Periodic Reporting for period 2 - POLYNICES (POLYmer based electro-optic PCB motherboard integration with Si3N4 Chiplets, InP Components and Electronic ICs enabling affordable photonic modules for THz Sensing and quantum computing applications)
Reporting period: 2024-07-01 to 2025-06-30
POLYNICES goal is to design and produce a low cost, scalable, high-performing electro-optic PCB platform that will drive the development of the next generation photonic modules. POLYNICES will heterogeneously integrate FhG-HHI’s PolyBoard hybrid photonic integration platform on PCBs, to realize a low-cost Electro-Optic PCB (EOPCB) motherboard with low-loss single mode waveguides and etched recesses to host chiplets from other photonic integration technologies such as silicon nitride and InP, as well as high-frequency elements such as rod-antennas.
Over Periods 1 and 2 (up to M30), the project focused on meeting its core scientific and technical objectives, with substantial progress.
In terms of system architecture and materials (Objectives 1 & 2), the consortium finalized the high-level functional layouts of the POLYNICES prototypes, supported by extensive simulation activities. FR-4 was confirmed as the EO-PCB substrate after a detailed comparative evaluation with Rogers 6002.
Progress in PIC development (Objective 3) was equally significant. The TriPleX-based Si₃N₄ platform advanced with Precursor-0A laser cavities fabricated and under test, while thermo-optically actuated cavity lasers and optical beamforming networks entered fabrication. For the quantum information processing demonstrators, 780 nm and 1550 nm laser submodules were designed, fabricated, and tested, complemented by the development of 8×8 Clements matrices now in production.
In the area of actuation technologies (Objective 4), PZT stress-optic actuators demonstrated stable operation with modulation bandwidths around 1 MHz and reliable performance at reduced operating voltages. However, given the different cross-section requirements and the greater maturity of thermo-optical phase shifters, the consortium decided to adopt thermo-optical actuators for the main demonstrators to reduce integration risks and improve reliability.
The work on antenna development (Objective 5) led to the design, simulation, and validation of both single and arrayed dielectric rod antennas on FR-4.
Concerning integration processes (Objective 6), major achievements include optimization of flip-chip techniques such as solder gap control (≤20 µm), Au stud bumping, and bonding methods.
Finally, in electronics co-packaging (Objective 7), significant progress was made in the development of driver and read-out electronics, including compact multi-channel PZT drivers with MHz-level bandwidths, providing scalability and performance for the upcoming demonstrators.
In terms of system architecture and materials (Objectives 1 & 2), the consortium finalized the high-level functional layouts of the POLYNICES prototypes, supported by extensive simulation activities. FR-4 was confirmed as the EO-PCB substrate after a detailed comparative evaluation with Rogers 6002.
Progress in PIC development (Objective 3) was equally significant. The TriPleX-based Si₃N₄ platform advanced with Precursor-0A laser cavities fabricated and under test, while thermo-optically actuated cavity lasers and optical beamforming networks entered fabrication. For the quantum information processing demonstrators, 780 nm and 1550 nm laser submodules were designed, fabricated, and tested, complemented by the development of 8×8 Clements matrices now in production.
In the area of actuation technologies (Objective 4), PZT stress-optic actuators demonstrated stable operation with modulation bandwidths around 1 MHz and reliable performance at reduced operating voltages. However, given the different cross-section requirements and the greater maturity of thermo-optical phase shifters, the consortium decided to adopt thermo-optical actuators for the main demonstrators to reduce integration risks and improve reliability.
The work on antenna development (Objective 5) led to the design, simulation, and validation of both single and arrayed dielectric rod antennas on FR-4.
Concerning integration processes (Objective 6), major achievements include optimization of flip-chip techniques such as solder gap control (≤20 µm), Au stud bumping, and bonding methods.
Finally, in electronics co-packaging (Objective 7), significant progress was made in the development of driver and read-out electronics, including compact multi-channel PZT drivers with MHz-level bandwidths, providing scalability and performance for the upcoming demonstrators.
Results and Potential Impacts:
By M30, the project has delivered a coherent set of technical and scientific results that lay the foundation for the FMCW THz spectrometer and QIP demonstrators. System-level activities established and refined the use cases, system requirements, and prototype specifications, ensuring that all updated functional layouts comply with the performance targets defined for their application domains. The system design was successfully translated into component-level specifications, and the packaging and assembly process flows were refined to minimize fabrication risks while supporting reliable integration.
On the materials and integration side, FR-4 was selected as the baseline PCB substrate for the EO-PCB motherboards, following comparative analysis with Rogers 6002. The planarization process was optimized to achieve sub-micron reproducibility (<1 µm deviation), enabling optical coupling with <1 dB loss. PolyBoard waveguides fabricated on FR-4 substrates achieved propagation losses around 1 dB/cm for both TE and TM polarizations, a significant improvement over the first iteration. Progress was also made in etching recesses and grooves for flip-chipping, with first results expected in the following period. Wafer runs for both precursor and final demonstrators have been scheduled, ensuring continuity toward system-level integration.
In photonics development, high-performance external cavity lasers (ECLs) at 780 nm and 1550 nm were designed, fabricated, and partially tested, with integration of gain sections and submodules ongoing. The 1550 nm laser submodules designed in collaboration with Tera-6G broadened the impact of the work and ensure alignment with emerging applications. Calibration methodologies were developed for both Blass matrix–based beamforming networks (for FMCW systems) and Clements matrix architectures (for QIP systems). Blass and Clements matrix chiplets were fabricated, and preliminary squeezer chiplets for quantum applications were also produced.
For actuation technologies, PZT phase actuators demonstrated stable and linear operation in the 10–30 V range, with modulation speeds around 1 MHz and improved electrical stability at reduced voltages. Although thermo-optical shifters were adopted for the main demonstrators to reduce risks and shorten fabrication cycles, the PZT results pave the way for future low-power, high-speed applications.
Work on antennas and THz integration yielded the design and simulation of single and arrayed dielectric rod antennas, with a 1×4 THz array showing promising gain and beam steering capabilities. Integration concepts for antennas with the EO-PCB and InP devices were refined, and deliverable D2.5 documented the updated design of the THz antenna array. InP-based THz emitters and receivers were fabricated according to new flip-chip design rules, with initial wafer-level characterization confirming successful performance.
The flip-chip and soldering processes advanced through extensive experimental studies of materials, alignment, and bonding. Au stud bumps, conductive epoxy, and dispensing techniques were validated, contributing to the final definition of the POLYNICES cross-section. Passive alignment guidelines were shared with partners, and initial packaging concepts for both precursor and final demonstrators were defined.
System-level progress was visible in the assembly and packaging of demonstrators. Precursor-0A was fabricated, packaged, and characterized. Testing methodologies and KPIs for FMCW THz and QIP demonstrators were defined, and early experiments validated theoretical analyses, including the first noise analysis of FMCW THz systems. Driving and readout electronics, including current sources, single-channel and 10-channel PZT drivers, and mode-hop–free laser sweeping algorithms, were developed and tested.
Potential Impacts:
The project’s results demonstrate the feasibility of an integrated EO-PCB-based approach for next-generation THz spectrometry and quantum information processing. The combination of low-loss PolyBoard waveguides, validated FR-4 integration, novel antenna designs, compact drivers, and maturing PIC technologies provides a strong technological platform with potential to address pressing needs in high-resolution sensing, secure communications, and scalable quantum photonics.
In the medium term, the adoption of wafer-scale integration processes and refined packaging flows de-risks industrial scalability. The methodologies for system calibration (Blass and Clements architectures) directly support reproducibility and standardization of system-level performance, while the progress on multi-channel drivers and modular electronics ensures compatibility with diverse application environments.
By M30, the project has delivered a coherent set of technical and scientific results that lay the foundation for the FMCW THz spectrometer and QIP demonstrators. System-level activities established and refined the use cases, system requirements, and prototype specifications, ensuring that all updated functional layouts comply with the performance targets defined for their application domains. The system design was successfully translated into component-level specifications, and the packaging and assembly process flows were refined to minimize fabrication risks while supporting reliable integration.
On the materials and integration side, FR-4 was selected as the baseline PCB substrate for the EO-PCB motherboards, following comparative analysis with Rogers 6002. The planarization process was optimized to achieve sub-micron reproducibility (<1 µm deviation), enabling optical coupling with <1 dB loss. PolyBoard waveguides fabricated on FR-4 substrates achieved propagation losses around 1 dB/cm for both TE and TM polarizations, a significant improvement over the first iteration. Progress was also made in etching recesses and grooves for flip-chipping, with first results expected in the following period. Wafer runs for both precursor and final demonstrators have been scheduled, ensuring continuity toward system-level integration.
In photonics development, high-performance external cavity lasers (ECLs) at 780 nm and 1550 nm were designed, fabricated, and partially tested, with integration of gain sections and submodules ongoing. The 1550 nm laser submodules designed in collaboration with Tera-6G broadened the impact of the work and ensure alignment with emerging applications. Calibration methodologies were developed for both Blass matrix–based beamforming networks (for FMCW systems) and Clements matrix architectures (for QIP systems). Blass and Clements matrix chiplets were fabricated, and preliminary squeezer chiplets for quantum applications were also produced.
For actuation technologies, PZT phase actuators demonstrated stable and linear operation in the 10–30 V range, with modulation speeds around 1 MHz and improved electrical stability at reduced voltages. Although thermo-optical shifters were adopted for the main demonstrators to reduce risks and shorten fabrication cycles, the PZT results pave the way for future low-power, high-speed applications.
Work on antennas and THz integration yielded the design and simulation of single and arrayed dielectric rod antennas, with a 1×4 THz array showing promising gain and beam steering capabilities. Integration concepts for antennas with the EO-PCB and InP devices were refined, and deliverable D2.5 documented the updated design of the THz antenna array. InP-based THz emitters and receivers were fabricated according to new flip-chip design rules, with initial wafer-level characterization confirming successful performance.
The flip-chip and soldering processes advanced through extensive experimental studies of materials, alignment, and bonding. Au stud bumps, conductive epoxy, and dispensing techniques were validated, contributing to the final definition of the POLYNICES cross-section. Passive alignment guidelines were shared with partners, and initial packaging concepts for both precursor and final demonstrators were defined.
System-level progress was visible in the assembly and packaging of demonstrators. Precursor-0A was fabricated, packaged, and characterized. Testing methodologies and KPIs for FMCW THz and QIP demonstrators were defined, and early experiments validated theoretical analyses, including the first noise analysis of FMCW THz systems. Driving and readout electronics, including current sources, single-channel and 10-channel PZT drivers, and mode-hop–free laser sweeping algorithms, were developed and tested.
Potential Impacts:
The project’s results demonstrate the feasibility of an integrated EO-PCB-based approach for next-generation THz spectrometry and quantum information processing. The combination of low-loss PolyBoard waveguides, validated FR-4 integration, novel antenna designs, compact drivers, and maturing PIC technologies provides a strong technological platform with potential to address pressing needs in high-resolution sensing, secure communications, and scalable quantum photonics.
In the medium term, the adoption of wafer-scale integration processes and refined packaging flows de-risks industrial scalability. The methodologies for system calibration (Blass and Clements architectures) directly support reproducibility and standardization of system-level performance, while the progress on multi-channel drivers and modular electronics ensures compatibility with diverse application environments.