nBn HOT-MW and SWIR on silicon towards 300 mm using interfacial misfit (IMF) arrays and direct bonding

The project goal is to develop interfacial misfit (IMF) arrays to integrate III-V semiconductor nBn mid-wave infrared (MWIR) detectors with large-format Si substrates. The MWIR region of the electromagnetic spectrum is critically important for naval defense applications including target tracking and identification in all weather conditions. The Navy is therefore eager to deploy widespread MWIR imaging and detection. Currently, the best low-noise MWIR detectors and focal plane arrays (FPAs) are based on the nBn device architecture that we first pioneered in 2006. nBn MWIR detectors are made from III-V semiconductor heterostructures grown on GaSb substrates. The drawback is that GaSb substrates are extremely expensive and are only available up to 6 in diameter, two factors limitingthe Navys ability to implement latly on Si wafers would overcome both of these issues with GaSb substrates. Si substrates are two orders of magnitude cheaper than GaSb substrates of the same size, and are available up to 12 in diameter. What is more, the use of large format Si wafers allows oneto take advantage of the high-volume manufacturing processes developed by the electronics industry. The problem is that due to differences in lattice constant, polarity, and thermal expansion, integrating III-V semiconductors with Si substrates is a long-standingchallenge.We propose to overcome this challenge through the use of IMF arrays. We can form these periodic arrays of in-plane misfitdislocations right at the III-V/Si interface by selecting very specific conditions for molecular beam epitaxy. IMF arrays efficiently relieve the large strain between the III-V and the Si, without generating a high density of threading dislocations. The result ishigh-quality III-V growth on Si without the need for prohibitively thick buffer layers or pre-growth substrate patterning. Since the IMF arrays self-assemble, this approach is inherently scalable, presenting us with a viable route to grow low-noise MWIR nBn FPAs on large format, 812 Si wafers. In parallel, we will also explore the feasibility of also integrating III-V SWIR and LWIR detectors with Si wafers using modified IMF techniques.To meet the project goal, we will 1) Create high-quality GaSb buffers on AlSb/Si(001)IMF virtual substrates; 2) Synthesize and test III-V nBn devices on Si for specific MWIR bands; 3) Explore IMF-based SWIR and LWIR devices on Si using IMF arrays; 4) Fabricate and test FPAs from III-V nBn detectors grown on Si; and 5) Define a scalable pathway towards III-V IR FPAs on 12 Si wafers using direct-bonding. We are well-qualified to perform this work. Our research team consists ofPI: Paul Simmonds (Boise State), an expert in dissimilar materials integration by molecular beam epitaxy; and co-I: Shimon Maimon (NetzVision), the original inventor of the nBn detector for low-noise FPAs. If successful, the anticipated research outcome would be to provide the Navy with high-operating temperature, low-noise MWIR FPAs, at significantly reduced cost and on larger format substrates than is currently possible. We will demonstrate nBn FPAs on 8 Si, and establish a trajectory for future scaling of this technology to the 12 Si platform. By overcoming the GaSb substrate bottleneck, Si-based nBn devices could revolutionize the thermal imagingquarely with N00014-21-S-SN01 TA1. Although certainly an ambitious project, given our expertise in the MBE of IMFs, nBn detector design and fabrication, key university and industry partnerships, and access to world-class facilities, we are confident of achieving our technical objective of producing scalable III-V IR detectors on Si.