Graphic LED chip technology three schools

I mentioned last year, LED competition is the competition of technology in the past two years will be, but cannot do without the two laws: "more and more advanced technology, especially enhance the optical efficiency; the cost will be more and more low", especially in the chip, reduce packaging and power cost, and of course the lamp assembly how can automation, reduce the manpower cost. In the LED industry for almost 16 years I saw a blue chip by a five dollars (in 1998) to one thousand now 5 dollars (5 RMB/k), the industry progress too fast, even faster than the integrated circuit, I witnessed its growth, I also participated in the his progress. Now for more than a decade of my experience with LED technology to share with you, especially in the upper reaches of the technical aspects.

I put the GaN LED technology school is divided into three parts: the first one is the vertical structure, with CREE and OSRAM as the representative of the metal substrate attached to the silicon substrate and Xuming, Puri Toshiba and crystal, of course, also can not forget the many Japanese manufacturers and Mr. Nakamura Shuji in the development of Gan homostructure. The second is sent (flip Chip), flip flip thought is of course PHILPS Luminled, of course, the current Taiwan Jingke and many chip factory in the development of this chip, even CREE has started doing this kind of product, so the technology and yield are constantly mature. Third of course is the sapphire substrate structure, the mainstream of the current LED, almost every company based on this structure make great improvement, said 16 years of progress is always one thousand times, according to the structure of continuous improvement.

The three schools have their own advantages and disadvantages:

Vertical structure:

A bit like Ming or sun moon god feeling, always is not the mainstream, but also have unique technology, to achieve the ultimate if like CREE or silicon carbide technology with mainstream contend, in Jin Yong's novels, whether it is Yang up Zhang Wuji, imprisoned or they can meet as equals to Invincible Eastern, Shaolin Wudang, but still not the mainstream or like a meteor disappeared. Vertical structure is the case, it has always been in the non mainstream line, has been published in the new technology, but in the market has always been difficult to find them. In addition to the vertical structure of silicon carbide and homogeneous substrate, need to process very complex to manufacture, regardless of silicon substrate or the metal substrate, and gallium nitride thermal mismatch problem has not been resolved, the structure yield is very low, although the good thermal conductivity, substrate than sapphire cheap, light emitting area is relatively large, but always not sapphire in the structure of cost and technology competition opponent. Because the upper and lower electrodes, the application is limited, especially not satisfy many requirements in the lamp circuit design need to string and can make high voltage chip, so the vertical structure of CREE's chips can be used only in specific market street price is high, the general lighting and backlighting of the two mainstream vertical market the structure of chip has been unable to intervene heavily. Of course, we can not ignore the homogeneous structure the Japanese company and Mr. Nakamura Shuji are developing, it is not mentioned shortcomings, but a fatal problem is enough to beat it all the advantages of the gallium nitride substrate is ridiculously expensive, two years ago was $1000, is now more than 500 dollars, LED will not be too extravagant? So now homogeneous substrate can only be used to make blue laser diode for HD-DVD player or a high quality PS3 game console. We will ask me whether there is room for price cuts? My answer is yes, but limited, currently only two kinds of mainstream GaN substrate approach, HVPE and ammonia amonothermal method by hydride vapor phase epitaxy, estimated to be reduced to 100 dollars will be competitive, but five years in accordance with the current technology are not up to.

[1] [2] next page