Obviousness

Obviousness Analysis of US Patent 12087871 under 35 U.S.C. § 103

This analysis assesses the obviousness of US patent 12087871, titled "Microstructure enhanced absorption photosensitive devices," under 35 U.S.C. § 103, using prior art explicitly referenced and described within the patent's "Definitions" section. The current date is April 26, 2026.

Summary of the Invention (Derived from Definitions)

The patent US12087871 describes a photodetector (including photodiodes (PDs) and avalanche photodiodes (APDs)) that incorporates a microstructure-enhanced photon absorbing semiconductor region. This region includes a plurality of microstructures (such as pillars, holes, and/or voids) that are dimensioned and positioned to increase the absorption of photons at a range of wavelengths, with at least one dimension equal to or shorter than the longest signal wavelength. The microstructures enhance absorption by forming an "absorbing mode high contrast grating" that utilizes resonance, scattering, near-field, sub-wavelength, and/or interference effects. The invention aims to achieve high data bandwidths (e.g., greater than 5 Gb/s or 10 Gb/s) and high quantum efficiencies (e.g., greater than 60% or 90%) at various wavelengths, including 850 nm, 980 nm, 1000 nm, and up to 1750 nm. A key aspect is that these microstructures also effectively reduce the capacitance of the photodetector, allowing for higher bandwidth due to reduced RC time constants. The semiconductor materials can include silicon, germanium, III-V materials, or combinations thereof.

Identified Prior Art

The "Definitions" section of US12087871 explicitly identifies and describes several pieces of prior art:

1. Conventional PIN Photodiode (e.g., FIG. 2): A standard silicon PIN photodiode structure is presented as conventional, where the absorption "I" region thickness ("d") dictates performance. The patent notes that for 90% absorption in Si at 850 nm, "d" is over 30 microns, limiting bandwidth to less than 2.5 Gb/s, and achieving 10 Gb/s results in quantum efficiency (QE) less than 40% (FIGS. 3A and 3B).
2. Silicon Nanowire/Nanohole Arrays for Photovoltaics:
   • Garnett et al., "Light trapping in silicon nanowire solar cells," Nano Letters, 2010: This reference is cited for experimentally demonstrating that an ordered array of silicon nanowires increased the optical path length of incident radiation by 73 times greater than bulk silicon, leading to an effective absorption coefficient 73 times that of bulk silicon. It also notes that the nanowire array reduced the effective capacitance compared to bulk material.
   • Kelzenberg et al., "Enhanced absorption and carrier collection in Si wire arrays for photovoltaic applications," Nature Materials, 2010: Cited for enhanced absorption and carrier collection in Si wire arrays for photovoltaic applications.
   • Li et al., "Optical absorption enhancement in silicon nanowire and nanohole arrays for photovoltaic applications," Proceeding of SPIE, 2010: Cited for optical absorption enhancement in silicon nanowire and nanohole arrays for photovoltaic applications.
     The patent clarifies that nanowires are "known to be used for light trapping in photovoltaic applications where the photogenerated carriers diffuse to the anode or cathode of a P-N junction with zero external bias operating at DC (direct current)".
3. High Contrast Gratings (HCGs): The patent refers to "Chang-Hasnain" in relation to HCGs, noting that resonant Q can be as high as 10^7 if the HCG has minimal absorption loss, implying known applications for high reflectivity.
4. Epitaxially Grown Ge/Si Avalanche Photodiodes (Kang et al., 2008): This reference describes a Ge/Si APD for 1310 nm wavelength operation, which achieved approximately 15 Gb/s bandwidth for a 30 μm diameter device, but with a QE of only 56% for a 1 μm Ge absorption length. The patent notes that Kang et al. "cannot extend the wavelength due to low absorption of the bulk material without sacrificing a significant reduction in bandwidth".

Obviousness Combinations

A person having ordinary skill in the art (PHOSITA) at the time of the invention (priority date May 22, 2013) would possess knowledge in semiconductor device physics, optoelectronics, and fabrication techniques for photodetectors and related devices.

Combination 1: Conventional Silicon Photodiode (FIG. 2) + Silicon Nanowire/Nanohole Arrays (Garnett et al., Kelzenberg et al., Li et al.)

• Prior Art Teachings:
  • The conventional silicon photodiode (FIG. 2, FIGS. 3A and 3B) is known to have limitations for high-speed datacom applications. Specifically, at 850 nm, achieving acceptable quantum efficiency (e.g., 90%) requires a thick absorption region (e.g., >30 microns), which severely limits bandwidth (e.g., <2.5 Gb/s) due to long carrier transit times. Conversely, reducing the thickness for higher bandwidth leads to unacceptably low quantum efficiency (e.g., <40% for 10 Gb/s). This presents a clear problem: improving both bandwidth and QE in silicon photodetectors, especially at datacom wavelengths.
  • Garnett et al., Kelzenberg et al., and Li et al. explicitly teach that silicon nanowire and nanohole arrays significantly enhance optical absorption and increase the optical path length within silicon in photovoltaic (solar cell) applications. Garnett et al. quantifies this, showing a 73x increase in effective absorption coefficient for Si nanowires at 850 nm, reaching levels higher than direct bandgap III-V materials. Crucially, Garnett et al. also indicates that these nanowire arrays lead to reduced effective capacitance.
• Motivation to Combine: A PHOSITA, aware of the bandwidth and QE limitations of conventional silicon photodetectors for high-speed datacom (e.g., 10 Gb/s at 850 nm), would be motivated to find ways to increase the effective absorption of silicon without increasing the physical thickness of the absorption region. The demonstrated success of silicon nanowire/nanohole arrays in enhancing absorption for solar cells would naturally suggest applying similar microstructures to the absorption region of a silicon photodiode. The motivation is to leverage the known optical absorption enhancement of these microstructures to reduce the necessary physical length of the "I" region.
• Predictable Result: By reducing the physical absorption length due to enhanced effective absorption, the carrier transit time would be reduced, leading to higher bandwidth. Furthermore, the known reduction in effective capacitance associated with such microstructures (as explicitly noted in the patent when discussing Garnett et al.) would further contribute to a reduction in the RC time constant, also increasing bandwidth. This combination would predictably address the core problem of simultaneously achieving high bandwidth and high quantum efficiency in silicon photodetectors. The modification of applying these known structures to an existing device (a photodiode instead of a solar cell) to achieve a known and desired property (enhanced absorption for speed/efficiency) would be obvious.

Combination 2: Conventional Ge-on-Si APD (Kang et al.) + Microstructures for Absorption Enhancement (Garnett et al., Kelzenberg et al., Li et al.)

• Prior Art Teachings:
  • Kang et al. teaches a Ge-on-Si APD capable of detecting 1310 nm light, achieving around 15 Gb/s bandwidth for a 30 μm diameter device. However, this device suffered from relatively low QE (56% for a 1 μm Ge absorption layer) and could not easily extend to longer wavelengths without sacrificing bandwidth, due to the low absorption of bulk germanium at these wavelengths. This highlights a problem with Ge-on-Si APDs: improving QE and extending wavelength range while maintaining high bandwidth.
  • Garnett et al., Kelzenberg et al., and Li et al. (as discussed above) teach that microstructures like nanowire and nanohole arrays significantly enhance optical absorption in silicon and other materials for light trapping applications.
• Motivation to Combine: A PHOSITA seeking to improve the QE and wavelength sensitivity of Ge-on-Si APDs for high-speed datacom (e.g., 30 Gb/s at 1300-1750 nm) would be motivated to address the "low absorption of the bulk material" limitation described for Kang et al.'s device. Given the well-established principle from the solar cell art that microstructures can drastically increase effective absorption, it would be obvious to apply similar microstructuring techniques (e.g., pillars or holes) to the germanium absorption layer of a Ge-on-Si APD. The patent itself provides this motivation: "Ge on Si microstructures are fabricated on the Ge to increase absorption which allows a shorter length of Ge to be used resulting in higher speed due to lower effective capacitance and shorter transit time for the carriers".
• Predictable Result: By incorporating microstructures into the germanium absorption layer, the effective absorption coefficient of the germanium would be increased. This allows for a shorter physical absorption length while maintaining high QE, thereby reducing carrier transit time and improving bandwidth. The reduction in effective capacitance due to the microstructures would further contribute to increased bandwidth by reducing the RC time constant. These predictable improvements directly overcome the limitations of conventional Ge-on-Si APDs as described by Kang et al..

In conclusion, the core claims of US12087871 regarding microstructure-enhanced absorption in photodetectors (PDs and APDs) for improved bandwidth and quantum efficiency would have been obvious to a PHOSITA in light of the prior art, particularly the teachings of light-trapping microstructures in silicon for solar cells and the known limitations of conventional photodetector designs. The motivation would be to apply known absorption enhancement and capacitance reduction techniques to solve known problems in photodetector performance.