The liberation of light

A member of staff at Philips sweeps her finger over the screen of her smartphone. All the rectangular lighting elements with gas-discharge lamps high in the ceiling of the truck assembly hall suddenly rotate a quarter turn, and LEDs (light emitting diodes) take their place. It’s like the rotating number plates on a James Bond car.

We are in one of the many demonstration areas for lighting at the Philips Light division in Eindhoven. The hostess from the electronics company is demonstrating the advantages of LEDs to a group of a dozen PhD students from Delft.

LEDs are energy-saving, you can dim them, they have a life of twenty thousand hours and they are small, so they can be worked into all manner of materials. According to Philips, the LED is the light source of the 21st century.

Not much needs to be explained to the young researchers. They are all involved in LED-related research in close collaboration with Philips. The visit is an interesting outing for them.

Retrofit is stupid

The PhD students’ supervisor is Prof. Kouchi Zhang, who has been part-time professor at the Delft Institute of Microsystems and Nanoelectronics (Dimes) for two years. Zhang, who also works for Philips, is a Messiah of the new light. “By 2020, three-quarters of all bulbs will be LEDs”, he says.

Will they all be LED bulbs that you have to screw into a socket? Zhang hopes not. “I predict the liberation of light. Retrofit is actually very stupid. If you power LEDs with batteries – and we’re getting better and better at that, because they’re becoming increasingly energy-efficient – then you no longer need any of the infrastructure with wires and sockets.”

LEDs are nothing at all like incandescent bulbs, energy-saving bulbs, gas-discharge lamps or fluorescent tubes. They are chips that work like a solar cell in reverse.

The semiconductor materials in solar cells absorb photons and current begins to flow, but in LEDs, flows of electrons through the semiconductor materials result in photons being emitted. And they currently produce 150 lumen per watt, ten times more efficient than an incandescent bulb.

LEDs with smoke sensors

The fact that LEDs are chips means that all sorts of electronics and sensors can be connected to them. What about LEDs fitted into window panes, to measure how much sunlight is shining into a room and provide light when it gets dark outside? Or LEDs with temperature and smoke sensors that mark out the safest escape route when there is a fire?

All these things could be possible, claims Zhang’s colleague, Dr Henk van Zeijl. “Ever since Edison invented electric light, it has been imprisoned inside glass. Until shortly after the Second World War, electronic functions were realised with glass; glass radio tubes. When transistors made it possible to make electronics without glass, this brought about a revolution. Something similar is now happening with lighting.”

The research of the Delft PhD students focuses on many different aspects relating to electronics and sensors for LEDs. But the housing for the chip – which to a large extent determines the colour of the light – and techniques for preventing the bulb from overheating are also important subjects for research. (See box ‘A selection from the LED research at Delft’.)

With a total of fifteen PhD students and post-docs, Zhang says his group is one of the major academic players in Europe. But if you search in the scientific literature using the keywords ‘light emitting diode’, you’ll find many other research groups elsewhere in Europe (e.g. Germany, Switzerland and France) that are publishing much more on the subject. And Eindhoven University of Technology is scoring well too in that respect.

Is the Dimes group really such an important player? “What we’re doing is micro-electronics and lighting systems integration,” responds Zhang. “That’s what our research focuses on, not on the chip itself. You’ll see that we lead the field in Europe in that respect. And many of our articles still haven’t been publishes yet, because the group was only established a few years ago.”

Prof. Paul Urbach of the Optics research group at the Faculty of Applied Sciences thinks that TU Delft is a major academic player in Europe in the field of LED research. He regularly has students and PhD candidates who are conducting research into LEDs, usually in collaboration with Philips. “With Philips just around the corner, it’s logical that we’re so big in this field,” he says.

Zhang wants to extend his group in the coming years, and work more closely with colleagues from Applied Sciences (including Prof. Urbach), 3mE and Architecture. He wants to work with them to set up a Centre for Solid State Lighting (another name for LEDs).

A long way to go

The group of visitors in Eindhoven has now reached a dark research area where researchers are testing whether lamps shine evenly in all directions. Despite all the fascinating possibilities, Philips’ efforts, as far as the consumer market is concerned, remain largely focused for the time being on LED bulbs that can be screwed into a socket. One such bulb – the L bulb – was recently tested here.

Philips developed the L bulb to replace the sixty- watt incandescent bulb. In 2011 it won the company ten million dollars (the Bright Tomorrow Lighting Prize), a prize awarded by the U.S. Department of Energy. The LED bulb consumes ten watts and should last for twenty thousand hours.

PhD student René Poelma has a similar LED bulb at home, he says. “It gives very fine light. Very diffuse. You can’t tell that it’s an LED bulb rather than an incandescent bulb.”

Yet Poelma doubts whether we are on the eve of an LED revolution. “Personally, I wouldn’t have bought the bulb. I think it’s still too expensive. In my view there’s a long way to go before they’re interesting to consumers.”

The award-winning Philips bulb still costs sixty euros. But according to Zhang, there are already high-quality LED bulbs on sale for much less than this. And very rapid advances are being made. “In China you can already buy a good bulb for ten euros.”

How does a light emitting diode work?

There is a chip the size of a grain of sand on the desk of Dr Henk van Zeijl at Dimes. That’s all it is. “When I pass current through it, you mustn’t look directly into the LED,” he explains. “The light is so bright that it would blind you.”

In the chip’s semiconductor material there are freely moving electrons and electron holes. When a current is passed through the material, the electrons move through the material until they come to a hole that they can fill. When the electron drops from a high energy state to the lower energy state of the electron hole, light is emitted.

Van Zeijl: “Think of the electrons as water, and the difference between the energy level of a moving electron and that of an electron that has just filled a hole as the height of a waterfall. The height determines the colour of the light that is emitted. For every ten buckets of water that flow down the waterfall, eight or nine are converted into light. The internal efficiency is therefore as much as eighty to ninety per cent.”

However, part of that light does not leave the chip; it is absorbed again soon after, reducing the efficiency further, to around thirty to forty per cent. But the expectation is that chips will become much more efficient.

LEDs have been used in electronics since the 1960s. For decades, the chips emitted only weak red light. Since the 1990s, following much experimentation with alloys, clear blue LEDs have been made using the semiconductor materials gallium nitride and indium gallium nitride. These blue LEDS can be combined with green and red LEDs to produce white light. But white light is usually created by coating the blue LED with phosphors (these are usually rare earth metals) that turn the blue light that is emitted into white light.

A selection from the LED research at Delft

LED bulbs have an average life of twenty thousand hours (twenty times longer than that of incandescent bulbs). But, thanks to technical advances, they could soon last for a hundred thousand hours, believes PhD student Sima Tarashioon. To achieve this we need to know what the life is of all the components of a bulb, and how they affect each other. Tarashioon is using simulations to study this. Her work focuses on the electronics of the bulb. Her field is known as ‘the physics of failure’.

PhD student René Poelma is interested in the mechanical behaviour of materials on the nanoscale. His work for Zhang’s group has focused on, among other things, the elastic properties of silicons. LEDs are produced on a wafer, to which silicon lenses are attached one by one, by hand. If all the lenses could be attached at the same time, the cost saving would be huge. But the lenses would have to release easily from the mould above. Poelma’s research is geared to finding the optimum conditions for this.

An incandescent light bulb has a filament that literally glows when heated. But LEDs can also become quite hot. More than half the energy you put into them is converted into heat. The heat is the limiting factor in the life of LEDs. Most LEDs have a passive cooling system: small metal structures that dissipate the heat. Msc Huaiyu Ye has come up with something quite different. The PhD student, who studied cooling techniques at the University of Shanghai, has developed a cooling system that uses cooling fluid. A few millilitres of water per second flow past the LED through a very small circular tube. This is enough to keep the temperature of the LED at a constant one hundred degrees Celsius. And the good thing is, this happens automatically. Due to the phase transition of the water, from liquid to gas and back again, the water flows by itself. As long as the light is on, of course.

Miniature retrofit halogen bulbs are small and give a great deal of light, but they are certainly energy-hungry. Philips wants to produce LED bulbs that look like halogen bulbs. The major challenge is to ensure that the LEDs, which are enclosed in a small space, produce enough light without overheating. There is no space in the bulb for cooling systems. Pan Liu believes she has found the solution. She is making foldable chips to hold LEDs. Five LEDs linked to form an L shape can be folded into an open cube that emits light on all sides. The open structure allows the heat that is created to dissipate.



Together with China

The Delft LED researchers are working in close collaboration with fellow researchers in China. TU Delft has had a research branch in Beijing since 2011. The branch has seven PhD students who are supervised from Delft, and the plan is to increase this number to ten in the course of this year. The collaboration works in both directions. For several months now, Dimes has housed an office of the Chinese State Key Laboratorium (SKL).

The collaboration seems to be a sensible move. “With a view to saving energy, China earmarked 22 billion yuan (2.8 billion euros) in 2011 for the promotion of green lighting, primarily LED lighting.