe period goes to infinity at the critical thickness.
604: %
605: For systems with $N > 2$ the critical
606: stripe period $D_{cr}$ has finite values.
607: %
608: The thin (red) line shows
609: the dependence $D(h)/l$ for the period of
610: F stripes in the case $N=4$.
611: %
612: The corresponding solutions $p(\Lambda)$
613: of Eq. (\ref{equationP}) for these parameters
614: are plotted in the Inset: F stripes (red) and
615: AF (blue) lines. Dashed line indicates
616: unstable solutions.
617: %
618: }
619: \end{figure}%
620: %
621: Typical solutions of Eq. (\ref{equationP})
622: are presented in Figs. \ref{FMPeriods}
623: and \ref{SolutionsD}.
624: %
625:
626: \textit{Ferro stripes}. These solutions
627: exist for any layer thickness Fig.~\ref{FMPeriods}.
628: %
629: In the limit of thick layers the period $D$ increases,
630: while it tends to infinity as $h$ tends to zero.
631: %
632: The role of the stray-field couplings changes
633: with the ratio of the interlayer thickness to
634: the thickness of the polarized layers, $\nu$.
635: %
636: For the limiting cases of small
637: and large
638: thickness of the spacer layer
639: given by the limits $\nu\ll 1$ or $\nu \gg 1$, respectively,
640: the dependence $D(h)$ approaches
641: a behaviour of isolated layers with
642: thickness $hN$ and $h$ correspondingly.
643: %
644: In the limit of small ratios $\nu$
645: the stripe period $D(h)$ approaches the period
646: for a F stripe state in a single layer with an effective
647: total polarized layer thickness $hN$.
648: %
649: For very large separation between the polarized layers,
650: i.e., for large ratios $\nu$ the stripe period is determined
651: by the properties of the decoupled single layers with thickness $h$.
652: %
653: The limiting solution for an isolated layer $D(h)$
654: (Inset Fig. \ref{FMPeriods})
655: obtained by \cite{MalekKambersky}
656: (see also Ref.~\onlinecite{Hubert98}) has
657: a minimum point with parameters
658: ($h_{min}/l = 0.96067$, $D_{min}/l = 16.3136$).
659: %
660: The nonmonotonic behaviour of $D(h)$
661: reflects the antagonistic role of magnetic charges
662: in the formation of the equilibrium stripes.
663: %
664: In the case of small domains $D \ll h$, which
665: is typical for classical systems,
666: the dipole interaction between different
667: surfaces of the layer is negligibly small,
668: and only the interaction between charges
669: on the same surface give a contribution
670: to the stray field energy.
671: %
672: For stripes with sizes $D \geq h$ the
673: interaction between charges from different
674: surfaces becomes a noticeable effect
675: and counteracts the interactions between
676: charges on the same surface.
677: %
678: This can be understood as a screening effect.
679: %
680: As the layer thickness decreases
681: this screening effect becomes stronger
682: and suppresses the stray field energy.
683: %
684: As a result, for $h < h_{min}$ the extension
685: of domains decreases both the sum of domain wall energies
686: and the stray field energy, and the domain period
687: increases exponentially with decreasing
688: layer thickness.
689: %
690: These simple energetical arguments demonstrate
691: that in nanoscale multilayers, where domain sizes
692: usually considerably exceed the thicknesses of
693: magnetic and interlayer, $h$ and $s$,
694: the interaction between magnetic poles through
695: the stack strongly influences the equilibrium
696: magnetic states.
697: %
698: %However, for finite values of $\nu$ the dependencies
699: %$D(h)$ markedly deviate from those of isolated layers \cite{Kiselev}.
700: %
701: Due to this effect
702: a non-monotonous dependence of stripe periods $D(h)$ arises
703: in multilayers between the limiting cases of large/small values
704: both of $\nu$ and $h$, as shown in \ref{FMPeriods}.
705: %
706:
707: \textit{AF stripes}. In this case the solutions
708: of Eq. (\ref{equationP}) $p(\Lambda)$ exist
709: on the finite interval $\Lambda < \Lambda_{cr}$
710: and consist of two branches with stable and unstable
711: solutions (Fig. \ref{SolutionsD}, Inset).
712: %
713: Correspondingly the equilibrium stripe states
714: exist only in a finite
715: range of the thickness $h < h_{cr}$ (Fig. \ref{SolutionsD}).
716: %
717: At a critical thickness $h_{cr}$ the period
718: reaches a critical value $D_{cr}$
719: for multilayer systems with $N > 2$ and tends to infinity
720: for two-layer system, $N=2$, (Fig. \ref{SolutionsD}).
721:
722:
723: %
724: \begin{figure}
725: %
726: % \vspace{70mm}
727: \includegraphics[width=7.5cm]{Stripes_Fig_4.eps}
728: %
729: \caption{
730: \label{gN}
731: The reduced values of the factor $g_N$ as
732: a function of $N$ and different values
733: of $\nu$: 1 ($\triangle$), 5 ($\square$), 0.1 ($\bigcirc$).
734: %
735: Inset gives functions $g_N (\nu)$
736: for different values $N$.
737: }
738: \end{figure}
739:
740: In multilayers with a ferromagnetic coupling ($J < 0$) both
741: the stray fields and exchange interactions favour parallel
742: arrangement of the magnetization across the stack.
743: %
744: Hence, the F-stripe mode is the ground state in such multilayers.
745: %
746: On the contrary, in systems with
747: antiferromagnetic interlayer couplings, $J> 0$,
748: the dipole and exchange forces have competing character.
749: %
750: As a result, three different equilibrium phases can exist in
751: such multilayers in zero field,
752: see phase diagram Fig.~\ref{Kappah}.
753: %
754: For sufficiently strong interlayer couplings
755: a monodomain phase with antiparallel correlations of adjacent layers
756: across the stack, and an AF stripe phase exist.
757: %
758: The transition between the monodomain and AF stripe states
759: for the two-layer system, $N=2$, is continuous,
760: as seen from the divergence of $D(h)$ (Fig.~\ref{SolutionsD}).
761: %
762: For systems with $N \geq 3$ the transition is first-order.
763: %
764: For weaker interlayer coupling $\varkappa$,
765: the F stripe phase becomes stable.
766: %
767: The transition between the F stripes and
768: the AF stripes or monodomain state is a topological
769: transition, and it is always first-order.
770: %
771: At a triple point, that weakly depends on
772: the multilayer repetitions $N$,
773: all three phases coexist.
774: %
775: Lines of first order transitions between the AF monodomain and
776: AF stripe phase in dependence on $N$ and layer thickness $h$
777: are shown in Fig.~\ref{hN}.
778: %
779: \begin{figure}
780: % \vspace{70mm}
781: \includegraphics[width=7.5cm]{Stripes_Fig_5.eps}
782: %
783: \caption{
784: \label{Kappah}
785: %
786: The phase diagram of magnetic states in variables
787: $h/l$ and the reduced coupling $\varkappa$ favouring
788: antiparallel orientation of adjacent layers for $N=4$.
789: %
790: Thick lines indicate the first-order transitions
791: between different phases.
792: %
793: They meet in a triple point.
794: %
795: The dashed line shows the stability limit
796: of the AF stripes.
797: %
798: Inset shows a difference between stripe periods
799: in F and AF modes.
800: %
801: Between the triple point
802: and the critical point AF domains
803: are metastable.
804: }
805: \end{figure}
806:
807: In the limit of large domains ($D \geq l$)
808: the expansion of the integral (\ref{Omega})
809: for $q=0$ yields
810: $\Omega (\omega) = 3/2 - \ln (p \omega/2)$
811: and energy (\ref{energy1}) can be transformed
812: in the form
813: %
814: \begin{eqnarray}
815: w_N = 1 - \frac{2 p_N}{\pi^2}
816: \left[ \frac{3}{2} -\ln \left( \frac{p_N}{2} \right)
817: -\tilde {\Lambda}_N \right]
818: + \varkappa \left( 1 - \frac{1}{N} \right),
819: \label{energy3}
820: \end{eqnarray}
821: %
822: where
823: $p_N =pN= 2 \pi (hN)/D$,
824: $ \tilde {\Lambda}_N = \Lambda/N + g_N(\nu)$.
825: %
826: Then, the energy $w_N$ (\ref{energy3}) has the
827: form of the energy for an \textit{isolated} layer
828: (see \cite{FTT80}) with ``effective'' thickness $hN$.
829: %
830: This means that in a system with large
831: domains ($D \gg h$)
832: the multilayer stack behaves as an effectively
833: coupled single layer and the domain period
834: approaches the solution for a single layer
835: with a total polarized thickness $hN$
836: with an effective characteristic parameter
837: $\tilde {\Lambda}$.
838:
839: %
840: The function $g_N(\nu)$ describes the influence
841: of finite spacers on the magnetic
842: properties of the multilayer
843: via the redistribution of the internal ``charges''
844: within the stack.
845: %
846: This function is given by
847: %
848: \begin{eqnarray}
849: g_N = - \frac{1}{N^2} \sum_{k=1}^{N-1} (N-k)\widetilde{G}(\nu)
850: -\ln N\,,
851: \label{gN}
852: \end{eqnarray}
853: %
854: where
855: $\widetilde{G}(\nu) = 2\tilde{g}(\tau k)-\tilde{g} (\tau k +1)
856: - \tilde{g} (\tau k -1) $
857: and
858: $\tilde{g}(\omega) = \omega^2 \ln(\omega)$.
859: %
860: The dependence of $g_N(\nu)$ on the number of layers $N$
861: and the ratio $\nu=s/h$ is shown in Fig.~\ref{gN}.
862: %
863: In the limit of small $\nu$
864: this function behaves as $g_N = a_N \nu$
865: where the factor
866: %
867: \begin{eqnarray}
868: a_N = 2 (1 -1/N ) \ln N
869: - 4 N^{-2} \sum_{k=1}^{N-1} k \ln k
870: \label{aN}
871: \end{eqnarray}
872: %
873: varies from $\ln(2)$ for $N =2$ to
874: unity as $N$ tends to infinity.
875: %
876: The equilibrium domain period in the limit of large $N$
877: is (cf.~\cite{FTT80})
878: %
879: \begin{eqnarray}
880: D = \pi hN \exp(\tilde {\Lambda}_N-1/2)\,.
881: \label{Dn}
882: \end{eqnarray}
883:
884: Recently, stripe domains have been investigated
885: in Co/Pt multilayers
886: \cite{Hamada02,Hellwig03,Hellwig05}.
887: %
888: AF stripes have been observed in Co/Pt multilayers \cite{Hamada02}.
889: %
890: F stripes have been investigated in a set of multilayers
891: [Co(4\AA) Pt (7\AA)]$_X$ (with $X$ from 5 to 160)
892: \cite{Hellwig05}.
893: %
894: The average domain width $D/2$ in our notation
895: plotted versus the total multilayer thickness is close
896: to the line typical for an isolated layer
897: (Inset, Fig.~\ref{FMPeriods}),
898: the minimum point $(h_{min},D_{min})$
899: is in the region $X \approx 20$ with a period $D$
900: about 150~nm.
901: %
902: This system has $\nu_f$ = 7/4=1.75.
903: However, the existing data
904: %(six points)
905: are insufficient for detailed analysis
906: and comparison with a theoretical dependencies
907: for $D(h)$ in the Fig.~\ref{FMPeriods}.
908: %
909: More substantial results have been obtained on the investigation
910: of antiferromagnetically coupled (via Ru) ferromagnetic blocks
911: [[Co(4\AA) Pt (7\AA)]$_{X-1}$Co(4\AA) Ru(9\AA)]$_N$
912: (with $N = 2$ to $10$ and $X= 2 $ to $12$), Ref.~[\onlinecite{Hellwig03}].
913: %
914: Strictly speaking magnetic properties of
915: such ferromagnetic multilayers may
916: strongly differ from those of single
917: layers (see Fig. \ref{FMPeriods}).
918: %
919: However, according to the results of \cite{Hellwig05}
920: these blocks can be modelled by a single effectively ferromagnetic
921: layer with total thickness $h = 11X -7$\AA.
922: %
923: Hence, $\nu (X) = 9/(11X-7)$ varies from 0.072 ($X$ =12)
924: to 0.6 ($X$ = 2).
925: %
926: The stripe periods, $D$ = 260 nm for
927: $X = 8$ \cite{Hellwig05}, are much larger than the
928: layer thicknesses $h$ ranging from 1.5 to 12.5~nm.
929: %
930: Thus, the approach Eq. (\ref{energy3}) can be applied to
931: describe these multidomain states.
932: %
933: In particular, the critical line $h_t(N)$ of the
934: first order transition between the
935: homogeneous antiferromagnetic state and F stripes
936: is derived from the following equation
937: %
938: \begin{eqnarray}
939: \pi^2 \varkappa(1-1/N)/2 =
940: \exp( - \pi l/(h N) -g_N \nu +1/2)
941: \label{transition}
942: \end{eqnarray}
943: %
944: and plotted in the Inset of Fig.~\ref{hN}.
945: %
946: It should be stressed that
947: despite the fact that energy (\ref{energy3})
948: has been reduced to the same functional form
949: as the energy for an isolated layer,
950: these two model have
951: different physical properties.
952: %
953: Namely the function $g_N(\nu)$ in
954: Eq.~(\ref{energy3}) describes
955: the influence of the internal ``charges''
956: on the equilibrium states of stripes.
957: %
958: The transition line $h_{t0} (N)$ between the antiferromagnetic
959: homogeneous state and F stripes for
960: an isolated layer ($g_N(\nu)$ =0)
961: is plotted for comparison
962: in the Inset of Fig.~\ref{hN}.
963: %
964: The difference between the two line
965: $(h_{t0}-h_{t})/h_{t0} = sN/(\pi l)$
966: depends on the ratio $s/l$ and
967: increases with increasing $N$.
968:
969: %
970: \begin{figure}
971: % \vspace{70mm}
972: \includegraphics[width=8.00cm]{Stripes_Fig_6.eps}
973: %
974: \caption{
975: \label{hN}
976: The phase diagram in variables
977: $h/l$ and $N$ with the
978: critical lines between the antiferro
979: single domain and the multidomain states.
980: %
981: The transitional line
982: between F stripes
983: and AF single domain state (Inset, solid points) is in a close agreement
984: with results obtained for magnetic Co/Pt multilayers
985: in \cite{Hellwig03}.
986: %
987: The line with hollow points indicates the transition
988: in a model with zero internal charges ($g_N =0$).
989: }
990: \end{figure}
991:
992: In conclusion, we have investigated a continuum approach
993: for the stripe multidomains that arise as regular
994: equilibrium depolarization structure of coupled multilayers.
995: %
996: The zero-field phase diagram displays transitions
997: between stripe states where adjacent layers are parallel
998: due to depolarization effects
999: and other phases where the layers are antiparallel
1000: under influence of short-range couplings.
1001: %
1002: In this AF case, it is possible to stabilize monodomain
1003: states in thin multilayer systems.
1004: %
1005: Furthermore, antiferro stripe phases can become stable
1006: in systems with sufficiently strong interlayer coupling.
1007: %
1008: The transitions between these phases are first-order.
1009: %
1010: In fact, within this model, the ferro stripe structure
1011: is metastable in the whole parameter range of the model.
1012: %
1013: Therefore, the different multidomain phases can coexist
1014: in multilayer systems and are transformed into each other
1015: by domain nucleation and growth processes.
1016: %
1017: Hence, severe hysteresis effects can arise
1018: in such multilayer systems subject to specific
1019: coercivity mechanisms.
1020: %
1021: In particular the AF monodomain state can arise in the form of
1022: two domains that transform into each other
1023: by a global reversal of the magnetization structure.
1024: %
1025: These domains can coexist out of equilibrium as remnant
1026: of an antiferro stripe state.
1027:
1028: % \begin{acknowledgments}
1029: %
1030: \textit{Acknowledgments}.
1031: %
1032: We thank I. Dragunov, O. Hellwig, N. Kiselev,
1033: J. McCord, V. Neu, and R. Sch\"afer
1034: for discussions.
1035: %
1036: A.N.B.\ thanks H.\ Eschrig for support and hospitality at the IFW Dresden.
1037: %
1038: This work was supported by Deutsche Forschungsgemeinschaft
1039: through project RO 2238/6.
1040: %
1041: %
1042: % \end{acknowledgments}
1043:
1044: % textbf{References}
1045:
1046: \begin{thebibliography} {99}
1047:
1048:
1049:
1050:
1051: % magnetic domains
1052:
1053: \bibitem{Kittel}
1054: C. Kittel,
1055: Phys. Rev. \textbf{70}, 965 (1946).
1056:
1057: \bibitem{Hubert98}
1058: A. Hubert, R. Sch{\"{a}}fer R.,
1059: {\textit{Magnetic Domains}}
1060: % The Analysis of Magnetic Microstructures}},
1061: (Springer-Verlag, Berlin, 1998);
1062: %
1063: V. G. Bar'yakhtar, et al.
1064: %A. N. Bogdanov, D. A. Yablonskii,
1065: Usp. Fiz. Nauk. {\textbf{156}}, 47 (1988),
1066: [Sov. Phys. Usp. {\textbf{31}}, 810 (1988)].
1067: % The physics of magnetic domains
1068:
1069: % ferroelectric domains
1070:
1071: \bibitem{Scott06}
1072: M. Dawber, K. M. Rabe, J. F. Scott,
1073: %Physics of thin-film ferroelectric oxides,
1074: Rev.\ Mod.\ Phys.\ \textbf{77}, 1083 (2005),
1075: J. F. Scott,
1076: Nanoferroelectrics: statics and dynamics
1077: J. Phys.:Cond. Mat. \textbf{18}, R361 (2006).
1078:
1079: % 180 domain in ferroelectric film
1080: %
1081: \bibitem{Streiffer02}
1082: S.K. Streiffer,
1083: J. A. Eastman, D. D. Fong, C.Thompson, A. Munkholm,
1084: M.V. Ramana Murty, O. Auciello, G. R. Bai, G. B. Stephenson,
1085: % Observation of Nanoscale 180 Stripe Domains
1086: %in Ferroelectric PbTiO3 Thin Films
1087: Phys.\ Rev.\ Lett.\ \textbf{89}, 067601 (2002);
1088: %
1089: D. D. Fong,
1090: G. B. Stephenson, S. K. Streiffer,
1091: J. A. Eastman, O.Auciello, P. H. Fuoss, C. Thompson,
1092: %Ferroelectricity in ultrathin perovskite films
1093: Science \textbf{304}, 1650 (2005);
1094: %
1095: A. Schilling, T. B. Adams, R. M. Bowman, J. M. Gregg,
1096: G. Catalan, J. F. Scott,
1097: %Scaling of domain periodicity with thickness
1098: % measured in BaTiO3 single crystal lamellae and
1099: % comparison with other ferroics
1100: Phys.\ Rev.\ B\ \textbf{47}, 024115 (2006).
1101: %
1102: %
1103: % elastic domains
1104: %
1105: \bibitem{Bratkovsky01}
1106: A. M. Bratkovsky, A. P. Levanyuk,
1107: %Phase Transitions, Stability, and Dielectric Response
1108: %of the Domain Structure in Ferroelectric-Ferroelastic Thin Films
1109: Phys.\ Rev.\ Lett.\ \textbf{86}, 3642 (2001);
1110: %Elastic domain structure and the transition between
1111: %polydomain and monodomain states in thin epitaxial films
1112: Phys.\ Rev.\ B\ \textbf{65}, 094102 (2002).
1113:
1114: \bibitem{Fullerton03}
1115: E. E. Fullerton et al.,
1116: % Margulies, D. T., Supper, N., Do, H., Schabes, M.,
1117: % Berger, A., Moser, A.
1118: % Antiferromagnetically coupled magnetic recording media.
1119: IEEE Trans. Magn. {\textbf{39}}, 639 (2003);
1120: J. $\dot{A}$kerman,
1121: % Toward a Universal Memory
1122: Science \textbf{308}, 508 (2005).
1123:
1124:
1125: \bibitem{MalekKambersky}
1126: Z. M\'{a}lek, V. Kambersk\'{y},
1127: Czech. J. Phys. \textbf{8}, 416 (1958).
1128:
1129: \bibitem{KooyEnz60}
1130: C.Kooy, U. Enz,
1131: % Experimental and theoretical study of the
1132: % domain configuration in thin layers of BaFe12O19
1133: Philips Res. Repts. \textbf{15}, 7 (1960).
1134:
1135: \bibitem{Suna85}
1136: A. Suna,
1137: J.\ Appl.\ Phys.\ \textbf{59}, 313 (1985);
1138: H. J. G. Draaisma, W. J. M. de Jonge,
1139: J.\ Appl.\ Phys.\ \textbf{62}, 3318 (1987).
1140: %
1141: %
1142: %
1143: \bibitem{Grunberg86}
1144: P. Gr\"unberg, R. Schreiber, Y. Pang, M. B. Brodsky, and H. Sowers,
1145: % Layered Magnetic Structures: Evidence for Antiferromagnetic Coupling
1146: % of Fe Layers across Cr Interlayers
1147: Phys.\ Rev.\ Lett.\ textbf{57}, 2442 (1986).
1148: %
1149: % ferroelectric and antiferroelectric superlattices
1150: % (anti)ferroelectric superlattices
1151: %
1152: \bibitem{Lee04}
1153: H.N. Lee,
1154: H.M. Christen, M.F. Chisholm,
1155: C.M. Rouleau, and D.H. Lowndes,
1156: %Strong polarization enhancement in asymmetric
1157: %three-component ferroelectric superlattices
1158: Nature \textbf{433}, 395 (2005);
1159: % ok
1160: %
1161: J. Sigman,
1162: D. P. Norton, H. M. Christen, P.H. Fleming, L.A. Boatner,
1163: %Antiferroelectric behavior in symmetric KNbO3/KTaO3 superlattices
1164: Phys.\ Rev.\ Lett.\ \textbf{88}, 097601 (2002);
1165: % ok
1166: H. M. Christen,
1167: E. D. Specht, S. S. Silliman, K. S. Harshavardhan,
1168: %Ferroelectric and antiferroelectric coupling in superlattices
1169: %of paraelectric perovskites at room temperature
1170: Phys.\ Rev. B.\ {\textbf{68}}, 020101R (2003).
1171: % ok
1172:
1173:
1174: \bibitem{Gehanno97}
1175: V. Gehanno, Y. Samson, A. Marty, B. Gilles, A. Chamberod,
1176: %Magnetic susceptibility and magnetic domain configuration
1177: %as a function of the layer thickness in epitaxial FePd(001)
1178: %thin films ordered in the L1(0) structure
1179: J.\ Magn.\ Magn.\ Mater. \ \textbf{172}, 26 (1997).
1180: %monolayers
1181: %
1182: % multidomain states in AFM multilayers
1183: %
1184: % Co/Pt perpendicular superlattices
1185:
1186: % observation of AF domains in Co/Ru multilayers
1187: \bibitem{Hamada02}
1188: S. Hamada, K. Himi, T. Okuno, K. Takanashi,
1189: %MFM observation of perpendicular magnetization and
1190: % antiferromagnetically coupled domains
1191: %in Co/Ru superlattices
1192: J.\ Magn.\ Magn.\ Mater. \ \textbf{240}, 539 (2002).
1193:
1194: \bibitem{Hellwig03}
1195: O. Hellwig, et al.
1196: %T. L. Kirk, J. B. Kortright, A.Berger, E. E. Fullerton,
1197: % A new phase diagram for layered antiferromagnetic films,
1198: Nature Mater. \textbf{2}, 112 (2003);
1199: %
1200: O. Hellwig, A. Berger, E. E. Fullerton,
1201: %Domain walls in antiferromagnetically coupled multilayer films,
1202: Phys.\ Rev.\ Lett.\ \textbf{91}, 197203 (2003);
1203:
1204:
1205: \bibitem{Hellwig05}
1206: %Magnetic reversal and domain structure in perpendicular
1207: %AF-coupled films,
1208: J. Magn. Magn. Mater. \textbf{290-291}, 1 (2005).
1209: %
1210: %other systems with perpendicular magnetization
1211: %
1212: \bibitem{Itoh03}
1213: H. Itoh, et al.
1214: % H. Yanagihara, K. Suzuki, E. Kita,
1215: %Coexistence of the uniaxial anisotropy and the
1216: %antiferromagnetic coupling in Co/Ir(111) superlattices,
1217: J. Magn. Magn. Mater. \textbf{257} 184 (2003);
1218: %
1219: Z.Y. Liu, S. Adenwalla,
1220: %Oscillatory Interlayer Exchange Coupling and Its Temperature Dependence
1221: %in [Pt/Co]3/NiO/[Co/Pt]3 Multilayers with Perpendicular Anisotropy,
1222: Phys.\ Rev.\ Lett.\ \textbf{91}, 037207 (2003).
1223: %ok
1224: %
1225: \bibitem{PRB04}
1226: U. K. R{\"o}{\ss}ler, A. N. Bogdanov,
1227: Phys.\ Rev. B.\ {\textbf{69}}, 094405 (2004);
1228: %Magnetic states and reorientation transitions in
1229: %antiferromagnetic superlattices
1230: %
1231: J. Magn. Magn. Mater. {\bf 269}, L287 (2004);
1232: % Synthetic metamagnetism -magnetic switching of
1233: % perpendicular antiferromagnetic superlattices
1234: %
1235: Phys.\ Rev. B.\ {\textbf{69}}, 184420 (2004).
1236: % Magnetic phases and reorientation transitions
1237: % in antiferromagnetically coupled multilayers
1238: %
1239: %theory of domains in ferroelectric superlattices
1240: %
1241: \bibitem{Bratkovsky00}
1242: A. M. Bratkovsky, A. P. Levanyuk,
1243: %Abrupt Appearance of the Domain Pattern and Fatigue
1244: % of Thin Ferroelectric Films
1245: Phys.\ Rev.\ Lett.\ \textbf{84}, 3177 (2000);
1246: % Very large dielectic response of thin ferroelectric
1247: %films with the dead layers
1248: Phys.\ Rev. B \textbf{63}, 132103 (2001);
1249: %
1250: V. A. Stephanovich, et al.
1251: %I. A. Lukyanchuk,M. G. Karkut,
1252: %Domain-Enhanced Interlayer Coupling in
1253: %Ferroelectric/Paraelectric Superlattices
1254: Phys.\ Rev.\ Lett.\ \textbf{94}, 047601 (2005).
1255: %
1256: %
1257:
1258: %Stripes in ferroelectric and antiferroelectric liquid crystals
1259: % for liquid crystal display
1260: %
1261: \bibitem{Lagerwall}
1262: S. T. Lagerwall. Ferroelectric and Antiferroelectric
1263: Liquid Crystals (Wiley-VHC, Weinheim, 1999),
1264: L.U.Ruibo, et al.
1265: % W. U. Shin-Tson, X. U. Keshu,
1266: %Fabrication of Surface-Stabilized Ferroelectric
1267: %Liquid Crystal Display with Stripe-Shaped Domain Structure
1268: Jpn. J. Appl. Phys. \textbf{42}, 1628 (2003).
1269: %
1270: \bibitem{Sayar03}
1271: M. Sayar, M. Olivera de la Cruz, S.I. Stupp,
1272: % Polar order in nanostructured organic materials
1273: Europhys. Lett. \textbf{61}, 334 (2003).
1274: %
1275: \bibitem{Ruzette05}
1276: A.-V. Ruzette, L. Leibler,
1277: % Block copolymers in tomorrow's plastics
1278: Nature Mater. \textbf{4} 19 (2005).
1279: %
1280: \bibitem{Gillijns05}
1281: W. Gillijns,
1282: A. Y. Aladyshkin, M. Lange, M. J. Van Bael, V.V. Moshchalkov,
1283: %Domain-Wall Guided Nucleation of Superconductivity in Hybrid
1284: %Ferromagnet-Superconductor-Ferromagnet Layered Structures
1285: Phys.\ Rev.\ Lett.\ \textbf{95}, 227003 (2005);
1286: % ok
1287: %
1288: V. Jeudy, C. Gourdon, and T. Okada,
1289: %Impeded Growth of Magnetic Flux Bubbles in
1290: %the Intermediate State Pattern of Type I Superconductors
1291: Phys.\ Rev.\ Lett.\ \textbf{92}, 147001 (2004).
1292: % ok
1293: %
1294: % shape memory alloys thin films
1295: %
1296: \bibitem{Dong04}
1297: J. W. Dong,
1298: J. Q. Xie, J. Lu, C. Adelmann, C. J. Palmstr{\o}m,
1299: J. Cui, Q. Pan, T. W. Shield, R. D. James,
1300: %Shape memory and ferromagnetic shape memory effects
1301: %in single-crystal Ni2MnGa thin films
1302: J.\ Appl.\ Phys.\ \textbf{95}, 2593 (2004);
1303: %
1304: A. N. Bogdanov,
1305: A. DeSimone, S. M\"uller, U.K. R\"o\ss ler,
1306: %Phenomenological theory of magnetic-field-induced
1307: %strains in ferromagnetic shape-memory materials
1308: J.\ Magn.\ Magn.\ Mater. \ \textbf{261}, 204 (2003).
1309:
1310: \bibitem{Choe99}
1311: S.-B. Choe, S.-C. Shin,
1312: % Stability of stripe magnetic domains in magnetic multilayers:
1313: % Prediction and observation
1314: Phys.\ Rev.\ B \textbf{59}, 142 (1999).
1315:
1316:
1317:
1318: %
1319: \bibitem{Choe99}
1320: S.-B. Choe, S.-C. Shin,
1321: % Stability of stripe magnetic domains in magnetic multilayers:
1322: % Prediction and observation
1323: Phys.\ Rev.\ B \textbf{59}, 142 (1999).
1324: %
1325: \bibitem{Labrune01}
1326: M. Labrune, A. Thiaville,
1327: % Micromagnetic structure in multilayer films with
1328: % moderate perpendicular anisotropy
1329: Eur.\ Phys.\ J.\ B \textbf{23}, 17 (2001).
1330: %
1331: \bibitem{FTT80}
1332: A. N. Bogdanov, D. A. Yablonskii.
1333: %Theory of the domain structure in ferrimagnets,
1334: Fiz. Tverd. Tela {\textbf{22}}, 680 (1980),
1335: [Sov. Phys. Solid State {\textbf{22}}, 399 (1980)].
1336: %
1337: \bibitem{PRBstripes}
1338: N.S. Kiselev, I.E. Dragunov, U. K. R{\"o}{\ss}ler, A. N. Bogdanov,
1339: to be published.
1340: % Phys.\ Rev. B. should be written soon.
1341: %
1342: \bibitem{Kiselev}
1343: The particular behaviour of $D(h)$
1344: was first obtained for a two-layer system
1345: in Ref.~[\onlinecite{PRBstripes}].
1346:
1347: \end{thebibliography}
1348:
1349: \end{document}
1350:
1351: %
1352: