Out of the red and into the blue: making the LED revolution cost-effective

British manufacturer Plessey Semiconductors is racing to be the first company to make energy-efficient LEDs for home lighting at a price that consumers will pay, and they’re using a technology developed by Cambridge researchers.

In 2012, Plessey acquired technology to grow a remarkable man-made material that can emit light in every part of the colour spectrum when electricity passes through it. They recommissioned a mothballed processing plant, created new jobs and hired three researchers from the University of Cambridge. Their aim: to put energy-efficient lighting within financial reach of the consumer.

Prototypes of their light-emitting diodes (LEDs) rolled off the production line later that year, and by April 2013 the company was gearing up to fulfil its first commercial orders. In just 15 months, Plymouth-based Plessey had gone from never having made an LED to being the world’s first manufacturer of commercially available LEDs made on large-diameter silicon substrates.

Today, the company is addressing a global market that, according to a report released in 2013 by WinterGreen Research, could be worth up to $42 billion by 2019.What gives Plessey an edge over its competitors is its ability to manufacture LEDs at a fraction of their costs, thanks to a unique process developed by Professor Sir Colin Humphreys in the Cambridge Centre for Gallium Nitride.

Blue and white gallium nitride (GaN) LEDs have been commercialised around the world since Shuji Nakamura in Japan developed a method of growing thin GaN layers on sapphire in the early 1990s. Although GaN LEDs are now expected to dominate the world market for lighting, their performance and cost both need to be improved.

Humphreys’ team has developed a way of growing GaN on the vastly cheaper substrate silicon and, crucially, a means of scaling this up for commercial purposes. “We’ve got lower costs for growing GaN LEDs on silicon than anyone else we know,” explained Humphreys. “Potentially, this is an advantage that puts Britain right at the forefront of LED research.”

Competition between manufacturers (including Toshiba and Samsung) to lead the market in competitively priced LEDs has been intense, driven by the increasing demand for energy-efficient lights.

LED bulbs have much longer working lives than any other forms of artificial lighting: LEDs can last for 100,000 hours compared with 10,000 hours for fluorescent tubes and 1,000 hours for tungsten filament light bulbs. LEDs in dashboards frequently outlive the life of the car; LED light bulbs in the home would probably have to be changed only once in a person’s lifetime.

LEDs also use less energy than other forms of lighting. UK homes use 20% of their energy on lighting and, because LEDs use 90% less energy than incandescent bulbs, Humphreys estimates that the superior energy efficiency of LEDs could save the UK £2 billion per year in energy, and reduce CO2 emissions. It’s little wonder that LEDs have been hailed as a lighting revolution.

Yet, few homeowners have invested in LEDs for home lighting. “A 48-watt equivalent LED bulb costs around £15. Although it will save people money over its lifetime, very few people will pay this. We think we can reduce the cost to £3,” said Humphreys.

Humphreys’ team has pioneered a technique for depositing successive layers of GaN and indium GaN, each only 5–10 atoms thick and growing at the speed of grass, on a six-inch silicon wafer. The wafer is then cut into up to 150,000 pieces, each of which forms the heart of a small LED. Using this technology, Plessey hopes to become the commercial leader in GaN-on-silicon LEDs, producing billions per year.

“Growing GaN on silicon is quite a complex process,” said Humphreys. A particular problem is the appearance of cracks on cooling from the growth conditions of 1,000°C. This has now been solved by the researchers through careful balancing of the tension in the material as it cools down.

Plessey acquired the technique when it bought the spin-out company CamGaN, set up by Humphreys and colleagues to commercialise the technology. The size of the silicon wafers is greater than the conventional two-inch sapphire wafers, meaning a greater number of LEDs can be made. Fortuitously, Plessey had a six-inch processing line that had been mothballed.

“When we launched CamGaN, we were contacted by companies all over the world wanting to utilise the technology,” said Humphreys. “The research has been funded by government money through the Engineering and Physical Sciences Research Council (EPSRC) and it was important to us that this research be exploited here in the UK.”

Meanwhile, the researchers are continuing their work with what Humphreys describes as a “truly remarkable” material. A new £1 million growth facility funded by the EPSRC and made by AIXTRON Ltd, a long-term collaborator of Humphreys’ research group, has been installed in Cambridge, where the researchers are adjusting minute aspects of the growth process to improve the efficiency of light emission.

The benefits of increased efficiency could go far beyond home lighting, and the researchers are now looking at applications that extend from biomedicine to power electronics.

In collaboration with the University of Manchester, they plan to build tiny LED devices that can be implanted by keyhole surgery in cancer patients being treated with radiotherapy. “If a patient moves while an X-ray or proton beam is directed at their tumour, then they risk healthy tissue being damaged,” explained Humphreys. “An LED attached to a sensor would detect movement at the site of the tumour in order to redirect the beam.”

Even tinier devices are being investigated as a means of firing single neurons in the brain. Working with US researchers involved in the Brain Activity Map Project – a flagship initiative of the Obama administration – the researchers will supply LEDs that can be implanted in the brain with the goal of mapping the activity of every neuron, to understand how the brain works both in health and disease.

Water purification is another area where LEDs could benefit the health of millions. “Life on earth has developed in the absence of deep UV and so we can kill bacteria and viruses by damaging their DNA with deep UV light,” he said. The idea is to have a ring of LEDs powered by solar cells on the inside of water pipes. “We now have LEDs with the right energy level to do this, we just need to increase the intensity.”

Collaborative projects with the Universities of Glasgow and Strathclyde are also investigating GaN transistors as power electronics in devices that manage electrical energy and LEDs as light-based Wifi (so-called Li-Fi). Humphreys believes that the day will arrive when GaN devices will light our homes, power our mobile phones, laptops, cars and aircraft engines, and connect us wirelessly to information transmitted from traffic lights and street lights.

“In addition to the £2 billion per year in electricity savings from GaN LEDs, GaN transistors could help the UK save £1 billion from power electronics and 15–20% in carbon emissions, and GaN UV LEDs could save millions of lives,” said Humphreys. “It’s the ultimate shopping list.”

Plessey, in the meantime, is focusing on product improvement and reduction in cost. “There are huge players in the market but we were first to market with GaN-on-silicon LEDs,” said Dr Keith Strickland Chief Technology Officer of Plessey. “Our continuing relationship with the Cambridge research group will help us push this technology to its highest potential.”