The race continues to produce the most efficient white, mid power LEDs for illumination. Tremendous progress in the last several years in die, phosphor, and package technology have seen the efficiency claims from numerous LED manufacturers climb to what seem to be impossibly high values. Is there a limit to how high they can go? Yes, physics imposes practical limits, which vary, depending on the color quality and color temperature of interest. LED manufacturers are always looking to stretch these limits. Figure 1 plots the efficiency of various types of lighting over time. You can see the steep efficiency gains achieved in recent years with LED lighting.
As LEDs get more and more efficient, their ability to replace traditional lighting technologies improves. Also, the number of LEDs needed to produce the same amount of light decreases, thereby lowering the costs of the bulbs and fixtures. An added benefit is the fact that a more efficient LED produces less waste heat, so the heatsinks in bulbs and fixtures can also be reduced. All of this makes the job of the lighting designer easier, and LED lighting even more attractive to the mass market.
White LED efficiency is typically expressed in lumens per watt (lm/W). Most white, mid power LEDs use a die that produces blue light, which is then converted to white light by a phosphor system. In order to improve the efficiency of an LED, manufacturers focus on improving the die and the phosphor. Blue dies have been getting better and better at converting electrical energy into blue light. While it is almost impossible to be 100% efficient at doing this, blue dies have moved up the efficiency curve year after year.
One way to increase the efficiency of LEDs is to put larger and larger dies in mid power packages and drive them at relatively low currents. This helps overall efficiency because the blue dies used in mid power parts tend to reach their highest efficiencies at lower drive currents. So one trend we have seen in recent years is increasing the die area in mid power LEDs.
While this may seem like an expensive way to improve efficiency, the fact is that die costs have actually been coming down quite rapidly due to an over-capacity in the global LED market. So there really isn't a significant cost disadvantage to using larger dies. However, clearly a more elegant and cost-effective way to improve blue dies is to make them inherently more efficient, which will allow manufacturers to use smaller dies to save cost.
Now that dies are getting more efficient at producing blue light, LED manufacturers then need to make sure as much of that blue light as possible is converted to white light. This is done by using phosphor materials -- typically green and red. Conventional or broadband phosphors (~80-100nm FWHM) are pretty good at doing this. We are seeing efficiency values of LEDs approaching 200 lm/W using these phosphors.
Improvements will continue in blue dies to the point where physics kicks in and dies just can't get any more efficient, except perhaps incrementally. The Holy Grail of warm, white LED efficiency has been narrow-band phosphors (<50nm FWHM) -- particularly red. The advantage of narrow-band phosphors is that they don't waste energy by producing light that isn't visible to the human eye, as conventional phosphors do at longer wavelengths. Extensive work is being done in this area, and soon we hope to see LEDs with phosphors that start to push through the 200 lm/W barrier for warm, white, CRI 80 parts.