The existence and uniqueness of the solution of the boundary-value problem for the one-dimensional drift-diffusion model of a semiconductor device (Q1803482)

A singularly perturbed boundary value problem for a system of ordinary differential equations modelling the one-dimensional current flow in a semiconductor device was stated in [\textit{M. P. Belyanin}, Zh. Vychisl. Mat. Mat. Fiz. 28, No. 1, 34-51 (1988)]. In the article under review the existence, uniqueness and boundedness of the solution of such a problem are proved for the special case of so-called shortened model of semiconductor device with a single \(p-n\) junction. The initial equations have the form (1) \(p'=Ep-{{J_ p} \over {\mu_ p}}\), \(n'=-En+{{J_ n} \over {\mu_ n}}\), \(\varepsilon E'=p-n+N\), \(J_ p'=-R(p,n)\), \(J_ n'=R(p,n)\). Here prime denotes the derivative with respect to the spatial variable \(t\in[0,1]\). The unknown functions \(p\) and \(n\) are the hole and electron concentrations, respectively, \(E\) is the electric field, \(J_ p\) and \(J_ n\) are the current densities of the holes and electrons. The given functions \(\mu_ p\), \(\mu_ n\), \(N\), \(R\) are differentiable with respect to \(t\), may be with the single exception of a point of \(p-n\) junction \(\theta\in(0,1)\) where they can have a finite discontinuity; \(\mu_ p\) and \(\mu_ n\) are the hole and electron mobilities respectively, \(N\) is the concentration of ionized impurities, \(R\) is the recombination rate and \(\varepsilon\) is a small positive parameter. Introducing the new variables \[ x = {{p+n} \over \varepsilon}, \quad y={{p-n} \over \varepsilon}, \quad z=E, \quad u={{J_ n-J_ p} \over \varepsilon}, \] \[ J = {{J_ n+J_ p} \over \varepsilon}, \quad H= {N \over \varepsilon}, \quad \mu_ 1= {{\mu_ n+\mu_ p} \over {2\mu_ n \mu_ p}}, \quad \mu_ 2= {{\mu_ n-\mu_ p} \over {2\mu_ n \mu_ p}} \] the author reduces the system (1) to the system of four ordinary differential equations in the unit interval \(I=[0,1]\) \[ \begin{aligned} x' &= yz+u\mu_ 1-J\mu_ 2, \\ y' &= xz+u\mu_ 2-J\mu_ 1,\\ z' &= y+H\\ u' &= {{(x+y)(x-y)+\gamma^ 2} \over {\tau_ p(x+y+\gamma)+\tau_ n (x- y+\gamma)}}, \end{aligned} \tag{2} \] where \(J<0\) and \(\gamma\), \(\tau_ p\), \(\tau_ n\) are some positive constants. The boundary values for the system (2) are (3) \(x(0)=(H^ 2(0)+ \gamma^ 2)^{1/2}\), \(y(0)=-H(0)\), \(x(1)=(H^ 2(1)+\gamma^ 2)^{1/2}\), \(y(1)=-H(1)\), the \(H\) function is assumed to be continuous in the interval \(I\), with the exception of the point \(\theta\) where it has a finite discontinuity; also it is assumed that \(\sup| H(t)|<\infty\), \(\sup| H'(t)|<\infty\) and \(H(t)>0\) for \(t\in[0,\theta)\), \(H(t)<0\) for \(t\in(\theta,1]\). So, namely the boundary value problem (2)--(3) is the subject of the author's consideration; the main result is the proof of the existence, uniqueness and boundedness of its solution.