Workshop Presentations and Materials

Nearly 900 researchers, manufacturers, industry insiders, and interested observers from the U.S. and abroad gathered for DOE’s 18th annual — and first virtual — Lighting R&D Workshop, held February 1–4, 2021. The workshop, a cornerstone of DOE Lighting R&D program planning, was once again co-sponsored by the Illuminating Engineering Society (IES), leveraging the duo’s long-term partnership to advance the quality and efficiency of lighting through science and engineering.

The workshop’s virtual format and free registration was an opportunity for DOE to reach a much larger audience, welcoming participation not only from stalwart supporters but also from new stakeholders in more sectors, such as small businesses, national laboratories, and research labs overseas. With an online audience nearly three times that of previous in-person events, DOE designed an agenda aimed at optimizing discussion and input while providing the thought-provoking content attendees have come to expect from DOE workshops.

Prevalent throughout the agenda was the theme of “meeting the moment.” Just as recent events have shifted how we live and work, lighting technology is shifting as well, redefining lighting and related energy savings, and improving our understanding of how light affects human well-being and what functions and services lighting can provide. From initial design through operation and maintenance, from devices to buildings to cities — what research is needed to meet this moment?


Day 1 began with introductory remarks from Brian Walker, Lighting R&D program manager, David Nemtzow, director of the DOE Building Technologies Office (BTO) within the Office of Energy Efficiency and Renewable Energy (EERE), and Brian Liebel, IES director of standards and research. Noting that buildings are key to decarbonizing the U.S. economy, Nemtzow pointed out that DOE has been working for years to research, develop, and deploy innovative technologies to make our buildings more efficient. A key part of that future will be driving advances in lighting and control technologies and connected lighting systems in commercial buildings—where long operating hours, controllability, and networked capabilities offer the greatest opportunity for energy, cost, and carbon savings. He also congratulated the winners of the 2021 student poster competition: Ruqayah Bhuiyan, University of Georgia; Yunping Huang, University of Washington; and J. Mundinger, Penn State University.


Chuck Swoboda, innovator and former CEO of Cree Lighting, gave the first keynote talk. Recounting lessons learned from Cree’s journey from startup to global technology company, he added pointed insight on the barriers to innovation and what, particularly, it will take to drive lighting to the next level. Acclaimed lighting designer Star Davis gave the second keynote, sharing her passion for process innovation derived from her experiences at Arup and most recently at WeWork, where she was global head of lighting.


DOE advisor Morgan Pattison moderated the first panel, asking four experts to compare existing conventional lighting systems against their ideal. As he described, new understanding of human physiological responses, productivity, and safety impacts of light require reconsideration of lighting performance targets, as do new lighting technological capabilities in terms of spectral power distribution, optical control, and intensity control. Moreover, new computational models are necessary to design and predict lighting performance based on these new capabilities to satisfy our understanding of ideal or optimum lighting conditions.  

Jennifer Scheib of the University of Colorado Boulder addressed how ideal office lighting must support the health and well-being of workers and reduce energy consumption to achieve zero-energy buildings. Although current office lighting meets existing lighting targets for visual function using efficient light sources, daylighting, and careful lighting design, future office lighting designs must consider higher light levels for health and well-being functions and climate change mitigation and resilience goals. Improved design metrics and technology improvements remain necessary to advance toward healthy, productive, and zero-energy buildings.

Tero Mäkinen of LEDiL discussed the role of optics in achieving ideal lighting by showing the range of lighting aspects that must be considered and how improved optical technologies can help with control and delivery. New optics can reduce unneeded light, thereby reducing glare, increasing uniformity, and increase system efficiency. He described how lower-efficacy lights with improved optical control can result in reduced energy consumption while reaching the target illuminance. He added that in order to achieve ideal optical control, we must understand where light is needed and balance source efficiency with optical control to deliver just the right amount of light at the right location.

Ron Gibbons of the Virginia Tech Transportation Institute explained that roadway lighting is directly linked to roadway safety, and ideal roadway lighting simultaneously provides for safe travel while minimizing skyglow and other light pollution impacts. Gibbons showed how roadway lighting with a correlated color temperature (CCT) of 4000K, compared to high-pressure sodium or LEDs with a warmer CCT, increased detection distance for drivers to see pedestrians on the road — a possible effect of increased color contrast enabled by the light spectrum. He also shared research thrusts related to lighting impacts on crime, human health, and ecological impacts to highlight the range of considerations for ideal roadway lighting. He concluded by stating that ideal roadway lighting must be a balance between intensity, spectrum, duration, and timing, and that adaptive lighting — dimming lights when less necessary — is a key component for achieving the best balance.

Rounding out the panel, Ian Ashdown of SunTracker Technologies focused on design software for ideal lighting. Current lighting design software provides physically realistic renderings and predicts lighting performance in terms of illuminance and luminance, glare, and daylighting. While this data is helpful, future software must better accommodate for controls, color changing and other dynamic lighting, and different lighting functions, including color vision, health impacts, germicidal light usage, and horticultural responses. Ashdown explained that all these features can be included today, but they need to be implemented and validated in commercial lighting design software, and there must be market demand for them first.


The next panel was an “Ask Me” session moderated by Kelly Gordon of Pacific Northwest National Laboratory (PNNL), focused on better understanding of light and health terminology that is used by researchers but not always clear to the broader lighting industry. Workshop attendees had the opportunity to submit questions in advance, and the expert panel lay the foundation for their answers with models, assumptions, and definitions referenced throughout the discussion.

Erin Flynn-Evans of NASA described the two-process model of sleep-wake regulation, consisting of Process S for sleep pressure, i.e., the longer you’re awake, the sleepier you get; and Process C for circadian wake drive, the 24-hour cycle of wakefulness and sleep. Light is the primary zeitgeber (time giver or time indicator) for the circadian wake drive: even if we were up all night, it is often difficult to fall asleep during the day, because the circadian wake drive is responding to time of day. Food, exercise, and other factors also play a role as zeitgebers, but are significantly less influential than light.

Céline Vetter of the University of Colorado Boulder explained we can conceptualize circadian disruption by considering how our rhythms are dampened or change with repeated phase shifts, such as shift work schedules. Oftentimes, we expect to see a certain relationship between rhythms. For example, melatonin is known to start to rise about three hours prior to sleep onset. If we force someone to go to bed much later than usual, this relationship has changed, and we call this circadian misalignment or disruption. This model is often used in the laboratory, where behavioral rhythms are timed to occur at different times of the circadian melatonin rhythm.  

Andrew Phillips from Monash University (Australia) reviewed the difference between visual and non-visual photoreceptors in the eye, noting that the intrinsically photosensitive retinal ganglion cells (ipRGCs) are very blue sensitive, with peak response at 480 nm, have prolonged activation, and sustain responsiveness for long periods of exposure. He explained that there are large differences among people in how light affects the circadian system, seen in a range of different non-visual responses (melatonin suppression, phase shifting, effects on sleep, etc.). For example, there can be a 50-fold difference melatonin suppression between individuals. As well as individual differences in the healthy population, there are systematic changes in non-visual light sensitivity with age, and differences in non-visual light sensitivity in certain clinical populations (e.g., depression, bipolar, delayed sleep).

Jamie Zeitzer of Stanford University listed the non-image-forming aspects of light that are relevant for human health and productivity — circadian impact, acute alertness, mood, perceptual comfort, and downstream effects on sleep, mental, and physical health — and noted we don’t actually care about the acute impact of melatonin suppression. He summarized what the current and proposed metrics quantitate: RPI Circadian Stimulus (CS) to quantify melatonin suppression; WELL Equivalent Melanopic Lux (EML) to quantify melanopsin action spectrum; and CIE Melanopic equivalent daylight illuminance (m-EDI) to quantify melanopsin action spectrum. WELL/CIE assume that outcomes of interest are driven solely by melanopsin; CS assumes that outcomes of interest are regulated identically to melatonin suppression. However, there is not just one ipRGC; different ipRGCs use different combinations of intrinsic (melanopsin) and extrinsic (rod/cone) input, so there is different spectral sensitivity. He added that more work is needed on the spectral sensitivity of phase shifting, alertness, etc., to complex (spectrum/temporality/magnitude) light stimuli.


Day 1 of the workshop concluded with virtual networking sessions, in which attendees could select from nine topics and join in informal discussion. Topics included overcoming barriers to deploying new technologies; solving problems in LED materials and device science; sharing new lighting technology; and sharing stories among manufacturers and lighting designers.



Day 2 of the workshop kicked off with a panel moderated by Brian Walker of DOE and Alex Baker of IES, who discussed the potential and challenges surrounding germicidal ultraviolet (GUV) lighting. 

Alex Baker described the IES’s role in developing GUV standards and best practices. He also reviewed problems the organization has seen with hyperbolic performance claims and safety implications with GUV products currently on the market.

Kyung Lee of Guidehouse analyzes technical potential and adoption of LED technology and the overall lighting market. He presented examples of DOE analysis efforts that could be used within the context of GUV and cited a recent Guidehouse study characterizing the energy impacts of agricultural lighting and the energy benefits of transitioning to more efficient LED light sources. Similar studies could develop understanding of the energy implications of widespread GUV usage and the impact of using sources with different power conversion efficiency levels. 

David Sliney, a medical physics consultant, discussed exposure limits and safety standards for GUV application and the particular germicidal application of ultraviolet-C (UV-C) to reduce transmission of SARS-CoV2, the virus that causes COVID-19. He explained that SARS-CoV2 is primarily transmitted through airborne transmission of aerosol droplets, so upper-room GUV application, which is safe for room occupants, could be the preferred application approach. Sliney went on to discuss occupational safety issues with UV-C, such as the risk of photokeratitis and erythema. He described recent and upcoming advancements in safety guidelines and the human damage spectrum with UV-C as well as the germicidal action spectrum for DNA.

Cameron Miller of the National Institute of Standards and Technology reviewed standards and implementation for GUV technology, including the need to measure the angular dependent intensity of UV sources to model the exposure on surfaces in a room and determine if there is sufficient flux. Miller described differences in source technology of low-pressure mercury vapor, excimer, and LED as well as discrepancies with UV sensors that could result in variations in flux readings. He also described experiments to determine pathogen sensitivity and the resulting deactivation of viruses.

Ruth Taylor of Pacific Northwest National Laboratory recounted the DOE CALiPER program — its history, why it was necessary, and how a similar program could be used for GUV lighting. Launched in the early days of LED technology, CALiPER tested LED products to compare their performance against product claims and against other lighting products to address the dearth of performance characterization standards. Taylor described the specific testing, objectives, and results that came out of CALiPER and how a similar program could benefit the GUV market. 

Brian Walker of DOE closed out the panel by sharing stakeholder feedback and requests for the Lighting R&D Program to foster collaborations and discussions on the GUV topic, and to pursue GUV validation and characterization activities similar to CALiPER and GATEWAY lighting efforts. Other stakeholder suggestions include DOE support for technology development of LED UV-C sources and metrology tools and techniques for characterization of GUV performance, and the development of safety and application standards driven by improved understanding of human and germ photobiological effects. 


In one of three concurrent sessions, Brian Liebel of IES moderated a discussion on how to better define and parameterize benefits from advanced lighting systems using healthcare, outdoor lighting, and horticultural lighting as examples.

Michael Myer from PNNL began the panel with a general discussion on quantifying lighting value as part of indoor environmental quality (IEQ). He emphasized that while the value of IEQ is being realized, more data and analysis is needed for lighting. He described a 10-year GSA study of 400 people that examined IEQ in a Chicago building. Because the office spaces already had reasonably good lighting, and there were issues with the HVAC system, the expected value of any lighting improvements was considerably less than thermal comfort. Myer is also chair of the IES Economics Committee and discussed the new matrix in development, which groups lighting by different areas of interest and applications. The matrix addresses the complex, non-visual values of lighting, along with the value of lighting as an asset. The economic analysis has also been expanded, focusing not only on energy savings and maintenance, but also grid demand response and load shaping. Myer asked the audience to contact him or IES directly with insight about lighting value and industries the committee should contact.

Shadab Rahman, from the Division of Sleep Medicine at Harvard Medical School, stated that in the last few decades, researchers have learned the physiological response to light spans from gene expression to protein expression, along with hormones, mood, and behavior. Metrics have primarily focused on mood, alertness, and psychological outcomes, but there are many others that should be considered. Rahman focuses on lighting in healthcare applications, with the key stakeholders being the patient, the care provider, and the institution itself. All three provide unique data sources: for the patient, health outcomes and satisfaction; for the care provider, work performance and worker satisfaction; for the institution, costs, customer satisfaction, and employee satisfaction. Rahman said it is critical to evaluate the impact of lighting on all three stakeholders, and that the optimal conditions for the patient are not always the optimal conditions for the care provider. He described a study at four Wisconsin care homes, where he found a 40% reduction in falls in the two care homes with a lighting intervention, compared to the two without. The lighting intervention used commercially available sources with higher CCT and light levels during the day and sources with lower CCT and light levels at night. (A report from PNNL is forthcoming which carefully documents the lighting at these four sites.) The other study was a hospital intensive care unit, changing the m-EDI from 61 to 208 in the nursing station. Rahman noted a 6% reduction in medical errors, a 10% reduction in potential errors, and a 13% reduction in all harmful errors. While these reductions were not considered statistically significant, he also noted a 33% reduction in high-severity harmful errors which was considered statistically significant. 

Leora Radetsky of the DesignLights Consortium (DLC) spoke on lighting for horticulture, emphasizing that there is not enough arable land to produce enough food globally, the demand for food is growing, and there are negative environmental impacts of food and agriculture. Benefits of growing food indoors include year-round production, weather-resistant facilities, and lower-water usage, although there is the potential for high energy loads. Radetsky stated that photosynthetic photon flux density should be measured instead of illuminance, and future metrics for food will be related to nutrition, flavor/aroma, pharmaceutical compounds, pest mitigation, and freshness. DLC’s Horticultural Lighting Qualified Products List helps stakeholders identify high-quality, energy-efficient products; as of January 13, 2021, the QPL has 241 horticultural lighting products from 67 manufacturers. 

Closing the panel, Rajaram Bhagavathula of the Virginia Tech Transportation Institute stated that standards and specifications for roadway lighting have lagged with the advent of LED technology. Metrics that can be used to quantify roadway lighting include crashes, light distribution, color or spectral power distribution, and light level. Bhagavathula shared studies showing the relationship between roadway light and safety but stated that crashes are not a good metric to evaluate lighting as they are rare and extreme scenarios with multiple causal factors. He noted the importance of the contrast between the object of interest and the surround, suggesting researchers study the probability distributions of detection in different lighting conditions and compare the detection distance to stopping sight distance. Bhagavathula also discussed roadway lighting’s effect on human health, and that the lighting in a car is much lower than photopic illuminance; further findings will be released later in 2021. In summary, he encouraged the use of metrics that are tailored to an application instead of a one-size-fits-all approach.


The second concurrent session, moderated by DOE advisor Monica Hansen, explored the latest innovations in LED emitter materials in the long wavelengths (green to red) and the short wavelength (UV-C) ranges. The use of tunnel junctions to eliminate pain points for LED device performance across the wavelength spectrum was also highlighted.    

Rob Armitage gave an overview of Lumileds’ DOE-funded project exploring improvements to green and amber LEDs. He reviewed the device mechanisms that limit efficiency and how the project is addressing the issue through a multi-disciplinary approach combining expertise in computational physics and recombination modeling, deep-level defect characterization, epitaxy and die fabrication, and tunnel junction design. The internal quantum efficiency droop in green LEDs must be improved to make the color mixing approach competitive for illumination applications. To do this, Armitage said, Lumileds will target improving active region design to involve more quantum wells and implement cascaded LED device designs via tunnel junctions.

Kirstin Alberi of the National Renewable Energy Laboratory discussed the pathways for emitter material improvements — the traditional pathways of LED heterostructure design and growth technology, doping, and defects, and additional pathways for future advances. Alberi mentioned leveraging existing materials know-how from adjacent industries, such as the metamorphic growth process used in photovoltaic devices, stating there is potential in using computational materials discovery to identify new compounds and refine known existing compounds for an application. She acknowledged the challenges of knowing when scientists may not be able to overcome fundamental material property limitations of existing emitter materials and when it is time to focus on new materials discovery. She closed by asking: How can we leverage R&D to build on breakthroughs in understanding and in materials control to collaboratively advance aspects that are beneficial to multiple technologies?

Michael Kneissl of Technische Universität Berlin spoke about the challenges and the current status of UV-C LED emitters, citing a steep efficiency drop-off as the UV-C wavelength decreases beyond 260 nm. The drop in efficiency occurs in many areas of the device design from defects (low internal quantum efficiency) to high resistivity device layers and contacts (low power conversion efficiency) to poor light extraction (refractive index limitations in existing materials). Although efficiency is low, there have been significant improvements over the past five years, and now more established LED manufacturers are entering the market, providing the critical mass of R&D investment required to make further breakthroughs. Kneissl said that improvements in defect density by improving substrate technology is one key thrust in improving UV-C LED performance, and the second is improving the light extraction from the device by developing UV-reflective contacts, a UV-transparent p-side epitaxial stack, and UV-transparent polymers for encapsulation.

Siddharth Rajan from The Ohio State University wrapped up the panel by describing the benefits of tunnel junction layers to device performance in LEDs and lasers. He explained how a cascaded LED device design (one with multiple active regions stacked upon each other) can circumvent current density droop in nitride-based LEDs, because the voltage is being scaled instead of current. He also described the state-of-the-art tunnel junctions grown by production metal organic chemical vapor deposition (MOCVD) processes and how the resistance has been minimized to allow improved device performance from visible LEDs and laser to improving UV LEDs.


In the third concurrent session, DOE advisor Lisa Pattison moderated a panel to address the key R&D challenge of OLED technology: the state of the art, needs, and recent advancements in blue materials.

To begin the session, Marina Kondakova of OLEDWorks discussed the state of the art of blue OLED materials and performance requirements of blue materials systems for industry. Kondakova summarized the strengths and weaknesses of trending blue emitter approaches including fluorescence, phosphorescence, thermally activated delayed fluorescence (TADF), and hyperfluorescence. She suggested that hyperfluorescent blue emissive layer (EML) seems the most promising approach for stable, efficient blue, but the components (host, emitter and/or TADF dopants) need to be developed in parallel to achieve the best performance. Further, improvements in charge injection and transport layers could allow simultaneously a reduction in absorption, voltage, and plasmonic losses. Kondakova concluded by emphasizing that any new materials need to be cost effective and tested in a state-of-the-art OLED lighting stack so that performance improvements can be verified. 

Mark Thompson from the University of Southern California presented on approaches to blue stability through molecular design. By reducing the emission lifetime, τ, device lifetime can be extended by the consequent reduction in triplet-triplet and triplet-polaron annihilation, which lead to molecular degradation. After discussing the limitations of TADF and phosphorescence, Thompson introduced his work on Cu-, Ag-, and Au-based complexes, which could be competitive alternatives to iridium phosphors. Preliminary results from carbene-Cu-carbazolyl planar emitters showed short decay lifetimes of ~1µs and tunable emission across the visible spectrum with respectable external quantum efficiency (EQE) of ~12% for blue and up to ~20% for green. Thompson went on to discuss work comparing Cu-based complexes to Ag-, and Au-centered complexes with which emission lifetime improvements were made. In Ag complexes, τ was reduced to 0.33µs and 0.5µs for green and blue, respectively. Ag shows the fastest TADF rate primarily due to the very small energy difference between the singlet and triplet excited state. Thompson proposed that to improve lifetime, host and transport materials need development.

Jun Yeob Lee from Sungkyunkwan University (South Korea) discussed advancements and approaches to host materials developments for blue emitter systems. High triplet energy hosts with bipolar charge transport properties that are stable need to be developed for blue OLED devices. Single- or mixed-host materials can be used, but mixed hosts are preferred as they allow longer device lifetime and easier charge balance. Lee outlined strengths and weaknesses of exciplex-free and exciplex-based mixed hosts for blue. From there, he focused on exciplex-based hosts, specifically TADF-type exciplex hosts, which can convert triplet excitons into singlet excitons and minimize the excited state lifetime of triplet excitons for stable devices. High triplet energy hosts were achieved with triplet energy of around 2.97 eV and excited state lifetime τ ~ 4.6 microsecond. Devices with these hosts show blue EQE as high as 20% and a lifetime improvement of 60%.

Russ Holmes of the University of Minnesota transitioned to a discussion of OLED efficiency and the factors that play into efficiency. For efficient emitters, peak efficiency is limited by light extraction losses and non-unity charge balance. At higher current density, efficiency roll-off reflects bimolecular exciton quenching. Exciton quenching is usually assumed to be negligible before turn-on, but Holmes’ group revisited this assumption. Using a typical green phosphorescent device, they measured photoluminescence and EQE vs. current density and found that photoluminescence (PL) is maximized at reverse-bias, well before turn-on. They attribute this exciton-polaron quenching with accumulated charges from spontaneous orientation polarization (SOP) at the interface. Molecules have a preferred dipole orientation during film deposition which induces sheet charge at layer interfaces, resulting in reduced hole injection voltage and increased hole accumulation. The increased hole density quenches PL by 20–35% before turn-on. The impact of this is that charge balance is near unity at peak EQE, and accounting for quenching is critical when quantifying max efficiency. Finally, Holmes said, experiments show that heating during deposition reduces orientation and charge accumulation, resulting in decreased quenching and higher EQE.

Chris Giebink of Pennsylvania State University finished the panel by pointing out that some of the properties considered “intrinsic” to OLED materials could be modified by diluting the organic semiconductor with a carefully selected insulator for improved OLED device performance. Typically, OLED materials have a refractive index n ~ 1.7, which limits optical outcoupling efficiency. Moreover, glass transition temperatures are typically Tg ~ 100–120°C, though OLED panels operate at elevated temperatures such that thermal stability is closely connected to catastrophic failure. Giebink then posed the question: What if we could change properties like these without reengineering OLED molecules? He explained that by diluting the organic semiconductor with a carefully selected insulator, one can change optical and thermal properties without sacrificing electrical transport. Giebink presented results of such an experiment using Teflon AF (n ~ 1.3) to dilute the hole-transporting N,N′-Di(1-naphthyl)-N,N′-diphenyl-(1,1′-biphenyl)-4,4′-diamine (NPD) organic semiconductor. For NPD hole-only devices, dilution lowered drive voltage and resulted in higher thermal stability. It also reduces built-in voltage, improving carrier injection at the interface. Finally, bulk mobility decreases due to lower energetic disorder and percolation. Giebink introduced the idea of tailoring the mobility and refractive index in every transport layer of a white OLED stack, with dilution molecules becoming a standard ingredient for OLEDs. Doing so may present an economic advantage by reducing the quantity of expensive stack materials required. Further, modeling shows internal extraction efficiency could be improved by 70% or more. However, Giebink said, a material other than Teflon is needed and should be selected to optimize the electrical and optical performance as well as stability of devices. 


The annual poster session concluded the second day of the workshop, where 50 researchers spanning the breadth of DOE’s project portfolio presented their work virtually, providing attendees with additional opportunities for one-on-one discussion, information exchange, and potential partnering. 

The poster session was preceded by a session called “Lighting Innovators,” hosted by Joel Chaddock of National Energy Technology Laboratory. Chaddock previewed the poster session by showcasing the following DOE-funded projects with potentially significant results:

  • High Performance Green LEDs for Solid State Lighting: University of California, Santa Barbara, presented by Cheyenne Lynsky
  • Flexible Low Cost Platform for Beam Pattern Optimization OR Antireflective Materials for Flexible SSL: Glint Photonics, presented by Peter Kozodoy
  • Improved Light Extraction for 130 lm/W OLED Lighting Panel: Pixelligent, presented by Selina Monickam
  • Improved Light Extraction by Engineering Molecular Properties of Square Planar Phosphorescent Emissive Materials: Arizona State University, presented by Jian Li

Chaddock also recognized the grand prize winners of the 2021 Student Poster Competition: 

  • Ruqayah Bhuiyan, University of Georgia: Lettuce Tolerates Fluctuating Light, Potentially Reducing Energy Costs in Controlled Environment Agriculture            
  • Yunping Huang, University of Washington: Green Syntheses of Stable and Efficient Organic Dyes for Organic Hybrid Light-Emitting Diodes
  • J. Mundinger, Pennsylvania State University: The Visual Experience of Fine Art Under Low Illuminance



Day 3 of the workshop began with a panel moderated by Monica Hansen, which considered how technology advances in LEDs, OLEDs, phosphors, and quantum dot materials for displays can be leveraged for lighting applications. The panel reviewed the strong technology overlaps that allow for synergistic R&D opportunities between the two applications. 

Mike Hack of Universal Display Corporation reviewed the many applications for OLEDs, ranging from displays of all sizes to automotive lighting to general illumination. Presenting a detailed comparison of performance features in displays versus lighting, he stated that luminance levels and reliability and lifetime requirements are more difficult for illumination applications compared to display requirements.

Rajiv Pathak from Lumileds described how lighting is moving to higher levels of functionality and technology integration. Pathak said the design of light sources for dynamic lighting applications with individually addressable LED arrays and projection optics can provide “the right light where and when it is needed.” He added that features such as high precision dynamic beam shaping, low glare, thin form factor, and fusion of basic projection display functionality (e.g., wayfinding, information display) require higher spatial and angular resolution, which can be achieved by LED miniaturization (mini or micro-LEDs).  

John Whiteman of Plessey Semiconductors reviewed the use of micro-LEDs in displays and the LED performance challenges as they are miniaturized. Displays require discrete color emitters with high color purity, narrow full width at half maximum, or FWHM (saturated colors), and uniformity across a display and wafer. Whiteman detailed the current performance levels for red, green, and blue micro-LEDs of differing sizes, showing how the peak EQE dropped and moved to higher current densities with decreasing micro-LED size due to the increased perimeter-to-area ratio leading to higher nonradiative surface recombination. With optimized epitaxial growth and device architecture, the EQE was increased  from 8% to 37% for an array of blue 4 μm pixels. The green gap was closed for small pixels. Unfortunately, with the high surface recombination velocity for aluminium gallium indium phosphide (AlGaInP) materials (leading to worse losses than gallium nitride [GaN] LEDs), the red micro-LED performance dropped precipitously with device size. The ‘red gap’ is the new ‘green gap.’ Whiteman finished by pointing to three competing approaches to addressing the red gap: improve AlGaInP surface recombination losses (passivation), improve the efficiency of red InGaN LEDs, or implementing color conversion with blue LEDs and quantum dots.

Jim Murphy of GE Research discussed evolution of phosphor materials in displays; narrow-band emitting phosphors have enabled more efficient, brighter, and wider color gamut displays. He opined that “the future of displays is size diversity,” which will require a wide range of sizes for both LEDs and phosphors. All-phosphor solutions, Murphy said, are the most economical in displays, but other hybrid solutions of phosphors and quantum dots down-converters can provide the best color gamut and efficiency. He finished by highlighting the great potential for energy savings in displays by moving away from inefficient LCD displays to self-emissive displays that are made possible with blue micro-LEDs and down-converter materials for the green and red channels.

Ray Ma of Nanosys concluded the panel by highlighting the prevalence of quantum dots (QDs) in display products and the different architectures used with them. QDs can be implemented in hybrid structures with blue OLEDs or combined with blue micro-LEDs. Blue OLEDs + QDs have better color and viewing angle and higher luminous efficiency, and single-color (blue) OLED layers provide design flexibility with no differential aging. Micro-LED architectures can provide low power consumption, high brightness, high contrast ratio, and low response time. The big challenge for QD materials, Ma said, is improving their stability to high temperature and blue photon fluxes for on-chip LED application. He finished by discussing the potential for QD electroluminescent displays as another category of self-emissive displays. 


The next panel, moderated by DOE advisor Norman Bardsley, discussed recent progress in our understanding of the human impact of glare and shared examples of the technologies that may enable the production of lightweight, conformable, glare-free diffuse light sources.

Nathaniel Jones of Arup explained that disability glare hinders the eye’s ability to form clear images, either because of excessive brightness or low contrast. Discomfort glare occurs when objects in the field of view are much brighter than the luminance to which the eye is adapted. Both effects depend on the instantaneous field of view and level of adaptation of the eye, as well as on the age and health of the observer. Jones said quantification of these effects is complicated by the existence of several glare metrics, two of which are favored for studying daylight and one (unified glare rating, or UGR) is used most often for electric light. Cameras are now able to measure the parameters that go into these metrics, but the interpretation of the results is difficult; tolerance of glare often depends on the activity engaged by the observer. Jones added that software is available to predict glare levels, and the results are often cited in luminaire specifications, but glare predictions need to take into account all light sources, the room or outdoor environment, and the viewer.

Dan Schwade of Acuity Brands illustrated the use of the UGR metric to study glare from OLED pendants, wall sconces, and ceiling-mounted fixtures. He stressed that studies of all the luminaires in a room are more valuable than data on individual lights. However, he added, UGR does not take into account all the factors that affect glare.

Eric Haugaard of Cree Lighting discussed the use of waveguides in spreading the light emerging from arrays of LEDs. He stated the substantial reduction in the size of LED packages increases the potential for glare if viewed directly but makes it easier to inject light into waveguides and broadens the options for secondary optics to control the distribution of the light. The effectiveness of total internal reflection means that light can be spread efficiently across a waveguide. Haugaard said the major challenge lies in the extraction of the light. Although surface microstructures can be designed to give excellent control of the uniformity of the light across the panel and the angular distribution of the emitted light, manufacturing the required patterns can be difficult and expensive. Haugaard showed examples of commercial products that distribute the light in attractive and effective ways.

Finally, Sergey Vasylyev of Lucent Optics described research, supported by the DOE Lighting R&D program, to adapt the ultra-thin waveguide technology used for many years in displays to general lighting applications. These efforts have led to conformable panels that can emit light from one or both surfaces with uniformity over 90% and optical efficiency above 80%. The panels can be transparent when off, or they can be patterned to produce decorative luminous surfaces. Vasylyev said the waveguides can be constructed from relatively inexpensive plastic materials and patterned by microprinting. He added the thin lightweight panels provide an excellent way for general lighting to benefit from the development of mini-LEDs.


Andrea Wilkerson from Pacific Northwest National Laboratory convened academic researchers and entrepreneurs, asking them to discuss the promise and limitations of new tools for lighting research, design, and manufacturing and what it will take to implement these tools in practice.

Steve Fotios of the University of Sheffield (England) began with the study question, “What are the key tasks for pedestrians?” He said the problem with this question is that participants give their stated preference, answers they think the researcher wants to hear, or socially accepted responses. Fotios suggested the use of eye-tracking as a more objective way to answer this question, because eye-movements are strategic, and instead asking, “How safe do people feel when walking?” He noted that in 1965 lighting researcher Peter Boyce used a haptic contact lens to quantify eye movements. Today, eye-tracking can be completed with a pair of spectacles with integrated cameras. Fotios carefully explained why critical fixations give a different result than looking at all fixations. He also shared his research using driver’s gaze to estimate adaptation luminance. 

Siobhan Rockcastle of the University of Oregon detailed her experience using virtual reality (VR) headsets to render environments, noting that even as she prepared her presentation, new product releases allowed for different workflows and techniques. Rockcastle initially was interested in VR to show controlled images and scenes to research participants, but now she leverages it for educational purposes. She acknowledged there is always a difference in how people respond to virtual environments than in real environments vs. immersive tools. She stepped through different workflows using a mix of software, showing the differences in the results produced by each of them. She also stated some of the challenges of using VR, many contributing to the inaccuracy of the light rendered in the environments.

Gary Trott of Acuity Brands discussed his experience with 3D printing and how it addressed his frustrations with product development, which included slow development timelines due to tooling, innovation limited to manufacturing practices, thus leading all luminaires to look similar, and the inability to customize. With 3D printing Trott was able to iterate almost daily, leading to new forms that can only be built this way. He also noted the opportunity for mechanical design simplification, replacing multiple parts with one part that was simpler and stronger. Trott predicted much more use of 3D printing by lighting companies over the next few years. He also spoke of his shift in focus to GUV, stating there were three big problems to solve: scaling from niche to mainstream, balancing safety with pathogen reduction, and disinfection while people are present. When applied properly, well-designed products can reduce pathogens safely, but a conservative and science-based approach to product design and application is required, and in-depth due diligence of potential suppliers and products is necessary. 

Rounding out the panel was Steven Rosen of Available Light, who shared his experience with controls throughout his career, and how he has heavily relied on his experience in theatrical lighting and translated it to architectural lighting. Rosen said there is an expanding experience design market—retailers are now creating and curating spaces for consumers to visit and experience the product first before buying online. He added that architectural lighting designers have always been interested in the aesthetic and technical aspects of light, but control systems design, in large part, has not received the same attention and priority. From Rosen’s perspective, creating a system manageable by end users, even if it means simply allowing remote monitoring and operation, is critical to successful lighting control systems, along with owner buy-in and multiple layers of owner and operator training. He shared several case studies including the Caesars Forum banquet hall in Las Vegas, where every one of the almost 5,000 luminaires in the room is individually programmable and controllable by an iPad.


Day 3 of the workshop concluded with a 90-minute topic table session, in which attendees participated in small-group discussion to provide input to DOE and IES that guides planning for future research activities. Each group examined the importance of a specific subject and its technical challenges, using a few “questions to consider” to help spur the discussion. A wide variety of LED and OLED topics was available ranging from technologies (materials, devices, quantum dots, controls) to data and metrics to of-the-moment applications (horticultural, physiological, animal response, sustainability).



The last day of the workshop began with a panel moderated by Michael Poplawski of Pacific Northwest National Laboratory, which explored the growing use of software in the building industry and how to incorporate data from current building systems to optimize performance and inform the design process for new building systems.

Leland Curtis from the design firm SmithGroup examined how data translates into knowledge and wisdom, with coding being the way data is refined into useful products. Curtis highlighted new visual scripting tools that enable a broader audience to code without in-depth knowledge of a coding language. Repetitious tasks are automated or augmented. He provided an example of using automated analysis for the design of a university science building façade, with the parametric analysis allowing for a more informed design than traditional methods. This automation generates data that feeds the analytics and algorithms for new design workflows, and Curtis challenged the audience to think about how data will change their workflows.

Lighting designer Star Davis returned for a second workshop talk, stating that innovation is a combination of design and engineering using iterative-type solutions based on knowledge from past projects. She noted that optimization has depreciating returns, and while LEDs have enabled customized solutions, they are inefficient and costly, recalling the creativity achieved with the fluorescent lamp form factor. She encouraged process automation, so the focus is on being creative where it matters, providing several supporting examples from her time at WeWork. Davis said one large problem is that the intelligent data set from design gets simplified to paper for construction, then the data set is rebuilt by the contractor, and then it’s again simplified to paper for building operation. She highlighted the need for a process that supports data continuity to provide feedback to designers and contractors, because without a feedback loop, it is best guesses and best intentions.

Finally, Cindy Zhu of Prescriptive Data, a smart building technology company, discussed optimizing the operation of commercial buildings through data-driven insights. Zhu described Prescriptive Data’s integrated, cloud-based Nantum operating system, which combines and analyzes data from different silos: building management system, occupancy, weather, water, air quality, emergency, and utilities. Zhu said that energy savings is key, and Nantum users see on average a 17% reduction in energy use, along with a 43% reduction in thermal comfort. She noted, however, that lighting integration is not something commonly requested, even by customers with smart lighting systems, and that a challenge her company faces is that “a lot of the lighting industry is not as open, or willing to play nice,” a sentiment shared by the other panelists.


In the first of two concurrent sessions, Morgan Pattison moderated a panel discussion on the evolution of the LED supply chain. While recent economic events including tariffs and the global pandemic have affected lighting manufacturing and supply, LED technology and manufacturing techniques have matured and evolved, opening up opportunities for new manufacturing concepts and advancements in lighting product design for performance and improved sustainability. 

Valerie Nubbe of Guidehouse Consulting provided an overview of the global supply chain based on analyses of value-added manufacturing portions and global trade flows of LED products and components. Nubbe showed trends in lighting product manufacturing and how tariffs have affected these production trends. Trade flows showed differences in LED package, lamp, and luminaire geographical production. LED packages are mostly produced in Asia, and LED lamps are almost entirely produced in China and were exempt from recent tariffs. However, Nubbe said, LED luminaire production shows more global balance with considerable imports to the United States from Mexico.  Opportunities for U.S. manufacturing include high-end luminaire and emerging lighting technology markets.

Monica Hansen of LED Lighting Advisors described a decade of change in LED lighting manufacturing.  LED packages and lighting products have dramatically matured. Lighting products have shifted from using high-brightness packages to mid-power packages, as mid-power efficiency has dramatically increased and prices have come down. Hansen noted that a decade ago, there was still some question as to the most effective substrate for GaN-based LEDs; the answer is now clear that sapphire substrates are, by far, the dominant choice. LED package types have proliferated over the last decade as well, providing sources and features for new lighting functionality, including control of color and optical distribution. Hansen said the trends in LED production and lighting function open up opportunities for new manufacturing technologies, with one clear opportunity being the application of additive/3D manufacturing at various portions of the manufacturing process to increase flexibility. 

John Trublowski of Eaton further discussed additive manufacturing for LED luminaires, including R&D opportunities and technology gaps. Additive manufacturing enables on-demand manufacturing, reduced part counts and inventory, and integrated mechanical and electronic functionality of luminaire components. While additive manufacturing is currently expensive, it does reduce the design cycle time, reduce tooling requirements, enable component integration, and localize manufacturing. These factors, Trublowski said, can offset the higher production cost of additive techniques. Additive manufacturing can also enable new designs and design features that cannot be manufactured with existing production techniques. Over the course of a DOE-funded R&D project, he said, Eaton has explored additive manufacturing techniques for luminaire heat sinks and mechanical structures and for reflective and refractive optics. 

Rounding out the panel was Aaron Smith of Finelite, who reviewed the opportunity offered by LED technology to make the production and use of lighting products much more sustainable. Smith said that LED lighting technology can enable a full circular economy for the use of lighting products: lighting products can be made from low-impact materials, used with high efficiency, and then broken down at end of life for reuse. This possibility, however, requires a commitment by manufacturers and consumers to pursue these benefits rather than continuing with the status quo. Smith added that LED technology is particularly well suited to foster a more sustainable mindset, because LEDs are more efficient and therefore require fewer thermal and different materials, but we must break away from the design and manufacturing approaches associated with previous lighting technologies. 


In the second concurrent session, Norman Bardsley and three OLED experts discussed the research needed to design structures that can be manufactured affordably, protect the fragile organic materials, and extract the light. The panel also reviewed recent manufacturing progress, opportunities for cost reduction, and analysis of market demands for diffuse lighting.

Jeff Spindler of OLEDWorks provided an overview of the methods currently used in OLED panel manufacture along with details on process windows, quality control, and fabrication costs. He stressed the need for a stronger supply chain, particularly with respect to integrated substrates, encapsulation, and back-end processing after the panels have been separated.

Christian May of Fraunhofer FEP in Dresden, Germany, described research and prototype production of flexible OLED lighting panels, most of which was carried out in collaboration with other European research institutions and companies with funding from the European Union (EU) as well as national and local governments. May said Germany has also benefited from a bi-national program with Japan. He went on to review current patterning techniques and roll-to-roll manufacturing. The difficulty of providing moisture barriers for plastic substrates has led Fraunhofer to favor processing on ultra-thin glass supplied by Corning as well as Japanese and German vendors. Following the success of the PI-SCALE and LAOLA projects, Fraunhofer is now working with small- and medium-sized enterprises to develop prototype OLED lighting products with funding through the EU’s SmartEEs and ACTPHAST projects.

Closing out the panel, Stephen Forrest described the University of Michigan’s DOE-funded project to develop roll-to-roll fabrication techniques. Forrest explained his team’s focus is on the simultaneous use of two deposition techniques for the organic layers, necessitating passage of the web between chambers at different pressure levels. Organic vapor phase deposition is used for the emitter layers and the traditional vacuum thermal evaporation for the transport and charge generation layers. The team also will explore the use of atomic layer deposition in the development of simpler encapsulation methods.


The final workshop session returned to the theme of “meeting the moment”—how will the current shift in lighting technology redefine lighting and related energy savings and evolve to fulfill future functions and services? Moderated by Morgan Pattison, the panel considered how future lighting will be more efficient, flexible, and controllable to improve well-being and performance and reduce negative environmental impacts.

Naomi Miller of Pacific Northwest National Laboratory emphasized the need for lighting systems to be more dynamic in order to respond to human physiological needs and achieve more extreme light levels in daytime and at night. Humans should wake up to morning-like lighting conditions and prepare for sleep “with only the stars and moon in the sky.” Miller discussed the possibility of luminaires that change both color and directionality to mimic changes in natural light over the course of a day. She added that controls are a critical part of the future of lighting and should be designed for simplicity and usability. She also stressed that working in source efficacy, intensity controls, and optical distribution can continue to reduce the energy consumption for lighting.

Tom Phoenix of CPL Architects and Engineers, representing ASHRAE, discussed the road to zero-energy buildings and the role of lighting in this objective. Phoenix said new lighting technologies are not only more efficient but reduce HVAC loads and are readily controllable with occupancy and daylight sensors to save even more energy. LED lighting also works well with DC power grids, which can facilitate renewable energy integration and the use of power storage. Because LED technology reduces energy consumption and is a technology ready for use, Phoenix said lighting is “low-hanging fruit” for achieving zero-energy buildings.

Next, Brad Koerner of Cima explained how sustainable lighting is a competitive advantage. Reducing the material in a luminaire, shrinking the luminaire, and using eco-friendly materials can all provide a performance and cost advantage. However, he said, lighting manufacturers must move beyond striving for lowest-possible pricing and avoid “greenwashing” products—claiming sustainable manufacturing when there is little actual sustainability benefit. Koerner proposed that sustainability in manufacturing and design of lighting products should be considered an R&D challenge for developing lasting, local manufacturing advantages. He described the wastefulness of legacy AC electrical connections that result in efficiency losses and additional, unnecessary hardware. Simplification of power delivery will also result in simplification of lighting controls and resulting simplification of the luminaire itself. Finally, Koerner described his award-winning sustainable lighting product concept that uses eco-friendly sustainable materials configured for ease of assembly and disassembly at end of life. The design leverages advances in LED technology to enable the use of new luminaire materials.  

The final talk of the session was from DOE’s Brian Walker, who described the Lighting R&D program’s approach for defining and accelerating lighting of the future. He explained that DOE gathers inputs from numerous stakeholders to develop aggressive expectations for advancements in lighting technology that result in significant energy savings and improved lighting performance for visual function and human well-being. This approach, along with DOE R&D funding, has resulted in significant energy savings and numerous patents and products. Walker presented a vision of the future of lighting that encompasses improvements to lighting source technology, better understanding of lighting science including physiological responses to light, and integration and validation of technology and processes brought together to achieve maximum benefit and meet DOE Lighting R&D goals of energy-efficient lighting that supports health, productivity, and well-being. Walker emphasized that the discussions and work happening today will define the future of lighting.