Periodic Reporting for period 1 - NEXG-UV (Towards next Generation UV Optical Waveguides)
Reporting period: 2022-11-01 to 2024-10-31
The NEXG-UV project addresses the challenge of integrating ultraviolet (UV) light into photonic systems. Integrated photonics, which involves the miniaturization of optical components like lasers, modulators, and detectors onto a single chip, has advanced in visible and infrared wavelengths. However, UV photonics remains underdeveloped due to material limitations. Conventional materials like silicon and silicon nitride (SiN), widely used in infrared and visible photonics, exhibit high absorption at shorter UV wavelengths, making them unsuitable for UV-based applications. This gap in technology limits the potential of UV photonics in fields such as biosensing, quantum computing, and advanced communications.
UV photonics has significant societal implications. UV light is essential for highly sensitive biosensing applications, enabling precise detection of biological molecules for medical diagnostics, and other critical uses. In quantum computing, UV photons are vital for manipulating qubits and advancing computation technologies that promise unprecedented processing power. Therefore, advancing UV photonic technology can lead to breakthroughs in healthcare, computing, and communications infrastructure, driving societal progress in critical areas.
Throughout the project’s timeline, most objectives and goals across the various work packages (WPs) were successfully achieved. The overall objective of the NEXG-UV project was to develop low-loss photonic waveguides for UV light on a miniaturized chip. This required the identification and integration of materials that could operate efficiently in the UV spectrum. The project explored two key strategies to achieve this:
Inverse Damascene Process: This method aimed to fabricate waveguides without direct etching of the material, avoiding sidewall roughness and reducing power losses, which are common in conventional etching processes.
Conventional Direct Etching: Utilizing established CMOS foundry techniques, this approach leveraged mature technologies used for silicon nitride (SiN) and silicon platforms, focusing on direct etching to create waveguides.
A key material identified was Aluminum Oxide (AlOx), a dielectric with a large bandgap and compatibility with CMOS fabrication. AlOx was chosen for its potential to function efficiently in the UV range, making it an attractive candidate for the development of UV photonic devices. By the project's conclusion, ultra-low-loss waveguides at UV wavelength were successfully fabricated in a CMOS foundry, marking significant progress in UV photonics and bringing the technology closer to commercial viability and widespread societal impact.
The following key project objectives and milestones were successfully addressed:
• Successful ALD deposition of high-quality AlOx thin films.
• Optimization of AlOx thin films for low-loss performance.
• Development of an inverse damascene process for AlOx waveguides.
• Exploration of thermal SiO2 etching techniques.
• CMP process development for AlOx waveguide.
• Creation of a PDK for AlOx-based passive elements.
• Completion of wafer-scale processing in imec’s 200mm p-line.
UV photonics has significant societal implications. UV light is essential for highly sensitive biosensing applications, enabling precise detection of biological molecules for medical diagnostics, and other critical uses. In quantum computing, UV photons are vital for manipulating qubits and advancing computation technologies that promise unprecedented processing power. Therefore, advancing UV photonic technology can lead to breakthroughs in healthcare, computing, and communications infrastructure, driving societal progress in critical areas.
Throughout the project’s timeline, most objectives and goals across the various work packages (WPs) were successfully achieved. The overall objective of the NEXG-UV project was to develop low-loss photonic waveguides for UV light on a miniaturized chip. This required the identification and integration of materials that could operate efficiently in the UV spectrum. The project explored two key strategies to achieve this:
Inverse Damascene Process: This method aimed to fabricate waveguides without direct etching of the material, avoiding sidewall roughness and reducing power losses, which are common in conventional etching processes.
Conventional Direct Etching: Utilizing established CMOS foundry techniques, this approach leveraged mature technologies used for silicon nitride (SiN) and silicon platforms, focusing on direct etching to create waveguides.
A key material identified was Aluminum Oxide (AlOx), a dielectric with a large bandgap and compatibility with CMOS fabrication. AlOx was chosen for its potential to function efficiently in the UV range, making it an attractive candidate for the development of UV photonic devices. By the project's conclusion, ultra-low-loss waveguides at UV wavelength were successfully fabricated in a CMOS foundry, marking significant progress in UV photonics and bringing the technology closer to commercial viability and widespread societal impact.
The following key project objectives and milestones were successfully addressed:
• Successful ALD deposition of high-quality AlOx thin films.
• Optimization of AlOx thin films for low-loss performance.
• Development of an inverse damascene process for AlOx waveguides.
• Exploration of thermal SiO2 etching techniques.
• CMP process development for AlOx waveguide.
• Creation of a PDK for AlOx-based passive elements.
• Completion of wafer-scale processing in imec’s 200mm p-line.
The first two WPs required extensive effort in clean room environments and measurement labs, while the third WP involved close collaboration across multiple departments and teams to align with both the NEXG-UV project’s goals and the broader objectives of imec, the host organization.
Work Package 1 (WP1): Material studies
In WP1, a comprehensive study was conducted to characterize AlOx thin films deposited using various ALD tools at imec. This involved examining the optical and physical properties of the films. In some cases, the film quality already met high standards, while in others, the deposition process was optimized to improve material quality. As a result, the optimized films achieved very low losses, measuring less than 1 dB/cm at UV wavelengths.
Work Package 2 (WP2): Fabrication and processing
WP2 focused on fabricating waveguides using the inverse damascene method. This process involved creating a channel through plasma etching, followed by the deposition of AlOx and planarization using Chemical Mechanical Polishing (CMP). The results were compared with conventional etching of AlOx films. WP2 began early in parallel with WP1 and continued nearly until the project's conclusion. The fabrication was successful, resulting in buried AlOx waveguides embedded in SiO2. However, the attenuation of the waveguides was higher than anticipated, likely due to issues in the planarization or deposition steps. Despite multiple trials and improvements, further optimization is needed to achieve the desired reduction in waveguide loss.
Work Package 3 (WP3): PDK and transfer to pilot line
The goal of WP3 was fully achieved, even though it was initially considered ambitious and not guaranteed to be completed, as outlined in the project proposal. A library of photonic elements (PDK) was successfully designed and fabricated using imec’s 200mm p-line. This resulted in AlOx waveguides with exceptionally low losses. Achieving this outcome required significant cross-collaboration among internal imec teams, which demanded additional resources beyond the scope of the MSCA project. The progress made in WP2 directly supported WP3 by providing insights into the performance of AlOx waveguides for UV light. The outcomes of WP3 open up the potential for commercialization, particularly in offering foundry services to external partners in fields like quantum computing and life sciences.
These results are a major step forward in the field of UV photonics, making the technology more commercially viable and scalable. The successful fabrication of waveguides in a CMOS foundry demonstrated the project’s ability to integrate UV photonics into existing semiconductor infrastructure, facilitating future development and potential applications in biosensing, quantum computing, and advanced communications.
The project’s results have been disseminated through various academic and industry platforms. Key publications include conference proceedings at the SSDM conference (September 2024, Japan), with an additional presentation scheduled at Photonics West 2025 (USA). Public outreach was conducted through channels like LinkedIn, as well as at events such as PIC Summit Europe (Eindhoven 2023) and the Frequency Comb Workshop (Ghent 2024). These dissemination efforts have helped raise awareness of the project’s breakthroughs and fostered engagement with the broader photonics and semiconductor communities.
Project management & training activities:
Alongside the research activities, the MSCA fellow participated in various training programs aimed at gaining experience with innovative simulation tools, enhancing project management skills, and fostering networking opportunities with a diverse range of partners from both industry and academia. The project's broad scope, which encompasses multiple layers of abstraction, significantly expanded the researcher’s areas of interest and expertise.
Work Package 1 (WP1): Material studies
In WP1, a comprehensive study was conducted to characterize AlOx thin films deposited using various ALD tools at imec. This involved examining the optical and physical properties of the films. In some cases, the film quality already met high standards, while in others, the deposition process was optimized to improve material quality. As a result, the optimized films achieved very low losses, measuring less than 1 dB/cm at UV wavelengths.
Work Package 2 (WP2): Fabrication and processing
WP2 focused on fabricating waveguides using the inverse damascene method. This process involved creating a channel through plasma etching, followed by the deposition of AlOx and planarization using Chemical Mechanical Polishing (CMP). The results were compared with conventional etching of AlOx films. WP2 began early in parallel with WP1 and continued nearly until the project's conclusion. The fabrication was successful, resulting in buried AlOx waveguides embedded in SiO2. However, the attenuation of the waveguides was higher than anticipated, likely due to issues in the planarization or deposition steps. Despite multiple trials and improvements, further optimization is needed to achieve the desired reduction in waveguide loss.
Work Package 3 (WP3): PDK and transfer to pilot line
The goal of WP3 was fully achieved, even though it was initially considered ambitious and not guaranteed to be completed, as outlined in the project proposal. A library of photonic elements (PDK) was successfully designed and fabricated using imec’s 200mm p-line. This resulted in AlOx waveguides with exceptionally low losses. Achieving this outcome required significant cross-collaboration among internal imec teams, which demanded additional resources beyond the scope of the MSCA project. The progress made in WP2 directly supported WP3 by providing insights into the performance of AlOx waveguides for UV light. The outcomes of WP3 open up the potential for commercialization, particularly in offering foundry services to external partners in fields like quantum computing and life sciences.
These results are a major step forward in the field of UV photonics, making the technology more commercially viable and scalable. The successful fabrication of waveguides in a CMOS foundry demonstrated the project’s ability to integrate UV photonics into existing semiconductor infrastructure, facilitating future development and potential applications in biosensing, quantum computing, and advanced communications.
The project’s results have been disseminated through various academic and industry platforms. Key publications include conference proceedings at the SSDM conference (September 2024, Japan), with an additional presentation scheduled at Photonics West 2025 (USA). Public outreach was conducted through channels like LinkedIn, as well as at events such as PIC Summit Europe (Eindhoven 2023) and the Frequency Comb Workshop (Ghent 2024). These dissemination efforts have helped raise awareness of the project’s breakthroughs and fostered engagement with the broader photonics and semiconductor communities.
Project management & training activities:
Alongside the research activities, the MSCA fellow participated in various training programs aimed at gaining experience with innovative simulation tools, enhancing project management skills, and fostering networking opportunities with a diverse range of partners from both industry and academia. The project's broad scope, which encompasses multiple layers of abstraction, significantly expanded the researcher’s areas of interest and expertise.
There are several main outcomes of the project:
• ALD-deposited AlOx thin films were optimized for photonic applications, resulting in the production of films with significantly lower internal losses.
• The project successfully employed the inverse damascene method for waveguide fabrication, which avoids the direct etching of the thin film. This approach is expected to produce smoother sidewalls, thereby reducing scattering losses in the waveguides.
• A library of passive photonic elements was designed and fabricated in a CMOS foundry, facilitating large-scale production at the 200mm wafer level. This development holds substantial potential for commercialization, particularly in providing foundry services to external partners in areas such as quantum computing and life sciences.
• ALD-deposited AlOx thin films were optimized for photonic applications, resulting in the production of films with significantly lower internal losses.
• The project successfully employed the inverse damascene method for waveguide fabrication, which avoids the direct etching of the thin film. This approach is expected to produce smoother sidewalls, thereby reducing scattering losses in the waveguides.
• A library of passive photonic elements was designed and fabricated in a CMOS foundry, facilitating large-scale production at the 200mm wafer level. This development holds substantial potential for commercialization, particularly in providing foundry services to external partners in areas such as quantum computing and life sciences.