Extendability of Equilibria of Nematic Polymers

Abstract

The purpose of this paper is to study the extendability of equilibrium states of rodlike nematic polymers with the Maier-Saupe intermolecular potential. We formulate equilibrium states as solutions of a nonlinear system and calculate the determinant of the Jacobian matrix of the nonlinear system. It is found that the Jacobian matrix is nonsingular everywhere except at two equilibrium states. These two special equilibrium states correspond to two points in the phase diagram. One point is the folding point where the stable prolate branch folds into the unstable prolate branch; the other point is the intersection point of the nematic branch and the isotropic branch where the unstable prolate state becomes the unstable oblate state. Our result establishes the existence and uniqueness of equilibrium states in the presence of small perturbations away from these two special equilibrium states.

1. Introduction

Liquid crystal polymers (LCPs) consisting of rigid rodlike macromolecules in a viscous solvent have wide technological applications [1–4]. The most common theoretical framework for modeling nematic polymers is to represent the nematogenic molecules as rigid rods and to describe the ensemble with an orientational probability density function (pdf) [5]. Equilibrium orientational distribution is related to the interaction potential by the Boltzmann relation. Recently, rigorous mathematical analysis on the equilibrium states of nematic polymers has garnered serious attention from mathematicians led by Constantin and Titi [6–22]. For example, various proofs for the axisymmetry of equilibrium states with only the Maier-Saupe interaction were given in [8, 19] for the 2-D case and in [9, 18, 20] for the 3-D case; the equilibrium states for the case where dipole-dipole interaction is coupled to the Maier-Saupe interaction were studied in [17, 21]. Not surprisingly, many mathematical issues of polymers are still unexplored. In particular, to our knowledge, the extendability of the equilibrium states has never been attempted.

In our previous study [21], it was revealed that for large values of the nematic strength and small perturbations, there exists at least one solution near the corresponding unperturbed pure nematic solution. In [21], we concluded the existence using complicated free energy arguments. Moreover, in [21] only existence was established whereas the uniqueness was not addressed at all. Our goal in the current study is to establish rigorously the extendability, both existence and uniqueness, of the equilibrium states of nematic polymers in the presence of small perturbations. Notice that all equilibrium states of nematic polymers are axisymmetric [9, 18, 20]. To facilitate the discussion of extendability, we introduce two new variables for the nonlinear system governing the equilibrium states: one variable is proportional to the order parameter of the pdf; the other variable is proportional to the biaxial order parameter, which measures the deviation of the pdf from being axisymmetric. One advantage of this approach is that the Jacobian matrix is diagonal at equilibrium states. We will show that the Jacobian determinant of the nonlinear system is nonzero everywhere except at two equilibrium states. These two equilibrium states correspond to two points in the phase diagram: one point is the folding point where the stable prolate branch folds into the unstable prolate branch; the other point is the intersection point of the nematic states and the isotropic states where the unstable prolate state becomes the unstable oblate state. The extendability of the equilibrium states except these two states follows immediately from the implicit function theorem.

2. Nonlinear System for Equilibrium States, Jacobian Matrix, and Its Determinant

We now briefly recall the mathematical description for equilibrium states of rigid rodlike nematic polymers. The orientation direction of each polymer rod is denoted by a unit vector 𝐦. In this study, by “pure nematic polymers,” we mean the case where the Maier-Saupe interaction potential is the only potential. The Maier-Saupe potential is given by 𝑈(𝐦)=−𝑏⟨𝐦𝐦⟩∶𝐦𝐦, (2.1) where the tensor product 𝐦𝐦 and the tensor double contraction 𝐀∶𝐁 are defined as𝑚𝐦𝐦≡¹𝑚₁𝑚₁𝑚₂𝑚₁𝑚₃𝑚₂𝑚₁𝑚₂𝑚₂𝑚₂𝑚₃𝑚₃𝑚₁𝑚₃𝑚₂𝑚₃𝑚₃,𝐀∶𝐁≡trace(𝐀𝐁). (2.2) In (2.1), 𝑏=3𝑁/2,𝑁 is the normalized polymer concentration describing the strength of intermolecular interactions, and ⟨𝐦𝐦⟩ is the second moment of the orientation distribution:⟨𝐦𝐦⟩≡_‖𝐦‖=1𝐦𝐦𝜌(𝐦)𝑑𝐦, (2.3)where 𝜌(𝐦) is the orientational probability density function (pdf) of the ensemble of polymer rods. For convenience, potential (2.1) has been normalized with respect to 𝑘_𝐵𝑇, where 𝑘_𝐵 is the Boltzmann constant and 𝑇 the absolute temperature. For pure nematic polymers, equilibrium states are described by the Boltzmann distribution [5]:1𝜌(𝐦)=𝑍exp[−𝑈(𝐦)], (2.4)where 𝑍=∫_𝑆exp[−𝑈(𝐦)]𝑑𝐦 is the partition function and 𝑆 is the unit sphere.

We choose the coordinate system such that the second moment ⟨𝐦𝐦⟩ is diagonal:⟨𝐦𝐦⟩=⟨𝑚²₁⟩000⟨𝑚²₂⟩000⟨𝑚²₃⟩. (2.5)As a consequence, the Maier-Saupe potential can be written as𝑈(𝐦)=−𝑏(⟨𝑚²₁⟩𝑚²₁+⟨𝑚²₂⟩𝑚²₂+⟨𝑚²₃⟩𝑚²₃). (2.6)The most significant conclusion for pure nematic polymers is that all equilibrium states are axisymmetric [9, 18, 20]. Since not all equilibrium states of a perturbed nematic polymer ensemble are necessarily axisymmetric, to study the extendability, we formulate the problem without using the axisymmetry so that non-axisymmetric perturbations are allowed. The axisymmetry will be useful in our analysis because the extendability is determined by the Jacobian determinant evaluated at the unperturbed equilibrium state, which is axisymmetric. The definition and derivation of the Jacobian matrix, however, require non-axisymmetric formulation. For pure nematic polymers, the total potential is completely specified by the second moment ⟨𝐦𝐦⟩. As a result of the Boltzmann relation (2.4), an equilibrium state is completely specified by the second moment ⟨𝐦𝐦⟩. Because of the constraint 𝑚²₁+𝑚²₂+𝑚²₃=1, an equilibrium state is completely specified by𝑠₁≡⟨𝑚²₁⟩,𝑠₂≡⟨𝑚²₂⟩. (2.7)At a pure nematic equilibrium state without perturbation, we have 𝑠₁=𝑠₂ or 𝑠₁=1−𝑠₁−𝑠₂ (which means 𝑠₁=𝑠₃) or 𝑠₂=1−𝑠₁−𝑠₂ (which means 𝑠₂=𝑠₃). At an equilibrium state, we select the coordinate system such that the equilibrium satisfies 𝑠₁=𝑠₂. In terms of (𝑠₁,𝑠₂), the Maier-Saupe potential can be expressed as𝑈(𝐦)=−𝑏(𝑠₁𝑚²₁+𝑠₂𝑚²₂+(1−𝑠₁−𝑠₂)𝑚²₃)1=−𝑏[2(𝑠₂−𝑠₁)(𝑚²₂−𝑚²₁3)+(1−2(𝑠₁+𝑠₂))𝑚²₃]+const. (2.8)Consequently, the equilibrium pdf is given by1𝜌(𝐦)=𝑍1exp{𝑏[2(𝑠₂−𝑠₁)(𝑚²₂−𝑚²₁3)+(1−2(𝑠₁+𝑠₂))𝑚²₃]}. (2.9)At an equilibrium state, (𝑠₁,𝑠₂) satisfies the nonlinear system:𝑠₁−⟨𝑚²₁⟩=0,𝑠₂−⟨𝑚²₂⟩=0. (2.10)The form of the equilibrium pdf (2.9) motivates us to introduce (𝜂₁,𝜂₂):𝜂₁3≡𝑏[1−2(𝑠₁+𝑠₂)],𝜂₂≡𝑏(𝑠₂−𝑠₁)2. (2.11)Here 𝜂₁ is proportional to the order parameter of the pdf while 𝜂₂ is proportional to the biaxial order parameter, which measures the deviation of the pdf from being axisymmetric. As we will see, one advantage of using (𝜂₁,𝜂₂) is that the Jacobian matrix is diagonal at equilibrium states. In terms of (𝜂₁,𝜂₂), the pdf has the expression𝜌(𝐦,𝜂₁,𝜂₂)=exp[𝜂₂(𝑚²₂−𝑚²₁)+𝜂₁𝑚²₃]∫_𝑆exp[𝜂₂(𝑚²₂−𝑚²₁)+𝜂₁𝑚²₃]𝑑𝐦. (2.12)The second advantage of using (𝜂₁,𝜂₂) is that the pdf does not depend on 𝑏 explicitly. Later on, this property will enable us to write 𝑏 as a function of 𝑟 which is the equilibrium value of 𝜂₁. At an equilibrium state, the nonlinear system for (𝜂₁,𝜂₂) is𝐹₁(𝜂₁,𝜂₂𝜂;𝑏)≡¹𝑏−12(3⟨𝑚²₃𝐹⟩−1)=0,₂(𝜂₁,𝜂₂𝜂;𝑏)≡²𝑏−12⟨𝑚²₂−𝑚²₁⟩=0. (2.13)In this paper, we study the extendability of equilibrium states of nematic polymers in the presence of small perturbations. We consider a perturbed version of system (2.13):𝐹₁(𝜂₁,𝜂₂;𝑏)=𝜖₁,𝐹₂(𝜂₁,𝜂₂;𝑏)=𝜖₂. (2.14)Mathematically, we study the existence and uniqueness of solution of system (2.14) near a solution of system (2.13) for small 𝜖₁ and 𝜖₂. According to the implicit function theorem, the existence and uniqueness are determined by the Jacobian determinant of (𝐹₁(𝜂₁,𝜂₂),𝐹₂(𝜂₁,𝜂₂)). To calculate the Jacobian matrix, we first calculate the derivatives of pdf (2.12):𝜕𝜕𝜂₁𝜌(𝐦,𝜂₁,𝜂₂)=(𝑚²₃−⟨𝑚²₃⟩)𝜌(𝐦,𝜂₁,𝜂₂𝜕),𝜕𝜂₂𝜌(𝐦,𝜂₁,𝜂₂)=(𝑚²₂−𝑚²₁−⟨𝑚²₂−𝑚²₁⟩)𝜌(𝐦,𝜂₁,𝜂₂). (2.15)Recall that at a pure nematic equilibrium state without perturbation, we have selected our coordinate system such that 𝜂₂=0. Evaluating the partial derivatives of 𝐹₁ and 𝐹₂ at the equilibrium (𝜂₁=𝑟,𝜂₂=0) yields𝜕𝜕𝜂₁𝐹₁(𝜂₁,𝜂₂;𝑏)|_{𝜂1=𝑟,𝜂2=0}=1𝑏−32⟨𝑚²₃(𝑚²₃−⟨𝑚²₃1⟩)⟩=𝑏−32var(𝑚²₃𝜕),𝜕𝜂₂𝐹₁(𝜂₁,𝜂₂;𝑏)|_{𝜂1=𝑟,𝜂2=0}3=−2⟨𝑚²₃(𝑚²₂−𝑚²₁−⟨𝑚²₂−𝑚²₁𝜕⟩)⟩=0,𝜕𝜂₁𝐹₂(𝜂₁,𝜂₂;𝑏)|_{𝜂1=𝑟,𝜂2=0}1=−2⟨(𝑚²₂−𝑚²₁)(𝑚²₃−⟨𝑚²₃𝜕⟩)⟩=0,𝜕𝜂₂𝐹₂(𝜂₁,𝜂₂;𝑏)|_{𝜂1=𝑟,𝜂2=0}=1𝑏−12⟨(𝑚²₂−𝑚²₁)(𝑚²₂−𝑚²₁−⟨𝑚²₂−𝑚²₁=1⟩)⟩,𝑏−12⟨(𝑚²₂−𝑚²₁)²⟩. (2.16)Here we point out that the Jacobian matrix of system (2.13) evaluated at an equilibrium state is diagonal, which is caused by the axisymmetry of the equilibrium state.

It follows that the determinant of Jacobian matrix of system (2.13) at the equilibrium isdet(𝜕(𝐹₁,𝐹₂)𝜕(𝜂₁,𝜂₂))|_{𝜂1=𝑟,𝜂2=0}1=(𝑏−32var(𝑚²₃1))⋅(𝑏−12⟨(𝑚²₂−𝑚²₁)²⟩). (2.17)In (2.17), all averages are evaluated using the equilibrium pdf𝜌(𝐦,𝜂₁,𝜂₂)|_{𝜂1=𝑟,𝜂2=0}=1𝑍exp(𝑟𝑚²₃). (2.18)For pure nematic polymers, an equilibrium state is completely specified by 𝑟. The governing equation for 𝑟 is the first equation of (2.13): 𝐹₁(𝑟,0;𝑏)=0, whereas the second equation of (2.13) is satisfied automatically when 𝜂₂=0. We introduce a new function:𝐹(𝜂;𝑏)≡𝐹₁𝜂(𝜂,0;𝑏)=𝑏−12(3⟨𝑚²₃⟩−1). (2.19)The governing equation for 𝑟 is 𝐹(𝑟;𝑏)=0. It follows directly from the definition that function 𝐹(𝜂;𝑏) is related to the (1,1) element of the Jacobian matrix at equilibrium as𝜕𝜕𝜂₁𝐹₁(𝜂₁,𝜂₂;𝑏)|_{𝜂1=𝑟,𝜂2=0}=𝜕𝜕𝜂𝐹(𝜂;𝑏)|_𝜂=𝑟. (2.20)As we will find later, this relation is a key tool for determining the sign of the (1,1) element of the Jacobian matrix at equilibrium. In the above, we have used variables 𝜂 and 𝑟 more or less interchangibly. Later on when necessary, we will use 𝜂 to denote the independent variable of functions and use 𝑟 to denote the particular value of 𝜂 that satisfies 𝐹(𝑟;𝑏)=0. For that purpose, we will continue using these two variables.

3. Extendability of Equilibria of Nematic Polymers

To obtain more specific properties of equilibrium states of nematic polymers, we rewrite the governing equation for 𝑟 using spherical coordinates. We select the 𝑚₃-axis as the pole of the spherical coordinate system. The equilibrium pdf given in (2.18) becomes𝜌(𝜙,𝜃;𝑟)=exp(𝑟cos²𝜙)2𝜋∫^𝜋₀exp(𝑟cos²𝜙)sin𝜙𝑑𝜙, (3.1)where 𝜙 is the polar angle and 𝜃 the azimuthal angle. Here we need to point out that 𝑟 is not the radius in the spherical coordinate system. It is a parameter of the equilibrium state. Substituting equilibrium pdf (3.1) into (1/2)(3⟨𝑚²₃⟩−1), using a change of variable 𝑢=cos𝜙, and integrating by parts, we have12(3⟨𝑚²₃∫⟩−1)=𝑟⋅¹⁰(𝑢²−𝑢⁴)exp(𝑟𝑢²)𝑑𝑢∫¹₀exp(𝑟𝑢²)𝑑𝑢. (3.2)Thus, 𝐹(𝜂;𝑏) defined in (2.19) becomes 𝐹(𝜂;𝑏)=𝜂[1/𝑏−𝑓(𝜂)], where 𝑓(𝜂) is defined as∫𝑓(𝜂)≡¹⁰𝑢²(1−𝑢²)exp(𝜂𝑢²)𝑑𝑢∫¹₀exp(𝜂𝑢²)𝑑𝑢. (3.3)The equation for 𝑟, 𝐹(𝑟;𝑏)=0, becomes1𝑟[𝑏−𝑓(𝑟)]=0. (3.4)Equation (3.4) has a trivial solution 𝑟=0 for arbitrary value of 𝑏, which corresponds to the isotropic branch of the nematic polymer phase diagram. At 𝑟=0, the equilibrium pdf is uniform, 𝜌(𝜙,𝜃;0)=1/4𝜋. It follows that⟨𝑚²_𝑗1⟩=3,⟨𝑚⁴_𝑗⟩=⟨𝑚⁴₃⟩=¹₀𝑢⁴1𝑑𝑢=5,⟨𝑚²_𝑖𝑚²_𝑗⟩=⟨𝑚²₁𝑚²₃1⟩=2¹₀𝑢²(1−𝑢²1)𝑑𝑢=15,𝑖≠𝑗,var(𝑚²₃)=⟨𝑚⁴₃⟩−⟨𝑚²₃⟩²=15−19=4,45⟨(𝑚²₂−𝑚²₁)²⟩=⟨𝑚⁴₂⟩−2⟨𝑚²₁𝑚²₂⟩+⟨𝑚⁴₁2⟩=5−2=415.15 (3.5)Substituting these results into (2.16), we obtain𝜕𝜕𝜂₁𝐹₁(𝜂₁,𝜂₂;𝑏)|_{𝜂1=0,𝜂2=0}=1𝑏−32var(𝑚²₃1)=𝑏−2,𝜕15𝜕𝜂₂𝐹₂(𝜂₁,𝜂₂;𝑏)|_{𝜂1=0,𝜂2=0}=1𝑏−12⟨(𝑚²₂−𝑚²₁)²1⟩=𝑏−2.15 (3.6)Therefore, when 𝑏≠15/2, the Jacobian matrix is nonsingular and thereby the isotropic equilibrium is extendable. At 𝑏=15/2, the isotropic branch intersects with the nematic branch.

The remaining solutions of (3.4), if any, satisfy1𝑏−𝑓(𝑟)=0. (3.7) In [20], it has been shown that the function 𝑓(𝜂) has the following properties: (1)𝑓(0)=2/15, lim_𝜂→+∞𝑓(𝜂)=0 and lim_{𝜂→−∞}𝑓(𝜂)=0;(2)𝑓(𝜂) attains its maximum at 𝜂=𝜂^∗=2.1782879748>0, where the maximum is 𝑓(𝜂^∗)=0.14855559992254>0;(3)𝑓^′(𝜂)>0 for 𝜂<𝜂^∗ and 𝑓^′(𝜂)<0 for 𝜂>𝜂^∗. A graph of 𝑓(𝜂) is shown in Figure 1(a). Thus, we draw conclusions below for (3.7) as follows. (i)The critical value of 𝑏 is 𝑏^∗=1/𝑓(𝜂^∗)=6.7314863965.(ii)For 𝑏<𝑏^∗, (3.7) has no solution.(iii)At 𝑏=𝑏^∗, (3.7) has one solution 𝑟(𝑏^∗)=𝜂^∗>0.(iv)For 𝑏^∗<𝑏<2/15, (3.7) has two solutions: 𝑟_𝑈(𝑏)>𝜂^∗>0 and 0<𝑟_𝑀(𝑏)<𝜂^∗.(v)At 𝑏=2/15, (3.7) has two solutions: 𝑟_𝑈(2/15)>𝜂^∗>0 and 𝑟(2/15)=0.(vi)For 𝑏>2/15, (3.7) has two solutions: 𝑟_𝑈(𝑏)>𝜂^∗>0 and 𝑟_𝐿(𝑏)<0. As mentioned before, we use subscript “𝑈” to denote to the “Upper” part of the phase diagram where 𝑟>𝜂^∗, subscript “𝑀” to denote to the “Middle” part of the phase diagram, where 0<𝑟<𝜂^∗, and subscript “𝐿” to denote to the “Lower” part of the phase diagram, where 𝑟<0. The phase diagram for nematic polymers is shown in Figure 1(b). In the terminologies of nematic polymers, curve segment 𝑟_𝑈(𝑏) is the stable part of the prolate branch; 𝑟_𝑀(𝑏) is the unstable part of the prolate branch; and 𝑟_𝐿(𝑏) is the unstable oblate branch.

Recall that 𝑏 is the strength of the Maier-Saupe interaction, which is proportional to the normalized polymer concentration and is inversely proportional to the temperature. 𝑟, the solution of 𝐹(𝑟;𝑏)=0, has the expression 𝑟=𝑏⋅(1/2)(3⟨𝑚²₃⟩−1), where (1/2)(3⟨𝑚²₃⟩−1) is the order parameter. Thus, from physical considerations, it is desirable to use 𝑏 as the independent variable and treat 𝑟 as a function of 𝑏. However, 𝑟(𝑏) is a multivalue function of 𝑏 and for 𝑏<𝑏^∗ function 𝑟(𝑏) is not even defined. Mathematically, it is much more convenient if we use 𝑟 as the independent variable and treat 𝑏 as a function of 𝑟. 𝑏(𝑟) is a single-value function of 𝑟 and is defined for all values of 𝑟 in (−∞,+∞). Below, we will adopt this new formulation of viewing 𝑏 as a function of 𝑟. In the new formulation, function 𝑏(𝑟) is determined from (3.7) as 𝑏(𝑟)=1/𝑓(𝑟). In terms of this new formulation, the stable part of the prolate branch 𝑟_𝑈(𝑏) can be simply represented as 𝑏(𝑟) for 𝑟>𝜂^∗; the unstable part of the prolate branch 𝑟_𝑀(𝑏) as 𝑏(𝑟) for 0<𝑟<𝜂^∗; and the unstable oblate branch 𝑟_𝐿(𝑏) as 𝑏(𝑟) for 𝑟<0. Now we discuss the extendability of these branches. Using relation (2.20), 𝐹(𝜂;𝑏)=𝜂[1/𝑏−𝑓(𝜂)], and 1/𝑏−𝑓(𝑟)=0, we arrive at𝜕𝜕𝜂₁𝐹₁(𝜂₁,𝜂₂;𝑏(𝑟))|_{𝜂1=𝑟,𝜂2=0}=𝜕𝜕𝜂𝐹(𝜂;𝑏(𝑟))|_𝜂=𝑟1=[𝑏(𝑟)−𝑓(𝜂)]|_𝜂=𝑟−𝜂𝑓^′(𝜂)|_𝜂=𝑟=−𝑟𝑓^′(𝑟)=>0,for𝑟>𝜂^∗<0,for0<𝑟<𝜂^∗>0,for𝑟<0. (3.8)To study the (2,2) element of Jacobian matrix, we first rewrite (1/2)⟨(𝑚²₂−𝑚²₁)²⟩ by substituting equilibrium pdf (3.1) into (1/2)⟨(𝑚²₂−𝑚²₁)²⟩ and using substitution 𝑢=cos𝜙,12⟨(𝑚²₂−𝑚²₁)²1⟩=4∫¹₀(1−𝑢²)²exp(𝑟𝑢²)𝑑𝑢∫¹₀exp(𝑟𝑢²)𝑑𝑢. (3.9)Using (2.16), 𝑏(1)=1/𝑓(𝑟), the expression of 𝑓(𝑟) given in (3.3), and result (3.9), we have𝜕𝜕𝜂₂𝐹₂(𝜂₁,𝜂₂;𝑏(𝑟))|_{𝜂1=𝑟,𝜂2=0}=1−1𝑏(𝑟)2⟨(𝑚²₂−𝑚²₁)²⟩=14∫¹₀(1−𝑢²)(5𝑢²−1)exp(𝑟𝑢²)𝑑𝑢∫¹₀exp(𝑟𝑢²)𝑑𝑢≡𝑔(𝑟). (3.10)It is straightforward to verify that 𝑔(0)=0. Below, we want to show that 𝑔(𝑟)>0 for 𝑟>0 and 𝑔(𝑟)<0 for 𝑟<0. To facilitate the analysis below, we write 𝑔(𝑟) as an average1𝑔(𝑟)=4⟨(1−𝑢²)(5𝑢²−1)⟩, (3.11)where the average is with respect to the pdf 𝜌(𝑢;𝑟)=exp(𝑟𝑢²)/∫¹₀exp(𝑟𝑢²)𝑑𝑢.

Lemma 3.1. Function 𝑔(𝑟) given in (3.11) has the property that 𝑔(𝑟)=0 implies 𝑔^′(𝑟)>0.

Proof. We first calculate the derivative of the pdf: (𝜕/𝜕𝑟)𝜌(𝑢;𝑟)=(𝑢²−⟨𝑢²⟩)𝜌(𝑢;𝑟), which leads to 𝑔^′1(𝑟)=4⟨(1−𝑢²)(5𝑢²−1)(𝑢²−⟨𝑢²=1⟩)⟩20⟨(1−𝑢²)(5𝑢²−1)²1⟩+5(1−5⟨𝑢²⟩)𝑔(𝑟). (3.12) Thus, whenever 𝑔(𝑟)=0, we have 𝑔^′(𝑟)=(1/20)⟨(1−𝑢²)(5𝑢²−1)²⟩>0.

Lemma 3.1 together with 𝑔(0)=0 leads to 𝑔(𝑟)>0 for 𝑟>0 and 𝑔(𝑟)<0 for 𝑟<0. Combining this result on 𝑔(𝑟) with result (3.8) on −𝑟𝑓^′(𝑟), we arrive atdet(𝜕(𝐹₁,𝐹₂)𝜕(𝜂₁,𝜂₂))|_{𝜂1=𝑟,𝜂2=0}=(−𝑟𝑓^′(𝑟))𝑔(𝑟)=>0,for𝑟>𝜂^∗,<0,for0<𝑟<𝜂^∗,<0,for𝑟<0. (3.13)Therefore, we conclude that all nematic equilibrium states (stable or unstable) are extendable except for the equilibrium state at (𝑟=0,𝑏=15/2) and the equilibrium state at (𝑟=𝜂^∗,𝑏=𝑏^∗).

4. Conclusions

In this work, we studied the extendability of equilibrium states of nematic polymers with the Maier-Saupe intermolecular potential. We found that the Jacobian matrix of the nonlinear system is nonsingular except at two special equilibrium states. The significance of this result is its implication on the existence and uniqueness of equilibrium states of a perturbed system, in the neighborhood of the unperturbed equilibrium states.

(a)

(b)

Figure 1 

(a) Graph of function 𝑓(𝜂); (b) phase diagram of nematic polymers.

Acknowledgments

H. Wang was partially supported by the National Science Foundation. H. Zhou was supported in part by the Air Force Office of Scientific Research under Grant no. F1ATA06313G003.