SBIR and STTR topics › DoW/NAVY
NAVY
Pre-release
SBIRITARCMMC Level 2 (Self)
Low-cost Mid-wave Infrared Focal-Plane Array (FPA) Fabrication through Direct-bond Interconnection
Topic # DON27BZ02-NV011 · DoW SBIR 2027 BAA · DoW/NAVY · released 2026-11-04
Opens
Dec 2, 2026
Opens in 14 dayscloses Jan 6, 2027
DSIP topics appsearch DON27BZ02-NV011 there
SBIR.gov page
Email me topics like this
Requirements to bid
CMMC Level 2 (Self)
Protection of Controlled Unclassified Information: the 110 requirements of NIST SP 800-171 Rev. 2. "Self" means you assess yourself every three years; "C3PAO" means an accredited third-party assessor certifies you every three years. Both need a yearly affirmation.
ITAR The work is export-controlled, which restricts who may work on it
(foreign nationals need authorisation).
A summary, not advice: the solicitation and 32 CFR Part 170 govern.
TPOC: Richard Espinola <[email protected]>; Myron Pauli <[email protected]>
Keywords: Focal plane array, FPA, direct-bond interconnection, DBI, 3D heterogeneous integration, 3D HI, infrared imaging, IR, semiconductor processing, read out integrated circuit, ROIC, sensor chip assembly, SCA, mid-wave infrared, MWIR, IR detector materials
Technology areas: None
Objective
Develop a focal-plane array (FPA) fabrication process based on wafer-scale 3D-heterogeneous integration that enables a mid-wave infrared (MWIR) III-V detector epi-layer wafer to be directly bonded to a Silicon-based read-out integrated circuit (Si-ROIC) wafer, thereby eliminating the inefficient die-level Indium (In) bump-bonding process currently required to assemble Mid-wave Infrared (MWIR) FPA sensor chips.
Description
Electro-Optic and Infrared (EO/IR) sensors are used in a wide variety of applications, such as navigation, wide-area surveillance, and targeting. Operating in the visible and near-IR bands, EO sensors are available with small pixels and in large arrays at low cost, because the Si-detector wafer and Si-ROIC wafer can be 3D-integrated directly through wafer-to-wafer bonding with high yield at high volume based on material and complementary metal-oxide-semiconductor (CMOS)-process compatibility. By contrast, sensors operating in the MWIR employ III-V detector materials, which are not CMOS-process compatible, while III-V epi-layer detectors cannot be epitaxially grown directly on Si, as required to produce the crystalline order of high-quality detectors [Refs 1-2]. Instead, the III-V detector epi-layer is grown on a lattice-compatible substrate wafer, which is then flip-chip bonded to a Si-ROIC through a dense array of sub-pixel indium bumps placed on each wafer, followed by detector-substrate removal. The technically demanding process of depositing, aligning, and compressing the dense arrays of indium bumps on detector and ROIC die, results in a labor-intensive, inefficient, low-yield process for IR FPA fabrication, which greatly increases cost.
In recent years, 3D-heterogeneous integration has enabled a single integrated chip of greater overall functionality to be realized at reduced cost by assembling multiple smaller chips, each having unique functionality, material, node size, and/or foundry origin. Direct-bond interconnection (DBI) [Refs 3-5] is a 3D-heterogeneous integration technique that can potentially enable low-cost MWIR FPA fabrication. In principle, DBI should enable very large format FPAs with small pixels to be fabricated, as the process is largely independent of wafer and pixel size, in contrast to conventional bump-bonding. Moreover, the DBI process utilizes a temperature range that is compatible with MWIR IR detector materials.
The goal of this SBIR topic is to demonstrate the use of DBI or a similar process for fabrication of high-performance, large format, small-pixel, MWIR FPAs at low cost with the objective of refining the technique sufficiently to potentially replace conventional bump-bonding.
Phase I
Plan a detailed concept for a FPA fabrication and characterization process, including a risk assessment. While the pixel unit cell design is flexible, a pixel size of less than or equal to 10 micro-meters (um) and format of at least 2K x 2K are required, but otherwise any existing or commercially available Si-based ROIC can potentially be used for this effort. The III-V detector epi-layer spectral response must cover at least the MWIR band (3-5 um). The specific foundry and fabrication process steps will be determined in Phase I. Develop a detailed FPA characterization plan, including measurement of inter-wafer voids, pixel operability, and yield. Perform a per-FPA cost analysis that compares DBI or the selected alternative process to conventional bump-bonding and includes an assessment of the cost impact of process selection trade-offs.
Phase II
Use the DBI or selected alternative process to fabricate multiple wafers of FPAs having less or equal to 10 um pixels and greater than or equal to 2K x 2K format. Dice and thoroughly characterize the FPAs. Integrate one or more FPA sensor chips into an imaging system and subsequent optical performance characterization (favorable but is not required under this effort.) Prepare a Phase III commercialization/transition plan.
Phase III / dual use
Build and test an imaging sensor employing a DBI-fabricated FPA with available Phase III funding upon successful completion of Phase II.
The advancement of MWIR imagers could support many dual-use applications, such as border security, law enforcement, and industrial process monitoring. In general, the advancement of 3D-HI and DBI specifically could facilitate the development of novel micro-electronic devices beyond MWIR imagers.
References
- Rogalski, A. "Recent progress in infrared detector technologies." Infrared Physics & Technology, 54(3), 2011, pp. 136-154. https://www.scirp.org/reference/referencepapers?referenceid=838480
- Martyniuk, P. and Rogalski, A. "HOT infrared photodetectors." Opto-Electronics Review, 21(2), 2013, pp. 239-257. https://www.researchgate.net/publication/257909096_HOT_infrared_photodetectors
- US Patent 9,082,627: Method for low temperature bonding and bonded structure (covers the core process of room or low-temperature dielectric-to-dielectric and metal-to-metal bonding). https://patents.google.com/patent/US9082627B2/zh
- US Patent 6,962,835: Method for room temperature metal direct bonding. https://patents.google.com/patent/US6962835B2
- US Patent 9,391,143: Method for low temperature bonding and bonded structure. https://patents.google.com/patent/US9391143B2
Copied from the official source and may be out of date. Confirm dates and requirements in the
official solicitation.