Strategies to approach high performance in Cr³+-doped phosphors for high-power NIR-LED light sources

The study addresses the need for broadband NIR-emitting phosphors efficiently excitable by blue light for compact, smart NIR light sources (e.g., smartphone-integrated systems) used in food/medicine monitoring, bioimaging, and night vision. Conventional NIR LED chips emit narrow bands, limiting applications. Phosphor-converted NIR LEDs (pc-NIR-LEDs) analogous to pc-white LEDs are promising, but require broadband NIR phosphors with high quantum efficiency and excellent thermal stability to withstand high-power operation. Cr³⁺-doped phosphors in garnet hosts have shown promise but prior CSSG:Cr³⁺ reports suffered from low IQE and thermal stability due to impurities and Cr³⁺ oxidation in air. This work proposes and tests a simple optimization strategy—enhancing crystallinity, tuning micromorphology, and maintaining Cr³⁺ valence via flux addition and reducing-atmosphere sintering—to realize high-performance CSSG:Cr³⁺ phosphors and high-power pc-NIR-LEDs.

Recent NIR phosphors include many Cr³⁺-activated hosts, with reported IQEs typically 58–75% and radiant powers of 14.7–54.29 mW at 100–130 mA. Progress has raised radiant power to ~65.2 mW at 350 mA, but high-current operation demands superb thermal stability to mitigate thermal quenching. Garnet hosts provide robust environments; CSSG is known for Ce³⁺ with high QE and thermal stability. Cr³⁺ in CSSG can be excited by blue light and emit broadband NIR, but previously showed low IQE (~12.8%) and poor thermal stability when synthesized in air due to impurities and Cr³⁺→Cr⁴⁺ oxidation. These gaps motivate strategies that improve crystallinity and preserve Cr³⁺ valence to unlock higher efficiency and stability.

- Host and activator: Cr³⁺-doped Ca₃Sc₂Si₃O₁₂ (CSSG:Cr³⁺), with Cr³⁺ expected to substitute Sc³⁺ in octahedral ScO₆ sites (weak crystal field). Photophysics: spin-allowed ⁴A₂g→⁴T₁g(F) (~460 nm) and ⁴A₂g→⁴T₂g(F) (~640 nm) excitations; broad NIR emission (~770 nm) from ⁴T₂g(F)→⁴A₂g. - Synthesis and optimization: Solid-state reaction with various fluxes (NH₄F, CaF₂, H₃BO₃, LiF, Li₂CO₃) under air and CO reducing atmospheres. Initial air-sintered CSSG:3%Cr³⁺ showed low IQE/EQE (~12.8%/4.8%). Flux screening under air indicated H₃BO₃, LiF, and Li₂CO₃ enhanced PL vs flux-free baseline; NH₄F and CaF₂ decreased it. Subsequent sintering under CO with these fluxes greatly boosted PL (2–3×). Li₂CO₃ was best; optimized at 1 wt%. With 1 wt% Li₂CO₃ and CO sintering, Cr³⁺ concentration was tuned; optimal at 6% Cr³⁺. - Structural and microstructural characterization: XRD to assess phase purity and crystallinity; SEM-EDS mapping to evaluate elemental homogeneity and identify impurity phases. Comparison of air- vs CO-sintered (with 1 wt% Li₂CO₃) samples showed reduced SiO₂ and Sc₂O₃ impurities and higher CSSG peak intensities after optimization, with more homogeneous element distributions. - Valence state diagnostics: XPS for elemental presence; Diffuse reflectance (DR) spectroscopy to identify Cr⁴⁺ absorption (~1140 nm) versus Cr³⁺ bands (~460, 640 nm); EPR to quantify Cr³⁺ signals (g≈3.8–3.9 for isolated Cr³⁺; g≈2 for Cr³⁺-Cr³⁺ pairs). CO-sintered samples showed suppressed Cr⁴⁺ absorption and stronger Cr³⁺ EPR features, indicating maintained/increased Cr³⁺. - Optical measurements: PL/PLE at room temperature and 77 K (high-resolution, 0.05 nm step) to determine excitation/emission features, zero-phonon line (ZPL), phonon sidebands, Stokes shift, and EPC parameters (Huang–Rhys factor). Temperature-dependent PL (25–300 °C) to assess thermal stability, with Arrhenius analysis to extract activation energy for thermal quenching. Time-resolved fluorescence to analyze decay dynamics (biexponential, ~190 µs at 77 K). - Device fabrication and testing: A high-power 460 nm blue LED chip combined with optimized CSSG:6%Cr³⁺ to fabricate a pc-NIR-LED. Electroluminescence (EL) spectra recorded versus drive current (100–600 mA). Optical power of total output and NIR band measured with a spectrometer (measurement range up to 850 nm). Conversion efficiencies computed: η_NIR/blue light (blue-to-NIR), η_NIR/input (input electrical-to-NIR), and η_blue light/input (blue chip photoelectric efficiency). Corrections applied for unmeasured NIR fraction beyond 850 nm.

- Efficiency gains: - With 1 wt% Li₂CO₃ flux and CO reducing atmosphere, CSSG:3%Cr³⁺ IQE and EQE increased from ~12.8% and 4.8% (air-sintered) to ~77.8% and 15.5% (measured up to 850 nm). Optimizing Cr³⁺ to 6% yielded measured EQE of 21.5%. - Accounting for the unmeasured NIR beyond 850 nm (~15.7%), the actual IQE and EQE for CSSG:6%Cr³⁺ are estimated at 92.3% and 25.5%, respectively. - Thermal stability: - Optimized CSSG:6%Cr³⁺ retains 97.4% of its 25 °C PL intensity at 150 °C (vs 85.6% for air-sintered). Arrhenius analysis gives higher activation energy for thermal quenching: ΔE = 0.336 eV (optimized) vs 0.220 eV (initial). - Emission peak redshifts from ~783 to ~807 nm (25–300 °C); FWHM increases from 1483 to 1551 cm⁻¹ (92.3–100.2 nm). - Spectroscopic parameters (77 K): - PLE peaks at ~453 nm (22,050 cm⁻¹, ⁴A₂g→⁴T₁g(F)) and ~636 nm (15,670 cm⁻¹, ⁴A₂g→⁴T₂g(F)); R-line at 698.3 nm (~14,320 cm⁻¹) in PLE; ZPL at ~713 nm (~14,030 cm⁻¹). Energy gap between ²E and ⁴T₂g ZPL ~290 cm⁻¹ indicates strong SOC and state mixing; decay is biexponential with ~190 µs lifetime. - PL peak at ~770 nm (12,980 cm⁻¹) with FWHM ~1560 cm⁻¹ (~93 nm); Stokes shift ~2700 cm⁻¹; phonon satellite ~730 nm with phonon energy ~330 cm⁻¹; Huang–Rhys factor S ~3–4, indicating notable EPC. - Mechanistic evidence for improvement: - XRD/SEM-EDS show reduced impurity phases (SiO₂, Sc₂O₃) and enhanced crystallinity in optimized samples. - DR/EPR show suppression of Cr⁴⁺ and increased Cr³⁺ population under CO sintering, maintaining desired valence state. - Device performance: - pc-NIR-LED total measured optical power reaches 97.8 mW at 520 mA; measured NIR power 64.7 mW. Considering unmeasured NIR beyond 850 nm (15.7%), actual NIR power ~76.8 mW and total optical power ~109.9 mW at 520 mA. - Conversion efficiencies: η_NIR/blue light decreases from 33.5% to 12.3% as current increases 100→600 mA; η_NIR/input decreases 7.2%→2.3%, largely limited by blue chip η_blue light/input (31.9%→19.4% over 100–600 mA).

The optimization strategy—using Li₂CO₃ flux and CO reducing atmosphere—addresses the two key bottlenecks: crystallinity and Cr valence state. Improved crystallinity reduces nonradiative recombination pathways associated with impurity phases (SiO₂, Sc₂O₃) and compositional inhomogeneity, as confirmed by stronger CSSG diffraction peaks and uniform EDS maps. Maintaining Cr³⁺ and suppressing Cr⁴⁺ (evidenced by diminished ~1140 nm absorption and stronger Cr³⁺ EPR signals) preserves the efficient ⁴T₂g→⁴A₂g NIR emission channel, boosting QE. The strong electron–phonon coupling in CSSG:Cr³⁺, together with a small ~290 cm⁻¹ gap between ²E and ⁴T₂g, explains the broadband emission, temperature-induced blueshift in the 650–725 nm region (increased ²E radiative contributions via SOC/EPC-assisted mixing), and robust thermal performance (higher activation energy for quenching). These material-level improvements translate into state-of-the-art device metrics for pc-NIR-LEDs, though overall electrical-to-NIR efficiency is currently limited by the blue chip’s photoelectric efficiency. Employing more efficient blue emitters should further enhance device-level η_NIR/input. The demonstrated high radiant power and thermal stability make the system suitable for high-power applications such as night vision and NIR imaging.

This work demonstrates a simple yet effective route—enhancing crystallinity via flux (optimum 1 wt% Li₂CO₃) and preserving Cr³⁺ valence through CO reducing atmosphere—to realize high-performance CSSG:Cr³⁺ NIR phosphors. The optimized CSSG:6%Cr³⁺ exhibits near-record internal and external quantum efficiencies (actual IQE ~92.3%, EQE ~25.5%), excellent thermal stability (97.4% retention at 150 °C), and enables a pc-NIR-LED with high optical power (~109.9 mW total, ~76.8 mW NIR at 520 mA). Spectroscopic and structural analyses attribute the improvements to reduced impurity phases, increased crystallinity, and suppression of Cr⁴⁺ formation, with EPC/SOC considerations explaining detailed emission behavior. The strategy is broadly applicable to other Cr³⁺-doped NIR phosphors and supports the development of next-generation smart NIR light sources. Future work should integrate higher-efficiency blue chips to further raise electrical-to-NIR conversion efficiency and explore long-term device reliability and extended spectral coverage.

- Measurement bandwidth limited to 850 nm led to underestimation of NIR power and QE; corrected values account for an unmeasured ~15.7% of NIR emission. - Device-level electrical-to-NIR efficiency is constrained by the moderate photoelectric efficiency of the blue LED chip (31.9%→19.4% from 100→600 mA). - Air sintering induces Cr³⁺ oxidation to Cr⁴⁺ and impurity phases; reducing atmosphere is required to maintain performance.