Ordering of the lamellar phase under a shear flow
Extracted source text ยท arXiv cond-mat/0210159
\documentstyle[twocolumn,prl,aps,epsfig]{revtex}
%\documentstyle[preprint,prb,aps,epsf]{revtex}
\def\inseps#1#2{\def\epsfsize##1##2{#2##1} \centerline{\epsfbox{#1}}}
\begin{document}
\draft
\input{psfig}
%\twocolumn[\hsize\textwidth\columnwidth\hsize\csname @twocolumnfalse\endcsname]
\title{Ordering of the lamellar phase under a shear flow}
\author{F. Corberi}
\address{Istituto Nazionale per la Fisica della Materia,
Unit\`a di Salerno {\rm and}
Dipartimento di Fisica, Universit\`a di Salerno,
84081 Baronissi (Salerno), Italy}
\author{G. Gonnella$^{(+)}$ and A. Lamura$^{(+)(*)}$}
\address{$^{(+)}$ Istituto Nazionale per la Fisica della Materia
{\rm and} Dipartimento di Fisica, Universit\`a di Bari {\rm and}
Istituto Nazionale di Fisica Nucleare, Sezione di Bari, via Amendola
173, 70126 Bari, Italy\\
$^{(*)}$ Istituto Applicazioni Calcolo, CNR, Sezione di Bari, via Amendola 122/I
- 70126 Bari - Italy}
\date{\today}
\maketitle
\begin{abstract}
The dynamics of a system
quenched into a state with lamellar order
and subject to an uniform shear flow
is solved in the large-$N$ limit.
The description is based on the Brazovskii free-energy and
the evolution
follows a convection-diffusion equation.
Lamellae order preferentially
with the normal along the vorticity direction.
Typical lengths grow as $\gamma t^{5/4}$ (with logarithmic corrections)
in the flow direction
and logarithmically in the shear direction.
Dynamical scaling holds in the two-dimensional case while
it is violated in $D=3$.
\end{abstract}
\pacs{PACS numbers: 64.75.+g; 05.70.Ln; 47.20.Hw}
Many systems in nature exhibit lamellar order.
One example is a diblock copolymer melt where
chains of type A and B covalently bonded end-to-end
in pairs segregate at low temperatures
with A-B junctions forming a stack of lamellae \cite{Bat}.
In ternary mixtures a lamellar phase is stable
with ordered sheets of surfactant
separating (say) oil and water domains \cite{GS94}.
Lamellar order is also observed in
Raleigh-B\'enard cells, where
convective rolls form
above the convective threshold \cite{CH}.
Further examples include smectic liquid crystals \cite{Deg},
dipolar fluids with long-range interactions \cite{SD}
and chemically reactive binary mixtures \cite{GC}.
A theoretical model for the general description of the
lamellar-disordered phase transition was proposed
by Brazovskii \cite{bra}
who showed the first-order character of
the transition induced by fluctuations.
Lamellar phases under an applied shear flow show a very rich
behavior which is relevant for many applications~\cite{larson}.
A variety of transitions in morphology and orientation
occur as shear rate and temperature are changed.
Stable configurations of lamellae
lying along the flow with the normals differently oriented
have been observed~\cite{Kop,Win}
and analyzed evaluating the effects of the flow on the fluctuation
spectra~\cite{cates,Fr95,MG,Fr}.
Non-stationary properties are far less considered \cite{KD96,Ba98,DCV,GIC}.
In this Letter we study
the effects of a simple planar shear flow on the ordering
of the lamellar phase in a system
quenched from an initially high temperature disordered state.
We present the first analytical
results
on the kinetics of this system by solving the Brazovskii
model in the limit of an
infinite number of components of the order parameter.
This is one of the few methods allowing an explicit
solution for phase ordering systems~\cite{B94}.
A similar approach for fluids under shear flow
has been used in~\cite{OK,Fre,FL,O87,daw,HM,BR,CGL}.
The behavior of quenched binary mixtures without imposed flows
is characterized by dynamical scaling: The structure factor obeys
$ C(\vec k, t) = R(t)^D f[k R(t)]$
where $R(t) \sim t^{\alpha}$ is the typical domain size
and $D$ is the space dimensionality~\cite{Gunton,B94}.
In the case of a fluid with lamellar order,
if the order parameter is not conserved as
in the Swift-Hohenberg model for
Raleigh-B\'enard convection \cite{SH}, regimes exhibiting dynamical
scaling have been found~\cite{BV}.
In models with conserved order parameter,
when diffusion is the only segregating physical mechanism,
simulations show the entanglement of the fluid
into frozen intertwined structures \cite{LG92,BO90}.
In this case the effects of hydrodynamical modes are crucial
for reaching order on large scales \cite{yeo}.
When shear is applied, our results
show that lamellae grow preferentially
with the perpendicular orientation~\cite{Fr95},
namely along the plane
formed by the flow and the shear (velocity gradient)
directions.
Their typical size, obtained from
the second momentum
of the structure factor, grows as $\gamma t^{5/4} \sqrt{\ln t}$ in the flow
direction and as $ \sqrt{\ln t}$ in the shear direction.
Surprisingly we find that
dynamical scaling is obeyed in two dimensions but not in $D=3$.
Our results concern
the cases of conserved and not conserved order parameter,
and apply to most of the systems mentioned above.
We consider the Brazovskii free energy
\begin{equation}
{\cal F}\{\phi \}=\int d\vec r \left \{ \frac {r}{2} \phi ^{2} +
\frac {u}{4} \phi ^{4} - \frac {b} {2} \vert \nabla \phi \vert ^{2}
+ \frac {1} {2} ( \nabla ^{2} \phi)^{2} \right \}
\label {freen}
\end{equation}
where $\phi(\vec r,t) $ is the order parameter field and $u > 0$.
With a negative value of $r$
and $b < 0$ the system orders
in one of the two minima of the local potential.
However,
when $b > 0 $ interfaces are favoured and a modulated state with
wavevector $ k_M = \sqrt{b/2}$ is stable~\cite{bra}.
When a flow is imposed, the kinetics
can be described by the convection-diffusion equation \cite{On97}
\begin{equation}
\frac {\partial \phi}{\partial t}+\vec \nabla \cdot (\phi \vec v)
=- \Gamma_p \frac {\delta {\cal F}} {\delta \phi} \quad, \hskip1cm p=0,2
\label{motion}
\end{equation}
where
\begin{equation}
\Gamma_0 = \Gamma, \qquad
\Gamma_2 = - \Gamma \nabla ^{2}
\label{c}
\end{equation}
and $\Gamma$ is a mobility coefficient.
$p=0 $ describes systems with non-conserved order parameter
(NCOP) and
corresponds to the Swift-Hohenberg equation \cite{SH};
the case $p=2$ is for conserved order parameter (COP)
and applies for example to copolymer melts.
For uniform shear flow in the $x$-direction
$ v_x =\gamma~y$, $ \gamma$ being the shear rate.
Eq.(\ref{motion}) neglects thermal fluctuations and possible effects due to
differences in viscosities between the two components~\cite{Fr95}.
The complete description should
take into account the coupling of
Eq.~(\ref{motion}) with the Navier-Stokes equation. However,
the study of Eq.~(\ref{motion}) is a prerequisite for any
more general theory and usually applies to intermediate temporal regimes,
as recognized for phase separation of
simple binary mixtures~\cite{B94}.
Eq.~(\ref{motion})
can be studied analytically generalizing the field $\phi$
to a vector order parameter with $N$ components and taking the
large-$N$ limit \cite{MZ85}.
One obtains
the following equation in Fourier space for the structure factor $C(\vec k,t)=
\langle \phi (\vec k,t) \phi (- \vec k,t)\rangle $
\begin{eqnarray}
\frac {\partial C(\vec k,t)} {\partial t} &-& \gamma k_x
\frac {\partial C(\vec k,t)} {\partial k_y}= \nonumber \\
&-& 2 \Gamma k^{p}\left [
r+u S(t)-bk^{2}+k^{4} \right ] C(\vec k,t)
\label{motoc}
\end{eqnarray}
where we have dropped the component indices due to internal
symmetry. $S(t)$ has to be computed self-consistently through
\begin{equation}
S(t)=\int _{\vert k \vert < \Lambda} \frac {d\vec k}{(2\pi)^D} C(\vec k,t),
\label{self}
\end{equation}
$\Lambda$ being an ultraviolet cut-off.
For a symmetric mixture
quenched from a high temperature homogeneous phase an appropriate initial
condition is $C(\vec k,0)=\Delta$, where $\Delta$ is a constant.
Eq.~(\ref{motoc}) can be integrated yielding
\begin{equation}
C(\vec k,t)=\Delta e^{-2\Gamma\int _0 ^t d\tau \vec {\cal K}^p(\tau)\left [ {\cal K}^4(\tau)
-b {\cal K}^2(\tau)+r+u S(t-\tau) \right ]}
\label{cc}
\end{equation}
where $ \vec {\cal K}(\tau)\equiv \vec k+\gamma \tau k_x \vec e_y$.
Defining
\begin{equation}
Q(t) \equiv \int_0^t d\tau [r + u S(t - \tau)]
\label{q}
\end{equation}
the analysis can be carried out asymptotically
through the {\it ansatz}
\begin{equation}
Q(t) = k_M^{4} t - \frac {1} {2 \Gamma k_M^p }
(2 \ln t - \ln \ln t + \ln v_p),
\label{ansatz}
\end{equation}
where $v_p$ is a constant to be determined.
Eq.(\ref{ansatz}) will be
justified {\it a posteriori} by proving
the solution of
the self-consistency problem. Specifically, the large time
value
$ S_{\infty}$ calculated through (\ref{self}),
inserted into (\ref{q}), must satisfy (\ref{ansatz}) asymptotically, namely
\begin{equation}
S_{\infty} = (k_M^{4}- r )/u
\label{sinfty}
\end{equation}
We now briefly illustrate this for COP in $ D = 3 $~\cite{nota1}
setting $\Gamma = 1 = u = b = -r $.
\begin{figure}
\epsfig{file=brazfig1.ps,bbllx=45pt,bblly=159pt,bburx=526pt,bbury=647pt,width=8.2cm,clip=}
\caption{The typical lengths as a function of the shear strain
in the $x$ ($\ast$), $y$ ($\circ$) and $z$ ($\bullet$) directions.
The lines are proportional to $\gamma t^{5/4} (\log t)^{1/2}$ and
$(\log t)^{1/2}$.}
\end{figure}
The symmetry of the problem suggests the use of the cylindrical set of
variables
$(k_x,k_\perp,\theta)$, where $k_y=k_\perp \cos \theta$ and
$k_z= k_\perp \sin \theta$.
Naive power counting applied to Eq.~(\ref{cc}) indicates that
wave vectors in the flow direction scale asymptotically as
$k_x\sim t^{-3/2}$ while in the perpendicular directions
$k_\perp - k_M \sim t^{-1/2}$.
Then, introducing the new variables $X=k_x L_\parallel(t)$,
$Q=(k_\perp-k_M)L_\perp (t)$, with
$L_\parallel(t)=2\sqrt {2/3} k_M^2 \gamma t^{3/2}$
and $L_\perp(t)=2\sqrt{2t}k_M^2$,
and taking into account Eq.~(\ref{ansatz}), Eq.~(\ref{cc}) reads as
\begin{equation}
C(\vec k, t) = v_2 \Delta
\frac{t^2}{\ln t} e^{-[f(X,Q,\theta)
+ g(X,Q,\theta ,t)]}
\label{cdikappa}
\end{equation}
with
\begin{equation}
f = Q^2+\sqrt{3}Q X\cos \theta +X^2 \cos ^2 \theta
\label{cscal}
\end{equation}
and $g(X,Q,\theta,t)= \sum_{j=1}^{16} t^{-j/2} g_j$, where
$g_j$ are polynomials in $X,Q,\theta, \log t$. For long times
$g$ can be shown to provide a relevant contribution
$g_2 \simeq \frac {9}{160} \frac{X^4}{k_M^6} - \frac {3}{4 k_M^6} X^2 \ln t$
only in the region $Q\sim 0, \cos \theta \sim 0$.
Then the integration over $Q$ and $\theta$ can be performed yielding
\begin{equation}
S(t) \simeq \frac {\sqrt 3 \Delta }{ 32 k_M^{3}\pi ^{\frac{3}{2}}}
\frac {v_2}{\gamma \ln t}
\int _{-\infty} ^\infty
dX e^{-\frac {X^2}{8} - \frac{g_2}{t} }
I_0({X^2}/{8})
\label{integral}
\end{equation}
where $ I_0(z)$ is a Bessel function.
The integral of Eq.~(\ref{integral}),
evaluated by
the asymptotic expansion of $I_0(z)$ at large $z$,
behaves as $ \ln t /\sqrt{\pi}$.
Hence at large times $S(t)$ approaches the constant $ S_{\infty}
= \frac{\sqrt{3} v_2 \Delta }
{32 \pi^2 k_M^3 \gamma} $.
A comparison with Eq.(\ref{sinfty}) fixes
$v_2 = \frac {32 \pi^2 k_M^3 \gamma (k^4_M + 1)}{
\sqrt{3} \Delta }$ and verifies the ansazt.
Due to the presence of $g$,
the function $C(\vec{k},t)$ of Eq.(\ref{cdikappa})
cannot be cast in a scaling form.
However, different scaling behaviors are obeyed in the regions
where the functions $f$ or $g$ can be respectively neglected.
In particular, for $Q\sim 0, \cos \theta \sim 0$, where $g$ dominates,
$C(\vec{k},t)$ scales in the $x$-direction with respect to the length
$l_\parallel(t) = (2/5)^{1/4}\sqrt{k_M} \gamma t^{5/4} $
different from $L_\parallel(t)$.
In $D=2$, on the other hand, where the contribution
of the function $g$ is always negligible,
one finds the scaling form
$ C(\vec k, t) = v_2 \Delta
t^2 e^{-[f(X,Q,0) ]}$.
\begin{figure}
%\vskip 20mm
\epsfig{file=brazfig2.ps,bbllx=80pt,bblly=235pt,bburx=550pt,bbury=755pt,width=8.2cm,clip=}
\caption{The structure factor
is shown for the COP case on the planes $k_x=0$, $k_y=0$, $k_z=0$,
$k_z=k_M$ and for the NCOP case on the planes $k_y=0$, $k_z=k_M$.}
\end{figure}
Ordering properties are usually inferred from the
momenta of the structure factor.
We define
$R_x = [\int
{d\vec k} C(\vec k,t) / \int
{d\vec k} k_x^2 C(\vec k,t)]^{1/2}
$ and similarly for the shear and vorticity directions.
In the absence of dynamical scaling
definitions of $R_x$ based on different
momenta of $C(\vec{k},t) $ may lead to different results. However in this case
it can be proven that changing the order of the momentum does not change
the growth exponent but only the logarithmic correction.
>From
Eq.~(\ref{cdikappa})
we find $R_x \sim \gamma t^{5/4} \sqrt{\ln t}$,
$R_y \sim \sqrt{\ln t}$ and
$R_z \sim k_M^{-1}$.
The behavior of $R_x$ shows the relevance of the scaling with respect to
$l_\parallel(t)$ for the ordering of the system.
The growth of $R_y$
indicates that lamellae order preferentially in the plane $x-y$.
In $D=2$ we find $R_x \sim \gamma t^{3/2}$
and $R_y \sim k_M^{-1}$.
The same results are obtained with non-conserved order parameter.
The behavior of $R_x, R_y, R_z$ resulting from the
numerical integration of Eq.~(\ref{motoc})
with an adaptive grid algorithm is shown in Fig.~1.
After the initial isotropic evolution,
the shear-induced anisotropy becomes evident for values
of the strain $\gamma t$ larger than one and agreement with the
analytical behavior is observed.
It is also useful to illustrate the behavior of the structure factor
(\ref{cdikappa}) shown in Fig.~2.
The maxima of $C(\vec{k},t)$ are located at
$ k_x^2 = 5 \frac {1}{k_M^4 \gamma^2} \frac{\ln t}{t^3}, k_y=0,
k_z^2=k_M^2$ for COP and at $ k_x=0, k_y^2 + k_z^2=k_M^2$ for NCOP.
The shape of $C(\vec{k},t) $ on the planes $k_x =0 $ and $k_z = 0 $
is qualitatively similar for COP and NCOP.
At $k_x = 0 $, in particular, the wavevectors are not affected by the flow
and the structure factor has a circular structure with
radius $k_M $.
At $ k_y =0 $ or $ k_z = k _M$ the behavior of $C(\vec{k},t)$ depends on the
conservation law. The maxima of $C(\vec{k},t)$
developed with NCOP are splitted with COP
into a pair of symmetric narrowing peaks. This pattern is
typical of the case with conserved dynamics and
has been observed in other segregating systems
under shear flow~\cite{CGLE}.
\begin{figure}
\epsfig{file=brazfig3.ps,bbllx=38pt,bblly=145pt,bburx=523pt,bbury=660pt,width=8.2
cm,clip=}
\caption{The excess viscosity as a function of the shear strain.
The straight line has slope -2.}
\end{figure}
Finally we turn to the study of the rheological properties.
In phase separation of binary mixtures the flow
acts against the surface tension inducing stretching of domains
followed by break-up processes and burst of small bubbles~\cite{OND,prl2}.
An excess viscosity $\Delta \eta $
is measured~\cite{KSH}: it reaches
a maximum generally for $\gamma t \geq 1 $
and later decays.
In our case we measure the evolution of the shear stress $\sigma_{xy}$
deducible in a general way from (\ref{freen}) in terms of the structure factor;
it is given by
$\sigma_{x y}(t) = \int
\frac {d\vec k}{(2\pi)^D} k_x k_y (2k^2-b)C(\vec k,t)$ \cite{daw}.
An excess viscosity can be introduced as
$\Delta \eta (t)= - \sigma_{x y}(t)/\gamma$ \cite{Onu}.
Using Eq.~(\ref{cdikappa}) we find $ \Delta \eta \sim
\gamma ^{-1}t^{-2}$.
The numerical results of
Fig.~3 show that, after reaching a maximum,
the decay of $\Delta \eta$ agrees with the analytical behavior.
The other rheological indicators can be shown to behave similarly.
In conclusion we have studied the ordering kinetics of
a lamellar phase in shear flow by solving the dynamics
of the convective-diffusion equation in the large-N limit.
Regarding the debated question of the stable orientation,
our results are in agreement with the expectation
of the stability of the perpendicular phase at high shear rates~\cite{Fr95}.
In $D=3$
a logarithmic growth law is found along the shear direction
while $R_x\sim \gamma t^{5/4}\sqrt {\ln t}$.
Interestingly, the same exponent $5/4$ is found for simple
binary fluids~\cite{BR,CGL}.
Our results show that the scaling depends
on dimensionality and is violated in $D=3$.
This is at variance with the case without shear where
the same approach gives a scaling form independently of $D$~\cite{fu}.
Violation of scaling is quite uncommon~\cite{CZ,B94}.
In this context this phenomenon
may be related to the existence of two lengths $L_\parallel $, $l_\parallel$
growing with different exponents.
A natural question is the relevance of the behavior of the large-N model
to the physical case with $N=1$.
As without flow, symmetry
arguments suggest an analogy between
the coarsening behavior of lamellar phases and that of
vectorial models and one would argue that the large-N
results give reliable indications also for the physical case~\cite{maz};
in particular the large-N exponent is the one expected
asymptotically \cite{BV}.
Simulations of the scalar case with shear
could elucidate this point;
however
strong
finite size effects make
the evaluation of the growth exponent difficult
(see \cite{prl2} and refs. therein).
We also mention that it will be important to study how
the presence of hydrodynamics affects the picture provided in this Letter.
~\\
We thank Mario Pellicoro and Marco Zannetti for useful discussions.
F.C. and G.G. acknowledge M.Cirillo and R. Del Sole
for hospitality in Rome University and
support by the PRA-HOP 1999 INFM and MURST (PRIN 1999).
\begin{references}
\bibitem{Bat}
S.F. Bates, Science {\bf 251}, 898 (1991).
\bibitem{GS94}
G. Gompper and M. Schick, {\em Phase transitions and critical
phenomena} vol. {\bf 16},
eds. C. Domb and J.L. Lebowitz, (Academic Press, New York,
1994).
\bibitem{CH}
M. C. Cross and P. C. Hohenberg, Rev. Mod. Phys.
{\bf 65}, 851 (1993).
\bibitem{Deg}
See, for example, P. de Gennes, {\it The Physics of Liquid Crystals}
(Clarendon, Oxford, 1974).
\bibitem{SD}
C. Roland and R. C. Desai, Phys. Rev. B {\bf 42}, 6658 (1990);
C. Sagui and R. Desai, Phys. Rev. Lett. {\bf 71}, 3995 (1993).
\bibitem{GC}
S. C. Glotzer, E. A. Di Marzio, and M. Muthukumar,
Polym. Mater. Sci. Eng. Proc. Am. Chem. Soc. {\bf 71},
645 (1994);
S. C. Glotzer, D. Stauffer, and N. Yan, Phys. Rev. Lett. {\bf 72},
4109 (1994).
\bibitem{bra}
S. A. Brazovskii, Sov. Phys. JETP {\bf 41}, 85 (1975).
\bibitem{larson}
See, e.g., R. G. Larson, {\it The structure and Rheology of Complex fluids}
(Oxford University Press, New York, 1999).
\bibitem{Kop}
K. A Koppi, M. Tirrell, F. S. Bates, K. Almdal, and R. H. Colby,
J. Phys. France {\bf 2}, 1941 (1992).
\bibitem{Win}
K. I. Winey, S. S. Patel, R. G. Larson, and H. Watanabe,
Macromolecules {\bf 26}, 2542 (1993).
\bibitem{cates}
M. E. Cates and S. T. Milner, Phys. Rev. Lett. {\bf 62}, 1856 (1989).
\bibitem{Fr95}G. Fredrickson, J. Rheol. {\bf 38}, 1045 (1994).
\bibitem{MG}
M. Goulian and S. T. Milner, Phys. Rev. Lett. {\bf 74}, 1775 (1995).
\bibitem{Fr}
A. N. Morozov, A. V. Zvelindovsky, and J. G. E. M. Frajie,
preprint cond-mat/0104339.
\bibitem{KD96}
H. Kodama and M. Doi, Macromolecules {\bf 29}, 2652 (1996).
\bibitem{Ba98}
M. Bahiana, Physica A {\bf 257}, 307 (1998).
\bibitem{DCV}
F. Drolet, P. Chen, and J. Vi\~nals, Macromolecules {\bf 32}, 8603 (1999).
\bibitem{GIC}
N. A. Gross, M. Ignatiev, and B. Chakraborty,
Phys. Rev. E {\bf 62}, 6116 (2000).
\bibitem{B94}
See, e.g., A. J. Bray, Adv. Phys. {\bf 43}, 357 (1994).
\bibitem{OK}
A. Onuki and K. Kawasaki, Ann. Phys. (N.Y.) {\bf 121}, 456 (1979).
\bibitem{Fre}
G. H. Fredrickson, J. Chem. Phys. {\bf 85}, 5306 (1986).
\bibitem{FL}
G. H. Fredrickson and R. G. Larson, J. Chem. Phys. {\bf 86}, 1553 (1987).
\bibitem{O87}
A. Onuki, J. Chem. Phys. {\bf 87}, 3692 (1987).
\bibitem{daw}
G. P\"{a}tzold and K. Dawson, Phys. Rev. E {\bf 54}, 1669 (1996).
G. P\"atzold and K. Dawson, J. Chem. Phys. {\bf 104}, 5932 (1996).
\bibitem{HM}
C. Huang and M. Muthukumar, J. Chem. Phys. {\bf 107}, 5561 (1997).
\bibitem{BR}
N. P. Rapapa and A. J. Bray, Phys. Rev. Lett. {\bf 83}, 3856 (1999).
\bibitem{CGL}
F. Corberi, G. Gonnella, and A. Lamura, Phys. Rev. Lett. {\bf 81},
3852 (1998).
\bibitem{Gunton}
J. Gunton, M. San Miguel, and P. Sahni, in {\it Kinetics
of first order phase transitions} Vol. 8 of {\it Phase
Transitions and Critical Phenomena}, eds. C. Domb
and J. Lebowitz (Academic, London, 1983).
\bibitem{SH}
J. Swift and P. C. Hohenberg, Phys. Rev. A {\bf 15}, 319 (1977).
\bibitem{BV}
K. R. Elder, J. Vi\~nals, and M. Grant, Phys. Rev. Lett. {\bf 68}, 3024 (1992);
see also D. Boyer and J. Vi\~nals, preprint cond-mat/0105543
and references therein.
\bibitem{LG92}
M. Laradji, H. Guo, M. Grant, and M. Zuckermann,
J.Phys. A: Math. Gen {\bf 24}, L629 (1991); J. Phys.:
Condens. Matter {\bf 4}, 6715 (1992).
\bibitem{BO90}
M. Bahiana and Y. Oono, Phys. Rev. A {\bf 41}, 6763 (1990).
\bibitem{yeo}
G. Gonnella, E. Orlandini, and J. Yeomans,
Phys. Rev. Lett. {\bf 78}, 1695 (1997).
\bibitem{On97}
See, e.g., A. Onuki, J. Phys.: Cond. Matt. {\bf 9}, 6119 (1997).
\bibitem{MZ85}
G. F. Mazenko and M. Zannetti, Phys. Rev. B {\bf 32}, 4565 (1985).
\bibitem{nota1}
The proof is similar for NCOP;
details will appear elsewhere.
\bibitem{CGLE}
C. K. Chan, F. Perrot, and D. Beysens, Phys. Rev. A {\bf 43},
1826 (1991); K. Migler, C. Liu, and D. J. Pine,
Macromolecules {\bf 29}, 1422 (1996);
F. Corberi, G. Gonnella, and A. Lamura,
Phys. Rev. E {\bf 61}, 6621 (2000).
\bibitem{OND}
T. Ohta, H. Nozaki, and M. Doi, Phys. Lett. A {\bf 145}, 304 (1990);
J. Chem. Phys. {\bf 93}, 2664 (1990).
\bibitem{prl2}
F. Corberi, G. Gonnella, and A. Lamura, Phys. Rev. Lett. {\bf 83}, 4057 (1999).
\bibitem{KSH}
A. H. Krall, J. V. Sengers, and K. Hamano, Phys. Rev. Lett. {\bf 69},
1963 (1992).
\bibitem{Onu}
A. Onuki, Phys. Rev. A {\bf 35}, 5149 (1987).
\bibitem{fu}
F. Corberi and U. Marini Bettolo Marconi, Europhys. Lett. {\bf 30},
349 (1995).
\bibitem{CZ}
A. Coniglio and M. Zannetti, Europhys. Lett. {\bf 10}, 575 (1989).
\bibitem{maz}
See, e.g., G. Mazenko in {\it Formation and interaction of topological
defects}, eds. A.C.Davis and R. Brandenberger (Plenum. New York, 1995).
\end{references}
\end{document}