Local Oxidation Isolation Punchthrough

isolationex01.in : Local Oxidation Isolation Punchthrough

Requires: SSuprem 4/DevEdit/S-Pisces
Minimum Versions: Athena 5.22.1.R, Atlas 5.22.1.R

This example simulates punchthrough between two n+ areas separated by an oxide isolation structure. The isolation structure is a local oxidation using Sealed Interface Local Oxidation (SILO) combined with a p-type field stop implant.

This example demonstrates the following features:

  • Processing of a local oxidation structure
  • Remeshing at the process to device simulation interface
  • Remeshing within device simulation
  • Ramping iv curves to a compliance limit

This example starts with process simulation using Athena, remeshes using DEVEDIT, and interfaces to Atlas. Atlas ramps the voltage up to a predefined drain current value, a further remesh in DEVEDIT is performed, and Atlas ramps to the punchthrough target current.

For details on local oxidation (bird's beak) process simulations, see the ATHENA_SSUPREM4 examples section.

The electrodes are placed at the end of the process flow and then subsequently named with electrode statements. Only the x coordinate needs to be specified in this case as the electrodes are on the surface of the structure. Should they be buried, both x and y coordinates need to be specified. Both polysilicon and metal regions are treated as possible electrodes.

The electrode name=substrate backside statement will automatically place an electrode on the bottom of the structure. This is the best and most efficient way of approximating a well contact.

After the process simulation, the structure is remeshed with DEVEDIT with the command: go devedit

At this point the solution only contains solution quantities from the process simulator, so that only impurity values are available as remeshing criteria. These are selected with the commands:

imp.refine imp="Arsenic" sensitivity=0.5
imp.refine imp="Boron" sensitivity=0.4
imp.refine imp="Phosphorus" sensitivity=0.5

Boron is weighted so that the mesh will be more sensitive to boron concentration gradients. In other words, the smaller the value for sensitivity, the more sensitive the mesh will be to concentration gradients and the denser the mesh will be.

In this example, constr.mesh commands control the maximum allowable angle within a triangle, specific to a material type. For example:
constr.mesh type=Insulator default max.angle=178
states that very obtuse triangles are allowed in all insulator (oxide) regions.

This is acceptable, as obtuse triangles are essentially only a problem in semiconductor regions when solving carrier transport equations. The reason for doing this is because by lowering the criteria for maintaining zero obtuse triangles, the more relaxed the mesh can be and the simulation will run faster.

Next the structure is passed into the device simulator Atlas for biasing and design parameter extraction. This is accomplished with the command go atlas.

As the device is simply punching though, the current is governed almost completely by the potential barrier and drain induced barrier lowering effects. Thus a special mobility model is not selected to speed up the calculation.

An interface charge of 1e11 is selected arbitrarily, and may be adjusted as a primary calibration parameter to match experimental results.

A sequence of solve commands ramps the drain contact up to 30V using a current compliance of 0.001uA. The simulation stops when either of these is exceeded. In the present example, this will be at about 14V drain voltage. It is assumed that the breakdown will not occur during this ramp. For the last bias point, a single structure file, containing both the mesh and the solution, is saved.

After the previous biasing of about 14 volts, the structure is remeshed again using DEVEDIT this time as a function of potential gradient, doping gradient and electron concentration gradient resulting in a more accurate mesh. The current compliance value has been chosen such that the resulting drain voltage has a value that will always be lower than the punchthrough voltage, but is high enough to offer a closer solution to the final IV area of interest.

Another log file is opened for the final punchthrough IV curve to be held. The solution is reloaded with the new mesh created by DEVEDIT and the drain bias is increased up to 30 volts or 0.1um/W/L, whichever is reached sooner.

Finally, a design parameter relating to the punchthrough voltage is extracted.

The command
extract name="n_isolation_vpt" x.val from curve(abs(v."drain"),abs(i."drain")) where y.val = 1.0e-7
should be allowed to run over two lines. Do not use a line continuation character in any extract statement. It defines the measure of punchthrough as that drain voltage at which the drain current is 0.1uA/um.

To load and run this example, select the Load example button in DeckBuild. This will copy the input file and any support files to your current working directory. Select the run button to execute the example.