1: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2: % manscr.: W. Jezewski, W. Kuczynski, and J. Hoffmann, %
3: % "Dielectric relaxation in chevron surface stabilized %
4: % ferroelectric liquid crystals" %
5: % %
6: % formatted in RevTeX 3.0 %
7: % submitted for publication in PRE %
8: % ========================================================================== %
9: % e-mail: jezewski@ifmpan.poznan.pl %
10: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
11: \documentstyle[pre,aps]{revtex}
12: %\documentstyle[seg,preprint,pre,aps]{revtex}
13: \def\btt#1{{tt$\backslash$#1}}
14: \hyphenation{SSFLC}
15:
16: \begin{document}
17: %\bibliographystyle{seg}
18: %\draft command makes pacs numbers print
19: \draft
20: \wideabs{
21: \title{Dielectric relaxation in chevron surface stabilized ferroelectric
22: liquid crystals}
23: \author{W. Je\.zewski, W. Kuczy\'nski, and J. Hoffmann}
24: \address{Institute of Molecular Physics, Polish Academy of Sciences,\\
25: Smoluchowskiego 17, 60-179 Pozna\'n, Poland}
26: \date{9 November 2005}
27: \maketitle
28:
29: \begin{abstract}
30: The dielectric response of surface stabilized ferroelectric liquid crystals
31: with chevron layer structure is studied within low and intermediate frequency
32: ranges, characteristic for collective molecular excitations. By analytically
33: solving the dynamic equation for collective molecular fluctuations under
34: a weak alternating electric field, it is demonstrated that chevron cells
35: stabilized by both nonpolar and polar surface interactions undergo at medium
36: frequencies two Debye relaxation processes, connected with two chevron slabs,
37: on opposite sides of the interface plane. This result is confirmed,
38: experimentally, making use of the electro-optic technique. Based on
39: qualitative arguments supported by microscopic observations of zigzag defects
40: at different frequencies and amplitudes of the external electric field, it is
41: shown that, at low frequencies, the electro-optic response of chevron samples
42: is determined by three kinds of motions of zigzag walls. The first two
43: dynamic categories are related to collective relaxation processes at weak
44: fields, within smectic A layers forming zigzag walls, and drift or creep
45: motions of thick walls occuring at stronger field amplitudes. Dynamic
46: processes of the third kind correspond to sliding of zigzag walls, which
47: appear at yet stronger field amplitudes, but below the switching threshold.
48:
49: \end{abstract}
50:
51: % insert suggested PACS numbers in braces on next line
52: \pacs{PACS numbers: 61.30.Dk, 61.30.Gd, 61.30.Hn, 81.40.Tv}
53: }
54: %\widetext
55: %\narrowtext
56: %\twocolumn
57:
58: \section{Introduction}
59: Thin ferroelectric liquid crystal cells with bookshelf and chevron structures
60: play the key role in developing high-resolution and large-area display screen
61: technologies\cite{la}. The main technical advantage of these
62: systems is the possibility of high-speed electro-optic switching between
63: bistable orientational states stabilized by surface
64: interactions\cite{c1,h1,h2,ri,xu,co}. Consequently, studies of the dynamic
65: behavior of surface stabilized ferroelectric liquid crystals (SSFLC) have
66: mostly been focused on ferroelectric switching processes. In the case of
67: the chevron geometry, a challenging problem is the inclusion of the process of
68: a rapid reorientation of molecules at the chevron interface in the theoretical
69: description of switching phenomena\cite{m1,m2,br,ha}. As compared to
70: investigations of the switching dynamics, the behavior of chevron
71: systems in the presence of the alternating external electric field of small
72: amplitudes, less than the switching threshold, has not been intensively studied.
73: Collective molecular excitations generated by a weak sinusoidal applied voltage
74: have been analyzed theoretically by solving a one-dimensional field equation
75: describing the dynamic distribution of the azimuthal angle of the molecular
76: $c$ director (the projection of the molecular director on the smectic plane)
77: with respect to the axis perpendicular to boundary plates, and then by
78: determining the dielectric permittivity\cite{p1,ka}. However, the solution to
79: this equation has been derived by assuming an unrealistic condition that, at the
80: chevron interface, the azimuthal angle of the $c$ director does not depend on
81: the external field\cite{p1,ka}. Clearly, such an assumption is equivalent to
82: the requirement that the strength of the interface anchoring interactions is
83: infinite.
84:
85: In this paper, the dynamic equation that describes collective molecular
86: excitations in uniform symmetric SSFLC with the chevron structure under the
87: influence of a weak sinusoidal electric field is solved for more natural
88: interface anchoring conditions, allowing the azimuthal orientation of
89: molecules to depend on the applied electric field also at the chevron
90: interface. The surface interactions are assumed to include both nonpolar and
91: polar couplings. In general, the existence of polar surface interactions, in
92: addition to the nonpolar ones, leads to a difference in anchoring energies on
93: two bounding plates. It is proved that the difference between the surface
94: energies can be modified by applying the electric field across boundary
95: plates. As will be shown, the alternating voltage induces in uniform SSFLC
96: with bookshelf and chevron structures a nonuniform contribution to the long
97: time average of the azimuthal angle. Such a change in the time average of
98: molecular orientation gives rise to a variation of anchoring energies on cell
99: plates, compared to the case of the stationary molecular orientation. It is
100: argued here that the differentiation of effective surface energies, changed by
101: a weak alternating voltage, causes collective relaxation processes in two
102: slabs of uniform chevron SSFLC to run in distinct manners, with separate
103: relaxation times. By imposing a constraint on the effective anchoring
104: energies, these relaxation times can be interrelated. Resulting calculated
105: dielectric loss spectra are shown to be in qualitative agreement with
106: experimental spectra, obtained by employing the electro-optic
107: method\cite{pi,ku}.
108:
109: Experimental investigations of chevron SSFLC in the low-frequency regime have
110: revealed a strong dependence of the electro-optic response on the voltage
111: amplitude, even for small amplitudes, much less than the switching
112: threshold\cite{p2}. By carrying out microscopic observations and by performing
113: measurements of the electro-optic loss, it is shown here that the low-frequency
114: electro-optic response of chevron samples is determined by the dynamics of
115: zigzag walls\cite{ri,c2,li}. Dynamical processes occuring at very weak fields
116: are identified with collective relaxation motions of molecules in smectic A
117: layers, which form thin and thick zigzag walls of various sizes. It is found
118: that, in the low-frequency regime, these motions can be described by a spectrum of
119: Debye processes characterized by a continuous, powerlike distribution of
120: relaxation times. The change of the low-frequency electro-optic loss as the
121: voltage increases is attributed to the appearance of creeping motions of thick
122: zigzag walls. Dynamic processes responsible for changes of the low-frequency
123: electro-optic loss under further growth of the voltage are ascribed to sliding
124: motions and gradual disappearance of zig-zag walls.
125: \section{Theoretical analysis of the dielectric response at intermediate
126: frequencies}
127: \subsection{Director equation of motion}
128: Cooperative dynamics in bookshelf and chevron cells is usually studied by
129: analyzing fluctuations of the azimuthal angle $\phi$ in the smectic plane,
130: along the surface normal direction $X$\cite{h2,ri,p1}, as illustrated in
131: Fig. \ref{f1}. The contribution to the dielectric response, due to
132: fluctuations of the angle $\phi$ under an applied alternating field, is
133: usually significant in a wide range of medium frequencies. Therefore, this
134: range is called here the intermediate frequency region. Assuming appropriate
135: boundary conditions for the azimuthal angle, its fluctuations can be described
136: independently in each chevron slab\cite{h2,ri}, by means of a dynamic field
137: equation. In the presence of the sinusoidal external electric field applied
138: along the $X$ axis and alternating with the frequency $\omega$, this equation
139: reads
140: \begin{equation}
141: \frac{\partial^2\phi}{\partial x^2}-\gamma\frac{\partial\phi}{\partial t}=
142: A\sin\phi\cos\omega t\,\,\,,
143: \end{equation}
144: where $\gamma=\gamma_\phi/K$ with $\gamma_\phi$ being the azimuthal viscosity
145: and $K$ being the twist elastic constant. The quantity $A=P_sE/K$ with $P_s$
146: denoting the local spontaneous polarization and $E$ being the electric field
147: amplitude. Boundary conditions for Eq. (1) should involve preferred
148: (zero-field) angle values at border surfaces as well as at the chevron
149: interface. In uniform chevron cells, the $c$ director is pretilted
150: (in the absence of the external field) by $\pm\phi_0$, where the convention
151: that $+$ and $-$ refer respectively to quantities on upper and lower chevron
152: slabs is used (this convention will henceforth be adopted). Consequently, for
153: the zero-field equilibrium state of uniform chevrons, the azimuthal angle $\phi$
154: is required here to obey the customary relation
155: \begin{equation}
156: \phi_\pm(0)=\mp\phi_0\,\,\,,
157: \end{equation}
158: where $\phi_\pm (0)$ refers to values of $\phi$ immediately above and below
159: the interface, respectively. The amount of the pretilt $\phi_0$ is, in general,
160: nonzero and depends on the molecular and layer tilt angles and surface
161: interaction strengths\cite{m1}. For the downward orientation of the
162: polarization (as shown in Fig. \ref{f1}), the anchoring surface energies can
163: be supposed in the form\cite{h1,h2,m1,na}
164: \begin{equation}
165: W_\pm(\phi)=-\gamma_1\cos^2(\phi\pm\phi_0)\pm\gamma_2\cos(\phi\pm\phi_0)\,\,\,,
166: \end{equation}
167: where $\gamma_1$ and $\gamma_2$ are respectively the nonpolar and polar
168: surface energy strengths, the same on both bounding plates. Hence, the boundary
169: conditions for $\phi$ near its zero-field equilibrium value $\phi_0$ can be
170: expressed in the case of downward-oriented chevron cells as\cite{h1,m1,m2}
171: \begin{equation}
172: \frac{\partial}{\partial x}\phi_{\pm}(x=\pm\frac{d}{2})\simeq\pm v_\pm
173: [\phi_\pm(x=\pm\frac{d}{2})\pm\phi_0]
174: \end{equation}
175: with $d$ denoting the thickness of a chevron cell and
176: \begin{equation}
177: v_\pm=2\lambda_1\mp\lambda_2
178: \end{equation}
179: being the combined surface interactions, where $\lambda_1=\gamma_1/K$ and
180: $\lambda_2=\gamma_2/K$. As the relations (4) and (5) show, the spatial
181: derivative of $\phi$ depends on the combined surface interactions (taking for
182: $\gamma_2\ne 0$ different values at top and bottom boundary plates). For weak
183: external fields and strong anchoring interactions at boundary and interface
184: surfaces, one can assume, in addition to the condition (4), that\cite{m1,m2}
185: \begin{equation}
186: \phi_\pm(\pm d/2)=\phi_\pm (0)\,\,\,.
187: \end{equation}
188:
189: In the static case ($\omega=0$), Eq. (1) reduces to the Euler-Lagrange equation
190: \begin{equation}
191: \frac{\partial^2}{\partial x^2}\phi=A\sin\phi\,\,\,.
192: \end{equation}
193: Using conditions (4) and (6), one finds in the case of downward oriented
194: chevron cells in a weak constant field the following solution of Eq. (4)
195: \begin{equation}
196: \phi_\pm(x)=\mp\phi_0+c f_\pm(x)+{\cal O}(A^2)\,\,\,,
197: \end{equation}
198: where
199: \begin{equation}
200: f_\pm(x)=\pm\frac{A}{2}\sin\phi_0(\frac{d}{2v_\pm}\pm\frac{d}{2}x-x^2)\,\,\,.
201: \end{equation}
202: According to Eqs. (5) and (9), the field-induced distribution of $\phi(x)$,
203: determined by the functions $f_\pm$, is different in the top and bottom
204: chevron slabs, provided that $A\ne 0$ (in particular,
205: $|\phi_+(0)|\ne|\phi_-(0)|$). This is a result of the difference between
206: $v_+$ and $v_-$.
207:
208: For weak alternating electric fields, the solution of Eq. (1), in the case of
209: downward polarization orientation, can be written as
210: \begin{eqnarray}
211: \phi_\pm(x,t)=&\mp&\phi_0+b_0^\pm+b_1^\pm x\nonumber\\
212: &+& F_\pm(x)\cos(\omega t-\beta_\pm)+{\cal O}(A^2)\,\,\,,
213: \end{eqnarray}
214: with $F_\pm(x)$ having series expansions
215: \begin{equation}
216: F_\pm(x)=\sum_{l=0}^{\infty}a_l^\pm x^l\,\,\,,
217: \end{equation}
218: where $\beta_\pm$ represents field phase shifts, different, in general,
219: in the top and the bottom chevron slabs, $b$'s and $a$'s are spatially and
220: temporary independent coefficients. Both $b_0^\pm$ and $b_1^\pm$ do not
221: depend also on $A$, while $a_l^\pm\sim A$, $l=0,1,2,...\,$. It is easy to verify
222: that Eq. (1) has no solution of the form (10) for $b_1^\pm=0$ and for any
223: nonzero $b_1^\pm$ such that $b_1^\pm\rightarrow 0$ as $A\rightarrow 0$.
224: Inserting (10) into (1), equating components with the same time functions and
225: with the spatial variable increased to the same powers, one obtains an
226: infinite sequence of relations
227: \parbox{7.5cm}{\begin{eqnarray*}
228: (l+2)!\,a_{l+2}^\pm\cos\beta_\pm-l!\,\gamma\omega a_l^\pm\sin\beta_\pm&=&A\rho_l^\pm
229: (b_1^\pm)^{l-2}\,\,\,,\\
230: (l+2)!\,a_{l+2}^\pm\sin\beta_\pm+l!\,\gamma\omega a_l^\pm\cos\beta_\pm&=&0\,\,\,,
231: \end{eqnarray*}} \hfill
232: \parbox{1cm}{\begin{eqnarray}\end{eqnarray}}\\
233: where $l=0,1,2,...,$ and
234: \begin{equation}
235: \rho_l^\pm=
236: \left\{
237: \begin{array}{cr}
238: (-1)^{(l+2)/2}\sin(\mp\phi_0+b_0^\pm)\,\,\,,\mbox{if}\,\,\,
239: l=0,2,4,...\,\,\,,\nonumber\\
240: \\
241: (-1)^{(l+1)/2}\cos(\mp\phi_0+b_0^\pm)\,\,\,,\mbox{if}\,\,\,
242: l=1,3,5,...\,\,\,.\nonumber
243: \end{array}
244: \right.
245: \end{equation}
246: Solving Eqs. (12) with respect to the coefficients $a$ yields
247: \begin{equation}
248: a_l^\pm=\frac{1}{l!}A\rho_l^\pm\cos\beta_\pm(b_1^\pm)^{l-2}\,\,\,,\,\,\,
249: l=0,1,2,...,
250: \end{equation}
251: with
252: \begin{equation}
253: (b_1^\pm)^2=\frac{\gamma\omega}{\tan\beta_\pm}\,\,\,.
254: \end{equation}
255: Thus, according to (13) and (14), the functions $F_\pm$ can be written in a
256: concise form as
257: \begin{equation}
258: F_\pm(x)=-A(b_1^\pm)^{-2}\cos\beta_\pm\sin(\mp\phi_0+b_0^\pm+b_1^\pm x)\,\,\,.
259: \end{equation}
260: The simple form of $F_\pm(x)$ is a result of the occurrence of field-independent
261: components (being a solution of the Laplace equation
262: $\frac{\partial^2}{\partial x^2}\phi=0$) in $\phi_\pm(x,t)$. The existence of
263: these linear components in (10) implies that the long time averages
264: $\bar\phi_\pm(x)=\lim_{T\rightarrow\infty}T^{-1}\int_{0}^{T}\phi_\pm(x,t)dt$
265: exhibit nonuniform distributions, given by
266: $\bar\phi_\pm(x)=\mp\phi_0+b_0^\pm+b_1^\pm x$. Such a dynamic splay is a
267: consequence of heterogeneous, throughout chevron cells, reaction of viscosity
268: interactions to the action of alternating electric field (note that
269: $b_1^\pm\rightarrow 0$, and thereby $b_0^\pm\rightarrow 0$, as
270: $\gamma\rightarrow 0$). Hence, average values of $\phi_\pm$ at boundary and
271: interface surfaces depend not only on boundary and interface anchoring
272: couplings but also on viscosity and elastic interactions (via $\gamma$).
273:
274: Since the dynamic splay is generated by the applied alternating voltage, one
275: can assume that, for small voltage amplitudes, this splay is not significant
276: and hence high-order components (including coefficients $a_l^\pm$ with large
277: powers of $b_1^\pm$) in expansions (11) can be neglected. Moreover, the
278: functions $F_\pm$ should approach their static counterparts $f_\pm$ as the
279: field frequency decreases. Consequently, $F_\pm$ can be approximated as
280: \begin{equation}
281: F_\pm(x)\simeq a_0^\pm+a_1^\pm x+a_2^\pm x^2\,\,\,.
282: \end{equation}
283: By means of relation(3), the boundary conditions for solution (10) are
284: given now by\cite{m1,m2}
285: \begin{equation}
286: \frac{\partial}{\partial x}\phi_{\pm}(x=\pm\frac{d}{2},t)\simeq\mp\tilde{u}_\pm\mp
287: u_\pm F_\pm(\pm\frac{d}{2})\cos(\omega t-\beta_\pm)\,\,\,,
288: \end{equation}
289: where
290: \begin{eqnarray}
291: \tilde{u}_\pm&=&\lambda_1\sin(2b_0^\pm\pm b_1^\pm d)\mp\lambda_2\sin(b_0^\pm\pm
292: b_1^\pm\frac{d}{2})\,\,\,,\\
293: u_\pm&=&2\lambda_1\cos(2b_0^\pm\pm b_1^\pm d)\mp\lambda_2\cos(b_0^\pm\pm
294: b_1^\pm\frac{d}{2})\,\,\,.
295: \end{eqnarray}
296: Then, imposing these boundary conditions on $\phi_{\pm}(x,t)$ and by analogy
297: with the condition (6), using the constraint that $F_\pm(\pm d/2)=F_\pm(0)$
298: gives
299: \begin{eqnarray}
300: &&b_1^\pm=\mp\tilde{u}_\pm\,\,\,,\\
301: &&(b_1^\pm)^2=\frac{4u_\pm}{d}\,\,\,,\\
302: &&\tan(\mp\phi_0+b_0^\pm)=\pm\frac{4}{b_1^\pm d}\,\,\,.
303: \end{eqnarray}
304: Relations (22), or alternatively (21), and (23) enable one to express the
305: constants $b_0^\pm$ and $b_1^\pm$ by the material parameters $\lambda_1$,
306: $\lambda_2$, and $\phi_0$. Using, in turn, Eqs. (14), (22), and (23), one
307: finds
308: \begin{equation}
309: F_\pm(x)\simeq-\frac{A}{2}\sin(\mp\phi_0+b_0^\pm)\cos\beta_\pm
310: (\frac{d}{2u_\pm}\pm\frac{d}{2}x-x^2)\,\,\,.
311: \end{equation}
312: The distribution of the azimuthal angle given by the above relation is
313: similar as in the static case [Eq. (9)], but involves modified, due to the
314: action of the alternating electric field, {\it effective} combined surface
315: interactions $u_\pm$, instead of $v_\pm$.
316:
317: In the case of strong surface anchoring, one can assume that, at boundary
318: surfaces, the effect of the dynamic splay is roughly insignificant. Then,
319: $\bar{\phi}_\pm(\pm d/2)\simeq\mp\phi_0$, and hence,
320: \begin{equation}
321: b_0^\pm\simeq\mp b_1^\pm\frac{d}{2}\,\,\,.
322: \end{equation}
323: Thus, the combined surface interactions $u_\pm$ are close to $v_\pm$,
324: respectively, for systems in weak applied fields and with strong anchoring of
325: molecules at surfaces. The relation (25), together with Eqs. (22) and (23),
326: leads to an approximate relationship between $\lambda_1$, $\lambda_2$, and
327: $\phi_0$. Unfortunately, this relationship is not uniquely determined, since
328: Eqs. (23) and (25) have many solutions for a given $\phi_0$. However, instead
329: of assuming the constraint $F_\pm(\pm d/2)=F_\pm(0)$, which refers to the
330: static case rather than to the dynamic case, one can use the condition that
331: the long time average of the interface energy takes its minimum value. This
332: condition concerns the dependence of the average interface energy on the
333: average azimuthal angle $\bar{\phi}_\pm(0)$. According to (25), the average
334: azimuthal angle takes at chevron interface the values
335: \begin{equation}
336: \bar{\phi}_\pm(0)\simeq\pm\phi_0\mp b_1^\pm\frac{d}{2}\,\,\,.
337: \end{equation}
338: Since, in general, $b_1^+\ne b_1^-$, the abrupt change of $\phi$ occurring for
339: $E=0$ at the interface is on average modified when the dynamic splay appears
340: (in the presence of applied alternating field). Such a modification of the
341: discontinuity of the average azimuthal angle at the interface is connected
342: with an increase of the average interface energy. For chevron structures with
343: small layer and molecular tilt angles (i.e., for typical chevron cells), the
344: growth of the density of this energy, $W_I$, can be determined by
345: \begin{equation}
346: \varepsilon\varepsilon_0W_I=P_s^2-P_x^+P_x^-+P_y^+P_y^-\,\,\,,
347: \end{equation}
348: where $\varepsilon$ is the dielectric permittivity, whereas $P_x^\pm$ and
349: $P_y^\pm$ denote $x$ and $y$ components of the spontaneous polarization of
350: molecules immediately above and below the interface plane, respectively.
351: Assuming that $|b_0^\pm|$ are small compared with $|\phi_0|$, allows $W_I$
352: to approximate as follows
353: \begin{equation}
354: \varepsilon\varepsilon_0W_I\simeq\frac{1}{2}(b_0^++b_0^-)^2-\frac{1}{4}
355: (b_0^+b_0^-)^2\,\,\,,
356: \end{equation}
357: By minimizing $W_I$ with respect to $b_0^-$ (for fixed $b_0^+$) one gets
358: \begin{equation}
359: b_0^-\simeq-\frac{b_0^+}{1-\frac{1}{2}(b_0^+)^2}\,\,\,.
360: \end{equation}
361: Consequently, by virtue of (22) and (25), the above constraint yields
362: \begin{equation}
363: u_-\simeq\frac{u_+}{1-du_+}\,\,\,.
364: \end{equation}
365: This approximate relation is valid for systems with strong, asymmetric
366: anchoring of molecules at bounding plates, i.e., for large
367: $\gamma_1$ and $\gamma_2$ (in comparison to $K$), such that
368: $2\gamma_1\ne\gamma_2$. As it is shown below, it is very useful for
369: determining the dielectric response of chevron samples.
370: \subsection{Dielectric response}
371: To determine the dynamic dielectric susceptibility and thereby the dynamic
372: dielectric response of chevron SSFLC, the solution (10) to Eq. (1) is applied.
373: Thus, in the case of a sinusoidal external electric field $E(t)$ acting along
374: the $X$ axis, the azimuthal mode contribution to the dynamic dielectric
375: susceptibility can be written as
376: \begin{eqnarray}
377: \chi(\omega)=
378: &&\lim_{E\rightarrow 0}\frac{P_s}{E(t)}{\big [}{\big <}
379: \cos\phi_+(x,t){\big >}_+-{\big <}\cos\bar{\phi}_+(x){\big >}_+\nonumber\\
380: &&+{\big <}\cos\phi_-(x,t){\big >}_--{\big <}\cos\bar{\phi}_-(x){\big >}_-
381: {\big ]}\,\,\,,
382: \end{eqnarray}
383: where $<\!\!...\!\!>_\pm$ denote space averages over respectively upper and lower
384: chevron arms, so that, e.g.,
385: \begin{equation}
386: {\big <}\cos\phi_\pm(x,t){\big >}_\pm=\pm\frac{2}{d}\int^{\pm d/2}_{0}\!\!\!
387: \cos\phi_\pm(x,t)dx\,\,\,.
388: \end{equation}
389: Using Eqs. (10), (22)-(25), and (31), and assuming that $|b_0^\pm|$ are small,
390: one derives
391: \begin{equation}
392: \frac{\chi(\omega)}{\varepsilon_0}=\varepsilon'(\omega)+
393: \varepsilon''(\omega)\tan(\omega t)\,\,\,,
394: \end{equation}
395: with the dielectric permittivity
396: \begin{equation}
397: \varepsilon'(\omega)=B(\cos^2\beta_++\cos^2\beta_-)
398: \end{equation}
399: and the dielectric loss permittivity
400: \begin{equation}
401: \varepsilon''(\omega)=B(\tan\beta_+\cos^2\beta_++\tan\beta_-\cos^2\beta_-)\,\,\,,
402: \end{equation}
403: where
404: \begin{equation}
405: B=\frac{Ad}{4u_+}\sin^2\phi_0\,\,\,.
406: \end{equation}
407: By means of Eqs. (15) and (22), the phase shifts $\beta_\pm=\omega\tau_\pm$
408: with the relaxation times
409: \begin{equation}
410: \tau_\pm=\frac{\gamma d}{4u_\pm}\,\,\,.
411: \end{equation}
412: Hence, the dielectric response of chevron cells stabilized by both nonpolar
413: and polar surface interactions can be described in the case of weak applied
414: electric field by two Debye processes. These processes are determined by
415: different relaxation times, each characterizing collective motions of
416: molecules on opposite sides of the chevron interface. Making use of
417: relation (30), valid for systems that display strong anchoring of molecules
418: at bounding surfaces, one can describe the dielectric response of chevron
419: samples by three parameters: $B$, $\tau_+$ (or equivalently $\tau_-$), and
420: $d$. Moreover, having determined the relaxation times, one obtains the
421: parameters $\gamma$, $u_+$, and $u_-$. A simple method to calculate the
422: combined surface interactions $u_+$ and $u_-$, based on experimental data,
423: is described below.
424: \subsection{Calculation of surface energy parameters}
425: The permittivity spectra (34) and (35) involve two Debye relaxation processes
426: with equal strengths. In such a case, the loss permittivity possesses a single
427: maximum if the ratio $\nu=\tau_+/\tau_-$ is $\nu\ge\nu_c$, where
428: $\nu_c=3-2\sqrt{2}\approx 0.172$, and two maxima of equal heights if
429: $\nu<\nu_c$. However, the case of the double maximum form of
430: $\varepsilon''(\omega)$ is not considered here because typical experimental loss
431: permittivity spectra obtained for chevron cells do not display equal, or nearly
432: equal, maxima that could be ascribed to collective azimuthal excitations of
433: molecules.
434:
435: In order to compare theoretical [Eqs. (34) and (35)] and experimental spectra
436: $\varepsilon(\omega)$, one has to determine combined surface interactions
437: $u_\pm$. By applying the approximate relation (30), one can describe
438: $\varepsilon(\omega)$ by one energy parameter, say $u_+$. Then, the position
439: of the maximum of $\varepsilon''(\omega)$ is given by
440: \begin{equation}
441: \omega_{\rm max}\tau_+=1/\sqrt{\nu}\,\,\,,
442: \end{equation}
443: where $\nu=1-du_+$. For a value of $\omega_{\rm max}$ found experimentally,
444: this relation enables one to express the relaxation time $\tau_+$ by $u_+$
445: and the cell thickness $d$. Next, introducing the rates
446: $\eta(\omega)=\varepsilon'(\omega)/\varepsilon''(\omega)$ and
447: $\sigma(\omega)=\omega/\omega_{\rm max}$, and applying (34) and (35), leads to
448: the following biquadratic equation with respect to $\nu$:
449: \begin{equation}
450: \sigma^2p^2-\eta\sigma(1+\sigma^2)p+2(1-\sigma^2)=0
451: \end{equation}
452: with
453: \begin{equation}
454: p=\nu+\frac{1}{\nu}\,\,\,.
455: \end{equation}
456: As a result, solving Eq. (39) for different values of $\omega$ and for
457: respective, experimentally obtained values of $\eta$, yields $u_+$ as a
458: function of $\omega$. Since $u_+$ is essentially independent of $\omega$,
459: this function should be constant or should at least exhibit a plateau in some
460: range of frequency. Thus, the considered procedure enables one not only to
461: determine the surface energy parameter $u_+$, but also to test the above
462: theoretical description of the dielectric response of chevron systems. In
463: particular, one can determine the frequency region where this description is
464: valid.
465: \section{Experimental}
466: To verify the presented theory, dielectric measurements on the
467: liquid-crystalline mixture Felix 15-100, commercially available from
468: Clariant have been carried out. The investigated mixture exhibits
469: ferroelectric smectic ${\rm C^*}$ phase at room temperature and transforms
470: into the paraelectric smectic A phase at $73$ $^{\rm o}{\rm C}$. The measurements
471: presented in this paper were performed at temperature $50$ $^{\rm o}{\rm C}$,
472: i.e., $23$ $^{\rm o}{\rm C}$ below the ferroelectric-paraelectric phase
473: transition. The investigated material was introduced into commercially
474: available measuring cells of various thicknesses ($5$ $\mu{\rm m}$, from Linkam,
475: UK, and $12.1$ $\mu{\rm m}$, $25.9$ $\mu{\rm m}$, $50$ $\mu{\rm m}$, from EHC,
476: Japan). The thickness of all cells used was less than or comparable to the
477: helical pitch of the investigated material ($\sim 20$ $\mu{\rm m}$), thus the
478: chevron texture was easily attained. The walls of measuring cells were
479: provided with indium-tin-oxide (ITO) semitransparent electrodes coated with
480: thin polymer layers. These layers were rubbed unidirectionally along
481: antiparallel directions on both plates. This promoted the orientation of
482: molecules more or less parallel to electrodes and reduced the ionic current
483: flowing across the sample. The measuring cell was placed inside a modified
484: Metler hot stage, in which the temperature was controlled using a Digi-Sense
485: temperature controller. The hot stage was installed between crossed polarizers
486: of a polarizing microscope (Biolar, from PZO, Poland). The angle between the
487: polarization plane of the incident light and the optic axis of the sample in
488: the smectic A phase was set to $22.5^{\rm ^o}$. The intensity of light passing
489: the sample and polarizers was registered using a photodiode connected to a
490: preamplifier and lock-in amplifier SR 850 from Stanford Research (USA).
491: Simultaneously, the electric capacitance of the sample was measured using the
492: Hewlett-Packard low-frequency impedance analyzer HP 4192A. The measurements
493: indicated that the dielectric increment (the difference between the electric
494: permittivity in the smectic ${\rm C^*}$ phase and in the smectic A phase) was
495: exactly proportional to the light modulation depth, detected by the photodiode,
496: as it was already demonstrated in Ref.\cite{kc}. Thus, the dielectric
497: measurements might be well replaced by electro-optic measurements, which are
498: more accurate. Moreover, they are less sensitive to ionic current, especially
499: at low frequencies\cite{kc}. This property enabled one to extend the measurement
500: range down to $\sim 10$ Hz. In this way, reliable results were obtained at much
501: lower frequencies than it might be possible using typical dielectric methods.
502: Therefore, in what follows, electro-optic experimental results obtained for
503: frequencies greater than 10 Hz are only discussed.
504: \section{Results and discussion}
505: The approach developed here to analyze the dielectric response of SSFLC with
506: chevron geometry requires determining the effective surface interaction $u_+$.
507: Using electro-optic data obtained for samples of different thicknesses and
508: employing the procedure of Sec. II C, the parameter $u_+$ has been derived as a
509: function of frequency in cases of various values of the voltage amplitude $U$.
510: Plots of $u_+$ are shown in Fig. \ref{f2} for a low voltage, $U=0.1$ ${\rm V}$.
511: It is seen that, for samples of different thicknesses, $u_+$ remains nearly
512: constant within rather wide frequency ranges, and quickly decreases outside these
513: ranges. Clearly, the rapid decay of $u_+$ in the low-frequency region is a
514: consequence of an overlapping, in the dielectric response, between
515: contributions arising from collective azimuthal excitations of molecules in
516: chevron slabs and contributions that have other origins than these collective
517: excitations. Similarly, a less sudden downfall of $u_+$ in high-frequency
518: regions can be ascribed to components of $\varepsilon(\omega)$ corresponding
519: to high-frequency processes\cite{go}, separated to a large extent from the
520: collective processes occurring at intermediate frequencies. Naturally, the
521: region of frequencies for which the parameter $u_+$ remains approximately
522: constant determines the range of the validity of the introduced theoretical
523: approach.
524:
525: The surface energy parameter $u_+$ has been found for cells of different
526: thicknesses by averaging parameter values calculated for frequencies
527: belonging to ranges in which these values are nearly constant. Results of
528: calculations of $u_+$ for $U=0.1$ ${\rm V}$ are given in Table I, where
529: corresponding values of the ratio $\nu$ and the frequency
530: $f_{\rm max}=\omega_{\rm max}/2{\rm \pi}$ are also given. Values of $u_+$
531: determined for various voltages demonstrate that $u_+$ scarcely depends on the
532: voltage amplitude if $U$ remains relatively small, below some value
533: $U_{\rm lin}$, different, in general, for individual samples
534: ($U_{\rm lin}\approx 3$ ${\rm V}$ for studied samples of the thickness
535: $d=5$ $\mu{\rm m}$). When the voltage amplitude exceeds this value, nonlinear
536: effects in $\varepsilon(\omega)$ become significant and Eq. (39) loses real,
537: positive solutions for medium frequencies. As is seen from Table I, values of
538: $\nu$, obtained in the low-voltage regime, are much greater than the threshold
539: value, above which $\varepsilon''(\omega)$ has a single maximum. This means
540: that, for the investigated SSFLC samples, differences between anchoring
541: energies on both boundary plates do not strongly affect the form of the
542: function $\varepsilon''(\omega)$. The dielectric response of studied chevron
543: cells displays a dependence on the sample thickness. In particular,
544: $f_{\rm max}$ decreases and thereby both the relaxation times $\tau_\pm$
545: increase, as $d$ grows. The increase of $\tau_\pm$ with $d$ is a consequence
546: of a stabilization of collective reorientations of molecules by strong surface
547: interactions. Since these interactions force the molecular ordering in entire
548: chevron cells, time needed to reach the ordered state in initially perturbed
549: samples is longer and longer as $d$ grows and/or as the surface interactions
550: become weaker. This is consistent with relation (37). Clearly, the
551: dependence of the effective surface interactions $u_\pm$ on the sample
552: thickness (as seen in Fig. 2 in the case of $u_+$) originates, in general,
553: from the dependence of surface energy strengths $\gamma_1$ and/or $\gamma_2$
554: on $d$. However, to explain such a behavior of $\gamma_1$ or $\gamma_2$ as $d$
555: is varied, one would analyze dynamic processes leading to the appearance of
556: surface interactions.
557:
558: To test the method introduced here for investigating dielectric response of
559: chevron cells, dielectric spectra have been determined using Eqs. (30), (34),
560: (35), (37), and (38) with values $\omega_{\rm max}$ and
561: $\varepsilon''(\omega_{\rm max})$ found experimentally. Additionally, a simple
562: relaxation model that involves a single Debye process, characterized by the
563: same values of $\omega_{\rm max}$ and $\varepsilon''(\omega_{\rm max})$, has
564: been considered. Both resulting theoretical dielectric loss spectra as well as
565: experimental spectra obtained for the voltage amplitude $U=0.1$ ${\rm V}$ are
566: plotted in Fig. \ref{f3} for intermediate frequencies. These plots show that
567: the experimentally determined dielectric loss spectra exhibit a broadening
568: near their maxima, compared to corresponding spectra obtained for single Debye
569: relaxation processes, and that such a broadening is well described by the
570: introduced model including two Debye processes. Consequently, the effect of
571: smearing of $\varepsilon''(\omega)$ near its maximum, located in the
572: intermediate frequency region, can be considered as a result of a difference
573: between anchoring energies at boundary surfaces. However, when $\omega$
574: decreases starting from $\omega_{\rm max}$, there appears a discrepancy
575: between measured and theoretically predicted dielectric loss permittivities,
576: as illustrated in Fig. \ref{f4}. This inconsistency is due to the occurrence
577: in chevron samples of molecular motions other than the collective molecular
578: reorientations taken into account in the theoretical model. Such additional
579: motions are called here low-frequency processes, although they markedly affect
580: the spectra $\varepsilon(\omega)$ even for $\omega$ relatively close
581: to $\omega_{\rm max}$. Thus, frequency regions, in which low- and
582: medium-frequency processes are significant, are not quite separated, and a
583: rigid determination of the corresponding low- and intermediate-frequency
584: ranges is not possible.
585:
586: To interpret the complicated low-frequency behavior of the dielectric loss
587: permittivity, it has been suggested that various dynamic mechanisms can be of
588: importance in low-frequency regime\cite{p2,hv}. However, no detailed analysis
589: of the complex dielectric response of chevron cells at low frequencies has
590: been reported yet. For low voltages (for which the dielectric response is
591: linear) contributions to $\varepsilon(\omega)$ that arise from dynamic
592: processes other than fluctuations of the azimuthal angle with chevron slabs
593: can easily be investigated by a subtraction of theoretical functions
594: $\varepsilon_t(\omega)$ involving two relaxation times from respective
595: experimental spectra $\varepsilon_e(\omega)$. Resulting extracted loss spectra
596: $\Delta\varepsilon''=\varepsilon_e''-\varepsilon_t''$ are shown in
597: Fig. \ref{f5} for samples of different thicknesses, in the case of
598: $U=0.1$ ${\rm V}$. It proves that, for each system under study here,
599: $\Delta\varepsilon''$ has a single maximum at a frequency $\omega_0$ of the
600: same order of magnitude. Evidently, $\Delta\varepsilon''$ displays for
601: $\omega>\omega_0$ a complex form, different from the shape of loss spectra
602: corresponding to single Debye processes or even to sums of a few single
603: processes. This suggests that the dielectric response of chevron cells can
604: approximately be described by a spectrum of Debye processes with a continuous
605: distribution of relaxation times. Then, for low frequencies, such that
606: $\omega\ge\omega_0$, the extracted loss permittivity can be expressed as
607: \begin{equation}
608: \Delta\varepsilon''(\omega)=\frac{1}{\omega}
609: \int^{\tau_2}_{\tau_1}\!\!\rho(\tau)\frac{\omega\tau}{1+\omega^2\tau^2}d\tau\,\,\,,
610: \end{equation}
611: where $\tau_2\le 1/\omega_0$ and $\tau_1$ are respectively maximal and minimal
612: relaxation times, and $\rho(\tau)$ denotes the distribution of relaxation times.
613: For $\omega\ge\omega_0$, the shape of the extracted spectra determined with the
614: use of experimental data can easily be recovered by assuming that
615: \begin{equation}
616: \rho(\tau)\sim\tau^{-2}\,\,\,.
617: \end{equation}
618: Thus, applying (41) and (42) yields
619: \begin{equation}
620: \Delta\varepsilon''(\omega)=C_1\ln{\Big[}\frac{\tau_2^2(1+\omega^2\tau_1^2)}
621: {\tau_1^2(1+\omega^2\tau_2^2)}{\Big]}
622: \end{equation}
623: with $C_1$ being a positive constant. Similarly, using (42), one finds the
624: extracted dielectric permittivity $\Delta\varepsilon'$ to be
625: \begin{eqnarray}
626: \Delta\varepsilon'(\omega)=C_2{\big[}&&\frac{1}{\omega}{\big(}\frac{1}{\tau_1}
627: -\frac{1}{\tau_2}{\big)}\nonumber\\
628: &&-\arctan(\omega\tau_2)+\arctan(\omega\tau_1){\big]}\,\,\,,
629: \end{eqnarray}
630: where $C_2$ denotes a constant. Results of fits of functions given by Eqs. (43)
631: and (44) (on adjusting the parameters $C_1$, $C_2$, $\tau_1$, and $\tau_2$) to
632: respective extracted spectra obtained with the use of experimental data are
633: shown in Figs. \ref{f5} and \ref{f6}. It is seen that the agreement between
634: postulated and experimental extracted spectra $\Delta\varepsilon''$ is good
635: (as long as $\omega\ge\omega_0$), but is less satisfactory in the case of the
636: spectra $\Delta\varepsilon'$. Nevertheless, these figures demonstrate that the
637: low-frequency dielectric response of chevron cells can indeed be characterized
638: by Debye processes with continuously distributed relaxation times. It should
639: also be noted that the measured spectra presented in Fig. \ref{f5}, contrary
640: to earlier results derived from a numerical decomposition of the dielectric
641: loss spectra\cite{ka}, do not reveal the existence of any contributions that
642: would correspond to surface fluctuations of the $c$ director (at frequencies
643: of an order magnitude greater than $\omega_{\rm max}$).
644:
645: Turning to observations of chevron samples through polarizing optical
646: microscope, it is argued here that a complex dielectric response of SSFLC is
647: connected to dynamics of zigzag defects\cite{la}. Usually, these defects
648: appear spontaneously, forming irregular patterns, when chevron cells are
649: cooled down\cite{la,h2,li,fu,wa}. Microscopic observations of chevron samples
650: performed in the low-frequency regime at weak voltages have displayed periodic
651: changes (according to a voltage alternation) of the color and the intensity of
652: light transmitted not only by uniform chevron regions (domains) but also by
653: thin and thick zigzag walls separating these regions, as illustrated in the
654: microphographs of Fig. \ref{f7} for a sample of the thickness
655: $d=5$ $\mu{\rm m}$. Thus, in addition to uniform domains, collective relaxation
656: processes caused by a weak voltage alternating with low frequencies occur also
657: inside zigzag walls of different widths. Since zigzag walls reveal various
658: shapes as well as sizes\cite{la,c2,li} and consist of smectic A layers having
659: different thicknesses, Consequently, being affected by surface
660: interactions with different strengths, one can infer that relaxation times
661: characterizing dynamic processes within these walls undergo a nonuniform
662: distribution. Then, the complexity of the dielectric response of chevron cells
663: to weak voltages alternating with low frequencies may, in fact, be considered as
664: a result of relaxation processes within zigzag walls.
665:
666: As the amplitude of the applied voltage exceeds a threshold $U_1$ (different,
667: in general, for various samples), thick zigzag walls start to creep. Such a
668: drift viscous motion of defect walls, induced by an alternating electric
669: field, is spatially confined by pinning ends of thick walls to ends of thin
670: walls. This is illustrated in Fig. \ref{f8} for the same sample of the width
671: $d=5$ $\mu{\rm m}$, in the case of the voltage amplitude $U=1.0$ ${\rm V}$ and
672: the frequency $f=100$ ${\rm Hz}$. The drift motion of thick walls
673: [Figs. \ref{f8}(a) and \ref{f8}(b)] is not periodic, although long time
674: averages of displacements $r(t)$ of these walls vanish, i.e.,
675: $\lim_{t\rightarrow\infty}r(t)=0$, as shown in Fig. \ref{f8}(c).
676:
677: When the voltage amplitude goes beyond an another threshold $U_2$ ($U_2>U_1$),
678: both thin and thick walls begin to slide. Such irreversible motions of zigzag
679: walls are accompanied with a gradual shrinkage of corresponding zigzag lines
680: on boundary surfaces and with gradual disappearing of the defect walls, as $U$
681: grows, remaining below the switching threshold $U_s$. Thus, the density of
682: zigzag defects decreases when $U$ grows (above the sliding threshold $U_2$).
683: The effect of motions of zigzag walls in the sliding regime is presented in
684: Fig. \ref{f9} for the studied sample of the thickness $d=5$ $\mu{\rm m}$, in
685: the case of the voltage amplitude $U=5$ ${\rm V}$ and the frequency
686: $f=100$ ${\rm Hz}$. Note that the threshold voltage at which the switching
687: process appears, found for samples of different thicknesses, increases with
688: $d$ (e.g., $U_s=10$ ${\rm V}$ for the case of $d=5$ $\mu{\rm m}$, while
689: $U_s=20$ ${\rm V}$ for the case of $d=25.9$ $\mu{\rm m}$). Analysis of the
690: dynamics of defects presented here indicates that both the threshold voltages
691: $U_1$ and $U_2$ (similarly to $U_s$) are different for samples of different
692: thicknesses. However, to answer the question of whether the dependence of these
693: threshold voltages on the sample thickness has a general character, and to
694: possibly determine such a general dependence, further work on a wider class
695: of uniform chevron cells is needed.
696:
697: Changes in the dynamic behavior of defect walls as $U$ crosses its threshold
698: values are markedly reflected in the low-frequency dielectric loss spectra,
699: as illustrated in Fig. \ref{f10}. It is seen that, in comparison to
700: relaxation processes inside zigzag walls, creep motions of thick walls give a
701: large contribution to $\varepsilon''(\omega)$, in almost the entire
702: low-frequency range, with a distinct maximum. Contrary to creep loss spectra,
703: sliding spectra (associated with sliding of defect walls) have very small
704: amplitudes for $f>10$ ${\rm Hz}$. This is a consequence of the fact that
705: sliding motions are very slow compared to creep motions. Hence, the
706: dielectric loss is dominated in the sliding regime by the low frequency tail
707: of contributions to $\varepsilon''(\omega)$ arising from collective
708: fluctuations of the azimuthal orientation of the $c$ director. It should be
709: noted that similar relaxation, creep, and slide motions of complex objects in
710: elastic media have recently been studied, both theoretically and experimentally,
711: in various contexts. In particular, thermally activated and/or field-induced
712: domain wall motions of these types have been investigated in impure
713: magnets\cite{nt}, polydomain relaxor-ferroelectric single crystals\cite{kl},
714: discontinuous metal-insulator multilayer systems\cite{ch}, and periodically
715: poled ferroelectric single crystals\cite{ba}. Transitions between different
716: dynamic states of domain walls occurring in these systems have been shown to
717: reveal a character of dynamic phase transitions, at critical field amplitudes
718: being some functions of the temperature $T$ and the frequency of the field
719: alternation. Clearly, one can expect that changes in the dynamics of zigzag
720: walls appearing in chevron SSFLC also display a character of dynamic phase
721: transitions, determined in the parameter space ($U-T-\omega$). A detailed
722: exploration of this question is, however, beyond the scope of this paper.
723: \section{Conclusions}
724: It has been shown in this paper that the dielectric response of ferroelectric
725: chevron systems stabilized by nonpolar and polar surface interactions is
726: determined at intermediate frequencies by two Debye relaxation processes,
727: corresponding to collective fluctuations of the azimuthal orientation of
728: molecules in two chevron slabs. This relaxation mechanism has been proved to
729: cause a broadening of the dielectric loss spectrum near its maximum, compared
730: to a spectrum obtained for a model involving a single Debye process. Such a
731: broadening effect has been confirmed experimentally, for chevron cells of
732: different thicknesses. To investigate low-frequency processes and possible
733: intermediate-frequency processes, other than cooperative fluctuations of
734: azimuthal orientations of the $c$ director, theoretically derived dielectric
735: response functions have been extracted from respective measured electro-optic
736: spectra. Resulting extracted spectra have not displayed any additional
737: processes at medium frequencies, even for relatively thick samples. This is
738: in contrast to an earlier suggestion\cite{ka} that, in a medium-frequency
739: region, the dielectric response of thick enough cells can be affected by
740: surfacelike fluctuations of the $c$ director. It has been demonstrated that
741: low-frequency dynamic processes have different character in three voltage
742: regimes and can be ascribed to relaxation processes inside thin and thick
743: zigzag walls, creeping of thick walls, and sliding as well as disappearing
744: of defect walls. Certainly, low-frequency dielectric response of chevron
745: cells can be affected by other kinds of defects or smectic layer
746: deformations\cite{la,fu,is}. However, the low-frequency dynamic mechanisms
747: discussed in this paper appear to be most important. Finally, it should be
748: remarked that the possibility of macroscopic observations of dynamic
749: behaviors of defect walls provides a useful tool of studying low-frequency
750: dynamic states of chevron SSFLC and dynamic phase transitions between these
751: states.
752: \acknowledgments
753: The authors are very grateful to J. Ma{\l}ecki for useful discussions.
754: This work was supported by Polish Research Committee (KBN) under grant
755: No. 2PO3B 127 22.
756:
757: \begin{references}
758: \bibitem{la} S.T. Lagerwall, {\it Ferroelectric and Antiferroelectric Liquid
759: Crystals} (Wiley-VCH, Weinheim, 1999).
760: \bibitem{c1} N.A. Clark and S.T. Lagerwall, Appl. Phys. Lett. {\bf 36}, 899
761: (1980).
762: \bibitem{h1} M.A. Handschy, N.A. Clark, and S.T. Lagerwall, Phys. Rev. Lett.
763: {\bf 51}, 471 (1983).
764: \bibitem{h2} M.A. Handschy and N.A. Clark, Ferroelectrics {\bf 59}, 69 (1984).
765: \bibitem{ri} T.P. Rieker, N.A. Clark, G.S. Smith, D.S. Parmar, E.B. Sirota,
766: and C.R. Safinya, Phys. Rev. Lett. {\bf 59}, 2658 (1987).
767: \bibitem{xu} J. Xue, N.A. Clark, and M.R. Meadows, Appl. Phys. Lett {\bf 53},
768: 2397 (1988).
769: \bibitem{co} D. Coleman, D. Mueller, N.A. Clark, J.E. Maclennan, R.-F. Shao,
770: S. Bardon, and D.M. Walba, Phys. Rev. Lett. {\bf 91}, 175505 (2003).
771: \bibitem{m1} J.E. Maclennan, N.A. Clark, M.A. Handschy, and M.R. Meadows,
772: Liq. Cryst. {\bf 7}, 753 (1990).
773: \bibitem{m2} J.E. Maclennan, M.A. Handschy, and N.A. Clark, Liq. Cryst.
774: {\bf 7}, 787 (1990).
775: \bibitem{br} C.R. Brown, P.E. Dunn, and J.C. Jones, Eur. J. Appl, Math.
776: {\bf 8}, 281 (1997).
777: \bibitem{ha} L.D. Hazelwood and T.J. Sluckin, Liq. Cryst. {\bf 31}, 683 (2004).
778: \bibitem{p1} Y.P. Panarin, Yu.P. Kalmykov, S.T. MacLughadha, H. Xu,
779: and J.K. Vij, Phys. Rev. E {\bf 50}, 4763 (1994).
780: \bibitem{ka} O.E. Kalinovskaya, J.K. Vij, Y.V. Tretyakov, and Y.P. Panarin,
781: Liq. Cryst. {\bf 26}, 717 (1999).
782: \bibitem{pi} P. Piera\'nski, E. Guyon, P. Keller, L. Liebert, W. Kuczy\'nski,
783: and P. Piera\'nski, Mol. Cryst. Liq. Cryst. {\bf 38}, 275 (1977).
784: \bibitem{ku} W. Kuczy\'nski, J. Hoffmann, and J. Ma{\l}ecki, Ferroelectrics
785: {\bf 150}, 279 (1993).
786: \bibitem{p2} Y.P. Panarin, H. Xu, S.T. MacLughadha, and J.K. Vij,
787: Jpn. J. Appl. Phys. {\bf 33}, 2648 (1994).
788: \bibitem{c2} N.A. Clark and T.P. Rieker, Phys. Rev. A {\bf 37}, 1053 (1988).
789: \bibitem{li} L. Limat, Europhys. Lett. {\bf 44}, 205 (1998).
790: \bibitem{na} M. Nagakawa, M. Ishikawa, and T. Akahane, Jpn. Appl. Phys.
791: {\bf 27}, 456 (1988).
792: \bibitem{kc} W. Kuczy\'nski, in {\it Relaxation Phenomena, Liquid Crystals,
793: Magnetic Systems, Polymers, High $T_c$-Superconductors, Metallic Glasses},
794: edited by W. Haase and S. Wr\'obel (Springer, New York, 2003).
795: \bibitem{go} F. Gouda, K. Skarp, and S.T. Lagerwall, Ferroelectrics {\bf 113},
796: 165 (1991).
797: \bibitem{hv} S. Havrilak Jr., J.K. Vij, and M. Ni, Liq. Cryst. {\bf 26}, 465
798: (1999).
799: \bibitem{fu} A. Fukuda, Y.Ouchi, H. Arai, H. Takano, K. Ishikawa, and
800: H. Takezoe, Liq. Cryst. {\bf 5}, 1055 (1989).
801: \bibitem{wa} C. Wang, R. Kurihara, P.J. Bos, and S. Kobayashi, J. Appl. Phys.
802: {\bf 90}, 4452 (2001).
803: \bibitem{nt} T. Nattermann, V. Pokrovsky, and V.M. Vinokur, Phys. Rev. Lett.
804: {\bf 87}, 197005 (2001).
805: \bibitem{kl} W. Kleemann, J. Dec, S. Miga, Th. Woike, and R. Pankrath,
806: Phys. Rev. B {\bf 65}, 220101(R) (2002).
807: \bibitem{ch} X. Chen, O. Sichelschmidt, W. Kleemann, O. Petracic, Ch. Binek,
808: J.B. Sousa, S. Cardoso, and P.P. Freitas, Phys. Rev. Lett.
809: {\bf 89}, 137203 (2002).
810: \bibitem{ba} Th. Braun, W. Kleemann, J. Dec, and P.A. Thomas, Phys. Rev. Lett.
811: {\bf 94}, 117601 (2005).
812: \bibitem{is} N. Ul Islam, N.J. Mottram, and S.J. Elston, Liq. Cryst. {\bf 26},
813: 1059 (1999).
814:
815: %\bibitem{tag} Fake bibitem.
816: \end{references}
817:
818: % figures follow here
819: %
820: \begin{figure}
821: \caption{Geometry of orientation states in a uniform chevron cell of thickness
822: $d$: (a) in the smectic layer plane ($X$-$Y$), and (b) in the ($X$-$Z$) plane
823: (perpendicular to the smectic layer plane and to the bounding plates). In (a)
824: the symbols $\rightarrow$ and $\vdash$ represent, respectively, projections of
825: the microscopic polarization $\vec{P}_s$ and the $c$ director $\vec{c}$ on the
826: plane ($X$-$Y$). In (b) the symbols $\rightarrow$ and $>\!\!\!-$ denote
827: respectively projections of $\vec{P_s}$ and the molecular director $\vec{n}$
828: on the plane ($X$-$Z$). Scales of projections of vectors $\vec{c}$ and $\vec{n}$
829: on the planes ($X$-$Y$) and ($X$-$Z$), respectively, are not preserved. The
830: abbreviations $T$ and $B$ indicate top and bottom bounding surfaces,
831: respectively, whereas $I$ denotes the interface plane. The azimuthal angle
832: $\phi$ takes pretilt values $\phi=\pm\phi_0$ in upper and lower chevron slab,
833: respectively.}
834: \label{f1}
835: \end{figure}
836: %
837: \begin{figure}
838: \caption{The surface interaction parameter $u_+$ determined as function of
839: the frequency of the alternating voltage of the amplitude $U=0.1$ ${\rm V}$, for
840: samples of different thicknesses: $d=5$ ${\rm\mu m}$ $(\bigcirc)$,
841: $d=12.1$ ${\rm \mu m}$ $(\triangle)$, $d=25.9$ ${\rm \mu m}$ $(\Box)$.}
842: \label{f2}
843: \end{figure}
844: %
845: \begin{figure}
846: \caption{Loss spectra determined experimentally (dots) within a medium
847: frequency range, for $U=0.1$ $V$, in cases of chevron cells of thicknesses
848: $d=5$ ${\rm\mu m}$ and $d=12.1$ ${\rm\mu m}$, and derived theoretically by the
849: use of the approach including two Debye processes (labeled 1) as well as by the
850: use of the single relaxation model (labeled 2). Theoretical plots were obtained for
851: respective, measured values of
852: $\omega_{\rm max}$ and $\varepsilon''(\omega_{\rm max})$.}
853: \label{f3}
854: \end{figure}
855: %
856: \begin{figure}
857: \caption{Experimental (dots) and theoretical (solid lines) loss spectra for
858: chevron cells of different thicknesses. Theoretical plots were obtained by the
859: use of the approach involving two Debye processes.}
860: \label{f4}
861: \end{figure}
862: %
863: \begin{figure}
864: \caption{Dielectric loss spectra obtained by a subtraction of theoretical loss
865: spectra $\varepsilon_t$ from experimental spectra $\varepsilon_e$, in cases of
866: samples of different thicknesses, for $U=0.1$ ${\rm V}$. Dots correspond to
867: experimental data, while the solid line refers to the postulated function of
868: Eq. (43).}
869: \label{f5}
870: \end{figure}
871: %
872: \begin{figure}
873: \caption{Extracted spectra $\Delta\varepsilon'$ obtained for a sample of the
874: thickness $d=5$ $\mu{\rm m}$, in the case of $U=0.1$ ${\rm V}$. Dots refer to
875: experimental data and the solid line represents the function given by
876: Eq. (44).}
877: \label{f6}
878: \end{figure}
879: %
880: \begin{figure}
881: \caption{(Color online) Zigzag walls in a chevron cell of the thickness
882: $d=5$ $\mu{\rm m}$, under the influence of a voltage of the amplitude
883: $U=0.5$ ${\rm V}$, alternating with the frequency $f=10$ ${\rm Hz}$. The
884: microphotographs were taken at a short time period, both at the speed shutter
885: $1/100$ ${\rm s}$. These micrographs give an evidence of a pulsation of light
886: transmitted not only by domains but also by defect walls of various widths.}
887: \label{f7}
888: \end{figure}
889: %
890: \begin{figure}
891: \caption{(Color online) Creep motions of thick zigzag walls in a sample of
892: the thickness $d=5$ $\mu{\rm m}$, induced by an alternating voltage of the
893: amplitude $U=1$ ${\rm V}$ and the frequency $f=100$ ${\rm Hz}$. The
894: microphotographs (a) and (b) were taken at the time period $1$ ${\rm s}$, both
895: at a high shutter speed $1/1000$ ${\rm s}$. Thin walls, seen on left and
896: right sides of these micrographs, remain fixed confining the motion of the
897: thick wall connected to the thin walls. (c) is a micrograph of the same
898: objects as in micrographs (a) and (b) but taken at a slow shutter speed
899: $1$ ${\rm s}$. This figure shows a thick-wall drift, averaged over a relatively
900: long time period.}
901: \label{f8}
902: \end{figure}
903: %
904: \begin{figure}
905: \caption{(Color online) Sliding of zigzag walls in a chevron sample of the
906: thickness $d=5$ $\mu{\rm m}$, for the voltage amplitude $U=5$ ${\rm V}<U_s$ and
907: the frequency $f=100$ ${\rm Hz}$. The micrographs (a) and (b) were taken at the
908: shutter speed $1/2000$ ${\rm s}$, in a time $1$ ${\rm s}$ [(b) after (a)]. The
909: arrows indicate an irreversible motion and a shortening of respective defect
910: walls in this lapse of time.}
911: \label{f9}
912: \end{figure}
913: %
914: \begin{figure}
915: \caption{Experimental loss spectra obtained for a chevron system of the
916: thickness $d=5$ $\mu{\rm m}$, in cases: $U=0.3$ ${\rm V}<U_1$ ($\bigcirc$),
917: $U_1<U=1.0$ ${\rm V}<U_2$ ($\triangle$), and $U_2<U=5$ ${\rm V}<U_s$ ($\Box$).
918: The shift of the maximum of $\varepsilon''$ in the case of $U=5$ ${\rm V}$
919: gives evidence of nonlinear effects in collective azimuthal excitations
920: of molecules.}
921: \label{f10}
922: \end{figure}
923: %
924: \begin{table}
925: \caption{Values of $u_+$, $\nu$, and $f_{\rm max}$ determined for chevron
926: samples of different thicknesses. The parameters $u_+$ and $\nu$ were
927: obtained by applying the procedure described in Sec. II and by using
928: electro-optic data.
929: \label{t1}}
930: \begin{tabular}{cccccccccc}
931: &$d\,\,[{\rm\mu m}]$& \hspace{14mm} &$u_+\,\,[{\rm m^{-1}}]$& \hspace{7mm}
932: & $\nu$& \hspace{7mm} & $f_{\rm max}\,\,[{\rm Hz}]$& \vspace{1mm}\\
933: \tableline
934: \vspace{-3mm}\\
935: &\,\,\,$5.0\,$& &$6.95\times 10^4$& &$0.348$& &$1349.3$&\\
936: &$12.1\,$& &$3.67\times 10^4$& &$0.556$& &$\,\,\,363.1$&\\
937: &$25.9\,$& &$1.89\times 10^4$& &$0.510$& &$\,\,\,186.5$&\\
938: &$50.0\,$& &$1.21\times 10^4$& &$0.395$& &$\,\,\,208.9$&
939: \end{tabular}
940: \end{table}
941: %
942: \end{document}
943:
944: % ****** End of file template.aps ******