Mask creation with Python (and gdspy)

This repository implements a framework for the generation of mask files (GDS-II) for semiconductor processing via python scripts. The framework is based on the gdspy python package.

The use case for the mask sets implemented in this repository is for the fabrication of simple wafer-wafer bonded particle sensor diodes, as part of the PhD thesis work of Johannes Wüthrich. The corresponding PhD thesis will be linked here in the future.

Implemented Mask Sets

Project	File	Notes
BRNC_Run1_202003	mask/sets/BRNC_Run1_202003.py	Mask set for the (failed) first 6" production run. Do not invoke directly, but use BRNC_Run1_202003_P and BRNC_Run1_202003_N.
BRNC_Run1_202003_P	mask/sets/BRNC_Run1_202003_P.py	P-wafer mask set of BRNC_Run1_202003.
BRNC_Run1_202003_N	mask/sets/BRNC_Run1_202003_N.py	N-wafer mask set of BRNC_Run1_202003.
BRNC_Run2_202105	mask/sets/BRNC_Run2_202105.py	Mask set for the (successful) second 4" production run. Do not invoke directly, but use BRNC_Run2_202105_P and BRNC_Run2_202105_N.
BRNC_Run2_202105_P	mask/sets/BRNC_Run2_202105_P.py	P-wafer mask set of BRNC_Run2_202105.
BRNC_Run2_202105_N	mask/sets/BRNC_Run2_202105_N.py	N-wafer mask set of BRNC_Run2_202105.

Framework Structure

Module	Notes
mask.config	Helper classes for layer and export handling. Used in the GLOBAL config (see mask.config.__init__) and the specific process definitions (see mask.processes).
mask.elements	Classes representing mask elements, for example an entire diode. Elements usually extend over multiple mask layers. And are thus linked to a specific process.
mask.forms	Classes representing simple geometrical forms, usually limited to a single (given layer).
mask.macros	Functions for creating parts of mask set.
mask.processes	Modules containing the definition of a given fabrication process. This usually includes the definition of available layers, the definition of layers to be exported individually and process specific macro functions.
mask.sets	Executable (__main__) files for creating a given mask set. The name of a given file without the .py suffix (e.g. BRNC_Run2_202105_N) represents a project name to be used with the Makefile.
mask.tools	Various utility function and classes for mask exporting, mask manipulation, mask verification etc.

Usage

The main usage for running the framework and to generate the output files is via the provided Makefile.

The name of the current project file (see mask.sets) needs to be defined via an environment variable (for example as a prefix to the make command, as PROJECT=[PROJECT] make [TARGET]). The following targets are available:

Target	Example	Notes
all	PROJECT=BRNC_Run2_202105_N make all	Create the GDS-II file for the given process.
show	PROJECT=BRNC_Run2_202105_N make show	Same as all but also show the generated file using KLayout
clean	PROJECT=BRNC_Run2_202105_N make clean	Remove all output files of the given project.
export	PROJECT=BRNC_Run2_202105_N make export	Export individual layers to be used with a laser writer for mask writing.
check	PROJECT=BRNC_Run2_202105_N make check	Check if the generated GDS-II file is compatible with the Heidelberg Laser-Writer.

Implement Custom Sets

The following is the general strategy for using this framework for creating new mask sets:

1. Implement the corresponding process definition in mask.processes.
2. Create new custom forms and elements in mask.forms and mask.elements.
3. Implement the mask set as a new project in mask.sets.
   • Use the existing files as a template.
   • Certain functions (such as exporting, file saving etc.) are directly implemented in the project file.

Output Files

The output files are placed in a ./files/[PROJECT]/ subfolder, where [PROJECT] is the project name as defined via the environment variable (see above).

Dependencies

Programs

KLayout

For viewing the generated .gds files KLayout can be used (via the make show command).

KLayout is available online (https://www.klayout.de/) and via the package manager of common Linux distributions.

Python

The code is written for Python 3 and was tested with Python 3.9.2.

The following python packages are necessary

• gdspy
• rich
• gitpython
• numpy
• matplotlib

The python requirements can be installed with

python -m pip install -r requirements.txt

The use of a python virtual environment is recommended.