InP Photonic Integration Technology Landscape 2026
InP Photonic Integration Technology Landscape 2026
Indium Phosphide photonic integration is at an inflection point as IMOS, heterogeneous InP/Si bonding, and THz electronic-photonic convergence converge. Filings from 2019–2026 represent approximately 60% of the relevant dataset.
Why InP Photonic Integration Matters in 2026
InP photonic integration encompasses monolithic and heterogeneous fabrication of active and passive optical components—including lasers, modulators, photodetectors, and amplifiers—on InP substrates or hybrid III-V/Si platforms. The field addresses high-speed optoelectronic circuits for telecommunications, sensing, and quantum photonics. The direct bandgap of InP enables efficient light emission and amplification, distinguishing it from silicon photonics.
InP-based materials including InGaAsP and InAlAs alloys support lasing, electro-absorption modulation, and photodetection in the 1.3–1.55 µm telecom window. The generic foundry model pioneered by COBRA at TU Eindhoven reduces R&D cost and time-to-market by more than an order of magnitude compared to custom processes, enabling SMEs to prototype complex PICs without internal process development.
In this dataset, the 2019–2026 window accounts for approximately 60% of relevant entries, with a clear clustering around heterogeneous integration architectures. Three distinct InP/Si integration approaches—wafer bonding, direct heteroepitaxy, and InP membrane transfer—each show credible 2021–2022 performance data, though no single approach has demonstrated clear manufacturing scalability to 300 mm wafers.
China is the dominant recent patent-filing jurisdiction in this dataset, with 10+ relevant Chinese-jurisdiction filings from 2019 to 2026. State-affiliated entities such as CETC 55th Research Institute and commercial players including Hisense Broadband Multimedia Technology and China Resources Microelectronics account for the majority of 2023–2026 filings. Western IP from AT&T, Agilent, and Avago is represented only by older, now-inactive filings.
Filing Trends and Technology Cluster Distribution
In this dataset, InP photonic integration activity spans 1976 to 2026, with the 2019–2026 period accounting for approximately 60% of relevant entries. Four principal technology clusters—native InP foundry, IMOS, monolithic InP/SOI, and electronic-photonic convergence—show distinct performance milestones and geographic concentrations.
InP Technology Cluster Distribution by Record Count
The IMOS and heterogeneous InP/Si clusters account for the largest share of 2019–2026 records, reflecting a shift from native foundry integration toward silicon-compatible architectures.
↗ Click bars to exploreInP Patent Filing Activity by Era (Dataset Records)
Filing and publication activity in this dataset accelerated sharply from 2019 onward, with 2024–2026 already contributing multiple emerging-direction records including SPAD, VCSEL, and diamond substrate integration.
↗ Click bars to exploreKey InP Photonic Integration Application Areas and Research Sites
InP photonic integration is deployed across optical interconnects, THz/beyond-5G electronics, LiDAR and single-photon detection, quantum photonics, and near-infrared sensing, with specific institutions and companies driving each domain.
TU Eindhoven / COBRA Generic Foundry
The COBRA group at TU Eindhoven pioneered the open-access generic InP foundry model, described in the 2014 literature review as reducing R&D cost and time-to-market by more than an order of magnitude. The platform supports DFB lasers, electro-absorption modulators, SOAs, and MMI couplers on standardized epitaxial stacks. Commercialized via JePPIX, Smart Photonics, and EFFECT Photonics, this approach enables SME access to complex PIC prototyping.
Generic InP FoundryInP Membrane on Silicon (IMOS) Platform
The 2019 paper “Indium Phosphide Membrane Nanophotonic Integrated Circuits on Silicon” demonstrated high SMSR lasers and ultrafast photodiodes in the IMOS architecture, with applications in quantum photonics and optical cross-connect. The 2020 InP membrane integrated photonics review reported that sub-micron InP membrane on silicon breaks speed, energy, and density bottlenecks. The 2021 epitaxial growth study achieved directly bonded III-V membrane on Si enabling regrowth without dislocation formation.
Heterogeneous IntegrationChina Resources Micro. SPAD Arrays
China Resources Microelectronics (Chongqing) Co., Ltd. filed planar InP-based SPAD patents in CN (2023), EP (2024), and US (2024) featuring isolation rings that suppress tunneling-induced dark counts, targeting aerospace communication and nuclear power fields. The dual-jurisdiction EP and US filings reflect an international prosecution strategy for radiation-hard photon-counting applications in LiDAR and space-grade sensors.
Single-Photon DetectionCETC 55th Institute THz Electronics
CETC 55th Research Institute filed the InP E/D multi-function chip (2023, CN) targeting THz-band monolithic integration of enhancement-mode and depletion-mode InP HEMTs for ultra-high-frequency MMIC circuits. The 2021 literature paper “Towards Monolithic InP-Based Electronic Photonic Technologies for beyond 5G” developed SPICE-compatible models of UTC-PDs and InP DHBTs for THz OEICs. CETC also filed LiNbO₃-InP optoelectronic integrated device patents on diamond substrates in 2025 and 2026 for thermal management.
THz Electronic-PhotonicFour Emergent InP Integration Directions (2024–2026)
Based on records filed or published in 2024–2026 in this dataset, four emergent directions are identifiable: LiNbO₃–InP on diamond substrates, InP VCSEL with 2D material integration, high-bandwidth hybrid InP/Si chips, and radiation-hard InP SPAD arrays.
LiNbO₃–InP Heterogeneous Integration on Diamond
CETC 55th Research Institute filed two patents in late 2025 and early 2026 combining LiNbO₃ electro-optic modulators with InP active devices on diamond substrates. This configuration directly addresses the heat dissipation bottleneck in high-integration-density photonic systems. Diamond substrates provide superior thermal management compared to InP bulk or SOI, making this architecture relevant for high-power, high-density integration scenarios.
High-Bandwidth Hybrid InP/Si Chips for Optical Modules
Hisense Broadband Multimedia Technology filed two CN patents in 2025 targeting ultra-heavily doped P-type InGaAs contact layers at doping concentrations of ≥2×10²⁰ cm⁻³ on bonded InP/Si platforms. This approach minimizes contact resistance and maximizes bandwidth for high-rate optical modules. The patents include RF traveling-wave electrode structures for high-speed optical data transmission targeting data center applications.
InP Integration Approach Comparison: Native Foundry vs. Heterogeneous InP/Si
Click any row to explore further.
| Dimension | Native InP Generic Foundry | Heterogeneous InP/Si (IMOS / Bonding) |
|---|---|---|
| Dimension | InP bulk substrate | Silicon or SOI wafer with InP membrane or bonded die |
| Key Wavelength Range | 1.3–1.55 µm telecom window | 1.2–1.65 µm (InP/SOI platform, 2021 data) |
| Photodetector Bandwidth | 28 Gb/s balanced APD (foundry PIC, 2019) | >40 GHz, 40 Gb/s on monolithic InP/SOI (2021) |
| Wafer Scale Compatibility | InP wafer scale; limited 300 mm scalability demonstrated | No single approach has demonstrated 300 mm scalability in this dataset |
| Integration Density | Up to 650 optical and electrical components on single chip (quantum PIC roadmap, 2022) | Sub-micron waveguide confinement enabling high density; specific component counts not reported in this dataset |
| Key Assignees (Dataset) | TU Eindhoven/COBRA, AT&T, American Telephone & Telegraph Company | Hisense Broadband, CETC 55th Research Institute, Agilent/Avago (buffer layer IP) |
| Thermal Management Approach | Standard InP substrate; thermal bottleneck at high density | Diamond substrate integration (CETC, 2025–2026) for superior heat dissipation |
| CMOS Compatibility”> | Limited; InP fabrication lines separate from Si CMOS | Targets convergence with CMOS infrastructure via Si substrate bonding or epitaxy |
Frequently Asked Questions: InP Photonic Integration Technology 2026
InP has a direct bandgap, enabling efficient light emission and amplification, which silicon cannot provide. InP-based materials including InGaAsP and InAlAs alloys support lasing, electro-absorption modulation, and photodetection in the 1.3–1.55 µm telecom window. Silicon photonics relies on indirect bandgap silicon and requires external III-V light sources.
InP Membrane on Silicon (IMOS) uses a sub-micron InP membrane bonded or grown on a silicon substrate for high-index-contrast nanophotonic waveguiding alongside active III-V functions. The 2019 paper demonstrated high SMSR lasers and ultrafast photodiodes in IMOS with quantum photonics and optical cross-connect applications. The 2020 review reported IMOS breaks speed, energy, and density bottlenecks versus conventional InP platforms.
In this dataset, CETC 55th Research Institute (4 filings including THz HEMTs and LiNbO₃-InP on diamond), China Resources Microelectronics (3 filings for planar InP SPAD across CN, EP, and US), and Hisense Broadband Multimedia Technology (2 filings for hybrid InP/Si chips) are the most active recent filers from 2023 to 2026.
According to the 2021 literature records in this dataset, high-performance III-V photodetectors on a monolithic InP/SOI platform achieved greater than 40 GHz bandwidth, 40 Gb/s operation, 0.55 nA dark current, and an operation range of 1240–1650 nm. The platform used dislocation-free selective growth of InP sub-micron wires on (001) SOI directly adjacent to the Si device layer.
Multiple 2025–2026 filings in this dataset explicitly address heat dissipation in high-density InP integration. CETC 55th Research Institute filed two patents combining LiNbO₃ electro-optic modulators with InP active devices on diamond substrates specifically to address the heat dissipation bottleneck. Diamond substrates provide superior thermal management, and IP around thermal interface materials and substrate choices is identified as increasingly strategic as integration density scales.
According to the dataset, no single InP/Si integration approach—wafer bonding, direct heteroepitaxy, or InP membrane transfer—has demonstrated clear manufacturing scalability to 300 mm wafers as of the records reviewed. The 2021 AlGaInAs laser paper targeted low-cost 300 mm Si wafer laser integration via InP-seed bonding and regrowth, but the dataset does not report a validated 300 mm process across all platform types.