Batteries are the most disappointing part of the modern technology experience. They don't last long enough, they charge slowly, and their capacity degrades over time. Technology moves at a rapid pace. Cell phones seem dated only 6 months after they debut. When newer phones come out, they are faster and have bigger, better screens but the battery life is always about the same.
Processors get exponentially better with time, as described by Moore's law. Magnetic storage has Kryder's law whereby there is an exponential increase in the amount of storage. But batteries just plod along. What does it take to make a better battery? It turns out that there are specific physical limits the properties of batteries, so doubling the energy density, for example, would be a significant undertaking.
Let's continue with the comparison to a processor. The fundamental unit for a processor is the transistor. Transistors decrease in size so that the density on a chip doubles every two years. For a sense of scale, consider 22 nm processing. The chips made using this process have a half-spacing between transistors of 22 nm, so the full spacing is 44 nm: this is equivalent to 200 silicon atoms laid end to end. That isn't very large, but there is room to shrink transistors and therefore increase their density on chips. Moore's law isn't going to fall just yet.
Compare this to Li-ion batteries: the fundamental unit is a single lithium atom. It's hard to get smaller than a single atom. How then can batteries be improved? To answer that, we need to know a little about how batteries store energy.
In a Li-ion battery, there are two electrodes, the anode and the cathode, separated by an electrolyte and connected in a circuit. Lithium ions start at one electrode, they flow through an electrolyte to the opposite electrode. Electrons can't move through the electrolyte that separates the electrodes, but instead move through a wire that is hooked up to whatever gadget you want powered and eventually end up at the opposite electrode. The equation that governs the energy density of a battery is essentially
Energy density = Electrical charge density x Voltage,
where voltage is the voltage difference between the anode and the cathode and the electrical charge density corresponds to the number of electrons in a certain volume.
The electrical charge corresponds to the frequently cited milliampere-hour (mAh) spec of a battery- a (more or less) steady current for over some amount of time. But a mAh is not a unit of energy, for that, you need to multiply by the voltage of the battery to arrive at units of milliwatt-hours (a mWh is like a kWh from the electric company but much smaller). The electron moving from low to high voltage releases energy and moving from high to low voltage stores energy. You can take the energy on a per volume or per weight basis to get energy density. If you want to increase the energy density of a battery, you need to increase the voltage or increase the density of electrons.
In the case of Li-ion batteries, there is one electron per lithium ion, so an obvious way to increase the energy density of a battery is to increase the density of lithium atoms in the electrodes.
The anode and the cathode are made of materials that can hold a lot of lithium atoms. One common anode material is graphite. Graphite is stable in the absence of lithium, but you can stuff a decent amount of lithium atoms between the graphitic layers. However, a good amount of the weight and volume of the anode is due to carbon, which limits the density. What could hold more lithium atoms? Pure lithium metal. Not only does lithium metal have the highest density of lithium atoms possible, but it also increases the voltage difference between the anode and the cathode compared to graphite. However, there is a certain danger for having lithium or even graphite anodes.
Batteries have been known from time to time to burst into flame. It doesn't take much imagination to see why this sort of failure could be disastrous on an airplane. The problem is when the anode is close in voltage to pure lithium, occasionally, instead of filtering into the anode material, lithium ions form lithium metal dendrites. These pointy growths can form across the electrolyte to the anode and create a short in the battery, which, as we know, can have explosive consequences. So instead of decreasing the anode voltage, you could try to increase the cathode voltage.
Now here is some potential progress to be made. Just double the voltage and you double the energy density. If only it were so easy. The Achilles heel for this approach is that electrolytes that we use now just can't handle much more than a 5 V potential difference. The molecules in the electrolyte literally begin to come apart. So it starts to become clear that researchers have quite a difficult set of constraints to deal with while coming up with new battery materials.
Will batteries ever be as energy dense as fossil fuels? Probably not, but it isn't a great comparison because of the energy that is wasted when you burn fossil fuels.
Currently, gasoline is about 30 to 100 times more energy dense than a battery, depending on how you measure it, so even there is a breakthrough that eventually makes it to market that increases energy density by a factor of 10, it still won't have the energy density of plain, old gasoline.
Gasoline is so full of energy, that if you could extract it all and used it to power your cell phone, according to Exxon-Mobile, it would last for 20 years. This doesn't take into account the efficiency of an engine burning the fuel to extract the energy, so this figure is a exaggerated.* While carbon based fuels store a lot of energy in their carbon-carbon and carbon-hydrogen bonds, the need to burn those fuels wastes a large portion of that energy. However, even considering the wasted energy of an engine isn't enough to make batteries on par with fossil fuels... yet.
And gas outperforms batteries in another area: refilling/recharging.
The charging rate of batteries is a limitation that is especially felt in electric cars. Electric cars don't have a large range and take hours to charge to full capacity. Long road trips powered purely by batteries aren't quite feasible yet. There have been new standards agreed upon to bring down charge rate to 15-20 minutes. This still is longer than the few minutes it takes to fill a gas tank, and batteries have to be filled more often than a gas tank because of the shorter range they have.
Another option is to just swap the battery out with a fully charged one. The process can take as little as five minutes. And there is plenty of ongoing research into ultra-fast battery charging. So don't despair, there is some hope for the bettering of batteries.
Battery powered cars are an emerging market, but batteries might not be the go to storage medium for the long term. In his interview with the Verge, Ford CEO Alan Mulally talked about their technology road map in which batteries might just be a step towards hydrogen fueled cars later on. However, the hurdles that hydrogen power faces are too numerous to list here.
Another possibility is the use syngas, which could be burned in an internal combustion engine and is manufactured using sequestered CO2 such that the process has a small or no carbon footprint.
As for batteries, there is a lot of work being done on Li-air batteries. These cells use lithium metal for the anode material. As mentioned before, lithium metal is as good as lithium density gets. And instead of a traditional, self contained cathode, these batteries do a bit of outsourcing by using external oxygen rather than an internal material so the volume of the cell is decreased and energy density goes up.
There are a lot of constraints working against attempts to improve humble battery and there are limitations as to how good it can get, but don't you fret, the battery will keep improving, slowly but surely. Just don't plan a cross country trip in your electric car any time soon.
*As far as I know. I crunched some of the numbers when I first saw this claim and arrived at about the same figure of 1 gallon lasting 20 years, but without including any energy loss of an engine.
This was originally posted here.