Mesoscopic description of network-forming clusters of weakly charged colloids

Systems composed of spherical charged particles in solvents containing counterions and inducing effective short-range attraction are studied in the framework of mesoscopic field-theory. We limit ourselves to mean-field approximation (MF) and to weak ordering. We discuss properties of potentials cons...

Full description

Saved in:
Bibliographic Details
Published in:Condensed Matter Physics
Date:2010
Main Authors: Ciach, A., Góźdź, W.T.
Format: Article
Language:English
Published: Інститут фізики конденсованих систем НАН України 2010
Online Access:https://nasplib.isofts.kiev.ua/handle/123456789/32094
Tags: Add Tag
No Tags, Be the first to tag this record!
Journal Title:Digital Library of Periodicals of National Academy of Sciences of Ukraine
Cite this:Mesoscopic description of network-forming clusters of weakly charged colloids / A. Ciach, W.T. Góźdź // Condensed Matter Physics. — 2010. — Т. 13, № 2. — С. 23603: 1-12. — Бібліогр.: 27 назв. — англ.

Institution

Digital Library of Periodicals of National Academy of Sciences of Ukraine
id nasplib_isofts_kiev_ua-123456789-32094
record_format dspace
spelling Ciach, A.
Góźdź, W.T.
2012-04-08T15:48:46Z
2012-04-08T15:48:46Z
2010
Mesoscopic description of network-forming clusters of weakly charged colloids / A. Ciach, W.T. Góźdź // Condensed Matter Physics. — 2010. — Т. 13, № 2. — С. 23603: 1-12. — Бібліогр.: 27 назв. — англ.
1607-324X
PACS: 61.20.Gy, 64.60.De, 82.70.Dd
https://nasplib.isofts.kiev.ua/handle/123456789/32094
Systems composed of spherical charged particles in solvents containing counterions and inducing effective short-range attraction are studied in the framework of mesoscopic field-theory. We limit ourselves to mean-field approximation (MF) and to weak ordering. We discuss properties of potentials consisting of strong short-range attraction and weak long-range repulsion (SALR) in the context of formation of nonuniform distribution of particles on a mesoscopic length scale instead of macroscopic phase separation. In earlier work it was found that spherical, cylindrical and slab-like clusters of particles are formed, and for low enough temperatures the clusters form ordered, periodic bcc, hexagonal and lamellar phases. In addition, a gyroid phase was predicted in which two interwoven regular network-like clusters branching in triple junctions are formed. At properly rescaled density and temperature, the coexistence lines between different ordered phases were found to be universal in MF, with the exception of the gyroid phase. Here the phase diagram is determined for two choices of the SALR potential, one corresponding to a large range of the attractive part of the potential, and the other one to a very small range of attraction. We find that the region of stability of the gyroid phase very weakly depends on the form of the SALR potential within the approximate theory.
У межах мезоскопічної польової теорії вивчаються системи, що складаються зі сферичних заряджених частинок у розчинниках, які містять контріони та індукують короткосяжне притягання. Розглянуто наближення середнього поля (СП) і слабке впорядкування, а також властивості потенціалів, що складаються із сильного короткосяжного притягання та слабкого далекосяжного відштовхування (SALR) у контексті формування неоднорідного розподілу частинок на мезоскопічних масштабах довжин замість макроскопічного фазового відокремлення. Було показано, що формуються сферичні, циліндричні та щілиноподібні кластери частинок, і для достатньо низьких температур ці кластери формують впорядковані періодичні об'ємоцентричні, гексагональні та ламеларні фази. Крім того, було передбачено гіроїдну фазу, в якій формуються два переплетені регулярні мережоподібні кластери, що розгалужуються в потрійному з'єднанні. Знайдено, що за відповідного масштабування густини та температури, лінії співіснування між різними впорядкованими фазами є універсальними в наближенні СП, за винятком гіроїдної фази. Визначено фазову діаграму для двох виборів потенціалу SALR: такого, що відповідає великій області притягувальної частини потенціалу, та такого, що відповідає малій області притягання. Виявлено, що область стабільності гіроїдної фази дуже слабо залежить від форми потенціалу SALR у межах наближеної теорії.
en
Інститут фізики конденсованих систем НАН України
Condensed Matter Physics
Mesoscopic description of network-forming clusters of weakly charged colloids
Мезоскопічний опис мережо-формуючих кластерів слабо заряджених колоїдів
Article
published earlier
institution Digital Library of Periodicals of National Academy of Sciences of Ukraine
collection DSpace DC
title Mesoscopic description of network-forming clusters of weakly charged colloids
spellingShingle Mesoscopic description of network-forming clusters of weakly charged colloids
Ciach, A.
Góźdź, W.T.
title_short Mesoscopic description of network-forming clusters of weakly charged colloids
title_full Mesoscopic description of network-forming clusters of weakly charged colloids
title_fullStr Mesoscopic description of network-forming clusters of weakly charged colloids
title_full_unstemmed Mesoscopic description of network-forming clusters of weakly charged colloids
title_sort mesoscopic description of network-forming clusters of weakly charged colloids
author Ciach, A.
Góźdź, W.T.
author_facet Ciach, A.
Góźdź, W.T.
publishDate 2010
language English
container_title Condensed Matter Physics
publisher Інститут фізики конденсованих систем НАН України
format Article
title_alt Мезоскопічний опис мережо-формуючих кластерів слабо заряджених колоїдів
description Systems composed of spherical charged particles in solvents containing counterions and inducing effective short-range attraction are studied in the framework of mesoscopic field-theory. We limit ourselves to mean-field approximation (MF) and to weak ordering. We discuss properties of potentials consisting of strong short-range attraction and weak long-range repulsion (SALR) in the context of formation of nonuniform distribution of particles on a mesoscopic length scale instead of macroscopic phase separation. In earlier work it was found that spherical, cylindrical and slab-like clusters of particles are formed, and for low enough temperatures the clusters form ordered, periodic bcc, hexagonal and lamellar phases. In addition, a gyroid phase was predicted in which two interwoven regular network-like clusters branching in triple junctions are formed. At properly rescaled density and temperature, the coexistence lines between different ordered phases were found to be universal in MF, with the exception of the gyroid phase. Here the phase diagram is determined for two choices of the SALR potential, one corresponding to a large range of the attractive part of the potential, and the other one to a very small range of attraction. We find that the region of stability of the gyroid phase very weakly depends on the form of the SALR potential within the approximate theory. У межах мезоскопічної польової теорії вивчаються системи, що складаються зі сферичних заряджених частинок у розчинниках, які містять контріони та індукують короткосяжне притягання. Розглянуто наближення середнього поля (СП) і слабке впорядкування, а також властивості потенціалів, що складаються із сильного короткосяжного притягання та слабкого далекосяжного відштовхування (SALR) у контексті формування неоднорідного розподілу частинок на мезоскопічних масштабах довжин замість макроскопічного фазового відокремлення. Було показано, що формуються сферичні, циліндричні та щілиноподібні кластери частинок, і для достатньо низьких температур ці кластери формують впорядковані періодичні об'ємоцентричні, гексагональні та ламеларні фази. Крім того, було передбачено гіроїдну фазу, в якій формуються два переплетені регулярні мережоподібні кластери, що розгалужуються в потрійному з'єднанні. Знайдено, що за відповідного масштабування густини та температури, лінії співіснування між різними впорядкованими фазами є універсальними в наближенні СП, за винятком гіроїдної фази. Визначено фазову діаграму для двох виборів потенціалу SALR: такого, що відповідає великій області притягувальної частини потенціалу, та такого, що відповідає малій області притягання. Виявлено, що область стабільності гіроїдної фази дуже слабо залежить від форми потенціалу SALR у межах наближеної теорії.
issn 1607-324X
url https://nasplib.isofts.kiev.ua/handle/123456789/32094
citation_txt Mesoscopic description of network-forming clusters of weakly charged colloids / A. Ciach, W.T. Góźdź // Condensed Matter Physics. — 2010. — Т. 13, № 2. — С. 23603: 1-12. — Бібліогр.: 27 назв. — англ.
work_keys_str_mv AT ciacha mesoscopicdescriptionofnetworkformingclustersofweaklychargedcolloids
AT gozdzwt mesoscopicdescriptionofnetworkformingclustersofweaklychargedcolloids
AT ciacha mezoskopíčniiopismerežoformuûčihklasterívslabozarâdženihkoloídív
AT gozdzwt mezoskopíčniiopismerežoformuûčihklasterívslabozarâdženihkoloídív
first_indexed 2025-11-24T19:09:26Z
last_indexed 2025-11-24T19:09:26Z
_version_ 1850493881097912320
fulltext Condensed Matter Physics 2010, Vol. 13, No 2, 23603: 1–12 http://www.icmp.lviv.ua/journal Mesoscopic description of network-forming clusters of weakly charged colloids A. Ciach, W.T. Góźdź Institute of Physical Chemistry, Polish Academy of Sciences, 01-224 Warszawa, Poland Received February 10, 2010, in final form March 18, 2010 Systems composed of spherical charged particles in solvents containing counterions and inducing effective short-range attraction are studied in the framework of mesoscopic field-theory. We limit ourselves to mean- field approximation (MF) and to weak ordering. We discuss properties of potentials consisting of strong short- range attraction and weak long-range repulsion (SALR) in the context of formation of nonuniform distribution of particles on a mesoscopic length scale instead of macroscopic phase separation. In earlier work it was found that spherical, cylindrical and slab-like clusters of particles are formed, and for low enough temperatures the clusters form ordered, periodic bcc, hexagonal and lamellar phases. In addition, a gyroid phase was predicted in which two interwoven regular network-like clusters branching in triple junctions are formed. At properly rescaled density and temperature, the coexistence lines between different ordered phases were found to be universal in MF, with the exception of the gyroid phase. Here the phase diagram is determined for two choices of the SALR potential, one corresponding to a large range of the attractive part of the potential, and the other one to a very small range of attraction. We find that the region of stability of the gyroid phase very weakly depends on the form of the SALR potential within the approximate theory. Key words: colloids, clusters, self-assembly, order-disorder phase transitions PACS: 61.20.Gy, 64.60.De, 82.70.Dd 1. Introduction Biologically relevant macromolecules, nanoparticles or colloids are typically charged, and there- fore repel each other. On the other hand, van der Waals and various solvent-mediated interactions lead to effective attraction between them. The range of attraction is usually shorter than the range of screened electrostatic repulsion. However, for weakly to moderately charged particles, the magni- tude of the attraction is significantly larger than the magnitude of repulsion. The resulting potential is negative for short distances, and positive for larger distances and is known as short-range attrac- tion long-range repulsion (SALR) potential [1–15]. Short separations of particles corresponding to the attraction are favorable, whereas the distances close to the maximum of repulsion lead to a significant increase in energy. Depending on the repulsion barrier and range, as well as on density ρ and temperature T , the random distribution of particles and macroscopic phase separation can be less favorable than the formation of clusters of various shape and size. When the average dis- tance between particles corresponds to strong repulsion, then it is energetically favorable to form clusters with particles strongly attracting each other, such that the separation between the clusters is larger than the range of repulsion. When separations between the particles forming the cluster are mainly within the range of attraction and the separation between clusters is larger than the range of repulsion, the energy takes low values. This effect may overcompensate the entropy loss associated with cluster formation. When the density increases such that the average separation between particles is too small to form spherical clusters with particles attracting each other, and with average separation between the clusters larger than the range of repulsion, then under the constraint of a given density, it is favorable to grow the cluster in one dimension, and keep the distance between elongated clusters larger than the range of repulsion. For still higher densities, slab-like clusters sufficiently far apart from each other may be created. Such scenario has been observed both in experiments and in simulations. In experimental systems the clusters have been c© A. Ciach, W.T. Góźdź 23603-1 http://www.icmp.lviv.ua/journal A. Ciach, W.T. Góźdź directly seen by confocal microscopy [3, 7]; for an increasing density, first spherical, then elongated and finally network-like clusters were observed. In the case of globular protein solutions, cluster formation is a subject of intensive debate [3, 16]. Simulation [9, 11] and theoretical studies [17, 18] of particular examples of the SALR potential show that indeed spherical, cylindrical and slab-like clusters are formed for an increasing density. In addition, for some range of particle density, lower than the density of crystallization, the clusters form ordered, periodic structures at low enough temperatures. The results of simulations and recent theoretical prediction [17, 18] indicate that the sequence of ordered phases is the same regardless of the detailed form of the SALR poten- tial. In particular, small [11] and very large [9] clusters are spherical, cylindrical and slab-like in shape for similar volume fractions of particles. Mean-field results [17] indicate that for a suitably rescaled temperature, the phase diagram is universal when weak ordering of particles occurs, i.e., in rescaled variables the phase-coexistence lines for different forms of the SALR potential collapse onto master curves. There is one exception, however. Between the stability regions of the hexago- nal phase of cylinders and lamellar phase of slabs, a thermodynamically stable gyroid phase was found [17], but its stability region is not universal, i.e., it may be different for different forms of the effective interactions. In this phase particles form two interwoven networks of clusters which branch in triple junctions. This phase may be related to the branched network of clusters observed experimentally and in simulations [6, 7, 14]. The prediction of stability of the gyroid phase is based on mean-field approximation to a mesoscopic field theory, and the results were obtained under assumption of weak ordering (i.e., strong fluctuations of particles around their average positions, leading to smooth concentration profiles with small amplitudes). Hence, further evidence is required to confirm that such a peculiar regular ’gel’ can be a thermodynamically stable phase. In this work we address an open issue how the stability region of this gyroid phase on the phase diagram depends on the shape of the SALR potential, and consequently on the size of the equilibrium clusters at strong dilution. To this end, we consider two forms of the SALR potential, one leading to very large and the other one to very small clusters, and determine the phase diagrams for these two versions of the SALR potential using the method developed in [17]. In section 2 we introduce a particular form of the SALR potential, consisting of two Yukawa terms, and discuss the range of parameters corresponding to phase separation and to clustering. In section 3 we briefly review the version of the statistical field theory introduced in [17]. This approach is suitable for a description of collective phenomena leading to ordering on the mesoscopic length scale. In section 4 we describe the calculations of the phase diagram in the simplest mean-field approximation in the case of weak ordering, and discuss our results. 2. The interaction potential and its properties We restrict our attention to colloids, nanoparticles or globular macromolecules that can be modeled as spherical hard cores with diameter σ, and with additional interactions present for distances r > σ. We limit ourselves to the interaction potential that has the form u(r) = uSR(r) + uLR(r) (1) with uSR(r) = { 0 if r > 1, ∞ if 0 < r < 1, (2) where r is in σ units. The long-range part uLR(r) is defined for r > 1, and for such separations we assume that uLR(r) has the form v(r) = −A1 r e−z1r + A2 r e−z2r, (3) where zi is the inverse range in σ−1 units. We consider the SR and the LR parts separately and treat the case with uLR(r) = 0 as a reference system. In microscopic theory we need an extension 23603-2 Mesoscopic description of network-forming clusters of weakly charged colloids of uLR(r) for r < 1 for calculational reasons. The contribution to the energy of the system coming from overlapping cores should be avoided, and therefore we assume uLR(r) = { v(r) if r > 1, 0 if r < 1. (4) Such an extension is not unique, and other choices made in [19–22] in different context lead to more accurate results. The simplest possibility of extension of the long-range part of the potential for r < 1 would be to assume the form (3) for all r > 0. In this case analytical results can be easily obtained. However, the self-energy would be included, and moreover, there would be no information on the length scale set by the hard cores. We shall compare the structure factor obtained within the simplest Gaussian approximation for uLR(r) = v(r) and for uLR(r) given by (4) in the next section; the form (4) leads to the structure factor that qualitatively agrees with experiments and simulations, whereas for uLR(r) = v(r) the second peak of the structure factor, associated with particle-particle correlations within the clusters is missing. In the mesoscopic theory [17, 18] the key role is played not by the potential alone, but rather by the potential multiplied by the microscopic correlation function, g(r), calculated for the frozen mesoscopic density distribution. The form of g(r)uLR(r) is given by (4) for any extension of uLR(r) for r > 1 for the simplest approximation g(r) ≈ θ(r − 1). We focus on the case of strong short-range attraction and weak long-range repulsion, and assume z1 > z2, and A1 > A2. In Fourier representation (4) takes the form ũLR(k) = 4π [ A2e −z2 z2 2 + k2 ( z2 sin k k + cos k ) − A1e −z1 z2 1 + k2 ( z1 sin k k + cos k ) ] . (5) We use the notation f̃(k) for the Fourier transform of the function f(r). Note that the energy associated with a plane wave of concentration, ρ∗(z) = Φ cos(kz) + ρ̄∗, is equal to V Φ2ũLR(k)/4, where V is the volume of the system and we introduced dimensionless density ρ∗ = ρσ3 (6) and denoted the space-averaged density by ρ̄∗. When ũLR(k) < 0, a concentration wave with the wavelength 2π/k leads to a decrease of energy compared to homogeneous distribution of parti- cles. The minimum of ũLR(k) at k = kb gives the most probable wavelength 2π/kb of the plane concentration wave, where layers with excess particle density of width π/kb are followed by re- gions of depleted particle density of the same width. Superpositions of plane waves in different directions may lead to nonuniform distribution of particles of different symmetries. The energy associated with the most probable wave of concentration, ũLR(kb) is a convenient energy unit, and we introduce dimensionless temperature T ∗ by T ∗ = kBT |ũLR(kb)| . (7) In [17] it was shown that in mean-field (MF) approximation, the spinodal lines are given by a universal curve in (T ∗, ρ∗) phase diagram, independently of the shape of the SALR potential. Thus, such units are appropriate for determination of the common features of different systems that form nonuniform structures on the mesoscopic length scale. The gas-liquid phase separation may occur when ũLR(0) < 0, and the necessary condition for the phase separation is A2 A1 < e−(z1−z2) ( z1 + 1 z2 + 1 )( z2 z1 )2 . (8) The behavior depends on whether ũLR(k) assumes a minimum or a maximum for k = 0. We find d2ũLR(k) dk2 |k=0 = 4π 3 [A1e −z1 ( z3 1 + 3z2 1 + 6z1 + 6 ) z4 1 − A2e −z2 ( z3 2 + 3z2 2 + 6z2 + 6 ) z4 2 ] . (9) 23603-3 A. Ciach, W.T. Góźdź For parameters Ai, zi corresponding to negative values of the above (maximum at k = 0) we find periodic ordering, because the second extremum, for k > 0, is a minimum. For z1 > 6 and z2 ≈ 1 this is typically the case if A2/A1 > 10−4. Hence, even very weak repulsion suppresses phase separation for very short-range of attraction. For longer range of attraction we need a sufficiently strong repulsion to suppress phase separation in favor of clustering. The parameters Ai and zi may be state-dependent, and in different systems the dependence on temperature and density may be different. In particular, for charged colloids in the presence of counterions, z2 is the inverse screening length which is a function of temperature and density of the form z2 ∝ √ ρ/T . Here we are not interested in any particular system, but rather try to determine the shape of the phase diagram in the universal variables. Note that when the potential depends on temperature and density, the same holds for the energy unit |ũLR(kb)|. We choose two sets of parameters for the potential given by (3). A1 = 1, A2 = 0.2, z1 = 1, z2 = 0.5, (10) A1 = 1, A2 = 0.05, z1 = 3, z2 = 0.5. The shape of the potential in the cases I and II is quite different. In the case I large clusters are formed, because kb ≈ 0.6088 whereas in the case II the clusters are rather small, because kb ≈ 1.7926. 3. Field theory Consider nonuniform density distributions ρ∗(x). The probability that the density has the given form ρ∗(x) in the mesoscopic description is given by [17] p[ρ∗(x)] = Ξ−1e−βΩMF[ρ∗(x)], (11) where β = 1/(kBT ), kB is the Boltzmann constant, and ΩMF[ρ∗] is the grand-thermodynamic potential in the system with the density constrained to have the given nonuniform form ρ∗(x), βΩMF[ρ∗] = 1 2 ∫ x ρ∗(x)βuLR(|x − x ′|)ρ∗(x′) + ∫ x ( βfh[ρ ∗|x) − µ∗ρ∗(x) ) , (12) where µ∗ = βµ/σ3, and we use the notation ∫ x ≡ ∫ dx. fh[ρ∗|x) is the free-energy density of the reference system (with uLR = 0) at x. In general, fh[ρ ∗|x) is a functional of ρ∗(x). Here we limit ourselves to the local-density approximation, where the free-energy density at the space position x is a function of the local density at x fh[ρ ∗|x) = fh(ρ ∗(x)). (13) We adopt the PY approximation for the free energy of hard spheres at density ρ∗, βfh(ρ∗) = ρ∗ ln(ρ∗) − ρ∗ + ρ∗ [ 3s(2− s) 2(1− s)2 − ln(1 − s) ] , (14) where s = ρ∗π/6. Note that in this theory the form of ΩMF[ρ∗(x)] depends on the choice for the extension of uLR(r) for r < 1 (see the first term in equation (12)). Let us first consider βΩMF for uniform states ρ∗ = const. βΩMF(ρ∗)/V = −β∗p∗ = ρ∗2 2 β∗ũ∗LR(0) + βfh(ρ∗) − µ∗ρ∗, (15) where we used the thermodynamic relation Ω = −pV , introduced ũ∗LR(k) = ũLR(k)/|ũLR(kb)|, β∗ = 1/T ∗, and p∗ = p/|ṽ(kb)|. 23603-4 Mesoscopic description of network-forming clusters of weakly charged colloids Figure 1. The LR part of the interaction potential, uLR(r) given by (4), for two choices of parameters. Left: A1 = 1, A2 = 0.2, z1 = 1, z2 = 0.5. Right: A1 = 1, A2 = 0.05, z1 = 3, z2 = 0.5. Figure 2. Fourier transform of the LR part of the interaction potential, ũLR(k) given by (5), for two choices of parameters. Left: A1 = 1, A2 = 0.2, z1 = 1, z2 = 0.5. Right: A1 = 1, A2 = 0.05, z1 = 3, z2 = 0.5. Minimum condition of ΩMF(ρ∗): ∂ΩMF ∂ρ∗ = 0 (16) has the explicit form ρ∗βũLR(0) + dβfh(ρ∗) dρ∗ = µ∗ (17) and ũLR(0) = ∫ x uLR(x) = 4π [ A2e −z2 z2 2 ( z2 + 1) − A1e −z1 z2 1 ( z1 + 1) ] . (18) Densities that satisfy the above for a given µ∗ and T ∗ correspond to the most probable density of 23603-5 A. Ciach, W.T. Góźdź the uniform system and are denoted by ρ∗ 0. The explicit form of (16) is µ∗ = ρ∗0β ∗ũ∗LR(0) + ln(ρ∗ 0) − ln(1 − s0) + 14s0 − 13s2 0 + 5s3 0 2(1 − s0)3 . (19) The correlation function for mesoscopic density in the Gaussian approximation is just given by G̃(k) = 1 C̃(k) = ρ∗0(1 − s)4 β∗ũ∗LR(k)ρ∗0(1 − s)4 + (1 + 2s)2 , (20) where C̃(k) = δ2βΩMF δρ̃∗(k)δρ̃∗(−k) = β∗ũ∗LR(k) + d2βfh(ρ ∗) dρ∗2 . (21) In figure 3 G̃(k) is shown for the first set of parameters of the SALR potential (see equation (10)), for the two extensions of ũ∗LR(r) for r < 1. The second maximum for r ≈ 6, present in experimental systems, occurs when equation (4) is used for ũ∗LR(r). Figure 3. Correlation function in Fourier representation, G̃(k) (equation (20)), in the diluted phase (structure factor is given by S(k) = G̃(k)/ρ∗) for A1 = 1, A2 = 0.2, z1 = 1 and z2 = 0.5 and ρ∗ = 0.1, and T ∗ = 1.58. The dashed and solid lines are for uLR = v and for uLR given by equation (4) respectively. The boundary of stability of ΩMF is given by C̃(kb) = 0. (22) Note that the boundary of stability of the functional (12) significantly depends on the form of ũ∗LR(r) for r < 1, where it is not uniquely defined. In the context of the mesoscopic theory, where ũ∗LR(r) in (12) should be replaced by ũ∗LR(r)g(r), the quantitative results depend on the assumed form of the correlation function g(r) with density constrained to have a fixed form on the mesoscopic length scale. Here we assume g(r) = θ(r− 1). The explicit expression for the boundary of stability of the uniform phase is β∗ = (1 + 2s)2 ρ∗(1 − s)4 . (23) The MF line of instability depends on the state dependence of the temperature unit |ũLR(kb)|. Instability with respect to separation of uniform phases occurs when kb = 0, and equation (23) represents the spinodal line in this case. Metastable separation occurs provided that the condition (8) is satisfied, and is given by (23) with ũ∗LR(kb) replaced by ũ∗LR(0). 23603-6 Mesoscopic description of network-forming clusters of weakly charged colloids 4. First order transition to microsegregated phases For weakly first-order transitions we assume that the density has the form ρ(x) = ρ∗ 0 + η(x). (24) Next we expand fh(ρ∗) about ρ∗0, and we obtain β∆ΩMF[ρ∗0, η(x)] = βΩMF[ρ∗0 + η(x)] − βΩMF[ρ∗0] = 1 2 ∫ k η̃(k)β∗ũ∗LR(k)η̃(−k) + ∞ ∑ n=3 An n! ∫ x η(x)n , (25) where An = ∂nβfh(ρ) ∂ρ∗n |ρ∗=ρ∗ 0 . (26) The stability requirement for η → ∞ allows us for truncating the expansion at even n > 4. Here we shall limit ourselves to n = 4. Explicit expressions for A3 and A4 are given in Appendix. The functional (25) has a well know structure [23–26] considered before in phenomenological approaches to block copolymers, highly charged colloids and microemulsions. We repeat the same procedure and assume that η(x) is periodic and can be written in the form η(x) = ∑ n Φngph n (x), (27) where gph n (x) represents orthonormal basis functions for the n-th shell that have the symmetry of the phase ph, and satisfy the normalization condition ∫ Vu gn(x)2 Vu = 1. (28) By ∫ Vu we denote an integration over the unit cell of the ordered structure. The volume of the unit cell is denoted by Vu. Φn is the n-th amplitude. Stable or metastable phases correspond to the global or a local minimum of ∆ΩMF[ρ∗0, η(x)]. At phase coexistence between two phases ∆ΩMF[ρ∗0, η(x)] for these phases takes equal values. At the transition between the fluid and the ordered phase ∆ΩMF[ρ∗0, η(x)] = 0. 4.1. One-shell approximation Let us first limit ourselves to the one-shell approximation. For structures possessing different symmetries the Fourier transform of g1 has the form g̃ph 1 (k) = (2π)d √ 2nph nph ∑ j=1 ( wδ(k − k j b) + w∗δ(k + k j b) ) , (29) where ww∗ = 1 and 2nph is the number of vectors k j b in the first shell of the structure ph. For the first shell |kj b | = kb. Since uLR(k) assumes a minimum for k = kb, the contribution to the grand potential from the first shell is lower than the contributions from the remaining shells. The explicit forms of the functions gph 1 in real-space representation for lamellar, hexagonal, bcc and gyroid phases are given in Appendix. Within the given symmetry, the amplitude Φ1 of the order parameter corresponds to the minimum ∆ΩMF ph of ∆ΩMF[ρ∗0 + Φ1g ph 1 ] with respect to Φ1. As already shown by Leibler [24], for A3 6= 0, the ordered phase, coexisting with the fluid, has the bcc symmetry in MF. For A3 = 0 the transition is to the striped (lamellar) phase [24]. In the considered system A3(ρ ∗) < 0 for ρ∗ < ρ∗c , A3(ρ ∗) = 0 for ρ∗ = ρ∗c , and A3(ρ ∗) > 0 for ρ∗ > ρ∗c , where ρ∗c 23603-7 A. Ciach, W.T. Góźdź is the critical density of the gas-liquid separation. In the PY approximation (14) ρ∗c ≈ 0.2457358. From (25)–(28) we obtain in the one-shell approximation ∆ΩMF = 1 2 C̃(kb)Φ2 1 + A3κ ph 3 3! Φ3 1 + A4κ ph 4 4! Φ4 1 , (30) where κph n = ∫ Vu gph 1 (x)n Vu . (31) From the extremum condition ∂∆ΩMF ∂Φ1 = C̃(kb)Φ1 + A3κ ph 3 2 Φ2 1 + A4κ ph 4 3! Φ3 1 = 0 (32) we obtain the amplitude Φ1+,− = √ 3 [ − √ 3A3κ3 ± √ 3(A3κ3)2 − 8A4κ4C̃(kb) ] 2A4κ4 (33) for each phase characterized by different geometric factors κn given in Appendix. Transition be- tween the disordered and the bcc phase, as well as the amplitude of the density profiles along the transition line are given by analytic expressions in [17], where the remaining transition lines between ordered phases, obtained numerically, are also shown. 4.2. Two-shell approximation The accuracy of the one-shell approximation depends on the shape of ũLR(k), in particular, on the ratio ũLR(kb2)/ũLR(kb), where kb2 is the wavenumber corresponding to the second shell. The approximation is the worse the smaller is the above ratio, since the second shell should not be disregarded when it leads to comparable energy as the first shell. For a given shape of the potential, the value of kb2 matters, and it is different for different structures. As discussed in detail in [17, 27], for the gyroid phase kb2 = kb2/ √ 3, whereas for the remaining phases kb2/kb is much larger, so the value of ũLR(kb2)/ũLR(kb) is much larger too. We thus consider the gyroid phase in the two-shell approximation, and the remaining phases in the one-shell approximation. In the two-shell approximation the grand potential assumes the form ∆ΩMF = 1 2 C̃(kb)Φ2 1 + κ3 A3 3! Φ3 1 + κ4 A4 4! Φ4 1 + 1 2 C̃(k2b)Φ 2 2 + κ0,3 A3 3! Φ3 2 + κ0,4 A4 4! Φ4 2 + κ2,1 A3 2 Φ2 1Φ2 + κ2,2 A4 2!2! Φ2 1Φ 2 2 , (34) where the geometric factors κph n,m = ∫ Vu gph 1 (x)ngph 2 (x)m Vu . (35) for the gyroid phase are given in Appendix. The minimum condition with respect to Φi gives Φ1± = ± √ − C̃(kb2)Φ2 + κ0,3 A3 2 Φ2 2 + κ0,4 A4 3! Φ3 2 κ2,1 A3 2 + κ2,2 A4 2 Φ2 , (36) Φ2± = −κ2,1A3 ± √ ∆ κ2,2A4 , (37) with ∆ = κ2 2,1A 2 3 − 2κ2,2A4 ( C̃(kb) + κ3,0 A3 2 Φ1 + κ4,0 A4 6 Φ2 1 ) . (38) 23603-8 Mesoscopic description of network-forming clusters of weakly charged colloids For ρ∗ < ρ∗c and ρ∗ > ρ∗c the minimum of ∆ΩMF corresponds to Φ1+ and Φ2−, and to Φ1− and Φ2+ respectively. It is important to note that the structure given by two-shell and by one-shell approximation is significantly different. Namely, in the one-shell approximation a single network of excess density is present, whereas for Φ1Φ2 < 0 two much thinner networks of excess density are formed. The phase diagrams calculated for the two potentials with parameters given in equa- tion (10) (see figures 1–2) are shown in figures 4, 5. Note that despite significantly different shape Figure 4. Phase diagram obtained in MF approximation with the bcc, hexagonal and lamellar phases considered in the one-shell and the gyroid phase in the two-shell approximation for A1 = 1, A2 = 0.2, z1 = 1 and z2 = 0.5. The outer solid lines are the coexistence lines between the disordered and the bcc crystal phases, the bcc and the hexagonal phases coexist along the dashed lines, along the dotted lines the hexagonal and the lamellar phases coexist, and inside the shaded regions the gyroid phase is stable. For densities lower than ρ∗c clusters of particles form ordered patterns, whereas for ρ∗ > ρ∗c the pattern is formed by bubbles – regions with depleted particle density. Figure 5. Phase diagram obtained in MF approximation with the bcc, hexagonal and lamellar phases considered in one-shell and the gyroid phase in the two-shell approximation for A1 = 1, A2 = 0.05, z1 = 3 and z2 = 0.5. The outer solid lines are the coexistence lines between the disordered and the bcc crystal phases, the bcc and the hexagonal phases coexist along the dashed lines, along the dotted lines the hexagonal and the lamellar phases coexist, and inside the shaded regions the gyroid phase is stable. For densities lower than ρ∗c clusters of particles form ordered patterns, whereas for ρ∗ > ρ∗c the pattern is formed by bubbles – regions with depleted particle density. of the interaction potential, the extent of the stability region of the gyroid phase on the phase diagram given in reduced variables (ρ∗, T ∗) is similar. Moreover, the phase diagram in variables (ρ∗, T ∗) is similar to the phase diagram obtained earlier [17] for the SALR potential characterized by the parameters A1 = 140 exp(8.4), A2 = 30 exp(1.55), z1 = 8.4 and z2 = 1.55, and corre- sponding to kb ≈ 1.94026. The ratio of ṽ(b2)/ṽ(kb) for the cases I and II and for the previously considered potential takes the values 0.9762993733, 0.8973863512, 0.9119505483 respectively. In the 23603-9 A. Ciach, W.T. Góźdź case of large clusters (case I, π/kb ≈ 5 ) the gyroid phase extents to higher temperatures and oc- cupies a larger range of densities than in the case of small clusters (case II, π/kb ≈ 1.7 and [17], π/kb ≈ 1.6). We conclude that in the simple MF approximation, the phase diagram in reduced variables, although not universal beyond the one-shell approximation, rather weakly depends on the shape of the effective interaction potential. The region of stability of the gyroid phase increases when the minimum of the interaction potential in Fourier representation becomes shallower, i.e., when ṽ(kb2)/ṽ(kb) increases. 5. Summary We have considered the SALR potential representing a wide class of biological and soft matter systems in MF approximation and in the case of weak ordering. We first discussed the properties of such kind of potential, and next focused on determination of the phase diagram in reduced variables. Our main goal was to determine the stability region of gyroid phase with Ia3d symmetry for various shapes of the SALR potential. We have chosen two qualitatively different cases, I and II, corresponding to very large clusters and to small clusters of colloids, respectively. We have found that the phase diagrams in reduced units in these two cases are similar. The gyroid phase is located between the hexagonal and the lamellar phase, and may be related to a branched network of clusters observed experimentally and in simulations [6, 7, 14].The smaller is the difference between the energy gain associated with the formation of a planar density wave with the wave number k ≈ kb and the gain for the most probable wavelength k = kb, the more favorable is the gyroid phase. We expect that the free-energy landscape contains many local minima corresponding to a distorted network, and that the formation of thermodynamically stable regular structure of the gyroid phase in experimental system may be preempted by a glass-like disordered branched network that locally resembles the gyroid phase. Confirmation of the existence of the predicted gyroid phase by experiment or simulation remains a challenge. Acknowledgement This work is dedicated to Prof. Ihor Mryglod on the occasion of his 50 birthday. We grate- fully acknowledge partial support by the Polish Ministry of Science and Higher Education, Grant No. NN 202 006034. 6. Appendix. Explicit expressions of functions and parameters Parameters in the functional (25) A3(ρ) = 12s3 + 20s2 + 5s − 1 ρ2(1 − s)5 , (39) A4(ρ) = 2 − 12s + 30s2 + 112s3 + 48s4 ρ3(1 − s)6 (40) geometric factors in equations (30) and (34) κbcc 3 = √ 4 3 , (41) κbcc 4 = 15 4 , (42) 23603-10 Mesoscopic description of network-forming clusters of weakly charged colloids κIa3d 3,0 = 1√ 6 , (43) κIa3d 4,0 = 17 8 , (44) κIa3d 0,3 = 2√ 3 , (45) κIa3d 0,4 = 15 4 , (46) κIa3d 2,1 = − √ 3 6 , (47) κIa3d 2,2 = 1 2 , (48) κhex 3 = √ 2 3 , (49) κhex 4 = 5 2 , (50) κlam 3 = 0 , (51) κlam 4 = 3 2 , (52) where κn,0 ≡ κn. Expressions for the functions g1(x). We use the notation x = (x1, x2, x3). g` 1(x) = √ 2 cos(kbx1), (53) ghex 1 (x) = √ 2 3 [ cos(kbx1) + 2 cos (kbx1 2 ) cos ( √ 3kbx2 2 ) ] , (54) gbcc 1 (x) = 1√ 3 ∑ i<j ( cos (kb(xi + xj)√ 2 ) + cos (kb(xi − xj)√ 2 ) ) , (55) gIa3d 1 (x) = √ 8 3 [ cos (kbx1√ 6 ) sin (kbx2√ 6 ) sin (2kbx3√ 6 ) + cos (kbx2√ 6 ) sin (kbx3√ 6 ) sin (2kbx1√ 6 ) + cos (kbx3√ 6 ) sin (kbx1√ 6 ) sin (2kbx2√ 6 ) ] . (56) Expression for the function g2(x) for the gyroid phase gIa3d 2 (x) = 1√ 3 ∑ i<j ( cos (2kb(xi + xj)√ 6 ) + cos (2kb(xi − xj)√ 6 ) ) . (57) References 1. Sear R.P., Gelbart W.M., J. Chem. Phys., 1999, 110, 4582. 2. Sciortino F., Mossa S., Zaccarelli E., Tartaglia P., Phys. Rev. Lett., 2004, 93, 055701. 3. Stradner A., Sedgwick H., Cardinaux F., Poon W.C.K., Egelhaaf S.U., Schurtenberger P., Nature, 2004, 432, 492. 4. Sedgwick H., Egelhaaf S., Poon W., J. Phys. Condens. Matter, 2004, 16, S4913. 5. Imperio A., Reatto L., J. Phys.: Cond. Mat., 2004, 16, 3769. 6. Sciortino F., Tartaglia P., Zaccarelli E., J. Phys. Chem. B, 2005, 109, 21942. 7. Campbell A.I., J.Anderson V., van Duijneveldt J.S., Bartlett P., Phys. Rev. Lett., 2005, 94, 208301. 8. Pini D., Parola A., Reatto L., J. Phys.: Cond. Mat., 2006, 18, S2305. 9. de Candia A., Del Gado E., Fierro A., Sator N., Tarzia M., Coniglio A., Phys. Rev. E, 2006, 74, 010403(R). 23603-11 A. Ciach, W.T. Góźdź 10. Archer A.J., Pini D., Evans R., Reatto L., J. Chem. Phys., 2007, 126, 014104. 11. Archer A.J., Wilding N.B., Phys. Rev. E, 2007, 76, 031501. 12. Archer A.J., Phys. Rev. E, 2008, 78, 031402. 13. Archer A., Ionescu, Pini D., Reatto L., J. Phys.: Cond. Mat., 2008, 20, 415106. 14. Toledano J.C.F., Sciortino F., Zaccarelli E., Soft Matter, 2009, 5, 2390. 15. Sennato S., Truzzolillo D., Bordi F., Sciortino F., Cametti C., Colloid Surface, 2009, 343, 34. 16. Shukla A., Mylonas E., Di Cola E., Finet S., Timmins P., Narayanan T., Svergun D.I., PNAS, 2008, 105, 5075. 17. Ciach A., Phys. Rev. E, 2008, 78, 061505. 18. Ciach A., AIP Conference Proceedings, 2009, 1198, 13. 19. Patsahan O., Condens. Matter Phys., 2004, 7, 35. 20. Patsahan O., Mryglod I., J. Phys.: Cond. Mat., 2004, 16, L235. 21. Patsahan O., Mryglod I., Condens. Matter Phys., 2004, 7, 755. 22. Patsahan O., Patsahan T., AIP Conference Proceedings, 2009, 1198, 124. 23. Brazovskii S.A., Sov. Phys. JETP, 1975, 41, 8. 24. Leibler L., Macromolecules, 1980, 13, 1602. 25. Ciach A., Góźdź W.T., Stell G., J. Phys. Cond. Mat., 2006, 18, 1629. 26. Ciach A., Góźdź W.T., Annu. Rep. Prog. Chem. C, 2001, 97, 269, and references therein. 27. Podneks V.E., Hamley I.W., Pis’ma Zh. Exp. Teor. Fiz., 1996, 64, 564. Мезоскопiчний опис мережо-формуючих кластерiв слабо заряджених колоїдiв А. Цях., В.Т. Гузьдзь Iнститут фiзичної хiмiї Польської Академiї Наук, Варшава, Республiка Польща В рамках мезоскопiчної польової теорiї вивчаються системи, що складаються iз сферичних заря- джених частинок у розчинниках, якi мiстять контрiони та iндукують короткосяжне притягання. Ми обмежуємося наближенням середнього поля (СП) i слабким впорядкуванням. Ми обговорюємо властивостi потенцiалiв, що складаються iз сильного короткосяжного притягання i слабкого дале- косяжного вiдштовхування (SALR) у контекстi формування неоднорiдного розподiлу частинок на мезоскопiчних масштабах довжин замiсть макроскопiчного фазового вiдокремлення. У попереднiй роботi показано, що формуються сферичнi, цилiндричнi i щiлиноподiбнi кластери частинок, i для достатньо низьких температур цi кластери формують впорядкованi перiодичнi об’ємоцентричнi, ге- ксагональнi i ламеларнi фази. Крiм того, було передбачено гiроїдну фазу, в якiй формуються два переплетенi регулярнi мережоподiбнi кластери, що розгалужуються в потрiйному з’єднаннi. Було знайдено, що при вiдповiдному масштабуваннi густини i температури, лiнiї спiвiснування мiж рiзни- ми впорядкованими фазами є унiверсальними в наближеннi СП, за винятком гiроїдної фази. У цiй статтi визначається фазова дiаграма для двох виборiв потенцiалу SALR: такого, що вiдповiдає ве- ликiй областi притягуючої частини потенцiалу, та такого, що вiдповiдає малiй областi притягання. Ми виявляємо, що область стабiльностi гiроїдної фази дуже слабо залежить вiд форми потенцiалу SALR у межах наближеної теорiї. Ключовi слова: колоїди, кластери, самоорганiзацiя, фазовi переходи лад-безлад 23603-12 Introduction The interaction potential and its properties Field theory First order transition to microsegregated phases One-shell approximation Two-shell approximation Summary Appendix. Explicit expressions of functions and parameters