Skip to content
Machine Learning Daily, home

Scalable Fabrication Platform Enables Flexible and Transparent Photonic Chips

MIT researchers have developed a new fabrication process that allows for the mass production of flexible, transparent silicon-photonics chips.

DERRICKHEALTH & BIO AI636 WORDS

A research team at MIT has established a scalable fabrication platform for creating silicon-photonics chips that are both mechanically flexible and optically transparent. This development marks a transition from rigid, opaque semiconductor architectures to a form factor capable of conforming to curved surfaces while maintaining high-performance light-based data processing.

The methodology leverages standard 300-millimeter foundry manufacturing techniques, a shift from previous laboratory-scale demonstrations that were limited by low throughput. By collaborating with engineers at NY Creates at the Albany NanoTech Complex, the researchers successfully integrated these processes into existing industrial workflows to ensure high-volume potential.

The fabrication sequence begins with the standard deposition and patterning of optical waveguides on a rigid silicon substrate. A temporary support layer is bonded to the top, allowing the team to invert the wafer and remove the original silicon base entirely. This leaves behind an ultrathin layer of oxide and waveguide material, measuring less than a tenth of a human hair in thickness.

Adhesive bonding to a transparent polyester film provides the necessary structural integrity for the final device. The team utilized low-temperature processes, maintained at or below 500 degrees Celsius, to manage thermal stress and prevent mechanical failure during the thinning process. A combination of industrial thinning and selective chemical etching ensures the delicate optical layers remain intact throughout the transition.

Tal Sneh, an EECS graduate student and lead author of the study, noted that the use of stable 300-millimeter foundry tools allows for the design of systems with a vast number of integrated devices. This reliability is essential for ensuring that the final flexible chips meet rigorous performance specifications required for commercial deployment.

Thomas Dyer of NY Creates highlighted the precision required during the removal phase, noting that improper strain management often leads to surface ripples or catastrophic shattering of the wafer. The team successfully navigated these physical constraints by optimizing the sequence of removal methods to maintain flatness across the entire diameter.

The research, published in the journal Optica, included extensive validation testing to confirm the platform’s durability and optical clarity. Experiments involved bending the chips around cylinders of varying diameters, including those as small as a screw, with no measurable degradation in waveguiding performance.

Jelena Notaros, the Robert J. Shillman Career Development Associate Professor of Electrical Engineering and Computer Science at MIT, emphasized the potential for this platform to be made accessible to the broader research community. The ability to produce these wafers at scale provides a foundation for integrating complex photonic components into non-traditional hardware environments.

The significance of this work lies in its potential to replace bulky, rigid optical systems currently found in heads-up displays and specialized sensors. By reducing the weight and physical footprint of these systems, the technology could enable more discreet health monitors and advanced augmented-reality visors for aviation or industrial use.

Maintaining optical performance while bending the substrate addresses a long-standing trade-off in photonics engineering. Previous attempts to achieve flexibility often sacrificed the precision of the waveguide, but this platform demonstrates that high-fidelity data transmission is possible even under mechanical stress.

This advancement effectively bridges the gap between traditional semiconductor manufacturing and the emerging demand for flexible electronics. By utilizing existing foundry infrastructure, the researchers have created a pathway for rapid adoption in sectors that require both high-speed data processing and physical adaptability.

Future development will focus on increasing the complexity of the integrated components and refining waveguide efficiency. The researchers intend to further optimize transparency metrics to ensure that the chips remain unobtrusive when placed in front of a user’s field of vision.

The project received funding from the National Science Foundation, the U.S. Defense Advanced Research Projects Agency, and a MathWorks Fellowship. As the team moves toward broader application, the focus remains on scaling these fabrication techniques to support the next generation of flexible photonic systems.

REFERENCED

  1. pmc.ncbi.nlm.nih.govThomas Dyer
  2. opg.optica.orgpublished in the journal
  3. eecs.mit.eduJelena Notaros
  4. nsf.govNational Science Foundation

FILED TO HEALTH & BIO AI · ALSO COMPUTE, RESEARCH

MORE IN HEALTH & BIO AI

ALL

THE DISPATCH

Applied machine learning, filed daily.

Model releases, silicon, clinical deployment, and the policy shaping them. No digest padding.