Periodic Reporting for period 1 - HDLN (High-Density Lithium Niobate Photonic Integrated Circuits)
Berichtszeitraum: 2023-05-01 bis 2024-04-30
Electro-optical (EO) modulators are crucial components in optical systems, translating high-speed electrical signals into the optical domain. They play a key role in various applications, including optical telecommunications, optical metrology and sensing, microwave photonics Terahertz (THz) signal processing, and quantum information processing.
Thin Film Lithium Niobate (TFLN) is an emerging material platform for photonic integrated circuits (PICs), with a significant impact on the next and beyond next-generation telecom and datacom industries. Lithium Niobate (LN) is a material that outperforms in many key metrics established platforms such as Silicon or Indium Phosphide. Some of these key metrics are its ultrafast modulation bandwidth (beyond 150 GHz), wide transparency window (from NIR to visible), no nonlinear absorption, tolerance to high power, low insertion loss, and most importantly, very efficient EO modulation. This makes it the ideal platform for building fast modulators.
Despite its potential, LN has not yet reached widespread adoption due to challenges in fabrication technology. LN is notoriously difficult to etch, and high-quality waveguide manufacturing has been demonstrated by only a few research groups, primarily using electron beam lithography methods. Our competitive advantage lies in a proprietary new etching technology for creating high-density, tightly integrated LN PICs. This patented process employs diamond-like carbon (DLC) to etch low-loss, straight sidewall, tightly confining LN integrated photonic circuits and is compatible with wafer-scale lithography.
The HDLN project aims to develop a commercial foundry service for TFLN, offering high production volumes, low optical losses, and high component density. We will establish a generic fabrication process for LN, together with a corresponding design manual and a qualified component library, which will be compiled into a process design kit (PDK). This PDK will be implemented in professional electronic photonic design automation (EPDA) software such as Synopsys, as well as in open-source alternatives like gdsfactory, featuring advanced capabilities like circuit-level simulation.
LN is applicable to many fields, from mature markets to emerging applications. Given its excellent properties for building modulators, its primary applications are in next-generation optical interconnects for telecom and data centers, where TFLN adoption is already strongly trending in various markets. Beyond mature communications markets, TFLN is being actively explored for emerging applications such as coherent LiDAR, photonic computing, AI accelerators, and photonic quantum computing.
By the end of the HDLN project, we will offer multi-project wafer runs to customers from industry and academia. We will also demonstrate the platform's viability through high-impact application demonstrations, such as ultra-fast optical telecom/datacom transceivers and THz- and mm-wave photonics for ultra-broadband signal processing and 6G networks.
Thin Film Lithium Niobate (TFLN) is an emerging material platform for photonic integrated circuits (PICs), with a significant impact on the next and beyond next-generation telecom and datacom industries. Lithium Niobate (LN) is a material that outperforms in many key metrics established platforms such as Silicon or Indium Phosphide. Some of these key metrics are its ultrafast modulation bandwidth (beyond 150 GHz), wide transparency window (from NIR to visible), no nonlinear absorption, tolerance to high power, low insertion loss, and most importantly, very efficient EO modulation. This makes it the ideal platform for building fast modulators.
Despite its potential, LN has not yet reached widespread adoption due to challenges in fabrication technology. LN is notoriously difficult to etch, and high-quality waveguide manufacturing has been demonstrated by only a few research groups, primarily using electron beam lithography methods. Our competitive advantage lies in a proprietary new etching technology for creating high-density, tightly integrated LN PICs. This patented process employs diamond-like carbon (DLC) to etch low-loss, straight sidewall, tightly confining LN integrated photonic circuits and is compatible with wafer-scale lithography.
The HDLN project aims to develop a commercial foundry service for TFLN, offering high production volumes, low optical losses, and high component density. We will establish a generic fabrication process for LN, together with a corresponding design manual and a qualified component library, which will be compiled into a process design kit (PDK). This PDK will be implemented in professional electronic photonic design automation (EPDA) software such as Synopsys, as well as in open-source alternatives like gdsfactory, featuring advanced capabilities like circuit-level simulation.
LN is applicable to many fields, from mature markets to emerging applications. Given its excellent properties for building modulators, its primary applications are in next-generation optical interconnects for telecom and data centers, where TFLN adoption is already strongly trending in various markets. Beyond mature communications markets, TFLN is being actively explored for emerging applications such as coherent LiDAR, photonic computing, AI accelerators, and photonic quantum computing.
By the end of the HDLN project, we will offer multi-project wafer runs to customers from industry and academia. We will also demonstrate the platform's viability through high-impact application demonstrations, such as ultra-fast optical telecom/datacom transceivers and THz- and mm-wave photonics for ultra-broadband signal processing and 6G networks.
Since the project's launch, the HDLN partners have focused on three key activities. First, we have developed and optimized the nanofabrication processes for high-density lithium niobate photonic integrated circuits. This effort has led to the demonstration of ultra-low-loss channel waveguides using a diamond-like-carbon (DLC) hard mask and a hydrogen-free cladding method. Additionally, we have developed and validated low-loss traveling wave electrodes on high-resistivity substrates using the copper damascene technique.
In parallel, we have designed the main building blocks (BB), including both passive and active components, to build different modulator architectures tailored specifically for this application.
Finally, the fabrication process was transferred from EPFL to LXT, where it was optimized for reproducibility and yield, and subsequently applied in the first engineering run.
Furthermore, the collaborative efforts of the partners resulted in the following achievements:
• Fabrication of ultra-low loss channel waveguides using a novel DLC hard mask has been successfully demonstrated.
• The copper damascene technique was developed and validated on several coplanar waveguide (CPW) electrodes.
• Active and passive BBs have been designed and the numerical results obtained are comparable to the defined performance.
• Reliability assessment: the first engineering run was carried out in the last quarter of 2023. The results of the run provided a good level of confidence in the technology validation (optical insertion loss < 10dB/m). As a result, the fabrication process has passed the initial performance check and is ready for PDK validation.
In parallel, we have designed the main building blocks (BB), including both passive and active components, to build different modulator architectures tailored specifically for this application.
Finally, the fabrication process was transferred from EPFL to LXT, where it was optimized for reproducibility and yield, and subsequently applied in the first engineering run.
Furthermore, the collaborative efforts of the partners resulted in the following achievements:
• Fabrication of ultra-low loss channel waveguides using a novel DLC hard mask has been successfully demonstrated.
• The copper damascene technique was developed and validated on several coplanar waveguide (CPW) electrodes.
• Active and passive BBs have been designed and the numerical results obtained are comparable to the defined performance.
• Reliability assessment: the first engineering run was carried out in the last quarter of 2023. The results of the run provided a good level of confidence in the technology validation (optical insertion loss < 10dB/m). As a result, the fabrication process has passed the initial performance check and is ready for PDK validation.
The fabrication of ultra-low loss channel waveguides using a novel DLC hard mask has been successfully demonstrated. The developed and optimized low-loss silicon tetrachloride cladding method has shown excellent compatibility with lithium niobate photonic integrated circuits, with no observable -OH absorption induced optical loss. The transfer of the fabrication process to LXT will allow the HDLN partners to demonstrate the viability of the technology platform in two highly relevant communication application fields: Ultra-fast optical telecom/datacom transceivers and THz- and mm-wave photonics.
In addition, the HDLN partners have created a library of critical optical components for building different HDLN TFLN modulator architectures, including the corresponding fabrication steps, marking the initial phase of supporting the fabrication process with a robust PDK. The first version of the PDK includes all the key building blocks, process layers, and verification rules expected of an industrial-grade PDK. This PDK paves the way for the commercialization of HDLN’s open-access manufacturing foundry offering for LNOI integrated photonic circuits, which is a key objective of the HDLN project.
In addition, the HDLN partners have created a library of critical optical components for building different HDLN TFLN modulator architectures, including the corresponding fabrication steps, marking the initial phase of supporting the fabrication process with a robust PDK. The first version of the PDK includes all the key building blocks, process layers, and verification rules expected of an industrial-grade PDK. This PDK paves the way for the commercialization of HDLN’s open-access manufacturing foundry offering for LNOI integrated photonic circuits, which is a key objective of the HDLN project.