Skip to main content

Overcoming the Limits of Diffraction with Superresolution Lighting on a Chip

Periodic Reporting for period 2 - ChipScope (Overcoming the Limits of Diffraction with Superresolution Lighting on a Chip)

Reporting period: 2019-01-01 to 2020-12-31

The physical laws of diffraction generally limit the spatial resolution of optical systems, being about 250 nm for light in the visible range. That is the reason why we usually cannot directly observe virus, single proteins, or molecules with conventional optical microscopes. Attempts to overcome this limit have led to “super-resolution” techniques, like STED, STORM, PALM, NSOM, etc., all of them based on bulky and expensive optical systems and complex sample preparations.
The approach to optical microscopy of ChipScope is different from existing solutions. In ChipScope, we build a lens-free optical microscope in which we use a set of light sources so small that are well below the diffraction limit. The intensity of light transmitted through the sample is then captured by a camera as each LED is lit, to build up a shadow of the sample pixel by pixel, where each LED denotes a pixel. Although the collected light is diffraction-limited, its origin is the lit LED, and that position is known. As a result, the LED size, not the diffraction limit, determines the resolution. In ChipScope, we have developed arrays of semiconductor nanoLEDs with a pixel-to-pixel spacing of less than the diffraction limit that can be turned on independently. This will be a leap forward in extreme miniaturization. Chipscope technology will make optical microscopes chip-sized, simple, affordable and ubiquitously available, not only for laboratories working in manifold research fields but also in everyday life.
The first challenge of ChipScope was downscaling the size of existing technology on GaN LEDs by a factor of ~100 in order to address the problem of sub-diffraction resolution. The second challenge was operating these nanoLED arrays in full synchronization with a CMOS light sensor. Finally, the third challenge dealed with the advanced theoretical studies regarding the fundamental physics of light-matter interaction at such a small scale that will serve us to better understand the further potential and limitations of the ChipScope approach.
Within the project, the first chip-sized ‘Chip-Scopes’ were developed, tested, calibrated, and compared with state-of-the-art microscopy systems.
Before ChipScope, the state of the art on addressable tiny LEDs was around 100 μm. Thanks to work at partner TUBS, ChipScope demonstrated the approach with GaN LEDs measuring 200 nm. Starting from a quick initial prototype of directly addressable 8x8 LED array with 5/10 µm pixel size/pitch (Led1), the features of the nanoLED array were improved towards pixel sizes below the Abbe’s limit. A 2x32 linear array of 200nm green-LEDs (Led3) was achieved by employing photo- and e-beam lithography. Furthermore, a matrix-addressable array of 32x32 10µm-LEDs (Led4) was developed, establishing the basis to achieve large GaN LED arrays with sub-µm sized LEDs. Supporting the LED fabrication, UNITOV formulated the design guidelines for the nano-LEDs based on the electrical models and optical simulations, identifying the critical aspects and providing optimized structures and materials. UNITOV also analyzed the resolution limits and image formation of the Chipscope approach.
To control these novel LED arrays, UB designed the driving electronics, first with off-the-shelf components for the earlier prototype (Mic1) and later with an integrated CMOS solution for the most advanced versions of the microscopes (Mic2 to Mic4). UB also designed an SPAD CMOS detector, capable of detecting single photons. In parallel, the different microscopes were assembled in the UB facilities, where two different setups were implemented based on sample mechanical movement and in microfluidics. Additional microscopes based on commercial µdisplays were built to demonstrate the Chipscope principle with large LED arrays, showing the potential to have chip-sized electronically activated scanning optical microscopes, with a high-resolution (~µm) and wide FOV without any movable part or image post-processing. Two additional setups based on the optical downscaling of the LEDs arrays were build to test the feasibility of using electronic control of nano-light sources for use in conventional fluorescence-based microscopy. Last but not least, Ymaging and the UB developed the image software to control the image acquisition and image reconstruction and post-processing for each microscope.
To characterize the Chipscope prototypes, LMU developed novel tools and fluorescence test samples based on DNA origami in order to adapt the size and structures of DNA origami nanorulers for the needs of ChipScope, improving the signal-to-noise ratio and the signal stability over time of DNA origami reference samples. The quantification of spatial resolution tests of the Chipscope microscopes were done for the different microscopes. Moreover, a way to bring specimens into the structured light source's close vicinity was vital for proper microscope operation. To do so, AIT optimized mixed PMMA and PDMS microfluidic chip fabrication processes and tested the fluidics on a prototype board for long-term cell studies. On the other hand, MUW established a cell seeding and attachment procedure, verified over LED chips equipped with a microfluidic chamber from AIT. MUW also provided biological samples of human lung fibroblasts to be observed with the microscopes.
The Chipscope results were disseminated to the general public and the scientific community through more than 48 conferences and with 27 peer-review publications, 14 wider public articles, and by organizing four public workshops in the framework of international conferences. Chipscope technology is being used on a new FET PROACTIVE EU-project (SMILE) and one spin-off (QubeDot GmbH) was founded in 2019.
The ChipScope approach to superresolution based on nano-LED arrays was 100% novelty. The viability of the ChipScope concept could only be studied and proven if an array of sufficiently small independently addressable light sources was fabricated, which was the key breakthrough addressed in this project. This goal was achieved with the fabrication of the smallest direct addressable nanoLED arrays. Besides, we required video in ‘real-time’. For that reason, ChipScope needed fast control of individual nanoLEDs, a major challenge considering the small sizes of the nanoLEDs with a pixel-to-pixel spacing of less than the diffraction limit. This requirement was also achieved through the CMOS integration of the nanoLED drivers, capable of switching on-off the LEDs up to 1MHz.
Although, the superresolution was not achieved due to the expansion of the spotlight from the light generation plane to the chip surface, being able to have a chip-sized scanning optical microscope (such as the µdisplay-based microscope) can have a great impact on our society. Due to the possibility of a high-volume production based on semiconductor technology, the ChipScope has opened the door to the widespread usage of optical microscopy in every laboratory and everyday life, e.g. for personal health monitoring or point-of-care applications. All this without forgetting that electronic control of nanometric light sources can provide numerous additional novel applications in diverse fields like nanofabrication, optical computing, communications, among others.
SEM image of etched pixels in the 200 nm structure
Basic principle of the ChipScope approach
Microdisplay based Chipscope microscope
200nm LED array and sample holder details on the microscope