1: %\documentclass[showpacs,preprintnumbers,amsmath,amssymb]{revtex4}
2: %\documentclass[preprintnumbers,amsmath,amssymb]{revtex4}
3: \documentclass[twocolumn,amsmath,amssymb]{revtex4}
4:
5: \usepackage{setspace}
6: \usepackage{graphicx}% Include figure files
7: \usepackage{dcolumn}% Align table columns on decimal point
8: \usepackage{bm}% bold math
9: \usepackage{ae}
10:
11: \usepackage{color}
12:
13: \begin{document}
14:
15: \newcommand{\eff}{{\text{eff}}}
16: \newcommand{\bare}{{\text{bare}}}
17: \newcommand{\back}{{\text{back}}}
18: \newcommand{\sat}{{\text{sat}}}
19: \newcommand{\res}{{\text{res}}}
20: \newcommand{\be}{\begin{equation}}
21: \newcommand{\ee}{\end{equation}}
22: \newcommand{\XXXX}{{\bf XXXX}}
23:
24:
25: \title{The renormalized jellium model for spherical and cylindrical colloids}
26:
27: \author{Salete Pianegonda}%
28: \affiliation{Laboratoire de Physique Th\'eorique et Mod\`eles Statistiques
29: (Unit\'e Mixte de Recherche UMR 8626 du CNRS), B\^atiment 100, Universit\'e de Paris-Sud, 91405 Orsay Cedex, France}
30: \affiliation{
31: Instituto de F\'{\i}sica, Universidade Federal do Rio Grande do Sul, CP 15051, 91501-970, Porto Alegre (RS), Brazil}
32:
33: \author{Emmanuel Trizac}
34: \affiliation{CNRS; Universit\'e
35: Paris-Sud, UMR 8626, LPTMS, Orsay Cedex, F-91405}
36: \affiliation{Center for Theoretical Biological Physics, UC San Diego,
37: 9500 Gilman Drive MC 0374 - La Jolla, CA 92093-0374, USA}
38:
39: \author{Yan Levin}%
40: \affiliation{
41: Instituto de F\'{\i}sica, Universidade Federal do Rio Grande do Sul, CP 15051, 91501-970, Porto Alegre (RS), Brazil}
42:
43:
44:
45:
46: \date{\today}
47:
48: \begin{abstract}
49: Starting from a mean-field description for a dispersion of
50: highly charged
51: spherical or (parallel) rod-like colloids, we introduce the simplification of
52: a homogeneous background to include the contribution of other polyions to the
53: static field created by a tagged polyion. The charge of this background is
54: self-consistently renormalized to coincide with the polyion effective charge,
55: the latter quantity thereby exhibiting a non-trivial density dependence, which directly enters into the equation of state
56: through a simple analytical expression.
57: The good agreement observed between the
58: pressure calculated using the renormalized jellium and Monte Carlo
59: simulations confirms the relevance of the
60: {renormalized} jellium model for theoretical and experimental purposes and provides an
61: alternative to the Poisson-Boltzmann cell model since it is free of some
62: of the intrinsic limitations of this approach.
63:
64: \end{abstract}
65:
66: %\pacs{82.70.Dd, 82.45.-h}
67: \maketitle
68:
69:
70:
71: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
72: \section{Introduction}
73: \label{sec:intro}
74: In the realm of soft matter, a quantitative description of systems with a
75: non vanishing density of mesoscopic constituents (colloids) is a difficult
76: task whenever long range Coulomb interactions are present \cite{Belloni}.
77: It is customary to
78: introduce a suitably defined Wigner-Seitz-like cell to render the situation
79: tractable. This considerable simplification allows for the computation
80: of various thermodynamic quantities (see e.g. \cite{Alfrey,Alexander}
81: and \cite{Hansenrevue,Deserno} for more recent accounts).
82: Transport properties may also be derived \cite{Kuwabara,Ohshima}.
83: Initially borrowed from solid state physics, the concept of a cell
84: nevertheless appears fruitful to describe the phase behaviour
85: of liquid phases (see e.g. \cite{LevinJPCM}). When a more microscopic
86: information such as effective interaction is sought,
87: there is however to date no evidence that the cell picture provides
88: accurate answers, through the approximate procedures that have
89: been proposed to infer solvent and micro-ions averaged
90: inter-colloid potentials of interaction \cite{Alexander}.
91: In the present paper, we adopt a
92: more "liquid-state" viewpoint~\cite{LevinL}
93: to describe the local and global screening properties of
94: microscopic ions around
95: highly charged rod-like or spherical colloids, taking due account of the finite
96: density of colloids. Our approach bears a strong resemblance with
97: a Jellium model put forward by Beresford-Smith, Chan and Mitchell \cite{Beresford},
98: with the important difference that the jellium under consideration
99: here is the ``renormalized'' counterpart of that studied in \cite{Beresford}.
100: A preliminary account with emphasis on the sedimentation of charged
101: colloids has been published in \cite{Jellium} and we note that our approach
102: has been recently tested with some success on liposome dispersions
103: \cite{Haro,HaroJFCM06,CastanedaCM}.
104:
105: The article is organized as follows. The model is defined in section
106: \ref{sec:model} and illustrated in section \ref{sec:nosalt} where salt-free suspensions
107: are considered and the numerical procedure exemplified with charged spheres.
108: The cylindrical geometry is also addressed which allows to
109: discuss the fate of classical Manning-Oosawa condensation
110: phenomenon \cite{Manning,Manning2006} within the present
111: framework. The effects of an added
112: electrolyte are investigated in section \ref{sec:salt} and concluding remarks are
113: presented in section \ref{sec:concl}. As will become clear below, our framework
114: provides a procedure to incorporate self-consistently charge
115: renormalization into the classical Donnan equilibrium description of suspensions.
116:
117:
118: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
119: \section{The model}
120: \label{sec:model}
121:
122: When immersed in a polar solvent, mesoscopic ``particles'' release small
123: ions in the solution that, together with other micro-ions resulting from
124: the dissociation of an added salt, form an in-homogeneous cloud
125: around each colloid.
126: Within the mean-field approximation to which we restrict here, neglect of
127: micro-ionic correlations allows to relate the local density of micro-species of
128: valency $z_i$ at point $\bm r$ to the local electrostatic potential $\varphi(\bm r)$
129: through $n_i({\bm r}) \propto \exp(-z_i e \beta \varphi(\bm r))$,
130: where $e$ is the elementary charge and
131: $\beta^{-1}=kT$ is the thermal energy \cite{Belloni,Levin}.
132: For any position of the $N$ colloids present, one needs to solve
133: the resulting Poisson-Boltzmann equation from which the electric potential
134: follows. This potential may then be inserted into the stress tensor \cite{Belloni}
135: to compute the force acting on the colloids.
136: Such a procedure, which
137: makes explicit use of the separation of time scales between colloids and
138: micro-ions, opens the way towards a complete description
139: of the statics and dynamics of the system, with e.g Monte Carlo or Molecular
140: Dynamics techniques to treat the colloidal degrees of freedom
141: \cite{Fushiki,Graz}. This treatment is however numerically
142: particularly demanding and much insight is gained from further approximations
143: that map the original problem with $N$ colloids onto a one-colloid situation.
144: The cell model is an option, where the finite density of colloids is
145: accounted for by an exclusion region. We propose here an alternative,
146: free of some of the limitations of the cell model,
147: that is equally simple to implement and solve.
148:
149:
150: A given colloid with bare charge $Z_\bare e$, assumed positive,
151: is tagged and fixed at the origin. The suspending medium
152: (solvent treated as a dielectric continuum with permittivity $\varepsilon$)
153: is taken as infinite, with a mean colloidal density $\rho$.
154: Following \cite{Beresford},
155: the colloids around the tagged particle are assumed to form
156: an homogeneous background, with charge density
157: $Z_\back e \rho$, so that the electrostatic potential around the tagged colloid
158: fulfills Poisson's equation
159: \be
160: \nabla^2 \varphi \,=\, -\frac{4 \pi}{\varepsilon} \,
161: \left[Z_\back \, \rho \, e + \sum_i n_i^0 z_i e\, e^{-\beta e z_i \varphi}
162: \right]
163: \label{eq:Poisson}
164: \ee
165: where the summation runs over all micro-species and the concentrations
166: $n_i^0$ are determined either from electro-neutrality
167: in the no salt case or from the osmotic equilibrium with a salt
168: reservoir in the semi-grand-canonical situation, both addressed
169: here.
170: At large distances ($r\to \infty$), the term in brackets
171: on the rhs of (\ref{eq:Poisson}) vanishes
172: %to ensure electro-neutrality. This
173: which imposes a
174: value $\varphi_\infty$ for the potential far from the colloid,
175: that may be called a Donnan potential.
176: The key point in our approach is that unlike in \cite{Beresford},
177: $Z_\back\neq Z_\bare$:
178: the background charge is not known {\it a priori} but is determined
179: self consistently as explained below.
180:
181: To illustrate the methodology, we consider a spherical colloid of radius $a$.
182: When $r \to \infty$, we may linearize (\ref{eq:Poisson}) around
183: $\varphi_\infty$, which results in a Helmholtz equation indicating that
184: \be
185: \varphi({\bm r}) \, \stackrel{r\to \infty}{\sim} \, \varphi_\infty \,+\,
186: \frac{Z_{\eff} \, e}{\varepsilon(1+\kappa a)r} \, e^{-\kappa(r-a)}
187: \label{eq:ff}
188: \ee
189: where the characteristic distance $\kappa^{-1}$ is given by
190: \be
191: \kappa^2 \,=\, 4 \pi \sum_i \frac{\beta e^2}{\varepsilon} n_i^0 z_i^2 \, e^{-\beta e z_i \varphi_\infty} = 4\pi \ell_B \sum_i z_i^2 n_i(\infty)
192: \ee
193: and $\ell_B=\beta e^2 /\varepsilon$ is the Bjerrum length.
194:
195: For very low bare charges, the solution (\ref{eq:ff}) holds for all
196: distances with $Z_\eff=Z_\bare$, and one can consider that $Z_\back=Z_\bare$.
197: However, typical colloidal charges are such that
198: $Z_\bare \ell_B/a \gg 1$, a regime for which
199: counterions become strongly
200: associated with the colloid and the charge renormalization effects
201: \cite{Belloni98,PRL2002,JCP2002} can not be ignored.
202: The counterion
203: condensation strongly affects the electrostatic far-field
204: so that the large distance signature involves an effective charge
205: [$Z_\eff$ in Eq. (\ref{eq:ff})] which significantly differs from the bare one.
206: As a result of non-linear screening, one has $Z_\eff \ll Z_\bare$
207: whenever $Z_\bare \ell_B/a \gg 1$.
208:
209: At this point, the effective charge arising in (\ref{eq:ff}) is a function
210: of both the background and the bare charge, other parameters being fixed:
211: $Z_\eff=Z_\eff(Z_\back,Z_\bare)$. As far as a static description is
212: pursued, for sufficiently strongly charged colloids the
213: bare charge is an irrelevant quantity far enough from the tagged
214: colloid, and we demand that $Z_\back$ coincides with $Z_\eff$, which
215: best characterizes the background charge resulting from smearing
216: out the other colloids contribution. We therefore
217: enforce the self-consistency constraint
218: \be
219: Z_\back \,=\, Z_\eff(Z_\back,Z_\bare)
220: \ee
221: to compute the {\it a priori} unknown background charge.
222: As we shall see below, this condition is readily implemented numerically
223: and for a given $Z_\bare$, leads to a unique value for
224: $Z_\back=Z_\eff$. This value is density dependent, which is also
225: the case of the inverse screening length $\kappa$.
226: Indeed, $\varphi_\infty$ depends on $\rho$ \cite{rque}
227: through the electro-neutrality
228: condition
229: \be
230: Z_\back \, \rho \, e + \sum_i n_i^0 z_i e\, \exp\left(-\beta e z_i \varphi_\infty\right)=0
231: \label{eq:electroneutrality}
232: \ee
233: which translates into a $\rho$ dependence for $\kappa$.
234: Considering now two colloids in the weak overlap
235: approximation (i.e. not too close), the effective potential of interaction
236: will take a DLVO form \cite{Belloni,Levin} with effective
237: parameters $\kappa$ and $Z_\eff$.
238:
239: The procedure outlined here incorporates non linear screening
240: together with finite $\rho$ effects. It is best suited to describe
241: low density systems since the
242: colloid-colloid pair distribution function $g_{cc}$
243: is implicitly considered to be unity for all distances.
244: This reduction, which has non trivial consequences,
245: is certainly of little relevance for high density suspensions
246: for which the cell model is presumably a better approximation.
247:
248: Before illustrating the method, we briefly consider the pressure
249: in the system, that is given by the densities of micro-ions
250: far from the tagged colloid:
251: \be
252: \beta P \,=\, \sum_i n_i(\infty) = \sum_i n_i^0 \, \exp\left(-\beta e z_i \varphi_{\infty} \right).
253: \label{eq:P}
254: \ee
255: The colloidal contribution is explicitly discarded \cite{Jellium}. This is well justified
256: in the low salt limit, which is a regime of counterions dominance
257: provided that $Z_\bare \gg 1$, which is easily achieved in practice.
258:
259:
260: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
261: \section{The no salt limit}
262: \label{sec:nosalt}
263:
264: %%%%%%%%%%%%%%%%%%%%
265: \subsection{Spherical colloids}
266: The simplest situation to investigate is that of de-ionized suspensions (no salt).
267: For simplicity, we consider counterions as monovalent. From (\ref{eq:Poisson})
268: it follows that the dimensionless potential $\phi = \beta e \varphi$ obeys the equation
269: \be
270: \frac{d^2 \phi}{d \tilde r^2} \,+\, \frac{2}{\tilde r} \frac{d \phi}{d\tilde r} \,=\,
271: 3 \eta \, \frac{Z_\back \ell_B}{a} \, \left(e^{\phi} -1 \right)
272: \label{eq:sphnosalt}
273: \ee
274: where $a$ is again the radius of the tagged particle from which the
275: dimensionless distance $\tilde r=r/a$ is defined, and $\eta=4 \pi \rho a^3/3$ is the
276: volume fraction. The boundary conditions are
277: \begin{eqnarray}
278: \phi \to 0 &&\quad \hbox{for} \quad \tilde r \to \infty
279: \label{eq:bc1}\\
280: \frac{d \phi}{d\tilde r} = -\frac{Z_\bare \ell_B}{a} &&\quad \hbox{for} \quad \tilde r =1.
281: \label{eq:bc2}
282: \end{eqnarray}
283: In writing (\ref{eq:sphnosalt}), use has been made of the
284: (global) electro-neutrality constraint
285: $n_-^0 e \exp(\phi_\infty) = Z_\back \rho e$ with the choice $\phi_\infty=0$.
286: For all values of $Z_\back$, the far-field of $\phi$ is governed by
287: $\kappa$ such that
288: \be
289: (\kappa a)^2 \,=\, 3 \eta \frac{Z_\back \ell_B}{a}.
290: \label{eq:kappasphnosalt}
291: \ee
292:
293: The above system is solved following similar lines as in \cite{Langmuir2003}.
294: We summarize here the main steps.
295: In practice, equation (\ref{eq:sphnosalt}) is solved numerically for
296: a finite system $\tilde r \in [1,\widetilde R]$, where $\tilde R$ needs to be large enough
297: (that is $\kappa a \widetilde R \gg 1$ but note that $\kappa$ is not known initially
298: but follows once the background charge is known).
299: a) The first and important step is to
300: rephrase the boundary value problem at hand as an initial value problem
301: with boundary conditions $\phi'(\widetilde R)=0$ (to ensure electro-neutrality)
302: and varying $\phi(\widetilde R)$. The volume fraction $\eta$ is fixed and
303: the background charge $Z_\back$
304: first assigned a guess value, to be modified iteratively (see below).
305: If $\phi(\widetilde R)$ is small enough,
306: the system then admits a solution.
307: b) From this solution, one computes $\phi'(\tilde r=1)$
308: to know the corresponding bare charge.
309: c) Changing $\phi(\widetilde R)$ \cite{rque2}, the targeted value
310: $\phi'(\tilde r=1)=-Z_\bare \ell_B/a$ is readily found by iteration.
311: d) The screening quantity $\kappa$
312: is subsequently computed from (\ref{eq:kappasphnosalt}) and the
313: effective charge associated to the particular couple $(Z_\back,Z_\bare)$
314: is deduced from the large $\tilde r$ behavior of $\phi$ [e.g
315: one needs to observe a clear-cut plateau for
316: $[\phi(\tilde r)-\phi(\widetilde R)] e^{\kappa a \tilde r} \tilde r $ plotted as a function of $\tilde r$
317: in the range $1 \ll \tilde r <\widetilde R$]. The first iteration ends here,
318: and the procedure is repeated with the $Z_\eff$ obtained as the next
319: trial value for $Z_\back$.
320: Alternatively, one may sample several trial values for $Z_\back$
321: and plot $Z_\eff$ versus $Z_\back$. As may be observed in Fig.
322: \ref{fig:Zsph1} where such a plot is displayed,
323: the dependence of $Z_\eff$ on $Z_\bare$
324: is very mild, which means that convergence towards
325: $Z_\back=Z_\eff$ is achieved in a few steps.
326: In the (artificial) limit where $Z_\back \to 0$, the problem at hand
327: reduces to an unscreened one (governed by Laplace equation)
328: with solution $\phi = Z_\bare \ell_B/r$: there is therefore no
329: renormalization of effective charge so that $Z_\eff \to Z_\bare$
330: (see Fig. \ref{fig:Zsph1}). {The inset shows how the
331: self-consistent background charge is determined, the other points
332: being unphysical.}
333:
334: {In the limit of a diverging bare charge, the procedure is well
335: behaved and yields a finite self-consistent effective charge. From the
336: previous discussion, we expect $Z_\eff$ to diverge at small $Z_\back$, which
337: is indeed the case (not shown).}
338:
339: {Once the physical solution to the problem has been located
340: (inset of Fig. \ref{fig:Zsph1}),
341: various quantities such as the pressure
342: may be computed. In the remainder, we will use the terms
343: ``effective'' and ``background'' charges to refer to the self-consistent
344: solution as obtained in Fig. \ref{fig:Zsph1}:
345: $Z_\eff=Z_\back$ is therefore
346: a function of $Z_\bare$ and volume fraction (possibly also salt concentration,
347: see section \ref{sec:salt}). For a particular density, this function
348: is shown in Fig. \ref{fig:Zsph3}. After the initial linear regime,
349: where no renormalization takes place, the effective charge slowly reaches
350: a saturation plateau as $Z_\bare \to \infty$. For this specific density
351: ($\eta=10^{-2}$) the effective charge saturates at $Z_\eff \ell_B/a \simeq 6.6$.}
352: The saturation phenomenon observed here is strongly reminiscent of
353: that observed in the classical Poisson-Boltzmann approach
354: (either in a cell, or in an infinite medium \cite{Alexander,JCP2002,Tellezsat}).
355: To assess quantitatively the possible difference with the results
356: obtained within the cell model, we compare in Fig.
357: \ref{fig:sphcompcellns} $Z_\eff$ derived in the cell
358: \cite{Alexander,Langmuir2003} to its {renormalized} jellium counterpart.
359: Both charges differ by a notable amount for $\eta >10^{-3}$
360: while the agreement at very low density is meaningless,
361: and follows from a divergence of the saturation effective charge
362: in both models: non-linear effects disappear when
363: $\eta \to 0$,
364: so that $Z_\eff \to Z_\bare$. This is a peculiarity of systems with colloidal spheres and counterions only, and it turns out that the behavior of charged cylinders is quite
365: different, see below.
366:
367:
368: Under the de-ionized conditions studied here, the pressure
369: takes the simple form $\beta P = Z_\eff \rho$, whereas
370: the corresponding expression in the cell model is less explicit
371: and does not directly involve the effective charge.
372: Remarkably, although there is a significant difference
373: between the effective charges calculated within {the Poisson-Boltzmann cell
374: model (PBC) and the renormalized jellium},
375: the pressures calculated using the two models are identical for $\eta<0.1$,
376: Fig. \ref{fig:Zsphpressns}. A similar agreement is found
377: at saturation \cite{Jellium}. We add here that a comparison
378: between the {renormalized} jellium equation of state and the ``exact'' results of the primitive
379: model obtained using the Monte Carlo
380: computations has been reported in \cite{Jellium},
381: with excellent agreement (the corresponding density range
382: is quite low such that PBC and {renormalized} jellium predictions agree).
383:
384: Before concluding this section, we emphasize that one must
385: carefully check that the results obtained do not depend on the particular
386: value chosen for the cutoff $\widetilde R$, e.g. by repeating
387: the analysis with an increased cutoff.
388:
389:
390:
391: %%%%%%%%%%%%%%%%%%%%%
392: \subsection{Cylindrical colloids}
393:
394: Consider a nematic phase of parallel infinite rods {($L \to \infty$)} with
395: line charge $\lambda_\bare$ (therefore no
396: positional order in the plane perpendicular to the main axis).
397: We may repeat the previous approach, tagging a given rod of radius $a$
398: and model the effects of the other rods by an homogeneous
399: background, with line charge $\lambda_\back$.
400: While the definition of $\kappa$ is unaffected compared to the
401: spherical case, the far-field potential now takes the form:
402: \be
403: \phi(\tilde{r}) \,=\, \phi_{\infty} \,+\, 2 \lambda_{\eff} \ell_{B} \,
404: \frac{K_{0}(\kappa a \tilde{r})}{\kappa a K_{1}(\kappa a)},
405: \ee
406: where $K_0$ (resp. $K_1$) denotes the zeroth (resp. first)
407: order modified Bessel function of the second kind.
408: In the spirit of the consistency requirement of section \ref{sec:model},
409: we impose $\lambda_\back=\lambda_\eff$ where again,
410: $\lambda_\eff$ follows from the large distance behavior of
411: the solution to Poisson's equation with background charge
412: $\lambda_\back$:
413: \be
414: \frac{d^2 \phi}{d \tilde r^2} \,+\, \frac{1}{\tilde r} \frac{d \phi}{d\tilde r} \,=\,
415: 4 \eta \, \lambda_\back \ell_B\, \left(e^{\phi} -1 \right).
416: \label{eq:cylnosalt}
417: \ee
418: Here the volume fraction is $\eta = \pi a^2\, n_{_{\text{2D}}}$ where
419: $n_{_{\text{2D}}}$ is the mean surface density of rods (in the plane perpendicular
420: to their axis). The boundary conditions are the same as (\ref{eq:bc1}) and (\ref{eq:bc2}),
421: and the numerical method identical to that used in the spherical case.
422:
423:
424:
425: {The effective charges calculated using the cell and the renormalized
426: jellium models are compared in Fig. \ref{fig:lambanosat} for
427: $\lambda_{bare}\ell_B=1$. The inset corresponds to the
428: saturation regime where $\lambda_\bare$ is very large ($\lambda_\bare \to \infty$). We observe a
429: substantial disagreement between the two effective charges.}
430: On the other hand, in the small bare charge regime where
431: non-linear effects are not at work, both quantities coincide (not shown),
432: which is a signature of
433: whenever non-linear effects come into
434: play (i.e. outside the small bare charge linear regime, which is the
435: case for both figures).
436: Beyond these differences, Manning-Oosawa
437: condensation \cite{Manning,Manning2006}, which is a key feature of 2D
438: electrostatics, is shared by both PBC and {renormalized} jellium models. As the colloid
439: density is decreased ($\eta \to 0^+$), the effective charge becomes independent
440: of the bare one, provided $\lambda_\bare$ exceeds the critical threshold
441: $1/\ell_B$. This feature is illustrated in Fig. \ref{fig:Manning}.
442: At the saturation plateau and again for $\eta \to 0^+$,
443: one has $\lambda_\eff \ell_B\simeq 0.47$,
444: a value that will be refined below.
445:
446: To be more quantitative, it is furthermore natural to compare
447: the corresponding functional forms of effective charges
448: versus bare ones, and versus density in both PBC (where it can be computed
449: analytically) and {renormalized} jellium models, where this information is accessed numerically.
450: To this end, we reconsider the
451: analytical results obtained in \cite{JCP2002, Langmuir2003}
452: where the effective charge in the cell model following Alexander {\it et al.}
453: prescription \cite{Alexander} reads:
454: \begin{eqnarray}
455: \lambda_{\eff} \ell_{B}&=&\frac{1}{2}K_{PB}^{2}aR_{ws}\{I_{1}(K_{PB}R_{ws})K_{1}(K_{PB}a)\nonumber \\
456: & & {} -I_{1}(K_{PB}a)K_{1}(K_{PB}R_{ws})\},
457: \label{eq:PBCMann}
458: \end{eqnarray}
459: with standard notation for the Bessel functions.
460: Here $R_{ws}\equiv a \eta^{-1/2}$ is the radius of the cell and $K_{PB}$
461: is the inverse screening length related to the micro-ionic density at the cell
462: boundary \cite{JCP2002}, which can be computed explicitly
463: from the analytical solution of \cite{Alfrey}. After some algebra, we find,
464: to leading order in density that when $\lambda_\bare > 1/\ell_B$
465: %---------------------------------
466: \begin{equation}
467: \lambda_{\eff}^\sat \ell_{B} \,\stackrel{\eta \to 0^+}{\sim} \, \frac{\sqrt{2}}{2}I_{1}(\sqrt{2}) \, +\, \pi^2
468: \frac{I_{0}(\sqrt{2})+\sqrt{2}I_{1}(\sqrt{2})+I_{2}(\sqrt{2})}{(2\xi-\log(\eta))^2},
469: \label{eq:form}
470: \end{equation}
471: %---------------------------------
472: where $\xi = \lambda_\bare/(\lambda_\bare -1/\ell_B )$.
473: We note that the leading term $ \sqrt{2}I_{1}(\sqrt{2})/2 \simeq 0.63$
474: differs from the value found in the {renormalized} jellium ($\simeq 0.47$, see Fig.
475: \ref{fig:Manning}).
476: Moreover, Eq. (\ref{eq:form}) also
477: suggests a fitting form to describe the saturation plateau in the low density
478: regime of the {renormalized} jellium model:
479: \be
480: \lambda_\eff^\sat \ell_B \,\stackrel{\eta \to 0^+}{\sim} \, A + \frac{B}{(C-\log \eta)^2}.
481: \ee
482: {The values of $A$, $B$ and $C$ can be obtained using a numerical fit. We find
483: that in the saturation limit $A\simeq 0.471$, $B\simeq16.87$ and $C\simeq
484: 0.843$ give an excellent agreement with the numerical data. We have also checked that
485: an equally good agreement is found at lower bare charges,
486: such as $\lambda_\bare \ell_B=4$, but with different values of $A$, $B$ and $C$.}
487: We conclude here that both models are described by the same
488: limiting law for low densities, at least beyond the condensation threshold.
489:
490:
491:
492: It is of interest to resolve the condensate structure once the counterion
493: condensation has set in. A useful measure of the condensate thickness
494: is provided by the so-called Manning radius $R_M$ \cite{Gueron}
495: that has been recently
496: worked out in the infinite dilution limit and for low salt content
497: \cite{OS,Manning2006}: in practice, the integrated charge
498: per unit length
499: $q(r)$ around
500: a rod has an inflection point
501: at $r=R_M$, when plotted as a function of $\log r$. This is exactly
502: the point where $q(R_M)\ell_B/e=1$.
503: We expect a similar
504: behavior for {the renormalized} jellium, given that in the vicinity of highly charged rods,
505: the (largely dominant) counterion
506: distribution should not be sensitive to the difference between
507: a uniform background as in the {renormalized} jellium model, and coions
508: as in the situation worked out in \cite{Manning2006}. The lower inset of
509: Figure \ref{fig:man1} shows that this is indeed the case.
510: In addition, from the analytical expressions derived
511: in \cite{Manning2006} and the fact that the relevant screening parameter
512: reads here $(\kappa a)^2 = 4 \eta \lambda_\eff \ell_B$, we expect
513: the scaling $\kappa R_M \propto (\kappa a)^{1/2}$, more precisely
514: \be
515: R_M \, \stackrel{\eta \to 0^+}{\propto} \, a\, \eta^{-1/4}
516: \exp\left(-\frac{1}{2(\lambda_\bare \ell_B-1)}
517: \right).
518: \label{eq:mann}
519: \ee
520: The dependence of $R_M$ on both density and bare charge embodied
521: in Eq. (\ref{eq:mann}) is fully supported by the numerical data, see
522: Fig. \ref{fig:man1}.
523:
524:
525: Finally, and much like for spherical colloids,
526: there is a good agreement between the osmotic pressure calculated using the cell model
527: and the renormalized jellium approximation, in spite of the different effective charges,
528: see Fig. \ref{press}. Discrepancies are observed only for volume fractions
529: $\eta>0.06$ and the agreement seems to be better at high charges.
530:
531:
532: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
533: \section{Effects of added salt}
534: \label{sec:salt}
535:
536: In this section, we consider systems dialyzed against an electrolyte reservoir
537: with the monovalent salt concentration $c_s$. The corresponding screening
538: parameter is
539: $\kappa_{\res}^2=8\pi\ell_{B}c_{s}$.
540: It is convenient to choose the reference potential so that micro-ionic
541: densities are $n_\pm({\bm r}) = c_s \exp[\mp \phi({\bm r})]$,
542: where the counterions are assumed to be monovalent.
543: Using Eq. (\ref{eq:electroneutrality}), the potential at infinity becomes
544: \be
545: \phi_\infty \,=\, \hbox{arcsinh}\left(\frac{Z_\back \rho}{2 c_s} \right).
546: \ee
547:
548:
549: It is important to keep in mind that
550: $n_{\pm}({\bf r})$ are not the physical microion
551: densities, but are the effective (renormalized) quantities satisfing
552: \be
553: \int d {\bm r} [n_+({\bm r}) -n_-({\bm r}) +Z_\eff
554: \rho]= - Z_\bare.
555: \ee
556: {Since the renormalization does not affect coions, their concentration inside the jellium with
557: one colloid fixed at $\bm r=0$ is}
558: \be
559: C_+ = \frac{1}{V}\int d {\bm r} n_+({\bm r}),
560: \ee
561: where it is understood that $V$ denotes the measure of a large
562: volume centered at $\bm r=0$.
563: The concentration of counterions, $C_-$, then
564: follows from the
565: overall charge neutrality inside suspension,
566: $C_- = C_+ +Z_\bare \rho $.
567:
568:
569:
570: Far from colloid, $n_+({\bm r})$ saturates at the bulk value
571: $\widetilde n_+$, so that in the
572: thermodynamic limit ($V \rightarrow\infty$)
573: \be
574: C_+ =\widetilde n_{+}.
575: \label{eq1}
576: \ee
577: Similarly, for $V\to \infty$
578: \be
579: \frac{1}{V}\int d {\bm r} [n_+({\bm r}) -n_-({\bm r}) +Z_\eff
580: \rho]= \frac{- Z_\bare}{V}\rightarrow 0 \;,
581: \ee
582: which means that
583: \be
584: C_+- \frac{1}{V}\int d {\bm r}n_-({\bm r})+Z_\eff\rho=0 \;.
585: \label{eq3}
586: \ee
587: The charge neutrality allows us to
588: rewrite Eq.~(\ref{eq3}) as
589: \be
590: \frac{1}{V} \int d {\bm r} n_-({\bm r})=C_--(Z_\bare-Z_\eff)\rho \;.
591: \label{counter}
592: \ee
593: Eq.~(\ref{counter}) provides a suggestive
594: interpretation of $n_-({\bm r})$
595: as the local density of free (uncondensed) counterions.
596: Far from colloid, $n_-({\bm r})$ saturates at its bulk value
597: $\widetilde n_-$, and in
598: thermodynamic limit Eq.~(\ref{counter}) reduces to
599: \be
600: C_-= \widetilde n_-+(Z_\bare-Z_\eff)\rho.
601: \label{eq2}
602: \ee
603: Eqs.~(\ref{eq1}) and (\ref{eq2}) allow us to calculate the ionic content
604: inside a suspension dialized against a salt reservoir. This
605: is particularly useful when comparing the results of the
606: {renormalized} jellium model, which is grand canonical in electrolyte,
607: with the Monte Carlo simulations, which are usually performed in a
608: canonical ensemble.
609: Knowledge of the assymptotic potential allows us to obtain the
610: concentrations of coions and {\it free}
611: counterions inside the suspension,
612: \be
613: \widetilde n_{\pm} \,=\, c_s \exp(\mp \phi_\infty).
614: \ee
615: These are precisely the densities that
616: govern screening within {the renormalized} jellium,
617: $\kappa^2 = 4 \pi \ell_B (\widetilde n_++\widetilde n_-)$.
618:
619:
620:
621:
622: %%%%%%%%%%%%%%%%%%%%
623: \subsection{Spherical colloids}
624:
625: In spherical geometry, Poisson equation (\ref{eq:Poisson}) now
626: takes the form
627: \be
628: \frac{d^2 \phi}{d \tilde r^2} \,+\, \frac{2}{\tilde r} \frac{d \phi}{d\tilde r} \,=\,
629: (\kappa_\res a)^2 \sinh \phi -
630: 3 \eta \, \frac{Z_\back \ell_B}{a} .
631: \label{eq:PSph}
632: \ee
633: We again solve it numerically as a boundary value problem in a (large enough)
634: finite cell with vanishing $\phi'$ at the boundary, increasing gradually
635: the boundary potential from the value
636: \be
637: \phi_\infty \,=\,
638: \hbox{arcsinh}\left[\frac{ 3 \eta \, Z_\back \ell_B/a}{(\kappa_\res a)^2}
639: \right],
640: \ee
641: which corresponds to a vanishing bare charge.
642:
643:
644: Linearizing Eq. (\ref{eq:PSph}) around $\phi_\infty$, it can be seen that at
645: large distances the potential takes the form of Eq. (\ref{eq:ff}), with a
646: screening constant $\kappa$ given by
647: \be
648: (\kappa a)^4 \,=\, (\kappa_\res a)^4 + \left(
649: \frac{3 \eta Z_\back \ell_B}{a} \right)^2.
650: \ee
651: For highly charged colloids and typical salt conditions,
652: the corresponding density dependence is shown
653: in Fig. \ref{fig:kapsphsalt}, while the effective charge (deduced from the condition
654: $Z_\eff = Z_\back$) is displayed in Fig. \ref{fig:Zsphsalt}.
655: When $\eta \to 0$, both quantities coincide with the infinite dilution limit
656: of the traditional
657: PB theory, as they should. The increase of $\kappa$
658: with the density of colloids reflects the increasing importance of counterion
659: screening. The effective charge shows a non-monotonous behaviour with respect to
660: density.
661:
662:
663: To compute the osmotic pressure, we subtract the reservoir pressure
664: ($2 c_s kT$) from the expression (\ref{eq:P}). Moreover,
665: it should be remembered that such a relation only provides the
666: ionic contribution to the pressure. In the presence of salt and at low
667: colloidal density
668: this contribution becomes smaller than the colloidal one.
669: The vanishing of the microion contribution to pressure is exponential in the cell model, while it is
670: algebraic for the jellium. Both models should then strongly disagree in the
671: low density limit.
672: To mimic the colloidal contribution, we add the ideal gas term $\rho k T$ to
673: (\ref{eq:P}), so that
674: the resulting osmotic pressure reads
675: \begin{equation}
676: \beta \Pi =\rho+\sqrt{Z_{\eff}^2\rho^2+4c_{s}^2}-2c_{s}.
677: \label{pressalt2}
678: \end{equation}
679: In the no salt case, addition of the ideal term is irrelevant since it is
680: always much smaller than
681: the micro-ionic one, provided that $Z_\eff$ is large
682: enough (this is the case for highly or even weakly charged colloids
683: since $a \gg \ell_B)$. Moreover,
684: addition of the ideal gas term breaks the scaling form valid in the
685: no salt case where $a^2 \ell_B \beta P $ only depends on
686: $\eta$ and reduced charge $Z_\bare \ell_B/a$. We therefore show the
687: osmotic pressure in Fig. \ref{fig:pressphsel} for two values
688: of colloid radius, within both the PBC and the {renormalized} jellium
689: frameworks. Apart from the expected deviations at small densities,
690: one observes compatible values at higher $\eta$.
691:
692:
693: There exist relatively little simulational data for the primitive model with salt,
694: where the bare Coulomb interactions between all charged species
695: --colloids and micro-ions--
696: are taken into account (with still an implicit solvent). A reference
697: equation of state with salt is provided in
698: \cite{Lobaskin},
699: with the simplification of a Wigner-Seitz cell, but explicit micro-ions.
700: The simulations were performed in canonical ensemble with
701: fixed salt content.
702: The amount of added salt is characterized by a ratio
703: of the overall added cation charge to the overall macroion charge,
704: $\beta_{L}=C_{+}/(Z_{\bare} \,\rho).$
705: We compute the densities $C_{\pm}$ corresponding to a given salt
706: content as discussed in section \ref{sec:salt}. In Fig. \ref{fig:Lobaskin}
707: the osmotic pressure $\beta P/\rho_{t}$ is plotted as a function of $\beta_{L}$, where $\rho_{t}$
708: is the total density of ionic species. As in the case of salt-free
709: suspensions the pressures calculated using the PBC and the {renormalized} jellium are in good agreement.
710:
711:
712: %%%%%%%%%%%%%%%%%%%%
713: \subsection{Rod-like colloids}
714:
715:
716: For completeness, we briefly report here results for cylindrical geometry.
717: Unlike salt-free case where $\lambda_\eff$ is a monotonic function
718: of density, a minimum appears in the {renormalized} jellium curve shown in Fig. \ref{fig:k0k1}.
719: The agreement between PBC and {renormalized} jellium at low $\eta$ signals
720: the region where the system is salt dominated (the colloid density
721: is too low and, consequently, counterions do not participate in screening).
722: Conversely, the inset indicates the density range where counterions do
723: dominate : for $\eta > 10^{-1}$, the results become independent of the reservoir
724: ionic strength and coincide with those obtained in the no salt limit.
725:
726: {Finally, the pressure (\ref{pressalt2}) for cylindrical colloids is given by
727: \begin{equation}
728: 4\pi\ell_B a^2\beta \Pi =4\eta\frac{\ell_B}{L}+\sqrt{(4\lambda_{eff}\ell_B\eta)^2+(\kappa_{res}a)^4}-(\kappa_{res}a)^2.
729: \label{pressaltsgi}
730: \end{equation}
731: Note that for infinite polyions ($L\to \infty$), the first term on the right
732: hand side of Eq. (\ref{pressaltsgi}) vanishes. In Fig. \ref{presscylsel2} we
733: plot the equation of state for polyions of $\lambda_\bare\ell_B=2$. One should note a strong
734: disagreement between the equation of state obtained using the renormalized
735: jellium model and the PBC theory. In the case of cylindrical polyions the
736: disagreement is exacerbated by the fact that the ideal gas contribution to
737: the equation of state, Eq. (\ref{pressaltsgi}), vanishes in the limit of $L\to
738: \infty$ considered in this work. For small $\eta$, the behavior predicted by
739: the renormalized jellium model is more realistic than that of the PBC.}
740:
741:
742:
743: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
744: \section{Conclusion}
745: \label{sec:concl}
746:
747: %To do : make a plot showing what happens if the background is not renormalized.
748: %Take say spheres, no salt, Z_bare lb/a=6, eta=3.10^-2 and solve with Z_back=Z_bare.
749: %Superimpose the effective charges obtained to those when Z_back=Z_eff.
750: %Also compare the pressures. Another possibility is that you take the parameters
751: %of Fig. 2 in the PRE paper with Yan, and you compute the pressure
752: %with Z_back=Z_bare. In short, it would be nice to find parameters where
753: %there is a noticeable difference between the naive and the renormalized approach.
754:
755:
756: Starting from a mean-field description in which a dispersion of $N$
757: spherical or rod-like polyions is treated using a $N$-body
758: Poisson-Boltzmann theory \cite{rque3}, we have introduced the simplification of
759: a homogeneous background to include the contribution of other colloids to the
760: static field created by a tagged colloid. The charge of this background is
761: consistently renormalized to coincide with the effective charge
762: governing the far-field potential. This results in a non-trivial density dependence
763: of the effective colloidal charge, which directly enters into the equation of state
764: through a simple analytical expression. The good agreement observed between the
765: pressure calculated using the renormalized jellium and the Monte Carlo
766: simulations confirms the relevance of the {renormalized}
767: jellium model for theoretical and experimental purposes and provides an
768: alternative to the Poisson-Boltzmann cell approach.
769: Furthermore, we note that the effective
770: charge calculated using the renormalized jellium model should be more relevant
771: for the study of the effective interaction between the colloids than its
772: Poisson-Boltzmann cell (PBC)
773: counterpart. This is particularly the case since
774: at finite colloidal density, the DLVO potential arises naturaly within the
775: jellium formalism, while it has to be introduced extraneously within the PBC.
776: In this work, we have left
777: untouched the question whether the pair potential calculated using jellium
778: is a potential of mean-force
779: or an effective pair potential (following the terminology of \cite{Belloni}).
780: Further work is required to answer this question.
781:
782: In a cylindrical geometry, the present approach implicitly subsumes an
783: alignment between infinite rods --which is also a prerequisite for the analysis of \cite{Alfrey}--
784: but contrary to the crystalline structure underlying the introduction of
785: the cell model, we consider here systems with no positional order
786: for the rods. As for spherical colloids the pressure in both approaches
787: is in good agreement up to relatively high densities, whereas the
788: effective charges differ significantly. We have also shown that the scenario for counterion
789: condensation is similar to that of the cell picture.
790:
791: Our approach, which is best suited to describe systems with low macro-ions
792: densities, may be easily extended to the case of asymmetric electrolytes.
793: One interesting aspect of {the renormalized} jellium is that the description of colloidal mixtures
794: (macro-ions with different sizes and charges) appears to be as straightforward
795: as for the mono-disperse systems reported here.
796: This is an important difference with the cell approach, which cannot be
797: easily extended to such systems.
798:
799: Among the possible refinements, it is possible to
800: consider an in-homogeneous jellium with, again, renormalized charge.
801: This should allow to extend the relevant range of densities where
802: the model holds. Another interesting extension deals with
803: the derivation of electro-kinetic properties. Work along these lines
804: is in progress.
805:
806:
807: \section{Acknowledgments} The financial support of Capes/Cofecub is gratefully acknowledged.
808: We would also like to thank M. Deserno, J. Dobnikar, H.H. von Gr\"unberg,
809: R. Casta\~neda Priego and L. Belloni for useful discussions.
810:
811: \newpage
812:
813: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
814: \begin{thebibliography}{99}
815:
816:
817: \bibitem{Belloni}
818: L. Belloni,
819: J. Phys. Condens. Matter {\bf 12}, R549 (2000).
820:
821: \bibitem{Alfrey}
822: T. Alfrey Jr, P.W. Berg and H. Morawetz,
823: J Polym. Sci. {\bf 7}, 543 (1951);
824: R.M. Fuoss, A. Katchalsky and S. Lifson,
825: P. Natl. Acad. Sci. USA {\bf 37}, 579 (1951).
826:
827: \bibitem{Alexander}
828: S. Alexander, P.M. Chaikin, P. Grant, G.J. Morales, P. Pincus, and D. Hone,
829: J. Chem. Phys. {\bf 80}, 5776 (1984).
830:
831: \bibitem{Hansenrevue}
832: J.-P. Hansen and H. L\"{o}wen,
833: Annu. Rev. Phys. Chem. {\bf 51}, 209 (2000).
834:
835: \bibitem{Deserno}
836: M. Deserno and C. Holm, in {\em Proceedings of NATO Advanced Study Institute
837: on Electrostatic Effects in Soft Matter and Biophysics}, edited by C. Holm,
838: P. Kekicheff, and R. Podgornik (Kluwer, Drodrecht, 2001), p.\ 27.
839:
840: \bibitem{Kuwabara}
841: S. Kuwabara,
842: J. Phys. Soc. Japan {\bf 14}, 527 (1959).
843:
844: \bibitem{Ohshima}
845: H. Ohshima,
846: Coll. Surf. B {\bf 38}, 139 (2004).
847:
848: \bibitem{LevinJPCM}
849: Y. Levin, E. Trizac and L. Bocquet,
850: J. Phys.: Condens. Matt. {\bf 15}, S3523 (2003).
851:
852: \bibitem{LevinL} Y. Levin, M.C. Barbosa, M.N. Tamashiro
853: Europhys. Lett. {\bf 41}, 123 (1998);
854: A. Diehl, M. C. Barbosa, and Y. Levin,
855: Europhys. Lett. {\bf 53}, 86 (2001)
856:
857: \bibitem{Beresford}
858: B. Beresford-Smith, D.Y. Chan and D.J. Mitchell,
859: J. Colloid Interface Sci. {\bf 105}, 216 (1984);
860:
861: \bibitem{Jellium}
862: E. Trizac and Y. Levin,
863: Phys. Rev. E {\bf 69}, 031403 (2004).
864:
865: \bibitem{Haro}
866: C. Haro-P\'{e}rez, M. Quesada-P\'{e}rez, J. Callejas-Fern\'{a}ndez, R. Sabate, J. Estelrich,
867: R. Hidalgo-\'{A}lvarez,
868: Coll. Surfaces A {\bf 270}, 352 (2005).
869:
870: \bibitem{HaroJFCM06}
871: C. Haro-P\'{e}rez, M. Quesada-P\'{e}rez, J. Callejas-Fern\'{a}ndez,
872: P. Schurtenberger, R. Hidalgo-\'{A}lvarez,
873: J. Phys.: Condens. Matter {\bf 18}, L363 (2006).
874:
875: \bibitem{CastanedaCM}
876: R. Casta\~{n}eda-Priego, L. F. Rojas-Ochoa, V. Lobaskin, J. C. Mixteco-S\'{a}nchez,
877: Cond-mat/0608163.
878:
879:
880: \bibitem{Manning}
881: G.S. Manning,
882: J.~Chem.~Phys.~\textbf{51}, 924 (1969) ; \textbf{51}, 934 (1969);
883: F.~Oosawa,
884: Polyelectrolytes, Dekker, New York (1971).
885:
886: \bibitem{Manning2006}
887: E. Trizac and G. T\'ellez,
888: Phys. Rev. Lett. {\bf 96}, 038302 (2006).
889:
890: \bibitem{Levin}
891: Y. Levin,
892: Rep. Prog. Phys. {\bf 65}, 1577 (2002).
893:
894: \bibitem{Fushiki}
895: M. Fushiki, J. Chem. Phys. {\bf 97}, 6700 (1992);
896: H. L\"owen, J.P. Hansen, and P.A. Madden, J. Chem. Phys. {\bf 98}, 3275
897: (1993).
898:
899: \bibitem{Graz}
900: J. Dobnikar, Y.Chen, R. Rzehak and H.~H.~von~Gr\"unberg,
901: J. Chem. Phys. {\bf 119},4971 (2003);
902: J. Dobnikar, D. Halo\v{z}an, M. Brumen, H.~H.~von~Gr\"unberg
903: and R. Rzehak,
904: Comput. Phys. Commun. {\bf 159}, 73 (2004).
905:
906: \bibitem{Belloni98}
907: L. Belloni,
908: Colloid Surf. A {\bf 140}, 227 (1998).
909:
910: \bibitem{PRL2002}
911: E. Trizac, L. Bocquet and M. Aubouy,
912: Phys. Rev. Lett. {\bf 89}, 248301 (2002).
913:
914: \bibitem{JCP2002}
915: L. Bocquet, E. Trizac, and M. Aubouy,
916: J. Chem. Phys. {\bf 117}, 8138 (2002).
917:
918: \bibitem{rque}
919: More precisely, it is the product $n_i^0 \exp(-\beta z_i e \varphi_\infty)$
920: which is physically relevant. One may always chose
921: $\varphi_\infty=0$, modulo a proper redefinition of the pre-factors
922: $n_i^0$. This convention is convenient for salt free systems,
923: but has not been adopted in practice in the presence of added
924: salt. In addition, Eq. (\ref{eq:electroneutrality}) appears to be an effective
925: electro-neutrality condition, which does not coincide with the physical one.
926: This aspect is discussed at the beginning of section \ref{sec:salt}.
927:
928: \bibitem{Langmuir2003}
929: E. Trizac, M. Aubouy, L. Bocquet, and H.H. von Gr\"unberg,
930: Langmuir {\bf 19}, 4027 (2003).
931:
932: \bibitem{rque2}
933: We emphasize that not all trial values of $\phi(\widetilde R)$ lead to
934: a solution. For a given $\widetilde R$,
935: there indeed exists a critical threshold $\phi^\sat(\widetilde R)$
936: beyond which no solution can be found. For small $\phi(\widetilde R)$
937: [i.e. $\phi(\widetilde R) \ll \phi^\sat(\widetilde R)$],
938: there is a linear relationship between $Z_\bare$ and $\phi(\widetilde R)$,
939: but when $\phi(\widetilde R)$ approaches $\phi^\sat(\widetilde R)$ from below,
940: the bare charge diverges. This is a consequence of the phenomenon
941: of effective charge saturation \cite{JCP2002,Tellezsat} that is ubiquitous
942: in mean-field treatments.
943:
944: \bibitem{Tellezsat}
945: G. T\'ellez and E. Trizac,
946: Phys. Rev. E {\bf 68}, 061401, 2003.
947:
948: \bibitem{Gueron}
949: M. Gueron and G. Weisbuch,
950: Biopolymers {\bf 19}, 353 (1980).
951:
952: \bibitem{OS}
953: B. O'Shaughnessy and Q. Yang,
954: Phys. Rev. Lett. {\bf 94}, 048302 (2005).
955:
956: \bibitem{rque3}
957: This mean-field approach discards correlations between microions,
958: that become prevalent at large electrostatic couplings
959: (see \cite{Grosberg}). In a solvent like water
960: at room temperature, PB theory nevertheless provides a good description
961: of monovalent ion systems, see e.g. \cite{DesernoMay} or
962: \cite{Levin}.
963:
964: \bibitem{Grosberg}
965: A.Y. Grosberg, T.T. Nguyen and B.I. Shklovskii,
966: Rev. Mod. Phys. {\bf 74}, 329 (2002).
967:
968: \bibitem{DesernoMay}
969: M. Deserno, C. Holm and S. May,
970: Macromolecules {\bf 33}, 199 (2000).
971:
972: \bibitem{Lobaskin}
973: V. Lobaskin and K. Qamhieh,
974: J. Phys. Chem. B {\bf 107}, 8022 (2003).
975:
976:
977:
978: \end{thebibliography}
979:
980:
981: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
982:
983:
984: \newpage
985: \section{Captions}
986:
987: Fig. (\ref{fig:Zsph1}): The effective charge as a function of
988: background charge for $\eta=10^{-2}$ and $Z_{\bare}\ell_{B}/a=4$
989: (spherical colloids, no salt). The physical
990: solution $Z_\eff=Z_\back$
991: to the problem is the point of intersection with the first
992: bisectrix (see inset, where a magnification of the relevant part
993: of the main graph is shown).\\
994:
995:
996:
997: %Fig. (\ref{fig:Zsph2}): Same as Fig. \ref{fig:Zsph1} in the limit of a highly
998: %charged colloid where $Z_\bare \to \infty$
999: %($\eta=10^{-2}$, spherical geometry, no salt).
1000: %As in Fig. \ref{fig:Zsph1}, the inset shows a zoom
1001: %around the self-consistent solution.\\
1002:
1003:
1004:
1005: Fig. (\ref{fig:Zsph3}): The effective charge (or equivalently, the background charge)
1006: as a function of
1007: bare charge for $\eta=10^{-2}$ (spherical colloids, no salt).
1008: The line has slope unity to emphasize the initial
1009: ``Debye-H\"uckel'' regime.\\
1010:
1011:
1012:
1013: Fig. (\ref{fig:sphcompcellns}): Comparison between the effective charges
1014: within the Poisson-Boltzmann cell (PBC) and the renormalized jellium model,
1015: as a function of the volume fraction. Here $Z_{\bare} \ell_{B}/a=6$
1016: (spherical colloids, no salt).\\
1017:
1018:
1019:
1020: Fig. (\ref{fig:Zsphpressns}): Pressure as a function of volume fraction within
1021: the cell and the renormalized jellium model,
1022: on a log-log scale. Here, $Z_{\bare}l_{B}/a=6$ (spherical colloids, no salt).
1023: The inset shows the same data on a linear scale.\\
1024:
1025:
1026:
1027: %Fig. (\ref{fig:lambasat}): The effective charge as a function of
1028: %volume fraction within the Poisson-Boltzmann cell (PBC) and the jellium models
1029: %(rod-like colloids, no salt) in saturation regime $\lambda_\bare \to \infty$.\\
1030:
1031:
1032:
1033: Fig. (\ref{fig:lambanosat}): Effective charge as a function of
1034: volume fraction within the PBC and the renormalized
1035: jellium model, for $\lambda_{\bare}\ell_B=1$. The inset corresponds to the saturation
1036: regime where $\lambda_{bare}\to\infty$ (rod-like colloids, no salt).\\
1037:
1038:
1039:
1040: Fig. (\ref{fig:Manning}): The effective charge as a function of the bare charge for
1041: different values of volume fractions (renormalized jellium model
1042: for charged rods). The inset shows a magnification of the
1043: main graph in the low charge
1044: regime. The present scenario is exactly that of the Manning-Oosawa
1045: counterion condensation occurring in the cell model.\\
1046:
1047:
1048:
1049: %Fig. (\ref{chargeadj}): The effective charge as a function of
1050: %volume fraction for the PBC (circles) and renormalized jellium (pluses)
1051: %models (rod geometry, no salt) in saturation regime $\lambda_\bare \to
1052: %\infty$. For PBC, the triplet $(A,B,C)$ is taken from Eq.
1053: %(\ref{eq:form}), while it is fitted in the jellium case, with values
1054: %$A\simeq 0.471$, $B\simeq16.87$ and $C\simeq 0.843$.\\
1055:
1056:
1057:
1058: Fig. (\ref{fig:man1}): Manning radius $R_M$ versus packing fraction for
1059: $\lambda_\bare \ell_B = 4.2$. Extremely low densities have
1060: been considered to see the predicted
1061: power law dependence $R_M\propto \eta^{-1/4}$,
1062: see Eq. (\ref{eq:mann}). The upper inset shows that the
1063: bare charge dependence of $R_M$ also follows
1064: the form given by Eq. (\ref{eq:mann}). The lower inset shows
1065: $q(\tilde r)\ell_B/e$ as a
1066: function of distance from the rod axis on a
1067: linear-log scale: as expected, the inflection point,
1068: indicated by the vertical arrow, coincides with the point where
1069: $q(\tilde R_M)\ell_B/e=1$.\\
1070:
1071:
1072:
1073: Fig. (\ref{press}): Pressure as a function of volume fraction within the PBC and
1074: the renormalized jellium model, for both a moderately charged, and a highly charged rods
1075: (saturation limit), without added salt.
1076: The inset shows the same data on a linear scale.\\
1077:
1078:
1079:
1080: Fig. (\ref{fig:kapsphsalt}): Ratio between $\kappa $ and $\kappa_{\res}$ as a function of volume
1081: fraction for $\kappa_{\res}a=1$ (saturation regime and spherical colloids).\\
1082:
1083:
1084:
1085: Fig. (\ref{fig:Zsphsalt}): The effective charge for spheres as a function of volume fraction
1086: within the PBC and the renormalized jellium model for $\kappa_{\res}a=1$ in
1087: the saturation regime $Z_\bare \to \infty$. The inset shows the same
1088: quantity on a linear scale.\\
1089:
1090:
1091:
1092: Fig. (\ref{fig:pressphsel}): The osmotic pressure as a function of volume fraction within the PBC
1093: and the renormalized jellium model in the saturation regime for $\kappa_{\res}a=1$. The
1094: inset shows the same data on linear scale (spherical colloids).\\
1095:
1096:
1097:
1098: Fig. (\ref{fig:Lobaskin}): Comparison of the PBC and the renormalized jellium equations of state
1099: with the one obtained in Ref. \cite{Lobaskin} from the Monte Carlo
1100: simulations. Here, the macro-ion volume fraction
1101: is $\eta=8.4 \,10^{-3}$ while $Z_{\bare} \ell_B/a \simeq 21.45$.\\
1102:
1103:
1104:
1105: Fig. (\ref{fig:k0k1}): The effective charge for highly charged cylindrical colloids
1106: (saturation regime) as a function of volume
1107: fraction within the PBC and the renormalized jellium model, for $\kappa_{\res}a=1$.
1108: The inset shows appearance of a minimum in the presence of salt.\\
1109:
1110:
1111:
1112: Fig. (\ref{presscylsel2}): Osmotic pressure as a function of volume fraction
1113: within the PBC and the renormalized jellium model
1114: for $\kappa_{\res}a=1$ and
1115: $\lambda_{bare}l_{B}=2$. The inset shows the same data on linear
1116: scale (cylindrical colloids).\\
1117:
1118:
1119:
1120: %Fig. (\ref{presscylsel}): Same as Fig. \ref{presscylsel2}, for $\lambda_\bare \to \infty$.\\
1121:
1122: \newpage
1123:
1124: %------------------------------
1125: \begin{figure}[h]
1126: \centering
1127: \resizebox{9cm}{!}{\includegraphics[clip]{zeffXzbackspheress.eps}}
1128: \caption{}
1129: \label{fig:Zsph1}
1130: \end{figure}
1131: %---------------------------------
1132:
1133: %\newpage
1134:
1135: %------------------------------
1136: %\begin{figure}[h]
1137: %\centering
1138: %\resizebox{9cm}{!}{\includegraphics[clip]{zeffXzbacksphereas.eps}}
1139: %\caption{}
1140: %\label{fig:Zsph2}
1141: %\end{figure}
1142: %---------------------------------
1143:
1144:
1145: %------------------------------
1146: \begin{figure}[h]
1147: \centering
1148: \resizebox{9cm}{!}{\includegraphics[clip]{zeffXzbaresphere.eps}}
1149: \caption{}
1150: \label{fig:Zsph3}
1151: \end{figure}
1152: %---------------------------------
1153:
1154:
1155: %------------------------------
1156: \begin{figure}[h]
1157: \centering
1158: \resizebox{9cm}{!}{\includegraphics[clip]{zeffxetazbare6.eps}}
1159: \caption{}
1160: \label{fig:sphcompcellns}
1161: \end{figure}
1162: %---------------------------------
1163:
1164:
1165: %------------------------------
1166: \begin{figure}[h]
1167: \centering
1168: \resizebox{9cm}{!}{\includegraphics[clip]{presspherez6.eps}}
1169: \caption{}
1170: \label{fig:Zsphpressns}
1171: \end{figure}
1172: %---------------------------------
1173:
1174: %\newpage
1175:
1176: %------------------------------
1177: %\begin{figure}[h]
1178: %\centering
1179: %\resizebox{9cm}{!}{\includegraphics[clip]{leffxetaasat.eps}}
1180: %\caption{}
1181: %\label{fig:lambasat}
1182: %\end{figure}
1183: %------------------------------
1184:
1185:
1186: %------------------------------
1187: \begin{figure}[h]
1188: \centering
1189: \resizebox{9cm}{!}{\includegraphics[clip]{leffxetalbare1.eps}}
1190: \caption{}
1191: \label{fig:lambanosat}
1192: \end{figure}
1193: %------------------------------
1194:
1195:
1196: %------------------------------
1197: \begin{figure}[h]
1198: \centering
1199: \resizebox{9cm}{!}{\includegraphics[clip]{cmanning2.eps}}
1200: \caption{}
1201: \label{fig:Manning}
1202: \end{figure}
1203: %------------------------------
1204:
1205: %\newpage
1206:
1207: %%------------------------------
1208: %\begin{figure}[h]
1209: %\centering
1210: %\resizebox{9cm}{!}{\includegraphics[clip]{chargeeffajust.eps}}
1211: %\caption{}
1212: %\label{chargeadj}
1213: %\end{figure}
1214: %---------------------------------
1215:
1216:
1217: %------------------------------
1218: \begin{figure}[h]
1219: \centering
1220: \resizebox{9cm}{!}{\includegraphics[clip]{rmxeta.eps}}
1221: \caption{}
1222: \label{fig:man1}
1223: \end{figure}
1224: %------------------------------
1225:
1226:
1227: %------------------------
1228: \begin{figure}[h]
1229: \centering
1230: \resizebox{9cm}{!}{\includegraphics[clip]{pressionlbare1m.eps}}
1231: \caption{}
1232: \label{press}
1233: \end{figure}
1234: %----------------------------
1235:
1236:
1237: %----------------------------
1238: \begin{figure}[h]
1239: \centering
1240: \resizebox{9cm}{!}{\includegraphics[clip]{kfraccan.eps}}
1241: \caption{}
1242: \label{fig:kapsphsalt}
1243: \end{figure}
1244: %----------------------------
1245:
1246:
1247: %----------------------------
1248: \begin{figure}[h]
1249: \centering
1250: \resizebox{9cm}{!}{\includegraphics[clip]{zeffsphereasel.eps}}
1251: \caption{}
1252: \label{fig:Zsphsalt}
1253: \end{figure}
1254: %----------------------------
1255:
1256:
1257: %----------------------------
1258: \begin{figure}[h]
1259: \centering
1260: \resizebox{9cm}{!}{\includegraphics[clip]{pressphereasel.eps}}
1261: \caption{}
1262: \label{fig:pressphsel}
1263: \end{figure}
1264: %----------------------------
1265:
1266:
1267: %----------------------------
1268: \begin{figure}[h]
1269: \centering
1270: \resizebox{9cm}{!}{\includegraphics[clip]{pressZ60.eps}}
1271: \caption{}
1272: \label{fig:Lobaskin}
1273: \end{figure}
1274: %----------------------------
1275:
1276:
1277: %----------------------------
1278: \begin{figure}[h]
1279: \centering
1280: \resizebox{9cm}{!}{\includegraphics[clip]{leffcylaselsat.eps}}
1281: \caption{}
1282: \label{fig:k0k1}
1283: \end{figure}
1284: %----------------------------
1285:
1286:
1287: %----------------------------
1288: \begin{figure}[h]
1289: \centering
1290: \resizebox{9cm}{!}{\includegraphics[clip]{presscylaselL2.eps}}
1291: \caption{}
1292: \label{presscylsel2}
1293: \end{figure}
1294: %----------------------------
1295:
1296: %\newpage
1297:
1298: %----------------------------
1299: %\begin{figure}[h]
1300: %\centering
1301: %\resizebox{9cm}{!}{\includegraphics[clip]{presscylaselsat.eps}}
1302: %\caption{}
1303: %\label{presscylsel}
1304: %\end{figure}
1305: %----------------------------
1306:
1307:
1308:
1309: \end{document}
1310: