LEDs: ZnO prepares to leap from lab to fab
As a material for making LEDs, ZnO has loads to recommend it. It is cheap, non toxic, and chemically stable; it is incredibly well understood; it can emit light very efficiently; and it is capable of making devices that span the ultraviolet to the infrared. However, despite all its promise, the ZnO LED has failed to make any commercial impact. That is partly because material issues have hampered device development — realizing p-type doping in this material has been a major challenge, and there have also been differences of opinion over the design of the active region needed to yield a light-emitting device. While these issues have undoubtedly hampered the progress of the ZnO LED, they are not the primary reason why this device is still confined to the lab, according to Lawrence Gasman, Principal Analyst at the market research firm NanoMarkets. He believes that if a large firm had financed a major effort at commercializing ZnO LEDs, they could have ironed-out the technical issues in just a year or so. “I see the obstacles as being on the business side,” says Gasman, who points out that winning venture capital investment for novel semiconductor businesses is far, far harder than it was before the credit crunch. However, that’s not to say that there is no funding for ZnO LED development. This device has ‘green’ credentials, and some start-ups have won funding on the back of that attribute. Gasman expects the first ZnO LEDs to hit the market in two to four years time. “The LED market is slated to grow very fast now, and if ZnO can knock off a segment of that market in a time starting in three or fours years from now, then that [sub-section of the] market can be very substantial.” He believes that ZnO LEDs will make an impact and net $415 million in 2015. That will make this device the biggest earner for ZnO, which will also be used to build transistors and provide a transparent, conductive oxide for displays and photovoltaics (see box “The great versatility of ZnO”). Starting with ultraviolet The pioneers of ZnO LED manufacture will begin by making ultraviolet emitters. This is the simplest form of the device to produce, because ZnO’s bandgap is 3.3 eV. “The market for [ultraviolet lighting] is not huge, but it’s a good place to start, because you are not completely reliant on the whims of everyday consumers,” claims Gasman. Today lamps with fairly short lifetimes are the most common providers of ultraviolet emission, which is used for water purification and medical treatments. ZnO LEDs promise to be a far more reliable source. Another reason why ZnO LED makers will target the ultraviolet market first is that it will allow them to get their businesses off of the ground without having to compete with huge, well-established GaN chipmakers. “If you are in the UV business, you are talking about a much more modest supply chain [than the general illumination business], which a small firm could conquer with a couple of decent sales guys and a business development guy,” says Gasman.
Once the ZnO LED manufacturers have enjoyed some success in this market, they will add phosphors to their devices to produce white emitters. These promise to be cheaper than the GaN incumbents, which are starting to penetrate general lighting but need to drop in price to fuel the rapid adoption of this technology.
“The hope is that ZnO will give a range of colors that are more attractive than CFLs and GaN LEDs,” says Gasman. This could help to spur the sales of these ZnO emitters, because, according to Gasman, many people have the perception that GaN based LEDs produce a cold and harsh form of white light. “The companies that are working on ZnO LEDs claim that you can get better color quality from these LEDs.”
According to Gasman, another strength of the ZnO LED is the abundance of its constituents: “There is plenty of nitrogen around, but maybe there is some constraint on gallium. Remember that there has been a dance between cost and supply of indium, for indium tin oxide.” Although incredibly small amounts of indium tin oxide are needed to make a display, billions and billions of screens are made every year.
One difficulty facing the pioneers of any innovative technology is the reluctance of potential customers who may be suspicious of adopting the new materials. However, this should not be a major issue for the ZnO pioneers. “Samsung has announced using carbon nanotubes for backlighting and displays, and they are actually producing that now,” says Gasman. He is also aware of companies working with silicon quantum dots, and he sees ZnO as just another emerging material in this sector.
According to Gasman, there might be up to a dozen companies working towards the commercialization of ZnO LEDs. They tend to keep a very, very low profile. “In some cases they are funded by famous names,” says Gasman.
The progress of all of these firms is held back by the materials and processing equipment. One issue is that there are only about ten producers of crystalline ZnO, and the substrate sizes produced by them are too small for mass production. For example, the material offered by the Atlanta-based firm Cermet is either shipped as 10 mm by 10 mm squares, or 25 mm-diameter circular substrates.
Although ZnO LED developers would prefer to manufacture their devices on native substrates, partly because this should offer the best route to high quality material, these firms might begin by using sapphire. The difference in lattice constant between ZnO and GaN is only a few percent and both materials share the wurtzite crystal structure, so many of the challenges of growing these wide bandgap materials on sapphire are similar.
ZnO LED developers are yet to reach a consensus on the best deposition technique for epitaxial growth of the heterostructure. For many of these firms, optimization of the growth technology, which will form a key part of their intellectual property, is their overriding goal.
It is possible to buy commercial ZnO deposition tools, such as the range of MOCVD reactors made by Structured Materials Industries of Piscataway, NJ, that feature high speed rotating discs. These Spin CVD tools — which are available as single wafer, 1 inch or 2-inch tools, or multi-wafer variants capable of accommodating up to 38 wafers with a 2-inch diameter — employ a uniformly heated deposition plane and are capable of growth rates of 10-20 nm/minute (for more details see “Nitride LEDs get brighter with transparent ZnO contacts,” Compound Semiconductor September 2007, p. 14).
However, according to Gasman, many ZnO developers are not buying commercial tools and using them off of the shelf. Instead, they are either building their own tools or adapting commercial ones. In some cases, they are also pioneering novel forms of growth. For example, ZnO device specialist MOXtronics, which is based in Columbia, MO, has developed a hybrid beam deposition process that it claimed to be comparable to MBE. The unique features of this MBE-related approach are a ZnO plasma source, which is produced by illuminating a polycrystalline target with either a pulsed laser or an electron beam, and a high-pressure, oxygen plasma created by a radio-frequency oxygen generator.
Another issue that ZnO LED developers might be struggling with is the design of the active region. Some researchers are employing a homojunction, while others are considering which pairing of materials is best for forming the active region. Some of the early work on ZnO light emitting structures involved the combination of ZnO and MgZnO, but more recent research has shown that this ternary tends to phase separate when the magnesium content exceeds 33 percent, due to differences in crystal structure – MgO is cubic. One promising alternative is BeZnO, which shares the hexagonal crystal structure of ZnO.
There is no doubt that all these ZnO developers face an up-hill task. However, they will be motivated by the dream of creating cheaper, high quality ultraviolet and white LEDs. And if companies in the supply chain can share their vision and play their part, then many of the manufacturing issues will fade away. For example, although the crystalline ZnO substrates available today are to small for high volume manufacture, the technique to produce them, hydrothermal growth, is well established and churns out tones and tones of quartz every year. If such efforts to scale up ZnO are applied alongside other activities to improve the supply chain, then maybe the solid-state lighting revolution will turn out to be a twopronged affair: GaN LEDs and their ZnO cousins.
Lawrence Gasman is the author of the report Zinc Oxide Markets, 2010 and beyond.
