Worked Examples

Reference

Four structures that ship with EpiSolve: one from each category on the Open Example menu, and a benchmark that checks the tool against published results. Each comes set up to run — open it, press Run Simulation, and read the result against the page that explains it.

01A p–n junction

Open Example → Semiconductor devices → GaAs PN Diode. The structure is 500 nm of n-type GaAs (10¹⁷ per cm³) on 500 nm of p-type GaAs (10¹⁷ per cm³), over the substrate. It opens with a Schottky top contact (0.7 eV barrier), 300 K, both carriers active and dopants fully ionised.

Press Run Simulation and look at the band diagram. The conduction and valence bands step across the junction at 500 nm by the built-in potential — ~1.3 volts for GaAs doped this way — with a depletion region straddling the junction and the carriers swept out of it. This is the plainest illustration of the basic output of the model: the equilibrium bands of a device, generated from a layer structure. Read it alongside The Results Window and The Plot Window.

02A Schottky C–V extraction

Open Example → C-V devices → Schottky N GaAs 1e16 Mott Schottky. A single 1000 nm n-type GaAs layer (10¹⁶ per cm³) under a Schottky contact, and it opens with CV calculation already ticked and a reverse sweep from −3 V to 0 V set up.

Press Run Simulation. The C-V Sweep Settings dialog appears pre-filled — press Run — and the C–V results open. On the 1/C² Analysis tab the straight-line fit returns an extracted doping close to the 10¹⁶ per cm³ the layer was given, together with a built-in potential from the intercept. This is the standard doping measurement, reproduced in the model: you put in a known doping and the analysis recovers it. See Capacitance–Voltage.

03A two-dimensional electron gas

Open Example → Quantum devices → AlGaAs GaAs 2D Electron Gas. A modulation-doped stack: 500 nm of undoped GaAs, then a thin undoped AlGaAs spacer at 60% aluminium, then 100 nm of n-type AlGaAs of the same composition. The doped barrier gives up its electrons, which fall into a triangular notch in the GaAs at the interface. It opens with Find quantised states ticked and the Schrödinger window set to bracket that interface, 400 nm to 600 nm.

Press Run Simulation. The band diagram shows the notch pulled below the Fermi level at the interface — the 2DEG — and a separate Schrödinger Region window lists the confined electron states and their sheet density. This example is also the clearest demonstration of a recurring point: the confined sheet density reported there is the true two-dimensional figure, and it is lower than the semiclassical density drawn on the main plot. See Quantised States.

04A benchmark against nextnano and Snider

Open Example → Quantum devices → Snider Nextnano QW Hemt. This is the 1-D Schrödinger–Poisson case from nextnano’s tutorial — an AlGaAs/GaAs structure whose donor and acceptor doping is an ion-implanted profile rather than a set of grown layers. An implant is a smooth curve, not a step; here it is approximated as a stack of uniformly doped layers whose net charge follows that curve, read layer by layer straight off the published profile plot.

It opens ready to run, with Find quantised states and Quantum feedback on, the Schrödinger window set over the well, and the dopants fully ionised at 300 K. Press Run Simulation: three confined electron states form a two-dimensional gas in the well, with a sheet density near 6.6×10¹¹ cm⁻².

The point is how well that lands despite the doping being estimated from a graph. The 2DEG sheet density comes out 6.63×10¹¹ cm⁻², against nextnano’s published 6.64×10¹¹ and Snider’s 1dpoi 6.36×10¹¹, and the three subband energies agree with both to a few meV. Running the same structure through 1dpoi locally reproduces it too — a graphical estimate of the implant is enough to recover the published electrostatics. See Quantised States.