Processing math: 60%
Research article

Layered solutions for a nonlocal Ginzburg-Landau model with periodic modulation

  • Received: 03 March 2022 Revised: 21 April 2023 Accepted: 12 May 2023 Published: 26 May 2023
  • We study layered solutions in a one-dimensional version of the scalar Ginzburg-Landau equation that involves a mixture of a second spatial derivative and a fractional half-derivative, together with a periodically modulated nonlinearity. This equation appears as the Euler-Lagrange equation of a suitably renormalized fractional Ginzburg-Landau energy with a double-well potential that is multiplied by a 1-periodically varying nonnegative factor g(x) with 101g(x)dx<. A priori this energy is not bounded below due to the presence of a nonlocal term in the energy. Nevertheless, through a careful analysis of a minimizing sequence we prove existence of global energy minimizers that connect the two wells at infinity. These minimizers are shown to be the classical solutions of the associated nonlocal Ginzburg-Landau type equation.

    Citation: Ko-Shin Chen, Cyrill Muratov, Xiaodong Yan. Layered solutions for a nonlocal Ginzburg-Landau model with periodic modulation[J]. Mathematics in Engineering, 2023, 5(5): 1-52. doi: 10.3934/mine.2023090

    Related Papers:

    [1] José Antonio Vélez-Pérez, Panayotis Panayotaros . Wannier functions and discrete NLS equations for nematicons. Mathematics in Engineering, 2019, 1(2): 309-326. doi: 10.3934/mine.2019.2.309
    [2] Federico Cluni, Vittorio Gusella, Dimitri Mugnai, Edoardo Proietti Lippi, Patrizia Pucci . A mixed operator approach to peridynamics. Mathematics in Engineering, 2023, 5(5): 1-22. doi: 10.3934/mine.2023082
    [3] Qiang Du, Tadele Mengesha, Xiaochuan Tian . $ L^{p} $ compactness criteria with an application to variational convergence of some nonlocal energy functionals. Mathematics in Engineering, 2023, 5(6): 1-31. doi: 10.3934/mine.2023097
    [4] Luca Formaggia, Alessio Fumagalli, Anna Scotti . A multi-layer reactive transport model for fractured porous media. Mathematics in Engineering, 2022, 4(1): 1-32. doi: 10.3934/mine.2022008
    [5] Luca Franzoi, Riccardo Montalto . Time almost-periodic solutions of the incompressible Euler equations. Mathematics in Engineering, 2024, 6(3): 394-406. doi: 10.3934/mine.2024016
    [6] Masashi Misawa, Kenta Nakamura, Yoshihiko Yamaura . A volume constraint problem for the nonlocal doubly nonlinear parabolic equation. Mathematics in Engineering, 2023, 5(6): 1-26. doi: 10.3934/mine.2023098
    [7] Giacomo Canevari, Arghir Zarnescu . Polydispersity and surface energy strength in nematic colloids. Mathematics in Engineering, 2020, 2(2): 290-312. doi: 10.3934/mine.2020015
    [8] Carmen Cortázar, Fernando Quirós, Noemí Wolanski . Decay/growth rates for inhomogeneous heat equations with memory. The case of large dimensions. Mathematics in Engineering, 2022, 4(3): 1-17. doi: 10.3934/mine.2022022
    [9] Christopher Chong, Andre Foehr, Efstathios G. Charalampidis, Panayotis G. Kevrekidis, Chiara Daraio . Breathers and other time-periodic solutions in an array of cantilevers decorated with magnets. Mathematics in Engineering, 2019, 1(3): 489-507. doi: 10.3934/mine.2019.3.489
    [10] Dmitrii Rachinskii . Bifurcation of relative periodic solutions in symmetric systems with hysteretic constitutive relations. Mathematics in Engineering, 2025, 7(2): 61-95. doi: 10.3934/mine.2025004
  • We study layered solutions in a one-dimensional version of the scalar Ginzburg-Landau equation that involves a mixture of a second spatial derivative and a fractional half-derivative, together with a periodically modulated nonlinearity. This equation appears as the Euler-Lagrange equation of a suitably renormalized fractional Ginzburg-Landau energy with a double-well potential that is multiplied by a 1-periodically varying nonnegative factor g(x) with 101g(x)dx<. A priori this energy is not bounded below due to the presence of a nonlocal term in the energy. Nevertheless, through a careful analysis of a minimizing sequence we prove existence of global energy minimizers that connect the two wells at infinity. These minimizers are shown to be the classical solutions of the associated nonlocal Ginzburg-Landau type equation.



    In this paper, we consider minimization of the following nonlocal energy functional:

    J(u):=α2R|u|2dx+Rg(x)W(u)dx+β4πRR[(u(x)u(y))2(xy)2(η(x)η(y))2(xy)2]dydx (1.1)

    in the set

    A={uH1loc(R):uηH1(R)}.

    Here α,β are positive constants, g(x) is a 1-periodic (a general period T>0 can be treated similarly) nonnegative function whose reciprocal satisfies the following integrability assumption

    101g(x)dx<. (1.2)

    The fixed function ηC(R) satisfies |η|1, η(x)=1 for x1, η(x)=1 for x1, W(u) is a double well potential satisfying

    W(u)>0 if u±1,W(±1)=W(±1)=0andW(±1)>0. (1.3)

    Formally, for uC2loc(R)L(R) the Euler-Lagrange equation associated with (1.1) is

    αu+β(d2dx2)12u+g(x)W(u)=0xR, (1.4)

    where

    (d2dx2)12u(x):=limε01π|xy|εu(x)u(y)(xy)2dy. (1.5)

    We are mainly interested in solutions of (1.4) that satisfy

    limx±u(x)=±1, (1.6)

    we call such solutions layered solutions.

    Equation (1.4) is a special case of the more general equation

    αΔu+β(Δ)su+g(x)W(u)=0xRn, (1.7)

    where 0<s<1. For uC2loc(Rn)L(Rn) the operator (Δ)s is the fractional Laplacian defined by

    (Δ)su(x):=Cn,sP.V.Rnu(x)u(y)|xy|n+2sdy=Cn,slimε0|xy|εu(x)u(y)|xy|n+2sdy,

    where Cn,s is a normalization constant to guarantee that the Fourier symbol of the resulting operator is |ξ|2s, see e.g., [25], Section 3 for more details.

    When g(x)=γ is a constant, (1.7) reduces to

    αΔu+β(Δ)su+γW(u)=0xRn. (1.8)

    This type of equation has attracted a lot of attention over the last twenty years (see e.g., [17,25,26,27,28,47,49,53,64]). In particular, the structure of layered solutions in the case β=0 (Allen-Cahn) or α=0 (fractional Allen-Cahn) is well understood at present. Here a layered solution of (1.8) is a bounded solution which is monotone in one direction. When β=0, De Giorgi conjecture posits that the level sets of such a layered solution are hyperplanes for n8. De Giorgi's initial conjecture was for W(u)=14(1u2)2. This conjecture was proved for any C2 function W(u) satisfying (1.3) by Ghoussoub and Gui [42] when n=2. When n=3, Ambrosio and Cabré [10] proved the conjecture for a large class of W(u) which includes the original De Giorgi's choice. Later, Alberti, Ambrosio and Cabré [2] extended their results to cover all C2 function W(u) with the properties specified in (1.3). Under the additional assumption of anti-symmetry of solutions, Ghoussoub and Gui [43] established the De Giorgi conjecture for n=4,5. Further developments on the conjecture can be found in [12]. De Giorgi conjecture was completely solved by Savin [60,61] for 4n8 under the additional assumption limxn±u(x)=±1. For dimensions n9, a counter-example was constructed by Del Pino, Kowalczyk and Wei [31]. A weaker version of the De Gigorgi conjecture, known as Gibbons conjecture, replaced monotonicity assumption by the stronger condition

    limxn±u(x)=±1 uniformly for (x1,,xn1)Rn1. (1.9)

    Gibbons conjecture was proved in all dimensions [12,13,40].

    De Giorgi's conjecture has also been extended to the fractional Allen-Cahn case. The fractional De Giorgi conjecture was proved in [25,26,27,63] for the case n=2,s(0,1), and in [22,23] for n=3 and s12. Under additional limit conditions, fractional De Giorgi conjecture was proved for n=3 and s(0,12) by Dipierro, Serra and Valdinoci in [37] and by Savin in [62] for 4n8 and s[12,1). The limit condition is removed in [34] for n=3 and s(0,12). Recently, Figalli and Serra [41] solved the De Giorgi conjecture for half-Laplacian when n=4 (such a result is not known for the classical case s=1). Based on all these results, when g(x) is a constant, solutions to (1.8) satisfying (1.6) reduces to the unique one-dimensional solution (modulo translation) which is monotone and the problem is essentially one-dimensional.

    When g(x) is not constant, but rather periodic, the continuous translational symmetry of layered solutions of (1.7) is broken and the structure of the set of solutions is much more complex. When β=0, the nonautonomous Allen-Cahn equation

    Δu+Wu(x,u)=0xRn  (1.10)

    with

     W(x+k,u)=W(x,u)kZn

    has been studied extensively over the last three decades. Equation (1.10) is a special case of a model problem initiated by Moser [48] for developing a PDE version of Aubry-Mather theory of monotone twist maps (see [11,14,15,57,58] for related work). A different motivation is to view (1.10) as a model for phase transitions. When n=1 and subject to homogeneous Neumann boundary conditions on the interval of x(0,1), the following results have been proved for (1.10) for various choices of the potential term W(x,u). Angenent, Mallet-Paret and Peletier [1] gave a complete classification of all stable equilibrium solutions to (1.10) for Wu(x,u)=u(1u)(ua(x)). Existence and stability of equilibrium solutions with a single transiton layer is proved in [44] for a general class of Wu(x,u)=f(x,u), with f satisfying f(x,0)=f(x,1)=0 and f(x+k,u)=f(x,u) for some k>0. Nakashima [50] proved existence of stable solutions with multiple transition layers for the case Wu(x,u)=(ua(x))(ub(x))(uc(x)). Existence and stability of multilayered solutions were provided in [51] for Wu(x,u)=h2(x)f(u). Nakashima and Tanaka [52] studied the one-dimensional case with a general potential W(x,u) and obtained existence of solutions with clustering layers. For higher dimensions and the special case of

    Δu+a(x)W(u)=0x=(x1,,xn)Rn, (1.11)

    Alessio, Jeanjean and Montecchiari [5] proved existence of infinitely many solutions which are distinct up to periodic translations and satisfy limx1±u(x1,x2)=±b uniformly in x2 for the case n=2 when a(x1,x2) is a positive, even, periodic function in x1,x2, and W(u) is a double well potential vanishing at u=±b for some b>0. For the same equation, Alessio and Montecchiari [7] showed existence of brake orbits type solutions, and Alessio, Gui and Montecchiari [3] proved existence and asymptotic behavior of saddle solutions. When a(x1,x2) depends only on one variable, existence of two-dimensional solutions was proved by Alessio and Montecchiari [4], and existence of infinitely many solutions can be found in [6,66]. Alessio and Montecchiari [8] proved existence of infinitely many solutions verifying limx1±u(x1,x2,x3)=±1 uniformly in (x2,x3) for n=3 and a=a(x1). For results on solutions to (1.10) for n=2 with general potentials, see the papers by Rabinowitz and Stredulinsky [55,56]. Existence of various (multi-layer, mountain pass or higher topological complexity) solutions to (1.11) for general n was obtained in a series of papers by Byeon and Rabinowitz [18,19,20,21]. A review on existence results for (1.10) is given in [54] (see book [59] for a more thorough review on extensions of Moser-Bangert theory).

    An extensive discussion on moving front solutions for time-dependent inhomogeneous Allen-Cahn equation can also be found in the literature. For example, Xin [65] considered propagating front solutions (which include stationary layered solutions) for

    ut=x(a(x)xu)+b(x)xu+f(u) (1.12)

    when a(x), b(x) are periodic and f(u) is bistable. Keener [46] studied propagation of waves in periodic media for the following model:

    ut=uxx+(1+g(xL))f(u)au (1.13)

    where g(x) is a 1-periodic function and obtained a nearly complete picture of propagation in periodic medium. In particular, his results show how wave front shape changes when the medium becomes more and more nonuniform, and how propagation failure occurs when the medium becomes sufficiently nonuniform. Pinning and de-pinning phenomena for front propagation in heterogenous media was discussed in [38]. Existence and qualitative properties of pulsating travelling wave solutions is proved in [33] for equation

    ut=(a(xL)ux)x+f(x,u). (1.14)

    where a(x) and f(x,u) are 1-periodic in x.

    Studies of layered solution in the fractional case when g(x) is not constant caught less attention. Existence of layered solutions to

    (d2dx2)su+g(x)W(u)=0xR (1.15)

    was obtained in [45] for s[12,1) when g>0 is an even, periodic function and W(u) is odd. Another related work is [39], where the authors studied existence of multi-layered solution to the following equation:

    ε2s(d2dx2)su+g(x)u(1u2)=0xR, (1.16)

    when g(x) is not constant and s(12,1). Existence of heteroclinic orbits was proved in [9] for equation

    (Δ)su+g(ϵx)W(u)=0xR,s(12,1), (1.17)

    in each of the three cases of g, namely, g (asymptotically) periodic, g coercive and g satisfying Rabinowitz's condition. For a more general nonlocal operator in the form

    Lu(x)=P.V.R[u(x)u(y)]K(xy)dy (1.18)

    where

    θ0|x|1+2sχ[0,r0](x)K(x)Θ0|x|1+2s for some Θ0θ0>0 and r0>0, (1.19)

    heteroclinic orbits were constructed for s(12,1] in [35] (see also [30]) and for s(14,12] in a later work [36] for the nonlocal equation

    Lu+g(x)W(u)=0xR (1.20)

    with an oscillatory g.

    The mixed case, with local and nonlocal operators (α0 and β0), has gained attention in the last years at least in the autonomous case. For example, layered solutions of Allen-Cahn type equations in the form of a sum of fractional Laplacians of different orders was addressed in [24]. Systematic study on regularity and maximum principles for mixed local and nonlocal operators in the form L=Δ+(Δ)s has recently been developed in [16]. The work in the current paper is partly motivated by a recent work by the authors [29] where we considered the following renormalized nonlocal Ginzburg-Landau energy

    Eε(u)=Rε2|u|2dx+RW(u)dx+RR[(u(x)u(y))2(xy)2(η(x)η(y))2(xy)2]dydx. (1.21)

    We proved existence, regularity, monotonicity and uniqueness (up to translation) of the minimizer of Eε(u) in A. Moreover, as ε0 we recovered the solution in [53] as the global minimizer (unique up to translations) of

    E0(u)=RW(u)dx+RR[(u(x)u(y))2(xy)2(η(x)η(y))2(xy)2]dydx. (1.22)

    The proof of existence and uniqueness of minimizers in [29] relies on an essential observation that a minimizer of Eε among all functions satisfying uηW1,20 (I) on any sufficiently large fixed interval I is monotone. Such conclusion follows from the key assumption that (1.21) is translation invariant. For model (1.1) with only discrete translation invariance, this argument fails and we need to seek a new method. The main difficulty to prove the existence of minimizer of (1.1) lies in two parts. Firstly, since η˚H12(R), it is not a priori clear that J(u) is bounded from below on A. Secondly, the energy bound does not necessarily imply the boundedness of u in a suitable Sobolev space in general. Therefore, we cannot a priori apply the direct method of calculus of variations to obtain a minimizer. To show that J is bounded from below on A, we divide the real line into the regions where u is close to ±1 and where u is away from ±1. By carefully matching the contributions from each region, all negative parts of the potential infinite energy are cancelled out.

    To prove the existence of a minimizer, our main idea is as follows. Given an arbitrary minimizing sequence {un}, we replace this sequence by another sequence {¯un} constructed via reflecting the negative parts of un outside suitable regions. Taking into account our energy estimates from the lower bound argument, we can carefully choose the region where we apply the reflection to un so that the energy J(¯un) differs only slightly from J(un). The sequence {¯un} satisfies |¯un(x)+sgn(x)|c>0 outside a uniformly bounded interval. For such a sequence, boundedness of energy implies boundedness of ¯unη in H1(R). From this and a lower semicontinuity argument, we obtain a limit function which attains a minimum of J(u) in A.

    Our main result is the following existence and regularity theorem.

    Theorem 1.1. Let α,β be positive constants. Assume gC(R) is a nonnegative 1-periodic function which satisfies (1.2), ηC(R) is a given function satisfying |η|1, η(x)=1 for x1, η(x)=1 for x1, W(u) is a double well potential satisfying (1.3). Then there exists a minimizer u0 of J(u) over A. Moreover, u0C2,12(R)L(R) and satisfies the Euler-Lagrange equation

    αu0+g(x)W(u0)+β(d2dx2)12u0=0, (1.23)

    and the condition at infinity

    limx±u0(x)=±1. (1.24)

    Here the fractional operator (d2dx2)12 is defined by (1.5).

    Remark 1.2. Part of the motivation for our choice of the assumptions on g is to study layer solutions in excitable media. Keener [46] introduced a model for calcium release in cardiac cells where the release sites are discrete, resulting in a production term that may degenerate in space. Here we consider a medium exhibiting bistability locally away from a suitable measure zero set, with one possible choice of g satisfying (1.2) being g(x)=(1+cos(2πx))1/4.

    Remark 1.3. When g(x)=1, the authors proved in [29] the monotonicity and uniqueness (up to translation) of minimizers uα of J(u) for each α>0. In addition, when α0, the unique miminzer with uα(0)=0 converges to a global minimizer of J(u) with α=0. The proof relies crucially on the translational invariance of J(u) when g is constant and does not extend to the periodic case studied here. In contrast, for g periodic it is not at all clear whether a layer solution exhibits monotonicity or uniqueness (up to discrete translation).

    Remark 1.4. When α=0, our argument can still be used to establish that the energy J(u) is bounded from below. It is not clear, though, which function space is a suitable choice for carrying out the minimization in this case. Indeed, for α>0 our argument uses the continuous representative of an H1(R) function. If α=0, however, one can no longer conclude a priori that the weak limit of the minimizing sequence remains in H1(R). When α0, it would be interesting to know whether the layer minimizers uα we constructed for α>0 converge to a function that minimizes J(u) with α=0 in some sense. In contrast to our earlier results in [29], due to the lack of monotonicity we were not able to derive uniform L2(R) bounds on uαη, and it is not clear how to pass to the limit.

    We prove Theorem 1.1 in three steps. We first check that J(u) is bounded from below. Let

    F(u):=Rg(x)W(u(x))dx+β4πRR[(u(x)u(y))2(xy)2(η(x)η(y))2(xy)2]dydx. (1.25)

    We show that F is bounded from below in Section 2. In the second step, we construct a global minimizer of J in A in Section 3. Regularity is treated in Section 4 and follows from a bootstrap argument, since u0=v0+η with v0H1(R). However, a priori it is not clear whether (d2dx2)12u0L2(R), and we handle this term separately when deriving the Euler-Lagrange equation.

    We note that our result and the variational approach are also closely related to those in [36]. There, Dipierro, Patrizi and Valdinoci prove existence of heteroclinic orbits for (1.20) and (1.21) in the case α=0 as follows. Given α,β(0,1], they considered a constrained double obstacle minimization of a perturbed renormalized energy with an additional penalization term in the following form

    Iα,β(u)=α2R|u(x)|2dx+β2R|u(x)η(x)|2dx+Rg(x)W(u(x))dx (1.26)
    +14RR[(u(x)u(y))2(η(x)η(y))2]K(xy)dydx. (1.27)

    The authors showed that the constrained minimizer becomes an unconstrained minimizer when the obstacles are far apart [36,Proposition 10.1]. The existence of heteoclinic orbits for (1.20) is obtained by first letting α0 and then sending β to zero. Our renormalized energy in this paper corresponds to β=0 in their setting and requires a much more delicate analysis of the tail behavior of the members of minimizing sequences.

    Let J(u), F(u) be defined by (1.1) and (1.25). We shall prove the following lower bound in this section.

    Proposition 2.1. There exists a positive constant C independent of u such that F(u)>C for any u A.

    Remark 2.2. The proof of the lower bound of F can be simplified if g(x) is a positive periodic function. In this case, we have g(x)a0>0 for all x. Thus for any uA,

    F(u)F0(u)=Ra0W(u)dx+RR[(u(x)u(y))2(xy)2(η(x)η(y))2(xy)2]dydx,

    and lower bound on F follows directly from the results in [29] since

    F(u)F0(u)F0(u0)>C,

    where u0 is the global minimizer of F0 obtained in [29]. The main difference in the argument employed in this paper is, therefore, to allow for a degeneracy in the energy occurring on a set of measure zero.

    The lower bound on J(u) follows directly from Proposition 2.1.

    First we observe that replacing u by ±1 whenever |u|>1, the energy is only getting smaller. Here for the nonlocal term, a direct calculation shows (u(x)u(y))2(˜u(x)˜u(y))2 where ˜u(x)=max{min(u(x),1),1}). Without loss of generality, we shall assume |u(x)| 1 on R throughout the paper. Moreover, the following Lemma can be proved by the same argument as in the proof of Lemma 2.2 from [29].

    Lemma 2.3. Given uA, there exists a sequence unA such that unηC0(R) and J(un)J(u) as n.

    We introduce the following subset of A:

    A0:={uA:|u(x)|1 on R and uη is compactly supported in R}.

    Letting

    f(x,y):=β4π[(u(x)u(y))2(xy)2(η(x)η(y))2(xy)2],

    we can write F(u) as

    F(u)=1g(x)W(u)dx+1g(x)W(u)dx+11g(x)W(u)dx+11f(x,y)dydx+11f(x,y)dydx+211f(x,y)dydx+2111f(x,y)dydx+2111f(x,y)dydx+1111f(x,y)dydx. (2.1)

    Direct calculation shows that the integrals

    111(η(x)η(y))2(xy)2dydx, 111(η(x)η(y))2(xy)2dydx, 1111(η(x)η(y))2(xy)2dydx

    are all bounded. To show F(u) is bounded from below, the question reduces to showing that

    11f(x,y)dydx+11f(x,y)dydx+211f(x,y)dydx+1g(x)W(u)dx+1g(x)W(u)dx>C, (2.2)

    for some C>0 independent of uA0. Since η˚H1/2(R), the term 11f(x,y)dydx could potentially be negative infinity. In particular, if we choose a sequence un(x) which oscillates between 1 and 1 on intervals which get larger and larger, it is not clear that we can have a uniform lower bound on the left-hand side in (2.2). Our idea is the following: if |u| stays away from 1 on a big portion of R, the term

    1g(x)W(u)dx+1g(x)W(u)dx

    would dominate

    11f(x,y)dydx.

    On the other hand, if |u|1 on R and u oscillates between 1 and 1, the sum

    11(u(x)u(y))2(xy)2dydx+11(u(x)u(y))2(xy)2dydx+211(u(x)u(y))2(xy)2dydx

    would approach infinity at the same order as

    11(η(x)η(y))2(xy)2dydx,

    thus eventually cancelling out the potential negative infinite energy. In both cases, we obtain a finite lower bound on F(u).

    To explain our ideas more precisely, recall that uηH1(R) and has compact support for every uA0. By Sobolev embedding theorem, uη and, therefore, u are continuous. Given any δ>0, we define the following decomposition of (,1][1,) with respect to u:

    I+δ:={x1:1u(x)1+δ},II+δ:={x1:1δu(x)1},III+δ:={x1:1+δ<u(x)<1δ}, (2.3)

    and

    Iδ:={x1:1u(x)1+δ},IIδ:={x1:1δu(x)1},IIIδ:={x1:1+δ<u(x)<1δ}. (2.4)

    Under these notations, we observe I+δ, IIδ, III+δ and IIIδ are all bounded sets. We show that there exists a constant C=C(δ,g,β,W)>0 and independent of uA0 such that

    11f(x,y)dydx+141g(x)W(u)dx+141g(x)W(u)dxI+δIδβπ(xy)2dydxII+δIIδβπ(xy)2dydxC, (2.5)

    and

    11f(x,y)dydx+11f(x,y)dydx+121g(x)W(u)dx+121g(x)W(u)dxI+δIδ2βπ(xy)2dydxII+δIIδ2βπ(xy)2dydx>C. (2.6)

    Throughout the paper, we will use C to represent a generic constant independent of uA0, and depending only on δ, g, β and W, which might change from line to line. A lower bound in (2.2) follows from (2.5) and (2.6).

    Since

    11f(x,y)dydx+11f(x,y)dydxβ2πI+δII+δ(u(x)u(y))2(xy)2dydx+β2πIδIIδ(u(x)u(y))2(xy)2dydx,

    the proof of (2.6) reduces to the following main technical inequalities:

    J+δ(u):=β4πI+δII+δ(u(x)u(y))2(xy)2dydxI+δIδβπ(xy)2dydx+141g(x)W(u)dx>C, (2.7)
    Jδ(u):=β4πIδIIδ(u(x)u(y))2(xy)2dydxII+δIIδβπ(xy)2dydx+141g(x)W(u)dx>C. (2.8)

    The proof of (2.7) and (2.8) uses a contradiction argument. We prove one bound, the other one can be proved similarly. Assume J+δ(un) for some sequence (un). Representing the decomposition of (,1][1,) with respect to un by adding index n in (2.3) and (2.4), decompose I+δ,n, Iδ,n and II+δ,n into union of disjoint intervals. We can estimate

    I+δ,nII+δ,n(un(x)un(y))2(xy)2dydx (2.9)

    and

    I+δ,nIδ,n4(xy)2dydx (2.10)

    in terms of summation of integral over those intervals. In particular, J+δ(un) implies I+δ,n[1,Rn] where Rn and (2.10) goes to infinity at most logarithmically in Rn. If |un| is bounded away from 1 on a large portion of [1,), then the term 1W(un) dominates (2.10). If un1 on a large portion of [1,), then (2.9) would approach infinity at the same logarithmic order as (2.10). In either case, we can always conclude that J+δ(un) is bounded from below, a contradiction.

    We prove Proposition 2.1 in several steps.

    We first state the following lemma.

    Lemma 2.4. Given uA0,, the following bounds hold:

    11g(x)W(u)dxC,1111(η(x)η(y))2(xy)2dydx4η2 (2.11)
    111(η(x)η(y))2(xy)2dydx2η2,111(η(x)η(y))2(xy)2dydx2η2 (2.12)

    Proof. The bounds in (2.11) are straightforward. By the definition of η, we have

    111(η(x)η(y))2(xy)2dydx=11(η(x)1)21xdx2η2.

    The second inequality in (2.12) follows from a similar argument.

    Lemma 2.4 implies that the terms involving integration on [1,1] in (2.1) are all bounded from below. The boundedness of F(u) from below would then follow from the following lemma.

    Lemma 2.5. There exists a constant C=C(δ,g,β,W)>0 such that for all uA0, the following lower bound holds:

    11f(x,y)dydx+11f(x,y)dydx+211f(x,y)dydx+1g(x)W(u)dx+1g(x)W(u)dx>C.

    Lemma 2.5 is proved in two steps. Under decompositions (2.3) and (2.4), we can write

    11f(x,y)dydx=I+δIδf(x,y)dydx+I+δIIδf(x,y)dydx+I+δIIIδf(x,y)dydx+II+δIδf(x,y)dydx+II+δIIδf(x,y)dydx+II+δIIIδf(x,y)dydx+III+δIδf(x,y)dydx+III+δIIδf(x,y)dydx+III+δIIIδf(x,y)dydx. (2.13)

    The following lower bound will be used in the proof of Lemma 2.5.

    Lemma 2.6. Let A[1,) and B(,1]. Assume either A or B is bounded, then there exists a constant C=C(β,g)>0 such that for all uA0, the following bounds hold:

    ABf(x,y)dydxεAg(1u)2dxεBg(1+u)2dyCε, (2.14)

    and

    ABf(x,y)dydxεAg(1+u)2dxεBg(1u)2dyCε (2.15)

    for any ε>0.

    Proof. We only show how to obtain (2.14), as the other inequality follows by a similar argument. Since (η(x)η(y))2=4 for xA and yB, we have

    f(x,y)=β4π[(u(x)u(y))24(xy)2]=β4π[(u(x)u(y)2)2+4(u(x)u(y)2)(xy)2].

    By (1.2) and periodicity of g(x), we conclude

    11g(x)(x+1)2dx+11g(x)(1x)2dx2101g(x)dxn=11(n+1)2<. (2.16)

    Hence we deduce

    ABf(x,y)dydx=β4π[AB(u(x)u(y)2)2(xy)2dydx4AB(1u(x))(xy)2dydx4AB(1+u(y))(xy)2dydx]β4π[AB(u(x)u(y)2)2(xy)2dydx4A1u(x)x+1dx4B1+u(y)1ydy]β4πAB(u(x)u(y)2)2(xy)2dydxεAg(1u)2dxεBg(1+u)2dyCε,

    where we applied Hölder's inequality and (2.16) in the last line. Notice that when either A or B is bounded, the integral AB(u(x)u(y)2)2(xy)2dydx is finite for any uA0, justifying the spliting of the integrals in the calculation above.

    The first step to prove Lemma 2.5 is the following Lemma.

    Lemma 2.7. For any 0<δ1, there exists a constant C=C(δ,g,β,W) such that for all uA0

    11f(x,y)dydx+141g(x)W(u)dx+141g(x)W(u)dxI+δIδβπ(xy)2dydxII+δIIδβπ(xy)2dydxC

    Proof. Recall that I+δ, IIδ, III+δ and III+δ are all bounded sets. By estimate (2.14) in Lemma 2.6, we have

    II+δIδf(x,y)dydxεII+δg(1u)2dxεIδg(1+u)2dyCε, (2.17)
    II+δIIIδf(x,y)dydxεII+δg(1u)2dxεIIIδg(1+u)2dyCε, (2.18)
    III+δIδf(x,y)dydxεIII+δg(1u)2dxεIδg(1+u)2dyCε, (2.19)

    and

    III+δIIIδf(x,y)dydxεIII+δg(1u)2dxεIIIδg(1+u)2dyCε. (2.20)

    Here for (2.17) we used the fact that the integral II+δIδf(x,y)dydx can be written as a sum of integrals of the form ABf(x,y)dydx, where either A or B is bounded when uA0.

    By estimate (2.15) from Lemma 2.6, we have

    I+δIIδf(x,y)dydxεI+δg(1+u)2dxεIIδg(1u)2dyCε, (2.21)
    I+δIIIδf(x,y)dydxεI+δg(1+u)2dxεIIIδg(1u)2dyCε, (2.22)

    and

    III+δIIδf(x,y)dydxεIII+δg(1+u)2dxεIIδg(1u)2dyCε. (2.23)

    Summing up (2.17)−(2.23) yields

    11f(x,y)dydx+141g(x)W(u)dx+141g(x)W(u)dyI+δIδβπ(xy)2dydxII+δIIδβπ(xy)2dydxCε2εI+δIδIIIδg(1+u)2dx2εII+δIIδIII+δg(1u)2dxεIIIδg(1u)2dyεIII+δg(1+u)2dx+14I+δIδg(x)W(u)dx+14II+δIIδg(x)W(u)dx+14III+δIIIδg(x)W(u)dx.

    Recall that W(±1)=W(±1)=0, W(u)>0 for |u|<1 and W(±1)>0. Picking δ1, we have the following estimates

    W(u(x))=12W((1θ(x))+θ(x)u(x))(1+u)214W(1)(1+u)2 for xI+δIδ, (2.24)
    W(u(x))=12W((1θ(x))+θ(x)u(x))(1u)214W(1)(1u)2 for xII+δIIδ, (2.25)

    Since

    W(u(x))min|u|1δW(u) for xIII+δIIIδ,

    the conclusion follows by taking ε=min(148min|u|1δW(u),132W(1),132W(1)).

    The second step to prove Lemma 2.5 is the following Lemma.

    Lemma 2.8. There exists a constant C=C(δ,g,β,W)>0 such that for all uA0 we have

    11+11f(x,y)dydx+121g(x)W(u)dx+121g(x)W(u)dxI+δIδ2βπ(xy)2dydxII+δIIδ2βπ(xy)2dydx>C.

    Lemma 2.7 and Lemma 2.8 imply Lemma 2.5.

    Decompositions and some basic estimates The proof of Lemma 2.8 is rather long and technical. First we decompose each set into intervals. By our assumption, for any given δ(0,1), there exists R1(u)>0 and R2(u)>0 such that

    u(x)=1 for all xR1(u) (2.26)

    and

    u(x)=1 for all xR2(u). (2.27)

    It follows that III+δ and IIIδ are open subsets of (1,R1(u)) and (R2(u),1) respectively. By the structure theorem of open sets in R and choosing u to be the continuous representative, there exist indices N±, and positive numbers α±i, β±i such that we can write III+δ and IIIδ as unions of disjoint open intervals in the following form.

    III+δ(1,R1(u))=N+i=1(α+i,β+i), IIIδ(R2(u),1)=Nj=1(βj,αj). (2.28)

    Without loss of generality, we can assume N± are finite and αi<βi<αi+1<βi+1 for all i. In fact, recall that uη in H10(R), write R=max(R1(u),R2(u)), we can obtain an approximation u1 in C(R) that is equal to η for |x|>R+1 and arbitrarily close to u in H1(R). Taking a linear interpolant u2 of u1 over a sufficiently fine partition X of [1R,1+R], we get a function that is arbitrarily close to u1 in W1,(R). Finally, shifting the values of the function u2 at the (finitely many) points of X by arbitrarily small amounts if necessary, we get a function u3 that is arbitrarily close to u2 in W1,(R) and u3 is non-zero a.e. in R. Hence on every interval of the partition X there is at most one point at which |u3|=1δ. From this, we conclude that each interval of the partition X intersects III+δ (or IIIδ) at most once. Relabling if necessary, we thus find finitely many disjoint intervals (αk,βk) (merge (αk,βk)(αk+1,βk+1) into (αk,βk+1) if βk=αk+1) where

    1+δ<u(x)<1δ on each (αk,βk).

    Renaming our endpoints we find disjoint intervals [ai,bi] , [cj,dj][1,R1(u)] and indices K,L such that

    I+δ=Ki=1[ai,bi] (2.29)

    and

    II+δ=Lj=1[cj,dj] (2.30)

    with

    1u(x)1+δ on I+δ,
    1δu(x)1 on II+δ,

    and

    1+δ<u(x)<1δ on [1,R1(u)](I+δII+δ)=III+δ.

    Here by a slight abuse of notation, we denote [cl,dl]=[cl,), and if a1=1 we replace [a1,b1] by (1,b1] in (2.29), or if c1=1, replace [c1,d1] by (1,d1] in (2.30). Similarly we write IIδ and Iδ as unions of disjoint intervals. For the rest of the paper, we write

    I+δ(u)=Ki=1[ai,bi], II+δ(u)=Lj=1[cj,dj]; (2.31)
    Iδ(u)=˜Ki=1[˜bi,˜ai], IIδ(u)=˜Lj=1[˜dj,˜cj]. (2.32)

    with the understanding that [˜d˜L,˜c˜L]=(,˜c˜L] and [˜b1,˜a1]=[˜b1,1) if ˜a1=1, or [˜d1,˜c1]=[˜d1,1) if ˜c1=1. Note that in this form, all ai,bi, cj,dj, ˜ai,˜bi, ˜cj,˜dj are greater or equal to 1.

    We first state some basic estimates.

    Lemma 2.9. The following estimates hold:

    I+δIδ1(xy)2dydx=ln(Ki=1˜Lj=1ai+˜djbi+˜djbi+˜cjai+˜cj), (2.33)
    I+δII+δ1(xy)2dydx=ln(Ki=1Lj=1bidjaidjaicjbicj) (2.34)
    IδIIδ1(xy)2dydx=ln(˜Ki=1˜Lj=1˜bi˜dj˜ai˜dj˜ai˜cj˜bi˜cj) (2.35)
    II+δIIδ1(xy)2dydx=ln(˜Ki=1Lj=1˜ai+dj˜bi+dj˜bi+cj˜ai+cj), (2.36)

    Proof. By (2.31) and (2.32), we have

    I+δIδ1(xy)2dydx=Ki=1˜Lj=1biai˜cj˜dj1(xy)2dydx=Ki=1˜Lj=1biai(1x+˜cj1x+˜dj)dx=Ki=1˜Lj=1(lnbi+˜cjai+˜cjlnbi+˜djai+˜dj)=ln(Ki=1˜Lj=1ai+˜djbi+˜djbi+˜cjai+˜cj).

    (2.3), (2.35) and (2.36) are proved similarly.

    An immediate corollary of Lemmas 2.9 is the following.

    Corollary 2.10. We have the following bounds.

    I+δIδ1(xy)2dydx<{ln2if˜c1>bKln2bKa1if˜c1<bK, (2.37)
    II+δIIδ1(xy)2dydx<{ln2ifc1>˜b˜Kln2˜b˜K˜a1ifc1<˜b˜K. (2.38)

    Proof.

    I+δIδ1(xy)2dydx=Ki=1˜Lj=1biai˜cj˜dj1(xy)2dydxbKa1˜c1d˜L1(xy)2dydx=ln(bK+˜c1a1+˜c1a1+d˜LbK+d˜L)ln(bK+˜c1a1+˜c1)

    If ˜c1>bK,

    ˜c1+bK˜c1+a1<2,

    if ˜c1<bK,

    ˜c1+bK˜c1+a1<2bKa1,

    so (2.37) follows. The estimate (2.38) follows from a similar argument.

    The following estimate on the potential term is important for the lower bound estimate.

    Lemma 2.11. Given any ε(0,1), there exists a positive constant τ=τ(ε,g) such that

    1g(x)W(u(x))dxτ(1ε)|IIIδ+|min|u|1δW(u)+I+δII+δg(x)W(u(x)dx (2.39)
    1g(x)W(u(x))dxτ(1ε)|IIIδ|min|u|1δW(u)+IδIIδg(x)W(u(x)dx (2.40)

    Proof. By (2.16). the zero level set of g E0={x[0,1]:g(x)=0} has measure zero. In particular, given any ε>0, we can find an open set Uε of (0,1) which covers E0 and |Uε|<ε. Let Oε=n=0{Uε+n}, then since g(x) is continuous and positive on [1,)Oε, there exists τ=τ(g,ε)>0 such that g(x)τ on [1,)Oε, therefore

    1g(x)W(u(x))dx=III+δg(x)W(u(x))dx+I+δII+δg(x)W(u(x)dxτ(1ε)|IIIδ+|min|u|1δW(u)+I+δII+δg(x)W(u(x)dx.

    (2.40) can be proved similarly.

    Remark 2.12. Fix an ε=ε0 and write γ=τ(ε0,g)(1ε0), we can write the lower bounds (2.39)–(2.40) as

    1g(x)W(u(x))dxγ|IIIδ+|min|u|1δW(u)+I+δII+δg(x)W(u(x)dx (2.41)
    1g(x)W(u(x))dxγ|IIIδ|min|u|1δW(u)+IδIIδg(x)W(u(x)dx (2.42)

    Asume I±δ, II±δ are written as unions of intervals in the form (2.31) and (2.32). Let ni=1[ai,bi]I+δ and mj=1[cj,dj]II+δ, ˜ni=1[˜bi,˜ai]Iδ and ˜mj=1[˜dj,˜cj]IIδ. If dj=, we write [cj,dj]=[cj,) and [˜dj,˜cj]=(,˜cj] if ˜dj=. Assume also

    0<a1<b1<a2<<bn1<an<bn<c1<d1<c2<<dm

    and

    0<˜a1<˜b1<˜a2<<˜a˜n<˜b˜n<˜c1<˜d1<˜c2<<˜d˜m.

    We have the following estimates.

    Lemma 2.13. Assume |c1bn|1, |˜c1˜b˜n|1 for all i and j. For δ1, there exist positive constants C=C(δ,g,β,W) and γ=γ(g) such that the following estimates hold for any uA0.

    β4πni=1mj=1biaidjcj(u(x)u(y))2(xy)2I+δIδβπ(xy)2dydx+141g(x)W(u)dxβπni=1mj=1lncjaicjbidjbidjaiβπln2bKa1+γmin|u|1δW(u)4|III+δ|C. (2.43)
    β4π˜ni=1˜mj=1˜ai˜bi˜cj˜dj(u(x)u(y))2(xy)2II+δIIδβπ(xy)2dydx+141g(x)W(u)dxβπ˜ni=1˜mj=1ln˜cj˜bi˜cj˜ai˜dj˜ai˜dj˜biβπln2˜b˜K˜a1+γmin|u|1δW(u)4|IIIδ|C. (2.44)

    Proof. We prove (2.43), (2.44) follows from a similar argument. By Corollary 2.10, Remark 2.12 and (2.24−2.25),

    β4πni=1mj=1biaidjcj(u(x)u(y))2(xy)2dydxI+δIδβπ(xy)2dydx+141g(x)W(u)dx=β4πni=1mj=1biaidjcj(u(x)+1+1u(y)2)24(xy)2dydx+ni=1mj=1biaidjcjβπ(xy)2dydxI+δIδβπ(xy)2dydx+141g(x)W(u)dxβπni=1mj=1biaidjcj(u(x)+1)(xy)2dydxβπni=1mj=1biaidjcj(1u(y))(xy)2dydx+βπni=1mj=1ln(cjaicjbidjbidjai)βπln(2bKa1)+γmin|u|1δW(u)4|III+δ|+14I+δII+δg(x)W(u)dxβπni=1biai(u(x)+1)mj=1(1cjx1djx)dxβπmj=1djcj(1u(y))ni=1(1yai1ybi)dy+βπni=1mj=1ln(cjaicjbidjbidjai)βπln(2bKa1)+γmin|u|1δW(u)4|III+δ|+W(1)16I+δg(1+u)2dx+W(1)16II+δg(1u)2dxW(1)16ni=1biaig(x)(u(x)+1)2dx4β2π2W(1)ni=1biai1g(x)(c1x)2dxW(1)16mj=1djcjg(y)(1u(y))2dy4β2π2W(1)mj=1djcj1g(y)(ya1)2dy+βπni=1mj=1ln(cjaicjbidjbidjai)βπln(2bKa1)+γmin|u|1δW(u)4|III+δ|+W(1)16I+δg(1+u)2dx+W(1)16II+δg(1u)2dxβπni=1mj=1ln(cjaicjbidjbidjai)βπln(2bKa1)+γmin|u|1δW(u)4|III+δ|C

    The last two steps follow from

    ni=1biai(u(x)+1)mj=1(1cjx1djx)dxϵ2ni=1biaig(x)(u(x)+1)2dx+12ϵni=1biai1g(x)(mj=1(1cjx1djx))2dxϵ2I+δg(x)(u(x)+1)2dx+12ϵni=1biai1g(x)(c1x)2dxϵ2I+δg(x)(u(x)+1)2dx+12ϵbna11g(x)(c1x)2dxϵ2I+δg(x)(u(x)+1)2dx+12ϵ([bn][a1]1g(x)(c1x)2dx+bn[bn]1g(x)dx)ϵ2I+δg(x)(u(x)+1)2dx+12ϵ101g(x)dxn=11n2ϵ2I+δg(x)(u(x)+1)2dx+C2ϵ

    and

    mj=1djcj(1u(y))ni=1(1yai1ybi)dyϵ2II+δg(y)(1u(y))2dy+C2ϵ.

    Proof of the main technical lemma Observe that

    11f(x,y)dydx+11f(x,y)dydx=β4π11(u(x)u(y))2(xy)2dydx+β4π11(u(x)u(y))2(xy)2dydx=β4πI+δI+δ(u(x)u(y))2(xy)2dydx+β4πII+δII+δ(u(x)u(y))2(xy)2dydx+β4πIII+δIII+δ(u(x)u(y))2(xy)2dydx+β4πIδIδ(u(x)u(y))2(xy)2dydx+β4πIIδIIδ(u(x)u(y))2(xy)2dydx+β4πIIIδIIIδ(u(x)u(y))2(xy)2dydx+β2πI+δII+δ(u(x)u(y))2(xy)2dydx+β2πII+δIII+δ(u(x)u(y))2(xy)2dydx+β2πI+δIII+δ(u(x)u(y))2(xy)2dydx+β2πIδIIδ(u(x)u(y))2(xy)2dydx+β2πIIδIIIδ(u(x)u(y))2(xy)2dydx+β2πIδIIIδ(u(x)u(y))2(xy)2dydx, (2.45)

    Lemma 2.8 would follow from (2.45) and the following Lemma.

    Lemma 2.14. There exists a constant C=C(δ,γ,β,W)>0 such that

    β4πI+δII+δ(u(x)u(y))2(xy)2dydx+β4πIδIIδ(u(x)u(y))2(xy)2dydx+141g(x)W(u)dx+141g(x)W(u)dxI+δIδβπ(xy)2dydxII+δIIδβπ(xy)2dydx>C

    Lemma 2.14 is a direct corollary of the following Lemma.

    Lemma 2.15. There exists a constant C=C(δ,γ,β,W)>0 such that for all uA0 the following bounds hold:

    β4πI+δII+δ(u(x)u(y))2(xy)2dydxI+δIδβπ(xy)2dydx+141g(x)W(u)dx>C, (2.46)
    β4πIδIIδ(u(x)u(y))2(xy)2dydxII+δIIδβπ(xy)2dydx+141g(x)W(u)dx>C. (2.47)

    Proof. We prove (2.46), as the proof of (2.47) is similar. We argue by contradiction and let

    J+δ(u):=β4πI+δII+δ(u(x)u(y))2(xy)2dydxI+δIδβπ(xy)2dydx+141g(x)W(u)dx

    We show that

    J+δ(un)

    implies

    limsupnmin1lKnanlbnl=1 (2.48)

    On the other hand, (2.48) and a diagonal argument would imply J+δ(un) is bounded from below, a contradiction. We prove (2.48) in a series of steps. We will only explain the first three steps in detail. The remaining steps can be proved similarly.

    Assume (2.46) fails. Then there would exist a sequence {un} such that J+δ(un)<n. We denote the decomposition of I±δ(un) and II±δ(un) as follows.

    I+δ,n=Kni=1[ani,bni], II+δ,n=Lni=1[cnj,dnj],Iδ,n=˜Kni=1[˜bni,˜ani], IIδ,n=˜Lni=1[˜dnj,˜cnj]

    Here we have dnLn= and cnLn>bnKn. We assume

    (bnKn,cnLn)II+δ,n=Ln1k=jn1[cnk,dnk],(bni1,ani)II+δ,n=jni11k=jni[cnk,dnk] for i=2,,Kn.

    Step 1: J+δ(un) implies there exists in1 satisfying jn1in1Ln such that

     liminfncnin1bnKn=limsupnanKnbnKn=1. (2.49)

    First we observe

    n>J+δ(un)=β4πI+δ,nII+δ,n(un(x)un(y))2(xy)2dydxI+δ,nIδ,nβπ(xy)2dydx+141g(x)W(un)dxI+δ,nIδ,nβπ(xy)2dydxβπln(2bnKnan1).

    Therefore, J+δ(un) implies

    limsupnbnKn=. (2.50)

    Case Ⅰ: Assume that for a subsequence, cnLnbnKn<1 (without relabeling for simplicity of notations). Then by Lemma 2.13,

    n>J+δ(un)β4πbnKnanKncnLn+1(un(x)un(y))2(xy)2dydxI+δ,nIδ,nβπ(xy)2dydx+141g(x)W(un)dxβπlnanKncnLn1bnKncnLn1βπln(2bnKnan1)+γmin|u|1δW(u)4|(bnKn,cnLn)III+δ,n|CβπlncnLnanKnbnKnβπln2+γmin|u|1δW(u)4|(bnKn,cnLn)III+δ,n|CβπlncnLnanKnbnKnC.

    Taking liminf on both sides, we have

    liminfncnLnbnKn=limsupnanKnbnKn=1

    For the remaining cases, we shall always assume cnLnbnKn1. Also whenever we need to work on a subsequence, we always use the original sequence for simplicity of notations.

    Case Ⅱ: Assume for a subsequence that there exists in1{ jn1, jn1+1, ,Ln1} such that cnin1bnKn<1. We also assume bnKnanKn>1 (otherwise (2.49) follows directly). Then by

    ga,b(x)=axblnxb(1lnba), for x>0, (2.51)

    and Lemma 2.13, we bound J+δ(un) as follows.

    n>J+δ(un)β4πbnKnanKncnLn(un(x)un(y))2(xy)2dydx+β4πbnKnanKn(bnKn, cnLn)II+δ,n(un(x)un(y))2(xy)2dydxI+δ,nIδ,nβπ(xy)2dydx+141g(x)W(un)dxβ4πbnKn1anKncnin1(un(x)un(y))2(xy)2dydxbnKnanKndnin1+|(bnKn, cnLn)III+δ,n|dnin1βπ(xy)2dydxI+δ,nIδ,nβπ(xy)2dydx+141g(x)W(un)dxβπlncnin1anKncnin1+1bnKnβπln(dnin1anKndnin1bnKndnin1+|(bnKn,cnLn)III+δ,n|bnKndnin1+|(bnKn,cnLn)III+δ,n|anKn)Cβπln(2bnKnan1)+γmin|u|1δW(u)4|(bnKn,cnLn)III+δ,n|βπlncnin1anKnbnKnβπln(1+|(bnKn,cnLn)III+δ,n|)+γmin|u|1δW(u)4|(bnKn,cnLn)III+δ,n|Cβπlncnin1anKnbnKnC.

    Taking liminf on both sides, we must have

    liminfcnin1bnKn=limsupanKnbnKn=1.

    Case Ⅲ: Assume no such in1 from case Ⅱ exists, then we must have cnjn1bnKn1. Thus

    n>J+δ(un)β4πbnKnanKn(cnjn1)II+δ,n(un(x)un(y))2(xy)2dydxI+δ,nIδ,nβπ(xy)2dydx+141g(x)W(un)dxβ4πbnKnanKncnjn1(un(x)un(y))2(xy)2dydxbnKnanKndnjn1+|(bnKn, cnLn)III+δ,n|dnjn1βπ(xy)2dydxI+δ,nIδ,nβπ(xy)2dydx+141g(x)W(un)dxβπlncnjn1anKncnjn1bnKnβπln(dnjn1anKndnjn1bnKndnjn1+|(bnKn,cnLn)III+δ,n|bnKndnjn1+|(bnKn,cnLn)III+δ,n|anKn)βπln(2bnKnan1)+γmin|u|1δW(u)4|(bnKn,cnLn)III+δ,n|Cβπlncnjn1anKnbnKnβπln(cnjn1 bnKn)βπln(1+|(bnKn,cnLn)III+δ,n|)+γmin|u|1δW(u)4|(bnKn,cnLn)III+δ,n|Cβπlncnjn1anKnbnKnC.

    The last step used (2.51) and  (bnKn,cnjn1)(bnKn,cnLn)III+δ,n. Taking liminf on both sides of the equation above, we must have

    liminfcnjn1bnKn=limsupanKnbnKn=1.

    Step 2: J+δ(un) implies there exists in2 satisfying jn2in2in1 such that

     liminfncnin2bnKn1=limsupnanKn1bnKn1=1 (2.52)

    First we observe

    n>J+δ(un)I+δ,nIδ,nβπ(xy)2dydx+141g(x)W(un)dxβπln(2bnKn1an1)βπlnbnKnanKn.

    Therefore, by Step 1 we have that J+δ(un) implies

    limsupbnKn1=.

    Case Ⅰ: If liminfnanKnbnKn1cnin1bnKn1=0, then limsupncnin1anKncnin1bnKn1=1. In this case, then we can replace (anKn1,bnKn1)(anKn,bnKn) by (anKn1,bnKn) and repeat our argument in Step 1 as follows.

    n>J+δ(un)β4πbnKn1anKn1(bnKn)II+δ,n(un(x)un(y))2(xy)2dydx+β4πbnKnanKn(bnKn)II+δ,n(un(x)un(y))2(xy)2dydxI+δ,nIδ,nβπ(xy)2dydx+141g(x)W(un)dxβ4πbnKnanKn1cnin1(un(x)un(y))2(xy)2dydxbnKnanKn1dnin1+|(bnKn, cnLn)III+δ,n|dnin1βπ(xy)2dydxanKnbnKn1cnin1βπ(xy)2dydxβπln2bnKn1an1βπlnbnKnanKn+141g(x)W(un)dxβ4πbnKnanKn1cnin1(un(x)un(y))2(xy)2dydxbnKnanKn1dnin1+|(bnKn,CnLn)III+δ,n|dnin1βπ(xy)2dydxβπlncnin1bnKn1cnin1anKnβπln2bnKn1an1βπlnbnKnanKn+141g(x)W(un)dxβπlncnin1anKn1cnin1bnKnβπlncnin1bnKn1cnin1anKnβπln(dnin1anKn1dnin1bnKndnin1+|(bnKn1,anKn)III+δ,n|bnKndnin1+|(bnKn1,anKn)III+δ,n|anKn1)βπln2bnKn1an1βπlnbnKnanKn+141g(x)W(un)dxCβπlncnin1anKn1bnKn1βπlncnin1bnKn1cnin1anKnβπln(cnin1bnKn)βπln(1+|(bnKn1,anKn)III+δ,n|)βπlnbnKnanKn+γmin|u|1δW(u)4|III+δ,n|Cβπlncnin1anKn1bnKn1βπlncnin1bnKn1cnin1anKnβπlnbnKnanKnC.

    Taking liminf on both sides, we get

    liminfcnin1bnKn1=limsupanKn1bnKn1=1.

    Case Ⅱ: If liminfnanKnbnKn1cnin1bnKn1>0, There are three cases.

    Case Ⅱ-ⅰ: If (bnKn1,anKn)II+δ,n=, then (bnKn1,anKn)III+δ,n,

    n>J+δ(un)β4πbnKn1anKn1(bnKn, )II+δ,n(un(x)un(y))2(xy)2dydx+β4πbnKnanKn(bnKn,)II+δ,n(un(x)un(y))2(xy)2dydxI+δ,nIδ,nβπ(xy)2dydx+141g(x)W(un)dxβ4πbnKn1anKn1cnin1(un(x)un(y))2(xy)2dydxbnKn1anKn1dnin1+|(bnKn, cnLn)III+δ,n|dnin1βπ(xy)2dydx+β4πbnKnanKncnin1(un(x)un(y))2(xy)2dydxbnKnanKndnin1+|(bnKn, cnLn)III+δ,n|dnin1βπ(xy)2dydxβπln2bnKn1an1βπlnbnKnanKn+141g(x)W(un)dxβπlncnin1anKn1cnin1bnKn1βπln(dnin1anKn1dnin1bnKn1dnin1+|(bnKn,cnLn)III+δ,n|bnKn1dnin1+|(bnKn,cnLn)III+δ,n|anKn1)+βπlncnin1anKncnin1bnKnβπln(dnin1anKndnin1bnKndnin1+|(bnKn,cnLn)III+δ,n|bnKndnin1+|(bnKn,cnLn)III+δ,n|anKn)βπln2bnKn1an1βπlnbnKnanKn+γmin|u|1δW(u)4|III+δ,n|Cβπlncnin1anKn1bnKn1+βπlncnin1+1anKncnin1+1bnKn1βπln(cnin1bnKn)2βπln(1+|(bnKn,cnLn)III+δ,n|)βπlnbnKnanKn+γmin|u|1δW(u)4|III+δ,n|Cβπlncnin1anKn1bnKn14ln(1+(anKnbnKn1))βπln(cnin1bnKn)2βπln(1+|(bnKn,cnLn)III+δ,n|)βπlnbnKnanKn+γmin|u|1δW(u)4|III+δ,n|Cβπlncnin1anKn1bnKn1βπlnbnKnanKnC, (2.53)

    where we used (2.51) and the facts

    (anKnbnKn1)|III+δ,n|, (cnin1bnKn)max(1,|(bnKn,cnLn)III+δ,n|).

    Taking liminf on both sides of (2.53), we conclude

    liminfncnin1bnKn1=limsupnanKn1bnKn1=1.

    Case Ⅱ-ⅱ: If (bnKn1,anKn)II+δ,n, there are two cases. If liminfncnin1anKn1=1, then (2.52) follows directly. We therefore assume liminfncnin1anKn1>1 for the remaining two cases.

    Case Ⅱ-ⅱ-a: There exists in2{jn2, jn2+1, , in1} such that cnin2bnKn1<1. We assume bnKn1anKn1>1 without loss of generality. Applying Corollary 2.10, Lemma 2.13 and (2.51), we bound J+δ(un) as follows.

    n>J+δ(un)β4πbnKn1anKn1(bnKn1, anKn)II+δ,n(un(x)un(y))2(xy)2dydx+β4πbnKn1anKn1(bnKn,)II+δ,n(un(x)un(y))2(xy)2dydxI+δ,nIδ,nβπ(xy)2dydx+141g(x)W(un)dxβ4πbnKn11anKn1anKncnin2(un(x)un(y))2(xy)2dydxbnKn1anKn1dnin2+|(bnKn1, anKn)III+δ,n|dnin2βπ(xy)2dydx+β4πbnKn1anKn1cnin1(un(x)un(y))2(xy)2dydxbnKn1anKn1dnin1+|(bnKn1, cnLn)III+δ,n|dnin1βπ(xy)2dydxβπln2bnKn1an1βπlnbnKnanKn+141g(x)W(un)dxβπln(cnin2anKn1cnin2+1bnKn1anKnbnKn1+1anKnanKn1)βπln(dnin2anKn1dnin2bnKn1dnin2+|(bnKn1,anKn)III+δ,n|bnKn1dnin2+|(bnKn1,anKn)III+δ,n|anKn1)+βπlncnin1anKn1cnin1bnKn1βπln(dnin1anKn1dnin1bnKn1dnin1+|(bnKn,cnLn)III+δ,n|bnKn1dnin1+|(bnKn,cnLn)III+δ,n|anKn1)βπln2bnKn1an1βπlnbnKnanKnC+γmin|u|1δW(u)4|III+δ,n|βπlncnin2anKn1bnKn1+βπln(cnin1anKn1cnin1bnKn1anKnbnKn1anKnanKn1)βπln(1+|(bnKn1,anKn)III+δ,n|)βπln(1+|(bnKn,cnLn)III+δ,n|)βπlnbnKnanKnCβπln(1+cnin2bnKn1)+γmin|u|1δW(u)4|III+δ,n|βπlncnin2anKn1bnKn1+βπlnanKnbnKn1cnin1bnKn1βπlnbnKnanKnC.

    Taking liminf on both sides, by Step 1 and liminfnanKnbnKn1anKnanKn1>0, we must have

    liminfcnin2bnKn1=limsupanKn1bnKn1=1.

    Case Ⅱ-ⅰ-b: No such in2 exists, then we must have cnjn2bnKn1. Applying Corollary 2.10, Lemma 2.13 and (2.51), we bound J+δ(un) as follows.

    n>J+δ(un)β4πbnKn1anKn1(bnKn1, anKn)II+δ,n(un(x)un(y))2(xy)2dydx+β4πbnKn1anKn1(bnKn)II+δ,n(un(x)un(y))2(xy)2dydxI+δ,nIδ,nβπ(xy)2dydx+141g(x)W(un)dxβ4πbnKn1anKn1anKncnjn2(un(x)un(y))2(xy)2dydxbnKn1anKn1dnjn2+|(bnKn1, anKn)III+δ,n|dnjn2βπ(xy)2dydx+β4πbnKn1anKn1cnin1(un(x)un(y))2(xy)2dydxbnKn1anKn1dnin1+|(bnKn1, cnLn)III+δ,n|dnin1βπ(xy)2dydxβπln2bnKn1an1βπlnbnKnanKn+141g(x)W(un)dxβπln(cnjn2anKn1cnjn2bnKn1anKnbnKn1anKnanKn1)βπln(dnjn2anKn1dnjn2bnKn1dnjn2+|(bnKn1,anKn)III+δ,n|bnKn1dnjn2+|(bnKn1,anKn)III+δ,n|anKn1)+βπlncnin1anKn1cnin1bnKn1βπln(dnin1anKn1dnin1bnKn1dnin1+|(bnKn,cnLn)III+δ,n|bnKn1dnin1+|(bnKn,cnLn)III+δ,n|anKn1)βπln2bnKn1an1βπlnbnKnanKnC+γmin|u|1δW(u)4|III+δ,n|βπlncnjn2anKn1bnKn1+βπln(cnin1anKn1cnin1bnKn1anKnbnKn1anKnanKn1)βπln(1+|(bnKn1,anKn)III+δ,n|)βπlnbnKnanKnCβπln(cnjn2bnKn1)+γmin|u|1δW(u)4|III+δ,n|βπlncnjn2anKn1bnKn1+βπlnanKnbnKn1cnin1bnKn1βπlnbnKnanKnC.

    Taking liminf on both sides, we must have

    liminfcnjn2bnKn1=limsupanKn1bnKn1=1.

    Step 3: J+δ(un) implies there exists in3 satisfying jn3in3in2 such that

     liminfncnin3bnKn2=limsupnanKn2bnKn2=1 (2.54)

    First we observe

    n>J+δ(un)=βπI+δ,nII+δ,n(u(x)u(y))2(xy)2dydxI+δ,nIδ,nβπ(xy)2dydx+141g(x)W(un)dxI+δ,nIδ,nβπ(xy)2dydxβπln(2bnKn2an1)βπlnbnKnanKnβπlnbnKn1anKn1.

    Therefore

    limsupbnKn2=.

    Case Ⅰ: liminfanKnbnKn2cnin1bnKn2=0. This implies

    liminfcnin1bnKn2cnin1anKn=1. (2.55)

    In this case, we can replace (anKn2,bnKn2)(anKn1,bnKn1)(anKn,bnKn) by (anKn2,bnKn), and repeat our argument in step 1.

    We estimate J+δ(un) as follows.

    n>J+δ(un)β4πbnKn2anKn2(bnKn,)II+δ,n(un(x)un(y))2(xy)2dydx+β4πbnKn1anKn1(bnKn,)II+δ,n(un(x)un(y))2(xy)2dydx+βπbnKnanKn(bnKn,)II+δ,n(un(x)un(y))2(xy)2dydxI+δ,nIδ,nβπ(xy)2dydx+141g(x)W(un)dxβ4πbnKnanKn2cnin1(un(x)un(y))2(xy)2dydxbnKnanKn2dnin1+|(bnKn,)III+δ,n|dnin1βπ(xy)2dydxanKnbnKn2cnin1βπ(xy)2dydxI+δ,nIδ,nβπ(xy)2dydx+141g(x)W(un)dxβπlncnin1anKn2cnin1bnKnβπln[dnin1anKn2dnin1bnKndnin1+|(bnKn,)III+δ,n|bnKndnin1+|(bnKn,)III+δ,n|anKn2]βπlncnin1bnKn2cnin1anKnβπln2bnKnan1+γmin|u|1δ4|III+δ,n|.

    By (2.55), we are back to the situation in Step 1 with (anKn,bnKn) replaced by (anKn2,bnKn) and we conclude

    liminfcnin1bnKn=limsupanKn2bnKn=limsupanKn2bnKn2=1.

    Case Ⅱ: liminfanKnbnKn2cnin1bnKn2>0 and liminfanKn1bnKn2cnin2bnKn2=0. This implies

    limsupcnin2anKn1cnin2bnKn2=1. (2.56)

    In this case, we replace (anKn2,bnKn2)(anKn1,bnKn1) by (anKn2,bnKn1) and repeat our argument in step 2.

    We bound J+δ(un) as follows.

    n>J+δ(un)β4πbnKn2anKn2II+δ,n(un(x)un(y))2(xy)2dydx+β4πbnKn1anKn1II+δ,n(un(x)un(y))2(xy)2dydx+β4πbnKnanKnII+δ,n(un(x)un(y))2(xy)2dydxI+δ,nIδ,nβπ(xy)2dydx+141g(x)W(un)dxβ4πbnKn1anKn2II+δ,n(cnin2,)(un(x)un(y))2(xy)2dydxanKn1bnKn2II+δ,n(cnin2,)βπ(xy)2dydx+βπbnKnanKnII+δ,n(un(x)un(y))2(xy)2dydxI+δ,nIδ,nβπ(xy)2dydx+141g(x)W(un)dxβπbnKn1anKn2II+δ,n(cnin2,)(un(x)un(y))2(xy)2dydx+β4πbnKnanKnII+δ,n(un(x)un(y))2(xy)2dydxI+δ,nIδ,nβπ(xy)2dydx+141g(x)W(un)dxβπlncnin2bnKn2cnin2anKn1.

    By (2.56), we are back to the case of step 2 with (anKn1,bnKn1) replaced by (anKn2,bnKn1) and we can conclude

    liminfcnin2bnKn1=limsupanKn2bnKn1=limsupanKn2bnKn2=1.

    Case Ⅲ: liminfanKnbnKn2cnin1bnKn2>0 and liminfanKn1bnKn2cnin2bnKn2>0. We discuss several cases.

    Case Ⅲ-ⅰ: (bnKn2,anKn1)II+δ,n=. Then (bnKn2,anKn1)III+δ,n. We also assume liminfnanKnbnKn1+1cnin1bnKn1>0.

    We have

    n>J+δ(un)β4πbnKn2anKn2(bnKn1,anKn)II+δ,n(un(x)un(y))2(xy)2dydx+β4πbnKn2anKn2(bnKn)II+δ,n(un(x)un(y))2(xy)2dydx+β4πbnKn1anKn1(bnKn1, anKn)II+δ,n(un(x)un(y))2(xy)2dydx+β4πbnKn1anKn1(bnKn)II+δ,n(un(x)un(y))2(xy)2dydxI+δ,nIδ,nβπ(xy)2dydx+141g(x)W(un)dxβ4πbnKn2anKn2anKncnin2(un(x)un(y))2(xy)2dydxbnKn2anKn2dnin2+|(bnKn1,anKn)III+δ,n|dnin2βπ(xy)2dydx+β4πbnKn2anKn2cnin1(un(x)un(y))2(xy)2dydxbnKn2anKn2dnin1+|(bnKn,cnLn)III+δ,n|dnin1βπ(xy)2dydx+β4πbnKn11anKn1anKncnin2(un(x)un(y))2(xy)2dydxbnKn1anKn1dnin2+|(bnKn1, anKn)III+δ,n|dnin2βπ(xy)2dydxβπln2bnKn2an1βπlnbnKn1anKn1βπlnbnKnanKn+141g(x)W(un)dxβπln(cnin2anKn2cnin2bnKn2anKnbnKn2anKnanKn2)βπln(dnin2anKn2dnin2bnKn2dnin2+|(bnKn1,anKn)III+δ,n|bnKn2dnin2+|(bnKn1,anKn)III+δ,n|anKn2)+βπln(cnin2anKn1cnin2+1bnKn1anKnbnKn1+1anKnanKn1)βπln(dnin2anKn1dnin2bnKn1dnin2+|(bnKn1,anKn)III+δ,n|bnKn1dnin2+|(bnKn1,anKn)III+δ,n|anKn1)+βπlncnin1anKn2cnin1bnKn2+βπlncnin1anKn1cnin1bnKn1βπln(dnin1anKn2dnin1bnKn2dnin1+|(bnKn,cnLn)III+δ,n|bnKn2dnin1+|(bnKn,cnLn)III+δ,n|anKn2)βπln2bnKn2an1βπlnbnKn1anKn1βπlnbnKnanKnC+γmin|u|1δW(u)4|III+δ,n|βπlncnin2anKn2bnKn2+βπln(cnin2anKn1cnin2bnKn2anKnbnKn2anKnanKn2)+βπln(cnin1anKn2cnin1bnKn2anKnbnKn1+1anKnanKn1)+βπlncnin1anKn1cnin1bnKn1βπln(cnin2bnKn1+1)2βπln(1+|(bnKn1,anKn)III+δ,n|)βπln(1+|(bnKn,cnLn)III+δ,n|)βπlnbnKn1anKn1βπlnbnKnanKnC+γmin|u|1δW(u)4|III+δ,n|βπlncnin2anKn2bnKn2+βπln(cnin2anKn1cnin2bnKn2)+βπln(anKnbnKn2cnin1bnKn2)+βπln(anKnbnKn1+1cnin1bnKn1)βπlnbnKnanKnβπlnbnKn1anKn1C.

    Taking liminf on both sides, we get

    \begin{equation*} \lim \inf \frac{c_{i_{2}^{n}}^{n}}{b_{K_{n}-1}^{n}} = \lim \sup \frac{ a_{K_{n}-2}^{n}}{b_{K_{n}-2}^{n}} = 1. \end{equation*}

    If \lim \inf_{n\rightarrow \infty }\frac{a_{K_{n}}^{n}-b_{K_{n}-1}^{n}+1}{ c_{i_{1}^{n}}^{n}-b_{K_{n}-1}^{n}} = 0, then \lim \sup_{n\rightarrow \infty } \frac{c_{i_{1}^{n}}^{n}-a_{K_{n}}^{n}}{c_{i_{1}^{n}}^{n}-b_{K_{n}-1}^{n}} = 1. We can modify the argument above by replacing \left(a_{K_{n}-1, }^{n}b_{K_{n}-1}^{n}\right) \cup \left(a_{K_{n}}^{n}, b_{K_{n}}^{n}\right) by \left(a_{K_{n}-1, }^{n}b_{K_{n}}^{n}\right) and same conclusion follows.

    Case Ⅲ-ⅱ: \left(b_{K_{n}-2}^{n}, a_{K_{n}-1}^{n}\right) \cap II_{\delta, n}^{+}\neq \emptyset. There are two cases.

    Case Ⅲ-ⅱ-a: There exists i_{3}^{n}\in \left\{ j_{3}^{n}, \text{ } j_{3}^{n}+1, \text{ }\cdots, j_{2}^{n}-1\right\} such that

    \begin{equation} c_{i_{3}^{n}}^{n}-b_{K_{n}-2}^{n} < 1. \end{equation} (2.57)

    We bound J_{\delta }^{+}\left(u_{n}\right) as follows.

    \begin{eqnarray} -n & > &J_{\delta }^{+}\left( u_{n}\right) \\ &\geq &{\frac{\beta}{4\pi}}\int_{a_{K_{n}-2}^{n}}^{b_{K_{n}-2}^{n}}\int_{ \left( b_{K_{n}-2}^{n},a_{K_{n}-1}^{n}\right) \cap II_{\delta ,n}^{+}}\frac{ \left( u_{n}\left( x\right) -u_{n}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}dydx+{\frac{\beta}{4\pi}} \int_{a_{K_{n}-2}^{n}}^{b_{K_{n}-2}^{n}}\int_{\left( b_{K_{n}-1}^{n},a_{K_{n}}^{n}\right) \cap II_{\delta ,n}^{+}}\frac{\left( u_{n}\left( x\right) -u_{n}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}dydx \\ &&+{\frac{\beta}{4\pi}}\int_{a_{K_{n}-2}^{n}}^{b_{K_{n}-2}^{n}}\int_{\left( b_{K_{n}}^{n}\infty \right) \cap II_{\delta ,n}^{+}}\frac{\left( u_{n}\left( x\right) -u_{n}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}} dydx-\int_{I_{\delta ,n}^{+}}\int_{I_{\delta ,n}^{-}}\frac{\beta}{{\pi} \left( x-y\right) ^{2}}dydx+\frac{1}{4}\int_{1}^{\infty }g(x)W\left(u_{n}\right)dx \\ &\geq &{\frac{\beta}{4\pi}}\int_{a_{K_{n}-2}^{n}}^{b_{K_{n}-2}^{n}-1} \int_{c_{i_{3}^{n}}^{n}}^{a_{K_{n}-1}^{n}}\frac{\left( u_{n}\left( x\right) -u_{n}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}} dydx-\int_{a_{K_{n}-2}^{n}}^{b_{K_{n}-2}^{n}} \int_{d_{i_{3}^{n}}^{n}}^{d_{i_{3}^{n}}^{n}+\left\vert \left( b_{K_{n}-2}^{n},a_{K_{n}-1}^{n}\right) \cap III_{\delta ,n}^{+}\right\vert } \frac{\beta}{{\pi}\left( x-y\right) ^{2}}dydx \\ &&+{\frac{\beta}{4\pi}}\int_{a_{K_{n}-2}^{n}}^{b_{K_{n}-2}^{n}} \int_{c_{i_{2}^{n}}^{n}}^{a_{K_{n}}^{n}}\frac{\left( u_{n}\left( x\right) -u_{n}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}} dydx-\int_{a_{K_{n}-2}^{n}}^{b_{K_{n}-2}^{n}} \int_{d_{i_{2}^{n}}^{n}}^{d_{i_{2}^{n}}^{n}+\left\vert \left( b_{K_{n}-1}^{n},a_{K_{n}}^{n}\right) \cap III_{\delta ,n}^{+}\right\vert } \frac{\beta}{{\pi}\left( x-y\right) ^{2}}dydx \\ &&+{\frac{\beta}{4\pi}}\int_{a_{K_{n}-2}^{n}}^{b_{K_{n}-2}^{n}} \int_{c_{i_{1}^{n}}^{n}}^{\infty }\frac{\left( u_{n}\left( x\right) -u_{n}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}} dydx-\int_{a_{K_{n}-2}^{n}}^{b_{K_{n}-2}^{n}} \int_{d_{i_{1}^{n}}^{n}}^{d_{i_{1}^{n}}^{n}+\left\vert \left( b_{K_{n}}^{n},c_{L_{n}}^{n}\right) \cap III_{\delta ,n}^{+}\right\vert } \frac{\beta}{{\pi}\left( x-y\right) ^{2}}dydx \\ &&-{\frac{\beta}{\pi}}\ln 2\frac{b_{K_{n}-2}^{n}}{a_{1}^{n}}-{\frac{\beta}{ \pi}}\ln \frac{b_{K_{n}-1}^{n}}{a_{K_{n}-1}^{n}}-{\frac{\beta}{\pi}}\ln \frac{b_{K_{n}}^{n}}{a_{K_{n}}^{n}}+\frac{1}{4}\int_{1}^{\infty }g(x)W\left(u_{n}\right)dx \\ &\geq &{\frac{\beta}{\pi}}\ln \left( \frac{c_{i_{3}^{n}}^{n}-a_{K_{n}-2}^{n} }{c_{i_{3}^{n}}^{n}-b_{K_{n}-2}^{n}+1}\cdot \frac{ a_{K_{n}-1}^{n}-b_{K_{n}-2}^{n}}{a_{K_{n}-1}^{n}-a_{K_{n}-2}^{n}}\right) -{ \frac{\beta}{\pi}}\ln \left( \frac{d_{i_{3}^{n}}^{n}-a_{K_{n}-2}^{n}}{ d_{i_{3}^{n}}^{n}-b_{K_{n}-2}^{n}}\cdot \frac{d_{i_{3}^{n}}^{n}+\left\vert \left( b_{K_{n}-2}^{n},a_{K_{n}-1}^{n}\right) \cap III_{\delta ,n}^{+}\right\vert -b_{K_{n}-2}^{n}}{d_{i_{3}^{n}}^{n}+\left\vert \left( b_{K_{n}-2}^{n},a_{K_{n}-1}^{n}\right) \cap III_{\delta ,n}^{+}\right\vert -a_{K_{n}-2}^{n}}\right) \\ &&+{\frac{\beta}{\pi}}\ln \left( \frac{c_{i_{2}^{n}}^{n}-a_{K_{n}-2}^{n}}{ c_{i_{2}^{n}}^{n}-b_{K_{n}-2}^{n}}\cdot \frac{a_{K_{n}}^{n}-b_{K_{n}-2}^{n}}{ a_{K_{n}}^{n}-a_{K_{n}-2}^{n}}\right) -{\frac{\beta}{\pi}}\ln \left( \frac{ d_{i_{2}^{n}}^{n}-a_{K_{n}-2}^{n}}{d_{i_{2}^{n}}^{n}-b_{K_{n}-2}^{n}}\cdot \frac{d_{i_{2}^{n}}^{n}+\left\vert \left( b_{K_{n}-1}^{n},a_{K_{n}}^{n}\right) \cap III_{\delta ,n}^{+}\right\vert -b_{K_{n}-2}^{n}}{d_{i_{2}^{n}}^{n}+\left\vert \left( b_{K_{n}-1}^{n},a_{K_{n}}^{n}\right) \cap III_{\delta ,n}^{+}\right\vert -a_{K_{n}-2}^{n}}\right) \\ &&+{\frac{\beta}{\pi}}\ln \frac{c_{i_{1}^{n}}^{n}-a_{K_{n}-2}^{n}}{ c_{i_{1}^{n}}^{n}-b_{K_{n}-2}^{n}}-{\frac{\beta}{\pi}}\ln \left( \frac{ d_{i_{1}^{n}}^{n}-a_{K_{n}-2}^{n}}{d_{i_{1}^{n}}^{n}-b_{K_{n}-2}^{n}}\cdot \frac{d_{i_{1}^{n}}^{n}+\left\vert \left( b_{K_{n}}^{n},c_{L_{n}}^{n}\right) \cap III_{\delta ,n}^{+}\right\vert -b_{K_{n}-2}^{n}}{d_{i_{1}^{n}}^{n}+ \left\vert \left( b_{K_{n}}^{n},c_{L_{n}}^{n}\right) \cap III_{\delta ,n}^{+}\right\vert -a_{K_{n}-2}^{n}}\right) \\ &&-{\frac{\beta}{\pi}}\ln 2\frac{b_{K_{n}-2}^{n}}{a_{1}^{n}}-{\frac{\beta}{ \pi}}\ln \frac{b_{K_{n}-1}^{n}}{a_{K_{n}-1}^{n}}-{\frac{\beta}{\pi}}\ln \frac{b_{K_{n}}^{n}}{a_{K_{n}}^{n}}-C+\frac{\gamma \min_{\left\vert u\right\vert \leq 1-\delta }W\left( u\right) }{4}\left\vert III_{\delta ,n}^{+}\right\vert \\ &\geq &{\frac{\beta}{\pi}}\ln \frac{c_{i_{3}^{n}}^{n}-a_{K_{n}-2}^{n}}{ b_{K_{n}-2}^{n}}+{\frac{\beta}{\pi}}\ln \left( \frac{ a_{K_{n}-1}^{n}-b_{K_{n}-2}^{n}}{a_{K_{n}-1}^{n}-a_{K_{n}-2}^{n}}\cdot \frac{ c_{i_{2}^{n}}^{n}-a_{K_{n}-2}^{n}}{c_{i_{2}^{n}}^{n}-b_{K_{n}-2}^{n}}\right) +{\frac{\beta}{\pi}}\ln \left( \frac{c_{i_{1}^{n}}^{n}-a_{K_{n}-2}^{n}}{ c_{i_{1}^{n}}^{n}-b_{K_{n}-2}^{n}}\cdot \frac{a_{K_{n}}^{n}-b_{K_{n}-2}^{n}}{ a_{K_{n}}^{n}-a_{K_{n}-2}^{n}}\right) \\ &&-{\frac{\beta}{\pi}}\ln \left( c_{i_{3}^{n}}^{n}-b_{K_{n}-2}^{n}+1\right) - {\frac{\beta}{\pi}}\ln \left( 1+\left\vert \left( b_{K_{n}-2}^{n},a_{K_{n}-1}^{n}\right) \cap III_{\delta ,n}^{+}\right\vert \right) -{\frac{\beta}{\pi}}\ln \left( 1+\left\vert \left( b_{K_{n}-1}^{n},a_{K_{n}}^{n}\right) \cap III_{\delta ,n}^{+}\right\vert \right) \\ &&-{\frac{\beta}{\pi}}\ln \left( 1+\left\vert \left( b_{K_{n}}^{n},c_{L_{n}}^{n}\right) \cap III_{\delta ,n}^{+}\right\vert \right) -{\frac{\beta}{\pi}}\ln \frac{b_{K_{n}-1}^{n}}{a_{K_{n}-1}^{n}}-{ \frac{\beta}{\pi}}\ln \frac{b_{K_{n}}^{n}}{a_{K_{n}}^{n}}-C+\frac{\gamma \min_{\left\vert u\right\vert \leq 1-\delta }W\left( u\right) }{4}\left\vert III_{\delta ,n}^{+}\right\vert \\ &\geq &{\frac{\beta}{\pi}}\ln \frac{c_{i_{3}^{n}}^{n}-a_{K_{n}-2}^{n}}{ b_{K_{n}-2}^{n}}+{\frac{\beta}{\pi}}\ln \left( \frac{ a_{K_{n}}^{n}-b_{K_{n}-2}^{n}}{c_{i_{1}^{n}}^{n}-b_{K_{n}-2}^{n}}\cdot \frac{ a_{K_{n}-1}^{n}-b_{K_{n}-2}^{n}}{c_{i_{2}^{n}}^{n}-b_{K_{n}-2}^{n}}\right) -{ \frac{\beta}{\pi}}\ln \frac{b_{K_{n}}^{n}}{a_{K_{n}}^{n}}-{\frac{\beta}{\pi}} \ln \frac{b_{K_{n}-1}^{n}}{a_{K_{n}-1}^{n}}-C. \end{eqnarray} (2.58)

    Taking liminf on both sides, we conclude

    \begin{equation*} \lim \inf \frac{c_{i_{3}^{n}}^{n}}{b_{K_{n}-2}^{n}} = \lim \sup \frac{ a_{K_{n}-2}^{n}}{b_{K_{n}-2}^{n}} = 1. \end{equation*}

    Case Ⅲ-ⅱ-b: no such i_{3}^{n} satisfying (2.57) exists. Then we must have

    \begin{equation} c_{j_{3}^{n}}^{n}-b_{K_{n}-2}^{n}\geq 1 \end{equation} (2.59)

    In this case, we can estimate J_{\delta }^{+}\left(u_{n}\right) in the same way as case Ⅲ-ⅰ with c_{i_{3}^{n}}^{n} replaced by c_{j_{3}^{n}}^{n} and conclude

    \begin{equation*} \lim \inf \frac{c_{j_{3}^{n}}^{n}}{b_{K_{n}-2}^{n}} = \lim \sup \frac{ a_{K_{n}-2}^{n}}{b_{K_{n}-2}^{n}} = 1. \end{equation*}

    Continuing this way, we conclude that J_{\delta }^{+}\left(u_{n}\right) \rightarrow -\infty implies

    \begin{equation*} \lim \sup \frac{a_{K_{n}}^{n}}{b_{K_{n}}^{n}} = \lim \sup \frac{ a_{K_{n}-1}^{n} }{b_{K_{n}-1}^{n}} = \cdots = \lim \sup \frac{a_{1}^{n}}{ b_{1}^{n}} = 1. \end{equation*}

    We now pick our subsequence as follows. Pick our first subsequence \left\{ u_{n}\right\} such that for its decomposition

    \begin{equation*} \ln \frac{a_{K_{n}}^{n}}{b_{K_{n}}^{n}} < \frac{1}{2}\text{ for all }n. \end{equation*}

    Next we pick a subsequence of the chosen subsequence such that

    \begin{equation*} \ln \frac{a_{K_{n}-1}^{n}}{b_{K_{n}-1}^{n}} < \frac{1}{4}\text{ for all }n. \end{equation*}

    Continuing this way, we pick our final subsequence using a diagonal argument. For simplicity of notations, we use the original sequence. For the final subsequence, we have for all n,

    \begin{equation*} \ln \frac{a_{K_{n}}^{n}}{b_{K_{n}}^{n}} < \frac{1}{2},\ln \text{ }\frac{ a_{K_{n}-1}^{n}}{b_{K_{n}-1}^{n}} < \frac{1}{4},\cdots ,\text{ ln}\frac{ a_{K_{n}-l}^{n}}{b_{K_{n}-l}^{n}} < \frac{1}{2^{l}},\cdots \end{equation*}

    it then follows that

    \begin{eqnarray*} \int_{I_{\delta ,n}^{+}}\int_{I_{\delta ,n}^{-}}\frac{4}{\left( x-y\right) ^{2}}dydx &\leq &4\ln \prod\limits_{i = 1}^{K_{n}}\frac{a_{i}^{n}+\widetilde{d} _{\widetilde{L}}}{b_{i}^{n}+\widetilde{d}_{\widetilde{L}}}\cdot \frac{ b_{i}^{n}+\widetilde{c}_{1}}{a_{i}^{n}+\widetilde{c}_{1}} \\ &\leq &4\sum\limits_{i = 1}^{K_{n}}\ln \frac{b_{i}^{n}}{a_{i}^{n}} < 4, \end{eqnarray*}

    a contradiction to the assumption that J_{\delta}^+\left(u_{n}\right) \rightarrow -\infty.

    In this section, we prove the existence of a minimizer of J\left(u\right). Observe that in general the boundedness of J\left(u_{n}\right) does not imply the boundedness of v_{n} = u_{n}-\eta in H^{1}\left(\mathbb{R} \right). A priori it is not clear whether we can obtain a suitable limit function from a minimizing sequence, using the direct method of calculus of variations. On the other hand, for a sequence \left\{ u_{n}\right\} satisfying \left\vert u_{n}\left(x\right) +sgn(x)\right\vert \geq c_{0} > 0 outside a uniformly bounded interval, the boundedness of J\left(u_{n}\right) implies the boundedness of \left\{ u_{n}-\eta \right\} in H^{1}\left(\mathbb{R}\right). Our main idea, therefore, is to show that we can replace a minimizing sequence \left\{ u_{n}\right\} by another one \{\overline{u}_{n}\} which satisfies \left\vert \overline{u}_{n}\left(x\right) +sgn(x)\right\vert \geq c_{0} > 0 for some constant c_{0} outside a uniformly bounded interval. The new sequence \left\{ \overline{u} _{n}\right\} has uniformly bounded energy J\left(\overline{u}_{n}\right) , and \left\{ \overline{u} _{n}-\eta \right\} is bounded in H^{1}\left(\mathbb{R}\right). From this, we obtain a subsequence which weakly converges to a limit function \overline{v}\in \mathcal{A} that achieves the minimum energy in \mathcal{A} . To construct the replacement sequence \left\{ \overline{u}_{n}\right\} , we use the interval decompositions of u_{n} from the previous section. We first construct \widetilde{u}_{n} by reflecting u_{n} over suitable regions. Keeping track of the energy contributions from each interval in the decomposition, we show that we can choose our regions of reflection so that the energy difference between J\left(\widetilde{u}_{n}\right) and J\left(u_{n}\right) is approaching zero as n\rightarrow \infty. Lastly, we define \overline{u} _{n} by a suitable translation of \widetilde{u}_{n} \ so that \left\vert \overline{u}_{n}\left(x\right) +sgn(x)\right\vert \geq c_{0} > 0 for some constant c_{0} outside a uniformly bounded interval. By periodic translation invariance of J and the above property of J\left(\widetilde{u}_{n}\right), we conclude that \left\{ \overline{u} _{n}\right\} is another minimizing sequence.

    We first state a translation invariant lemma.

    Lemma 3.1. Given any c\in \mathbb{Z}, let u_{c}\left(x\right) = u\left(x+c\right), then J\left(u_{c}\left(x\right) \right) = J\left(u\left(x\right) \right).

    Proof. Since the first two terms are translation invariant for c\in \mathbb{Z} , J(u_{c}) = J(u)+D(\eta _{c}, \eta) , where

    \begin{equation*} D\left( \eta _{c},\eta \right) = \int_{\mathbb{R}}\int_{\mathbb{R}}\left( \frac{\left( \eta _{c}\left( x\right) -\eta _{c}\left( y\right) \right) ^{2} }{\left( x-y\right) ^{2}}-\frac{\left( \eta \left( x\right) -\eta \left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}\right) dydx. \end{equation*}

    By Lemma 2.1 in [29], we have D\left(\eta _{c}, \eta \right) = 0 for any constant c. The conclusion of the Lemma then follows.

    Let \left\{ u_{n}\right\} be a minimizing sequence. By Lemma 2.3, we may assume that u_n-\eta is compactly supported in \mathbb{R} . By Lemma 3.1 and our assumption on the behavior of \left\{ u_{n}\right\} at infinity, after a suitable translation by an integer there exists c_{n}\in [ 0, 1) such that u_{n}\left(1+c_{n}\right) = 0 and

    \begin{equation*} u_{n}\left( x\right) \leq 0\text{ for }x\leq 1+c_{n}. \end{equation*}

    Throughout this section, we assume that on \left[ 1+c_{n}, \infty \right), u_{n} has a decomposition

    \begin{equation} I_{\delta ,n}^{+} = \bigcup\limits_{i = 1}^{K_{n}}\left[ a_{i,}^{n}\text{ } b_{i}^{n}\right] \end{equation} (3.1)

    and

    \begin{equation} II_{\delta ,n}^{+} = \bigcup\limits_{j = 1}^{L_{n}}\left[ c_{j}^{n},d_{j}^{n} \right] . \end{equation} (3.2)

    Here we understand d_{L_{n}}^{n} = \infty. Throughout this section, we fix \delta \ll 1 such that W^{\prime \prime }\left(u\right) \geq \frac{1}{4} W^{\prime \prime }\left(1\right) when \left(1-u\right) \leq \delta and W^{\prime \prime }\left(u\right) \geq \frac{1}{4}W^{\prime \prime }\left(-1\right) when 1+u\geq \delta. Since W\left(u\right) > 0 for u\in \left(-1, 1\right), there exists C_{\delta } > 0 such that

    \begin{equation} W\left( u\right) \geq C_{\delta }\left( 1-u\right) ^{2} \ \text{when } 1+u\geq \delta \end{equation} (3.3)

    and

    \begin{equation} W\left( u\right) \geq C_{\delta }\left( 1+u\right) ^{2} \ \text{when } 1-u\geq \delta .\text{ } \end{equation} (3.4)

    In this case, we prove the following proposition.

    Proposition 3.2. Let \delta > 0 be such that W\left(u\right) satisfies (3.3) and (3.4). Let u_n be a minimizing sequence for J(u) in \mathcal{A}_0 with decompositions (3.1) and (3.2). If there exists a constant M > 1 such that b_{K_{n}}^{n} < M for all n , then a subsequence of \left\{ u_{n}\right\} converges weakly to a minimizer u_{0} of J\left(u\right) in \mathcal{A} .

    Proof. Without loss of generality, we can assume \left\vert u_{n}\right\vert \leq 1. Consider v_{n}: = u_{n}-\eta. Decomposition (3.1) and our assumption imply

    \begin{equation} u_{n}+1\geq \delta \text{ for }x\geq M \end{equation} (3.5)

    and

    \begin{equation} u_{n}-1\leq -1\text{ for }x\leq 1+c_{n},\text{ }c_{n}\in [ 0,1) \end{equation} (3.6)

    We write J\left(u_{n}\right) in terms of v_{n} as follows:

    \begin{eqnarray} J\left( u_{n}\right) & = &\int_{\mathbb{R}}\left[ {\frac{\alpha }{2}} \left\vert u_{n}^{\prime }\right\vert ^{2}+g\left( x\right) W\left( u_{n}\right) \right] dx+{\frac{\beta }{4 \pi}} \int_{ \mathbb{R}}\int_{ \mathbb{R}}\left( \frac{\left( u_{n}\left( x\right) -u_{n}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}-\frac{\left( \eta \left( x\right) -\eta \left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}\right) dydx \\ & = &\int_{\mathbb{R}}\left[ {{\frac{\alpha }{2}}} \left\vert u_{n}^{\prime }\right\vert ^{2}+g\left( x\right) W\left( u_{n}\right) \right] dx+{{\frac{ \beta }{4 \pi}}}\int_{\mathbb{R}}\int_{ \mathbb{R}}\frac{\left( v_{n}\left( x\right) -v_{n}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}dydx \\ &&-{{\frac{\beta }{2 \pi}}}\int_{\mathbb{R}}\int_{\mathbb{R}}\frac{\left( v_{n}\left( x\right) -v_{n}\left( y\right) \right) \left( \eta \left( x\right) -\eta \left( y\right) \right) }{\left( x-y\right) ^{2}}dydx. \end{eqnarray} (3.7)

    We have

    \begin{eqnarray} &&-\int_{\mathbb{R}}\int_{\mathbb{R}}\frac{\left( v_{n}\left( x\right) -v_{n}\left( y\right) \right) \left( \eta \left( x\right) -\eta \left( y\right) \right) }{\left( x-y\right) ^{2}}dydx \\ & = &-2\int_{1}^{\infty }\int_{-1}^{1}\frac{\left( v_{n}\left( x\right) -v_{n}\left( y\right) \right) \left( \eta \left( x\right) -\eta \left( y\right) \right) }{\left( x-y\right) ^{2}}dydx-2\int_{1}^{\infty }\int_{-\infty }^{-1}\frac{\left( v_{n}\left( x\right) -v_{n}\left( y\right) \right) \left( \eta \left( x\right) -\eta \left( y\right) \right) }{\left( x-y\right) ^{2}}dydx \\ &&-2\int_{-\infty }^{-1}\int_{-1}^{1}\frac{\left( v_{n}\left( x\right) -v_{n}\left( y\right) \right) \left( \eta \left( x\right) -\eta \left( y\right) \right) }{\left( x-y\right) ^{2}}dydx-\int_{-1}^{1}\int_{-1}^{1} \frac{\left( v_{n}\left( x\right) -v_{n}\left( y\right) \right) \left( \eta \left( x\right) -\eta \left( y\right) \right) }{\left( x-y\right) ^{2}}dydx \\ &\geq &-4\int_{1}^{\infty }\frac{v_{n}\left( x\right) }{x+1} dx+4\int_{-\infty }^{-1}\frac{v_{n}\left( y\right) }{1-y}dy-\frac{1}{4} \int_{1}^{\infty }\int_{-1}^{1}\frac{\left( v_{n}\left( x\right) -v_{n}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}dydx \\ &&-4\int_{1}^{\infty }\int_{-1}^{1}\frac{\left( \eta \left( x\right) -\eta \left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}dydx-\frac{1}{4} \int_{-\infty }^{-1}\int_{-1}^{1}\frac{\left( v_{n}\left( x\right) -v_{n}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}dydx \\ &&-4\int_{-\infty }^{-1}\int_{-1}^{1}\frac{\left( \eta \left( x\right) -\eta \left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}dydx-\frac{1}{4} \int_{-1}^{1}\int_{-1}^{1}\frac{\left( v_{n}\left( x\right) -v_{n}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}-\int_{-1}^{1}\int_{-1}^{1} \frac{\left( \eta \left( x\right) -\eta \left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}} \\ &\geq &-\varepsilon \int_{1}^{\infty }g(x)v_{n}^{2}\left( x\right) dx-\varepsilon \int_{-\infty }^{-1} g(y) v_{n}^{2}\left( y\right) dy-\frac{C}{ \varepsilon }-C\left( \left\Vert \eta ^{\prime }\right\Vert _{L^{\infty }}\right) -\frac{1}{4}\int_{1}^{\infty }\int_{-1}^{1}\frac{\left( v_{n}\left( x\right) -v_{n}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}dydx \\ &&-\frac{1}{4}\int_{-\infty }^{-1}\int_{-1}^{1}\frac{\left( v_{n}\left( x\right) -v_{n}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}dydx- \frac{1}{4}\int_{-1}^{1}\int_{-1}^{1}\frac{\left( v_{n}\left( x\right) -v_{n}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}dydx. \end{eqnarray} (3.8)

    The last inequality in (3.8) follows from Hö lder inequality, (1.2) and bounds on

    \begin{equation*} \int_{-\infty }^{-1}\int_{-1}^{1}\frac{\left( \eta \left( x\right) -\eta \left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}dydx, \qquad \int_{1}^{\infty }\int_{-1}^{1}\frac{\left( \eta \left( x\right) -\eta \left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}dydx. \end{equation*}

    By (3.3), (3.4), (3.7) and (3.8), we have

    \begin{eqnarray*} C &\geq &J\left( u_{n}\right) \geq \int_{\mathbb{R}}{{\frac{\alpha }{2}}} \left\vert u_{n}^{\prime }\right\vert ^{2}dx+\int_{1}^{\infty } g W\left( u_{n}\right) dx+\int_{-\infty }^{-1} g W\left( u_{n}\right) dx \\ &&-{{\frac{\beta }{2 \pi}}}\varepsilon \int_{1}^{\infty }g(x)v_{n}^{2}\left( x\right) dx-{{\frac{\beta }{2 \pi}}}\varepsilon \int_{-\infty }^{-1} g(y)v_{n}^{2}\left( y\right) dy-{\frac{C\beta}{2\pi\varepsilon }}-C\left( \left\Vert \eta ^{\prime }\right\Vert _{L^{\infty }}\right) \\ &&+{\frac{3\beta}{8\pi}}\int_{-\infty }^{-1}\int_{-1}^{1}\frac{\left( v_{n}\left( x\right) -v_{n}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}dydx+{{\frac{\beta }{4 \pi}}}\int_{-\infty }^{-1}\int_{-\infty }^{-1} \frac{\left( v_{n}\left( x\right) -v_{n}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}dydx \\ &&+{{\frac{\beta }{4 \pi}}}\int_{1}^{\infty }\int_{1}^{\infty }\frac{\left( v_{n}\left( x\right) -v_{n}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}dydx+{\frac{\beta}{8\pi}}\int_{-1}^{1}\int_{-1}^{1}\frac{\left( v_{n}\left( x\right) -v_{n}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}dydx \\ &&+{\frac{3\beta}{8\pi}}\int_{1}^{\infty }\int_{-1}^{1}\frac{\left( v_{n}\left( x\right) -v_{n}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}dydx+{{\frac{\beta}{2\pi}}}\int_{-\infty }^{-1}\int_{1}^{\infty }\frac{ \left( v_{n}\left( x\right) -v_{n}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}dydx \\ &\geq &{{\frac{\alpha }{2}}}\int_{\mathbb{R}}\left\vert u_{n}^{\prime }\right\vert ^{2}dx+C_{\delta}\int_{\mathbb{R}} gv_{n}^{2}dx+{\frac{ \beta}{8\pi}}\int_{\mathbb{R}}\int_{\mathbb{R}}\frac{\left( v_{n}\left( x\right) -v_{n}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}} dydx-C\left( M,\left\Vert \eta ^{\prime }\right\Vert _{L^{\infty }}\right) . \end{eqnarray*}

    From this, we conclude that v_{n} is bounded in H^{1}\left(\mathbb{R} \right) , and, hence, there exists a subsequence v_{n_{k}}\rightharpoonup v\in H^{1}\left(\mathbb{R}\right) and

    \begin{equation*} \int_{\mathbb{R}}\int_{\mathbb{R}}\frac{\left( v_{n_k}\left( x\right) -v_{n_k}\left( y\right) \right) \left( \eta \left( x\right) -\eta \left( y\right) \right) }{\left( x-y\right) ^{2}}dydx\rightarrow \int_{\mathbb{R} }\int_{\mathbb{R}}\frac{\left( v\left( x\right) -v\left( y\right) \right) \left( \eta \left( x\right) -\eta \left( y\right) \right) }{\left( x-y\right) ^{2}}dydx. \end{equation*}

    Here the convergence above follows from the identity

    \begin{equation*} \int_{\mathbb{R}}\int_{\mathbb{R}}\frac{\left( v_{n_k}\left( x\right) -v_{n_k}\left( y\right) \right) \left( \eta \left( x\right) -\eta \left( y\right) \right) }{\left( x-y\right) ^{2}}dydx = 2\int_{\mathbb{R} }v_{n_k}(x)\lim\limits_{\varepsilon\rightarrow 0}\int_{|x-y|\geq \varepsilon}\frac{ \eta(x)-\eta(y)}{|x-y|^2}dydx, \end{equation*}

    the fact that

    \begin{equation*} \lim\limits_{\varepsilon \rightarrow 0}\int_{|x-y|\geq \varepsilon}\frac{ \eta(x)-\eta(y)}{|x-y|^2}dy \end{equation*}

    lies in L^2(\mathbb{R}) (see the discussion around (4.5) in Section 4) and the weak convergence of v_{n_k} in L^2(\mathbb{R}) .

    Let u_{0} = v+\eta, then u_{0}\in \mathcal{A} and

    \begin{eqnarray*} &&\lim \inf J\left( u_{n}\right) \\ & = &\lim \inf \left\{ \int_{\mathbb{R}}\left[ {{\frac{\alpha }{2}}}\left\vert u_{n}^{\prime }\right\vert ^{2}+g\left( x\right) W\left( u_{n}\right) \right] dx\right. \\ &&\left. +{{\frac{\beta }{4 \pi}}}\int_{\mathbb{R}}\int_{\mathbb{R}}\left[ \frac{\left( v_{n}\left( x\right) -v_{n}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}-2\frac{\left( v_{n}\left( x\right) -v_{n}\left( y\right) \right) \left( \eta \left( x\right) -\eta \left( y\right) \right) }{\left( x-y\right) ^{2}}\right] dydx\right\} \\ &\geq &\int_{\mathbb{R}}\left[ {{\frac{\alpha }{2}}}\left\vert u_{0}^{\prime }\right\vert ^{2}+g\left( x\right) W\left( u_{0}\right) \right] dx+{{\frac{ \beta }{4 \pi}}}\int_{\mathbb{R}}\int_{\mathbb{R}}\frac{\left( v\left( x\right) -v\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}dydx \\ &&-{{\frac{\beta }{2 \pi}}}\int_{\mathbb{R}}\int_{\mathbb{R}}\frac{\left( v\left( x\right) -v\left( y\right) \right) \left( \eta \left( x\right) -\eta \left( y\right) \right) }{\left( x-y\right) ^{2}}dydx \\ & = &J\left( u_{0}\right) . \end{eqnarray*}

    We prove the following proposition.

    Proposition 3.3. If \lim \sup b_{K_{n}}^{n} = \infty, we can find a new minimizing sequence \left\{ \overline{u}_{n}\right\} and \ \overline{c} _{n}\in \left[ 0, 1\right) such that

    \begin{equation*} \overline{u}_{n}\leq 0 \ \ \ \mathit{{for}} \ \ \ \ x\leq 1+\overline{c}_{n} \end{equation*}

    and

    \begin{equation*} \left[ 1+\overline{c}_{n},\mathit{{}}\infty \right) \cap I_{\delta ,n}^{+} = \cup _{i = 1}^{\overline{K}_{n}}\left[ a_{i}^{n},b_{i}^{n}\right] \end{equation*}

    with \lim \sup \overline{b}_{\overline{K}_{n}}^{n} < \infty .

    Proof. Let \rho _{i}^{n} < a_{i}^{n} be the biggest zero point of u_{n} that lies to the left of a_{i}^{n} and \sigma _{i}^{n} > b_{i}^{n} be the smallest zero of u_{n} lying to the right of b_{i}^{n}. Set

    \begin{equation*} A_{n}^{+} = \left\{ x\geq 1+c_{n},\text{ }u_{n}\left( x\right) \geq 0\right\} , \end{equation*}

    and

    \begin{equation*} A_{n}^{-} = \left\{ x\geq 1+c_{n},\text{ }u_{n}\left( x\right) \leq 0\right\} , \end{equation*}

    We construct our replacement minimizing sequence \left\{ \widetilde{u} _{n}\right\} in two cases.

    Case Ⅰ: There exists 0\leq l\leq K_{n}, such that \lim \sup_{n}\frac{a_{K_{n}-l}^{n}}{b_{K_{n}-l}^{n}} = \nu _{l} < 1 and \lim \sup_{n}\frac{a_{K_{n}-i}^{n}}{b_{K_{n}-i}^{n}} = 1 for all i = 0, 1, \cdots, l-1.

    The main idea to construct a replacement minimizing sequence in this case is to show that we must have \lim \sup_{n}\left(a_{K_{n}-l}^{n}-b_{1}^{n}\right) < \infty and \lim \sup_{n}\frac{a_{1}^{n} }{b_{K_{n}-l}^{n}} = 0. We then reflect the positive part of u_{n} defined on \left[ 1+c_{n}, \rho _{K_{n}-l}^{n}\right] to -u_{n} . It can be shown that the energy of the resulting function differs from initial minimizing sequence by a small amount, a suitable translation of this reflected minimizing sequence satisfies the assumption in Proposition 3.2 and we can obtain a limit function from this replacement minimizing sequence. To illustrate our main idea, we first assume l = 0.

    Case Ⅰ-ⅰ: \lim \sup_{n}\frac{a_{K_{n}}^{n}}{b_{K_{n}}^{n}} = 0. By definition of \rho _{i}^{n}, \sigma _{i}^{n}, we have \lim \sup_{n} \frac{\rho _{K_{n}}^{n}}{\sigma _{K_{n}}^{n}} = 0. In this case, we consider the sequence \left\{ \widetilde{u}_{n}\right\} defined by

    \begin{equation} \widetilde{u}_{n}\left( x\right) = \left\{ \begin{array}{cc} -u_{n}\left( x\right) & x\in \left[ 1+c_{n},\rho _{K_{n}}^{n}\right] \cap A_{n}^{+} \\ u_{n}\left( x\right) & \text{ otherwise} \end{array} .\right. \end{equation} (3.9)

    Then

    \begin{eqnarray} {{\frac{4\pi}{\beta}}}\left[J\left( u_{n}\right) -J\left( \widetilde{u} _{n}\right)\right] & = &\int_{\mathbb{R}}\int_{\mathbb{R}}\left[ \frac{\left( u_{n}\left( x\right) -u_{n}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}-\frac{\left( \widetilde{u}_{n}\left( x\right) -\widetilde{u}_{n}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}\right] dydx \\ & = &-8\int_{\left[ 1+c_{n},\rho _{K_{n}}^{n}\right] \cap A_{n}^{+}}\int_{ \mathbb{R}\backslash \left( \left[ 1+c_{n},\rho _{K_{n}}^{n}\right] \cap A_{n}^{+}\right) }\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ &\geq &-8\int_{1+c_{n}}^{\rho _{K_{n}}^{n}}\int_{\sigma _{K_{n}}^{n}}^{\infty }\frac{1}{\left( x-y\right) ^{2}}dydx \\ & = &-8\ln \frac{\sigma _{K_{n}}^{n}-1-c_{n}}{\sigma _{K_{n}}^{n}-\rho _{K_{n}}^{n}}. \end{eqnarray} (3.10)

    Since

    \begin{equation*} \lim \sup\limits_{n}\frac{\rho _{K_{n}}^{n}}{\sigma _{K_{n}}^{n}} = 0, \end{equation*}

    (3.10) implies the subsequence of \left\{ \widetilde{u}_{n}\right\} is also a minimizing sequence. Let s_{n} = \left[ \sigma _{K_{n}}^{n}\right] be the largest integer smaller than \sigma _{K_{n}}^{n}. By periodic translation invariance of the energy, we define \overline{u}_{n}\left(x\right) = \widetilde{u}_{n}\left(x+s_{n}-1\right) . Then J\left(\overline{u}_{n}\right) = J\left(\widetilde{u}_{n}\right) and \overline{u}_{n} satisfies

    \begin{equation*} \overline{u}_{n}\left( x\right) \leq 0\text{ for }x\leq 1+\sigma _{K_{n}}^{n}-s_{n}, \end{equation*}

    and

    \begin{equation*} \overline{u}_{n}\left( x\right) \geq -1+\delta \text{ for }x\geq 1+\sigma _{K_{n}}^{n}-s_{n}. \end{equation*}

    We conclude from Proposition 3.2 that \overline{u}_{n} is bounded in H^{1}\left(\mathbb{R}\right), and \overline{u}_{n} converges weakly to a minimizer in \mathcal{A} .

    Case Ⅰ-ⅱ: \lim \sup_{n}\frac{a_{K_{n}}^{n}}{b_{K_{n}}^{n}} = \nu < 1. Then we must have \lim \sup_{n}\left(a_{K_{n}}^{n}-b_{1}^{n}\right) < \infty and \lim \sup_{n}\frac{ a_{1}^{n}}{b_{K_{n}}^{n}} = 0. In this case, we prove that we essentially get back to the same situation as case Ⅰ-ⅰ, with a_{K_{n}}^{n} replaced by a_{1}^{n}. By estimates in Section 2 (we use the same notations in this section)

    \begin{eqnarray} J_{\delta }^{+}\left( u_{n}\right) & = &{\frac{\beta}{4\pi}}\int_{I_{\delta ,n}^{+}}\int_{II_{\delta ,n}^{+}}\frac{\left( u_{n}\left( x\right) -u_{n}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}} dydx-\int_{I_{\delta ,n}^{+}}\int_{I_{\delta ,n}^{-}}\frac{\beta}{{\pi} \left( x-y\right) ^{2}}dydx+\frac{1}{4}\int_{1}^{\infty }g(x)W\left( u_{n}\right) dx \\ &\geq &{\frac{\beta}{4\pi}}\int_{a_{K_{n}}^{n}}^{b_{K_{n}}^{n}} \int_{c_{i_{1}^{n}}^{n}}^{\infty }\frac{\left( u_{n}\left( x\right) -u_{n}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}dydx-{\frac{\beta }{4\pi}}\int_{a_{K_{n}}^{n}}^{b_{K_{n}}^{n}} \int_{d_{i_{1}^{n}}^{n}}^{d_{i_{1}^{n}}^{n}+\left\vert \left( b_{K_{n}}^{n},\infty \right) \cap III_{\delta ,n}^{+}\right\vert }\frac{ \left( u_{n}\left( x\right) -u_{n}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}dydx \\ &&-\int_{I_{\delta ,n}^{+}}\int_{I_{\delta ,n}^{-}}\frac{\beta}{{\pi}\left( x-y\right) ^{2}}dydx+\frac{1}{4}\int_{1}^{\infty }g(x)W\left( u_{n}\right) dx \\ &\geq &{\frac{\beta}{\pi}}\ln \frac{c_{i_{1}^{n}}^{n}-a_{K_{n}}^{n}}{ c_{i_{1}^{n}}^{n}-b_{K_{n}}^{n}}-{\frac{\beta}{\pi}}\ln \left( 2\frac{ b_{K_{n}}^{n}}{a_{1}^{n}}\right) +\frac{\gamma \min_{\left\vert u\right\vert \leq 1-\delta }W\left( u\right) }{4}\left\vert III_{\delta ,n}^{+}\right\vert -{\frac{\beta}{\pi}}\ln \left( 1+\left\vert \left( b_{K_{n}}^{n},\infty \right) \cap III_{\delta ,n}^{+}\right\vert \right) -C \\ &\geq &{\frac{\beta}{\pi}}\ln \frac{c_{i_{1}^{n}}^{n}-a_{K_{n}}^{n}}{ b_{K_{n}}^{n}}-C+\frac{\gamma \min_{\left\vert u\right\vert \leq 1-\delta }W\left( u\right) }{4}\left\vert III_{\delta ,n}^{+}\right\vert -{\frac{\beta }{\pi}}\ln \left( c_{i_{1}^{n}}^{n}-b_{K_{n}}^{n}\right) \\ &&-{\frac{\beta}{\pi}}\ln \left( 1+\left\vert \left( b_{K_{n}}^{n},\infty \right) \cap III_{\delta ,n}^{+}\right\vert \right) . \end{eqnarray} (3.11)

    Assuming \lim \sup_{n}\frac{a_{K_{n}}^{n}}{b_{K_{n}}^{n}} = \nu, (3.11) and the boundedness of J_{\delta }^{+}\left(u_{n}\right) imply \left\vert III_{\delta, n}^{+}\right\vert \leq C. We construct \widetilde{u}_{n} as follows.

    \begin{equation} \widetilde{u}_{n} = \left\{ \begin{array}{c} -u_{n}\left( x\right) \ \ \ x\in \left[ 1+c_{n},\rho _{K_{n}}^{n} \right] \cap A_{n}^{+} \\ u_{n}\left( x\right) \ \ \ \ \ \ \ \text{elsewhere}\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \end{array} .\right. \end{equation} (3.12)

    We first show that there does not exist an s such that \lim \sup_{n}\left(a_{K_{n}}^{n}-a_{s+1}^{n}\right) = A_{s} < \infty, \lim \sup_{n}\left(a_{K_{n}}^{n}-b_{s}\right) = \infty .

    Otherwise, letting t_{n} = \left\vert \left[ 1+c_{n}, \rho _{K_{n}}^{n}\right] \cap A_{n}^{+}\right\vert, t_{s}^{n} = \left\vert \left[ b_{s}^{n}, a_{s+1}^{n}\right] \cap II_{\delta, n}^{+}\right\vert, we can write

    \begin{eqnarray} &&{{\frac{4\pi }{\beta}}}\left[J\left( u_{n}\right) -J\left( \widetilde{u} _{n}\right)\right] \\ & = &\int_{\mathbb{R}}\int_{\mathbb{R}}\left[ \frac{\left( u_{n}\left( x\right) -u_{n}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}-\frac{ \left( \widetilde{u}_{n}\left( x\right) -\widetilde{u}_{n}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}\right] dydx \\ & = &-8\int_{\left[ 1+c_{n},\text{ }\rho _{K_{n}}^{n}\right] \cap A_{n}^{+}}\int_{\mathbb{R}\backslash \left( \left[ 1+c_{n},\rho _{K_{n}}^{n} \right] \cap A_{n}^{+}\right) }\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ & = &-8\int_{\left[ 1+c_{n},\text{ }\rho _{K_{n}}^{n}\right] \cap A_{n}^{+}}\int_{\sigma _{K_{n}}^{n}}^{\infty }\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx-8\int_{\left[ 1+c_{n}, \text{ }\rho _{K_{n}}^{n}\right] \cap A_{n}^{+}}\int_{-\infty }^{1+c_{n}} \frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ &&-8\int_{\left[ 1+c_{n},\text{ }\rho _{K_{n}}^{n}\right] \cap A_{n}^{+}}\int_{\left[ 1+c_{n},\text{ }\rho _{K_{n}}^{n}\right] \cap A_{n}^{-}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx-8\int_{\left[ 1+c_{n},\text{ }\rho _{K_{n}}^{n}\right] \cap A_{n}^{+}}\int_{\rho _{K_{n}}^{n}}^{\sigma _{K_{n}}^{n}}\frac{ u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ &\geq &-8\int_{\rho _{K_{n}}^{n}-t_{n}}^{\rho _{K_{n}}^{n}}\int_{\sigma _{K_{n}}^{n}}^{\infty }\frac{1}{\left( x-y\right) ^{2}}dydx-8\int_{\left[ 1+c_{n},\rho _{K_{n}}^{n}\right] \cap A_{n}^{+}}\int_{\rho _{K_{n}}^{n}}^{\sigma _{K_{n}}^{n}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ &\geq &-8\int_{\rho _{K_{n}}^{n}-t_{n}}^{\rho _{K_{n}}^{n}}\int_{\sigma _{K_{n}}^{n}}^{\infty }\frac{1}{\left( x-y\right) ^{2}}dydx-8\int_{\left[ b_{s}^{n},a_{s+1}^{n}\right] \cap II_{\delta ,n}^{+}}\int_{a_{K_{n}}^{n}}^{b_{K_{n}}^{n}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ &\geq &-8\ln \frac{\sigma _{K_{n}}^{n}-\rho _{K_{n}}^{n}+t_{n}}{\sigma _{K_{n}}^{n}-\rho _{K_{n}}^{n}}+8\int_{b_{s}^{n}}^{b_{s}^{n}+t_{s}^{n}} \int_{a_{K_{n}}^{n}}^{b_{K_{n}}^{n}}\frac{\left( 1-\delta \right) ^{2}}{ \left( x-y\right) ^{2}}dydx \\ & = &-8\ln \frac{\sigma _{K_{n}}^{n}-\rho _{K_{n}}^{n}+t_{n}}{\sigma _{K_{n}}^{n}-\rho _{K_{n}}^{n}} \\ &&+8\left( 1-\delta \right) ^{2}\ln \left( \frac{ b_{K_{n}}^{n}-b_{s}^{n}-t_{s}^{n}}{b_{K_{n}}^{n}-b_{s}^{n}}\cdot \frac{ a_{K_{n}}^{n}-b_{s}^{n}}{a_{K_{n}}^{n}-b_{s}^{n}-t_{s}}\right) . \end{eqnarray} (3.13)

    Recall that

    \begin{equation} \frac{t_{n}}{\sigma _{K_{n}}^{n}-\rho _{K_{n}}^{n}}\leq \frac{\rho _{K_{n}}^{n}}{\sigma _{K_{n}}^{n}-\rho _{K_{n}}^{n}}, \end{equation} (3.14)
    \begin{equation} \frac{b_{K_{n}}^{n}-b_{s}^{n}-t_{s}^{n}}{b_{K_{n}}^{n}-b_{s}^{n}}\geq \frac{ b_{K_{n}}^{n}-\rho _{K_{n}}^{n}}{b_{K_{n}}^{n}-b_{s}^{n}}\geq 1-\frac{\rho _{K_{n}}^{n}}{b_{K_{n}}^{n}}, \end{equation} (3.15)

    and

    \begin{eqnarray} \lim \sup\limits_{n}\left( a_{K_{n}}^{n}-b_{s}^{n}-t_{s}^{n}\right) & = &\lim \sup\limits_{n}\left( a_{K_{n}}^{n}-a_{s+1}^{n}+a_{s+1}^{n}-b_{s}^{n}-t_{s}^{n}\right) \\ &\leq &A_{s}+\left\vert \left[ b_{s}^{n},a_{s+1}^{n}\right] \cap III_{\delta ,n}^{+}\right\vert \leq C. \end{eqnarray} (3.16)

    Taking liminf on both sides of (3.13), we have

    \begin{equation*} \lim \inf\limits_{n}\left[ J\left( u_{n}\right) -J\left( \widetilde{u}_{n}\right) \right] = \infty , \end{equation*}

    contradicting the assumption that \left\{ u_{n}\right\} is a minimizing sequence and the fact that J is bounded from below. Therefore we must have \lim \sup_{n}\left(a_{K_{n}}^{n}-b_{1}^{n}\right) = a_{0} < \infty .

    Next we show that \lim \sup_{n}\frac{a_{1}^{n}}{b_{K_{n}}^{n}} = 0. If \lim \sup_{n}\frac{a_{1}^{n}}{b_{K_{n}}^{n}} > 0 and \kappa _{n} : = \left\vert \left[ 1+c_{n}, \rho _{K_{n}}^{n}\right] \cap II_{\delta, n}^{+}\right\vert , then

    \begin{equation*} t_{n}-\kappa _{n}\leq \left\vert III_{\delta ,n}^{+}\right\vert \leq C, \end{equation*}

    and

    \begin{eqnarray*} \kappa _{n} & = &\rho _{K_{n}}^{n}-\left( 1+c_{n}\right) -\left\vert \left[ 1+c_{n},\rho _{K_{n}}^{n}\right] \cap I_{\delta ,n}^{+}\right\vert \\ &&-\left\vert \left[ 1+c_{n},\rho _{K_{n}}^{n}\right] \cap III_{\delta ,n}^{+}\right\vert \\ &\leq &a_{K_{n}}^{n}-\left( 1+c_{n}\right) -\left( b_{1}^{n}-a_{1}^{n}\right) . \end{eqnarray*}

    Therefore, we have

    \begin{eqnarray} &&{{\frac{4\pi }{\beta}}}\left[J\left( u_{n}\right) -J\left( \widetilde{u} _{n}\right)\right] \\ & = &\int_{\mathbb{R}}\int_{\mathbb{R}}\left[ \frac{\left( u_{n}\left( x\right) -u_{n}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}-\frac{ \left( \widetilde{u}_{n}\left( x\right) -\widetilde{u}_{n}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}\right] dydx \\ & = &-8\int_{\left[ 1+c_{n},\text{ }\rho _{K_{n}}^{n}\right] \cap A_{n}^{+}}\int_{\mathbb{R}\backslash \left( \left[ 1,\text{ }\rho _{K_{n}}^{n}\right] \cap A_{n}^{+}\right) }\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ & = &-8\int_{\left[ 1+c_{n},\text{ }\rho _{K_{n}}^{n}\right] \cap A_{n}^{+}}\int_{\sigma _{K_{n}}^{n}}^{\infty }\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx-8\int_{\left[ 1+c_{n}, \text{ }\rho _{K_{n}}^{n}\right] \cap A_{n}^{+}}\int_{-\infty }^{1+c_{n}} \frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ &&-8\int_{\left[ 1+c_{n},\text{ }\rho _{K_{n}}^{n}\right] \cap A_{n}^{+}}\int_{\left[ 1+c_{n},\text{ }\rho _{K_{n}}^{n}\right] \cap A_{n}^{-}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx-8\int_{\left[ 1+c_{n},\text{ }\rho _{K_{n}}^{n}\right] \cap A_{n}^{+}}\int_{\rho _{K_{n}}^{n}}^{\sigma _{K_{n}}^{n}}\frac{ u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ &\geq &-8\int_{\rho _{K_{n}}^{n}-t_{n}}^{\rho _{K_{n}}^{n}}\int_{\sigma _{K_{n}}^{n}}^{\infty }\frac{1}{\left( x-y\right) ^{2}}dydx-8\int_{\left[ 1+c_{n},\rho _{K_{n}}^{n}\right] \cap II_{\delta ,n}^{+}}\int_{\left[ 1+c_{n},\rho _{K_{n}}^{n}\right] \cap I_{\delta ,n}^{+}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ &&-8\int_{\left[ 1+c_{n},\rho _{K_{n}}^{n}\right] \cap II_{\delta ,n}^{+}}\int_{a_{K_{n}}^{n}}^{b_{K_{n}}^{n}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}} \\ &\geq &-8\ln \frac{\sigma _{K_{n}}^{n}-\rho _{K_{n}}^{n}+t_{n}}{\sigma _{K_{n}}^{n}-\rho _{K_{n}}^{n}}+8\left( 1-\delta \right) ^{2}\int_{1+c_{n}}^{1+c_{n}+\kappa _{n}}\int_{a_{K_{n}}^{n}-b_{1}^{n}+a_{1}^{n}}^{b_{K_{n}}^{n}}\frac{1}{\left( x-y\right) ^{2}} \\ & = &-8\ln \frac{\sigma _{K_{n}}^{n}-\rho _{K_{n}}^{n}+t_{n}}{\sigma _{K_{n}}^{n}-\rho _{K_{n}}^{n}} \\ &&+8\left( 1-\delta \right) ^{2}\ln \frac{b_{K_{n}}^{n}-1-c_{n}-\kappa _{n}}{ b_{K_{n}}^{n}-1-c_{n}}\cdot \frac{a_{K_{n}}^{n}-b_{1}^{n}+a_{1}^{n}-1-c_{n}}{ a_{K_{n}}^{n}-b_{1}^{n}+a_{1}^{n}-1-c_{n}-\kappa _{n}}. \end{eqnarray} (3.17)

    Since

    \begin{equation} \lim \sup\limits_{n}\frac{\sigma _{K_{n}}^{n}-\rho _{K_{n}}^{n}+t_{n}}{\sigma _{K_{n}}^{n}-\rho _{K_{n}}^{n}}\leq \lim \sup\limits_{n}\frac{\sigma _{K_{n}}^{n}}{ \sigma _{K_{n}}^{n}-\rho _{K_{n}}^{n}} < \infty , \end{equation} (3.18)
    \begin{equation} \lim \inf\limits_{n}\frac{b_{K_{n}}^{n}-1-c_{n}-\kappa _{n}}{b_{K_{n}}^{n}-1-c_{n}} \geq 1-\lim \sup\limits_{n}\frac{a_{1}^{n}-1-c_{n}+a_{K_{n}}^{n}-b_{1}^{n}}{ b_{K_{n}}^{n}-1} > 1-\nu > 0, \end{equation} (3.19)

    and

    \begin{equation} a_{K_{n}}^{n}-b_{1}^{n}+a_{1}^{n}-1-c_{n}-\kappa _{n}\leq a_{0}+\left\vert III_{\delta ,n}^{+}\right\vert \leq C, \end{equation} (3.20)

    taking liminf on both sides of (3.17), it follows from (3.18), (3.19) and (3.20) that

    \begin{equation*} \lim \inf\limits_{n}\left( J\left( u_{n}\right) -J\left( \widetilde{u}_{n}\right) \right) = \infty , \end{equation*}

    a contradiction.

    Lastly we estimate the energy difference between u_{n} and \widetilde{u} _{n}. We have

    \begin{eqnarray} &&{\frac{4\pi}{\beta}}\left[J\left( u_{n}\right) -J\left( \widetilde{u} _{n}\right) \right] \\ & = &\int_{\mathbb{R}}\int_{\mathbb{R}}\left[ \frac{\left( u_{n}\left( x\right) -u_{n}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}-\frac{ \left( \widetilde{u}_{n}\left( x\right) -\widetilde{u}_{n}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}\right] dydx \\ & = &-8\int_{\left[ 1+c_{n},\rho _{K_{n}}^{n}\right] \cap A_{n}^{+}}\int_{ \mathbb{R}\backslash \left( \left[ 1+c_{n},\rho _{K_{n}}^{n}\right] \cap A_{n}^{+}\right) }\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ & = &-8\int_{\left[ 1+c_{n},\rho _{K_{n}}^{n}\right] \cap A_{n}^{+}}\int_{\sigma _{K_{n}}^{n}}^{\infty }\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx-8\int_{\left[ 1+c_{n},\rho _{K_{n}}^{n}\right] \cap A_{n}^{+}}\int_{-\infty }^{1+c_{n}} \frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ &&-8\int_{\left[ 1+c_{n},\rho _{K_{n}}^{n}\right] \cap A_{n}^{+}}\int_{\left[ 1+c_{n},\rho _{K_{n}}^{n}\right] \cap A_{n}^{-}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx-8\int_{\left[ 1+c_{n},\rho _{K_{n}}^{n}\right] \cap A_{n}^{+}}\int_{\rho _{K_{n}}^{n}}^{\sigma _{K_{n}}^{n}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ &\geq &-8\int_{1+c_{n}}^{\rho _{1}^{n}}\int_{\sigma _{K_{n}}^{n}}^{\infty } \frac{1}{\left( x-y\right) ^{2}}dydx-8\int_{\sigma _{1}^{n}}^{\rho _{K_{n}}^{n}}\int_{\sigma _{K_{n}}^{n}}^{\infty }\frac{1}{\left( x-y\right) ^{2}}dydx \\ &\geq &-8\ln \frac{\sigma _{K_{n}}^{n}-1-c_{n}}{\sigma _{K_{n}}^{n}-\rho _{1}^{n}}-8\ln \frac{\sigma _{K_{n}}^{n}-\rho _{K_{n}}^{n}+\rho _{K_{n}}^{n}-\sigma _{1}^{n}}{\sigma _{K_{n}}^{n}-\rho _{K_{n}}^{n}}. \end{eqnarray} (3.21)

    Recalling that

    \begin{eqnarray*} \lim \sup \frac{\rho _{K_{n}}^{n}}{\sigma _{K_{n}}^{n}} &\leq &\lim \sup \frac{a_{K_{n}}^{n}}{b_{K_{n}}^{n}} = \nu < 1,\text{ }\lim \sup \frac{\rho _{1}^{n}}{\sigma _{K_{n}}^{n}}\leq \lim \sup \frac{a_{1}^{n}}{b_{K_{n}}^{n}} = 0,\text{ } \\ && \end{eqnarray*}

    and

    \begin{equation*} \lim \sup \left( \rho _{K_{n}}^{n}-\sigma _{1}^{n}\right) \leq \lim \sup \left( a_{K_{n}}^{n}-b_{1}^{n}\right) = a_{0} < \infty , \end{equation*}

    we conclude that

    \begin{eqnarray*} \lim \sup\limits_{n}\frac{\sigma _{K_{n}}^{n}-\rho _{K_{n}}^{n}+\rho _{K_{n}}^{n}-\sigma _{1}^{n}}{\sigma _{K_{n}}^{n}-\rho _{K_{n}}^{n}} & = &1, \\ \lim \inf\limits_{n}\frac{\sigma _{K_{n}}^{n}-1-c_{n}}{\sigma _{K_{n}}^{n}-\rho _{1}^{n}} & = &1. \end{eqnarray*}

    Thus by (3.21) we have that \left\{ \widetilde{u }_{n}\right\} is also a minimizing sequence. Defining

    \begin{equation*} \overline{u}_{n} = \widetilde{u}_{n}\left( x+s_{n}-1\right) , \end{equation*}

    where s_{n} = \left[ \sigma _{K_{n}}^{n}\right] is the largest integer smaller than \sigma _{K_{n}}^{n}, then \left\{ \overline{u}_{n}\right\} is a minimizing sequence satisfying

    \begin{equation*} \overline{u}_{n}\left( x\right) \leq 0 \ \text{for }x\leq 1+\sigma _{K_{n}}^{n}-s_{n}\text{ and }\overline{u}_{n}\left( x\right) \geq -1+\delta \ \text{for }x > 1+\sigma _{K_{n}}^{n}-s_{n}. \end{equation*}

    Proposition 3.2 applies to \overline{u}_{n} , from which we can extract a converging subsequence to a minimizer u_0 \in \mathcal{A}_0 .

    Case Ⅰ-ⅲ: There exists l > 1 such that \lim \sup_{n} \frac{a_{K_{n}-j}^{n}}{b_{K_{n}-j}^{n}} = 1 for j = 0, 1, \cdots, l-1 and \lim \sup_{n}\frac{a_{K_{n}-l}^{n}}{b_{K_{n}-l}^{n}} < 1. Then we must have \lim \sup_{n}\left(a_{K_{n}-l}^{n}-b_{1}^{n}\right) < \infty and \lim \sup_{n}\frac{a_{1}^{n}}{b_{K_{n}-l}^{n}} = 0. We construct \widetilde{u}_{n} as follows.

    \begin{equation} \widetilde{u}_{n} = \left\{ \begin{array}{c} -u_{n}\left( x\right) \ \ \ x\in \left[ 1+c_{n},\rho _{l}^{n}\right] \cap A_{n}^{+} \\ u_{n}\left( x\right) \ \ \ \ \ \ \ \text{elsewhere} \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \end{array} .\right. \end{equation} (3.22)

    We also need

    \begin{equation*} \int_{I_{\delta ,n}^{+}}\int_{I_{\delta ,n}^{-}}\frac{4}{\left( x-y\right) ^{2}}dydx\leq 4\ln \left( \prod\limits_{j = 0}^{l-1}\frac{b_{K_{n}-j}^{n}}{ a_{K_{n}-j}^{n}}\right) +4\ln \left( 2\frac{b_{K_{n}-l}^{n}}{a_{1}^{n}} \right) . \end{equation*}

    We can follow a similar argument in Case Ⅰ-ⅰ and Case Ⅰ-ⅱ by replacing any estimates on \left[ a_{K_{n}}^{n}, b_{K_{n}}^{n}\right] in (3.13), (3.21) and (3.17) by estimates on \left[ a_{K_{n}-l}^{n}, b_{K_{n}-l}^{n}\right] , using \overline{u}_{n} = \widetilde{u}_{n}\left(x+s_{n, l}-1\right), where s_{n, l} = \left[ \sigma _{l}^{n}\right] is the largest integer less than or equal to \sigma _{l}^{n}.

    Case Ⅱ: No such l exists, i.e., \lim \sup_{n} \frac{a_{j}^{n}}{b_{j}^{n}} = 1 for all j where \lim \sup_{n}b_{j}^{n} = \infty. Let l be such that \lim \sup_{n}b_{K_{n}-j}^{n} = \infty for all j < l and \lim \sup_{n}b_{K_{n}-l}^{n} < \infty. In this case, we will reflect the negative part of u_{n} to -u_{n} outside a big portion of \left(b_{K_{n}-l}^{n}, a_{K_{n}-l+1}^{n}\right) \cap II_{\delta, n}^{+}. Following notations in section 2, we write

    \begin{equation*} \left( b_{K_{n}-l}^{n},a_{K_{n}-l+1}^{n}\right) \cap II_{\delta ,n}^{+} = \cup _{j = j_{l+1}^{n}}^{j_{l}^{n}-1}\left[ c_{j}^{n},\text{ }d_{j}^{n}\right] . \end{equation*}

    First by

    \begin{eqnarray} C & > &J_{\delta }^{+}\left( u_{n}\right) \geq -\int_{I_{\delta ,n}^{+}}\int_{I_{\delta ,n}^{-}}\frac{\beta}{{\pi}\left( x-y\right) ^{2}} dydx+\frac{1}{4}\int_{1}^{\infty }g(x)W\left( u_{n}\right) dx \\ &\geq &-{\frac{\beta}{\pi}}\ln \left( 2\frac{b_{K_{n}-l}^{n}}{a_{1}^{n}} \right) -\sum\limits_{i = K_{n}-l+1}^{K_{n}}{\frac{\beta}{\pi}}\ln \frac{b_{i}^{n}}{ a_{i}^{n}}+\frac{\gamma \min_{\left\vert u\right\vert \leq 1-\delta }}{4} \left\vert III_{\delta ,n}^{+}\right\vert \end{eqnarray} (3.23)

    we conclude that

    \begin{equation*} \left\vert III_{\delta ,n}^{+}\right\vert \leq C. \end{equation*}

    Case Ⅱ-ⅰ: There exists j\left(l\right) \in \left\{ j_{l+1}^{n}, \cdots, j_{l}^{n}-1\right\} such that

    \begin{equation} \lim \sup\limits_{n}\left( d_{j(l)}^{n}-c_{j(l)}^{n}\right) = \infty . \end{equation} (3.24)

    Let T_{n, l} = \left[ d_{j\left(l\right) }^{n}, c_{L_{n}}^{n}\right] \cap A_{n}^{-}, M_{n, l} = \left\vert T_{n, l}\cap III_{\delta, n}^{+}\right\vert. We define

    \begin{equation*} \overline{u}_{n}\left( x\right) = \left\{ \begin{array}{c} -u_{n}\left( x\right) \ x\in \left[ d_{j\left( l\right) }^{n},c_{L_{n}}^{n}\right] \cap A_{n}^{-} \\ u_{n}\left( x\right) \ \ \ \ \ \ \text{otherwise} \end{array} \right. . \end{equation*}

    Since

    \begin{equation*} \frac{u_{n}\left( y_{1}\right) }{\left( x-y_{1}\right) ^{2}}\geq -\frac{ u_{n}\left( y_{2}\right) }{\left( x-y_{2}\right) ^{2}}\text{ } \end{equation*}

    for x\in T_{n, l}, y_{1}\in \left[ c_{j\left(l\right) }^{n}, d_{j\left(l\right) }^{n}\right] and y_{2}\in \left[ b_{K_{n}-l}^{n}, \text{ } c_{j\left(l\right) }^{n}\right] \cap A_{n}^{-}, together with (3.24) and the fact that

    \begin{equation*} \left\vert \left[ b_{K_{n}-l}^{n},\text{ }c_{j\left( l\right) }^{n}\right] \cap A_{n}^{-}\right\vert \leq \left\vert III_{\delta ,n}^{+}\right\vert \leq C, \end{equation*}

    we conclude that

    \begin{equation*} \int_{T_{n,l}}\int_{\left[ b_{K_{n}-l}^{n},\text{ }c_{j\left( l\right) }^{n} \right] \cap A_{n}^{-}}^{{}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx+\int_{T_{n,l}}\int_{c_{j\left( l\right) }^{n}}^{d_{j\left( l\right) }^{n}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx\leq 0. \end{equation*}

    It then follows that

    \begin{eqnarray*} &&{\frac{4\pi}{\beta}}\left[J\left( \overline{u}_{n}\right) -J\left( u_{n}\right)\right] \\ & = &8\int_{T_{n,l}}\int_{\mathbb{R}\backslash T_{n,l}}^{{}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ &\leq &8\int_{T_{n,l}}\int_{-\infty }^{c_{j\left( l\right) }^{n}}\frac{ u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}} dydx+8\int_{T_{n,l}}\int_{\left[ d_{j\left( l\right) }^{n},\text{ } c_{L_{n}}^{n}\right] \cap A_{n}^{+}}^{{}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ &&+8\int_{T_{n,l}}\int_{c_{j\left( l\right) }^{n}}^{d_{j\left( l\right) }^{n}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx+8\int_{T_{n,l}}\int_{c_{L_{n}}^{n}}^{\infty }\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ &\leq &8\int_{T_{n,l}}\int_{-\infty }^{b_{K_{n}-l}^{n}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx+8\int_{T_{n,l}} \int_{\left[ b_{K_{n}-l}^{n},\text{ }c_{j\left( l\right) }^{n}\right] \cap A_{n}^{-}}^{{}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ &&+8\int_{T_{n.,l}}\int_{c_{j\left( l\right) }^{n}}^{d_{j\left( l\right) }^{n}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ &\leq &\int_{-\infty }^{c_{j\left( l\right) }^{n}}\int_{d_{j\left( l\right) }^{n}}^{M_{n,l}+d_{j\left( l\right) }^{n}}\frac{8}{\left( x-y\right) ^{2}} dydx+\sum\limits_{i = 0}^{l-1}\int_{-\infty }^{b_{K_{n}-l}^{n}}\int_{\rho _{K_{n}-i}^{n}}^{\sigma _{K_{n}-i}^{n}}\frac{8}{\left( x-y\right) ^{2}}dydx \\ &&+8\int_{T_{n,l}}\int_{\left[ b_{K_{n}-l}^{n},\text{ }c_{j\left( l\right) }^{n}\right] \cap A_{n}^{-}}^{{}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx+8\int_{T_{n,l}}\int_{c_{j\left( l\right) }^{n}}^{d_{j\left( l\right) }^{n}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ &\leq &8\ln \frac{M_{n,l}^{{}}+d_{j\left( l\right) }^{n}-c_{j\left( l\right) }^{n}}{d_{j\left( l\right) }^{n}-c_{j\left( l\right) }^{n}} +\sum\limits_{i = 0}^{l-1}\ln \frac{\sigma _{K_{n}-i}^{n}-b_{K_{n}-l}^{n}}{\rho _{K_{n}-i}^{n}-b_{K_{n}-l}^{n}}\rightarrow 0, \end{eqnarray*}

    i.e., \left\{ \overline{u}_{n}\right\} is also a minimizing sequence with \overline{b}_{\overline{K}_{n}}^{n} = b_{K_{n}-l}^{n} satisfying \lim \sup \overline{b}_{\overline{K}_{n}}^{n} < \infty.

    Case Ⅱ-ⅱ:

    \begin{equation*} \lim \sup\limits_{n}\left[ \max\limits_{j_{l+1}^{n}\leq j\leq j_{l}^{n}-1}\left( d_{j}^{n}-c_{j}^{n}\right) \right] \bf{ < \infty .} \end{equation*}

    Let S_{0} = \left\{ \text{ }j_{l+1}^{n}+1, \text{ }\cdots, \text{ } j_{l}^{n}-1\right\}. We define S_{0}^{+} as follows.

    \begin{equation*} S_{0}^{+} = \left\{ j\in S_{0}:d_{j}^{n}-c_{j}^{n} > c_{j}^{n}-d_{j-1}^{n}\right\} \end{equation*}

    There exists k = k\left(n\right) and indices p\left(1\right), \cdots, p\left(k\right) \in S_{0} , q\left(1\right), \cdots, q\left(k\right) \in S_{0} such that i\in S_{0}^{+} if p\left(s\right) \leq i\leq q\left(s\right) and i\notin S_{0}^{+} if q\left(s\right) \leq i\leq p\left(s+1\right) for s = 1, \cdots, k. We write S_{1} = \left\{ 1, \cdots, k\right\}.

    Case Ⅱ-ⅱ-1: There exists s\leq k such that

    \begin{equation*} \lim \sup\limits_{n}\sum\limits_{i = p\left( s\right) }^{q\left( s\right) }\left( d_{i}^{n}-c_{i}^{n}\right) \rightarrow \infty . \end{equation*}

    Let

    \begin{eqnarray*} T_{n,l,s} & = &\left[ d_{q\left( s\right) }^{n},c_{L_{n}}^{n}\right] \cap A_{n}^{-} \\ M_{n,l,s}^{{}} & = &\left\vert III_{\delta ,n}^{+}\cap T_{n,l,s}\right\vert . \end{eqnarray*}

    We define

    \begin{equation*} \overline{u}_{n}\left( x\right) = \left\{ \begin{array}{c} -u_{n}\left( x\right) \ x\in \left[ d_{q\left( s\right) }^{n},c_{L_{n}}^{n}\right] \cap A_{n}^{-} \\ u_{n}\left( x\right) \ \ \ \ \ \ \text{otherwise} \end{array} \right. . \end{equation*}

    Since

    \begin{equation*} \left\vert \left( c_{j}^{n},d_{j}^{n}\right) \right\vert > \left\vert \left( d_{j-1}^{n},c_{j}^{n}\right) \right\vert \text{ for }p\left( s\right) \leq j\leq q\left( s\right) ,\text{ }\cup _{j = p(s)}^{q\left( s\right) }\left( d_{j-1}^{n},c_{j}^{n}\right) \subset III_{\delta ,n}^{+} \end{equation*}

    and

    \begin{equation*} \lim \sup\limits_{n}\sum\limits_{i = p\left( s\right) }^{q\left( s\right) }\left( d_{i}^{n}-c_{i}^{n}\right) \rightarrow \infty ,\text{ }\left\vert III_{\delta ,n}^{+}\right\vert \leq C, \end{equation*}

    together with the observation that

    \begin{equation*} \frac{u_{n}\left( y_{1}\right) }{\left( x-y_{1}\right) ^{2}} > - \frac{ u_{n}\left( y_{2}\right) }{\left( x-y_{2}\right) ^{2}}\text{ } \end{equation*}

    for x\in T_{n, l, s}, y_{1}\in \left(c_{j}^{n}, d_{j}^{n}\right), y_{2}\in \left(d_{j-1}^{n}, c_{j}^{n}\right), p\left(s\right) \leq j\leq q\left(s\right) , and

    \begin{equation*} \frac{u_{n}\left( y_{1}\right) }{\left( x-y_{1}\right) ^{2}} > - \frac{ u_{n}\left( y_{2}\right) }{\left( x-y_{2}\right) ^{2}}\text{ } \end{equation*}

    for x\in T_{n, l, s}, y_{1}\in \cup _{j = p\left(s\right) }^{q\left(s\right) }\left(c_{j}^{n}, d_{j}^{n}\right), y_{2}\in III_{\delta, n}^{+}\cap \left(b_{K_{n}-l}^{n}, c_{p\left(s\right) }^{n}\right), we conclude that

    \begin{eqnarray*} &&\int_{T_{n,l,s}}\int_{\left[ b_{K_{n}-l}^{n},\text{ }c_{p\left( s\right) }^{n}\right] \cap A_{n}^{-}}^{{}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx+\sum\limits_{j = p\left( s\right) }^{q\left( s\right) }\int_{T_{n,l,s}}\int_{c_{j}^{n}}^{d_{j}^{n}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ &&+\sum\limits_{j = p\left( s\right) }^{q\left( s\right) }\int_{T_{n,l,s}}\int_{d_{j-1}^{n}}^{c_{j}^{n}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ &\leq &0. \end{eqnarray*}

    Therefore

    \begin{eqnarray*} &&{\frac{4\pi}{\beta}}\left[J\left( \overline{u}_{n}\right) -J\left( u_{n}\right)\right] \\ & = &8\int_{T_{n,l,s}}\int_{\mathbb{R\diagdown }T_{n,l,s}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ & = &8\int_{T_{n,l,s}}\int_{-\infty }^{d_{q\left( s\right) }^{n}}\frac{ u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}} dydx+8\int_{T_{n,l,s}}\int_{\left[ d_{q\left( s\right) }^{n},\text{ } c_{L_{n}}^{n}\right] \cap A_{n}^{+}}^{{}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ &&+8\int_{T_{n,l,s}}\int_{c_{L_{n}}^{n}}^{\infty }\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ &\leq &8\int_{T_{n,l,s}}\int_{-\infty }^{c_{p\left( s\right) }^{n}}\frac{ u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}} dydx+8\int_{T_{n,l,s}}\int_{c_{p\left( s\right) }^{n}}^{d_{q\left( s\right) }^{n}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ &\leq &8\int_{T_{n,l,s}}\int_{-\infty }^{b_{K_{n}-l}^{n}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx+8 \int_{T_{n,l,s}}\int_{\left[ b_{K_{n}-l}^{n},\text{ }c_{p\left( s\right) }^{n}\right] \cap A_{n}^{-}}^{{}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ &&+8\int_{T_{n,l,s}}\int_{c_{p\left( s\right) }^{n}}^{d_{q\left( s\right) }^{n}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ &\leq &\int_{-\infty }^{c_{p\left( s\right) }^{n}}\int_{d_{q\left( s\right) }^{n}}^{M_{n,l,s}^{{}}+d_{q\left( s\right) }^{n}}\frac{8}{\left( x-y\right) ^{2}}dydx+\sum\limits_{i = 0}^{l-1}\int_{-\infty }^{b_{K_{n}-l}^{n}}\int_{\rho _{K_{n}-i}^{n}}^{\sigma _{K_{n}-i}^{n}}\frac{8}{\left( x-y\right) ^{2}}dydx \\ &&+8\int_{T_{n,l,s}}\int_{\left[ b_{K_{n}-l}^{n},\text{ }c_{p\left( s\right) }^{n}\right] \cap A_{n}^{-}}^{{}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx+8\sum\limits_{j = p\left( s\right) }^{q\left( s\right) }\int_{T_{n,l,s}}\int_{c_{j}^{n}}^{d_{j}^{n}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ &&+8\sum\limits_{j = p\left( s\right) }^{q\left( s\right) }\int_{T_{n,l,s}}\int_{d_{j-1}^{n}}^{c_{j}^{n}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ &\leq &\int_{-\infty }^{c_{p\left( s\right) }^{n}}\int_{d_{q\left( s\right) }^{n}}^{M_{n,l,s}^{{}}+d_{q\left( s\right) }^{n}}\frac{8}{\left( x-y\right) ^{2}}dydx+\sum\limits_{i = 0}^{l-1}\int_{-\infty }^{b_{K_{n}-l}^{n}}\int_{\rho _{K_{n}-i}^{n}}^{\sigma _{K_{n}-i}^{n}}\frac{8}{\left( x-y\right) ^{2}}dydx \\ &\leq &8\ln \frac{M_{n,l,s}^{{}}+d_{q\left( s\right) }^{n}-c_{p\left( s\right) }^{n}}{d_{q\left( s\right) }^{n}-c_{p\left( s\right) }^{n}} +8\sum\limits_{i = 0}^{l-1}\ln \frac{\sigma _{K_{n}-i}^{n}-b_{K_{n}-l}^{n}}{\rho _{K_{n}-i}^{n}-b_{K_{n}-l}^{n}} \\ &\rightarrow &0 \end{eqnarray*}

    Case Ⅱ-ⅱ-2:

    \begin{equation*} \sup\limits_{s}\left( \lim \sup\limits_{n}\sum\limits_{i = p(s)}^{q(s)}\left( d_{i}^{n}-c_{i}^{n}\right) \right) < \infty. \end{equation*}

    We consider

    \begin{equation*} S_{1}^{+} = \left\{ \alpha \in S_{1}:\sum\limits_{i = p\left( \alpha \right) }^{q\left( \alpha \right) }\left( d_{i}^{n}-c_{i}^{n}\right) > \sum\limits_{i = 1+q\left( \alpha -1\right) }^{q\left( \alpha \right) }\left( c_{i}^{n}-d_{i-1}^{n}\right) \right\} . \end{equation*}

    There exists m = m\left(n\right) \leq k and p_{1}\left(\tau \right), q_{1}\left(\tau \right) \in S_{1} for each \tau \leq m and \tau \in \mathbb{N} such that \gamma \in S_{1}^{+} if p_{1}\left(\tau \right) \leq \gamma \leq q_{1}\left(\tau \right) and \gamma \notin S_{1}^{+} if q_{1}\left(\tau \right) \leq \gamma \leq p_{1}\left(\tau +1\right). Let S_{2} = \left\{ 1, \cdots, m\right\}

    Case Ⅱ-ⅱ-2-a: There exists \tau = \tau \left(n\right) such that

    \begin{equation} \lim \sup\limits_{n}\sum\limits_{\gamma = p_{1}\left( \tau \right) }^{q_{1}\left( \tau \right) }\sum\limits_{i = p\left( \gamma \right) }^{q\left( \gamma \right) }\left( d_{i}^{n}-c_{i}^{n}\right) \rightarrow \infty. \end{equation} (3.25)

    Then we consider

    \begin{equation*} \overline{u}_{n}\left( x\right) = \left\{ \begin{array}{c} -u_{n}\left( x\right) \ x\in \left[ d_{q\left( q_{1}\left( \tau \right) \right) }^{n},c_{L_{n}}^{n}\right] \cap A^{-} \\ u_{n}\left( x\right) \ \ \ \ \ \ \text{otherwise} \end{array} \right. \end{equation*}

    Let T_{n, l, q_{1}\left(\tau \right) } = \left[ d_{q\left(q_{1}\left(\tau \right) \right) }^{n}, \sigma _{K_{n}}^{n}\right] \cap A^{-}. M_{n, l, q_{1}\left(\tau \right) }^{{}} = \left\vert \left[ d_{q\left(q_{1}\left(\tau \right) \right) }^{n}, \sigma _{K_{n}}^{n}\right] \cap A^{-}\cap III_{\delta, n}^{+}\right\vert. Observe

    \begin{equation*} \frac{u_{n}\left( y_{1}\right) }{\left( x-y_{1}\right) ^{2}}\geq -\frac{ u_{n}\left( y_{2}\right) }{\left( x-y_{2}\right) ^{2}} \end{equation*}

    for x\in T_{n, l, q_{1}\left(\tau \right) }, y_{1}\in \left(c_{j}^{n}, d_{j}^{n}\right), y_{2}\in \left(d_{j-1}^{n}, c_{j}^{n}\right) when p\left(p_{1}\left(\tau \right) \right) \leq j\leq q\left(q_{1}\left(\tau \right) \right). The same inequality also holds for x\in T_{n, l, q_{1}\left(\tau \right) }, y_{1}\in \cup _{j = p\left(p_{1}\left(\tau \right) \right) }^{q\left(q_{1}\left(\tau \right) \right) }\left(c_{j}^{n}, d_{j}^{n}\right), y_{2}\in \left[ b_{K_{n}-l}^{n}, \text{ } c_{p\left(p_{1}\left(\tau \right) \right) }^{n}\right] \cap A_{n}^{-}. Moreover, by (3.25),

    \begin{equation*} \left\vert \left[ b_{K_{n}-l}^{n},\text{ }c_{p\left( p_{1}\left( \tau \right) \right) }^{n}\right] \cap A_{n}^{-}\right\vert \leq \left\vert III_{\delta ,n}^{+}\right\vert \leq C, \end{equation*}

    and

    \begin{equation*} \cup _{j = p\left( p_{1}\left( \tau \right) \right) }^{q\left( q_{1}\left( \tau \right) \right) }\left( d_{j-1}^{n},c_{j}^{n}\right) \subset III_{\delta ,n}^{+}, \end{equation*}

    we conclude that

    \begin{eqnarray*} &&\int_{T_{n,l,q_{1}\left( \tau \right) }}\int_{\left[ b_{K_{n}-l}^{n},\text{ }c_{p\left( p_{1}\left( \tau \right) \right) }^{n}\right] \cap A_{n}^{-}}^{{}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx+\sum\limits_{j = p\left( p_{1}\left( \tau \right) \right) }^{q\left( q_{1}\left( \tau \right) \right) }\int_{T_{n,l,q_{1}\left( \tau \right) }}\int_{c_{j}^{n}}^{d_{j}^{n}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ &&+\sum\limits_{j = p\left( p_{1}\left( \tau \right) \right) }^{q\left( q_{1}\left( \tau \right) \right) }\int_{T_{n,l,q_{1}\left( \tau \right) }}\int_{d_{j-1}^{n}}^{c_{j}^{n}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx\leq 0. \end{eqnarray*}

    Therefore

    \begin{eqnarray*} &&{\frac{4\pi}{\beta}}\left[J\left( \overline{u}_{n}\right) -J\left( u_{n}\right)\right] \\ & = &8\int_{T_{n,l,q_{1}\left( \tau \right) }}\int_{\mathbb{R\diagdown } T_{n,l,q_{1}\left( \tau \right) }}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ & = &8\int_{A_{2}^{-}}\int_{-\infty }^{d_{q\left( q_{1}\left( \tau \right) \right) }^{n}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx+8\int_{T_{n,l,q_{1}\left( \tau \right) }}\int_{\left[ d_{q\left( q_{1}\left( \tau \right) \right) }^{n},\text{ }c_{L_{n}}^{n} \right] \cap A_{n}^{+}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{ \left( x-y\right) ^{2}}dydx \\ &&+8\int_{T_{n,l,q_{1}\left( \tau \right) }}\int_{c_{L_{n}}^{n}}^{\infty } \frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ &\leq &8\int_{T_{n,l,q_{1}\left( \tau \right) }}\int_{-\infty }^{b_{K_{n}-l}^{n}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx+8\int_{T_{n,l,q_{1}\left( \tau \right) }}\int_{b_{K_{n}-l}^{n}}^{d_{q\left( q_{1}\left( \tau \right) \right) }^{n}} \frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ &\leq &\int_{-\infty }^{c_{p\left( p_{1}\left( \tau \right) \right) }^{n}}\int_{d_{q\left( q_{1}\left( \tau \right) \right) }^{n}}^{M_{n,l,q_{1}\left( \tau \right) }^{{}}+d_{q\left( q_{1}\left( \tau \right) \right) }^{n}}\frac{8}{\left( x-y\right) ^{2}}dydx+\sum\limits_{i = 0}^{l-1} \int_{-\infty }^{b_{K_{n}-l}^{n}}\int_{\rho _{K_{n}-i}^{n}}^{\sigma _{K_{n}-i}^{n}}\frac{8}{\left( x-y\right) ^{2}}dydx \\ &&+8\int_{T_{n,l,q_{1}\left( \tau \right) }}\int_{\left[ b_{K_{n}-l}^{n}, \text{ }c_{p\left( p_{1}\left( \tau \right) \right) }^{n}\right] \cap A_{n}^{-}}^{{}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx+8\sum\limits_{j = p\left( p_{1}\left( \tau \right) \right) }^{q\left( q_{1}\left( \tau \right) \right) }\int_{T_{n,l,q_{1}\left( \tau \right) }}\int_{c_{j}^{n}}^{d_{j}^{n}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ &&+8\sum\limits_{j = p\left( p_{1}\left( \tau \right) \right) }^{q\left( q_{1}\left( \tau \right) \right) }\int_{T_{n,l,q_{1}\left( \tau \right) }}\int_{d_{j-1}^{n}}^{c_{j}^{n}}\frac{u_{n}\left( x\right) u_{n}\left( y\right) }{\left( x-y\right) ^{2}}dydx \\ &\leq &8\ln \frac{M_{n,l,q_{1}\left( \tau \right) }^{{}}+d_{q\left( q_{1}\left( \tau \right) \right) }^{n}-c_{p\left( p_{1}\left( \tau \right) \right) }^{n}}{d_{q\left( q_{1}\left( \tau \right) \right) }^{n}-c_{p\left( p_{1}\left( \tau \right) \right) }^{n}}+8\sum\limits_{i = 0}^{l-1}\ln \frac{\sigma _{K_{n}-i}^{n}-b_{K_{n}-l}^{n}}{\rho _{K_{n}-i}^{n}-b_{K_{n}-l}^{n}} \\ &\rightarrow &0. \end{eqnarray*}

    Continuing this way if necessary, we can define the set S_{i} inductively by each m\in S_{i}, p_{i}\left(m\right), q_{i}\left(m\right) \in S_{i-1} such that any p_{i}\left(m\right) \leq p\leq q_{i}\left(m\right), p\in S_{i-1}^{+}, if q_{i}\left(m\right) \leq p\leq p_{i}\left(m+1\right), p\in S_{i-1}\backslash S_{i-1}^{+}. Here

    \begin{eqnarray*} &&S_{i}^{+} : = \\ &&\left\{ m\in S_{i}:\sum\limits_{l_{i} = p_{i}\left( m\right) }^{q_{i}\left( m\right) }\sum\limits_{l_{i-1} = p_{i-1}\left( l_{i}\right) }^{q_{i-1}\left( l_{i}\right) }\cdots \sum\limits_{l_{1} = p\left( l_{2}\right) }^{q\left( l_{2}\right) }\left( d_{l_{1}}^{n}-c_{l_{1}}^{n}\right) > \sum\limits_{l_{i} = p_{i}\left( m\right) }^{q_{i}\left( m\right) }\sum\limits_{l_{i-1} = p_{i-1}\left( l_{i}\right) }^{q_{i-1}\left( l_{i}\right) }\cdots \sum\limits_{l_{1} = p\left( l_{2}\right) }^{q\left( l_{2}\right) }\left( c_{l_{1}}^{n}-d_{l_{1}-1}^{n}\right) \right\} . \end{eqnarray*}

    By the definition of S_{i}, we have \left\vert S_{i}\right\vert \leq \left\vert S_{i-1}\right\vert \leq \cdots \leq \left\vert S_{0}\right\vert. Since \left\vert III_{\delta, n}^{+}\right\vert is uniformly bounded and \lim \sup \left(a_{K_{n}-l+1}^{n}-b_{K_{n}-l}^{n}\right) = \infty, we would be able to find r_{n} such that S_{r_{n}} = S_{r_{n}}^{+}. Let \mu = \left\vert S_{r_{n}}\right\vert and define

    \begin{equation*} \overline{u}_{n}\left( x\right) = \left\{ \begin{array}{c} -u_{n}\left( x\right) \ x\in \left[ d_{q\left( q_{1}\left( \cdots q_{r_{n}}(\mu \right) \right) }^{n},c_{L_{n}}^{n}\right] \cap A_{n}^{-} \\ u_{n}\left( x\right) \ \ \ \ \ \ \text{otherwise} \end{array} \right. . \end{equation*}

    By a similar argument, we can show that \{\overline{u}_{n}\} is a minimizing sequence which is close to \pm 1 away from a uniformly bounded interval.

    Proof of the first half of Theorem 1.1.Given a minimizing sequence \left\{ u_{n}\right\}, if \lim \sup_{n}b_{K_{n}}^{n} < \infty we obtain a minimizer by Proposition 3.2. If \lim \sup b_{K_{n}}^{n} = \infty, we obtain a new minimizing sequence \left\{ \overline{u}_{n}\right\} which satisfies \lim \sup \overline{b}_{\overline{K}_{n}}^{n} < \infty by Proposition 3.3, existence then follows from Proposition 3.2.

    Proof of the second half of Theorem 1.1.

    Proposition 4.1. Any minimizer u_{0} of J over \mathcal{A} is a C^{2, \frac{1}{2} }\left(\mathbb{R}\right) solution of

    \begin{equation*} -\alpha u_{0}^{\prime \prime }+g\left( x\right) W^{\prime }\left( u_{0}\right) +\beta \left( -\frac{d^{2}}{dx^{2}}\right) ^{\frac{1}{2} }u_{0} = 0, \end{equation*}

    where we understand the fractional operator in the sense of (1.5).

    Proof. Let v_{0} = u_{0}-\eta . We write J\left(u_{0}\right) in terms of v_{0} as

    \begin{eqnarray*} J\left( v_{0}+ \eta \right) & = &\frac{\alpha }{2}\int_{\mathbb{R}}\left\vert v_{0}^{\prime }+\eta ^{\prime }\right\vert ^{2}dx+\int_{\mathbb{R}} g\left( x\right) W\left( v_{0}+\eta \right) dx+\frac{\beta }{4\pi }\int_{\mathbb{R} }\int_{\mathbb{R}} \frac{\left( v_{0}\left( x\right) -v_{0}\left( y\right) \right) ^{2}}{\left( x-y\right) ^{2}}dydx \\ &&+\frac{\beta }{2\pi }\int_{\mathbb{R}}\int_{\mathbb{R}}\frac{\left( v_{0}\left( x\right) -v_{0}\left( y\right) \right) \left( \eta \left( x\right) -\eta \left( y\right) \right) }{\left( x-y\right) ^{2}}dydx. \end{eqnarray*}

    Consider now variations v_{\varepsilon } = v_{0}+\varepsilon \varphi , where \varphi is any smooth compactly supported function. Since u_{0} is a minimizer, we must have

    \begin{eqnarray} 0 & = &\left. \frac{d}{d\varepsilon }J\left( v_{\varepsilon } + \eta \right) \right\vert _{\varepsilon = 0} = \int_{\mathbb{R}}\left( \alpha u_{0}^{\prime }\varphi ^{\prime }+g\left( x\right) W^{\prime }\left( v_{0}+\eta \right) \varphi \right) dx \\ && + \frac{\beta }{2\pi }\int_{\mathbb{R}}\int_{\mathbb{R}}\frac{\left( v_{0}\left( x\right) -v_{0}\left( y\right) \right) \left( \varphi \left( x\right) -\varphi \left( y\right) \right) }{\left( x-y\right) ^{2}}dydx \\ &&+\frac{\beta }{2\pi }\int_{\mathbb{R}}\int_{\mathbb{R}}\frac{\left( \eta \left( x\right) -\eta \left( y\right) \right) \left( \varphi \left( x\right) -\varphi \left( y\right) \right) }{\left( x-y\right) ^{2}}dydx. \end{eqnarray} (4.1)

    Since v_{0}\in H^{1}\left(\mathbb{R}\right), we can define \left(- \frac{d^{2}}{dx^{2}}\right) ^{\frac{1}{2}}v_{0} via Fourier transform as (see e.g., [32] Proposition 3.3)

    \begin{equation*} \widehat{\left( -\frac{d^{2}}{dx^{2}}\right) ^{\frac{1}{2}}v_{0}}\left( \xi \right) = \left\vert \xi \right\vert \widehat{v_{0}}\left( \xi \right), \end{equation*}

    and write the second term in (4.1) (see [32] Remark 3.7) as

    \begin{equation*} {\frac{\beta }{2 \pi}} \int_{\mathbb{R}}\int_{\mathbb{R}}\frac{\left( v_{0}\left( x\right) -v_{0}\left( y\right) \right) \left( \varphi \left( x\right) -\varphi \left( y\right) \right) }{\left( x-y\right) ^{2}} dydx = \beta \int_{\mathbb{R}} \varphi \left( x\right) \left( -\frac{d^{2}}{ dx^{2}}\right) ^{\frac{1}{2} }v_{0}\left( x\right) . \end{equation*}

    Since \eta \in C^{\infty }\left(\mathbb{R}\right), for x > 1 take \varepsilon \ll 1 such that

    \begin{equation} \int_{\left\vert x-y\right\vert \geq \varepsilon }\frac{\eta \left( x\right) -\eta \left( y\right) }{\left( x-y\right) ^{2}}dy = \int_{-\infty }^{1}\frac{ \eta \left( x\right) -\eta \left( y\right) }{\left( x-y\right) ^{2}}dy\leq \frac{2}{x-1}, \end{equation} (4.2)

    and for x < -1 take \varepsilon \ll 1 such that

    \begin{equation} \int_{\left\vert x-y\right\vert \geq \varepsilon }\frac{\eta \left( x\right) -\eta \left( y\right) }{\left( x-y\right) ^{2}} dy = \int_{-1}^{\infty }\frac{ \eta \left( x\right) -\eta \left( y\right) }{\left( x-y\right) ^{2}}dy\leq \frac{2}{x+1}. \end{equation} (4.3)

    For -1\leq x\leq 1, we can write

    \begin{eqnarray*} \lim\limits_{\varepsilon \rightarrow 0}\int_{\left\vert x-y\right\vert \geq \varepsilon } \frac{\eta \left( x\right) -\eta \left( y\right) }{\left( x-y\right) ^{2}}dy & = &\frac{1}{2}\lim\limits_{\varepsilon \rightarrow 0}\int_{\left\vert y\right\vert \geq \varepsilon }\frac{\eta \left( x+y\right) +\eta \left( x-y\right) -2\eta \left( x\right) }{y^{2}}dy \\ & = &\frac{1}{2}\int_{\mathbb{R}}\frac{\eta \left( x+y\right) +\eta \left( x-y\right) -2\eta \left( x\right) }{y^{2}}dy, \end{eqnarray*}

    where the last step follows from the fact that \eta \in C^{\infty }\left(\mathbb{\ R}\right).

    For each x\in \mathbb{R} we have

    \begin{eqnarray} \left\vert \int_{\mathbb{R}}\frac{\eta \left( x+y\right) +\eta \left( x-y\right) -2\eta \left( x\right) }{y^{2}}dy\right\vert &\leq &\int_{1}^{\infty }\frac{4}{y^{2}}+\int_{-\infty }^{-1}\frac{4}{y^{2}} +\left\vert \int_{-1}^{1}\frac{\eta \left( x+y\right) +\eta \left( x-y\right) -2\eta \left( x\right) }{y^{2}}dy\right\vert \\ &\leq &8+2\left\Vert D^{2}\eta \right\Vert _{L^{\infty }}. \end{eqnarray} (4.4)

    Combining (4.2), (4.3) and (4.4), we conclude that the function

    \begin{equation} \lim\limits_{\varepsilon \rightarrow 0}\int_{\left\vert x-y\right\vert \geq \varepsilon } \frac{ \eta \left( x\right) -\eta \left( y\right) }{\left( x-y\right) ^{2}}dy \end{equation} (4.5)

    belongs to L^2(\mathbb{R}) . Thus the third term in (4.1) can be written as

    \begin{eqnarray*} \int_{\mathbb{R}}\int_{\mathbb{R}}\frac{\left( \eta \left( x\right) -\eta \left( y\right) \right) \left( \varphi \left( x\right) -\varphi \left( y\right) \right) }{\left( x-y\right) ^{2}}dydx & = &\lim\limits_{\varepsilon \rightarrow 0}\int_{\mathbb{R}}\int_{\left\vert x-y\right\vert \geq \varepsilon }\frac{\left( \eta \left( x\right) -\eta \left( y\right) \right) \left( \varphi \left( x\right) -\varphi \left( y\right) \right) }{\left( x-y\right) ^{2}}dydx \\ & = &2\int_{\mathbb{R}}\varphi \left( x\right) \lim\limits_{\varepsilon \rightarrow 0}\int_{\left\vert x-y\right\vert \geq \varepsilon }\frac{ \eta \left( x\right) -\eta \left( y\right) }{\left( x-y\right) ^{2}}dydx. \end{eqnarray*}

    We now introduce the notation

    \begin{equation*} \left( -\frac{d^{2}}{dx^{2}}\right) ^{\frac{1}{2}}u_{0} : = \left( -\frac{ d^{2}}{ dx^{2}}\right) ^{\frac{1}{2}}v_{0}+\lim\limits_{\varepsilon \rightarrow 0} \frac{1}{\pi }\int_{\left\vert x-y\right\vert \geq \varepsilon }\frac{ \eta \left( x\right) -\eta \left( y\right) }{\left( x-y\right) ^{2}}dy, \end{equation*}

    where the fractional operator in the right-hand side is understood via Fourier transform. Since \varphi is arbritrary, we conclude from (4.1) that u_{0} satisfies the following equation in the distributional sense:

    \begin{equation} -\alpha u_{0}^{\prime \prime }+g\left( x\right) W^{\prime }\left( u_{0}\right) +\beta \left( -\frac{d^{2}}{dx^{2}}\right) ^{\frac{1}{2} }u_{0} = 0. \end{equation} (4.6)

    Since \left\vert u_{0}\right\vert \leq 1 and v_{0} = u_{0}-\eta \in H^{1}\left(\mathbb{R}\right), we have W^{\prime }\left(u_{0}\right) \in L^{2}\left(\mathbb{R}\right) and \left(-\frac{d^{2}}{dx^{2}}\right) ^{ \frac{1}{2}}v_{0} \in L^{2}\left(\mathbb{R}\right). Thus (4.5) implies that \left(-\frac{d^{2}}{dx^{2}}\right) ^{\frac{1 }{2}}u_{0}\in L^{2}\left(\mathbb{R}\right). By elliptic estimates, we then conclude that u_{0}\in W^{2, 2}\left(\mathbb{R}\right).

    Weakly differentiating (4.6) yields

    \begin{equation*} -\alpha u_{0x}^{\prime \prime }+g^{\prime}\left( x\right) W^{\prime }(u_{0})+g\left( x\right) W^{^{\prime \prime }}\left( u_{0}\right) u_{0x}+\beta \left( -\frac{d^{2}}{dx^{2}}\right) ^{\frac{1}{2}}u_{0x} = 0 \end{equation*}

    in the sense of distributions. Here we used the facts that

    \begin{equation} \frac{d}{dx}\left( -\frac{d^{2}}{dx^{2}}\right) ^{\frac{1}{2}}v_{0} = \left( - \frac{d^{2}}{dx^{2}}\right) ^{\frac{1}{2}}v_{0x} \end{equation} (4.7)

    and

    \begin{equation} \frac{d}{dx}\lim\limits_{\varepsilon \rightarrow 0}\int_{\left\vert x-y\right\vert \geq \varepsilon }\frac{ \eta \left( x\right) -\eta \left( y\right) }{\left( x-y\right) ^{2}}dy = \lim\limits_{\varepsilon \rightarrow 0} \int_{\left\vert x-y\right\vert \geq \varepsilon }\frac{\eta ^{\prime }\left( x\right) -\eta ^{\prime }\left( y\right) }{\left( x-y\right) ^{2}}dy, \end{equation} (4.8)

    which follow from the properties of Fourier transform of Sobolev functions and the following calculation:

    \begin{eqnarray*} &&\frac{d}{dx}\lim\limits_{\varepsilon \rightarrow 0}\int_{\left\vert x-y\right\vert \geq \varepsilon }\frac{ \eta \left( x\right) -\eta \left( y\right) }{\left( x-y\right) ^{2}}dy \\ & = &\lim\limits_{h\rightarrow 0}\frac{1}{h}\left[ \lim\limits_{\varepsilon \rightarrow 0}\int_{\left\vert x+h-y\right\vert \geq \varepsilon }\frac{\eta \left( x+h\right) -\eta \left( y\right) }{\left( x+h-y\right) ^{2}} dy-\lim\limits_{\varepsilon \rightarrow 0}\int_{\left\vert x-y\right\vert \geq \varepsilon }\frac{ \eta \left( x\right) -\eta \left( y\right) }{ \left( x-y\right) ^{2}}dy\right] \\ & = &\lim\limits_{h\rightarrow 0}\frac{1}{h}\left[ \lim\limits_{\varepsilon \rightarrow 0}\int_{\left\vert x-z\right\vert \geq \varepsilon }\frac{\eta \left( x+h\right) -\eta \left( z+h\right) }{\left( x-z\right) ^{2}} dz-\lim\limits_{\varepsilon \rightarrow 0}\int_{\left\vert x-y\right\vert \geq \varepsilon }\frac{ \eta \left( x\right) -\eta \left( y\right) }{ \left( x-y\right) ^{2}}dy\right] \\ & = &\lim\limits_{h\rightarrow 0}\frac{1}{h}\lim\limits_{\varepsilon \rightarrow 0}\int_{\left\vert x-y\right\vert \geq \varepsilon }\frac{\eta \left( x+h\right) -\eta \left( x\right) -\eta \left( y+h\right) +\eta \left( y\right) }{\left( x-y\right) ^{2}}dy \\ & = &\lim\limits_{\varepsilon \rightarrow 0}\int_{\left\vert x-y\right\vert \geq \varepsilon }\lim\limits_{h\rightarrow 0}\frac{1}{h}\left[ \frac{\eta \left( x+h\right) -\eta \left( x\right) -\eta \left( y+h\right) +\eta \left( y\right) }{\left( x-y\right) ^{2}} \right] dy \\ & = &\lim\limits_{\varepsilon \rightarrow 0}\frac{1}{\pi }\int_{\left\vert x-y\right\vert \geq \varepsilon }\frac{\eta ^{\prime }\left( x\right) -\eta ^{\prime }\left( y\right) }{\left( x-y\right) ^{2}}dy. \end{eqnarray*}

    The same arguments as in the case of (4.5) can be used to prove that the function

    \begin{equation*} \lim\limits_{\varepsilon \rightarrow 0}\frac{1}{\pi }\int_{\left\vert x-y\right\vert \geq \varepsilon }\frac{\left( \eta ^{\prime }\left( x\right) -\eta ^{\prime }\left( y\right) \right) }{\left( x-y\right) ^{2}}dy \end{equation*}

    belongs to L^2(\mathbb{R}) as well. Define

    \begin{equation*} \left( -\frac{d^{2}}{dx^{2}}\right) ^{\frac{1}{2}}u_{0x} = \left( -\frac{d^{2} }{dx^{2}}\right) ^{\frac{1}{2}}v_{0x}+\frac{d}{dx}\lim\limits_{\varepsilon \rightarrow 0}\frac{1}{\pi }\int_{\left\vert x-y\right\vert \geq \varepsilon }\frac{ \eta^{\prime }\left( x\right) -\eta^{\prime }\left( y\right) }{ \left( x-y\right) ^{2}}dy. \end{equation*}

    Since W\in C^{2, 1}\left(\mathbb{R}\right), we have g^{\prime } W^{\prime }(u_{0})+gW^{^{\prime \prime }}\left(u_{0}\right) u_{0x}\in L^{2}\left(\mathbb{R}\right), \left(-\frac{d^{2}}{dx^{2}}\right) ^{ \frac{1}{2}}v_{0x}\in L^{2}\left(\mathbb{R}\right), we have \left(- \frac{d^{2}}{dx^{2}}\right) ^{\frac{1}{2}}u_{0x}\in L^{2}\left(\mathbb{R} \right). Thus elliptic estimates imply u_{0x}\in W^{2, 2}\left(\mathbb{R} \right), i.e., u_{0}\in W^{3, 2}\left(\mathbb{R}\right) \subset C^{2, \frac{1}{2}}\left(\mathbb{R}\right). Thus u_{0} is a classical solution of (4.6). Moreover, since u_{0}\in C^{2, \frac{1 }{2} }\left(\mathbb{R}\right), we can write

    \begin{equation*} \left( -\frac{d^{2}}{dx^{2}}\right) ^{\frac{1}{2}}u_{0}\left( x\right) = \lim\limits_{\varepsilon \rightarrow 0}\frac{1}{\pi }\int_{\left\vert x-y\right\vert \geq \varepsilon }\frac{ u_{0}\left( x\right) -u_{0}\left( y\right) }{\left( x-y\right) ^{2}}dy. \end{equation*}

    The second half of Theorem 1.1 follows.

    The authors declare they have not used Artificial Intelligence (AI) tools in the creation of this article.

    We thank the referee for careful reading of our manuscript and many helpful suggestions for improving this paper. In particular, the referee pointed out Remark 2.2 when g(x) is nondegenerate. The work of CM was supported, in part, by NSF via grants DMS-1614948 and DMS-1908709. XY's research was sponsored by a Research Excellence Grant from University of Connecticut, CLAS Dean's summer research grant and Simons Collaboration Grant #947054.

    The authors declare no conflict of interest.



    [1] S. B. Angenent, J. Mallet-Paret, L. A. Peletie, Stable transition layers in a semilinear boundary value problem, J. Differ. Equations, 67 (1987), 212–242. https://doi.org/10.1016/0022-0396(87)90147-1 doi: 10.1016/0022-0396(87)90147-1
    [2] G. Alberti, L. Ambrosio, X. Cabré, On a long-standing conjecture of E. De Giorgi: symmetry in 3D for general nonlinearities and a local minimality property, Acta Appl. Math., 65 (2001), 9–33. https://doi.org/10.1023/A:1010602715526 doi: 10.1023/A:1010602715526
    [3] F. Alessio, C. Gui, P. Montecchiari, Saddle solutions to Allen-Cahn equations in doubly periodic media, Indiana U. Math. J., 65 (2016), 199–221. https://doi.org/10.1512/iumj.2016.65.5772 doi: 10.1512/iumj.2016.65.5772
    [4] F. Alessio, L. Jeanjean, P. Montecchiari, Stationary layered solutions in \mathbb{R}^{2} for a class of non autonomous Allen-Cahn equations, Calc. Var., 11 (2000), 177–202. https://doi.org/10.1007/s005260000036 doi: 10.1007/s005260000036
    [5] F. Alessio, L. Jeanjean, P. Montecchiari, Existence of infinitely many stationary layered solutions in \mathbb{R}^{2} for a class of periodic Allen-Cahn equations, Commun. Part. Diff. Eq., 27 (2002), 1537–1574. https://doi.org/10.1081/PDE-120005848 doi: 10.1081/PDE-120005848
    [6] F. Alessio, P. Montecchiari, Entire solutions in \mathbb{R }^{2} for a class of Allen-Cahn equations, ESIAM: COCV, 11 (2005), 633–672. https://doi.org/10.1051/cocv:2005023 doi: 10.1051/cocv:2005023
    [7] F. Alessio, P. Montecchiari, Brake orbits type solutions to some class of semiliear elliptic equations, Calc. Var., 30 (2007), 51–83. https://doi.org/10.1007/s00526-006-0078-1 doi: 10.1007/s00526-006-0078-1
    [8] F. Alessio, P. Montecchiari, Layered solutions with multiple asymptotes for non autonomous Allen-Cahn equations in \mathbb{R} ^{3}, Calc. Var., 46 (2013), 591–622. https://doi.org/10.1007/s00526-012-0495-2 doi: 10.1007/s00526-012-0495-2
    [9] C. O. Alves, V. Ambrosio, C. E. Torres Ledesma, Existence of heteroclinic solutions for a class of problems involving the fractional Laplacian, Anal. Appl., 17 (2019), 425–451. https://doi.org/10.1142/S0219530518500252 doi: 10.1142/S0219530518500252
    [10] L. Ambrosio, X. Cabré, Entire solutions of semilinear elliptic equations in three dimensions and a conjecture of De Giorgi, J. Amer. Math. Soc., 13 (2000), 725–739. https://doi.org/10.1090/S0894-0347-00-00345-3 doi: 10.1090/S0894-0347-00-00345-3
    [11] V. Bangert, On minimal laminations on the torus, Ann. Inst. H. Poincaré Anal. Non Linéaire, 6 (1989), 95–138. https://doi.org/10.1016/S0294-1449(16)30328-6 doi: 10.1016/S0294-1449(16)30328-6
    [12] M. Barlow, R. Bass, C. Gui, The Liouville property and a conjecture of De Giorgi, Commun. Pure Appl. Math., 53 (2000), 1007–1038. https://doi.org/10.1002/1097-0312(200008)53:8<1007::AID-CPA3>3.3.CO;2-L doi: 10.1002/1097-0312(200008)53:8<1007::AID-CPA3>3.3.CO;2-L
    [13] H. Berestycki, F. Hamel, R. Monneau, One-dimensional symmetry for some bounded entire solution of some elliptic equations, Duke Math. J., 103 (2000), 375–396. https://doi.org/10.1215/S0012-7094-00-10331-6 doi: 10.1215/S0012-7094-00-10331-6
    [14] U. Bessi, Many solutions of elliptic problems on \mathbb{R} ^{n} of irrational slope, Commun. Part. Diff. Eq., 30 (2005), 1773–1804. https://doi.org/10.1080/03605300500299992 doi: 10.1080/03605300500299992
    [15] U. Bessi, Slope-changing solutions of elliptic problems on \mathbb{R}^{n}, Nonlinear Anal., 68 (2008), 3923–3947. https://doi.org/10.1016/j.na.2007.04.031 doi: 10.1016/j.na.2007.04.031
    [16] S. Biagi, S. Dipierro, E. Valdinoci, E. Vecchi, Mixed local and nonlocal elliptic operators, Commun. Part. Diff. Eq., 47 (2022), 585–629. https://doi.org/10.1080/03605302.2021.1998908 doi: 10.1080/03605302.2021.1998908
    [17] C. Bucur, E. Valdinoci, Nonlocal diffusion and applications, Cham: Springer, 2016. https://doi.org/10.1007/978-3-319-28739-3
    [18] J. Byeon, P. H. Rabinowitz, On a phase transition model, Calc. Var., 47 (2013), 1–23. https://doi.org/10.1007/s00526-012-0507-2 doi: 10.1007/s00526-012-0507-2
    [19] J. Byeon, P. H. Rabinowitz, Asymptotic behavior of minima and mountain pass solutions for a class of Allen-Cahn models, Commun. Inf. Syst., 13 (2013), 79–95. https://doi.org/10.4310/CIS.2013.v13.n1.a3 doi: 10.4310/CIS.2013.v13.n1.a3
    [20] J. Byeon, P. H. Rabinowitz, Solutions of higer topological type for an Allen-Cahn model equation, J. Fixed Point Theory Appl., 15 (2014), 379–404. https://doi.org/10.1007/s11784-014-0190-3 doi: 10.1007/s11784-014-0190-3
    [21] J. Byeon, P. H. Rabinowitz, Unbounded solutions for a periodic phase transition model, J. Differ. Equations, 260 (2016), 1126–1153. https://doi.org/10.1016/j.jde.2015.09.024 doi: 10.1016/j.jde.2015.09.024
    [22] X. Cabré, E. Cinti, Energy estimates and 1-D symmetry for nonlinear equations involving the half-Laplacian, Discrete Contin. Dyn. Syst., 28 (2010), 1179–1206. https://doi.org/10.3934/dcds.2010.28.1179 doi: 10.3934/dcds.2010.28.1179
    [23] X. Cabré, E. Cinti, Sharp energy estimates for nonlinear fractional diffusion equations, Calc. Var., 49 (2014), 233–269. https://doi.org/10.1007/s00526-012-0580-6 doi: 10.1007/s00526-012-0580-6
    [24] X. Cabré, J. Serra, An extension problem for sums of fractional Laplacians and 1-D symmetry of phase transitions, Nonlinear Anal., 137 (2016), 246–265. https://doi.org/10.1016/j.na.2015.12.014 doi: 10.1016/j.na.2015.12.014
    [25] X. Cabré, Y. Sire, Nonlinear equations for fractional Laplacians Ⅰ: regularity, maximum principles, and Hamiltonian estimates, Ann. Inst. H. Poincaré Anal. Non Linéaire, 31 (2014), 23–53. https://doi.org/10.1016/j.anihpc.2013.02.001 doi: 10.1016/j.anihpc.2013.02.001
    [26] X. Cabré, Y. Sire, Nonlinear equations for fractional Laplacians Ⅱ; existence, uniqueness, and qualitative properties of solutions, Trans. Amer. Math. Soc., 367 (2015), 911–941. https://doi.org/10.1090/S0002-9947-2014-05906-0 doi: 10.1090/S0002-9947-2014-05906-0
    [27] X. Cabré, J. Solà-Morales, Layer solutions in a half-space for boundary reactions, Commun. Pure Appl. Math., 58 (2005), 1678–1732. https://doi.org/10.1002/cpa.20093 doi: 10.1002/cpa.20093
    [28] L. Caffarelli, L. Silvestre, An extension problem related to the fractional Laplacian, Commun. Part. Diff. Eq., 32 (2007), 1245–1260. https://doi.org/10.1080/03605300600987306 doi: 10.1080/03605300600987306
    [29] K.-S. Chen, C. Muratov, X. Yan, Layer solutions of a one-dimensional nonlocal model of Ginzburg-Landau type, Math. Model. Nat. Phenom., 12 (2017), 68–90. https://doi.org/10.1051/mmnp/2017068 doi: 10.1051/mmnp/2017068
    [30] M. Cozzi, S. Dipierro, E. Valdinoci, Nonlocal phase transitions in homogeneous and periodic media, J. Fixed Point Theory Appl., 19 (2017), 387–405. https://doi.org/10.1007/s11784-016-0359-z doi: 10.1007/s11784-016-0359-z
    [31] M. del Pino, M. Kowalczyk, J. Wei, On De Giorgi's conjecture in dimension N\geq 9, Ann. Math., 174 (2011), 1485–1569. https://doi.org/10.4007/annals.2011.174.3.3 doi: 10.4007/annals.2011.174.3.3
    [32] E. Di Nezza, G. Palatucci, E. Valdinoci, Hitchhiker's guide to the fractional Sobolev spaces, Bull. Sci. Math., 136 (2012), 521–573. https://doi.org/10.1016/j.bulsci.2011.12.004 doi: 10.1016/j.bulsci.2011.12.004
    [33] W. Ding, F. Hamel, X. Zhao, Bistable pulsating fronts for reaction-diffusion equations in a periodic habitat, Indiana Univ. Math. J., 66 (2017), 1189–1265. https://doi.org/10.1512/iumj.2017.66.6070 doi: 10.1512/iumj.2017.66.6070
    [34] S. Dipierro, A. Farina, E. Valdinoci, A three-dimensional symmetry result for a phase transition equation in the genuinely nonlocal regime, Calc. Var., 57 (2018), 15. https://doi.org/10.1007/s00526-017-1295-5 doi: 10.1007/s00526-017-1295-5
    [35] S. Dipierro, S. Patrizi, E. Valdinoci, Chaotic orbits for systems of nonlocal equations, Commun. Math. Phys., 349 (2017), 583–626. https://doi.org/10.1007/s00220-016-2713-9 doi: 10.1007/s00220-016-2713-9
    [36] S. Dipierro, S. Patrizi, E. Valdinoci, Heteroclinic connections for nonlocal equations, Math. Mod. Meth. Appl. Sci., 29 (2019), 2585–2636. https://doi.org/10.1142/S0218202519500556 doi: 10.1142/S0218202519500556
    [37] S. Dipierro, J. Serra, E. Valdinoci, Improvement of flatness for nonlocal phase transitions, Amer. J. Math., 142 (2020), 1083–1160. https://doi.org/10.1353/ajm.2020.0032 doi: 10.1353/ajm.2020.0032
    [38] N. Dirr, N. K. Yip, Pinning and de-pinning phenomena in front propagation in heterogeneous media, Interfaces Free Bound., 8 (2006), 79–109. https://doi.org/10.4171/IFB/136 doi: 10.4171/IFB/136
    [39] Z. Du, C. Gui, Y. Sire, J. Wei, Layered solutions for a fractional inhomogeneous Allen-Cahn equation, Nonlinear Differ. Equ. Appl., 23 (2016), 29. https://doi.org/10.1007/s00030-016-0384-z doi: 10.1007/s00030-016-0384-z
    [40] A. Farina, Symmetry for solutions of semilinear elliptic equations in \mathbb{R}^{N} and related conjecturs, Atti Accad. Naz. Lincei Cl. Sci. Fis. Mat. Natur. Rend. Lincei Matem. Appl., 10 (1999), 255–265.
    [41] A. Figalli, J. Serra, On stable solution for boundary reactions: a De Giorgi type result in dimension 4+1, Invent. Math., 219 (2020), 153–177. https://doi.org/10.1007/s00222-019-00904-2 doi: 10.1007/s00222-019-00904-2
    [42] N. Ghoussoub, C. Gui, On a conjecture of De Giorgi and some related problems, Math. Ann., 311 (1998), 481–491. https://doi.org/10.1007/s002080050196 doi: 10.1007/s002080050196
    [43] N. Ghoussoub, C. Gui, On De Giorgi's conjecture in dimensions 4 and 5, Ann. Math., 157 (2003), 313–334. https://doi.org/10.4007/annals.2003.157.313 doi: 10.4007/annals.2003.157.313
    [44] J. K. Hale, K. Sakamoto, Existence and stability of transition layers, Japan J. Appl. Math., 5 (1988), 367–405. https://doi.org/10.1007/BF03167908 doi: 10.1007/BF03167908
    [45] Y. Hu, Layer solutions for a class of semilinear elliptic equations involving fractional Laplacians, Bound. Value Probl., 2014 (2014), 41. https://doi.org/10.1186/1687-2770-2014-41 doi: 10.1186/1687-2770-2014-41
    [46] J. P. Keener, Propagation of waves in an excitable medium with discrete release sites, SIAM. J. Appl. Math., 61 (2000), 317–334. https://doi.org/10.1137/S0036139999350810 doi: 10.1137/S0036139999350810
    [47] J. Lu, V. Moroz, C. B. Muratov, Orbital-free density functional theory of out-of-plane charge screening in graphene, J. Nonlinear Sci., 25 (2015), 1391–1430. https://doi.org/10.1007/s00332-015-9259-4 doi: 10.1007/s00332-015-9259-4
    [48] J. Moser, Minimal solutions of variational problems on a torus, Ann. Inst. H. Poincaré C Anal. Non Linéaire, 3 (1986), 229–272. https://doi.org/10.1016/S0294-1449(16)30387-0 doi: 10.1016/S0294-1449(16)30387-0
    [49] C. B. Muratov, X. Yan, Uniqueness of one-dimensional Néel wall profiles, Proc. R. Soc. A, 472 (2016), 20150762. https://doi.org/10.1098/rspa.2015.0762 doi: 10.1098/rspa.2015.0762
    [50] K. Nakashima, Stable transition layers in a balanced bistable equation, Differential Integral Equations, 13 (2000), 1025–1038. https://doi.org/10.57262/die/1356061208 doi: 10.57262/die/1356061208
    [51] K. Nakashima, Multi-layered solutions for a spatially inhomogeneous Allen-Cahn equations, J. Differ. Equations, 191 (2003), 234–276. https://doi.org/10.1016/S0022-0396(02)00181-X doi: 10.1016/S0022-0396(02)00181-X
    [52] K. Nakashima, K. Tanaka, Clustering layers and boundary layers in spatially inhomogeneous phase transition problems, Ann. Inst. H. Poincaré Anal. Nonlinéaire, 20 (2003), 107–143. https://doi.org/10.1016/S0294-1449(02)00008-2 doi: 10.1016/S0294-1449(02)00008-2
    [53] G. Palatucci, O. Savin, E. Valdinoci, Local and global minimizers for a variational energy involving a fractional norm, Annali di Matematica, 192 (2013), 673–718. https://doi.org/10.1007/s10231-011-0243-9 doi: 10.1007/s10231-011-0243-9
    [54] P. H. Rabinowitz, Single and multitransition solutions for a family of semilinear elliptic PDE's, Milan J. Math., 79 (2011), 113–127. https://doi.org/10.1007/s00032-011-0139-6 doi: 10.1007/s00032-011-0139-6
    [55] P. H. Rabinowitz, E. Stredulinsky, Mixed states for an Allen-Cahn type equation, Commun. Pure Appl. Math., 56 (2003), 1078–1134. https://doi.org/10.1002/cpa.10087 doi: 10.1002/cpa.10087
    [56] P. H. Rabinowitz, E. Stredulinsky, Mixed states for an Allen-Cahn type equation Ⅱ, Calc. Var., 21 (2004), 157–207. https://doi.org/10.1007/s00526-003-0251-8 doi: 10.1007/s00526-003-0251-8
    [57] P. H. Rabinowitz, E. Stredulinsky, On some results of Moser and Bangert, Ann. Inst. H. Poincaré Anal. Nonlinéaire, 21 (2004), 673–688. https://doi.org/10.1016/j.anihpc.2003.10.002 doi: 10.1016/j.anihpc.2003.10.002
    [58] P. H. Rabinowitz, E. Stredulinsky, On some results of Moser and Bangert Ⅱ, Adv. Nonlinear Stud., 4 (2004), 377–396. https://doi.org/10.1515/ans-2004-0402 doi: 10.1515/ans-2004-0402
    [59] P. H. Rabinowitz, E. Stredulinsky, Extensions of Moser-Bangert theory: locally minimal solutions, Boston, MA: Birkhäuser, 2011. https://doi.org/10.1007/978-0-8176-8117-3
    [60] O. Savin, Regularity of flat level sets in phase transitions, Ann. Math., 169 (2009), 41–78. https://doi.org/10.4007/annals.2009.169.41 doi: 10.4007/annals.2009.169.41
    [61] O. Savin, Some remarks on the classification of global solutions with asymptotically flat level sets, Calc. Var., 56 (2017), 141. https://doi.org/10.1007/s00526-017-1228-3 doi: 10.1007/s00526-017-1228-3
    [62] O. Savin, Rigidity of minimizers in nonlocal phase transitions, Anal. PDE, 11 (2018), 1881–1900. https://doi.org/10.2140/apde.2018.11.1881 doi: 10.2140/apde.2018.11.1881
    [63] Y. Sire, E. Valdinoci, Fractional Laplacian phase transitions and boundary reactions: a geometric inequality and a symmetry result, J. Funct. Anal., 256 (2009), 1842–1864. https://doi.org/10.1016/j.jfa.2009.01.020 doi: 10.1016/j.jfa.2009.01.020
    [64] P. R. Stinga, Fractional powers of second order partial differential operators: extension problem and regularity theory, Ph. D. Thesis, Universidad Autónoma de Madrid.
    [65] J. X. Xin, Existence and nonexistence of travelling waves and reaction-diffusion front propagation in periodic media, J. Stat. Phys., 73 (1993), 893–926. https://doi.org/10.1007/BF01052815 doi: 10.1007/BF01052815
    [66] Z. Zhou, Existence of infinitely many solutions for a class of Allen-Cahn Equations in \mathbb{R}^{2}, Osaka J. Math., 48 (2011), 51–67.
  • Reader Comments
  • © 2023 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Metrics

Article views(1507) PDF downloads(132) Cited by(0)

Other Articles By Authors

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return

Catalog