In this paper, we present analytical results for the so-called forward lateral photovoltage scanning (LPS) problem. The (inverse) LPS model predicts doping variations in crystal by measuring the current leaving the crystal generated by a laser at various positions. The forward model consists of a set of nonlinear elliptic equations coupled with a measuring device modeled by a resistance. Standard methods to ensure the existence and uniqueness of the forward model cannot be used in a straightforward manner due to the presence of an additional generation term modeling the effect of the laser on the crystal. Hence, we scale the original forward LPS problem and employ a perturbation approach to derive the leading order system and the correction up to the second order in an appropriate small parameter. While these simplifications pose no issues from a physical standpoint, they enable us to demonstrate the analytic existence and uniqueness of solutions for the simplified system using standard arguments from elliptic theory adapted to the coupling with the measuring device.

The Wigner quasi-density function allows a phase-space formulation of statistical quantum mechanics that is of fundamental importance in theoretical investigation and in applications. This work contributes to these tasks with the formulation and analysis of an optimal control problem for the Wigner equation, which describes the time evolution of the quasi-density function. For this purpose, two possible control mechanisms are considered, and, correspondingly, a detailed analysis in weighted Sobolev spaces for the controlled nonhomogeneous Wigner equation is presented. Further theoretical results are reported concerning existence of optimal controls and differentiability of the control-to-state map and of the ensemble cost functional, which allows the derivation of the optimality system that characterizes the optimal controls sought.

In this work we propose a weighted hybridizable discontinuous Galerkin method (W-HDG) for drift-diffusion problems. By using specific exponential weights when computing the $L^2$ product in each cell of the discretization, we are able to mimic the behavior of the Slotboom variables, and eliminate the drift term from the local matrix contributions, while still solving the problem for the primal variables. We show that the proposed numerical scheme is well-posed, and validate numerically that it has the same properties as classical HDG methods, including optimal convergence, and superconvergence of postprocessed solutions. For polynomial degree zero, dimension one, and vanishing HDG stabilization parameter, W-HDG coincides with the Scharfetter-Gummel finite volume scheme (i.e., it produces the same system matrix). The use of local exponential weights generalizes the Scharfetter-Gummel scheme (the state-of-the-art for finite volume discretization of transport dominated problems) to arbitrary high order approximations.

This paper concerns with a compact network model combined with distributed models for semiconductor devices. For linear RLC networks containing distributed semiconductor devices, we construct a mathematical model that joins the differential‐algebraic initial value problem for the electric circuit with multidimensional parabolic‐elliptic boundary value problems for the devices. We prove an existence and uniqueness result and the asymptotic behavior of this mixed initial boundary value problem of partial differential‐algebraic equations.

The lateral photovoltage scanning method (LPS) can be used to detect undesired impurities which appear in silicon crystals during growth. Our goal is to make a digital twin of the LPS method. To this end, we replace inflexible blackbox code with a physics preserving finite volume discretization, confirming three theoretical results via a new simulation strategy. By making the simulation transparent, it becomes easier to trace intermediate results and gain theoretical insights.

AbstractWe discuss recent progress in the mathematical modeling of semiconductor devices. The central result of this paper is a combined quantum-classical model that self-consistently couples van Roosbroeck’s drift-diffusion system for classical charge transport with a Lindblad-type quantum master equation. The coupling is shown to obey fundamental principles of non-equilibrium thermodynamics. The appealing thermodynamic properties are shown to arise from the underlying mathematical structure of a damped Hamitlonian system, which is an isothermal version of socalled GENERIC systems. The evolution is governed by a Hamiltonian part and a gradient part involving a Poisson operator and an Onsager operator as geoemtric structures, respectively. Both parts are driven by the conjugate forces given in terms of the derivatives of a suitable free energy.

When solving elliptic partial differential equations in a region containing immersed interfaces (possibly evolving in time), it is often desirable to approximate the problem using an independent background discretisation, not aligned with the interface itself. Optimal convergence rates are possible if the discretisation scheme is enriched by allowing the discrete solution to have jumps aligned with the surface, at the cost of a higher complexity in the implementation. A much simpler way to reformulate immersed interface problems consists in replacing the interface by a singular force field that produces the desired interface conditions, as done in immersed boundary methods. These methods are known to have inferior convergence properties, depending on the global regularity of the solution across the interface, when compared to enriched methods. In this work we prove that this detrimental effect on the convergence properties of the approximate solution is only a local phenomenon, restricted to a small neighbourhood of the interface. In particular we show that optimal approximations can be constructed in a natural and inexpensive way, simply by reformulating the problem in a distributionally consistent way, and by resorting to weighted norms when computing the global error of the approximation.

Modern electronic devices would be unthinkable without semiconductor materials. Few inventions have shaped our modern society like they have. There are emerging fields such as organic semiconductors, where there is still the demand to develop models and new theory, and there are many well-established fields, e.g. particle detector simulations, where it is mostly required to develop advanced numerical methods.

We present a mathematical and numerical framework for the optimal design of doping profiles for optoelectronic devices using methods from mathematical optimization. With the goal to maximize light emission and reduce the threshold of an edge-emitting laser, we consider a drift-diffusion model for charge transport and include modal gain and total current into a cost functional, which we optimize in cross sections of the emitter. We present 1D and 2D results for exemplary setups that point out possible routes for device improvement.

We derive an optoelectronic model based on a gradient formulation for the relaxation of electron-, hole- and photon-densities to their equilibrium state. This leads to a coupled system of partial and ordinary differential equations, for which we discuss the isothermal and the non-isothermal scenario separately.

We present the results of an error analysis of a B-spline based finite-element approximation of the stream-function formulation of the large scale wind-driven ocean circulation. In particular, we derive optimal error estimates for h-refinement using a Nitsche-type variational formulations of the two simplied linear models of the stationary quasigeostrophic equations, namely the Stommel and Stommel–Munk models. Numerical results obtained from simulations performed on rectangular and embedded geometries confirm the error analysis.

We derive an optoelectronic model based on a gradient formulation for the relaxation of electron-, hole- and photon-densities to their equilibrium state. This leads to a coupled system of partial and ordinary differential equations, for which we discuss the isothermal and the non-isothermal scenario separately.

Driven by applications like organic semiconductors there is an increased interest in numerical simulations based on drift-diffusion models with arbitrary statistical distribution functions. This requires numerical schemes that preserve qualitative properties of the solutions, such as positivity of densities, dissipativity and consistency with thermodynamic equilibrium. An extension of the Scharfetter–Gummel scheme guaranteeing consistency with thermodynamic equilibrium is studied. It is derived by replacing the thermal voltage with an averaged diffusion enhancement for which we provide a new explicit formula. This approach avoids solving the costly local nonlinear equations defining the current for generalized Scharfetter–Gummel schemes.

We present the error analysis of a B-spline based finite-element approximation of the stream-function formulation of the large scale winddriven ocean circulation. In particular, we derive optimal error estimates for h-refinement using a Nitsche-type variational formulations of the two simplied linear models of the stationary quasigeostrophic equations, namely the Stommel and Stommel–Munk models. Numerical results on rectangular and embedded geometries confirm the error analysis.

A methodology for the geometrical and physical optimization of a photovoltaic cell is proposed, which makes use of a detailed distributed model for the device simulation and a genetic algorithm. For the numerical simulation of the device, a TCAD simulator is used, appropriately interfaced with the genetic algorithm. Since the parameters to be optimized are geometrical, each simulation requires a different mesh grid, which is automatically set within the genetic algorithm optimization cycle. The evaluation of the fitness function requires the post-processing of the output of the device simulation, which is performed by another external software, also interfaced with the genetic algorithm. The feasibility of this methodology is assessed on a homogeneous emitter solar cell, with some relevant free parameters, related to the number of fingers in a cell and to the doping profile of the emitter. The parameters which maximize the efficiency of the cell are determined by using the proposed procedure.

We consider an elliptic partial differential-algebraic model which arises in the modeling of an electric network that contains semiconductor devices. In this context, the electric network is described by linear differential-algebraic equations, while the semiconductor devices are described by nonlinear elliptic partial differential equations. The coupling takes place through the source term for the network equations and the boundary conditions for the device equations. Under the assumption that the fully coupled model has tractability index 2, we prove an existence result for this system. The extension of the notion of tractability index to a specific class of coupled systems is a further result of this paper.

We consider a MEP hydrodynamical model obtained from a set of transport equations for the distribution functions of electrons in conduction band and phonon. We assume that the MEP model contains equations for the electron density fluxes and energy fluxes, and for the phonons energy fluxes. For this system we introduce a small parameter, related to the transition probabilities in the collision terms, and a diffusive scaling at the level of the Lagrangian multipliers appearing in the closure relations. In the diffusive limit, as the small parameter tends to zero, we obtain a model that can be physically interpreted in the framework of linear irreversible thermodynamics.

We consider a class of nonlinear partial-differential-algebraic equations, where the nonlinearity is present only in the PDEs and in the coupling conditions, and some additional structural conditions hold. For this special class of PDAEs, we introduce and characterize simple algebraic conditions which lead to a notion of extended tractability index, and exemplify its application to coupled systems arising from microelectronics.

We consider an electric network modeled via Modified Nodal Analysis, which is refined by a nonlinear multi-dimensional elliptic partial differential equations as semiconductor device model. This coupled system is a partial differential-algebraic equation. Starting from an analysis of the pure network equations (without semiconductor) and its index-2 conditions, we propose additional conditions which assure that coupled system is index-2. Then we proof an existence result for the coupled system.