Monday, August 30, 2010

A Brief History of Batteries- Part 1

When I was around 13 years old I remember learning about primary batteries in school.   This was about the time when I had also encounter lead-acid batteries and had realized that the car starter battery is being charged periodically.  I began to wonder why some batteries could not be recharged while other could.  Thus began my fascination with batteries. 

The above story is probably not true.  But it’s etched in my memory probably because it provides me with a romanticized notion of why I work on batteries today.  In reality, I was probably contemplating something more useful like “what is the point of learning math?”.   

Nonetheless, the question posed is an interesting one in that if we want to understand how to make a rechargeable battery that lasts a long time over many cycles, we first need to understand why some batteries do not recharge.  

The development of batteries is deeply rooted in this understanding (atleast one hopes it is and that it isn’t a series of accidents!).  Many of the ideas that people are proposing today for better batteries are in some ways trying to beat the fundamental limitations that prevent recharging of these primary batteries. 

If you don’t want to make the mistakes of the past, you better understand the past.  This post is an attempt to do just that. 

The topic I’m discussing is technical.  I’ve tried my best to make it accessible.  In doing so, I lose some technical accuracy; but this is the price of making it understandable.    

This blog post is written in the style of a novel.  The story of rechargeability can be written in a very linear fashion which steps through time to illustrate the rules one after another.  But this would make it easy to understand. Where is the fun in that!

So, I’ve borrowed a page from Quentin Tarantino’s movie Pulp Fiction and scrambled the narrative.  Each storyline is distinct, and they are all interrelated.  If the stories were arranged chronologically as the systems developed we would go Chapter 3, 2, 4, 1, Prologue, Epilogue.  I personally think the scrambled version reads better. 

This week I will start with the Prologue and Chapters 1 and 2.  Next week I will finish with Chapters 3 and 4 and the Epilogue. 

Fortunately, these stories are easier to follow when compared to Pulp Fiction.

And there is no swearing or violence. 

Prologue:  Do I smell sulfur in the air?  

There has been a lot of interest over the last few years on two systems that have been around for years.  These batteries are sold as going “beyond lithium-ion”.   These systems use lithium metal as the negative electrode and air (rather, the oxygen in the air) or sulfur as the positive electrode.  Some consider these systems to the Holy Grail of battery research. 

The promise that these battery chemistries hold is enormous.  In the case of lithium-air, the gravimetric energy is an order of magnitude higher than today’s batteries.  Think cars that can drive 500+ miles, cell phones that actually last all day, laptops that can last the complete transatlantic flight...  The volumetric energy is not that great, but hey, let us focus on the positives, shall we.

But there is a hitch (there always is).

In the case of lithium-air, it turns out that recharging lithium metal is a problem.  And the oxygen in the air reacts to form a compound that is basically pretty insoluble. 

In the case of sulfur, the lithium metal continues to be a problem.  And the sulfur electrode undergoes a series of reactions giving products that are either highly soluble, or insoluble.  All the products are a real pain to deal with and history shows us that dealing with them may not be for the faint of heart. 

If we are to succeed in getting these batteries to work, we first need to understand why similar concepts have failed in the past. 

Chapter 1:  A business plan that can’t fail- Take cheese; make milk.
You may remember my previous blog post where I alluded to the lithium thionyl chloride battery.  

This chemistry is used for missile applications.  In this battery lithium dissolves from Li metal on the negative electrode and the resulting lithium ions combine with thionly chloride (SOCl2) to form lithium chloride (LiCl) in the positive electrode. 

This is the process that occurs on discharge and it works pretty well.  If you want to recharge this battery, you will need to take the lithium chloride and convert is into thionyl chloride.   This is where the problem occurs. 

Lithium chloride is an insoluble solid.  Once you form it, it pretty much sits around and starts clogging up the whole cell.  Converting lithium chloride back to thionyl chloride is like trying to find a Quentin Tarantino movie with no violence- its not hard to find; its impossible to find!

Think of this as taking a pot full of hot milk and adding lemon juice to it.  The milk curdles and you get cheese.  There is pretty much no way to get back to milk and lemon juice from cheese (as far as I know).  Same story with lithium chloride. 

If, on the other hand, the lithium chloride were soluble in the electrolyte and dissociated in the solution, one can think of trying to recharge the system.  However, because lithium chloride is not soluble, there is no way to recharge the battery. 

Moral of the story:  If the product of your reaction leads to an insoluble product, in some batteries, recharging can be a problem.

The caveat “in some batteries”, is kind of important, but that is not for this post.    

Something similar happens in the lithium carbon monofluoride system (Li-CFx).  In this system, much like the thionly chloride system above, the lithium dissolves and reacts with the carbon monofluoride to form lithium fluoride and carbon; both solids.  Trying to go back to the starting chemicals by reacting the lithium fluoride with the carbon is pretty much impossible.  It’s the cheese to milk problem. 

So solubility is something we ought to be paying attention to when it comes to recharging. 

But what if the product of the reaction is highly soluble (think salt in water)?  This must be the best thing possible for recharging, would it not? 

Turns out that this is not true either.  This would be the story of the zinc electrode.  As in the zinc-manganese oxide alkaline battery (the Energizer-bunny battery).

Chapter 2:  If only the Dead Sea weren’t as salty. 

The alkaline battery is actually a rechargeable battery.  It’s just that it is not a very good one.  Some of you may remember that there were rechargeable alkaline batteries in the market that promised a few cycles. These all appear to be gone from the market.  I guess no one wants to carry around a charger. 

The positive electrode (MnO2) in these batteries can recharge thousands of times if you are careful to avoid certain phases.  I personally believe that this chemistry is worth revisiting. 

It’s the zinc electrode that is the problem.  The reaction involves zinc electrochemically reacting and forming a complex called zincate.  As the concentration of zincate increases, it precipitates out as zinc oxide.   Think of this as adding salt into water.  You can only add so much salt before the solution gets saturated and any more salt you add stays as crystals. 

This is the simple version. 

The whole reaction scheme is actually pretty complicated.  Turns out that five other reactions can happen in this battery depending on various conditions.  Some years ago, I was involved in modeling of the zinc-MnO2 battery and it is by far the most complicated battery chemistry there is.  Lithium batteries are so much easier to understand with compared to this system. 

The zinc oxide has pretty high solubility in the electrolyte (potassium hydroxide). What this means is that zinc oxide easily goes into solution to form zincate in the electrolyte.  One would think this is great for recharging and it is.  But it turns out that when you try to deposit the zinc from the zincate, it does not deposit in the same place where it dissolved. 

This leads to the structure of the zinc electrode changing continuously as you charge and discharge the battery repeatedly.  This is referred to as shape change.  Shape change leads to the zinc depositing as a solid mass in the one part of the electrode.  And this solid mass is hard to react.  Slowly, as the solid mass gets more pronounced, the zinc electrode fades. 

Reduce the solubility of zinc oxide and you can decrease the shape change and increase the cycle life. There is some wonderful data that shows that. This is like somehow making salt less soluble in water.  Turns out that in reality this approach also kills the power of the battery and so this is not a real solution. 

The reason why the shape change happens is a bit complicated, and probably not relevant at this stage.   What does matter is that, in some batteries, very high solubility can be a bad thing and can decrease the ability to recharge. 

Moral of the story:  High solubility may not necessarily be a good thing for recharging.

So we have a problem.  Neither no solubility nor high solubility appears to be a good idea.  So how do we make a rechargeable battery?

We shall look at that in the next chapter.  Next week. 

Now… isn’t that a nice tease!  Stay tuned. 


Monday, August 16, 2010

One battery chemistry to rule them all?

Darwinism is alive and kicking in the battery world (despite the threat the theory faces in other spheres).

Let me elaborate.

The Bible (or Koran, or Gita or… take your pick) for most battery-related topics is the book “Handbook of Batteries” by David Linden. It’s page after page of different chemistries that allow us to store charge. Some of these are primary batteries (not rechargeable) and others are secondary (rechargeable). Some operate at high voltages, others at low. Some are water-based, others are non-aqueous. Some operate at room temperature, others at higher (much higher) temperatures.

What is not listed in the book are a bunch of chemistries that can still be made into a battery but are not worth listing because no one is interested in using them.

One can (literally) go to the periodic table, pick up two elements from it, and make a battery from these elements. In all probability the resulting battery will not be very good, but you can certainly make something (think lemon battery).

Of these numerous combinations of anodes, cathodes, and electrolytes that make a battery, natural selection kicks in and we identify the ones that give us favorable performance. From this Darwinian process we end up with the batteries that we all know and love (although love may be too strong a word to describe our relationship with our energy storage device!).

Volta and Daniell could be playing with the Zn electrode, but it took putting it with a MnO2 electrode to make it a commercial success that endures to this day. Similarly, we could play all we want with a Li metal anode but it took pairing an oxide cathode with a graphite anode to make, what is come to be called, a lithium-ion battery for this concept to be a successful rechargeable battery.

Along the way, we have relegated a bunch of batteries to niche applications. No one has heard about a lithium-thionyl chloride battery because it’s used in obscure military applications involving missiles, nuclear bombs, remote launching, and a button in the White House (or so I’m told. Who knows). Similarly no one knows about the nickel-hydrogen battery, despite it being the chemistry that has the highest cycle/calendar life of all batteries we know of (the catch: it costs 100 times that of lithium-ion batteries. And you thought lithium-ion was expensive!). But it sits quietly in space powering our satellites day in and day out.

But natural selection does not happen overnight. There is a period of instability where it is not clear if a particular species (or battery chemistry) will make it or if it will left to the ash heap of history (being dodo-ed, so to speak). And similar to evolution, in battery chemistries, there is also such a thing as relegation to a niche (I will not give any examples from the animal kingdom because of the risk of offending someone). To survive you don’t have to be the fittest, you only need to be fit enough to have a market share, however small the share may be!

But let us stop this history lesson and get back to the modern day where I believe we are in the cusp of one (or two) such evolutionary change(s).

One is in the area of lithium batteries for vehicles.

The class of batteries that we have come to call lithium batteries is unlike the previous battery chemistries that we have encountered in that it does not represent a single anode/cathode/electrolyte combination.

When someone refers to a lead acid battery, they are talking about a battery that has a lead dioxide positive electrode and a lead negative electrode with sulfuric acid electrolyte. You may be using some carbon in the plate, or using a different current collector grid, but fundamentally you have not changed the voltage or capacity of this battery. In the same way a Ni-Cd or Ni-MH battery has a clear meaning in defining the anode/cathode/electrolyte.

When someone says they are using a lithium-ion battery that only tells you that the class of battery where the charge is carried by a lithium ion. The anode, cathode and (to a lesser extent) electrolyte are not specified. It could be a graphite/cobalt oxide battery that your cell phone uses or a graphite/lithium manganese oxide battery that your powertool uses. Both are called lithium-ion batteries.

In what I consider to be a public-relation disaster, the lithium battery community decided to club all these chemistries into one term, lithium-ion. If, instead of calling the graphite/cobalt oxide battery a lithium-ion battery, they had called it a, say, carbon-cobalt battery, today we would be talking about a carbon-manganese battery, a carbon-iron battery, a carbon-nickel battery, a titanate-nickel battery and so on and so forth.

If we had the foresight to do that (I use “we” to refer to the royal pronoun. I should not blame myself for this because I was but only a teenager when this happened) then it would seem like batteries are changing all the time. Instead, we get asked uncomfortable questions like “Lithium batteries have been around for 20 years. So… are you guys doing anything new?”. Not that I’m bitter or anything.

But let’s get back to the topic at hand.

Its becoming increasing clear that lithium-ion will be the choice for PHEVs in the near-term. What is not clear is which of these numerous chemistries will ultimately be the winner.

Would the low-cost, safe manganese system (the chemistry picked by LG Chem/CPI and Dow Kokam) win over the longer-life (presumably, although I have not seen such data) safe iron phosphate system (A123)? Would the higher energy nickelate system (Johnson Control and Saft) be the choice because of its higher energy despite its safety issues? Would the fast-charge capability of the titanate anode (Enerdel and Altair nano) be of any use in a world where even slow charging could be an infrastructure nightmare?

Each chemistry has its pros and cons. There are four criteria one looks for in a battery: Cost, life, performance, and safety. No one chemistry is the magic bullet that satisfies all these criteria. Each choice leads to a compromise. As of today, it appears hard to predict the winner.

We the consumer may be the ultimate deciding authority.

Will we choose a car with more trunk space and decide that we need to use the highest energy density battery we can get our hands on even if its only lasts 4-5 years or would we prefer something that may be a bit cheaper but that may be a small car that is only useful for a short commute?
Or maybe we will end up saying that gasoline is fine and that all this battery technology stuff is hype with no substance (now, that’s a blog post that is well worth making!).

We are at a time when we are unable to clearly see which one (or two or three…or maybe even none!) of these will be our future. GM and Nissan have chosen the manganese oxide system for the Volt and the Leaf, respectively, but this does not mean the race is finished. We live in interesting times, indeed.

Something similar is happening in the emerging market for grid-level electricity storage. With applications ranging from frequency regulation to storage of renewable applications, the time of discharge can range from less than a second all the ways to hours to days. It is clear that one battery will not serve as the solution for all these applications.

There are already many systems that are in various stages of deployment for grid applications. Lead-acid, sodium-sulfur, Ni-Cd, and Ni-MH are being deployed along with lithium-ion. And flow batteries are trying to make inroads. And similar to lithium batteries, flow batteries are a class and can represent anything from vanadium-based, halogen-based, or iron-based (I wonder if we are repeating our mistake with the lithium-ion by calling something a flow battery as opposed to, say, a hydrogen-halogen battery). Some of these systems have been around for a decades, while others are being developed as we speak.

And it’s not just batteries that compete in this application. Flywheels, capacitors, compressed air, and superconductors are jostling for space. Over the next decade, systems will be narrowed down and down-selected to ones that make sense. Something similar happened in the 90’s and in the early 2000’s in the vehicle space and technologies like electrochemical capacitors (or supercaps or pseudocaps) were slowly deemphasized. This is bound to happen in the grid storage space.

50 years from now, when we open the Handbook of Batteries (in whatever form we may be reading at that time), all these chemistries will be in the book. Question is, which chemistries will have the biggest chapters.


Tuesday, August 10, 2010

A shout-out to the separator

Anybody who has paid attention to batteries (especially, lithium batteries) and/or read this blog knows that in most batteries the anode and cathode materials are the main players for holding charge. There is a lot of research in trying to find new materials that hold more charge at high voltages. But as I have pointed out in my previous posts, we need a few other materials to ensure that we tap into this charge.

Things like current collectors, separators, and the electrolyte.

All these play as important a role as the electrode materials. In some cases, they are actually more important. I have decided to spend sometime giving them the credit they deserve.

I will start with a shout-out to the separator.

Most electrochemical systems (and, yes, batteries fall in the class of electrochemical systems) require some way to separate the anode and the cathode. One tries to keep these electrodes in very close proximity to decreases resistances for ions to travel between them while preventing shorts. An ideal way to achieve this is via the use of a separator. It’s a (arguably) simple physical barrier between the two electrodes that lets ions go through, but not electrons.

However, in some cases, the separator serves a larger purpose. For example, it also ensures that the anode and cathode reactants/products don’t mix. If you are trying to electrolyze water to make hydrogen and oxygen, it helps to not have them mix together (trust me). Separators help ensure that.

Be it a fuel cell, a flow battery or a containerized battery (like a lithium battery), a lot of effort is spent on the separator to make sure that it does its job. For the flow battery that we are planning to work on with a recent ARPA-E award, the separator will be an integral part of our developmental effort.

In the battery space, the separator has always had its part to play. In a lead-acid battery the absoptive glass mat (AGM) separator helps increase the life of the battery. In batteries that use zinc or lithium metal, the separator may help prevent dendritic shorts by retarding the growth of the dendrite. And in the Ni-MH battery it can help decrease the rate of self discharge.

Which bring me to the first news item that caught my eye.

After coming up with the magical iPhone and the magical iPad, Apple has recently unveiled the Magic Trackpad. Interestingly, Apple also announced that they were selling rechargeable batteries with a (magical?) charger for the trackpad (which operates on Bluetooth). Trust Apple to make a Ni-MH battery with a charger sound cool. Will the magic never stop?

In general, the Ni-MH battery is a terrible battery for Bluetooth applications. This battery has notoriously high self-discharge. A typical Ni-MH battery can discharge by as much as 20% of its capacity in 2 weeks in the SF Bay Area and 50% in balmy India, in summer. Bluetooth devices are used sparingly (hopefully your job does not require you to type 24 h a day), so the self-discharge can be a killer.

Apple, on the other hand, is promising 20% capacity loss in 1 year. Magical you think?

Not really. The answer, my friend, is (partly) a separator blowing in the wind.

There are three reasons why Ni-MH batteries self discharge. The first is oxygen evolution on the nickel electrode, the second is hydrogen evolution on the metal hydride electrode, and the third is a nitrate redox shuttle across the two electrodes. All three are forms of internal leaks that discharge the battery.

Mother Nature dictates the first two. It can be hard to beat Mother Nature, especially for mere mortals like me (and, yes… even Steve Jobs), but there are some things we can do to decrease the rate of these gas evolution reactions.

The third mechanism is what interests me in this post and it involves using a separator that traps the nitrates and prevents the ion from shuttling across. This prevents the battery from slowly leaking and keeps the battery charged. Very simple, yet very effective.

These developments in this mature chemistry are only 5 years old and have resulted in a significant decrease in the rate of self-discharge. Which takes me back to my post on how we tend to ignore older chemistries (read non-lithium) in most, if not all, R&D projects in this country.

So here is a shout-out to the humble separator.

Separators for lithium-ion batteries are more crucial in that they can be the difference between an iPhone that is plagued by dropped calls because of antenna issues and one that is burning your pant pocket.

Separators have a checkered history when it comes to lithium batteries. Remember that the volume occupied by this layer is excess space that is wasted. There have been many moves to try to decrease the thickness of the separator, but attempts at making this layer less than 20 microns result in the electrode shorting during a winding process that is part of battery assembly. Shorting a battery is typically not a good idea! Most separators today are 20-25 microns in thickness.

Moreover, these separators have what is called a “shut down” layer. This layer, made of a polymer that melts and shuts the pores if the temperature increases too much, is a mechanism by which reactions are stopped if a battery goes into thermal runaway (or “spontaneous disassembly”, as the industry calls it). However, there are some that say that when this melting occurs, the structural integrity of the separator decreases and the electrodes end up shorting with each other. Statements regarding shorting being a bad idea apply. This issue is still being played out.

As an aside, a couple of researchers at LBNL are doing something interesting with the separator. Tom Richardson and Guoying Chen incorporated a conducting polymer that prevents the battery from going to overcharge. Overcharge causes the thermal runaway in lithium batteries. The idea is to prevent the overcharge and hence make the battery safer.

But lets get back to the separator we use today.

You may remember that YouTube video’s of burning laptops. You may also remember that the cause was attributed to metal particles falling into the battery during assembly and leading to shorting. A way of dealing with this issue is to make a stronger separator; one that will prevent shorting even if particles fall into the battery. Some manufacturers are testing ceramic coatings on the (presently-used) polymer separators to see if this will increase the puncture resistance. This issue is also still being played out.

Obviously all these problems will go away if the electrodes were not kept so close to each other by using a thicker separator. But this decreases the energy density of the battery and decreases the power. Obviously, no one wants that!

Other than being crucial from a safety perspective, separators are also one of the culprits in making lithium batteries expensive.

Battery costs are impossible to find with any clarity (The US military can learn from battery companies on how to keep secrets and prevent incidents like the one with Wikileaks). But, estimates suggest that material costs can range from 50-80% of battery costs. And 25% of the material cost is the cost of the separator!

Think about this. This simple polymer layer, very similar to the polymer used to make grocery bags, can be as much as 20% of the cost of the battery! At the sake of repeating myself, batteries are expensive and we have to decrease the cost significantly to get any widespread penetration of EVs and PHEVs.

Part of the reason why separators are expensive is because of a process that creates the pores. And it appears that the market for separators does not have enough competition to drive down costs.

Which brings me to second news item.

Dupont just announced that they would be getting into the battery separator game by manufacturing their line of lithium battery separators. Information is scarce, but they appear to be using a different process than what their competitors use and promise higher power, and a higher operating temperature. No word on cost, but now there is one more player in this game bringing some competition. That can only be good.

Now if someone can come up with a way to make a really strong separator that is say, 5 microns thick, has a open path for ions, can withstand the winding process, does not puncture even when there are metal particles in the battery, and costs less than $1/square meter, then we should be all set.

For the uninitiated, the paragraph above is like hoping that Microsoft comes up with an operating system that does not crash all the time. It seems doable, but for some reason it never seems to happen!

Until then, let’s thank the separator that we do have.


Saturday, August 7, 2010

TWiB met TWiT and TWiB got 500 hits

This blog post has nothing to do with anything that happened in the battery world. But it happened this week, and it happened to me, and it’s my blog so I will make this post.

Regular readers (all 3 of you) of this blog know that the name of this blog is inspired by the popular podcast This Week in Tech (TWiT) run by Leo Laporte. You can see more in my first post.

Without getting into details I got to go to the TWiT cottage (the place where they record the show) to watch a recording of episode 259 this Sunday afternoon. The name of the show is “Next Stop: Gilbert, Arizona”. You can listen to it here.

The whole process of making the episode was sort of fascinating. The TWiT cottage is in Petaluma (i.e., middle-of-nowhere North San Francisco Bay). It is basically a 4 room cottage that has been converted to a recording studio. There were 3 “watchers”, one guest on the show and Leo (there were 3 off-site guests on Skype). All 5 of us were jammed into a small room full of cameras, and recording mics. Leo was twiddling on knobs, turning things up and down, typing on his computer, AND talking all at the same time. Amazing!

Needless of say, everything said in the room is recorded and is streamed live.

The inhouse guest was Brian Brushwood (pictured below), who is a magician and fire eater (no, really!). The other guests were John C. Dvorak (who is a regular on the show and is a tech guru), Baratunde Thurston (the web editor of The Onion. You know The Onion, don’t you?), and Felicia Day (who is a star on the web series, The Guild).

When I entered the room, they were setting up for the show. I was introduced and immediately Leo and Brian started asking me questions on batteries: is there enough lithium?, is there a moore’s law of batteries?, what is happening to battery development? When the off-site guests heard who I was, they started with “should I discharge my lithium batteries all the way?”. Apparently, they don’t read my blog (sigh). Anyway, it was the perfect opportunity to plug TWiB.

Remember that all this is live. There are people out there who are listening to all this. They (apparently) heard me plug TWiB. Anyway, I was checking the stats on my blog afterwards and at 3 PM (around the time I was plugging TWiB), 500+ people logged into the site! Ah… the power of the plug.

At the end of the show, Leo volunteered to take pictures with the watchers. Here is mine with Leo. Brian is in the back doing something funny. The idea is apparently called Photo-bombing where someone in the crowd decides to ruin the great picture of your kids by doing something strange in the background. He was imitating that scenario.

Anyway, you can sum this experience up in one sentence: TWiB met TWiT and TWiB got 500 hits.