Long-term performance of narrow-band red emitters used in SSL devices
Rountree, K., Davis, J. L., McCombs, M. S., Pope, R. H., Riter, K. C., & Hegarty-Craver, M. (2021). Long-term performance of narrow-band red emitters used in SSL devices. U.S. Department of Energy, Office of Energy Efficiency and Renewable Energy. https://www.energy.gov/sites/default/files/2021-10/ssl-rti-red-emitters-sept2021.pdf
Abstract
Solid-state lighting (SSL) technologies that use light-emitting diodes (LEDs) as illumination sources are efficient at converting electric current into optical radiation with wavelengths (λ) between 380 nanometers (nm) and 740 nm. SSL technologies usually produce a minimal amount of infrared radiation, in contrast to standard illumination technologies such as incandescent lamps that typically emit significant amounts of infrared. SSL technologies do produce waste energy due to inefficiencies in the conversion of electric current to emitted photons. In general, this waste energy manifests as increased LED junction temperatures (Tj) and is often removed from the LED package by a combination of thermal conduction (e.g., heat sink) and convection.
SSL technologies that use traditional phosphor-converted LEDs (pc-LEDs) provide an increase in spectral efficiency over incandescent lamps because almost all of the radiation they produce is in the visible range. However, conventional red phosphors used in white, pc-LEDs have broad emission peaks (full-width at half-maximum [FWHM] near 100 nm) that significantly spill over into the deep red and near-infrared regions (i.e., above 740 nm), where the human eye is not sensitive. By decreasing the FWHM of the red emitter (e.g., using a narrow-band [NB] red emitter) or shifting the red emission peak to lower wavelengths, a significant increase in luminous efficacy and spectral efficiency can be realized with potential impacts to the color rendering properties of the light source. One solution to managing the trade-off between spectral efficiency and color rendering is to use a blue LED pump, a broad emitting green-yellow phosphor, and a NB red emitter. This report focuses on a sampling of available SSL products that use NB red emitters to provide a benchmark of these technologies, and the performances of these NB products are compared in terms of luminous efficacy of radiant flux (LER), color rendering, and correlated color temperature (CCT). Conventional SSL products that use pc-LEDs of similar CCT values are also included in the comparison to investigate the potential energy savings of using NB red emitters instead of phosphors with broader red emission peaks, as is often done in conventional SSL devices.
The selected NB red emitter products investigated in this report used different architectures to decrease deep red and near-infrared emissions. One product was a 2-ft, Type A replacement, LED tube (Product NB-1) that contained magnesium-doped potassium fluorosilicate (PFS) phosphor, a NB technology that was first used in displays. Product NB-1 contains an LED module with 42 mid-power LEDs (MP-LEDs) encased in a glass tube and the product was tested in this form. The other two products (NB-2 and NB-3) that were benchmarked in this study contained red quantum dots (QDs) mixed with conventional green phosphors. QDs have recently made significant market gains in display and television applications. Both Product NB-2 and NB-3 were tested as light engines with two LED modules mounted to a common heat sink. No secondary optics were used with Products NB-2 and NB-3. The LED modules of Product NB-2 contain 21 MP-LEDs whereas the LED modules of Product NB-3 contain 72 MP-LEDs each.
This report summarizes the overall findings from up to 12,000 hours (hrs) of accelerated stress testing (AST) on the Product NB-1 DUTs and up to 11,000 hrs of AST on the Product NB-2 DUTs. The AST procedures used in this study included a room temperature operational life (RTOL) test, an operational life test conducted at 75°C (75OL), and a wet high-temperature operational life test performed at 75°C and 75% relative humidity (7575). Data for Product NB-3 is presented through 7,000 hours of operation in RTOL. During the ASTs described herein, separate populations of each product (three DUTs in each population for Product NB-1 and four DUTs in each population for Product NB-2) were subjected to power cycling of 1 hr on and 1 hr off. Four DUTs for Product NB-3 were operated continuously in RTOL. The performance of these NB products is compared to findings from two downlights that produce spectra that can be switched between different, preselected CCT values (e.g., nominal switching CCT values were 2,700 K, 3,000 K, 3,500 K, 4,000 K, and 5,000 K).
In addition, the light sources from representative samples of all DUTs were decapped using a chemical solution, which allowed an estimation of the external quantum efficiency (EQE) of the blue LED (η_(blue,EQE)) and the phosphor (η_(phos,EQE)). The η_(blue,EQE) estimate provides a guide to the efficiency of the product in converting electrical current into emitted photons. The η_(phos,EQE) values provide an estimate of the efficiency of blue photon conversion to broadband white light (e.g., green, red).
The key findings from this study include the following:
The radiant efficiency of the devices incorporating NB phosphors was generally higher than the conventional pc-LED technology benchmarks at all CCT values tested. Further improvements in radiant efficiency of these NB products appear possible with the use of more efficient LED emitters.
The narrowing of the red emission bands for the products in this study also led to significant gains in LER compared to the conventional pc-LED products. The best LER performance (329 lumens/watt [lm/W]) was found for the PFS samples (Product NB-1) with a nominal CCT of 3,500 K. This LER value represents a 15% improvement over the conventional pc-LED benchmark operated at 3,500 K.
The QD-containing products (NB-2 and NB-3) also exhibited an increase in LER over the conventional pc-LED benchmarks. The LER performance for the 4,000 K product (Product NB-3) was 310 lm/W (8% improvement over the pc-LED benchmarks at 4,000 K). For the 2,700 K product (Product NB-2), the LER was 294 lm/W, a 2% improvement over the pc-LED benchmarks.
The EQE value of the phosphor (η_(phos,EQE)) was estimated to be the highest for the PFS phosphor (0.99). The QD-phosphor composite was estimated to have excellent EQE values for both the 2,700 K and 4,000 K products, although the η_(phos,EQE) value was estimated to be higher for Product NB-3 (0.97) than Product NB-2 (0.89).
The color rendering properties of Products NB-1, NB-2, and NB-3 were excellent (Rf ≥ 91, Rg = 99), supporting the finding that increases in spectral efficiency in the red region can be achieved in a manner that does not adversely impact color rendering.
The temperature stability of the products in this study was generally good. The luminous flux maintenances (LFMs) of Product NB-1 remained above 0.93 in RTOL and 75OL over the test duration (12,000 hrs). For Product NB-2, the LFM value of the DUTs in RTOL and 75OL remained above 0.97 after 11,000 hrs of testing. Product NB-3 was only tested in RTOL, and its LFM value exceeded 0.99 after 7,000 hrs of testing.
The magnitude of chromaticity shift (u'v') remained small for Products NB-1, NB-2, and NB-3 in the RTOL and 75OL test conditions (u'v' < 0.001), demonstrating excellent color stability in these conditions.
The products in this study were greatly affected by the presence of humidity in the 7575 environment. LFM decreased drastically for both Products NB-1 and NB-2 and reached 0.80 by 6,000 hrs of exposure for Product NB-1 and 0.83 after 9,000 hrs of exposure for Product NB-2. Chromaticity shift also increased substantially for both products, with Product NB-1 (PFS phosphor) experiencing a shift in the blue direction (u'v' = 0.005) after 6,000 hrs and all DUTs of Product NB-2 (phosphor-QD mixture) failing parametrically due to excessive chromaticity shift (u'v' > 0.007) in the green direction by 3,000 hrs.
The phosphor-QD mixture used in Product NB-2 was especially susceptible to the humid environment of 7575, and the emission peak of the mixture shifted by 10 nm toward a lower value by the end of the test in the 7575 conditions.
The results in this report serve to benchmark the reliability of narrow-band red emitters and direct future research. In this report, narrow-band red emitters are shown to provide significant gains in spectral efficiency without compromising color fidelity. The products studied had good stability in low-humidity environments but improvements can be made in high-humidity environments, particularly for the phosphor-QD mixture.
To contact an RTI author, request a report, or for additional information about publications by our experts, send us your request.