OUSD (R&E) MODERNIZATION PRIORITY: Microelectronics; Network Command, Control and Communications; General Warfighting Requirements
TECHNOLOGY AREA(S): Ground Sea; Electronics; Materials
OBJECTIVE: Develop a high purity, ultra-high vacuum surface activated bonded wafer fusion (UHV-SAB) tool for heterogeneous integration of dissimilar materials for electronic materials integration and thermal management solutions. The target of the topic is to demonstrate a wafer fusion system that is able to bond wafers using an in-vacuo low damage surface preparation process (i.e., atomic surface activation step) in order to produce a high density of dangling bonds prior to wafer fusion in a high purity, high vacuum (i.e., free of interfacial species) environment .
DESCRIPTION: Future Navy ships and other DoD assets will require high power converters for systems such as Air and Missile Defense Radar (AMDR) and propulsion on DDG-51 size ship platforms. High voltage, high efficiency power switches are required to achieve enhanced Size, Weight, and Power enhancements, and doing so affordably (SWaP-C). Ga2O3 has six crystalline phases that have been researched for more than eight decades . The monoclinic phase of Ga2O3 (β-Ga2O3) is the only thermodynamically stable phase with an UWBG of (4.85 eV) and high critical field (Ec = 6-8 MV/cm), resulting in a nearly ten-fold higher Baliga FOM than that of 4H-SiC (BFOM of Ga2O3 = 3444, BFOM of 4H-SiC = 300) . Recent major advances in Ga¬2O3 melt-growth techniques have resulted in commercial substrate technology, which as of late 2018 has progressed to the 6” (150-mm) wafer scale. This critical milestone in semiconductor technology, combined with its attractive electronic properties, have positioned Ga2O3 to offer savings in both energy and cost in high-power, high-temperature electronic device applications.
The goal of this capability is to improve the efficiency of thermal transport across heterogeneous interfaces because of the promotion of newly-discovered phonon modes observed at the interface of high quality epitaxial materials . To extend this concept to dissimilar materials such as GaN, Ga2O3, and diamond, an advanced wafer bonding capability must be developed. For example, the Ultra-Wide Bandgap (UWBG) semiconductor Gallium Oxide (Ga2O3) semiconductor has an extraordinarily high critical electric field breakdown strength but suffers from low thermal conductivity. To achieve a superior thermal solution, fusion or bonding of Ga2O3 to a high thermal conductivity substrate such as Silicon Carbide (SiC) or diamond may present the Navy/DoD with an aggressive approach to very high performance power electronics. In another example, Wide Bandgap (WBG) Gallium Nitride High Electron Mobility Transistors (GaN HEMTs) technology suffers thermally and presents a trade-off between millimeter wave output power and reliability. Fusion of GaN HEMTs to diamond substrates for example could boost thermal performance tremendously by eliminating interface layers that cause an increase in thermal barrier resistance (TBR.) Such a breakthrough would improve millimeter wave output power without sacrificing reliability.
Wafer bonding also permits two or more optically-dissimilar materials to be smoothly interfaced over long distances with low optical losses. This new hybrid material enables previously unattainable optical properties. For example, the silicon waveguides used for photonic integrated circuits lack the intrinsic electro-optic and nonlinear properties required for amplification or high-fidelity modulation. Wafer-bonding an active material such as LiNbO3 or InGaAs to the silicon waveguide layer allows state-of-the-art optical functionality compatible with CMOS based semiconductor fabrication. Also, wafer-bonding permits the placement of materials with very different refractive indices in close proximity. This enables birefringent phase-matching, dispersion engineering, and ultra-high mode intensities for efficient nonlinear frequency mixing.
The leading approach to wafer bonding presently uses direct wafer bonding which requires annealing at elevated temperatures to develop a strong interface bond. This approach does not address bonding of materials with dissimilar coefficient of thermal expansion such as between Ga2O3 and diamond or GaN and diamond. State of the art (SOA) in-situ surface bond activation via plasma bonding or sputter cleaning leave undesirable damage to the surface which does not produce an atomically sharp bonding interface. Other SOA bonding approaches add monolayers of metal to assist in bonding the dissimilar surfaces; this is not desirable either since the bonding interlayer could result in an electrically conductive interface. Deposition of dielectrics is not desirable because it increases TBR. Instead, an in-vacuo subtractive clean is desirable to activate covalent bonds and induce direct bonding of the dissimilar materials at the atomic level. Upon bringing the two wafers into contact at room temperature, strongly covalently bonded surfaces with minimal thickness of amorphous damage layers are required. In-vacuo outgassing annealing or low power plasma cleaning of the wafers prior to bonding pair may be required. UWBG Ga2O3 semiconductor material has a high Baliga power figure of merit (FOM) needed for high voltage, high efficiency power switches; however, previously states, Ga2O3 material has a poor thermal conductivity. One goal of the effort is to wafer bond Ga2O3 wafers to a high thermal conductivity SiC wafer with a low damage and low electrical resistance wafer bond interface. The objective, then, is to develop a system for low surface damage, ultra-high vacuum, room temperature atomic surface-activated wafer bonding of WBG and UWBG semiconductor wafers.
The proposed surface-activated bonded (SAB) system should strive to achieve the following desired characteristics
• Sample size 2” to 6” diameter, 0.1 to 5 mm thick
• Vacuum Level > 1x10-7 Torr
• In-vacuo surface preparation approach to ensure ultra-low damage, ultra-clean and sharp bonded interface. Such approaches could include various plasma sources and fast atom beams.
• Bonding pressure up to 3000 Newtons
• Annealing temperature up to 1000°C
• Vacuum load lock for insertion of wafers into main vacuum chamber
• In-situ monitoring capability including residual gas analyzer (RGA)
PHASE I: Determine feasibility and establish a plan for the design and development of UHV-SAB system to activate and bond wafers of Si, SiC, GaN, Ga2O3, AlN, and diamond.
The system should be designed to meet as many of the desired characteristics listed above as possible, providing heat treatment regimes, necessary for successful wafer bonding. The contractor shall provide a Phase II development plan addressing technical risk reduction.
PHASE II: Develop a fully-functional surface-activated bonding (SAB) system having all parameter monitoring and control tools and capable of producing ultra-low TBR bonded pairs by minimizing plasma damage from the in-situ pre-bonding cycle. The system should demonstrate uniform, void-free bonding of up to 8” wafers as required in the technical specification. A prototype of the fully operational system with appropriate control software will be delivered to the Navy/DoD and is required by the end of Phase II for evaluation.
PHASE III DUAL USE APPLICATIONS: Phase III shall address the commercialization of the product developed as a prototype in Phase II. The small business is expected to work with suitable industrial partners for this transition to military programs and civilian applications by identifying the expected final state of the technology, its use, and the platform it will be used on.
- F. Mu et al., ACS Appl. Mater. Interfaces 2019, 11, 36, 33428–33434.
- Z. Cheng et al. Nat Commun 12, 6901 (2021).
- S.J. Pearton, Appl. Phys. Rev. 5, 011301 (2018).
- M. Higashiwaki et al., Appl. Phys. Lett. 100, 013504 (2012).
KEYWORDS: Wafer Fusion; Ultra-Wide Bandgap; Power Switches; Wafer Bonding; Surface Activated Bonding (SAB); Residual Gas Analyzer (RAG)