Multiscale Device Simulator

• Introduction »
• Device Section »
• Model section »
• Solver section »
• Physics section »
• Simulation section »
• Output »

Introduction

In this example we will see a simple thermal simulation of a 1D Si pn diode. The Seebeck and Peltier effects are negleted.

The thermal boundary conditions are represented by two heat reservoirs placed at the two ends of the simulation domain that, at least in this case, corresponds to the drift diffusion domain.

In order to execute correctly the example you should have the following files in the working directory:
tut_02.tib : input file for TiberCAD
tut_02.msh: mesh file produced by GMSH from the script tut_02.geo

Device section (up)

Model section (up)

The drift diffusion simulation is performed over the whole device. Therefore, in option subsection we have to indicate physical_regions = all

The non-radiative recombination model used is the Shockley-Reed-Hall model. The related physical model section looks as

Direct bias applied to the diode is obtained by using the following boundary conditions

As in the drift diffusion case, the thermal simulation is performed over the whole domain.

The connection with the drift diffusion simulation is specified in physical_model heat_source. The model name is drift_diffusion_dissipation. Furthermore, we have to specify the name of the drift-diffusion simulation (dd).

Heat reservoir boundary conditions are chosen with the keyword type = heat_reservoir (which is the default type).

We fix temperature at the boundary with the keyword temperature. The default value is taken from the simulation section.

Solver section (up)

The section sweep in this case defines a loop where both the thermal and the driftdiffusion simulations are performed for each bias step.

However, while the drift diffusion simulation is performed at each bias point at a constant temperature (as defined in the simulation section), the thermal simulation takes into account the drift diffusion results at each bias step.
In a furher tutorial we will see how to perform complete self-consistent calulations, where the particle transports is treated in non-isothermal conditions.

The simulation names listed in the simulation statement above have to appear in the proper order. In this case, in order to include the particle transport effect on the heat source, the thermal simulation (tt) must appear after the drift diffusion simulation name (dd in this case).

Physics section (up)

Here we indicate the statistics (FD means Fermi-Dirac).

Simulation section (up)

Through the keyword thermal we refer to all the thermal quantities, i.e the temperature, the heat sources and the power fluxes.

The heat source data and the power flux data are contained in a unique element-based file.

All files, except sweep_dd.dat, are written after the sweep simulation.

The filenames comprise the simulation name. For example, the temperature data is named tt_nodal.dat.

Output (up)

All results are referred to a bias of 1.2 V.

Band structure (dd_nodal.dat):

Temperature profile (tt_nodal_dat):

Heat Sources (tt_elemental.dat):

As one can see, the recombination effect is higher at the interface, since around the interface region the system is very far from equilibrium.

The discontinuity at the interface is due to the different non radiative recombination times of the electrons and holes.

Power fluxes (tt_elemental.dat):