Tiny LED could turn phone cameras into high-res microscopes

Researchers from the Disruptive & Sustainable Technologies for Agricultural Precision (DiSTAP) and the Critical Analytics for Manufacturing Personalised-Medicine (CAMP) Interdisciplinary Research Groups (IRG) of Singapore-MIT Alliance for Research and Technology (SMART), have developed a small LED (light-emitting diode) that enabled the conversion of existing mobile phone cameras into high-resolution microscopes. Smaller than the wavelength of light, the new LED was used to build a holographic microscope, paving the way for existing cameras in everyday devices such as mobile phones to be converted into microscopes via modifications to the silicon chip and software. This technology also represents a step forward in the miniaturisation of diagnostics for indoor farmers and sustainable agriculture.

This breakthrough was supplemented by the researchers’ development of a neural networking algorithm that can reconstruct objects measured by the holographic microscope, thus enabling enhanced examination of microscopic objects such as cells and bacteria without the need for advancement in photonics — the building of a powerful on-chip emitter that is smaller than a micrometre, which has been a challenge in the field.

The light in most photonic chips originates from off-chip sources, which leads to low overall energy efficiency and fundamentally limits the scalability of these chips. To address this issue, researchers have developed on-chip emitters using various materials such as rare-earth-doped glass, Ge-on-Si, and heterogeneously integrated III-V materials. While emitters based on these materials have shown promising device performance, integrating their fabrication processes into standard complementary metal-oxide-semiconductor (CMOS) platforms remains challenging.

While silicon (Si) has shown potential as a material for nanoscale and individually controllable emitters, Si emitters suffer from low quantum efficiency because of the indirect bandgap, and this disadvantage combined with the limitations set by the available materials and fabrication tools has hindered the realisation of a small native Si emitter in CMOS.

In a paper published in Nature Communications, SMART researchers described their development of the small Si emitter with a light intensity comparable to that of state-of-the-art Si emitters with much larger emission areas. In a related development, SMART researchers also revealed their construction of a novel, untrained deep neural network architecture capable of reconstructing images from a holographic microscope in a paper published in the journal Optica.

The novel LED is a CMOS-integrated sub-wavelength scale LED at room temperature exhibiting high spatial intensity (102 ± 48 mW/cm²) and possessing a small emission area (0.09 ± 0.04 μm²). In order to demonstrate a potential practical application, the researchers integrated this LED into an in-line, centimetre-scale, and all-silicon holographic microscope requiring no lens or pinhole, integral to a field known as lensless holography. A commonly faced obstacle in lensless holography is computational reconstruction of the imaged object. Traditional reconstruction methods require detailed knowledge of the experimental set-up for accurate reconstruction and are sensitive to difficult-to-control variables such as optical aberrations, the presence of noise, and the twin image problem.

The researchers also developed a deep neural network architecture to improve the quality of image reconstruction. This untrained deep neural network incorporates total variation regularisation for increased contrast and takes into account the wide spectral bandwidth of the source. Unlike traditional methods of computational reconstruction, this neural network eliminates the need for training by embedding a physics model within the algorithm. The neural network also offers blind source spectrum recovery from a single diffracted intensity pattern, which marks a departure from all previous supervised learning techniques.

The untrained neural network demonstrated in this study allows researchers to use novel light sources without prior knowledge of the source spectrum or beam profile, such as the novel Si LED described above, fabricated via fully commercial, unmodified bulk CMOS microelectronics. The researchers believe that this combination of CMOS micro-LEDS and the neural network can be used in other computational imaging applications, such as a compact microscope for live-cell tracking or spectroscopic imaging of biological tissues such as living plants. This work also demonstrates the feasibility of next-generation on-chip imaging systems.

Iksung Kang, lead author of the Optica paper and Research Assistant at MIT at the time of this research, said, “Our breakthrough represents a proof of concept that could be hugely impactful for numerous applications requiring the use of micro-LEDs.”

Rajeev Ram, Principal Investigator at SMART CAMP and co-author of both papers, said that the new LED has a range of other applications alongside its potential in lensless holography. “Because its wavelength is within the minimum absorption window of biological tissues, together with its high intensity and nanoscale emission area, our LED could be ideal for bio-imaging and bio-sensing applications, including near-field microscopy and implantable CMOS devices,” Ram said.

Image credit: iStock.com/dhdezvalle