Efficient Infrared Upconversion Imaging

• Infrared upconversion imaging is a process that converts infrared photons to visible light using advanced nonlinear optical techniques and engineered materials.
• Techniques like four-wave mixing, sum-frequency generation, and triplet–triplet annihilation enable high quantum efficiency, low noise, and superior spatial resolution.
• Integrating components such as metasurfaces, nanoparticles, and graphene achieves broadband, low-noise performance for applications in imaging, sensing, and quantum information.

High-efficiency infrared upconversion imaging refers to the set of physical processes, materials engineering, device architectures, and imaging system designs that enable the conversion of infrared (IR) photons into higher-energy visible photons with high quantum efficiency, low noise, high fidelity, and spatial resolution. This capability bridges the spectral gap between conventional IR and visible detectors, enabling enhanced imaging, sensing, and quantum information applications across scientific, biomedical, and technological domains.

1. Physical Mechanisms and Key Configurations

High-efficiency infrared upconversion imaging capitalizes on nonlinear optical or quantum photophysical mechanisms that mediate photon energy transfer from the IR regime to the visible. Principal mechanisms include:

• Nonlinear Wave Mixing in Atomic and Solid-State Systems:
  • Four-wave mixing (FWM) in hot vapor cells with ladder-type atomic configurations, as demonstrated in $^{85}$Rb, achieves high upconversion efficiency by exploiting a three-level ladder system (|3⟩ = 5S$_{1/2}$, |2⟩ = 5P$_{3/2}$, |1⟩ = 4D$_{5/2}$) and large single-photon detunings to suppress linear absorption and enhance nonlinear $\chi^{(3)}$ (Ding et al., 2012, Ding et al., 2012).
  • Sum-frequency generation (SFG) or third-harmonic generation (THG) in nonlinear crystals or metasurfaces, particularly those supporting high-Q resonances such as quasi–bound states in the continuum (quasi–BIC), dramatically enhance the local field and the conversion efficiency (Liu et al., 29 Aug 2025, Camacho-Morales et al., 2021, Mrejen et al., 2019).
  • Triplet-sensitization and triplet–triplet annihilation (TTA) in organic bulk heterojunctions, where NIR photons are harvested by sensitizers (Y6, perovskites) and upconverted via TTA (often enhanced by plasmonics or thin-film optics), resulting in all-passive and low-threshold upconversion (Nienhaus et al., 2019, Hamid et al., 27 Nov 2024).
• Energy Transfer Upconversion in Nanostructures:
  • Holmium-doped nanoparticles (HoNPs) use a four-photon energy transfer cascade to upconvert 2 μm SWIR photons to 640 nm emission. Plasmonic nanocavities are used to further enhance the emission intensity and radiative decay rate via the Purcell effect (Arul et al., 29 Nov 2024).
• Electronic Devices and Hybrid Structures:
  • HIWIP–LED architectures combine a homojunction interfacial workfunction internal photoemission (HIWIP) IR detector and an integrated LED, achieving ultra-broadband response (visible, MIR, THz) and pixel-less imaging (Li et al., 2022).
  • Graphene layer/LED (PGLIP–LED) heterostructures rely on pixelless upconversion by leveraging the high photoconductive gain and interband absorption in polycrystalline graphene (Ryzhii et al., 2017).

2. Efficiency Determinants and Enhancement Strategies

Efficiency of infrared upconversion imaging is dictated by nonlinear susceptibility, resonance engineering, and system losses. The main approaches for maximizing efficiency are:

• Resonance and Field Localization:
  • High-Q resonances (e.g., quasi-BIC in silicon metasurfaces) scale local field enhancement as $Q$ and can increase $\chi^{(n)}$-mediated THG or SFG efficiency even for modest pump intensities (Liu et al., 29 Aug 2025).
  • In atomic vapor FWM, large single-photon detunings decouple absorptive loss while two-photon near-resonance conditions maximize nonlinear dispersion (Ding et al., 2012).
• Material and Device Engineering:
  • Core–shell nanoparticle architectures (e.g., HoNP@NaGdF$_4$) reduce surface quenching and extend lifetimes, favoring sequential photon absorption and reducing nonradiative loss (Arul et al., 29 Nov 2024).
  • Plasmonic nanocavities both enhance the density of optical states and decrease emission lifetime (up to $>10^4$ Purcell factor), yielding up to 32× upconversion intensity enhancement (Arul et al., 29 Nov 2024).
  • Disordered metasurfaces with hybrid Mie-plasmonic cavities broaden the absorption spectrum and localize the near field, with reported increases of 2.6-fold IR absorption and 3.9-fold local field intensity compared to periodic analogues (Chen et al., 16 Mar 2025).
• Optical System Optimization:
  • Flat-top pump beams (as opposed to Gaussian) yield more uniform conversion across the image, enhancing segmentation and spatial fidelity in SFG-based imaging (Yang et al., 2019).
  • Adiabatic poling of nonlinear crystals (adiabatic SFG) extends the phase-matching bandwidth for multicolor imaging (2–4 μm in a single shot) and maintains high efficiency ($\sim$20%) across the band (Mrejen et al., 2019).

3. Fidelity, Spatial Resolution, and System Aberrations

Maintaining spatial fidelity through the upconversion process is essential for imaging applications:

• Phase Matching and Spatial Transfer:
  • In 4-$f$ systems, FWM or SFG upconversion preserves image information by coherent transfer of the spatial mask onto the visible field, with a frequency conjugation relationship $$U_\mathrm{output}(x', y') \propto U_\mathrm{input}(-x, -y)$$ (Ding et al., 2012).
  • Metasurface-based approaches maintain fidelity by minimizing angular (spatial-frequency) dependency in the transfer function $H(k)$; Fourier-plane imaging is implemented to compensate for angular dispersion and nonlocality (Molina et al., 28 May 2024).
  • Experimental systems with Siemens star targets attain $\sim$6 μm spatial resolution in silicon quasi–BIC metasurfaces (Liu et al., 29 Aug 2025), or $\sim$100 lines/mm in TTA-based all-passive systems (Hamid et al., 27 Nov 2024).
• Analytical Treatment of Aberrations:
  • Derived models for the depth of field (DOF) $d_\mathrm{DOF} = \lambda/\mathrm{NA}^2$ and astigmatic aberration $$d_a = M^2 [f_1 / (\sin^2\theta/n_1 - 1)] + [f_2 / (\sin^2(\theta/M)/n_2 - 1)]$$ allow system optimization for finite crystal apertures, high angular acceptance, and pump waist control, with close agreement between theory and experimental measurement (Han et al., 30 May 2025).
• Mode-Selective and Polarization-Preserving Approaches:
  • Tailored pump fields in nonlinear crystals (using SLMs) allow selective upconversion with extinction ratios >18 dB between desired and orthogonal spatial modes, beneficial for compressive or quantum imaging (Kumar et al., 2018).
  • Sagnac-type nonlinear interferometers achieve spatial polarization independent upconversion, fully preserving the vector (spin–orbit) structure of all spatial polarization components (Wu et al., 2020).

4. Enabling Materials, Devices, and Imaging Architectures

Cutting-edge materials and device configurations shape the performance limits and versatility of infrared upconversion imaging:

Material/System	Process	Efficiency/Resolution	Notable Features
$^{85}$Rb vapor	Ladder-type FWM	η ≈ 54% at 1529.4 nm→780 nm	High $\chi^{(3)}$, phase-matched
Si metasurface	Quasi-BIC THG	η = 3×10⁻⁵ (10 GW/cm²)	6 μm resolution, CMOS compatible
LiNbO₃ metasurface	Resonant SFG	η = 1.93×10⁻⁵ cm/GW (Q=40)	Edge-detection via nonlocality
Ho-doped nanoparticles	4-photon ETU	15.2% (max n_uc/n_dc)	Emission lifetime <1 ns (Purcell)
PGLIP–LED (graphene)	Interband absorption→LED	High gain, pixelless imaging	Contrast transfer tunable
Perovskite–Rubrene	TTA	>3% (785 nm)	Bilayer, sub-500 mW/cm² threshold
HIWIP–LED (GaAs, 20×)	FCA+IVBA→LED	0.14 A/W (10.5 μm)	Broadband (Vis–THz), pixelless
UCNPs+disordered metasf.	Multi-photon/field enh.	0.22 A/W (1550 nm), EQE 17.6%	19× current enhancement

Specific system choices are dictated by application (e.g., wavelength band, imaging speed, device integration requirements), with trade-offs in quantum yield, background signal, and fabrication complexity.

5. Imaging Modes, Applications, and Integration Prospects

• Imaging Modes and Sensing Performance:
  • Hyperspectral imaging with upconverted single photons (via cavity-enhanced SPDC and cascaded SFG) achieves high-contrast, shot-noise-limited MIR imaging (2.9–3.6 μm) of polymers and biological tissue with Si-SPAD detection (Meng et al., 27 Aug 2025).
  • High-resolution imaging in the 10 μm band for thermal targets is enabled at room temperature, with portable implementation via upconversion into the visible accessible to conventional detectors (Han et al., 30 May 2025).
  • Edge detection in upconversion devices can be built into the metasurface structure (e.g., by introducing a phase dislocation for simultaneous imaging and image processing) (Molina et al., 28 May 2024).
• Application Domains:
  • Night vision, surveillance, autonomous navigation: Upconversion enables uncooled, low-noise, high-sensitivity IR vision for security and autonomous platforms (Molina et al., 28 May 2024, Li et al., 2022).
  • Biomedical and chemical imaging: Label-free, non-destructive spectral imaging of tissues or chemicals with hyperspectral and spatial specificity, especially when photodamage must be minimized (Meng et al., 27 Aug 2025).
  • Quantum information: Photon-number-resolving imagers (e.g., Skipper-CCD, with sub-electron readout noise) are relevant for quantum metrology, astronomical detection, and quantum communication (Stefano et al., 2023, Kumar et al., 2018).

6. Limitations, Challenges, and Future Outlook

Key challenges and future directions include:

• Thermal Requirements and Integration:
  • Most state-of-the-art electronic upconversion devices (e.g., HIWIP–LED) operate at cryogenic temperatures; raising operation temperatures without severe loss of responsivity is an active area (Li et al., 2022).
  • Extraction efficiency is often limited by optical outcoupling; metasurface engineering for enhanced light extraction is an area of advancement (Arul et al., 29 Nov 2024).
• Bandwidth and Spectral Versatility:
  • Conventional nonlinear processes face narrow phase-matching constraints; adiabatic SFG and polychromatic metasurface engineering expand the spectral range (Mrejen et al., 2019, Liu et al., 29 Aug 2025).
  • Next-generation devices aim for multi-color or simultaneous hyperspectral imaging capabilities, as already demonstrated via advanced upconversion schemes (Mrejen et al., 2019, Meng et al., 27 Aug 2025).
• Material and Device Scalability:
  • Device architectures based on silicon or all-solid-state platforms offer scalability, CMOS compatibility, and ease of integration into chip-scale imaging systems (Liu et al., 29 Aug 2025, Chen et al., 16 Mar 2025).
  • All-passive, low-threshold upconversion elements are moving toward energy-autonomous night vision and remote sensing workflows (Hamid et al., 27 Nov 2024).

The field continues to evolve with innovations spanning quantum nanophotonics, material science, and device/system engineering, with future progress expected from synergistic optimization of nonlinear interactions, resonant photonics, and low-noise/high-sensitivity detection for comprehensive, broadband, and robust infrared upconversion imaging platforms.