Nano for batteries: the challenge of volumetric density

In my previous post I discussed some of the pros and cons of using our skills making materials in various nanoarchitectures in Li-ion battery technology, mainly related to the increase in surface area and shortening of diffusion paths. Today I will concentrate in other aspects: volume and density.

Most probably, you have heard how silicon-based electrodes can boost the storage of Li-ion batteries. And you may have even noticed that a lot of the releases announcing breakthroughs in this area have the word nano in them (here's a very recent example). Indeed, one of the most serious problems of silicon electrodes is that they expand and contract enormously upon cycling because they can uptake so much lithium (hence the boost in storage capacity). In brief, what happens with big particles is that they crack into much smaller ones due to these expansion/contraction cycles (think of freeze/thaw cycles in the winter) and lose contact with the electrode additives and the electrolyte, cripling the electrode life. So starting small (i.e., nano) bypasses this issue. With less volume in a particle, it is less likely to break in several domains.

Usually, electrodes composed of nanoparticles of silicon cycle better than those composed of bigger particles... provided that we are good at building the electrode structure so that all these new particles are well connected to the current collectors (and, subsequently, the battery leads) through the electrode additives. However, issues arise when using this strategy. Guess what one of the problems is? Yes, our friends the side reactions; silicon reacts outside the voltage of stability of the electrolyte. A second problem, related to the increase in surface, is that silicon nanoparticles are much more reactive with air than micron-size particles, bringing up the need for additional safety controls during handling. In fact, some people have even proposed primary silicon/air batteries.

It is quite likely that silicon will have to be used in nanometric form to make it a viable electrode with very high capacity. But be careful with overstating the numbers. A typical strategy to keep a good electrical contact is to use more conductive additives (carbon, basically) because there are more particles to connect to each other and the current collector. Increasing the amount of polymeric binder is another strategy that is commonly used to keep the particles together. And more carbon and/or binder means less silicon, which means that we are reducing the total capacity of the electrode. Yes, in terms of the silicon only, the capacity is unchanged, but we have to count everything when building the battery! Pay attention to what companies that announce silicon-based batteries show as capacity gains and you'll see they are more modest than would be expected theoretically.

In addition, nanoparticles have a very annoying tendency to form aggregates that is very difficult to control. These aggregates form rather disorderly and, therefore, leave a lot of dead space within, which can be helpful for electrolyte wetting but also be unnecessarily high. Finally, a lot of the strategies to alleviate volume expansions in silicon electrodes rely on placing the particles far from each other, so that they don't crush against each other upon expansion. Obviously, this also helps with the aggregation, but now we have even more inactive space between particles!

The result of all these approaches is always the same: the density (mass per unit of volume) of the electrode is lowered considerably with respect to an electrode made using more traditional methods with bigger particles. In general, it is very difficult to achieve bulk packing densities of materials using nanostructured electrodes, which may, after all, be something we have to live with, in some instances, if we want them to work.

Unfortunately, lower packing density also means lower total volumetric energy density of the electrode/battery. And volumetric energy density is no laughing matter when thinking of batteries for electric vehicles. There is very limited volume available for the battery (especially in a hybrid). Using certain nanostructures as electrodes can lead to very long life, but very modest volumetric densities, so that you still need a bigger battery to power your car. When volume is factored in, the gains of using nanoparticles are systematically much more modest or even totally offset. This is a problem that is still unsolved. Scientists are getting better at synthesizing nanoparticles and we are starting to be able to control aggregation and assembly, so there are possibilities that are being explored. Now, the methods associated may imply an increase in cost of manufacturing. But I'll leave this for another day.

More reading materials for those interested in the science of silicon electrodes:

Larcher et al., Recent findings and prospects in the field of pure metals as negative electrodes for Li-ion batteries, Journal of Materials Chemistry 17, 3759 - 3772 (2007)

What can nano do for batteries?

First and foremost, I must express my gratitude to Venkat for his introduction and for encouraging me to participate in this new blog-adventure. Now that he cannot hear us, I will tell you that I learned quite a few things myself with his posts. It is a pleasure for me to offer a different perspective to his and I hope you will find it informative. Since he has already introduced me, I’ll move on to the fun stuff.

Nanotechnology is certainly one of the fields of research that has witnessed greatest progress in the last decade (although the concept itself is not all that new; Michael Faraday can be considered one of the first nanoscientists!). When it comes to functional materials, among other things, nanoscience offers the promise of enhanced catalytic properties, controllable band gaps that can have an impact on the efficiency of a solar cell and novel medical diagnostics and treatment tools (not to mention that most of suntan lotions contain nanoparticles nowadays!).

It is easy to wonder whether nanotechnology would also be a helpful tool toward better batteries. There are a few arguments in favor of using nanoparticles in a battery, particularly as part of the electrodes. We will just concentrate on a couple of them for now (and leave the rest for another day if I am still considered a guest after today!). Although the discussion relies on the Li-ion technology for examples, the ideas are pretty much applicable to many battery technologies.

One of the first incentives one can think of is the shorter distances for ions and electrons to travel through. Broadly speaking, think of it as you having to cross your living room as opposed to a football field. The second one has to with the surface area. Let’s imagine that the square in the picture below is made of 1 um edges:

divide each edge ten times and you now have 100 squares with edges of 100 nm:

All those new lines you see are new surfaces (if you want to entertain yourself, expand the exercise to a 3D cube and reduce it to particles of 10 nm). In a battery, all of this surface could ideally be exposed to the electrolyte, which is the media through which ions transfer from electrode to electrode (the key to battery operation).

Now we have (lithium) ions that can access the active material much more extensively (more surface) and have shorter distances to travel through the solid (remember the living room). Diffusion in the solid is typically, but not always, slower than in the liquid, so, in theory, the result is a lower resistance to ion transport and, therefore, better utilization of the material (i.e., higher stored charge/energy) and, potentially, higher rates of charge and discharge (i.e., shorter charging times and higher power).

There are many reports available that show that samples composed of nanoparticles, all things equal, can lead to better performance than (what we call) bulk counterparts. In materials with poor ion conduction, this effect has had an important impact. The now world famous lithium iron phosphate owes some of its glamour to the chemists that were able to nanostructure it. The same happens with an anode material that is getting a lot of exposure as of late: lithium titanate.

So in a sense, nanotechnology has an important role in the field of (Li-ion) batteries, and, in fact, there are many R&D projects that concentrate on exploiting the advantages of nanotechnology to developing advanced materials, including some in our very own Batteries for Advanced Transportation Technologies program (go through our recent scientific reports to see what's up) at the Department of Energy. So, it is settled! Nanotechnology is the way to go to make batteries with five times more energy density!

Hmmm... not so quick... Apart from the fact that geometrically increasing the energy density is much trickier than it seems (we’ll also leave this for another day), there is a key subtlety in the case of lithium iron phosphate and lithium titanate that makes the whole trick work: they are both active within the voltage window of stability of the components of the liquid electrolyte. Venkat has already told you that bad things happen when we fall outside this window. Unfortunately, one of the ways of increasing the energy density is precisely by increasing the voltage (remember, energy = voltage x charge). Graphite electrodes, for instance, ubiquitous as they may be, react outside that window.

If we start using materials that are only slightly outside the thermodynamic window of the electrolyte, the undesired reactions may happen at a rate that is slow enough for us to live with. But, hey, remember that increased surface area of nanoparticles? Yeah, it is increased for every component of the electrolyte, not just the ions. The immediate consequence of reducing the particle size of our electrode materials is that side reactions tend to be exacerbated. And those side reactions produce insoluble products that deposit, just our luck, precisely on the surface of the particles, covering them and producing layers that are resistive to the diffusion of ions. Automatically, the advantage of using nanoparticles is lost. And things can be so bad that we may be better off using slightly less active electrodes!

In conclusion, you can choose to have a lithium iron phosphate/lithium titanate battery that operates extremely well and has a long life thanks to the use of nanoparticles... but what if I told that such battery only has a voltage output of around 2 V (yeah, those lead-acid batteries don't look so bad anymore)? And if I mentioned that, in addition, the lithium titanate has about half the storage capacity of graphite, so the energy density of this great battery is lower than that in your cellphone Li-ion cells? If we are to increase the energy density of current batteries to make them more application friendly, we have to come up with inventive ways of using materials operating at high and low voltages; preferably, not in the form of nanoparticles… unless we find a way of stopping those annoying side reactions without killing the ion transport at the same time.

There are other issues that make nano only a partial answer (sorry, no Moore's law for batteries) to the performance barriers of batteries, among which are higher associated manufacturing or processing costs and lower packing densities. But this is for another post (or two!). If you are left wanting more, you can get even more knitty-gritty details by reading some scientific literature (if you have access to it, of course), which will also offer additional shades of grey to the arguments developed here:

Aricò et al., Nanostructured materials for advanced energy conversion and storage devices, Nature Materials 4, 366 - 377 (2005)

Bruce et al., Nanomaterials for Rechargeable Lithium Batteries, Angewandte Chemie International Edition 47, 2930 – 2946 (2008)