MAlSiN₃:Eu²⁺ (M = Ca, Sr) is commonly used in high-power phosphor-converted white-light-emitting diodes and laser diodes to promote their color-rendering index. However, the wide application of this phosphor is limited by the degradation of its luminescent properties in high-temperature, high-humidity, and high-sulfur-content environment. Here, the degradation mechanism of the (Sr,Ca)AlSiN₃:Eu²⁺ (SCASN) red phosphor under thermal-moisture-sulfur coupling conditions is investigated. Furthermore, by performing first-principles calculations, the hydrolysis mechanism on an atomic scale is assessed. The adsorption energy (E_ads) and charge transfer (ΔQ) results showed that H₂O chemically adsorbed on the (0 1 0), (3 1 0), and (0 0 1) surfaces of the CaAlSiN₃ (CASN) host lattice. The energy barrier for H₂O dissociation is only 29.73 kJ mol⁻¹ on the CASN (0 1 0) surface, indicating a high dissociation probability. The formation of NH₃, Ca(OH)₂, and CaAl₂Si₂O₈ is confirmed by H⁺ tended to combine with surface N atoms, while OH⁻ combined with the surface Al/Si or Ca atoms. Moreover, ab initio molecular dynamics simulations were performed to further understand the hydrolysis process. This work offers a guidance on the design and applications of luminescent materials in LED packages with higher reliability and stability requirements in harsh environment.

@en

High-power white light-emitting diodes (LEDs) have demonstrated superior efficiency and reliability compared to traditional white light sources. However, ensuring maximum performance for a prolonged lifetime use presents a significant challenge for manufacturers and end users, especially in safety–critical applications. Thus, identifying functional anomalies and predicting the remaining useful lifetime (RUL) is of enormous importance in the operational longevity of the device. To address such challenges, this study proposes a combination of distance-based Mahalanobis distance (MD), entropy generation rate (EGR), and deep learning models for improved anomaly detection and RUL prediction accuracy. Unlike conventional health indicators based on luminous flux data that are challenging to monitor relevant optical performance, the MD and EGR methods are employed to extract in-situ monitored thermal and electrical data as new health indicators. Long short-term memory recurrent neural networks (LSTM-RNN) and convolutional neural networks (CNN) are established to detect anomalies and predict the RUL. The accelerated degradation tests of 3 W high-power white LED have been conducted, and the online and offline collected experimental data are deployed for model development and performance evaluation. The performance of the proposed methods is compared against the Illuminating Engineering Society of North America (IESNA) TM-21 method. The results indicate that LSTM-RNN, when combined with either MD or EGR, can detect anomalies with significantly fewer data (70 %) than is typically required. Furthermore, a significant improvement in prediction accuracy in RUL prediction based on MD and EGR-constructed time series health indicators and employed with the LSTM-RNN model demonstrates the effectiveness of the proposed methods.

@en

Spectral power distribution (SPD) is the radiation power intensity at different wavelengths, containing the most basic photometric and colorimetric performance of the illuminant, which is able to predict the lifetime of LEDs. This paper proposes an SPD model assisted by machine learning algorithms to detect the early failure of white LEDs. The SPD features of 3W high-power white LEDs were firstly extracted by the statistical models of Gaussian, Lorentz, and Asym2sig functions. An unsupervised learning method, principal component analysis (PCA), was then used to reduce the extracted features parameters’ dimensions. Next a K-nearest neighbor (KNN)-based method was used to detect LEDs’ anomalies by dividing the main cluster into groups, and estimating the distance from the center of mass of each cluster to the test point. The results showed the following: (1) for selected white LEDs, the Asym2sig function has a better fitting result than Gaussian and Lorentz functions; (2) machine learning methods can significantly assist in LED anomaly detection and can decrease the amount of anomaly detection time to 789.6 h, compared to the 1311 h when lumen maintenance degradation reaches 70% as required by IES TM21.@en

The (Ca, Sr) AlSiN₃:Eu²⁺(CSASN:Eu) red phosphor is widely used to improve color rendering of high-power phosphor-converted lighting diode (pc-WLED), but it is always unstable under high temperature and high humidity environments. Therefore, the studies on the temperature and humidity resistance of red phosphors and their aging mechanism have become essential to evaluate its reliability in harsh applications. In this paper, the pressure cooker test (PCT) and 85°C&85% RH aging test were carried out for the CSASN:Eu red phosphors. And, its hydrolysis reaction-driven degradation mechanism was simulated and analyzed based on first-principle calculation, in which the optimized adsorption of simplified CaAlSiN3(CASN) and H2 O was simulated based on Density Function Theory (DFT) and the specific aging process was analyzed by the charge density difference and ab initio molecular dynamics (AIMD). The experimental results showed that the photoluminescence performance of CSASN:Eu red phosphor dropped gradually and finally disappeared under PCT aging, and its temperature-dependent degradation kinetics followed the Arrhenius model well. Meanwhile, the simulation results indicate that the CASN, reacted with H2 O when the H atoms had a tendency to approach N atoms. Both the temperature and humidity could accelerate the hydrolysis reaction rate.@en

The LED degradation failure is highly dependent on temperature and this degradation failure is an irreversible energy dissipation process in thermodynamics. In this paper, the entropy generation is used to quantify the energy dissipation, which is regarded as one of the main performance characteristics of LED's degradation process. Considering the thermodynamic characteristics of entropy generation in the LED failure, a hybrid degradation prediction model based on the permutation entropy (PE) and data-driven methods was proposed. Firstly, a thermal aging test was designed for white LEDs in which the entropy generation rates (EGRs) of LEDs were extracted from the online collected thermoelectric performance parameters. Then, the EGRs of LEDs were treated as a time-series signal to perform phase space reconstruction and calculate PEs. Finally, both neural network model and Wiener process based data-driven methods were used to process the PEs. This hybrid model links the thermodynamic entropy of LEDs with its optical performance. The results show that: (1) Entropy generation based on thermodynamics can characterize the degradation process of LEDs; (2) The proposed hybrid degradation prediction model based on the PE and Wiener method can achieve early failure warning of LEDs before the actual failure occurs.@en