Let Ω \Omega Ω be the domain and let u u u and v v v denote the bulk and boundary phase variables, with T u T u T u the trace of u u u on ∂ Ω \partial\Omega ∂ Ω . Set
V : = B V ( Ω ; { a , b } ) × B V ( ∂ Ω ; { α , β } ) . \mathcal{V}:=BV(\Omega;\{a,b\})\times BV(\partial\Omega;\{\alpha,\beta\}). V := B V ( Ω ; { a , b }) × B V ( ∂ Ω ; { α , β }) .
For m m m , σ \sigma σ , L L L , V V V , and ζ \zeta ζ as in the source, define
c : = inf { ζ ∫ − ∞ ∞ ∫ − ∞ ∞ ∣ f ′ ( x ) − f ′ ( y ) ∣ 2 ∣ x − y ∣ 2 d x d y + ∫ − ∞ ∞ V ( f ( x ) ) d x : f ∈ H l o c 3 / 2 ( R ) , lim x → ∞ f ( − x ) = α , lim x → ∞ f ( x ) = β } , c:=\inf\left\{\zeta\int_{-\infty}^{\infty}\int_{-\infty}^{\infty}\frac{|f'(x)-f'(y)|^2}{|x-y|^2}\,dx\,dy+\int_{-\infty}^{\infty}V(f(x))\,dx:\ f\in H_{\mathrm{loc}}^{3/2}(\mathbb{R}),\ \lim_{x\to\infty}f(-x)=\alpha,\ \lim_{x\to\infty}f(x)=\beta\right\}, c := inf { ζ ∫ − ∞ ∞ ∫ − ∞ ∞ ∣ x − y ∣ 2 ∣ f ′ ( x ) − f ′ ( y ) ∣ 2 d x d y + ∫ − ∞ ∞ V ( f ( x )) d x : f ∈ H loc 3/2 ( R ) , x → ∞ lim f ( − x ) = α , x → ∞ lim f ( x ) = β } ,
where
ζ : = inf { ∬ R × R + ∣ D 2 u ( x , y ) ∣ 2 d x d y ∫ R ∫ R ∣ g ′ ( x ) − g ′ ( y ) ∣ 2 ∣ x − y ∣ 2 d x d y : u ∈ H 2 ( R × R + ) , T u ( ⋅ , 0 ) = g in R } . \zeta:=\inf\left\{\frac{\iint_{\mathbb{R}\times\mathbb{R}^{+}}|D^2u(x,y)|^2\,dx\,dy}{\int_{\mathbb{R}}\int_{\mathbb{R}}\frac{|g'(x)-g'(y)|^2}{|x-y|^2}\,dx\,dy}:u\in H^2(\mathbb{R}\times\mathbb{R}^{+}),\ Tu(\cdot,0)=g\ \text{in }\mathbb{R}\right\}. ζ := inf ⎩ ⎨ ⎧ ∫ R ∫ R ∣ x − y ∣ 2 ∣ g ′ ( x ) − g ′ ( y ) ∣ 2 d x d y ∬ R × R + ∣ D 2 u ( x , y ) ∣ 2 d x d y : u ∈ H 2 ( R × R + ) , T u ( ⋅ , 0 ) = g in R ⎭ ⎬ ⎫ .
Higher-order line-tension conjecture. Under the hypotheses of Theorem, the sequence { F ε } ε > 0 \{\mathcal{F}_{\varepsilon}\}_{\varepsilon>0} { F ε } ε > 0 Γ \Gamma Γ -converges as ε → 0 + \varepsilon\to0^+ ε → 0 + to
F 0 ( u , v ) : = { m Per Ω ( E a ) + ∑ z = a , b ∑ ξ = α , β σ ( z , ξ ) H N − 1 ( { T u = z } ∩ { v = ξ } ) + c L Per ∂ Ω ( F α ) , ( u , v ) ∈ V , ∞ , otherwise. \mathcal{F}_0(u,v):=\begin{cases}
\displaystyle m\operatorname{Per}_{\Omega}(E_a)+\sum_{z=a,b}\sum_{\xi=\alpha,\beta}\sigma(z,\xi)\mathcal{H}^{N-1}(\{Tu=z\}\cap\{v=\xi\})+cL\operatorname{Per}_{\partial\Omega}(F_\alpha),& (u,v)\in\mathcal{V},\\
\infty,&\text{otherwise.}
\end{cases} F 0 ( u , v ) := ⎩ ⎨ ⎧ m Per Ω ( E a ) + z = a , b ∑ ξ = α , β ∑ σ ( z , ξ ) H N − 1 ({ T u = z } ∩ { v = ξ }) + c L Per ∂ Ω ( F α ) , ∞ , ( u , v ) ∈ V , otherwise.
Moreover, ζ \zeta ζ is independent of g ∈ H l o c 3 / 2 ( R ) g\in H_{\mathrm{loc}}^{3/2}(\mathbb{R}) g ∈ H loc 3/2 ( R ) satisfying lim x → ∞ g ( − x ) = α \lim_{x\to\infty}g(-x)=\alpha lim x → ∞ g ( − x ) = α and lim x → ∞ g ( x ) = β \lim_{x\to\infty}g(x)=\beta lim x → ∞ g ( x ) = β . The conjecture identifies the missing boundary-transition energy as a line-tension term proportional to the perimeter of the boundary phase interface; the surrounding discussion explains that the bulk and boundary contributions are known, while this nonlocal higher-order contribution is expected but not established.