nanohub-padre

A Python library for creating and running PADRE semiconductor device simulations.

Overview

nanohub-padre provides a Pythonic interface to generate PADRE input decks, making it easier to set up complex device simulations programmatically. PADRE is a 2D/3D device simulator that solves the drift-diffusion equations for semiconductor devices.

Features

• Pythonic Interface: Define meshes, regions, doping profiles, and solver settings using Python objects
• Device Factory Functions: Pre-built functions to create common devices (PN/PIN diode, NIN/PIP isotype diode, MOSFET, BJT, solar cell, etc.)
• Complete PADRE Support: Covers mesh generation, material properties, physical models, and solve commands
• Validation: Built-in parameter validation and helpful error messages
• Examples: Ready-to-run examples for common device structures

Installation

pip install nanohub-padre

Or install from source:

git clone https://github.com/denphi/nanohub-padre.git
cd nanohub-padre
pip install -e .

Quick Start

Using Device Factory Functions (Recommended)

from nanohubpadre import create_mosfet, Solve, Log

# Create an NMOS transistor with one line
sim = create_mosfet(
    channel_length=0.05,
    device_type="nmos",
    temperature=300
)

# Add solve commands
sim.add_solve(Solve(initial=True))
sim.add_log(Log(ivfile="idvg"))
sim.add_solve(Solve(v3=0, vstep=0.1, nsteps=15, electrode=3))

# Generate the input deck
print(sim.generate_deck())

Building from Scratch

from nanohubpadre import (
    Simulation, Mesh, Region, Electrode, Doping,
    Contact, Models, System, Solve
)

# Create simulation
sim = Simulation(title="Simple PN Diode")

# Define mesh
sim.mesh = Mesh(nx=100, ny=3)
sim.mesh.add_x_mesh(1, 0)
sim.mesh.add_x_mesh(100, 1.0)
sim.mesh.add_y_mesh(1, 0)
sim.mesh.add_y_mesh(3, 1)

# Define silicon region
sim.add_region(Region(1, ix_low=1, ix_high=100, iy_low=1, iy_high=3, silicon=True))

# Define electrodes
sim.add_electrode(Electrode(1, ix_low=1, ix_high=1, iy_low=1, iy_high=3))
sim.add_electrode(Electrode(2, ix_low=100, ix_high=100, iy_low=1, iy_high=3))

# Define doping
sim.add_doping(Doping(p_type=True, concentration=1e17, uniform=True, x_right=0.5))
sim.add_doping(Doping(n_type=True, concentration=1e17, uniform=True, x_left=0.5))

# Set contacts
sim.add_contact(Contact(all_contacts=True, neutral=True))

# Configure models
sim.models = Models(temperature=300, srh=True, conmob=True, fldmob=True)
sim.system = System(electrons=True, holes=True, newton=True)

# Solve
sim.add_solve(Solve(initial=True))

# Generate and print the input deck
print(sim.generate_deck())

Device Factory Functions

The library includes factory functions for common devices:

Examples

Tutorial notebooks (examples/notebooks/)

A twelve-part course that puts every device's textbook theory side by side with a PADRE simulation and states, each time, what differs and why (constants, approximations, 2-D effects, simulator limitations). The analytic formulas use PADRE's own material constants, so remaining differences are physics.

Device scripts (examples/devices/)

Short, runnable companions to the notebooks (pn_diode, schottky_diode, mosfet, bjt, mos_capacitor, mesfet, nin_diode, solar_cell). Each builds the default device, runs PADRE if it is on your PATH (otherwise prints the input deck), and prints the key textbook-vs-PADRE comparisons:

python examples/devices/pn_diode_example.py

Deck translations (examples/*.py)

Python equivalents of classic PADRE input decks (pndiode.py, moscap.py, mosfet_equivalent.py, mesfet.py, single_mosgap.py), each paired with its .in file:

PYTHONPATH=. python3 examples/pndiode.py > pndiode.inp
padre < pndiode.inp > pndiode.out

Supported Commands

nanohub-padre supports all major PADRE commands:

Documentation

Full documentation is available at https://nanohub-padre.readthedocs.io/

Testing

Run the test suite:

pytest tests/ -v

License

MIT License - see LICENSE file for details.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.