Multi-machine learning methods to predict spatial variation characteristics of total nitrogen at watershed scale: Evidences from the largest watershed (Yangtze River Watershed), Asian.

Nitrogen pollution has emerged as a significant threat to the health of global river systems, garnering considerable attention. However, numerous challenges persist in understanding the characteristics and predicting the spatial changes of total nitrogen (TN) at the catchment scale. We leveraged data from 530 monitoring sections to calculate a land-use composite index and perform statistical analyses to explore the primary factors influencing nitrogen enrichment in the Yangtze River Watershed. We developed three machine learning models to forecast future TN concentrations at monitoring points. Our results showed that agricultural activities and rainfall were the primary drivers of monthly variations in TN concentrations. The upstream region of the watershed exhibited larger variations in TN concentrations (0.097 to 11.099 mg/L), significantly higher than the middle and downstream areas (0.348 to 6.844 mg/L). Microbial-mediated organic matter decomposition in sediment and changes in land-use were identified as key contributors to regional differences in nitrogen enrichment. Potential nitrogen sources include sediment release, urban sewage, and agricultural fertilization. Random Forest model achieved a prediction accuracy of 77.6 %, surpassing the BP and LSTM models. We identified 37 high-risk areas of nitrogen enrichment, concentrated in the Chengdu-Chongqing, Yunnan-Central urban cluster, and the Chaohu Lake sub-watershed. Increased urban land-use and industrial inputs primarily influenced nitrogen enrichment in the upstream area, while agricultural inputs were the main drivers in the middle and downstream regions. Our multi-machine learning models identified the relationship between TN and influencing factors, providing a reliable method for assessing nitrogen enrichment risk in the watershed.

Host specificity of plant-associated bacteria is negatively associated with genome size and host abundance along a latitudinal gradient.

Host specialization plays a critical role in the ecology and evolution of plant-microbe symbiosis. Theory predicts that host specialization is associated with microbial genome streamlining and is influenced by the abundance of host species, both of which can vary across latitudes, leading to a latitudinal gradient in host specificity. Here, we quantified the host specificity and composition of plant-bacteria symbioses on leaves across 329 tree species spanning a latitudinal gradient. Our analysis revealed a predominance of host-specialized leaf bacteria. The degree of host specificity was negatively correlated with bacterial genome size and the local abundance of host plants. Additionally, we found an increased host specificity at lower latitudes, aligning with the high prevalence of small bacterial genomes and rare host species in the tropics. These findings underscore the importance of genome streamlining and host abundance in the evolution of host specificity in plant-associated bacteria along the latitudinal gradient.

Inverse relationship between species competitiveness and intraspecific trait variability may enable species coexistence in experimental seedling communities.

Theory suggests that intraspecific trait variability may promote species coexistence when competitively inferior species have higher intraspecific trait variability than their superior competitors. Here, we provide empirical evidence for this phenomenon in tree seedlings. We evaluated intraspecific variability and plastic response of ten traits in 6750 seedlings of ten species in a three-year greenhouse experiment. While we observed no relationship between intraspecific trait variability and species competitiveness in competition-free homogeneous environments, an inverse relationship emerged under interspecific competition and in spatially heterogeneous environments. We showed that this relationship is driven by the plastic response of the competitively inferior species: Compared to their competitively superior counterparts, they exhibited a greater increase in trait variability, particularly in fine-root traits, in response to competition, environmental heterogeneity and their combination. Our findings contribute to understanding how interspecific competition and intraspecific trait variability together structure plant communities.

Evolution origin analysis and health risk assessment of groundwater environment in a typical mining area: Insights from water-rock interaction and anthropogenic activities.

Coal mining changes groundwater environment, results in deterioration of water quality and endangering human health in the mining area. However, the comprehensive study of groundwater evolution and its potential impact in mining area is still insufficient. In this study, 95 groundwater samples were collected from 2019 to 2020 in a typical mining area of China. Ion ratio coefficients, isotopic tracing technology, Entropy-weighted water quality index (EWQI) and human health risk assessment model (HHRA) were applicated to investigate the hydrochemical variation reasons, groundwater quality and its potential health risk in the study area. Results showed that the groundwater hydrochemical types changed from HCO3∙SO4-Ca∙Mg type to SO4-Ca∙Mg and SO4∙Cl-Ca∙Mg type. Water-rock interaction, agricultural activities, manure and sewage input, precipitation and evaporation controlled the groundwater hydrochemical composition. Groundwater quality showed a trend of fluctuation with an average EWQI of 59.23, 68.92, 63.75, 58.02 and 64.92, respectively. 91.6% of the water samples was fair and acceptable for drinking. The groundwater health risk of nitrate in the study area ranged from 0.03 to 17.80. Infants had the highest health risk and nitrate concentration was the most sensitive parameter. The results will present a comprehensive research of groundwater evolution and potential impacts through a typical mining area example. Thereby offering valuable insights into the influencing factors identification, hydrochemical processes evolution, protection and utilization of groundwater in global mining areas.

Planar Cation Passivation on Colloidal Quantum Dots Enables High-Performance 0.35-1.8 µm Broadband TFT Imager.

Solution-processed colloidal quantum dots (CQDs) are promising candidates for broadband photodetectors from visible light to shortwave infrared (SWIR). However, large-size PbS CQDs sensitive to longer SWIR are mainly exposed with nonpolar (100) facets on the surface, which lack robust passivation strategies. Herein, an innovative passivation strategy that employs planar cation, is introduced to enable face-to-face coupling on (100) facets and strengthen halide passivation on (111) facets. The defect density of CQDs film (Eg ≈ 0.74 eV) is reduced from 2.74 × 1015 to 1.04  × 1015 cm-3, coupled with 0.1 eV reduction in the activation energy of defects. The resultant CQDs photodiodes exhibit a low dark current density of 14 nA cm-2 with a high external quantum efficiency (EQE) of 62%, achieving a linear dynamic range of 98 dB, a -3dB bandwidth of 103 kHz and a detectivity of 4.7 × 1011 Jones. The comprehensive performance of the CQDs photodiodes outperforms previously reported CQDs photodiodes operating at >1.6 µm. By monolithically integrated with thin-film transistor (TFT) readout circuit, the broadband CQDs imager covering 0.35-1.8 µm realizes the functions including silicon wafer perspectivity and material discrimination, showing its potential for wide range of applications.

Colloidal quantum dot materials for next-generation near-infrared optoelectronics.

Colloidal quantum dots (CQDs) are a promising class of materials for next-generation optoelectronic devices, such as displays, LEDs, lasers, photodetectors, and solar cells. CQDs can be obtained at low cost and in large quantities using wet chemistry. CQDs have also been produced using various materials, such as CdSe, InP, perovskites, PbS, PbSe, and InAs. Some of these CQD materials absorb and emit photons in the visible region, making them excellent candidates for displays and LEDs, while others interact with low-energy photons in the near-infrared (NIR) region and are intensively utilized in NIR lasers, NIR photodetectors, and solar cells. In this review, we have focused on NIR CQD materials and reviewed the development of CQD materials for solar cells, NIR lasers, and NIR photodetectors since the first set of reports on CQD materials in these particular applications.

Dual stable isotopes to rethink the watershed-scale spatiotemporal interaction between surface water and groundwater.

The interaction between groundwater and surface water, including their recharge relationship and ratio, is crucial for water cycling, management, and pollution control. However, accurately estimating their spatiotemporal interaction at the watershed scale remains challenging. In this study, we used dual stable isotopes (δ18O, δ2H, d-excess, and lc-excess) and hydrochemistry methods to rethink spatiotemporal interaction at the Yiluo River watershed in central China. We collected 20 groundwater and 40 surface water samples over four periods in two seasons (dry and wet). Our results showed that in the downstream region, groundwater recharged surface water in the dry season while surface water recharged groundwater in the wet season, with average recharge ratios of 89.82% and 90.02%, respectively. In the midstream region, surface water recharged groundwater in both seasons with average ratios of 93.79% and 91.35%. In contrast, in the upstream region, groundwater recharged surface water in both seasons with ratios of 67.35% and 76.89%. Seasonal changes in the recharge relationship between surface water and groundwater in the downstream region also been found. Our findings provide valuable insights for watershed-scale water resource and pollution management.

Inhibition of voltage-gated Hv1 alleviates LPS-induced neuroinflammation via regulation of microglial metabolic reprogramming.

A growing body of evidence highlights the crucial role of metabolic reprogramming in activated immune cells, significantly contributing to both the initiation and progression of neuroinflammation and neurodegenerative diseases. The voltage-gated H channel (Hv1) has been reported to be involved in microglial activation and acts as a key driver of neuroinflammation. This study aimed to explore how Hv1-mediated metabolic reprogramming contributes to neuroinflammation and to assess the therapeutic potential of the Hv1 inhibitor 2-GBI in a model of lipopolysaccharide (LPS)-induced neuroinflammation. We investigated the influence of 2-GBI on the generation of ROS, metabolic reprogramming, and pro-inflammatory mediator production in vitro and examined the therapeutic effect of 2-GBI on microglial activation and hippocampal neuroinflammation in vivo. The results indicated that 2-GBI attenuated the LPS-induced pro-inflammatory response and aerobic glycolysis in microglia, specifically mitigating HIF1α-mediated upregulation of glycolysis. 2-GBI exerted a protective effect against LPS-induced neuroinflammation through HIF1α pathway-regulated aerobic glycolysis. Using a transwell coculture system, we demonstrated that 2-GBI reversed PC12 cell death caused by BV2-mediated neuroinflammation. In vivo experiments further suggested that 2-GBI mitigated neuroinflammatory processes and cognitive dysfunction via microglial metabolic reprogramming. Collectively, our results highlight the potential of Hv1 inhibition as a therapeutic strategy for alleviating LPS-induced neuroinflammation by modulating microglial metabolic reprogramming.

Comprehensive treatment of deep frostbite of multiple fingers after trauma: A case report.

BACKGROUND: Frostbite is becoming increasingly common in urban environments, and severe cases can lead to tissue loss. The treatment goal is to preserve tissue and function; the sooner appropriate treatment is administered, the more tissue can be saved. However, not every patient with deep frostbite seeks medical care promptly. CASE SUMMARY: We report the case of a 73-year-old male patient who was lost in the wilderness for 2 d due to trauma and confusion. He experienced deep frostbite on multiple fingers. Treatment should not be discontinued for patients with deep frostbite who present after the optimum treatment timing. Bullae that no longer form (bloody) blisters within 24 h of aspiration should be removed. Mucopolysaccharide polysulfate cream has clinical value in frostbite treatment. The patient was transferred to Chinese Academy of Medical Sciences and Peking Union Medical College Hospital 12 h after being rescued. The patient had contraindications for thrombolysis, the most effective treatment, due to intracranial hemorrhage and presenting past the optimum treatment timing. We devised a comprehensive treatment plan, which involved delayed use vasodilators and high-pressure oxygen therapy at day 49 post-injury. We experimented with mucopolysaccharide polysulfate cream to treat the frostbite. The aim of the treatment was to safeguard as much tissue as possible. In the end, the fingers that suffered from frostbite were able to be partially preserved. CONCLUSION: The case indicated that patients with severe frostbite who missed the optimal treatment time and had contraindications for thrombolysis could still partially preserve the affected limbs through comprehensive treatment.

Diversity and biogeography of plant phyllosphere bacteria are governed by latitude-dependent mechanisms.

Predicting and managing the structure and function of plant microbiomes requires quantitative understanding of community assembly and predictive models of spatial distributions at broad geographic scales. Here, we quantified the relative contribution of abiotic and biotic factors to the assembly of phyllosphere bacterial communities, and developed spatial distribution models for keystone bacterial taxa along a latitudinal gradient, by analyzing 16S rRNA gene sequences from 1453 leaf samples taken from 329 plant species in China. We demonstrated a latitudinal gradient in phyllosphere bacterial diversity and community composition, which was mostly explained by climate and host plant factors. We found that host-related factors were increasingly important in explaining bacterial assembly at higher latitudes while nonhost factors including abiotic environments, spatial proximity and plant neighbors were more important at lower latitudes. We further showed that local plant-bacteria associations were interconnected by hub bacteria taxa to form metacommunity-level networks, and the spatial distribution of these hub taxa was controlled by hosts and spatial factors with varying importance across latitudes. For the first time, we documented a latitude-dependent importance in the driving factors of phyllosphere bacteria assembly and distribution, serving as a baseline for predicting future changes in plant phyllosphere microbiomes under global change and human activities.

Comparison of multi-DLM approaches for predicting daily runoff: evidence from the data-driven model in one of China's largest wheat production-bases.

Runoff forecasting is extremely important for various activities of water pollution research and agricultural. Data-driven models have been proved an effective approach in predicting daily runoff when combining deep learning methods (DLM). However, predicting accuracy of daily runoff still need improved. Here, we firstly proposed a combined model of Gate Recurrent Unit (GRU) and Residual Network (ResNet) and compared with one shallow learning method (Back Propagation Neural Network, BPNN) and one deep learning method (GRU) with data from 2010 to 2020 in three stations in daily runoff forecasting in the Yiluo River watershed. The results showed that the combined model with precipitation data and runoff data as input has the highest prediction accuracy (NSE = 0.9325, 0.8735, 0.9186, respectively). Input data with precipitation have higher prediction accuracy than that without. The performance of the model was better in the dry season than the wet season. The topographic and geomorphic factors may also the main factors affecting runoff forecast. Those results of this study can provide useful strategies to predict short runoff and manage watershed scale water resources especially in the important agriculture region.

Propofol inhibits neuroinflammation and metabolic reprogramming in microglia in vitro and in vivo.

Microglial activation-induced neuroinflammation is closely related to the development of sepsis-associated encephalopathy. Accumulating evidence suggests that changes in the metabolic profile of microglia is crucial for their response to inflammation. Propofol is widely used for sedation in mechanically ventilated patients with sepsis. Here, we investigate the effect of propofol on lipopolysaccharide-induced neuroinflammation, neuronal injuries, microglia metabolic reprogramming as well as the underlying molecular mechanisms. The neuroprotective effects of propofol (80 mg/kg) in vivo were measured in the lipopolysaccharide (2 mg/kg)-induced sepsis in mice through behavioral tests, Western blot analysis and immunofluorescent staining. The anti-inflammatory effects of propofol (50 µM) in microglial cell cultures under lipopolysaccharide (10 ng/ml) challenge were examined with Seahorse XF Glycolysis Stress test, ROS assay, Western blot, and immunofluorescent staining. We showed that propofol treatment reduced microglia activation and neuroinflammation, inhibited neuronal apoptosis and improved lipopolysaccharide-induced cognitive dysfunction. Propofol also attenuated lipopolysaccharide-stimulated increases of inducible nitric oxide synthase, nitric oxide, tumor necrosis factor-α, interlukin-1ß and COX-2 in cultured BV-2 cells. Propofol-treated microglia showed a remarkable suppression of lipopolysaccharide-induced HIF-1α, PFKFB3, HK2 expression and along with downregulation of the ROS/PI3K/Akt/mTOR signaling pathway. Moreover, propofol attenuated the enhancement of mitochondrial respiration and glycolysis induced by lipopolysaccharide. Together, our data suggest that propofol attenuated inflammatory response by inhibiting metabolic reprogramming, at least in part, through downregulation of the ROS/PI3K/Akt/mTOR/HIF-1α signaling pathway.

Transparent and Flexible Composite Films with Excellent Electromagnetic Interference Shielding and Thermal Insulating Performance.

As the working environment becomes more complex, the visualization of windows in electronic devices increasingly requires transparent and flexible electromagnetic interference (EMI) shielding films. There is a need for materials with EMI shielding properties, while maintaining excellent high light transmission and good thermal insulation. However, the preparation of such multifunctional materials remains challenging due to the respective mechanisms of action of the different properties. Herein, a multilayer structure strategy is proposed to fabricate transparent and flexible indium tin oxide (ITO)/silver nanowire (AgNW) composite films, achieving a multifunctional integration of high light transmission, strong EMI shielding, and good thermal insulation properties of the composite films. Simultaneously, the layered structure was designed and the potential optimization mechanism of the EMI shielding performance of the composite film was analyzed, providing great flexibility for the preparation of transparent composite films. The combination of excellent EMI shielding performance, outstanding near-infrared shielding performance, and high light transmittance makes the ITO/AgNW (IA) composite films promising for abundant potential applications.

Increased Expression of TMPRSS11E Is Involved in LPS- or Poly(I:C)-mediated Inflammation.

Transmembrane serine proteases are a group of enzymes that are implicated in diverse biological processes. Transmembrane serine protease 11E (TMPRSS11E) is expressed in epithelial cells. Previous studies on TMPRSS11E mainly focused on its function in tumor progression. In this study, we investigated the association of TMPRSS11E with the inflammatory response. We found that TMPRSS11E levels were upregulated in the BAL fluid of patients with acute respiratory distress syndrome and LPS-stimulated cultured epithelial cells. In an LPS-induced acute lung injury mouse model and a cercal ligation and puncture-induced sepsis model, increased expression levels of TMPRSS11E were also observed in the lung tissues. Knockdown of TMPRSS11E suppressed the proinflammatory cytokine release and alleviated lung injury. In addition, increased TMPRSS11E expression was detected in lung tissues of poly(I:C)-challenged mice, and overexpression of TMPRSS11E aggregated the lung injury. Unexpectedly, ectopic TMPRSS11E expression in macrophages was observed. Protease-activated receptor-1 was proteolytically activated by TMPRSS11E and enhanced the levels of proinflammatory cytokines. Taken together, our results indicate that TMPRSS11E can be induced during inflammatory response triggered by infection. Therefore, interventions with TMPRSS11E can aid in the treatment of pulmonary inflammation.

Multi-methods to investigate spatiotemporal variations of nitrogen-nitrate and its risks to human health in China's largest fresh water lake (Poyang Lake).

Nitrogen-nitrate contamination has been recognized as the main threaten to the large lake surrounding with the intensive agriculture activities. However, there are still insufficient studies on the interannual evolution of nitrogen-nitrate, source and its impact on the environment and human health in the large fresh water lake. In this study, 248 samplings were collected in Poyang Lake from 2013 to 2018, multi-methods (mathematical statistics method, grey correlation analysis, person correlation analysis and human health risk assessment) to investigate the spatiotemporal variations, impact factors and potential health risks of NO2--N, NO3--N and NH4+-N. The results showed that the middle region had the highest NO2- concentration (mean 0.04µg/l), the northern region had the highest NO3- concentration (mean 1.12 mg/l), and the southern region had the highest NH4+ concentration (mean 0.48 mg/l). For NO3- and NH4+, the concentration was higher than in the wet season. While the concentration of NO2- had the reverse trend. Grey correlation analysis and person correlation analysis results indicate that nitrogen fertilizer, waste water, pH, CODMn and temperature were main factors affecting the nitrogen concentration in Poyang Lake. Health risk assessment results revealed that potential hazards in the study area were acceptable (HR < 1). NO3- provided the highest health risks, and oral ingestion is the major source of local nitrogen health risk. Those results can provide the reference for developing the treatment methods of the international large freshwater lake.

Responses of tree seedlings to understory filtering by the recalcitrant fern layer in a subtropical forest.

The recalcitrant understory fern layer is an important ecological filter for seedling regeneration, yet how the fern layer influences seedling regeneration dynamics remains unclear. Here we transplanted 576 seedlings of four dominant tree species, Castanopsis fargesii, Lithocarpus glaber, Schima superba and Hovenia acerba, to the treatments of Diplopterygium glaucum retention and removal under an evergreen broad-leaved forest in eastern China. We monitored the survival, growth and biomass data of these seedlings for 28 months, and then used generalized linear mixed models to evaluate the treatment effects on seedling survival, growth, biomass and root-shoot ratio. Our results showed that fern retention significantly inhibited the seedling establishment of all four species. During the seedling development stage, the seedling relative growth rate of L. glaber decreased under fern retention, which was not the case for the other three species. Root-shoot ratio of C. fargesii and L. glaber increased significantly under fern retention. Our findings provide new evidence of the filtering effect of a recalcitrant fern understory. Notably, we observed that the response of tree seedlings to the recalcitrant fern understory was more sensitive in the establishment stage. Finally, our work highlights that the filtering effect of the recalcitrant fern understory changes depending on the regeneration stages, and that shade-tolerant species, C. fargesii and L. glaber were even more affected by fern disturbed habitats, suggesting that effective management should attempt to curb forest fern outbreaks, thus unblocking forest recruitment.

Isoschaftoside Inhibits Lipopolysaccharide-Induced Inflammation in Microglia through Regulation of HIF-1α-Mediated Metabolic Reprogramming.

Isoschaftoside is a C-glycosyl flavonoid extracted from the root exudates of Desmodium uncinatum and Abrus cantoniensis. Previous studies suggested that C-glycosyl flavonoid has neuroprotective effects with the property of reducing oxidative stress and inflammatory markers. Microglia are key cellular mediators of neuroinflammation in the central nervous system. The aim of this study was to investigate the effect of isoschaftoside on lipopolysaccharide-induced activation of BV-2 microglial cells. The BV-2 cells were exposed to 10 ng/ml lipopolysaccharide and isoschaftoside (0-1000 µM). Isoschaftoside effectively inhibited lipopolysaccharide-induced nitric oxide production and proinflammatory cytokines including iNOS, TNF-α, IL-1ß, and COX2 expression. Isoschaftoside also significantly reduced lipopolysaccharide-induced HIF-1α, HK2, and PFKFB3 protein expression. Induction of HIF-1α accumulation by CoCl2 was inhibited by isoschaftoside, while the HIF-1α specific inhibitor Kc7f2 mitigated the metabolic reprogramming and anti-inflammatory effect of isoschaftoside. Furthermore, isoschaftoside attenuated lipopolysaccharide-induced phosphorylation of ERK1/2 and mTOR. These results suggest that isoschaftoside can suppress inflammatory responses in lipopolysaccharide-activated microglia, and the mechanism was partly due to inhibition of the HIF-1α-mediated metabolic reprogramming pathway.

ADCK1 is a potential therapeutic target of osteosarcoma.

We here showed that ADCK1 (AarF domain-containing kinase 1), a mitochondrial protein, is upregulated in human osteosarcoma (OS) tissues and OS cells. In primary and established OS cells, ADCK1 shRNA or CRISPR/Cas9-induced ADCK1 knockout (KO) remarkably inhibited cell viability, proliferation and migration, and provoked apoptosis activation. Conversely, ectopic ADCK1 overexpression exerted pro-cancerous activity by promoting OS cell proliferation and migration. ADCK1 depletion disrupted mitochondrial functions in OS cells and induced mitochondrial membrane potential reduction, ATP depletion, reactive oxygen species production. Significantly, ADCK1 silencing augmented doxorubicin-induced apoptosis in primary OS cells. mTOR activation is important for ADCK1 expression in OS cells. The mTOR inhibitors, rapamycin and AZD2014, as well as mTOR shRNA, potently decreased ADCK1 expression in primary OS cells. In nude mice, the growth of subcutaneous pOS-1 xenografts was largely inhibited when bearing ADCK1 shRNA or ADCK1 KO construct. Moreover, ADCK1 KO largely inhibited pOS-1 xenograft in situ growth in proximal tibia of nude mice. ADCK1 depletion, apoptosis activation and ATP reduction were detected in pOS-1 xenografts bearing ADCK1 shRNA or ADCK1 KO construct. Together, the mitochondrial protein ADCK1 is required for OS cell growth and is a novel therapeutic target of OS.

Effects of enhanced hygiene measures on severe diarrhea and anastomotic leak after colorectal cancer surgery: the experience of a tertiary referral hospital in China.

Background: The 2019 novel coronavirus (COVID-19) global pandemic has greatly changed the mode of hospital admissions. This study summarized and analyzed the incidence of severe diarrhea and anastomotic leakage during different periods for colorectal cancer surgery. Methods: From January 2017 to September 2020, 2,619 colorectal operations were performed in Peking Union Medical College Hospital. In contrast with previous years, enhanced hand hygiene training, more frequent ventilation of the wards, and separate bed treatments for patients were implemented in 2020. Data on incidence of severe diarrhea and anastomotic leakage were retrieved and collected. Results: The number of cases of severe diarrhea after colorectal surgery was 32 (4.60%), 24 (3.33%), 32 (3.83%), and 11 (2.99%) in 2017, 2018, 2019, and 2020 respectively, while the incidence of anastomotic leakage was 3.30% (23/696), 3.75% (27/720), 2.87% (24/835), and 2.17% (8/368), respectively. There was no significant difference in the incidence of postoperative severe diarrhea or anastomotic leakage across the various years. Conclusions: The number of colorectal surgeries in 2020 was significantly decreased due to the COVID-19 pandemic. Among the different years, no difference was observed regarding the incidence of postoperative flora disorder or anastomosis leakage. Enhanced hygiene measures during the COVID-19 epidemic partially contributed to the decrease of severe diarrhea and anastomotic leakage.

Deep learning enabled inverse design of nanocrystal-based optical diffusers for efficient white LED lighting.

Angular color uniformity and luminous flux are the most important figures of merit for a white-light-emitting diode (WLED), and simultaneous improvement of both figures of merit is desired. The cellulose-nanocrystal (CNC)-based optical diffuser has been applied on the WLED module to enhance angular color uniformity, but it inevitably causes the reduction of luminous flux. Here we demonstrate a deep-learning-based inverse design approach to design CNC-coated WLED modules. The developed forward neural network successfully predicts two figures of merit with high accuracy, and the inverse predicting model can rapidly design the structural parameters of CNC film. Further explorations taking advantage of both forward and inverse neutral networks can effectively construct the coating layer for WLED modules to reach the best performance.