China seeks LED solution to an energy-resource problem

04 Feb 2005

Despite being a vast country, China's energy resources are relatively scarce. As a result, the government sees LED manufacturing as a key part of the country's future as its rapid economic expansion puts an increasing strain on power. Michael Hatcher reports.

With its fast-growing economy, China is using energy at an ever-increasing rate. The total electricity consumption in the world's most populous country is estimated at 2000 billion kWh this year, with 10-12% of that figure devoured by lighting applications.

With electricity use growing at 10% per year on average, it's not surprising that the Chinese government sees efficient semiconductor-based lighting as a key part of its - and, for that matter, the wider world's - economic future. In 5-10 years, China's lighting energy demand is expected to double.

Energy savings

Until about a month ago, Gordon Liu was the president of China-based Lumei Optoelectronics, the company that acquired the LED manufacturing technology previously owned by US-based substrate supplier AXT last year. Speaking at the recent CS-MAX conference in Monterey, CA, Liu estimated that, if widely implemented, LEDs could save China 400 billion kWh of energy over the next decade.

To put this figure in perspective, it is worth noting that China's six operational nuclear power stations have a combined annual output of around one-tenth of this figure. The three-gorge dam power plant, the biggest of its kind in the world, is expected to produce 33 billion kWh in 2004.

This requirement for greater energy-efficiency, plus the Chinese government's motivation to compete with Taiwan's successful entry into the global optoelectronics industry, has propelled LEDs to the top of the political agenda. What the Chinese government is desperate to avoid is a repeat of the way the silicon industry's IC manufacturing model developed, where China was left behind as Taiwanese firms such as United Microelectronics Corporation and Taiwan Semiconductor Manufacturing Company became two of the biggest device foundries in the world.

Already established as a prime location for light-bulb manufacture, China has now set up four geographical bases with a remit to develop semiconductor-based lighting. These bases are located in Dalian, Xiamen, Shanghai and NanChang (see map), and so far China has more than 10 chip-manufacturing firms and more than 200 LED packagers. Liu now expects to see more companies established as an acceleration in government funding over the next five years kicks in.

Of the 12 chip manufacturers identified by Liu, six are said to be producing devices in volume, while three are in pilot production. The remaining three companies are at an earlier stage in their development (see table).

Although "traditional" LED manufacturing was first established in China back in the 1980s, it wasn't until 1996 that the country's first AlGaInP project was initiated. Volume production followed in 1998, and a similar nitride project that kicked off in 2000 spawned the country's first foray into volume production of GaN-based devices only last year.

Lumei, which was spun off from AXT in September 2003, finished construction of its manufacturing facilities a year ago and began pilot production in May, says Liu. It has since ramped up its production volumes of blue and green high-brightness LEDs (HB-LEDs; Lumei's devices emit in the 450-540 nm range) and is now said to be making and selling these devices in large volumes.

Boom time

With the formidable size of the domestic market, plus the backing of the Chinese government and the likely positive effects of events such as the 2008 Summer Olympics in Beijing, the opportunities for Chinese LED makers appear to be very good. However, Liu does warn against some negative aspects of the emerging industry that must be addressed before China can develop into the dominant force that it threatens to become. These include the lack of a complete industry food chain in the country; a system that tends to alienate some high-quality staff; and a substantial lack of intellectual property originating from within the country. Allied to this, says Liu, is the failure to convert some high-quality research at the country's universities and research centers into industrial processes.

Perhaps most tellingly, there is as yet no sign of a company emerging to become the Chinese equivalent of a Cree or a Lumileds; no true technology pioneer that could become a leading brand among high-power chip manufacturers with high-end products to compete with those of the leading companies in Japan, Europe and the US.

Lumei's brightest devices emit up to 290 mcd in the green region at 525 nm, and could therefore be justifiably rated as HB-LEDs, but their brightness still lags well behind that of the leading power chips produced by the likes of Nichia. The Chinese government is trying to change that situation with the sponsorship of technological programs to develop power LEDs, including the Ministry of Science and Technology's semiconductor lighting program set up in late 2003.

Kate Conway is an independent technical consultant for energy-efficient lighting, specializing in LEDs. She has been involved in helping to draft the Chinese government's "five-year plan" for energy efficiency, which begins in 2006. Conway says that the government's plan is to quickly fund a market for high-power LEDs, while simultaneously building up the technological capability to meet that demand through local production. Although building up the market through funding shouldn't prove problematic, Conway is less confident about the ability of the Chinese to meet technological requirements.

China already has a voluntary program aimed at building market share for energy-efficient lighting. The China Green Lights project has been running since 1995, and Conway is hoping that its remit will soon be extended for another five years. She says that if it is, then from 2006 the program would focus on leapfrogging incandescent and fluorescent lamps with LED-based technologies.

Similar programs include the Efficient Lighting Initiative (ELI), which resumes in 2005 with its administrative headquarters moved to China. It should help to address the quality concerns over Chinese products with its quality export mark for devices such as LEDs used in traffic signals. The ELI is funded by the International Finance Corporation and the Global Environment Fund.

The World Bank has also been heavily involved in a project that is seeking to bring solar-powered LED lighting to families in rural China who do not yet have access to an electricity grid. Purchasers of a photovoltaic system receive a fluorescent lamp and a 1 W LED lamp made by Sunpu Solar PV Technology of Beijing under the scheme.

Changing attitudes

iSuppli analyst Jagdish Rebello believes that China's reputation as a source of only low-value LED products will soon change. "Chinese LED manufacturers will not continue to be bottom feeders," Rebello told Compound Semiconductor. "It would be foolish to expect the region to remain focused on low-value devices." Having used its low-cost advantage to gain entry into LED manufacturing, Rebello says that it is only a matter of time before China competes on both price and performance through licensing deals and the development of proprietary technologies.

As always with China, there is a question mark over whether that technology will be developed "ethically" through licensing deals and proprietary technologies. Certainly it appears that high import taxes are currently deterring Chinese companies from working with those in the West . But one important area where there has been some progress is in product-testing procedures, with Conway saying that Chinese companies involved in LED test and measurement are keen to employ standard CIE tests within the next year.

Rebello estimates that around 7-10% of LED chips are currently made in China, and that the industry will follow the precedent set by companies in the telecoms industry, where Chinese operations that started off making low-value passive products have now moved into higher-value active components.

Although massive investment from the government would undoubtedly be required before China can become a serious producer of high-power chips for general lighting, Liu is certain that the increase in production capacity will eventually pay dividends. He says that if semiconductor technology became responsible for 20% of all lighting in the country, the energy demand from LEDs would be 100 billion kWh per year. The power needed for this would be 50 GW.

His estimate is that to produce this output from LEDs in one year would demand a $7.5 billion investment in MOCVD equipment. A huge expense, for certain, but the alternative facing China is to build more facilities like the three-gorge dam and at $120 billion, the cost of that really is prohibitive.