SWIR SPADs on Silicon Platforms: Heterostructure Integration, Performance Benchmarking, and Future Perspectives
Chips · 19.11.2026 · 5 (4) · 42
DOI: 10.3390/chips5040042 ↗Kurz gefasst
Fachartikel in der Zeitschrift Chips, online erschienen am 19.11.2026. Verfasst von Linpeng Gu und 6 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Leading innovation and entrepreneurship Project of Zhejiang Province, Hangzhou Leading Talent Program.
Abstract (Originalsprache)
Short-wave infrared (SWIR) single-photon avalanche diodes (SPADs) are essential for quantum communication, eye-safe light detection and ranging (LiDAR), and low-light biomedical imaging. Silicon-based SWIR SPADs combine the low-noise avalanche multiplication of Si with SWIR-absorbing materials to extend detection beyond Si’s intrinsic absorption edge. This review examines Si-based SWIR SPAD technologies from a heterostructure integration perspective, distinguishing three fundamentally different integration objectives: direct formation of III–V/Si avalanche heterostructures, transfer of prefabricated III-V SPADs onto Si photonic platforms, and direct Ge-on-Si epitaxy. For III–V-on-Si integration, we analyze direct epitaxy, bonding (including direct bonding and interlayer-mediated bonding), and micro-transfer printing, assessing their trade-offs among material quality, process complexity, thermal budget, and device performance. For Ge-on-Si integration, we trace the evolution from cryogenic mesa devices to room-temperature, waveguide-coupled, and CMOS-compatible arrays, highlighting the role of electric-field engineering in suppressing dark counts. A cross-platform comparison against application-specific requirements reveals that Ge-on-Si is best positioned for cost-sensitive, array-scale applications at 1310 nm, while III–V-on-Si bonding currently offers the highest reported 1550 nm photon detection efficiency (PDE) among the Si-based SPADs reviewed here, making it a potential device-level platform for performance-driven applications, provided that stringent DCR and afterpulsing requirements can be met in future system-level development. The two platforms are complementary rather than competing. Across all routes, the three shared challenges are defect-mediated dark noise, electric-field coupling to defects, and the thermal-budget–performance trade-off. Future progress will require coordinated advances in defect passivation, electric-field management, low-thermal-budget processing, and scalable assembly, ultimately enabling heterogeneous photonic systems that integrate multiple material platforms.
Förderung
- Leading innovation and entrepreneurship Project of Zhejiang Province
- Hangzhou Leading Talent Program
Zitieren
APA
Gu, L., Wen, X., Tang, Z., Fang, C., Liu, Y., Han, G., & Zhang, J. (2026). SWIR SPADs on Silicon Platforms: Heterostructure Integration, Performance Benchmarking, and Future Perspectives. Chips, 5(4), 42. https://doi.org/10.3390/chips5040042
Vancouver
Gu L, Wen X, Tang Z, Fang C, Liu Y, Han G, et al. SWIR SPADs on Silicon Platforms: Heterostructure Integration, Performance Benchmarking, and Future Perspectives. Chips. 2026;5(4):42. doi:10.3390/chips5040042
BibTeX
@article{Gu2026,
author = {Gu, Linpeng and Wen, Xinrong and Tang, Zhenrong and Fang, Cizhe and Liu, Yan and Han, Genquan and Zhang, Jishen},
title = {SWIR SPADs on Silicon Platforms: Heterostructure Integration, Performance Benchmarking, and Future Perspectives},
journal = {Chips},
year = {2026},
volume = {5},
number = {4},
pages = {42},
doi = {10.3390/chips5040042}
}
EUVIMED