VITALY PODZOROV’S desk is decked with transparent boxes containing brightly coloured crystals hovering on hair-fine silver wires. Each is a transistor, similar to the ones that switch current in the circuits of every computer. The difference is that they are not made of silicon but organic materials – crystals of hydrocarbons, in fact. And their path from idea to reality could not be stranger.
Four years ago, a Bell Labs physicist was dazzling his peers with his work on electronic devices using organic crystals. At the height of his career, Hendrik Schön was publishing results at the remarkable rate of once every 8 days. Then it all collapsed in scandal.
When other researchers found they were unable to reproduce his results, Schön’s work eventually came under intense scrutiny. An investigation revealed that Schön had fabricated the results in around a dozen papers. He was fired, and the field of organic electronics seemed to be in tatters. But it wasn’t.
Advertisement
While some were keen to airbrush Schön out of history, a few determined researchers refused to give up: Schön’s work had fired their imagination. “The work in my lab was inspired by Schön,” says Mike Gershenson, who runs the lab where Podzorov works at Rutgers University in Piscataway, New Jersey. “Without him, I would not have dived into this field.”
In a discipline where careers can be tainted just by association, that is a brave statement. But Gershenson is simply acknowledging the facts, he says. There was a very good reason why Schön was pursuing organic transistors, however dishonest his methods: they offer a radical and potentially lucrative new way to build cheap, light and flexible electronics. And that reason still stands.
The silicon transistors on today’s circuit boards work by means of an electric field called a gate. When the gate field is on, it brings charges from within the silicon crystal up to the surface. When the density of charge on the surface is high enough, the crystal begins to conduct electricity there, so the circuit is on. Switch off the gate and the charges fall back into the crystal’s interior, breaking the circuit.
It should work in organic crystals too. And if it does, it could lead to stronger, lighter and flexible plastic alternatives to rigid silicon chips. Several researchers at Bell Labs were interested in testing the idea in organic crystals in the late 1990s, including Bertram Batlogg. In 2000 Schön published a paper describing the impressive properties of a new transistor made of a semiconducting crystal of pentacene. The crystal had a thin insulating coating of aluminium oxide to shield it from the gate electrode and prevent short-circuiting.
“Results were being published at the remarkable rate of once every 8 days. Then it all collapsed in scandal”
Gershenson remembers this paper being the talk of the condensed matter physics department at Rutgers. And as he was interested in the behaviour of charges in materials, he decided it was time for his lab to start work on organic electronics. He encouraged Podzorov to postpone the completion of his PhD – on an entirely different topic – and try his hand at making organic crystal transistors.
The same thing was happening in many physics departments around the world. Unlike some others, Podzorov’s department had no background in organic electronics, so he started from scratch. He learned to grow organic crystals himself, instead of obtaining them from other people. And when, like everyone else, he could not get devices based on Schön’s reported design to work, he combed earlier papers for clues. Eventually he found a description of transistor-like behaviour in an organic crystal from 1998. In that paper Schön used a crude method of insulating the crystal: he merely stuck a thick layer of a polymer called Kapton around it. Podzorov found this technique worked for him, too.
The aluminium oxide coating was notorious as the aspect of Schön’s design that had proved most difficult to reproduce. Schön’s method was to expose a disc of aluminium oxide, along with the organic crystal, to a cloud of plasma. The plasma vaporises the aluminium oxide and “sputters” it everywhere, including onto the crystal. Many researchers tried fine-tuning their sputtering machines in the hope of getting the experiment to work for them, but without success. It was Podzorov who worked out what the problem was.
One day late in 2001, he took a crystal that he knew worked fine when insulated with Kapton. “I put it in a plasma for a fraction of a second,” he recalls. Then he tried the crystal in a circuit. “It was destroyed.”
Podzorov concluded that an organic crystal that has been in a sputtering chamber cannot be used to make a transistor. “This is exactly the right scientific process, to do it one step at a time,” says Horst Störmer, the Nobel prize-winning physicist who has spent many months trying to understand sputtering at Bell Labs.
So what to do instead? Podzorov remembered that, while working as an intern one summer at Honeywell Technologies in Morristown, New Jersey, he had seen circuits packaged in a thin clear layer of a polymer called parylene. The coating was applied simply by condensing parylene vapour onto the circuit. Podzorov tried it with another hydrocarbon crystal called rubrene and two months later he had a working transistor (see Graphic).
This was early 2002. But Podzorov couldn’t publish immediately. Reviewers wanted more evidence for his findings. It took Podzorov another 8 months to convince reviewers of his results. Meanwhile Schön’s highest-profile papers were being investigated and retracted. In some cases Schön’s co-authors noted, “this paper may contain some legitimate ideas and contributions”, though the fiasco forced many researchers to leave the field in disgust.
While Podzorov struggled to get his work published, a similar story was unfolding in Europe. Ruth de Boer at Delft University of Technology in the Netherlands was sputtering aluminium oxide on organic crystals without success. As she handled the crystals, using glass microscope slides, she hit a problem. “I noticed the crystals always stuck to the slides. I would pick them up with tweezers, but mostly they would rather break than come loose. It was annoying,” she says. The cause was static electricity, and she realised that rather than being a problem, it pointed to a solution. With such a strong attraction between an insulator, in this case glass, and the crystal, de Boer realised there was no need to use a sputtering chamber at all.
All she had to do was lay down her gate electrode first. Then she covered it with an insulating layer of glass, and only then did she add the crystal. Sure enough, it stuck fast, and soon she was making her first measurements. “It was 14 June 2002,” says de Boer. “I remember the date, because I drew balloons and fireworks in my logbook.”
It was not long before other people wanted a slice of the action. One of them was Christian Kloc, who had grown the crystals for Schön’s experiments at Bell Labs: though his name appeared on the retracted papers he, Batlogg, and all other co-authors were cleared of any misconduct.
“Organic transistors offer a radical new way to build cheap and flexible electronics”
Kloc is famed worldwide for his proficiency at growing crystals. His lab looks like a candy store, full of all kinds of shiny orange, black and purple fragments, with dates on them going back to 1998, the year in which he and his late friend Bob Laudise taught the world the simplest and best way to grow crystals. In a detailed paper in the Journal of Crystal Growth (vol 187, p 449), the pair described putting an organic chemical into one end of a glass tube, heating it, and allowing the vapour to be carried in a stream of pure hydrogen to the cooler end of the tube, where the vapour crystallises. Some tricks are not mentioned in the paper, though. For example, Kloc’s equipment is covered with signs reading “Do not bump”.
Chemists from all over the world bombard Kloc with compounds they have synthesised, hoping to have them transformed from dull powders into sparkly crystals. But after the Schön scandal, he decided he was no longer just a crystal grower and started trying to make transistors. “I wanted to do it myself,” he says.
He started out optimistically using the sputtering technique Schön had described but he failed to make it work. Only after hearing about Podzorov and Gershenson’s work did he finally switch from aluminium oxide to parylene, and in 2003 made his first transistor. Now he is making them on crystals of rubrene and other materials.
Today all these groups are making transistors with many of the properties first claimed in Schön’s reports. For physicists, the attraction of making organic transistors from single crystals is the ability to study how current flows through plastics. The groundwork for this was laid in the 1980s by Norbert Karl at the University of Stuttgart, who measured the speeds of charges moving through organic crystals. He found signs that the crystals conducted electricity in a completely different way to metals and inorganic materials such as silicon.
In silicon, charge is carried by the flow of positive or negative charges. But in organic crystals, these charges also induce clouds of polarised charge in the atoms of the crystal lattice. These clouds, called polarons, move through the crystal, and they can carry most of the current.
All three groups produced results backing up the idea that polarons carry electrical current in organic crystals, not only inside organic crystals, as Karl found, but also on their surfaces – a breakthrough, considering that surfaces are often messy, dirty places. So has Jun Takeya at the energy company CRIEPI in Tokyo, who independently had the same idea as de Boer at the same time while working with Batlogg at the technical university in Zurich, where he moved in 2000. “The findings of the various groups substantiate each other,” says Batlogg, who moved to the Swiss Federal Institute of Technology in Zurich in 2000.
The new transistors fulfil a long-standing scientific goal, but they also throw up some puzzles. For example, changing the insulator seems to change a transistor’s performance. This was a shock: previously, researchers had assumed that the insulating layer would have no effect on an electrical circuit. Alberto Morpurgo, who works with de Boer, thinks this is because polarons are very sensitive to their immediate environment – the crystal surface.
And last December Podzorov reported a novel phenomenon: he had noticed that rubrene crystals are actually true conductors. Unlike other organic crystals, they conduct a small amount of electricity even when there is no gate field on them. This means that a rubrene transistor will leak slightly when it switched off.
As he tried to find a way around this problem, he became intrigued by the idea of moving the gate electrode to the opposite side of the crystal while at the same time reversing its polarity – making it positive instead of negative. He naively thought this would have no effect because the electric field inside the crystal would be the same as before. But when he tried it, he was surprised to discover he had inadvertently made a new type of transistor – one that switches off when a gate voltage is applied. One advantage of the new design is that it remains off in spite of rubrene’s conductivity.
By chance, he found the transistor could be switched on again by shining light on it. A light-triggered reaction was freeing up electrons held in the crystal by the positive gate voltage, and these made it conductive once more. So some of Podzorov’s transistors are light-switchable, a development that could eventually lead to optical circuits based on organic materials.
In the end, a small but robust core of research triggered by Schön did survive the misconduct scandal and inspire others to take the ideas forward. But many people are troubled by it. They worry that researchers will be encouraged to fake results if they think that science could eventually vindicate them. Others play down Schön’s role. “This is just science getting on with its life after a perturbation,” says one physicist, who prefers not to be named. Störmer is certain that physicists studying the properties of organic crystals would have figured out how to turn them into transistors without Schön.
No one condones what Schön did, yet there is something defiant about the researchers who decided not to quit. “If you went into the field because of the hype, you will leave,” Morpugo says. “But if you went in because you thought about it, you will have worked seriously and got somewhere.”
Just as Kloc has done. Just before I leave his lab, he hands me a sheaf of papers. On the bottom is his recipe for growing organic crystals. On the top is a paper written with Gershenson and Podzorov. In the middle is a single sheet of retractions. There again is the statement, “this paper may contain some legitimate ideas and contributions”. These days, it looks different, highlighted by Kloc with a fat green marker.
