Feb 20, 2017: Like any new technology, LED lighting must demonstrate a compelling value to buyers before it begins to win sizeable market share from the incumbent technologies it will replace. Over the past decade, research and development have yielded impressive improvements in the cost, color performance, light output, efficacy, reliability, lifetime, and manufacturability of LED products.
Looking at LED packages specifically, the cost in dollars per kilolumen ($/klm) has been declining rapidly since 2005 at a rate of around 20% per year. As a result, LED lighting products have become competitive in almost every lighting application, but there is still room for improvement.
Despite the rapid pace of its development, LED lighting has not yet come close to achieving its full potential. Significant work remains to be done to further improve performance and reduce costs.
LEDs are semiconductor devices that produce light when an electrical current flows through them. Optimizing efficiency in LED lighting will hinge on ongoing improvements to light-generating materials (LEDs and phosphors) and system integration. LED package prices have declined to the point where they are typically no longer the primary cost component. Now LED package and luminaire system integration improvements can have a larger impact on the cost of LED lighting. However, solving technical challenges at the LED can still have a big impact on cost and efficiency, and enable new lighting value and applications.
One package-integration direction has been to increase light output from a given package size to decrease cost, while another has been to drive LED material at lower current densities and use lower-cost packaging materials.
DOE’s cost and efficacy goals
There are still many fundamental technical challenges that need to be met in order to achieve DOE’s cost and efficacy goals. For example, better green and red LEDs can reduce phosphor conversion losses and enable color control. And the process of down-conversion–whether with phosphors or quantum dots–offers room for improvement. The alternative approach for creating white light using color-mixed LEDs offers another path to high efficiency and the possibility of color tunability.
In addition, work remains to be done in such areas as current efficiency droop, color shift, and system reliability, as well as in light distribution, dimming, thermal management, and driver and power supply performance. Luminaire cost, performance, and design flexibility can also be improved by reducing efficiency losses at higher temperature and drive current operation conditions. And continuing to expand the range of LED package diversity is critical for addressing an expanding variety of lighting applications.
Challenges in LED R&D: Current droop
There’s a problem that is often talked about with efficacy of LEDs, it’s called droop. The higher the current density of the LED, the worse the performance. Droop refers to a reduction in efficiency of LEDs that occurs at very high currents, particularly in three nitride LEDs. There’s been a lot of debate over the years exactly as to what the source of that is, but at this point, there’s generally consensus that this is due to auger recombination. When we flow charge carriers into an LED, they have the option of either recombining and emitting light, which we call radiative, or recombining in a form that forms into heat, which would be non-radiative, and auger is one of the latter mechanisms.
So there’s really two reasons that we want to eliminate droop, and part of this is application driven. So there’s this desire to go to higher and higher flux, from a given chip. So we want very small, bright sources for many applications. And when we try to do that now, we run into droop, the efficiency begins to drop off. The other side of this is cost. If we have these very bright sources, we can have fewer of them, so at a system level we can reduce the cost overall. The thing about improving droop is that the gains that we would achieve are increasing at higher and higher current. So what we end up with is a situation where, we can run a device that used to be running at say, one amp, at some efficiency, now we can move that to 1.5 amps or higher, at the same efficiency. So the actual light output gain is tremendous.
Challenges in LED R&D: System reliability
System reliability is a challenge for LED products, in part because they’re a new technology, whereas conventional lighting technologies have been around for decades. So there is a body of experience that people have.
The incandescent light bulb that before was two or three years of life expectancy, now, the customer is looking for five, 10, 15 years, even 20. So that’s why you hear of 25,000 hours, 75,000 hours. With solid state lighting that’s a challenge, because you have more complexity, more opportunities for failure.
We’ve got many pieces that go together to build this luminaire. There, of course, is the LED component, and that’s the chip. And that’s covered with a phosphor. And there’s generally some type of a silicone material on that chip to pull the light out of the chip. Then the heat sink for the product is there to pull the heat away from the LED when it’s operating.
The LED has to be attached to a board and mounted on to the heat sink. Then there’s the electronics that are required to drive this light out of the LED. So there’s this driver component, which includes resistors and capacitors. They have to be mated to the LEDs. And then, oftentimes, there are secondary optics that are placed over these LEDs.
Any of those factors could impact the reliability of the system. At the LED luminaire level, we often look at the LEDs, which perform very well. But we may look at factors like, how they maintain their light output or what’s called lumen maintenance, or do they stay roughly the same color over the entire lifetime of the product?
For the drivers, we look at the electrical components. And so there you’re dealing with issues like you would for computers and other devices like that.
So if we look at today’s manufacturing, it can be very complex to get all of these individual components put together into a neat, efficient package.
Challenges in LED R&D: Color shift
LED products are specified to keep a constant color point over their lifetime. And color point is defined by a couple of different factors. You have the wavelength of blue LED. You have the absorption and emission characteristics of the phosphor. The design of the package and then any optical elements. All of these together need to be designed to remain stable over the full lifetime of the product under all its operating conditions. And when one of them starts to degrade, the color point can start to drift away from its original point and you get color shifts.
The human eye is very good at distinguishing variations in color, especially for light sources that are close together. Now, the extent of color shift that is acceptable depends on the amount of shift, of course, but then also on the application. So if you look at streetlights for example, a small amount of color shift might be tolerable. However, if you go to museum lighting or retail environments, then basically as soon as color shift is noticeable it’s usually unacceptable. And that means that the product has effectively reached the end of its life.
Now for the manufacturer, the challenge is to design a product so that they’re robust over the entire lifetime. That’s not easy to do because you want to project a long lifetime but at the same time you have only a very short time over which you can test a product. What makes it more complicated is that in the long term color shift is something that you cannot easily extrapolate from a short term test. And that’s because there are different mechanisms that can play and they occur at different color shift rates, and color shift may be in different directions.
There is a couple of things that the LED manufacturer can do. On the materials development side, develop more stable phosphor materials especially for the red phosphors. And then also protection of package materials against corrosive environments. And in general, thermal management inside the LED package is crucial because most of the degradation mechanisms are thermally activated. Also for the luminaire maker, that manufacturer needs to be aware of the operating boundaries of the LED in terms of drive current and temperature. So that it stays within its operating limits by using proper heat sinks, et cetera. So that actually you get the performance that the LED was designed for. You get robust performance over a long lifetime.
Industry input guides SSL R&D agenda
The DOE SSL R&D program is guided by the DOE SSL R&D Plan. All funding opportunity announcements (FOAs) and project selections align with this document, which is updated annually in collaboration with industry partners.
Source: US Department of energy