Maskless photolithography can simplify micro-patterning when the cost and lead time of producing physical photomasks are a barrier to experimentation. A system developed at the Semiconductor Materials Laboratory (SML), RRCAT, uses a digital projector and stereo microscope to expose patterns directly onto photoresist-coated substrates. The approach demonstrates how a relatively simple optical arrangement can support semiconductor device prototyping without a dedicated mask for each design.
Why consider maskless photolithography?
Photolithography transfers patterns onto a substrate, and those patterns help define the structure and function of semiconductor devices. Conventional ultraviolet photolithography uses a physical photomask and can provide high throughput, making it well suited to industrial production. However, masks must be designed and fabricated, and their production adds both cost and time before a design can be tested.
That delay can constrain development when engineers need to revise a layout or evaluate several geometries. A maskless process takes a different approach: patterned light is projected directly onto the substrate. Because the pattern is supplied optically rather than by a physical mask, designs can be changed without commissioning a new mask for every iteration. This makes the method relevant to selected research, development and prototyping workflows, while not replacing the throughput advantages of mask-based production.
How the projector-and-microscope arrangement works
The SML system directs light from a projector through suitable lenses into the trinocular head of a stereo microscope. The microscope objective then focuses the projected image onto a photoresist-coated sample held on an x-y-z controlled stage. The microscope eyepiece is used to check focus and align the incident light with the substrate.
The projector described in the system has a resolution of 1920 × 1080, brightness of 3500 lumens and a contrast ratio of 1:13000. Its output is focused through the microscope over an area of approximately 8 × 6 mm. The reported minimum pixel size at the exposure plane is up to 5.5 µm. With microscope magnification set at around 2.5×, the projected image is focused onto the sample’s exposure area.
These figures describe different aspects of the setup: pixel size refers to the projected image at the exposure plane, while the smallest feature actually produced also depends on the photoresist process and exposure conditions. In the reported work, the minimum feature size achieved was 20 µm.
Photoresist preparation and exposure
Pattern quality depends not only on the optics but also on preparing and processing the photoresist consistently. In the reported procedure, Shipley S1813 photoresist was spin-coated onto GaAs substrates at 3500 rpm for one minute. The coated substrates were then soft-baked at 95 °C.
For alignment, red patterned light was first used to position the image on the sample. The corresponding pattern was then exposed using white light. Exposure time and projector intensity were optimized through repeated trials to obtain suitable developed patterns. The reported process used a two-second exposure followed by one minute of development in MFCD26 developer.
These settings are specific to the described materials and optical arrangement; they should not be treated as universal process parameters. A different substrate, resist, magnification or illumination setup may require its own process optimization.
Applications demonstrated with the system
The arrangement was used to pattern several types of semiconductor structures, including laser diodes, photodetectors and Hall bar samples. It was also used to produce patterns for electronic transport measurements, including Hall bar and Corbino disc geometries. Since the exposure pattern is generated by the projector, the system can accommodate different layouts within the capabilities of its optics and process.
Measurements on patterned two-dimensional electron gas (2DEG) heterostructures showed magnetic-field-dependent changes in longitudinal and transverse resistivity, identified as Rxx and Rxy. The work also reports classical and quantum Hall effect measurements, as well as an inverse spin Hall effect signal in a patterned sample. These examples illustrate the system’s role in fabricating test structures for device and materials characterization.
Practical considerations for engineering teams
For engineers evaluating a maskless patterning setup, the reported system highlights several factors to consider:
- Optical resolution and feature size: Projector pixel dimensions at the sample are not the only determinant of the final feature. The reported 5.5 µm minimum pixel size and 20 µm minimum fabricated feature describe distinct outcomes.
- Exposure-area trade-offs: Microscope magnification affects the exposure area and the projected pattern dimensions. Changing magnification can therefore alter how much of a substrate can be patterned at one time.
- Alignment and focus: The controlled stage and microscope viewing path help position the pattern relative to the sample and establish focus before exposure.
- Process development: Projector intensity and exposure time require optimization alongside photoresist coating, baking and development conditions.
The original system was described as a low-cost build, with a stated cost within 2 lakhs. That figure belongs to the reported development and should be considered in its original context rather than assumed to represent the cost of every implementation.
A flexible option for selected patterning tasks
Maskless photolithography offers a way to create and revise micro-patterns without fabricating a separate physical mask for each design. The projector-and-microscope approach described by SML produced structures for semiconductor devices and transport measurements, while illustrating the importance of optical alignment, magnification and resist-process optimization.
For industrial automation and semiconductor engineering teams, the practical takeaway is to match the patterning method to the task. A maskless setup can support flexible development and characterization work where design changes matter; mask-based photolithography remains suited to applications requiring its high-throughput capabilities. Understanding that distinction helps teams choose an appropriate route from prototype pattern to production process.