1: %\documentstyle[aps,graphics,preprint]{revtex}
2: \documentstyle[aps,graphics]{revtex}
3: \newcommand{\qq}[1]{\sc #1}
4: \def\r{{\bf r}}
5: \def\k{{\bf k}}
6: \def\fm{f_{\rm{m}}}
7: \def\q{{\bf \hat q}}
8: \def\dq{\frac{d^3{\bf q}}{(2\pi)^3}}
9: \def\ie{{\it i.e.}}
10: \def\m{{\bf m}}
11: \def\ez{{\bf \hat e}_z}
12: \def\re{{\rm Re}}
13: \def\im{{\rm Im}}
14:
15: \addtolength{\topmargin}{20mm}
16: \begin{document}
17: \draft
18: \title{Effective index of refraction, optical
19: rotation, and circular dichroism in isotropic chiral liquid
20: crystals}
21: \author{D. Lacoste$^1$ and P. J. Collings$^2$ and T. C. Lubensky$^1$}
22: \address{$^1$ Department of Physics,
23: University of Pennsylvania, Philadelphia, PA 19104, USA \\
24: $^2$ Department of Physics and Astronomy, Swarthmore College,
25: Swarthmore, PA 19081, USA}
26: \date{\today}
27: \maketitle
28: \begin{abstract}
29: This paper concerns optical properties of the isotropic phase
30: above the isotropic-cholesteric transition and of the blue phase
31: BP III. We introduce an effective index, which describes spatial
32: dispersion effects such as optical rotation, circular dichroism,
33: and the modification of the average index due to the fluctuations.
34: We derive the wavelength dependance of these spatial dispersion
35: effects quite generally without relying on an expansion in powers
36: of the chirality and without assuming that the pitch of the
37: cholesteric $P$ is much shorter than the wavelength of the light
38: $\lambda$, an approximation which has been made in previous
39: studies of this problem. The theoretical predictions are supported
40: by comparing them with experimental spectra of the optical
41: activity in the BP III phase.
42: \end{abstract}
43: \pacs{PACS number(s):61.30.Mp,42.70.Df,78.20.Ek}
44:
45: \section{Introduction}
46: Chirality in liquid crystals produces a fascinating variety of
47: phases, such as the blue phases. Three blue phases (BPs)
48: designated BP I, BP II, and BP III have been identified, and their
49: structures are now well understood. BP I and BP II exhibit
50: long-range periodic order at the half micron scale and Bragg
51: scatter visible light. For these reasons, the optical study of
52: blue phases is an active field of research, with a particular
53: emphasis on spatial dispersion effects such as the optical
54: activity.
55:
56:
57: The first experiments on the optical activity in the
58: pretransitional region of the isotropic phase above the
59: isotropic-cholesteric transition were carried out by Cheng and
60: Meyer \cite{cheng}. Using a general formulation due to de Gennes,
61: they calculated and confirmed experimentally that
62: the pretransitional optical activity depends on the temperature as $(T-T^*)^{-0.5}$,
63: where $T^*$ is the metastability temperature of the isotropic
64: phase. Two years later, Brazovskii and Dmitriev developed the
65: first complete theory of phase transitions in cholesteric liquid
66: crystals, and they predicted the existence of the blue phases
67: \cite{brazovskii}. The optical activity in the pretransitional
68: region was derived using this approach by Dolganov et al.
69: \cite{dolganov1}. At this time, a detailed Landau theory of the
70: cholesteric blue phases was obtained by Hornreich and Shtrikman,
71: who also provided an outstanding study of the light scattering in
72: the blue phases including a treatment of the polarization of the
73: light, based on the formalism of the M\"{u}ller matrices
74: \cite{hornreich1,hornreich2}. Bensimon, Domany, and Shtrikman
75: studied the optical activity in the pretransitional regime and in
76: the blue phases \cite{bensimon}, confirming and extending the work
77: of Dolganov. In their paper, the blue phase BP III was considered
78: as an amorphous polycrystalline structure distinct from the
79: isotropic phase, and their treatment of the optical activity
80: relied on the long wavelength approximation. On the experimental
81: side, the wavelength dependence of the optical activity in the
82: blue phases was measured by Collings \cite{collings2,collings3}.
83: The structure of BP III still remained mysterious until
84: experiments clearly showed a continuous transition between BP III
85: and the isotropic phase \cite{collings4,keyes} and the existence
86: of a critical point terminating a line of coexistence of the two
87: phases \cite{collings5}. The same year, Lubensky and Stark
88: developed a theory for the isotropic to BP III transition with a
89: chiral liquid-gas like critical point \cite{lubensky}, which was
90: extended to include scaling theory in Ref.~\cite{collings8}. Very
91: recently, there has been a renewal of interest in the study of the
92: blue phases with the discovery of the smectic blue phases
93: \cite{pansu}, exhibiting isotropic-BP III coexistence terminating
94: at a critical point \cite{jamee}.
95:
96: In section \ref{sec:theory} of this paper, we introduce an
97: effective index for the propagation of light in the isotropic
98: phase and in BP III, which contains pretransitional effects such
99: as the optical activity and the circular dichroism. In section
100: \ref{sec:long_wavelength}, we present a standard derivation which
101: is based on the long wavelength approximation \cite{bensimon}. In
102: section \ref{sec:de Vries}, we make a digression on the optical
103: properties of periodic chiral media for comparison with the case
104: of non-periodic chiral media. By evaluating our weak-scattering
105: expression for the optical dielectric constant tensor in a
106: cholesteric phase, we find the well-known de Vries formula, which
107: describes the optical activity of light propagating in this medium
108: along the pitch axis. In section \ref{sec:exact_wavel} we come
109: back to isotropic chiral media and obtain the optical activity for
110: arbitrary wavelength and chirality. This approach is new in that
111: it is general, not limited to the long wavelength approximation,
112: and applicable in particular to the resonant region where the
113: wavelength of the light is of the order of the pitch of the
114: cholesteric, a regime which has not been considered in detail in
115: previous studies. In the long wavelength approximation, we collect
116: contributions to the optical activity up to order $(P/\lambda)^4$.
117: Though various contributions up to this order have appeared
118: separately in the literature
119: \cite{brazovskii,dolganov1,bensimon,dolganov2} and very recently
120: in \cite{sing}, these references are not fully consistent with
121: each other. We therefore think that it is important to collect and
122: discuss all these contributions in one place. We then generalize
123: our method and obtain the wavelength dependance of the circular
124: dichroism and of the symmetric part of the index, which to our
125: knowledge have not appeared in the literature and have never been
126: measured. Finally in section \ref{sec: Theory&exp}, our
127: theoretical predictions are supported by experimental spectra of
128: the optical activity for different values of the chirality in the
129: BP III phase \cite{collings7}. The agreement between theory and
130: experiments is very good. Experiments confirm the presence of a
131: broad maximum in the magnitude of the optical activity when the
132: wavelength is of the order of the pitch and for positive
133: (respectively negative) pitch. This feature is the pretransitional
134: signature in isotropic chiral media of the divergence arising from
135: the long-range periodicity of the cholesteric phase, BP I and BP
136: II.
137:
138: \section{Theory}
139: \label{sec:theory} Let $\epsilon$ be the relative complex-valued
140: dielectric constant of the cholesteric liquid crystal in the
141: isotropic phase or in the blue phase BP III with respect to the
142: average dielectric constant:
143: \begin{equation}
144: \epsilon_{ij}(\r)=\delta_{ij} + \delta \epsilon_{ij}(\r),
145: \end{equation}
146: where $\delta \epsilon_{ij}(\r)$ is a relative anisotropic small
147: local fluctuation of the dielectric constant: $\langle \delta
148: \epsilon_{ij}(\r) \rangle=0$, $|\delta \epsilon_{ij}(\r)| \ll 1$,
149: and $\langle..\rangle$ denotes a thermal average. To simplify the
150: notations, we take the velocity of light in the medium to be $1$,
151: and we do not indicate explicitly the dependence of the dielectric
152: constant on the frequency of light in the medium denoted $\omega$.
153: Maxwell equations in the absence of sources lead to the following
154: Helmholtz equation for the electric field ${\bf E}(\r,\omega)$:
155: \begin{equation}
156: \label{helmoltz} \nabla \times \left( \nabla \times {\bf
157: E}(\r,\omega) \right) = \omega^2 \epsilon \cdot {\bf
158: E}(\r,\omega).
159: \end{equation}
160: From Eq.~(\ref{helmoltz}), we deduce that the propagation of light
161: in a homogeneous medium in the absence of fluctuations is
162: characterized by the following Green's function
163: \begin{equation}\label{Green_libre}
164: G_{ij}^0 (\k,\omega)=\frac{\Delta_{ij}}{ -\omega^2 + k^2 - i0^{+}}
165: - \frac{k_i k_j}{\omega^2 k^2},
166: \end{equation}
167: where $\Delta_{ij}=\delta_{ij}-k_i k_j /k^2$ is a projector on the
168: space transverse to the wave vector $\k$. The first term of
169: Eq.~(\ref{Green_libre}) is the transverse part describing
170: travelling wave solutions of Maxwell equations and the second term
171: is the longitudinal part which describes non-propagating modes.
172: The scattering by the randomly fluctuating part of the dielectric
173: function $\delta \epsilon_{ij}(\r)$ is described by the following
174: 4-rank tensor \cite{stark}
175: \begin{equation}\label{def_tensor_B}
176: B_{ijkl}(\r)=\omega^4 \langle \delta \epsilon_{ik}(\r) \delta
177: \epsilon_{jl}(0) \rangle.
178: \end{equation}
179: Unless specified otherwise, we consider in this paper only
180: non-absorbing media in which the tensor $B_{ijkl}(\r)$ is real.
181: The averaged Green's function $G(\k,\omega)$ follows from Dyson's
182: equation,
183: \begin{equation}\label{Dyson}
184: G^{-1}=(G^{0})^{-1} -\Sigma.
185: \end{equation}
186: In the weak scattering approximation,
187: the tensor $\Sigma$ can be
188: calculated using the following equation \cite{kats}:
189: \begin{equation}\label{Sigma}
190: \Sigma_{ij}(\k,\omega)=\int \dq B_{ijkl}({\bf q}) G_{kl}^0({\bf
191: k-q},\omega),
192: \end{equation}
193: in terms of the Fourier transform $B_{ijkl}({\bf q})$ of the
194: tensor defined in Eq.~(\ref{def_tensor_B}). Note that
195: Eq.~(\ref{Sigma}) is general and applies also to periodic media
196: for which $B_{ijkl}({\bf q})$ has delta-function peaks in which
197: case this equation takes the form of a discrete sum over Bloch
198: waves \cite{galatola}.
199:
200: The general form of $\Sigma(\k,\omega)$ in an isotropic chiral
201: medium is:
202: \begin{equation}\label{def_sigma12}
203: \Sigma_{ij}(\k,\omega)=\Sigma_0(k,\omega)\delta_{ij} +
204: \frac{i\epsilon_{ijl}{k_l}}{\omega} \Sigma_1(k,\omega) +
205: \Sigma_2(k,\omega) \frac{k_i k_j + k_j k_i}{2\omega^2}.
206: \end{equation}
207: If $\kappa$ denotes a chiral parameter of the medium,
208: $\Sigma_{ij}(\k,\omega,\kappa)$ satisfy the general symmetry
209: relation (valid in any medium in the absence of a magnetic field
210: \cite{landau})
211: \begin{equation}\label{symmetry_general}
212: \Sigma_{ij}(\k,\omega,\kappa)=\Sigma_{ji}(-\k,\omega,\kappa),
213: \end{equation}
214: and the chiral symmetry relation
215: \begin{equation}\label{chiral_symmetry}
216: \Sigma_{ij}(\k,\omega,\kappa)=\Sigma_{ij}(-\k,\omega,-\kappa).
217: \end{equation}
218: The chiral symmetry relation implies that $\Sigma_0$ and
219: $\Sigma_2$ are odd functions of $\kappa$ and that $\Sigma_1$ is an
220: even function of $\kappa$. The dispersion relation for light
221: propagation modified by fluctuations is obtained from $\det \left[
222: G_0^{-1}(\k,\omega) -\Sigma(\k,\omega) \right]=0$. In a basis
223: composed of two vectors perpendicular to $\k$, the diagonal
224: elements of $\Sigma$ are equal to $\Sigma_0$, and the off-diagonal
225: elements are $\pm ik \Sigma_1 /\omega$. In such a basis,
226: longitudinal terms such as the last term in
227: Eq.~(\ref{def_sigma12}) vanish,
228: and the following relation is obtained
229: \begin{equation}\label{dispersion}
230: \omega^2-k^2-\Sigma_0(k,\omega)=\pm \frac{k}{\omega} \Sigma_1(k,\omega).
231: \end{equation}
232: This relation determines $k^+(\omega)$ and $k^-(\omega)$ as a
233: function of the frequency $\omega$. Writing $k=\omega+\delta k$
234: and expanding Eq.~(\ref{dispersion}) to first order in $\delta k$,
235: one finds
236: \begin{equation}
237: \label{disper} \delta k^{\pm} = \frac{\Sigma_0(\omega,\omega) \pm
238: \Sigma_1(\omega,\omega)} {-2\omega \mp
239: \Sigma_1(\omega,\omega)/\omega \mp \partial \Sigma_1/\partial k -
240: \partial \Sigma_0/\partial k}.
241: \end{equation}
242: To first order in $\Sigma$, this gives $k^{\pm}(\omega)= \omega -
243: \left[ \Sigma_0(\omega,\omega) \pm \Sigma_1(\omega,\omega)
244: \right]/2\omega$. The optical activity $\Phi$ is defined as the
245: angle through which the polarization vector of linearly polarized
246: light has turned when traversing a medium of length $L$. It is
247: given by $\Phi=\re (k^+(\omega)-k^-(\omega))L/2$. Similarly the
248: circular dichroism, measuring the difference in transmission from
249: left and right circularly polarized waves, is $\Psi=\im
250: (k^+(\omega)-k^-(\omega))L/2$. To first order in $\Sigma$, we have
251: therefore
252: \begin{equation}\label{Eq:Phi}
253: \Phi=\re \frac{\Sigma_1(\omega,\omega)L}{2\omega},
254: \end{equation}
255: and
256: \begin{equation}\label{Eq:Psi}
257: \Psi=\im \frac{\Sigma_1(\omega,\omega)L}{2\omega}.
258: \end{equation}
259: Note that these relations are valid at arbitrary frequency
260: $\omega$ when Eq.~(\ref{Sigma}) is valid. The fluctuations also
261: modify the symmetric part of $\Sigma$, \ie, the average dielectric
262: constant. The relative change of the average dielectric constant
263: within the same approximation is $\Delta \epsilon_0=-
264: \Sigma_0(\omega,\omega)/\omega^2$.
265:
266: We use in this paper the form of the Landau-de Gennes free energy
267: of Ref.~\cite{hornreich2},
268: \begin{equation}\label{landaufreeenergy}
269: F=V^{-1} \int d^3 \r \{ \frac{1}{2} \left[ a \epsilon_{ij}^2 + c_1
270: \epsilon_{ij,l}^2 + c_2 \epsilon_{ij,i} \epsilon_{lj,l} -2d
271: e_{ijl} \epsilon_{in} \epsilon_{jn,l} \right] -\beta \epsilon_{ij}
272: \epsilon_{jl} \epsilon_{li} + \gamma (\epsilon_{ij}^2)^2 \},
273: \end{equation}
274: where $\epsilon_{ij,l}=\partial \epsilon_{ij} /
275: \partial x_l$ and $e_{ijk}$ is the Levi-Civita fully antisymmetric tensor.
276: The coefficient $a$ is proportional
277: to the reduced temperature $t$, whereas all other coefficients are
278: assumed to be temperature independent. As explained in
279: Refs.~\cite{hornreich2} and \cite{lubensky}, it is convenient to
280: write this free energy in dimensionless form, by expressing all
281: lengths in units of the order parameter correlation length
282: $\xi_R=(12\gamma c_1/\beta^2)^{1/2}$. The chirality is then
283: $\kappa_0=q_c \xi_R$ where $q_c=d/c_1$ is the wave number of the
284: cholesteric phase which determines the pitch $P=4 \pi /q_c$. The
285: reduced temperature is defined by $t=(T-T_N^*)/(T_N-T_N^*)$ where
286: $T_N^*$ is the temperature corresponding to the limit of
287: metastability of the isotropic phase \cite{tom}, and
288: $\rho=c_2/c_1$ is the ratio of the two Landau coefficients, which
289: is of order 1. Using the Gaussian approximation for the
290: fluctuations \cite{bensimon}, the dielectric anisotropy
291: correlation tensor defined in Eq.~(\ref{def_tensor_B}) takes the
292: following form
293: \begin{equation}\label{tensor_B}
294: B_{ijkl}({\bf q})=\omega^4 \sum_{m=-2}^{m=2} \Gamma_m(q) T_{ik}^m
295: (\q) T_{jl}^{-m} (\q),
296: \end{equation}
297: where $\Gamma_m(q)$ is evaluated from the equipartition theorem
298: \begin{equation}\label{def_Gamma}
299: \Gamma_m(q)=\frac{k_B T}{t-m\kappa q +q^2 \left[ 1+ \frac{\rho}{6}
300: (4-m^2) \right] }.
301: \end{equation}
302: Note that $q$ denotes the dimensionless wavevector measured in
303: units of $1/\xi_R$, and that Eq.~(\ref{def_Gamma}) has been
304: corrected from the expression given in Ref.~\cite{lubensky} (the
305: factor $\rho/4$ has been replaced by $\rho/6$ in agreement with
306: Ref.~\cite{hornreich2}). In Eq.~(\ref{def_Gamma}), the parameter
307: $\kappa$ represents an effective chirality at the temperature $t$.
308: Except at the critical point, $\kappa$ is different from
309: $\kappa_0$, and the difference $\kappa-\kappa_0$ is proportional
310: to the order parameter of the BP
311: III-isotropic phase transition \cite{lubensky}. The index $m$
312: in Eq.~(\ref{tensor_B}) denotes the five independent
313: modes of the symmetric and traceless part of the dielectric tensor
314: $\delta \epsilon_{ij}(\r)$. Note that Eq.~(\ref{def_Gamma}) is
315: equivalent to
316: \begin{equation}\label{deftm}
317: \frac{k_B T}{\Gamma_m(q)}=t-m\kappa q + \Delta_m q^2 = \Delta_m
318: (q-q_m)^2+\tau_m,
319: \end{equation}
320: with $\Delta_m=1+ \frac{\rho}{6} (4-m^2)$, $q_m=m
321: \kappa/2\Delta_m$, $t_m=\Delta_m q_m^2$ and $\tau_m=t-t_m$. We
322: have introduced $t_m$, the transition temperature of the mode $m$,
323: and $q_m$, the wavevector which minimizes the energy of the mode
324: $m$. Denoting $r=1+\rho/2$, we have explicitly
325: $t_1=t_{-1}=\kappa^2/4r$ and $q_1=-q_{-1}=\kappa /2r$ for the
326: modes $\pm1$, $t_2=t_{-2}=\kappa^2$ and $q_2=-q_{-2}=\kappa$ for
327: the modes $\pm2$, $t_0=0$ and $q_0=0$ for the mode $0$. A plot of
328: $\Gamma_m(q)$ as a function of $q$ is shown in
329: Fig.~\ref{Fig:Gamma} for the modes $m=0,1$ and $2$, a chirality
330: $\kappa=0.2$ and a temperature $t=0.05$. This particular plot
331: shows the dominance of the mode $m=2$, because of the choice of
332: the temperature $t>t_2>t_1$ and $t$ close to $t_2$. More generally
333: it can be shown that as $t \rightarrow t_m$, $\Gamma_m(q)$ becomes
334: a Dirac function localized at $q=q_m$.
335:
336: The tensor $T_{ij}^m$ introduced in Eq.~(\ref{def_tensor_B}) are
337: eigenvectors of the tensor $B_{ijkl}({\bf q})$, with eigenvalues
338: $\omega^4 \Gamma_m(q)$ \cite{brazovskii}. These tensors are
339: \begin{equation}\label{def_T0}
340: T^0(\q)= \frac{1}{\sqrt{6}} \left( 3 \q \q -1 \right),
341: \end{equation}
342: \begin{equation}\label{def_T1}
343: T^1(\q)= \frac{1}{\sqrt{2}} \left[ \q \m(\q) + \m(\q) \q
344: \right]=T^{-1}(\q)^*,
345: \end{equation}
346: \begin{equation}\label{def_T2}
347: T^2(\q)= \m(\q)\m(\q) =T^{-2}(\q)^*,
348: \end{equation}
349: with $\m=(1/\sqrt{2}) ({\bf \hat \xi} + i {\bf \hat \eta})$ and
350: $({\bf \hat \xi},{\bf \hat \eta},\q)$ forming a right-handed
351: system of orthonormal vectors. We choose the vectors ${\bf \hat
352: \xi}$ and ${\bf \hat \eta}$ such that ${\bf \hat \xi}(-\q)={\bf
353: \hat \xi}(\q)$ and ${\bf \hat \eta}(-\q)=-{\bf \hat \eta}(\q)$, so
354: that
355: \begin{equation}\label{m}
356: \m(-\q)=\m(\q)^*.
357: \end{equation}
358:
359:
360: \section{Dispersion effects in the long wavelength approximation}
361: \label{sec:long_wavelength} In this section the optical activity
362: and the circular dichroism of isotropic chiral media is derived
363: using the method of Ref.~\cite{bensimon}, which is valid in the
364: long wavelength approximation and to linear order in $\kappa$. In
365: section \ref{sec:exact_wavel}, we will present another derivation
366: of these results, which does not rely on an expansion in
367: $1/\lambda$ or in $\kappa$. We present here the method of
368: Ref.~\cite{bensimon} because it has been widely used in the
369: literature. Following Ref.~\cite{bensimon}, we use an expansion to
370: first order in ${\bf k}$ of the Green's function
371: \begin{equation}\label{Green_exp}
372: G^0_{ij}(\k - {\bf q},\omega) \simeq G^0_{ij}({\bf q},\omega)
373: -k_l \frac{\partial G^0_{ij}({\bf q},\omega)}{\partial q_l},
374: \end{equation}
375: with
376: \begin{equation}\label{DG}
377: \frac{\partial G^0_{ij}({\bf q})}{\partial q_l}= -\frac{q_i
378: \Delta_{jl}+q_j \Delta_{il}}{\omega^2 (\omega^2-q^2+i0^+)}+
379: \frac{2q_l \Delta_{ij}}{(\omega^2-q^2+i0^+)^2}.
380: \end{equation}
381: Using this expression in Eq.~(\ref{Sigma}), we obtain
382: $\Sigma_1(k\rightarrow0,\omega)$ which was defined in
383: Eq.~(\ref{def_sigma12}). After integration over $\q$, we obtain
384: \begin{equation}\label{SigmaR}
385: \Sigma_1(k\rightarrow0,\omega)=\int dq q^2 {\frac
386: {{\omega}^{3}q\left[-\Omega_1(q){\omega}^{2}+\Omega_1(q){q}^{2}-
387: 2\,\Omega_2(q){\omega}^{2}\right]}{ 12{\pi
388: }^{2}\left({\omega}^2-{q}^2+i0^+ \right)^{2}}},
389: \end{equation}
390: where $\Omega_m(q)=\Gamma_m(q)-\Gamma_{-m}(q)$. In the integration
391: of Eq.~(\ref{SigmaR}), we take the upper limit of integration over
392: $q$ to be infinity. The continuous model used in this paper is
393: valid only up to wave vectors $q_{max}=2\pi/a$,
394: where $a$ is an intermolecular distance, but
395: in the case of Eq.~(\ref{SigmaR}), the
396: integral is not sensitive to the value of $q_{max}$
397: as discussed in Refs.~\cite{dolganov1} and \cite{lubensky}.
398:
399: Expanding Eq.~(\ref{SigmaR}) in powers of $\omega$ and integrating
400: over $q$, we obtain the complex-valued $\Sigma_1(\omega)$ in the
401: long wavelength limit. Using Eqs.~(\ref{Eq:Phi}) and
402: (\ref{Eq:Psi}), we deduce the optical activity
403: \begin{equation}\label{AO}
404: \frac{\Phi}{L}={\frac {\kappa\,\omega^{2} k_B T}{ 48 r^{3/2}\sqrt
405: \tau_1\pi }}+ \left (\,{\frac {1}{4\sqrt r
406: \,\tau_1^{3/2}}}-\,{\frac {1}{\tau_2^{3/2} }} \right
407: )\frac{\omega^{4} k_B T \kappa}{12 \pi},
408: \end{equation}
409: and the circular dicroism
410: \begin{equation}\label{CD}
411: \frac{\Psi}{L}= \left (-{\frac 1 {\tau_2^{2}}}+\,{\frac 1{6
412: \,\tau_1^{2}}}\right ) \frac{\omega^{5} k_B T \kappa}{4\pi}.
413: \end{equation}
414: Deriving Eqs.~(\ref{AO}) and (\ref{CD}), we have assumed that
415: $\kappa\ll \sqrt{\tau_2}$ and $\kappa\ll \sqrt{4r\tau_1}$, and the
416: equations are valid for temperatures such that $t_1<t_2<t$. Within
417: these approximations, the optical rotation and circular dichroism
418: are proportional to the chirality $\kappa$. The next order will be
419: of order $\kappa^3$ as imposed by the symmetry relation of
420: Eq.~(\ref{chiral_symmetry}). The first term in Eq.~(\ref{AO}) is
421: identical to the expression of Refs.~\cite{bensimon} and
422: \cite{dolganov1} for the optical activity in the pretransitional
423: region. This term is associated with the modes $m=\pm1$ and gives
424: the $(T-T_1^*)^{-0.5}$ dependence observed in many experiments. As
425: first noted by Filev, the modes $m=\pm 2$ need to be considered in
426: addition to the contribution from the modes $m=\pm 1$, in a region
427: very close to the transition in highly chiral liquid crystals
428: \cite{filev}. This is the origin of the third term in
429: Eq.~(\ref{AO}), whose contribution has the opposite sign with
430: respect to the contribution from the $m=\pm1$ modes. This term
431: produces a change of sign of the optical activity close to the
432: transition as shown in Fig.~(\ref{Fig:AOvsT}). There has been some
433: debate in the literature concerning the temperature dependance of
434: this term, whether it should be in $(T-T_2^*)^{-1/2}$ or
435: $(T-T_2^*)^{-3/2}$. The very short paper of Ref.~\cite{filev} did
436: not provide enough explanations to answer this point, and that of
437: Ref.~\cite{dolganov2} also lacks explanations and contains some
438: errors or misprints (in particular in Eq.~(4) of
439: Ref.~\cite{dolganov2}). In Ref.~\cite{collings1}, Collings et al.
440: studied experimentally systems of high chirality and found good
441: agreement with Filev's model with a $1/2$ exponent. Unfortunately
442: due to a large number of fitting parameters, the data could also
443: have been explained with an exponent $-3/2$ \cite{collings2}. We
444: shall come back to this point in section \ref{sec:exact_wavel}. In
445: a study of the wavelength dependence of the optical activity in
446: the long wavelength regime \cite{collings3}, Collings et al.
447: confirmed that the modes $m=\pm 2$ contribute to higher order than
448: the modes $m=\pm1$ in agreement with Eq.~(\ref{AO}). The presence
449: of a second order correction in $\omega^4$ for the modes $m=\pm1$
450: with a dependence $(T-T_1^*)^{-3/2}$, which would correspond to
451: the second term in Eq.~(\ref{AO}) has not been proved or disproved
452: experimentally, although quite good fits to the data can be
453: obtained with this term included as found in Refs.~\cite{sing} and
454: \cite{sing2}.
455:
456:
457:
458: \section{De Vries formula}
459: \label{sec:de Vries} We present in this section a derivation of
460: the de Vries formula, which is the well known solution of
461: Maxwell's equations for light propagating in a cholesteric liquid
462: crystal along the pitch axis \cite{deGennes}. The reason of this
463: digression into the optics of periodic chiral media will become
464: clear in the next section, where we explore the relation between
465: optical activity in isotropic chiral liquid crystals and the de
466: Vries formula for the cholesteric phase. We shall take the
467: $z$-axis to be the helical axis of the cholesteric phase, and
468: (${\bf \hat e}_x$, ${\bf \hat e}_y$, $\ez$) to be a right handed
469: frame. In the cholesteric phase, the order parameter, the
470: anisotropy in the dielectric constant, is
471: \begin{equation}\label{cholesteric_orderparam}
472: \delta \epsilon_{ij}= \epsilon_a \left( n_i n_j - \frac{1}{3}
473: \delta_{ij} \right),
474: \end{equation}
475: where $\epsilon_a=\epsilon_\parallel-\epsilon_\perp$, and $n=\cos
476: (\kappa z /2) {\bf \hat e}_x +\sin (\kappa z /2) {\bf \hat e}_y$.
477: Using Eq.~(\ref{cholesteric_orderparam}) and the definitions
478: (\ref{def_T0}-\ref{def_T2}), it is simple to show that
479: \begin{equation}\label{cholesteric_eps}
480: \delta \epsilon_{ij}(\r)= \delta
481: \epsilon_{ij}(z)=\frac{\epsilon_a}{2} \left[ e^{i\kappa z}
482: T_{ij}^{-2} (\ez) + e^{-i\kappa z} T_{ij}^{2} (\ez) \right] -
483: \frac{\epsilon_a}{\sqrt{6}} T_{ij}^0 (\ez).
484: \end{equation}
485: From this, the tensor $B_{ijkl}(\r)=B_{ijkl}(z)$ defined in
486: Eq.~(\ref{def_tensor_B}) can be constructed. To calculate its
487: Fourier transform $B_{ijkl}({\bf q})$, it is convenient to use a
488: spherical coordinate system with ${\bf q}=q(\sin \theta \cos
489: \phi,\sin \theta \sin \phi, \cos \theta)$,
490: \begin{equation}\label{cholesteric_B}
491: B_{ijkl} \left( {\bf q} \right) =\omega^4 \int dx e^{i q \sin
492: \theta \cos \phi x} \int dy e^{i q \sin \theta \sin \phi y} \int
493: dz e^{i q \cos \theta z} \delta \epsilon_{ik}(z) \delta
494: \epsilon_{jl} (0),
495: \end{equation}
496: which can be simplified using Eq.~(\ref{cholesteric_eps})
497: \begin{eqnarray} \label{cholesteric_B_simple} \nonumber
498: B_{ijkl} \left( {\bf q} \right) & = & \frac{\omega^4 \epsilon_a^2
499: \left( 2\pi \right)^3}{4 q^2 \sin \theta} \delta(\phi) \delta
500: (\theta) \left[ \delta \left( q+\kappa \right) T_{ik}^{-2}
501: T_{jl}^{2} + \delta \left( q-\kappa \right) T_{ik}^{2} T_{jl}^{-2}
502: \right] + A_{ijkl} ({\bf q}), \\
503: & =& \frac{\omega^4 \epsilon_a^2 \left( 2\pi \right)^3}{4} \left[
504: \delta^3 \left( {\bf q}+\kappa \ez \right) T_{ik}^{-2} T_{jl}^{2}
505: + \delta^3 \left( {\bf q}-\kappa \ez \right) T_{ik}^{2}
506: T_{jl}^{-2} \right] + A_{ijkl} ({\bf q}),
507: \end{eqnarray}
508: where $T_{ij}^{\pm 2}=T_{ij}^{\pm 2}(\ez)$, and $A_{ijkl}$ is a
509: linear combination of tensors of the form $T_{ik}^m T_{jl}^{m'}$
510: with $m$ and $m'$ being $0,2$ or $-2$. The Bragg peaks of the
511: cholesteric phase located at ${\bf q}=\pm \kappa \ez$ correspond
512: to the first two terms of Eq.~(\ref{cholesteric_B_simple}). Note
513: that in a periodic medium considered in this section, $\Sigma_1$
514: is a function of $\k$, whereas in an isotropic medium to be
515: considered in the next section, $\Sigma_1$ is only a function of
516: $k=|\k|$. In a periodic medium $\Sigma_1(\k,\omega)$ can be
517: derived generally from $\Sigma_{ij}(\k,\omega)$ using
518: \begin{equation}\label{Delta_Sigma}
519: \frac{\Sigma_1(\k,\omega)}{\omega}=\frac{ \left[
520: \Sigma_{ij}(\k,\omega)- \Sigma_{ij}(-\k,\omega) \right]
521: e_{ijl}k_l}{4ik^2}.
522: \end{equation}
523: Using Eqs.~(\ref{Sigma}), (\ref{cholesteric_B_simple}) and
524: (\ref{Delta_Sigma}), we find
525: \begin{equation}\label{Sigma1_Vries}
526: \frac{\Sigma_1(\k,\omega)}{\omega}=\frac{\omega^4 \epsilon_a^2
527: e_{ijm}k_m}{16 i k^2} \left[ G_{kl}^0 \left( \k + \kappa \ez
528: \right) - G_{kl}^0 \left( -\k + \kappa \ez \right) \right] \left[
529: T_{ik}^{-2} T_{jl}^{2} - T_{ik}^{2} T_{jl}^{-2} + \tilde{A}_{ijkl}
530: \right],
531: \end{equation}
532: where $\tilde{A}_{ijkl}= \left( -T_{ik}^{-2} + T_{ik}^{2} \right)
533: T_{jl}^0 / \sqrt{6}$. In Eq.~(\ref{Sigma1_Vries}), the tensors $T$
534: can be eliminated using Eqs.~(\ref{def_T0}), (\ref{def_T2}) and
535: (\ref{Relation1}).
536:
537: We now assumes that $\k$ is along the $z$-axis. In this case, it
538: is simple to show that the tensor $\tilde{A}_{ijkl}$ does not
539: contribute to $\Sigma_1$. According to the analysis of section
540: \ref{sec:theory}, to linear order in $\Sigma$ the optical activity
541: and the circular dichroism are proportional to the real and
542: imaginary part of $\Sigma_1(\k,\omega)$ evaluated at
543: $|\k|=\omega$. With these assumptions, Eq.~(\ref{Sigma1_Vries})
544: gives
545: \begin{equation}\label{Sigma1_Vries3}
546: \frac{\Sigma_1(\k=\omega \ez,\omega)}{\omega}=\frac{\omega^4
547: \epsilon_a^2 \kappa}{\left( \kappa^2 + 2 \omega \kappa -i0^+
548: \right) \left( 2\omega \kappa - \kappa^2 + i0^+ \right)}.
549: \end{equation}
550: The real part of Eq.~(\ref{Sigma1_Vries3}) leads to de Vries
551: formula \cite{deGennes}
552: \begin{equation}\label{Vries_f}
553: \frac{\Phi}{L} = \frac{\pi \epsilon_a^2}{16 P \lambda'^2
554: \left[1-\lambda'^2 \right]},
555: \end{equation}
556: where $\lambda'=\lambda/P$ and $P=4 \pi \xi_R/\kappa$, the pitch
557: of the cholesteric phase, and the circular dichroism $\Psi$ is
558: zero within the same approximations. Note the following features
559: \cite{deGennes}: there is a dispersion anomaly at the Bragg
560: reflection $\lambda'=1$, but both Eq.~(\ref{Vries_f}) and $\Psi=0$
561: break down near the Bragg reflection. Indeed close to the Bragg
562: reflection, terms of higher order in $\Sigma$ contribute to the
563: optical rotation and the circular dichroism, and for this reason
564: Eqs.~(\ref{Vries_f}) and $\Psi=0$ are only valid in the domain
565: $\epsilon_a \lambda \ll ||P|-\lambda|$ \cite{kats2}. The sign of
566: the optical rotation is such that a right-handed helix
567: ($\kappa>0$) produces a positive optical rotation (the material is
568: laevogyric) when $\lambda \ll P$ and a negative one (the material
569: is dextrogyric) when $\lambda \gg P$. In the long wavelength limit
570: $\lambda \ll P$, the de Vries optical activity $\Phi/L$ is of
571: order $P^3/\lambda^4$ and depends on the light propagation
572: direction as opposed to the isotropic-BP III phases where it is of
573: order $P/\lambda^2$ and is independent of the light propagation
574: direction.
575:
576:
577:
578: \section{Wavelength dependance of spatial dispersion
579: effects in an isotropic chiral medium} \label{sec:exact_wavel} In
580: this section, we compute the exact wavelength dependance of the
581: optical activity, the circular dichroism and the average index in
582: an isotropic chiral medium, without relying on the long wavelength
583: approximation and at any order in the chirality $\kappa$. The
584: method has been pioneered by Dolganov \cite{dolganov1,dolganov2}
585: in his study of BP I and BP II, but this reference does not
586: present all the details of the calculation. Instead of using the
587: expansion Eq.~(\ref{Green_exp}), which is only valid in the long
588: wavelength approximation, we calculate exactly
589: $\Sigma_1(k,\omega)$ from Eqs.~(\ref{Sigma}) and
590: (\ref{Delta_Sigma}). Separating the contribution from the modes
591: $m=2$ and $m=1$, we obtain (see appendix \ref{AppendixA} for
592: details)
593: \begin{equation}\label{Vries}
594: \frac{\Sigma_1^{m=\pm2}(k,\omega)}{\omega}=\int \dq \frac{\mp q
595: \omega^2 c^2
596: \Gamma_{\pm2}(q) \left( 2\omega^2-k^2 + k^2 c^2 \right)} {2
597: \left(k^2+2k q c +q^2-\omega^2-i0^+ \right) \left( k^2-2kqc
598: +q^2-\omega^2 -i0^+ \right) },
599: \end{equation}
600: and
601: \begin{equation}\label{noVries}
602: \frac{\Sigma_1^{m=\pm1}(k,\omega)}{\omega}=\int \dq \frac{\mp q
603: \omega^2 \Gamma_{\pm1}(q) \left( c^2 -1 \right) \left( \omega^2+
604: 4k^2 c^2 - k^2 -q^2 \right)} {4 \left(k^2+2kqc+q^2-\omega^2-i0^+
605: \right) \left( -k^2+2kqc-q^2+\omega^2+i0^+ \right) },
606: \end{equation}
607: with $c={\bf \hat k} \cdot \q$. From these equations, the
608: contribution of the modes $m=\pm1$ and $m=\pm2$ of
609: Eq.~(\ref{SigmaR}) of section \ref{sec:long_wavelength} is
610: recovered in the limit $\k \rightarrow 0$. According to the
611: analysis of section \ref{sec:theory} however, the correct
612: procedure to obtain the optical activity and the circular
613: dichroism to leading order in $\Sigma$ is from
614: $\Sigma_1(k,\omega)$ evaluated at $k=\omega$. As will become clear
615: later, this does not give the same result in general when compared
616: to section \ref{sec:long_wavelength} where the limit $\k
617: \rightarrow 0$ is taken in Eq.~(\ref{Green_exp}). Using
618: $k=\omega$, Eqs.~(\ref{Vries}) and (\ref{noVries}) can be
619: simplified
620: \begin{equation}\label{Vries2}
621: \frac{\Sigma_1^{m=\pm2}(\omega,\omega)}{\omega}=\int \dq \frac{\mp
622: \omega^4 c^2
623: \Gamma_{\pm2}(q) \left( 1 + c^2 \right)} {2
624: \left(q + 2\omega c -i0^+ \right) \left( q-2 \omega c -i0^+
625: \right)q },
626: \end{equation}
627: and
628: \begin{equation}\label{noVries2}
629: \frac{\Sigma_1^{m=\pm1}(\omega,\omega)}{\omega}=\int \dq \frac{\mp
630: \omega^2 \Gamma_{\pm1}(q)}{4q} \left( c^2 -1 \right).
631: \end{equation}
632: Note that the integrand of Eq.~(\ref{noVries2}) vanishes for $\k
633: \parallel {\bf q}$ but is non-zero otherwise. This means that the
634: modes $m=\pm1$ contribute only if the light is propagating off
635: axis or if the pitch axis and the director are not perpendicular
636: to each other as in SmC$^*$ for instance. Eq.~(\ref{Vries2}) is
637: the analog for isotropic chiral media of the de Vries formula of
638: Eq.~(\ref{Sigma1_Vries3}) valid for periodic media. The two
639: expressions become identical with the replacements $c\rightarrow1$
640: since the light is propagating along the pitch axis, and
641: $q\rightarrow \pm\kappa$ according to
642: Eq.~(\ref{cholesteric_B_simple}). Other periodic and chiral media
643: like BP I and BP II can be treated as the cholesteric phase in
644: section \ref{sec:de Vries}. Therefore we expect in BP I and BP II
645: a divergence of the $m=\pm2$ contribution to the optical activity
646: at the Bragg condition, {\it i.e.} when the denominator of
647: Eq.~(\ref{Vries2}) is zero, and a change of sign of the optical
648: activity when crossing this point. There should be no divergence
649: for the modes $m=\pm1$ as can be inferred from
650: Eq.~(\ref{noVries2}).
651:
652: The recent reference of Hunte and Singh \cite{sing} contains a
653: derivation of the pretransitional optical activity in isotropic
654: chiral media, which uses a method similar to the one presented in
655: this article although the authors have only applied it to the long
656: wavelength regime. We believe that the authors have made an error
657: in deriving Eqs.~(30) and (31) of Ref.~\cite{sing} which represent
658: respectively the contribution of the modes $m=\pm1$ and of
659: $m=\pm2$ to the optical activity. The integrand of Eq.~(30) of
660: Ref.~\cite{sing} should vanish when $c=1$ as is true for the
661: integrand of Eq.~(\ref{noVries2}) but it does not, and Eq.~(31) of
662: Ref.~\cite{sing} should reproduce the de Vries as is true for
663: Eq.~(\ref{Vries2}) but does not. Therefore we think that the
664: results of the derivation of the optical activity in
665: Ref.~\cite{sing} are incorrect although the method used is valid.
666:
667:
668: A straightforward evaluation of Eq.~(\ref{noVries2}) gives
669: $\Sigma_1$ for the modes $m=\pm1$ and for an arbitrary frequency
670: $\omega$:
671: \begin{equation}\label{Mode1}
672: \frac{\Sigma_1^{m=\pm1}(\omega,\omega)}{\omega}=\frac{\omega^2
673: \kappa k_B T}{24 \pi r^{3/2} \sqrt{\tau_1}}.
674: \end{equation}
675: The $m=\pm 1$ contribution of the optical activity has a trivial
676: frequency dependance of $\omega^2$. The reason is the following:
677: the modes $m=\pm1$ unlike the modes $m=\pm 2$ are longitudinal,
678: therefore these modes are associated with the longitudinal part of
679: the Green's function (also called near-field part) which is the
680: second term in the r.h.s of Eq.~(\ref{Green_libre}). The only pole
681: of the near-field Green's function is at $\omega=0$, which is the
682: reason for the absence of a complex wavelength dependance for the
683: modes $m=\pm1$ and also the reason for which these modes give the
684: leading contribution to the optical activity at small $\omega$.
685:
686:
687: Therefore the most interesting part of the wavelength dependance
688: of the optical activity comes from the modes $m=\pm 2$. It is also
689: the dominant contribution when $t$ is close to $t_2$. The
690: integration in Eq.~(\ref{Vries2}) can be done analytically for
691: $t>t_2$, and the final result (see appendix \ref{AppendixB} for
692: details of the derivation) takes the following form
693: \begin{equation}\label{OAG}
694: \re
695: \frac{\Sigma_1^{m=\pm2}(\omega,\omega)}{\omega}=\frac{-\omega^3}{32
696: \pi \sqrt{\tau_2}} \left[ x_1 f(x_1) - x_3 f(x_3) \right] k_B T,
697: \end{equation}
698: where $x_1=(\kappa+i\sqrt{\tau_2})/2\omega$ and
699: $x_3=(-\kappa+i\sqrt{\tau_2})/2\omega$. Note that Eq.~(\ref{OAG})
700: agrees with the symmetry relation of Eq.~(\ref{chiral_symmetry}).
701: We have introduced the function
702: \begin{equation}\label{function_f}
703: f(x)=-2x^2-\frac{8}{3}+(x+x^3)\ln \left( \frac{x+1}{x-1} \right).
704: \end{equation}
705: This function tends to $-8/3$ as $x\rightarrow\infty$, is
706: equivalent to $16/15x^2$ as $x\rightarrow0$, and diverges at
707: $x=\pm 1$ as a result of the divergence of the initial expression
708: of Eq.~(\ref{Vries2}). Eqs.~(\ref{OAG}) and (\ref{function_f})
709: fully agree with Filev's results in Ref.~\cite{filev} quoted for
710: $\tau_2=0$. Similar functions but not exactly identical functions
711: are present in Refs.~\cite{brazovskii}, \cite{dolganov2},
712: \cite{dolganov3}, and \cite{sing} which could indicate misprints
713: or errors regarding this particular point.
714:
715:
716: It is interesting to consider two particular cases of
717: Eq.~(\ref{OAG}), the case where $\omega \ll \kappa$, which defines
718: the long wavelength approximation and the opposite case where
719: $\omega \gg \kappa$. In the first case, it is easy to deduce from
720: Eq.~(\ref{OAG}) that $\re \Sigma_1^{m=\pm2}(\omega,\omega)/\omega=
721: -2\omega^4 \kappa k_B T /15\pi \sqrt{\tau_2} t$. Therefore the
722: optical activity in the long wavelength approximation for a
723: temperature $t>t_2>t_1$ is
724: \begin{equation}\label{AO_lw}
725: \frac{\Phi}{k_B T L}=\frac{\omega^2 \kappa} {48\pi r^{3/2}
726: \sqrt{\tau_1}} -\frac{\omega^4 \kappa}{15\pi \sqrt{\tau_2}
727: t}+o(\omega^4).
728: \end{equation}
729: Note that the contribution of the modes $m=\pm2$ goes as
730: $(T-T^*_2)^{-3/2}$ if $\kappa \ll t$, but should go as
731: $(T-T^*_2)^{-1/2}$ otherwise. As discussed before,
732: Eq.~(\ref{AO_lw}) is not identical with Eq.~(\ref{AO}) because
733: $\lim_{\omega \rightarrow 0} \Sigma(\omega,\omega) \neq \lim_{k
734: \rightarrow 0} \Sigma(k,\omega)$. Since the modes $m=\pm1$ and
735: $m=\pm2$ contribute with different signs in Eq.~(\ref{AO_lw}),
736: there is a competition between these two modes, which leads to a
737: maximum in the optical activity as a function of the temperature
738: $t$ as shown in Fig.~\ref{Fig:AOvsT} for a chirality $\kappa=0.2$.
739: In the opposite case of $\omega \gg \kappa$, we find using
740: Eq.~(\ref{OAG}) that $\re \Sigma_1^{m=\pm2}(\omega,\omega)/\omega=
741: \omega^2 \kappa k_B T/12\pi \sqrt{\tau_2}$.
742:
743: We now consider the complete wavelength dependance of the optical
744: activity for the modes $m=\pm 2$. In Figs.~\ref{Fig:AOvslambda}a
745: and \ref{Fig:AOvslambda}b, the optical activity and the circular
746: dichroism are shown as a function of the wavelength expressed in
747: units of $\xi_R$, as calculated with Eqs.~(\ref{OAG}) and
748: (\ref{Im_Sigma}). In Fig.~\ref{Fig:AOvslambda}a $\tau_2=10^{-3}$
749: and in Fig.~\ref{Fig:AOvslambda}b $\tau_2=10^{-5}$. Note that in
750: all these figures, a positive value of $\kappa$ has been chosen
751: corresponding to a right-handed helix in the cholesteric phase,
752: but that an opposite optical activity and circular dichroism would
753: have been found with a left-handed helix. These figures clearly
754: show a maximum in the magnitude of the optical activity when the
755: wavelength is equal to the pitch $P$ of the cholesteric phase,
756: which is about $62.8 \xi_R$ for $\kappa=0.2$ in the case of
757: Fig.~\ref{Fig:AOvslambda}. This maximum at this wavelength is the
758: remains in the isotropic phase of the divergence present in the
759: optical activity of BP I and BP II and in the de Vries formula in
760: Eq.~(\ref{Vries_f}). As $t$ gets closer to $t_2$, $\Gamma_2(q)$
761: becomes peaked at $q=\kappa$, and the width of the maximum in the
762: magnitude of the optical activity gets smaller as can be seen by
763: comparing Figs.~\ref{Fig:AOvslambda}a and \ref{Fig:AOvslambda}b.
764:
765: These figures also show the circular dichroism which to our
766: knowledge has never been discussed theoretically or measured
767: directly in isotropic chiral media. The absence of experiments is
768: probably due to the difficulty separating the contribution from
769: pretransitional fluctuations and the contribution from the
770: absorption bands. According to Fig.~\ref{Fig:AOvslambda}, circular
771: dichroism is best observed in the region where $\omega \gg \kappa$
772: as it gets very small when $\omega \ll \kappa$. The curve shown in
773: the figure is the contribution from pretransitional fluctuations,
774: which is valid away from absorption bands. Note that
775: experimentally this circular dichroism translates in the multiple
776: light scattering regime into a difference in the scattering mean
777: free paths for circularly polarized waves. This means that the
778: transport properties of light in the multiple light scattering
779: regime should be different for the different states of circular
780: polarization. Such a difference has indeed been observed in static
781: and dynamic light scattering measurements with circularly
782: polarized light in the isotropic phase and in BP III
783: \cite{collings6}.
784:
785: Finally we have applied our method to the modification of the
786: average dielectric constant due to the fluctuations. In section
787: \ref{sec:theory}, we have defined this quantity as $\Delta
788: \epsilon_0=-\Sigma_0(\omega,\omega)/\omega^2$, where $\Sigma_0$ is
789: the isotropic part of $\Sigma$. The effect of the fluctuations on
790: the symmetric part of the dielectric constant has to our knowledge
791: never been calculated or measured in the pretransitional region.
792: We find that all the five modes contribute to this average
793: dielectric constant. In order to illustrate a case where the
794: corrections can be large, we focus here on the contribution of the
795: modes $m=\pm2$, which as noted before is dominant when $t$ is
796: sufficiently close to $t_2$. Using Eqs.~(\ref{Sigma}) and
797: (\ref{def_sigma12}), we find
798: \begin{equation}\label{Indice}
799: \Sigma_0^{m=\pm2}(\omega,\omega)=\frac{\omega^4}{2 (2\pi)^3} \int
800: d^3 {\bf q} \frac{ \left( 1+c^2 \right) \Lambda_2(q)}{2q\omega c
801: -q^2 + i0^+},
802: \end{equation}
803: with $\Lambda_2(q)=\Gamma_2(q)+\Gamma_{-2}(q)$. After integrating
804: over $c$ and $q$, one obtains
805: \begin{equation}\label{ReIndice}
806: \re \Delta \epsilon_0^{m=\pm2} = \frac{\omega^3}{64 \pi
807: \sqrt{\tau_2} \kappa} \left[ g(x_1,t)-g(x_3,t) \right] k_B T,
808: \end{equation}
809: with \begin{equation} g(x,t)=-\frac{t+(2\omega x)^2}{\omega^2}
810: \left[ \left( 1+x^2 \right) \ln \left( \frac{-1+x}{1+x} \right)
811: +2x \right],
812: \end{equation}
813: and
814: \begin{equation}\label{ImIndice}
815: \im \Delta \epsilon_0^{m=\pm2} = \frac{\omega^3}{4 \pi} \int_0^1
816: \left( 1+x^2 \right) x dx \Lambda_2(2\omega x).
817: \end{equation}
818: As imposed by the symmetry relation of
819: Eq.~(\ref{chiral_symmetry}), $\Delta \epsilon_0$ of
820: Eq.~(\ref{ReIndice}) is an even function of $\kappa$. In
821: Fig.~\ref{Fig:indice}, the real and imaginary part of the average
822: dielectric constant $\Delta \epsilon_0$ is shown as a function of
823: the wavelength, as calculated using Eqs.~(\ref{ReIndice}) and
824: (\ref{ImIndice}) for the modes $m=\pm2$. Similarly to the case of
825: the optical activity, there is a maximum in $\Delta \epsilon_0$ as
826: a function of the wavelength when the wavelength is of the order
827: of the pitch of the cholesteric.
828:
829:
830: \section{Theory vs. Experiment} \label{sec: Theory&exp}
831: In this section, we compare the prediction of our model with the
832: experimental spectra of the optical activity (ORD spectra) in the
833: BP III phase taken from Ref.~\cite{collings7}. This reference is
834: the only one known to us that reports measurements of the optical
835: activity in the isotropic phase and in BP III which are not
836: limited to the long wavelength regime. The ORD spectra are
837: measured in the BP III phase for pure cholesteryl myristate (CM),
838: pure cholesteryl nonanoate (CN) and mixtures of CN with small
839: quantities of cholesteryl chloride (CC), a compound of opposite
840: chirality. The sign of the optical rotation agrees with section
841: \ref{sec:de Vries}: all the samples have a dominant left-handed
842: character ($\kappa <0$ and $P<0$), and the optical activity is
843: indeed found to be negative when $\lambda \ll |P|$. The optical
844: activity presents a broad maximum at $\lambda \simeq -P$, which
845: grows smaller and broader and shifts to higher wavelength as $|P|$
846: is increased. The curves are fitted with a theoretical expression,
847: which is the sum of Eq.~(\ref{Mode1}) (for the modes $m=\pm1$) and
848: Eq.~(\ref{OAG}) (for the modes $m=\pm2$), and as can be seen in
849: Fig.~\ref{Fig:OA_fit}, the fits are excellent. The spatial
850: dispersion of the index, which affects the wavelength of the light
851: in the medium $\lambda$, is incorporated in this fit using the
852: measurements on CN reported in Ref~\cite{pelzl}. There are four
853: free parameters in this fit $A_1$,$A_2$,$A_3$ and $A_4$. These
854: parameters are $A_1=\pi^2 k_B T \xi_R^3 /8\sqrt{\tau_2}$,
855: $A_2=-\kappa/4\pi \xi_R=1/|P|$, $A_3=\sqrt{\tau_2}/4\pi \xi_R$,
856: and $A_4=\kappa k_B T \xi_R^2 \pi / 12\sqrt{\tau_1} r^{3/2}$ so
857: that $x=(\pm A_2+iA_3)\lambda$ where $\lambda$ is expressed in nm
858: so that $A_2$ and $A_3$ have units of nm$^{-1}$. The values of
859: these parameters for the different samples are shown in Table
860: \ref{tab:coeff}. The parameter $A_2$ obtained from the fits should
861: be identical to the absolute value of the inverse pitch of the
862: cholesteric phase. This is indeed the case as shown in
863: Fig.~\ref{Fig:chirality_fit}, where the points correspond to the
864: measurements of Fig.~\ref{Fig:OA_fit} and the solid line is a
865: linear fit which gives a slope of $0.98\pm0.04$ and an intercept
866: of $(-1.2\pm1.6) \cdot 10^{-4}$.
867:
868: \begin{table}[h]
869: \centering
870: {\par\centering
871: \resizebox*{6.5in}{1.2in}{\rotatebox{0}{\includegraphics{tableau.eps}}}
872: \par}
873: \caption{Parameters of the fit of the optical activity spectra for the different
874: samples: pure cholesteryl myristate (CM),
875: pure cholesteryl nonanoate (CN) and mixtures of CN with 5 and 10
876: mol$\%$ of cholesteryl chloride (CC).}\label{tab:coeff}
877: \end{table}
878:
879:
880:
881:
882:
883: \section{Conclusion}
884: In this article, we have discussed spatial dispersion effects
885: (optical activity, circular dichroism and average index) for light
886: propagation in the isotropic-BP III phase, by introducing an
887: effective index which describes the fluctuations. We have obtained
888: the wavelength dependance of the spatial dispersion effects
889: without relying on the long wavelength approximation, on which
890: previous studies on this problem have been based. This new
891: approach allows us to discuss the optical properties in the
892: resonant region when the wavelength of the light is of the order
893: of the pitch of the cholesteric. In the vicinity of this point,
894: there is a divergence of the magnitude of the optical activity in
895: the cholesteric, BP I, and BP II phases and a broad maximum in the
896: case of the BP III and isotropic phases. These features are
897: confirmed by measurements of spectra of the optical activity in
898: the BP III phase for different values of the chirality. We have
899: also provided predictions for the wavelength dependance of the
900: circular dichroism and for the symmetric part of the effective
901: index. We hope that this work will motivate experimentalists to
902: study the
903: optical properties of periodic and non-periodic chiral media.\\
904:
905: We acknowledge many stimulating discussions with B. Pansu, P.
906: Ziherl and P. Galatola. This work was supported in part by the
907: MRSEC program under grant NSF DMR00-79909. D. Lacoste received
908: support by a grant from the French Ministry of Foreign Affairs.
909:
910: \appendix
911: \section{Derivation of $\Sigma_1(k,\omega)$}
912: \label{AppendixA} Using Eqs.~(\ref{Sigma}) and
913: (\ref{Delta_Sigma}), we find that
914: \begin{equation}\label{Sigma_n}
915: \frac{\Sigma_1(k,\omega)}{\omega}=\frac{e_{ijm}k_m}{4ik^2} \int
916: \dq \left[ B_{ijkl}(-{\bf q}) - B_{ijkl}({\bf q}) \right]
917: G_{kl}^0({\bf k+q},\omega).
918: \end{equation}
919: For the modes $m=\pm2$, $B_{ijkl}({\bf q})=\omega^4
920: \Gamma_2(q) T_{ik}^2 (\q) T_{jl}^{-2}
921: (\q)=\omega^4 \Gamma_2(q) m_i m_k m_j^* m_l^*$ according to
922: Eqs.~(\ref{tensor_B}) and (\ref{def_T2}). Using the relation
923: \begin{equation}\label{Relation1}
924: e_{ijl} k_l m_i m_j^*=-i \k \cdot \q,
925: \end{equation}
926: together with Eq.~(\ref{m}), Eq.~(\ref{Sigma_n}) takes a simpler
927: form
928: \begin{equation}\label{Sigma_12}
929: \frac{\Sigma_1^{m=\pm2}(k,\omega)}{\omega}=\frac{\omega^4 \k \cdot
930: \q}{4k^2} \int \dq \Gamma_2(q) \left( m_k^* m_l + m_k m_l^*
931: \right) G_{kl}^0({\bf k+q},\omega).
932: \end{equation}
933: With the definition Eq.~(\ref{Green_libre}), this expression
934: reduces to Eq.~(\ref{Vries}).
935:
936: Similarly, $B_{ijkl}({\bf q})=\omega^4
937: \Gamma_1(q) T_{ik}^1 (\q) T_{jl}^{-1}
938: (\q)=\omega^4 \Gamma_1(q) \left( {\hat q}_i m_k + m_i {\hat q}_k
939: \right) \left( {\hat q}_j m_l^* + m_j^* {\hat q}_l \right)$ for
940: the modes $m=\pm1$ according to Eqs.~(\ref{tensor_B}) and
941: (\ref{def_T1}). Using the relation
942: \begin{equation}\label{Relation2}
943: e_{ijl} m_i {\hat q}_j k_l = i \k \cdot \m,
944: \end{equation}
945: Eq.~(\ref{Sigma_n}) takes the form
946: \begin{equation}\label{Sigma_11}
947: \frac{\Sigma_1^{m=\pm1}(k,\omega)}{\omega}=\frac{-\omega^4}{2k^2}
948: \int \dq \Gamma_1(q) \left( \k \cdot \m^* \, m_k {\hat q}_l + \k
949: \cdot \m \, {\hat q}_k m_l^* -\k \cdot \q \, {\hat q}_k {\hat q}_l
950: \right) G_{kl}^0({\bf k+q},\omega),
951: \end{equation}
952: which after simple algebra reduces to Eq.~(\ref{noVries}).
953:
954:
955: \section{Derivation of Eq.~(\ref{OAG})} \label{AppendixB}
956: Using the change of variables $x=q/2\omega$, Eq.~(\ref{Vries2})
957: can be written as
958: \begin{equation}\label{ReS}
959: \frac{\Sigma_1^{m=\pm 2}(\omega,\omega)}{\omega}=\mp
960: \frac{\omega^4}{8 \pi^2} \int_0^{\infty} x dx \Gamma_{\pm 2}(2
961: \omega x) K(x),
962: \end{equation}
963: where $K(x)$ denotes the integral
964: \begin{equation}\label{K}
965: K(x)=\int_{-1}^{1} \frac{c^2 \left(1+c^2 \right) dc}{ (x-i0^+
966: -c)(x-i0^+ +c)},
967: \end{equation}
968: which can be decomposed in
969: \begin{equation}\label{ReK}
970: \re K(x)=-2x^2-\frac{8}{3}+(x+x^3)\ln \frac{|x+1|}{|x-1|},
971: \end{equation}
972: and
973: \begin{equation}\label{ImK}
974: \im K(x)= \pi x \left( 1+x^2 \right),
975: \end{equation}
976: if $-1 \leq x \leq 1$ and $0$ otherwise. The integration over $q$
977: in Eq.~(\ref{ReS}) can be carried out in the complex plane with
978: the function $f$ defined in Eq.~(\ref{function_f}) which is
979: identical with $\re K$ except for the absence of absolute values
980: and for this reason $f$ is analytical. $\Gamma_2(q)$ has only two
981: poles in the upper half plane $\kappa+i \sqrt{\tau_2}$ and
982: $\kappa-i \sqrt{\tau_2}$, which correspond to $x_1$ and $x_3$ in
983: Eq.~(\ref{OAG}).
984:
985: Similarly, the imaginary part of
986: $\Sigma_1^{m=\pm2}(\omega,\omega)$ can be obtained from
987: Eqs.~(\ref{ReS}) and (\ref{ImK})
988: \begin{equation}\label{Im_Sigma}
989: \im \frac{\Sigma_1^{m=\pm2}(\omega,\omega)}{\omega}=\mp
990: \frac{\omega^4}{8\pi} \int_0^1 x dx \Gamma_{\pm2}(2\omega x) x
991: (1+x^2),
992: \end{equation}
993: which can also be calculated analytically as a function of
994: $\tau_2$, $\kappa$ and $\omega$.
995:
996:
997:
998: \begin{figure}
999: {\par\centering
1000: \resizebox*{3.6in}{3.6in}{\rotatebox{0}{\includegraphics{Bp4.eps}}}
1001: \par}
1002: \caption{Amplitude $\Gamma_m(q)/k_B T$ of the dielectric
1003: anisotropy correlation tensor as a function of the dimensionless
1004: wave vector $q$ for the modes $m$=0,1, and 2. $\kappa$ is $0.2$,
1005: $\rho=1$, and the normalized temperature is $t=0.05$. The curves
1006: show the predominance of the mode $m=2$, because of the choice
1007: $t>t_2>t_1$ and $t$ close to $t_2$.} \label{Fig:Gamma}
1008: \end{figure}
1009:
1010: \begin{figure}
1011: {\par\centering
1012: \resizebox*{4.6in}{3in}{\rotatebox{0}{\includegraphics{bp3.eps}}}
1013: \par}
1014: \caption{Optical activity $\Phi/Lk_BT$, the dark curve, and
1015: circular dichroism $\Psi/Lk_BT$, the light curve as a function of
1016: the temperature $t$ for $\kappa=0.2$. The wavelength is
1017: $\lambda=100 \xi_R$, which corresponds to the long wavelength
1018: regime where Eq.~(\ref{AO_lw}) is applicable. The curve of the
1019: optical activity illustrates the competition between the modes
1020: $m=\pm1$ and $m=\pm2$, the latter been responsible for the
1021: decrease (and the change of sign) of the optical activity when $t$
1022: approaches $t_2$.} \label{Fig:AOvsT}
1023: \end{figure}
1024:
1025: \begin{figure}
1026: {\par\centering
1027: \resizebox*{4.6in}{3.2in}{\rotatebox{0}{\includegraphics{bp2.eps}}}
1028: \par} {\par\centering
1029: \resizebox*{4.6in}{3.2in}{\rotatebox{0}{\includegraphics{bp1.eps}}}
1030: \par} \caption{The contribution from the modes $m=\pm2$ to the
1031: optical activity $\Phi/Lk_BT$, the dark curve, and the circular
1032: dichroism $\Psi/Lk_BT$, the light curve,
1033: are shown as function of the wavelength $\lambda$ for $\kappa=0.2$.
1034: (a) The temperature $t$ is such that $\tau_2=10^{-5}$, and (b)
1035: $\tau_2=10^{-3}$. The optical activity is the darker curve, and
1036: the wavelength is expressed in units of $\xi_R=25$nm.}
1037: \label{Fig:AOvslambda}
1038: \end{figure}
1039:
1040: \begin{figure}
1041: {\par\centering
1042: \resizebox*{4.6in}{3.2in}{\rotatebox{0}{\includegraphics{bp5.eps}}}
1043: \par} \caption{Real (in dark) and imaginary (in light) part of the $m=\pm2$ contribution
1044: to the average dielectric constant $\Delta \epsilon_0$ as a
1045: function of the wavelength which is expressed in units of
1046: $\xi_R=25$nm. Note that $\kappa=0.2$ and $\tau_2=10^{-5}$.}
1047: \label{Fig:indice}
1048: \end{figure}
1049:
1050: \begin{figure}
1051: {\par\centering
1052: \resizebox*{4.6in}{3.2in}{\rotatebox{0}{\includegraphics{OptAct.eps}}}
1053: \par}
1054: \caption{Measured ORD spectra (points) in the BP III phase as
1055: function of the wavelength of light in vacuum $\lambda_0$,
1056: together with a fit (solid curve) using Eqs.~(\ref{Mode1}) and
1057: (\ref{OAG}). The ORD spectra are shown for pure cholesteryl
1058: myristate (CM) with circles, for pure cholesteryl nonanoate (CN)
1059: with squares, for mixtures of CN with $5$ and $10$ mol\% of
1060: cholesteryl chloride (CC) with upper and lower triangles,
1061: respectively.} \label{Fig:OA_fit}
1062: \end{figure}
1063:
1064: \begin{figure}
1065: {\par\centering
1066: \resizebox*{4.6in}{3.2in}{\rotatebox{0}{\includegraphics{kappa.eps}}}
1067: \par}
1068: \caption{Chirality parameter $A_2$ deduced from the fit of
1069: Fig.~\ref{Fig:OA_fit} as a function of the absolute value of the
1070: inverse pitch of the cholesteric phase $1/|P|$.}
1071: \label{Fig:chirality_fit}
1072: \end{figure}
1073:
1074: %\bibliographystyle{unsrt}
1075: %\bibliography{bp}
1076:
1077: \begin{thebibliography}{10}
1078:
1079: \bibitem{cheng}
1080: J.~Cheng and R.~B. Meyer.
1081: \newblock {\em Phys. Rev. A}, 9:2744, 1974.
1082:
1083: \bibitem{brazovskii}
1084: S.~A. Brazovki\u{i} and G.~Dmitriev.
1085: \newblock {Z}h. {E}ksp. {T}eor. {F}iz. {\bf 69}, 979 (1975).
1086: \newblock {\em Sov. Phys.-JETP}, 42(3):497, 1976.
1087:
1088: \bibitem{dolganov1}
1089: V.~K. Dolganov, S.~P. Krylova, and V.~M. Filev.
1090: \newblock {Z}h. {E}ksp. {T}eor. {F}iz. {\bf 78}, 2343 (1980).
1091: \newblock {\em Sov. Phys.-JETP}, 51(6):1177, 1980.
1092:
1093: \bibitem{hornreich1}
1094: R.~M. Hornreich and S.~Shtrikman.
1095: \newblock {\em Phys. Rev. A}, 28(3):1791, 1983.
1096:
1097: \bibitem{hornreich2}
1098: H.~Grebel, R.~M. Hornreich, and S.~Shtrikman.
1099: \newblock {\em Phys. Rev. A}, 28(2):1114, 1983.
1100:
1101: \bibitem{bensimon}
1102: D.~Bensimon, E.~Domany, and S.~Shtrikman.
1103: \newblock {\em Phys. Rev. A}, 28:427, 1983.
1104:
1105: \bibitem{collings2}
1106: J.~Ennis, J.~E. Wyse, and P.~J. Collings.
1107: \newblock {\em Liquid Crystals}, 5(3):861, 1989.
1108:
1109: \bibitem{collings3}
1110: P.~J. Collings.
1111: \newblock {\em Modern Physics Letters B}, 6(8):425, 1992.
1112:
1113: \bibitem{collings4}
1114: Z.~Kutnjak, C.~W. Garland, J.~L. Passmore, and P.~J. Collings.
1115: \newblock {\em Phys. Rev. Lett.}, 74(24):4859, 1995.
1116:
1117: \bibitem{keyes}
1118: E.~P. Koistinen and P.~H. Keyes.
1119: \newblock {\em Phys. Rev. Lett.}, 74(22):4460, 1995.
1120:
1121: \bibitem{collings5}
1122: Z.~Kutnjak, C.~W. Garland, C.~G. Schatz, P.~J. Collings, C.~J.
1123: Booth, and J.~W.
1124: Goodby.
1125: \newblock {\em Phys. Rev. E}, 53(5):4955, 1996.
1126:
1127: \bibitem{lubensky}
1128: T.~C. Lubensky and H.~Stark.
1129: \newblock {\em Phys. Rev. E}, 53(1):714, 1996.
1130:
1131: \bibitem{collings8}
1132: M.~A. Anisimov, V.~A. Agayan, and P.~J. Collings.
1133: \newblock {\em Phys. Rev. E}, 57(1):582, 1998.
1134:
1135: \bibitem{pansu}
1136: B.~Pansu, E.~Grelet, M.~H. Li, and H.~T. Nguyen.
1137: \newblock {\em Phys. Rev. E}, 62(1):658, 2000.
1138:
1139: \bibitem{jamee}
1140: P.~Jam\'ee, G.~Pitsi, M.~H. Li, H.~T. Nguyen, G.~Sigaud, and
1141: J.~Thoen.
1142: \newblock {\em Phys. Rev. E}, 62(3):3687, 2000.
1143:
1144: \bibitem{dolganov2}
1145: E.~I. Demikhov and V.~K. Dolganov.
1146: \newblock {K}ristallografika {\bf 34}, 1198 (1989).
1147: \newblock {\em Sov. Phys. Cristallogr.}, 34(5):723, 1989.
1148:
1149: \bibitem{sing}
1150: C.~Hunte and U.~Singh.
1151: \newblock {\em Phys. Rev. E}, 64:031702--1, 2001.
1152:
1153: \bibitem{collings7}
1154: P.~J. Collings.
1155: \newblock {\em Phys. Rev. A}, 30(4):1990, 1984.
1156:
1157: \bibitem{stark}
1158: H.~Stark and T.~C. Lubensky.
1159: \newblock {\em Phys. Rev. E}, 55(1):514, 1997.
1160:
1161: \bibitem{kats}
1162: E.~I. Kats.
1163: \newblock {Z}h. {E}ksp. {T}eor. {F}iz. {\bf 65}, 2487 (1973).
1164: \newblock {\em Sov. Phys.-JETP}, 38:1242, 1974.
1165:
1166: \bibitem{galatola}
1167: P.~Galatola.
1168: \newblock {\em Phys. Rev. E}, 55(4):4338, 1997.
1169:
1170: \bibitem{landau}
1171: L.~D. Landau and E.~M. Lifshitz.
1172: \newblock {\em Electrodynamics of continuous media}.
1173: \newblock Pergamon Press, New York, 1960.
1174:
1175: \bibitem{tom}
1176: T.~C. Lubensky and P.~Chaikin.
1177: \newblock {\em Principles of Condensed Matter Physics}.
1178: \newblock Cambridge University Press, New York, 1995.
1179:
1180: \bibitem{filev}
1181: V.~K. Filev.
1182: \newblock {Z}h. {E}ksp. {T}eor. {F}iz. {\bf 37}, 589 (1983).
1183: \newblock {\em Sov. Phys.-JETP Lett.}, 37:703, 1983.
1184:
1185: \bibitem{collings1}
1186: P.~R. Battle, J.~D. Miller, and P.~J. Collings.
1187: \newblock {\em Phys. Rev. A}, 36(1):369, 1987.
1188:
1189: \bibitem{sing2}
1190: C.~Hunte, U.~Singh, and P.~Gibbs.
1191: \newblock {\em J. Phys. II}, 6:1291, 1996.
1192:
1193: \bibitem{deGennes}
1194: P.~G. de~Gennes and J.~Prost.
1195: \newblock {\em The {P}hysics of {L}iquid {C}rystals}.
1196: \newblock Oxford Science Publications, Oxford, 1993.
1197:
1198: \bibitem{kats2}
1199: E.~I. Kats.
1200: \newblock {Z}h. {E}ksp. {T}eor. {F}iz. {\bf 59}, 1854 (1970).
1201: \newblock {\em Sov. Phys.-JETP}, 32(5):1004, 1971.
1202:
1203: \bibitem{dolganov3}
1204: V.~A. Belyakov, E.~I. Demikhov, V.~E. Dmitrienko, and V.~K.
1205: Dolganov.
1206: \newblock {Z}h. {E}ksp. {T}eor. {F}iz. {\bf 89}, 2035 (1985).
1207: \newblock {\em Sov. Phys.-JETP}, 62(6):1173, 1985.
1208:
1209: \bibitem{collings6}
1210: U.~Singh, P.~J. Collings, C.~J. Booth, and J.~W. Goodby.
1211: \newblock {\em J. Phys. II France}, 7:1683, 1997.
1212:
1213: \bibitem{pelzl}
1214: G.~Pelzl and H.~Sackmann.
1215: \newblock {\em Z. Phys. Chem. (Leipzig)}, 254:354, 1973.
1216:
1217: \end{thebibliography}
1218:
1219: \end{document}
1220: