Redefining artificial lighting through spectral engineering of light sources for well-being | Scientific Reports

Scientific Reports volume 14, Article number: 26298 (2024) Cite this article Metrics details Light-emitting diodes (LEDs) have revolutionized artificial lighting, but also exposed the detrimental

Scientific Reports volume 14, Article number: 26298 (2024) Cite this article

Metrics details

Light-emitting diodes (LEDs) have revolutionized artificial lighting, but also exposed the detrimental health effects that stem from insufficient exposure to natural light. Human-centric artificial lighting requires both visual quality and circadian lighting performance that mimics daylight’s evolving spectral power distribution (SPD). Here, we present a color-tunable LED-based light source that achieves SPDs similar to various conditions of daylight and incandescent lighting over the range of visible wavelengths. This light source is comprised of a linear combination of light converter channels containing dyes exhibiting thermally activated delayed fluorescence (TADF) fabricated through additive manufacturing, photoexcited by violet-emitting LEDs (VLED). This hybrid light source establishes a new benchmark for state-of-the-art artificial lighting at approximating daylight, with Illuminating Engineering Society (IES) color gamut index Rg values within 3%, IES color fidelity index Rf values within 7% through CCT values ranging from 4277 K to 22,333 K. We propose efficiency metrics to accurately quantify similarity between light sources and the respective reference daylight spectrum encompassing visual and circadian effects, facilitating WLED benchmarking. The efficiency metrics pertaining to circadian lighting performance remain within 10% over the same CCT range. These results advance lighting science to address simultaneously the grand challenges of health and sustainability.

Light is vital to human life. Artificial lighting has evolved over thousands of years from using candlelight, to electric incandescent light sources over the last century, and most recently, ushering in the LED revolution of the past 30 years, driven by the aspect of sustainability pertaining to energy efficiency as well as increased awareness over the physiological and/or psychological effect upon humans1,2,3,4,5.

Although natural daylight is essential for optimal health6,7,8, Americans spend over 90% of their entire lives indoors according to the Environmental Protection Agency, a figure even higher for countries with harsh climate, such as the United Arab Emirates9. It is not always feasible to provide outdoor exposure to all building occupants across varying climates, nor daylighting in energy efficient buildings due to heating and glare considerations10,11,12.

Throughout the day, daylight’s evolving spectral power distribution (SPD) provides cues that regulate vital neurobiological functions, by way of the light incident on the back of the eye. These effects go beyond the visual perception of color, based on light reflected by surfaces and objects, which can induce fatigue and eye strain13. Health effects from circadian health disruption include metabolic, endocrine, and circadian cycle regulation14, leading to affected mood, alertness15, cancer rates16. Therefore, a human-centric lighting solution must possess high fidelity of both visual and circadian effects with respect to natural light.

Despite progress in color-tunable technologies to date, no lighting technology can replicate all of the characteristics of daylight, such as illuminance values exceeding 100,000 lx17. For this study, the scope is limited only to reproducing the lighting performance metrics corresponding to the SPD of natural light, over the range of visible wavelengths at a luminance relevant to indoor daytime lighting, i.e. 250 photopic lux. Furthermore, achieving high performance as measured by standard metrics used in evaluating artificial lighting, such as the correlated color temperature (CCT) and color rendering index (CRI), is insufficient to develop artificial lighting that approximates the characteristics of natural daylight18. State-of-the-art color mixed LEDs (cm-LED) can be electronically controlled to match the color coordinates for a given CCT, but differ significantly in circadian lighting and color rendering.

For instance, when an amber-emitting phosphor-converted LED (pc-LED) channel is added to a 3-channel cm-LED, producing an RGBA (red, green, blue, alpha) LED, circadian lighting match to an incandescent 2700 K SPD is improved to suitable levels for minimizing light pollution and circadian disruption, but color rendering at CCT of 6504 K remains low with IES (Illuminating Engineering Society) color fidelity index Rf values near 70. Even when a fifth LED channel is added, the CRI remains under 90, and the energy related to circadian lighting exceeds that of daylight with CCT of 6504 K by over 10%, at the same luminance and CCT. Linear combination of pc-LED channels can achieve tunable circadian lighting, such as the product named BIOS, but their CCT can only be tuned between 2700 K and 3500 K, with a reported CRI of 80 for the 3500 K condition. Other pc-LEDs can approximate sunlight with various CCT adofferings by employing around 12 phosphors, accurately reproducing color rendering and circadian lighting performance, but the CCT value of the LED chip is fixed at manufacturing time, prohibiting SPD tunability during operation. Similarly, LEDs using Cu(I) halides conversion layers reach CRI values ranging from 62.3 to 73.9 but lack SPD tunability19. Hybrid organic/inorganic LEDs employing III-V blue LEDs with organic light converters20,21,22,23 can achieve very high CRI values of 95.720, but lack SPD tunability. On the other hand, white organic light-emitting diodes (WOLED)24 show advantages in reduction of blue light at night and a diffuse light distribution25,26, but their tandem layer stack architecture prevents SPD tunability.

(a) images of light converters for each component TADF-WLED, (b) Photoluminescence (PL) spectrum of each TADF light-converter including electroluminescent (EL) spectrum of 415 nm VLED with accompanying photographs of devices under 415 nm VLED illumination, and (c) rendering of device cross-section showing representation of remote TADF dye suspended in photopolymer matrix, VLED chip model used for illumination, path length dL between light source and light converter layer of 2 mm, and total light converter volume of 13 mm3, and (d) image of prototype device with current driver, and diffuser layer for mixing the channel light outputs.

The cool white WOLED option’s SPD reaches a melanopic equivalent daylight illuminance (melanopic EDI) that is 68% of that of daylight with a CCT value of 6,504 K, at the same luminance of 250 photopic lux. These limitations for WLED and WOLED technologies in enabling human-centric lighting, are intrinsic to the photophysics of their materials and device architectures.

Here, we present a novel electronically spectrally-tunable hybrid organic/inorganic light source, consisting of a linear combination of spectra emitted by light converter channels based on thermally-activated delayed fluorescence (TADF) dyes photoexcited by violet-emitting LEDs (VLED). Furthermore, these light converters are fabricated through additive manufacturing to enable ease of integration. Hereon, we refer to this hybrid device as TADF-WLED, shown in Fig. 1. TADF-WLED differs from state-of-the-art color-tunable WLEDs in that they allow approximating targeted α-opic EDI where α = s, m,l, indexes corresponding to short (s), medium (m), and long (l) wavelength receptors, and α can also refer to the melanopsin-containing or rhodopsin receptors in the human eye, relevant to color rendering and circadian lighting performance.

This linear combination is mathematically described in Eq. (1) below with device compositions found in the “Methods” section, Table 1.

In the TADF-WLED all current is injected into VLED with a peak wavelength of 415 nm, exploiting the higher wall-plug efficiency of VLED compared to green- and amber-emitting LEDs, lower by 53.9% and 85.5% respectively27, serving as a proof of concept establishing a roadmap towards high efficiency lighting systems surpassing 390 lm/W based on the high Stokes Shift that is characteristic of TADF emitters compared to traditional fluorescent and phosphorescent emitters, as shown graphically in (Supplementary Figs. 1 and 2) and in tabulated form in (Supplementary Table 1), with high color-rendering and circadian lighting performance. This addresses the current tradeoff between color rendering and energy efficiency28, with the procedure detailed in Supplementary Note 1 and the TADF-WLED theoretical power efficacy values tabulated on (Supplementary Table 2) for different light-converter efficiency values.

The TADF molecules were chosen due to their potential of achieving high photoluminescence efficiency, reduced dimerization favorable for photostability23,29,30,31, and because their inherent disorder and 3D molecular shapes result in broad spectral emission compared to fluorescent emitters, which is favorable for solid-state lighting. Additionally, both TADF molecules and stereolithography are reputed to be low-cost methods due to decreased complexity in processing and low cost of raw materials32,33. The reason for the spectrally broad emission can be attributed to the charge-transfer characteristic of the lowest energy excited state34.

TADF-WLED achieves an SPD approximating that of daylight as described by CIE Standard Illuminant D, with CCT ranging from 4277 K to 22,333 K, corresponding to sunlight at different times of the day, varying climates, and weather conditions such as clear morning, mid-day, sunset, overcast daylight, and the northern sky. In addition, the TADF-WLED provides an approximation comparable to the state-of-the-art color tunable LEDs to an incandescent light source as described by CIE Standard Illuminant A, an illuminant of interest towards reducing sky glow, light pollution, and disruption of nocturnal ecosystems. To the best of our knowledge, no other color-tunable LED technology achieves this function. All reference SPDs were sourced as outlined in Supplementary Note 2 and scaled to a luminance value of 250 photopic lux, shown in (Supplementary Fig. 4).

The degree to which an LED’s SPD delivers energy to each relevant photoreceptor in the eye can be quantified by metrics introduced over the last decade by the CIE, such as α-opic EDI in units of (cdm− 1), defined in CIE Standard S 026 (Supplementary Note 3, Eq. 13 through 16). However, for benchmarking purposes to compare different light sources, we propose the introduction of a new figure of merit that describes the α-opic EDI Efficiency in units of (%), or how closely the EDI of the light source under evaluation approximates that of the relevant reference illuminant, given the same illuminance of X photopic lux (defined in Eq. 2 below).

These metrics must be computed for each photoreceptor within the human eye relevant to the application in order to validate that these metrics capture visual and circadian characteristics of lighting. Color rendering was characterized using IES Standard TM-30-15, which measures both color fidelity index Rf and color gamut index Rg, to describe how a light source renders the hue of objects, as well as the shift in hue with respect to reference light source, evaluated over 99 color samples representing many objects. The color fidelity index, Rf, describes how accurately the hue of color will be reproduced similar to CRI but covering a larger color sample pallet, and the color gamut index Rg describes how saturated or de-saturated the colors appear with respect to the reference illuminant. Values for Rf and Rg are found in Supplementary Table 3. CRI, or CIE General Color Rending Index Ra, is reported in Supplementary Table 4 but priority is given to Rf and Rg metrics, as CRI is limited to a single value for fidelity and uses only 8 color samples that are less representative of the infinite color possibilities encountered by users. No further processing is performed on Rf and Rg as they are already a ratio between the light source in evaluation and the reference illuminant. Values of Rf and Rg can exceed 100, but this is not desirable in this application, since it translates to higher-than-natural color hue distortion or saturation. Details of computation of all color rendering metrics are described in Supplementary Note 4. The high values in Rf seen with the TADF-WLED translate to high CRI values, so the same conclusions can be drawn from both values.

Circadian lighting performance was characterized through a figure of merit based on circadian stimulus (CS), circadian lighting (CLA), and equivalent melanopic lux (EML). These metrics are defined based on standards1,4,35 that account for sensitivity of melanopsin photopigment contained within the ipRGC, and CS and CLA take into account the effect of the transduction of light through the front of the eye. Supplementary Note 5 contains the procedures for computing the CS, CLA, and EML metrics. To better enable benchmarking how closely the light source approximates the corresponding reference illuminant, we define a set of figures of merit describing as CS Efficiency, CLAEfficiency, and EML Efficiency given by Eqs. (3)–(5).

The TADF light converters were fabricated individually as shown in “Methods” section, “Materials” and “Fabrication” subsections, and the linear combination of the channel SPDs was achieved in practice by an embedded computing platform built in-house and reported by our group36, utilizing open-source hardware and software. The SPD tuning was achieved by manually tuning the currents in MATLAB for simulated SPDs with luminance scaled to 250 photopic lux.

The choice of utilizing additive manufacturing expands this technology’s ability to serve a broader base of stakeholders. It contributes towards advancing the goal of the US Department of Energy (DOE) to develop all-3D printed luminaires as well as reducing carbon footprint by 75% in material processing, 28% at transport, and 27% at end of life compared to traditional machined luminaire manufacturing at those stages respectively, according to a study conducted by Signify for the DOE, utilizing the standard Impact Assessment Carbon footprint IPCC 2013 GWP 10037. TADF-WLED employ all-organic light-emitting molecules that carry a smaller energy footprint and toxicity for the device through its full life cycle. The materials used in this study for example, were synthesized at low temperatures under 200 °C38,39,40,41 and are free of rare-earth elements. 3D printing opens avenues towards democratizing lighting manufacturing, promoting equity by empowering makers and lighting designers to deploy lighting systems, whether they are makers at home, independent designers, local manufacturers, or large firms. For lighting scientists, additive manufacturing of light sources provides degrees of freedom to experiment with different light distributions. Optical structures can be printed directly on device to improve light delivery, outcoupling, and scattering36, enabling limitless form factors and larger surface areas.

For benchmarking to quantify the performance of TADF-WLED, the SPD of the TADF-WLED was first compared to the SPD of an RGBA LED as found by the IES-TM-13 Color Rendering Toolbox. The SPDs were compared under low and a high CCT conditions, with the reference spectra of an incandescent bulb and daylight with a CCT of 6504 K respectively. The SPDs were equalized to a luminance of 250 photopic lux using MATLAB, representative of office lighting conditions. Rhodopsin sensitivity was excluded because it corresponds to vision in low-light conditions under 1 lx. The results of this benchmark are found in Fig. 2.

Bar graphs of the s-opic EDI Efficiency and melanopic EDI Efficiency were the focus of this analysis as m-opic Efficiency and l-opic EDI Efficiency were comparable since they contribute spectrally primarily to photopic vision, and luminance was fixed. For the incandescent SPD, the results show that the TADF-WLED s-opic EDI Efficiency is near 100% while for RGBA it is around 38%. In consequence, both Rf and Rg show a gain of nearly 20% for the TADF-WLED. Both have comparable CS Efficiency and CLAEfficiency, showing that no undesirable energy content is being introduced that can exacerbate melatonin suppression for nighttime use. For daylight at CCT of 6504 K, there are also comparable gains in color rendering. Melanopic EDI Efficiency is improved by over 20%, improving CS Efficiency and CLAEfficiency by 5% and 10% respectively, which is correlated to melatonin suppression potential. Full lists of values are found in Supplementary Table 3 for TADF-WLED approximating incandescent light, Supplementary Table 5 for TADF-WLED approximating daylight 6504 K, and Supplementary Table 7 for RGBA.

The next benchmark that was performed was the stability of the α-opic EDI Efficiency. For this benchmark, strictly III-V LED channels were used for the cm-LED as they are the long-term strategy for efficient color-tunable LED technology. This analysis is shown in Fig. 3. Fifty sampled SPD variants were synthesized numerically in MATLAB by varying the optical power of each channel within a tolerance of 10%. The TADF-WLED had comparable standard deviation in the m-opic EDI Efficiency, l-opic EDI Efficiency, and melanopic EDI Efficiency, while retaining more stability in the s-opic EDI Efficiency, with a standard deviation of melanopic EDI Efficiency, or σmel of 3.8 compared to 6.1 for cm-LED. This is likely the byproduct of operating the blue III-V LED at a higher optical power to supply the correct amount of energy for s-cone photoreceptor, a necessary design choice to overcome the small FWHM of the emission spectrum of the LED. Supplementary Note 6 details the effect of variations of component channel optical power for an SPD comprised of narrow FWHM channels compared to the TADF-WLED design.

Overall, the α-opic EDI Efficiency is slightly more robust to variations in the optical power of the TADF-WLED channels as shown in Fig. 3. This is relevant because during operation, differential efficiency droops of the component III-V LEDs, differential aging, and LED driver variations all create a challenge for controls, posing a roadblock to adoption of network connected lighting systems.

This analysis was performed on III-V semiconductor 5 channel component cm-LED. While the devices resulted in improved performance α-opic EDI Efficiency metrics, color rendering showed limitations, with Rf values of 86, Rg of 106, and CRI of 83.9. The full list of performance metric values for 4 channel and 5 channel cm-LEDs can be found in Supplementary Table 8. These values were achieved in spite of excellent values for distance to black-body locus Duv = − 4 \(\:\times\:\) 10− 4 and CCT matching reference daylight spectrum of 6504 K, metrics often used for simplicity due to correlation to the visual color appearance of a light source and its proximity to the color appearance of the reference black-body radiator.

SPD of illuminance scaled to 250 photopic lux for TADF-WLED and RGBA LED with reference SPD at same photopic lux used for evaluation of performance metrics, with (a) incandescent SPD with CCT of 2855 K and (b) daylight SPD with CCT 6504 K, and bar graphs of results of benchmark for (c) and (d) s-opic EDI Efficiency and melanopic EDI Efficiency; (e) and (f) color rendering performance metrics compared; and (g) and (h) CS Efficiency, CLAEfficiency, and EML Efficiency benchmarked to reference SPDs of incandescent 2855 K (c), (e), and (g) and daylight 6504 K (d), (f), and (h), with target value of figure of merit 100 denoted by horizontal lines.

Plots showing (a) SPD variations for TADF-WLED and (b) for cm-LED, with insets showing standard deviation of variation per channel optical power across all 50 variant SPDs, and (c) box-and-whisker data for each of the α-opic EDI Efficiencies with corresponding standard deviation for TADF-WLED and (d) box-and-whisker data for each of the α-opic EDI Efficiencies with corresponding standard deviation for cm-LED.

For the final benchmark, the TADF-WLED was compared to a set of illuminants representing natural light (Supplementary Fig. 3). These included measured sun and skylight SPD at sunset with CCT of 4277 K and Illuminant D with CCTs of 5002 K (horizon daylight), 5499 K (mid-morning daylight), 7005 K, and 7507 K (overcast daylight). The SPD of the northern sky was included because it may be of importance to inhabitants of the northern hemisphere. Illuminant A and the black-body spectrum with temperature of 3500 K were chosen because these spectra are useful for minimizing melatonin suppression, light pollution for coastal ecosystems and migratory birds, and astronomical observatories.

The SPDs attained by the TADF-WLED utilized the same 4 light conversion layers and varied only the optical power output by each channel. Figure 4 shows the α-opic EDI Efficiency, color rendering, and CS Efficiency, CLAEfficiency, and EML Efficiency metrics for TADF-WLED approximating the performance of a subset of the illuminants, specifically the early morning (5002 K), mid-morning (5499 K), overcast (7507 K), and sunset (4277 K), illustrating that no value has below a 95% match for these diverse set of conditions, with many of the α-opic EDI Efficiency values staying within ± 3%, and Rg is 100 for 3 of the 4 conditions. This analysis also shows that the EML Efficiency mirrors the melanopic EDI Efficiency, consistent with the definitions based solely on corresponding sensitivity curve.

The plots of the SPD for the TADF-WLED with reference natural light SPD are found in Supplementary Figs. 5 and 6. Supplementary Table 3 contains all the α-opic EDI Efficiency metrics, CS Efficiency, CLAEfficiency, EML Efficiency, Rf, and Rg for the full selection of daylight and incandescent SPDs ranging from CCT of 2700 K to 22,333 K, showing that all of the demonstrated SPD attain high Rf and Rg values above 90 except for an incandescent illuminant.

Plots showing performance metrics of TADF-WLED compared to reference spectra of daylight in early morning (CCT 5002 K), mid-morning (CCT 5499 K), overcast daytime (CCT 7507 K), and sunset skylight (CCT 4277 K) including (a) α-opic EDI Efficiency, (b) color rendering and (c) CS Efficiency, CLAEfficiency, and EML Efficiency.

To further place the performance metrics demonstrated by this prototype, the TADF-WLED was benchmarked against over 1500 state-of-the-art commercially-available WLED including pc-LED and RGBA42, WOLED, and a state-of-the-art lab prototype hybrid rare-earth free WLED with high color rendering20, for the fixed spectrum of daylight with CCT 6504 K. All SPDs were equalized to a luminance of 250 photopic lux. The TADF-WLED performance was comparable with the single top-performing, fixed spectrum pc-LED on the market from Seoul Semiconductors. These analysis results for the top performing LEDs are included in Fig. S7 and Tables S6 and S7.

For manufacturability for a user-product, a resin that can withstand high irradiance without yellowing or photobleaching is needed. Future studies should look at materials classes used in LED packaging for the TADF-containing matrix. The luminous efficiency, defined as the ratio of the illuminance (lm) divided by the total radiant power (W) was computed and appears in Supplementary Table 9. These numbers match what is expected, as with RGBA LEDs there are significant deficits in the energy spectrum caused by their narrow emission spectrum, so there is less optical power emitted over the full visible range of wavelengths. Supplementary Table 10 includes the R9 values for the LEDs in this study, showing that future designs of a TADF-WLED may benefit from an additional red and green channel to better approximate the incandescent spectrum.

The TADF-WLEDs demonstrate the potential for light converters containing TADF molecules fabricated through additive manufacturing for human-centric lighting applications. There are substantial gains compared to state-of-the-art cm-LED in terms of α-opic EDI Efficiency and subsequently color rendering and circadian lighting performance metrics resulting from their broad emission spectrum and independently controlled channels. For future evaluation of light sources, we propose metrics that capture the efficiency of delivering energy to each receptor in the eye compared to the corresponding reference illuminant at the same illuminance, ensuring high color rendering and circadian lighting performance. There are additional advantages in sustainability and technology life cycle. We believe that this work constitutes an unmatched performance milestone for artificial lighting for the benefit of the human condition.

Synthesis of 5,5’-(2,3,5,6-tetra(9 H-carbazol-9-yl)-1,4-phenylene)bis(2-(4-(tert-butyl)phenyl)-1,3,4-oxadiazole), or TCZPBOX, was discussed in40. Synthesis of 5-(2-(4,6-diphenyl-1,3,5-triazin-2-yl)phenyl)-5 H-benzofuro[3,2-c]carbazole, or oBFCzTrz, was discussed in41. 2-[4-(Diphenylamino)phenyl]-10,10-dioxide-9 H-thioxanthen-9-one, or TXO-TPA, was purchased from Luminescence Technologies Corporation. 2,7-bis(9,9-dimethylacridin-10(9 H)-yl)-9,9-dimethyl-9 H-thioxanthene 10,10-dioxide, or DMTDAc, was purchased and used as is from Luminescence Technologies Corporation.

Additional considerations for the materials chosen for this device include photostability and efficiency. Some of the emitters selected for this study have been used for high stability OLED architectures, such as TCZPBOX40 DMTDAc43,44, TXO-TPA38, oBFCzTrz45 have been demonstrated to have low aggregation as well, an approach that may result in high photostability. This approach paves the way for higher device stability since the emitter dyes are operated in photoluminescence (PL), where the usual OLED degradation mechanisms of high-energy triplet interactions30,46,47 and carrier imbalance43 are not present. Future work should study TADF emitters with larger FWHM, as most emitting materials have been developed for display applications requiring narrow spectra with low FWHM for color purity according to a study conducted by the organization NHK STRL for the Japanese government on trends in OLED research and development48.

The device fabrication process begins with creating the CAD file for the encapsulation that houses the light-conversion layer based on the dimensions of the beam spot size of the excitation LED. This process itself also allows the creation of built-in beam-steering and focusing structures, such as lenses. The design was printed in Clear v4 resin by FormLabs, using a FormLabs Form 2 SLA system. With this method, up to 60 samples were fabricated per batch.

Resin formation consisted of blending the Clear v4 resin from FormLabs with the TADF compounds in powder form blended by magnetic stirring bar after sublimation and purification. A magnetic steering plate was used for 12–16 h depending on the material until the resin looked homogenous. All samples were stirred in amber vials to prevent photoinitiated cross-linking. This resin was chosen because it was readily available, relatively inexpensive, and compatible with high-throughput, high yield scalable additive manufacturing. After resin was filled into the encapsulation, the encapsulation was cured under 405 nm LED irradiance for 75 min at a temperature of 60 °C using a FormLabs Form Cure station.

Each resin sample was characterized by utilizing a 415 nm peak wavelength LED, for photoexcitation which was also used as the excitation source for the final WLED. The spectrum was then collected utilizing an OceanOptics USB2000 spectrometer. For each lab bench prototype device consisting of a VLED and encapsulated TADF light conversion layer was controlled by Keithley 2400. VLED model A007-UV410-48 was used, though in principle, other VLEDs with emission peak wavelength near 415 nm suitable for high power applications can be used.

Samples e-3, e-34, e-89, and e-100 were irradiated with an estimated 30.3 mW, 30.3 mW, 30.3 mW, and 57.8 mW respectively for spectrum characterization, corresponding to operating currents of 1 to 2 mA. Fabrication is depicted in Supplementary Fig. 8.

Data set of over 1500 WLED SPD data was downloaded as provided by measurements42.

WOLED SPD data was collected by measuring OLEDWorks OLED panels, measured in-house using OceanOptics USB2000 and powered by the OLEDWorks SDK which includes current driver and power supply.

Seoul Semiconductor Sunlike WLED data was digitized from datasheet using Origin 2020, datasheet available online at http://www.seoulsemicon.com/en/product/spec/_91_SunLike_92_%20SAWS0661A/37files/1469/37.html.

Hybrid rare earth free WLED consists of 2 organic emitters in PMMA matrix, data was digitized using Origin 2020 from plots in source publication20.

All datasets were equalized by scaling such that luminous power was 250 photopic lux.

All the data that support the findings of this study are included in the main text and “Supplementary Information.” The data are available from the corresponding authors upon reasonable request.

Rea, M. S., Figueiro, M. G., Bierman, A. & Bullough, J. D. Circadian light. J. Circadian Rhythms. 8, 2. https://doi.org/10.1186/1740-3391-8-2 (2010).

Article PubMed PubMed Central Google Scholar

Royer, M. P., Wei, M., Wilkerson, A. & Safranek, S. Experimental validation of colour rendition specification criteria based on ANSI/IES TM-30-18. Light Res. Technol. 52, 323–349. https://doi.org/10.1177/1477153519857625 (2019).

Article Google Scholar

Miller, N. J. & Irvin, A. C. LD + Magazine 50, 12–14 (2020).

Google Scholar

Lucas, R. J. et al. Measuring and using light in the melanopsin age. Trends Neurosci. 37, 1–9. https://doi.org/10.1016/j.tins.2013.10.004 (2014).

Article CAS PubMed Google Scholar

Ohno, Y. Practical use and calculation of CCT and Duv. LEUKOS. 10, 47–55. https://doi.org/10.1080/15502724.2014.839020 (2014).

Article Google Scholar

Lam, R. W. Seasonal Affective Disorder and beyond: Light Treatment for SAD and non-SAD Conditions (American Psychiatric, 1998).

Shishegar, N. & Boubekri, M. Lighting up living spaces to improve mood and cognitive performance in older adults. J. Environ. Psychol. 82, 101845. https://doi.org/10.1016/j.jenvp.2022.101845 (2022).

Article Google Scholar

Fernandez, F. X. Current insights into optimal lighting for promoting sleep and circadian health: brighter days and the importance of sunlight in the built environment. Nat. Sci. Sleep. 14, 25–39. https://doi.org/10.2147/NSS.S251712 (2022).

Article PubMed PubMed Central Google Scholar

Jung, C. & Awad, J. The improvement of indoor air quality in residential buildings in Dubai, UAE. Buildings. 11, 250–265 (2021).

Article Google Scholar

Reinhart, C., Dogan, T., Jakubiec, J. A., Rakha, T. & Sang, A. In Proceedings of BS: 13th Conference of International Building Performance Simulation Association, 476–483 (2013).

Shen, H. & Tzempelikos, A. Sensitivity analysis on daylighting and energy performance of perimeter offices with automated shading. Build. Environ. 59, 303–314 (2013).

Article Google Scholar

Johnson, R. et al. Glazing energy performance and design optimization with daylighting. Energy Build. 6, 305–317 (1984).

Article Google Scholar

Hou, D., Ge, Y. & Lin, Y. In 2018 15th China International Forum on Solid State Lighting: International Forum on Wide Bandgap Semiconductors China1–4 (SSLChina: IFWS, 2018). https://doi.org/10.1109/IFWS.2018.8587405.

Mason, I. C. et al. Light exposure during sleep impairs cardiometabolic function. Proc. Natl. Acad. Sci. U.S.A. 119, e2113290119. https://doi.org/10.1073/pnas.2113290119 (2022).

Smolders, K. C. H. J., de Kort, Y. A. W. & Cluitmans, P. J. M. A higher illuminance induces alertness even during office hours: findings on subjective measures, task performance and heart rate measures. Physiol. Behav. 107, 7–16. https://doi.org/10.1016/j.physbeh.2012.04.028 (2012).

Article CAS PubMed Google Scholar

Masri, S. & Sassone-Corsi, P. The emerging link between cancer, metabolism, and circadian rhythms. Nat. Med. 24, 1795–1803. https://doi.org/10.1038/s41591-018-0271-8 (2018).

Article CAS PubMed PubMed Central Google Scholar

Kim, C. H. & Kim, K. S. Development of sky luminance and daylight illuminance prediction methods for lighting energy saving in office buildings. Energies. 12, 592 (2019).

Article Google Scholar

Durmus, D. Correlated color temperature: use and limitations. Lighting Res. Technol. 54 (4), 363–375. https://doi.org/10.1177/14771535211034330 (2022).

Article Google Scholar

Fang, Y. et al. A systematic approach to achieving high performance hybrid lighting phosphors with excellent thermal- and photostability. Adv. Funct. Mater. 27, 1603444. https://doi.org/10.1002/adfm.201603444 (2017).

Article CAS Google Scholar

Menéndez-Velázquez, A., Morales, D. & García-Delgado, A. B. Sunlike white light-emitting diodes based on rare-earth-free luminescent materials. Materials. 15, 1680 (2022).

Article ADS PubMed PubMed Central Google Scholar

Wen, Y. et al. Introduction of red-green-blue fluorescent dyes into a metal-organic framework for tunable white light emission. Adv. Mater. 29, 1700778. https://doi.org/10.1002/adma.201700778 (2017).

Article CAS Google Scholar

Findlay, N. J. et al. An organic down-converting material for white-light emission from hybrid LEDs. Adv. Mater. 26, 7290–7294. https://doi.org/10.1002/adma.201402661 (2014).

Article CAS PubMed PubMed Central Google Scholar

Di Martino, D. et al. Straightforward fabrication of stable white LEDs by embedding of inorganic UV-LEDs into bulk polymerized polymethyl-methacrylate doped with organic dyes. Sci. Rep. 4, 4400. https://doi.org/10.1038/srep04400 (2014).

Article CAS PubMed PubMed Central Google Scholar

Tang, C. W. & VanSlyke, S. A. Organic electroluminescent diodes. Appl. Phys. Lett. 51, 913–915. https://doi.org/10.1063/1.98799 (1987).

Article ADS CAS Google Scholar

Jou, J. H., Hsieh, C. Y., Chen, P. W., Kumar, S. & Hong, J. Candlelight style organic light-emitting diode: a plausibly human-friendly safe night light. J. Photonics Energy. 4, 043598 (2014).

Article ADS CAS Google Scholar

Mou, X., Narendran, N., Zhu, Y. & Freyssinier, J. P. SPIE. In Fifteenth International Conference on Solid State Lighting and LED-based Illumination Systems, vol. 9954, 995403. https://doi.org/10.1117/12.2240465 (2016).

Pattison, M. P., Hansen, M. & Tsao, J. Y. LED lighting efficacy: status and directions. C. R. Phys. 19, 134–145. https://doi.org/10.1016/j.crhy.2017.10.013 (2018).

Article ADS CAS Google Scholar

Royer, M. Evaluating tradeoffs between energy efficiency and color rendition. OSA Contin. 2, 2308–2327. https://doi.org/10.1364/OSAC.2.002308 (2019).

Article CAS Google Scholar

Kotadiya, N. B., Blom, P. W. M. & Wetzelaer, G. J. A. H. Efficient and stable single-layer organic light-emitting diodes based on thermally activated delayed fluorescence. Nat. Photonics. 13, 765–769. https://doi.org/10.1038/s41566-019-0488-1 (2019).

Article ADS CAS Google Scholar

Dias, F. B., Penfold, T. J. & Monkman, A. P. Photophysics of thermally activated delayed fluorescence molecules. Methods Appl. Fluoresc. 5, 012001. https://doi.org/10.1088/2050-6120/aa537e (2017).

Article ADS CAS PubMed Google Scholar

Chen, X. K., Kim, D. & Brédas, J. L. Thermally activated delayed fluorescence (TADF) path toward efficient electroluminescence in purely organic materials: molecular level insight. Acc. Chem. Res. 51, 2215–2224. https://doi.org/10.1021/acs.accounts.8b00174 (2018).

Article CAS PubMed Google Scholar

Xie, X. et al. Molecular engineering enables TADF emitters well suitable for non-doped OLEDs with external quantum efficiency of nearly 30%. Adv. Funct. Mater. 32, 1616–3028. https://doi.org/10.1002/adfm.202112881 (2022).

Article CAS Google Scholar

Lee, M. et al. Development of a 3D printer using scanning projection stereolithography. Sci. Rep. 5, 9875. https://doi.org/10.1038/srep09875 (2015).

Article CAS PubMed PubMed Central Google Scholar

Ansari, R., Shao, W., Yoon, S. J., Kim, J. & Kieffer, J. Charge transfer as the key parameter affecting the color purity of thermally activated delayed fluorescence emitters. ACS Appl. Mater. Interfaces. 13, 28529–28537. https://doi.org/10.1021/acsami.1c02943 (2021).

Article CAS PubMed Google Scholar

Rea, M. S. Value Metrics for Better Lighting Appendix, vol. 3 (SPIE, 2012).

Lubin, L., Nwokedi, I., Diwan, Y., Moreno, O. & Kippelen, B. In 2022 Opportunity Research Scholars Symposium, 49–52 (IEEE, 2022).

Waumans, L. I. 2022 Solid-State Lighting Workshop (U.S. DOE, 2022).

Wang, H. et al. Novel thermally activated delayed fluorescence materials–thioxanthone derivatives and their applications for highly efficient OLEDs. Adv. Mater. 26, 5198–5204. https://doi.org/10.1002/adma.201401393 (2014).

Article CAS PubMed Google Scholar

Adachi, C. Third-generation organic electroluminescence materials. Jpn. J. Appl. Phys. 53, 060101 (2014).

Article ADS Google Scholar

Zhang, X. et al. Host-free yellow-green organic light-emitting diodes with external quantum efficiency over 20% based on a compound exhibiting thermally activated delayed fluorescence. ACS Appl. Mater. Interfaces. 11, 12693–12698. https://doi.org/10.1021/acsami.8b18798 (2019).

Article CAS PubMed Google Scholar

Zhang, X. et al. High performance blue-emitting organic light-emitting diodes from thermally activated delayed fluorescence: a guest/host ratio study. J. Appl. Phys. 124, 055501. https://doi.org/10.1063/1.5041447 (2018).

Article ADS CAS Google Scholar

Kokka, A. et al. Development of white LED illuminants for colorimetry and recommendation of white LED reference spectrum for photometry. Metrologia. 55, 526–534. https://doi.org/10.1088/1681-7575/aacae7 (2018).

Article ADS CAS Google Scholar

Cai, X. & Su, S. J. Marching toward highly efficient, pure-blue, and stable thermally activated delayed fluorescent organic light‐emitting diodes. Adv. Funct. Mater. 28, 1802558 (2018).

Article Google Scholar

Lee, I. & Lee, J. Y. Molecular design of deep blue fluorescent emitters with 20% external quantum efficiency and narrow emission spectrum. Org. Electron. 29, 160–164 (2016).

Article CAS Google Scholar

Abroshan, H., Cho, E., Coropceanu, V. & Brédas, J. L. Suppression of concentration quenching in Ortho-substituted thermally activated delayed fluorescence emitters. Adv. Theory Simul. 3, 1900185. https://doi.org/10.1002/adts.201900185 (2020).

Article CAS Google Scholar

Adachi, J., Kakizoe, H., Tsang, P. K. D. & Endo, A. In Soc. Inf. Disp. Int. Symp., vol. 50, 95–98. https://doi.org/10.1002/sdtp.13399 (2019).

Endo, A. et al. Thermally activated delayed fluorescence from Sn4+–porphyrin complexes and their application to organic light-emitting diodes - a novel mechanism for electroluminescence. Adv. Mater. 21, 4802–4806. https://doi.org/10.1002/adma.200900983 (2009).

Article CAS PubMed Google Scholar

Shimizu, T. & Fukagawa, H. Trends in research and development of OLED. Broadcast. Technol. NHK STRL. 74, 3 (2018).

Google Scholar

Download references

The authors would like to acknowledge Dr. Yadong Zhang and Dr. Seth Marder for synthesis of some of the TADF dyes used in this study, Jingwei Yang for assistance with building database of materials. Shashwati Da Cunha for assistance with some of the samples fabricated, and Dr. Silja Abraham for suggestions with processing parameters for the resins. This work was enabled in part by the U.S. Department of Energy’s Office of Energy Efficiency and Renewable Energy (EERE) under the Solid-State Lighting Program (Award Number DE-EE0008205), and by a Small Bets Seed Grant Award by the Office of the Executive Vice-President of Research at Georgia Tech.

Center for Organic Photonics and Electronics (COPE), School of Electrical and Computer Engineering, Georgia Institute of Technology, Atlanta, GA, 30332, USA

O. Moreno & B. Kippelen

Department of Electrical and Computer Engineering, Northeastern University, Boston, MA, 02115, USA

C. Fuentes-Hernandez

You can also search for this author in PubMed Google Scholar

Conceptualization: O.M., C.F., B.K. Methodology: O.M., C.F. Fabrication process: O.M. Visualization and graphics: O.M. Figures: O.M. Data processing and analysis: O.M. Funding acquisition: C.F., B.K., O.M. Project administration: C.F. Supervision: O.M., B.K. Manuscript preparation: O.M., B.K., C.F. Proofread: O.M., B.K.

Correspondence to B. Kippelen.

The authors declare no competing interests.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Below is the link to the electronic supplementary material.

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

Moreno, O., Fuentes-Hernandez, C. & Kippelen, B. Redefining artificial lighting through spectral engineering of light sources for well-being. Sci Rep 14, 26298 (2024). https://doi.org/10.1038/s41598-024-78315-4

Download citation

Received: 24 June 2024

Accepted: 30 October 2024

Published: 01 November 2024

DOI: https://doi.org/10.1038/s41598-024-78315-4

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

Govee COB Strip Light Pro Review: An Elegant Light For Moods and Vibes

The best gaming PC 2024: Take the pre-built route to greatness | GamesRadar+

News

Redefining artificial lighting through spectral engineering of light sources for well-being | Scientific Reports