Human Body Model in a MOSFET

esdex03.in : Human Body Model in a MOSFET

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

In this example, transient simulation of Electrostatic Discharge (ESD) on an NMOS transistor using the Human Body Model (HBM) is performed. During the ESD event significant local heating is produced. The solution of local lattice temperature is included. This example shows:

  • Formation of a MOS structure in Athena
  • Interface to Atlas
  • Selection of lattice heat flow models
  • Use of current boundary conditions and non-linear transient pulse to simulate a HBM ESD event.
  • Analysis of both the temperature distribution in the device and non-isothermal IV curve.

The MOS structure is constructed using Athena process simulation. This structure is passed to Atlas for a HBM test simulation. The NMOS transistor is a 0.8um LDD device using oxide spacers. For a more complete description of the MOS process simulation see the MOS examples.

The Atlas syntax shows a simple and effective test procedure for HBM ESD simulation that may be used with any initial NMOS structure.

The Atlas simulation begins with definition of the models and material parameters of the device. The contact statement is used to define the workfunction of the polysilicon electrode. The material statement is used to define the capture times (electron and hole lifetimes) in the semiconductor.

The physical models used in this simulation reflect the different physical effects important to ESD device simulation. The mobility model 'analytic' accounts for concentration and the temperature dependencies.

The mobility model 'fldmob' accounts for the electric field dependency. In addition to the Shockley-Read-Hall recombination model (SRH), the recombination model 'auger' is included to take into account the high injection level effects. Band gap narrowing is taken into account by means of the bgn parameter. The continuity equations for both carriers are selected by default. The impact ionization model is enabled using the impact selb statement.

A nonisothermal approach is used, which means that the heat flow equation is solved in addition to the semiconductor equations and all physical parameters become temperature dependent. The syntax models lat.temp enables the solution of the heat flow equation.

The definition of the thermal boundary conditions is very important in all non-isothermal simulations. Thermal boundary conditions are defined in the thermcontact statement. A value of the thermal conductance determined from the heat conductivity of the substrate is specified at the thermocontact located along the substrate. Thermal isolation is assumed where no thermal contacts are specified. Here thermal isolation conditions are assumed on all other surfaces besides the bottom.

To simulate the device interaction with the simplified HBM test circuit the current pulse is applied to the drain of the MOS structure in the reverse direction. The current pulse risetime is 10ns and the exponential decay time is 150ns, which corresponds to the discharge of the 100pF capacitor through the 1500 Ohm resistor into the test device. These values are the definition of the Human Body Model.

The transient current/voltage characteristics are saved in the LOG file. Using TonyPlot it is possible to observe the maximum temperature in the device versus time. The temperature increases with time and peaks significantly after the peak current. The value of the maximum temperature can be extracted and used as a figure of merit for comparing the ESD protection capability of various device designs.

The probe statement is used to define quantities to be measured at each bias step. The values from these quantities are saved in the log file and can be plotted versus time and bias. The first probe statement saves the lattice temperature at the drain contact and the second electric field across the gate oxide. The former can be used to determine the likelihood of metal melting and the latter to determine likelihood of gate dielectric breakdown.

The solution at a time of 10ns and at the final stage are saved. All internal distributions can be observed using TonyPlot. Most interesting is that the temperature distribution in the MOSFET shows the location and value of the hot spot.

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.