Lithium still rules in battery design but nano and biotechnologies could massively improve on what we have. E&T examines.
Things don't change quickly in battery design. The choice of materials to make better batteries is pretty limited and the core material - lithium - has been a feature of batteries with high energy densities for a long time already. The first lithium battery appeared in the early 1970s, although they were not rechargeable systems.
"Why lithium batteries?" asks Yi Cui of Stanford University rhetorically. "Well, lithium has the best energy density of all for batteries."
The problem for the battery industry is that, with arguably the best material already in place, improvements in energy storage come much more slowly than in the technologies that the batteries feed. "Compared with semiconductors, the improvement is not a whole lot. There has been a four-fold improvement in energy density over the past 150 years. The energy density improves only 8 per cent per year in terms of Watt-hours. And that is mainly through better packaging."
The key to a better battery, argues Cui, lies in the two electrodes. This is where there is potentially a greater choice in materials and the greatest room for improvement. Today's graphite electrodes have a reversible capacity of up to 370mAh/g. But other materials such as silicon, Cui reckons, can push that reversible capacity much higher, particularly for the cathode.
"Sulphur is a good choice for the cathode. How come the battery community has not used it yet?" Cui wonders aloud. The problem, he says, is the way that the electrode materials expand as the battery is charged and lithium ions are sucked into the structure.
"After one charge, you lose capacity and the materials start to break," explains Cui, adding that his proposed solution is to use nanowire structures. "They also see the volume expansion but they are small enough that they don't break."
The work so far has concentrated on replacing carbon with silicon in the electrode. Each silicon atom can combine with more and more lithium ions as the cell charges.
"There is a dramatic change in nanowire width, but they don't break," Cui claims. "The nanowires start as single crystals but as they combine with the lithium, part of the structure becomes amorphous. And after cycling, the nanowire remains amorphous.
"Another thing we find is that when you take the lithium out, the nanowire shrinks a little but it doesn't come all the way back. It means that there are a lot of pores in the nanowires. They allow the electrolyte ions to come in faster."
Seeing this happen, Cui's team considered what might happen if they moved to a two-layer structure. Build a crystalline core for stability and then coat with an amorphous form of silicon. "The single-crystal phase in the core remains intact and retains core stability. You can charge the battery extremely quickly this way."
Cui claims it is possible to charge a battery in less than 10 minutes using the nanowire electrodes. The charging capacity currently approaches 1000mAh/g, "which is already three times higher than [graphite] powder. Replacing carbon with silicon alone can almost double the capacity of the battery. We don't need to use high-purity silicon - certainly not solar grade. And there is plenty of silicon to make batteries."
A further benefit of a silicon-electrode battery may be safety, an important consideration when batteries keep getting recalled because some suddenly catch fire. "I believe silicon may be safer than carbon. The charging potential is slightly higher than that of carbon. This prevents lithium dinitride forming, which can cause shorts and explosions. And silicon produces sand when it burns; carbon produces CO2 gas."
The challenge now is to find a material for the negative electrode, Cui says, although the silicon electrode shows enough promise to start thinking about commercialisation. "The main obstacle now is scaling up to production," he claims, as well as safety testing.
Using conventional chemical vapour deposition techniques, similar to those used in chipmaking, Cui's team can grow nanowires at the rate of 10µm per minute, which is fast enough for making batteries commercially.
Biology: Viruses that build batteries
Biotechnology provides another route to nanowire electrodes. At the Massachusetts Institute of Technology (MIT), Angela Belcher has led a team that uses viruses to form the wires. Other teams are looking at using DNA as a way of forming templates for electrodes structures made from polymers.
The argument for using biology to build batteries is that the production processes can be much more benign than the harsh chemistry that might be needed for technologies that use techniques from the chipmaking business. Publishing their latest work in the 2 April 2009 issue of Science, Belcher's team claimed their approach works at room temperature and requires no toxic organic solvents. This may be a consideration in the automotive market where a drive toward hybrid systems is focusing attention on the environmental cost of battery production.
Three years ago, an MIT team led by Belcher reported that it had engineered viruses that could build an anode by coating themselves with cobalt oxide and gold and then self-assembling into a nanowire. The M13 viruses they use are good for building wires because they are long and thin.
In the latest work, the team focused on building the more problematic cathode structure. To achieve that, the researchers genetically engineered viruses that first coat themselves with iron phosphate, then grab hold of pre-made carbon nanotubes to create a network of highly conductive material.
Because proteins that encapsulate the viruses recognise and bind specifically to certain materials such as carbon nanotubes in this case, each iron phosphate nanowire can be electrically 'wired' to conducting carbon nanotube networks. Electrons can travel along the carbon nanotube networks, percolating throughout the electrodes to the iron phosphate and transferring energy in a very short time.
The viruses infect bacteria - and have the advantage of not killing their hosts, so that they can keep on producing M13 - but are harmless to humans. The team found that incorporating carbon nanotubes increases the cathode's conductivity without adding too much weight to the battery. In lab tests, batteries with the new cathode material could be charged and discharged at least 100 times without losing any capacitance; the first discharge capacity was measured at 165mAh/g. That is fewer charge cycles than currently available lithium-ion batteries, but "we expect them to be able to go much longer", Belcher said.
Qian Wang and colleagues at the University of South Carolina are working with M13 viruses as well as the rod-like tobacco mosaic virus to made nanowire based on purely organic conductors, such as polyaniline. Instead of using mutant form of M13, Wang's group use chemical agents to alter the proteins on the outer shell of the virus so that they bind more readily to the actual wiring molecules.
DNA templating offers an alternative route to constructing nanowires that, because they don't have a chain of viruses sitting inside, are much smaller. The ability to synthesise and then replicate DNA with specific sequences opens the door to more complex wiring structures as different conductor molecules will bind to different parts of the sequence. This may make it easier to build complex cross-linked or percolated structures that work well in battery electrodes, as with the nanotube-phosphate complexes that Belcher and her team grew.
Techniques like these may give the slow-moving world of batteries some much needed acceleration in the coming decade.
Modifying a virus to attach itself to different molecules or particles takes several steps. In principle, it should be possible to design proteins that the virus will use as a coat. In practice, protein design has not reached this stage.
So, designing a new protein coat involves the production of many different mutant forms of M13 and then screening for those that bind well to a target - in the MIT team's case either iron phosphate or carbon nanotubes.
Once the team identified the best versions of coat proteins that could bind to the target molecules, they inserted the DNA that codes for them into the M13 virus that could then go on and infect bacteria to produce more copies. Simply mixing the virus particles with the saw the viruses form into structures. By having iron phosphate attach to a protein that shrouds the long sides of the virus and having a different protein found only on the caps attach to nanotubes, the mosaic virus particles formed themselves into a complex network of fibres.