Jack Rickard's Lithium-Ion Battery Tech

We’ve been sifting through the video archives recently and stumbled upon one of Jack’s very special tech talks about batteries from 2013. It's prime Jack and it is relevant today.

It’s at 21:07 in the video. We think you're going to love it. Let's take a look.

I threatened to talk about batteries today. Last week. I like to talk about batteries.

Actually, I like batteries better than I like electric cars. Originally we did an electric car while I was playing around with batteries. The first thing we're gonna do today is take one apart and show it to you.

So cinch up your depends old guys. We're going in. Let me see if I can find a knife.

Ah, here next to my heart, heart, heart. This is an A123 cell. It is a lithium iron phosphate cell which is what I think everyone should use.

For automotive purposes. This one is ostensibly ruined and at zero volts. I'm gonna take this metal knife and cut into it and we'll know pretty quick, huh? And I'm just gonna scribe around here and see if I can pull off the, it's kind of a mica, what do you call it? Not mica, the mylar pouch.

And inside we have, this is a separator and some little green tapes holding it together. The A123 cells are known for relatively high power. And I'm gonna show you why.

I don't know if you'll be able to tell from another cell. This is a propylene separator. Just like a piece of very thin garbage bag.

It is electrically insulating so it keeps the foils from touching each other and shorting but it has, it's microporous and absorbs electrolyte as do the anode and the cathode. And this is kind of in here in a zigzag between these foils. Here is a foil and it is copper.

As you can see, and copper is our anode and this dull material on here is a very fine graphite powder. Graphite and carbon are an excellent conductor at least along one plane of their hexagon matrix, crystal matrix. And this is what our lithium ions intercalate into when we charge the cell.

And the purpose of the cell is to store energy. And we do that by taking electrons from the anode or the cathode and storing them on this anode. Let's see, what's our next sheet here? It's shiny.

And so this is actually, I'll scrape off a little bit so you can see, this is actually aluminum. And you can see on both sides of this, it's a shiny black. Now this side has a lot of copper on it and that's a sign of an over discharged cell.

And we'll talk more about that later. But when we over discharge the cell below a certain potential, the copper that the anode's on begins to come apart and it migrates throughout the battery and can form shunts and dendrites to shorten the cell, but it also deposits elsewhere. Here in this case, you can see it deposited on the cathode.

Well, every place this copper is deposited now no longer can intercalate lithium ions. The copper blocks that. And we've got quite a bit on one side, a little bit on the other.

Maybe there's some lithium plating on the other side. This is kind of shiny. Most of our cells aren't, but that is because lithium ion phosphate actually doesn't conduct at all.

And so in order to get it to conduct and get a good current flow through the cell, they mix about 5% carbon in with the lithium ion phosphate. In the case of A123, they actually make a nanopowder granule of lithium ion phosphate and then coat that with carbon and make a slurry and put it on this aluminum foil on both sides in a kind of like printing. They put a slurry on it and then they mechanically press it between rollers.

This is like in an argon atmosphere. And then they heat treat it and dry it out. The binder is kind of a polyvinylene material that is sort of like conductive glue.

And that's what binds it to that aluminum thing. So that's basically what we have, is a aluminum cathode with lithium ion phosphate on both sides. And in this case, a little bit of copper, which is why we're at zero volts.

And then our anode is copper with graphite, graphene, carbon black, hard carbon, or some mixture of all that, again, in a very fine powder that's mixed with the same polyvinylene binder and basically printed on each side of these very thin foils. Then the foils are gathered at the top and connected to these tabs. And they kind of have tabs on them.

These are torn off, but that alternate to where all your copper ones are connected to your negative terminal and all your aluminum are connected to the positive. This is your polyethylene separator, and it's in a mylar pouch. And that makes up a lithium ion phosphate battery.

For a larger prismatics, they're no different. They're not quite as shiny. They use a little bigger particles, usually from a lease, a company in Taiwan that have two sizes, again, mixed with carbon.

And the same kind of leaves are stuffed down into a plastic box. And then there's a heavier terminal that clamps really tabs just like this from the foils to the terminals that you screw into. And that is the physical construction and layout of a lithium ion phosphate cell.

Let's make this thing go away. Let's review the basics first. Got my little diagram over here.

You have a cathode, which I just showed you. It's made of aluminum, and it has layers of lithium ion phosphate in it. And we'll talk a little more about that.

The lithium is intercalated in that. And when we charge, we take an electron off this cathode and that lowers its potential to a more positive. It actually makes it more positive.

And we take that around the top with our charger and we go over and we pump that into our anode. A couple of things happen there. Let's talk about conduction and conductivity because there's actually a pretty broad oversimplification of this that's okay, but let's go a little deeper.

If we take a non-conductive item like plastic or mylar or ultra high molecular weight Teflon or something like that, those are very long chain polymers. And they have a number of elements bound up together and almost no free electrons. And so they cannot conduct electricity.

Metals, on the other hand, can. And the reason for that is that metals will have a number of electrons in the outer valence band somewhat loosely tied to the atom. As we may have said that depending on the level of the valence band will be a different number of electrons that are said to be full.

If you have more than half of those electrons, they tend to bind pretty tightly to the atom and they will readily accept electrons from other atoms. If you have less than half, then we refer to them as minus electrons where they will give them up quite easily. And all metals tend to have less than half the full outer valence band and will give up electrons.

Lithium is an extreme case of this. It has one electron that it gives up. We call it the party girl of batteries.

It gives it up very easily. When you take a metal, a pure metal, or even a metal alloy, what you do is take the element, copper or iron or lithium or whatever the metal is, and it forms a matrix, a crystal. And structure by sharing electrons.

In so doing, there's a number of them left over, extra electrons or we normally term that free electrons. They exist in the metal, but they're not connected to any of the atoms. They're simply a cloud of free electrons in that device.

I'd like you to think of that as a gas under pressure. Electrons, as we know, all have a negative charge and they repel each other. And so think of it as a gas in a container.

Your wire is actually a container and it's full of electrons, free electrons, and they're a gas and they push out against all edges of that wire to the skin because they're trying to get away from each other. They repel each other. And the number of electrons in there is a function of that pressure.

And so if we introduce an electron by force with a power supply in one end of that wire, we increase that pressure by one electron. And that's instantly felt through the whole wire. And so we say that electricity travels with the speed of light.

Okay, I'm not too sure. It's pretty much instantaneous. That pressure is exhibited through the whole length of the wire so quickly that we think it's close to 300 million meters a second.

And so that happens instantaneously. If you put it in one end of a 1,000 foot wire, you increase the pressure through the whole wire and that's felt at the far end instantly. If we put a power supply on the two ends and we put electrons in one end and take them out the other, we do have electron flow.

And we have current flow in one end and out the other, but along the way, all these electrons seek to repel each other. The actual electron flow, the movement of an electron through the wire, would you believe, and this varies depending on the temperature, the size of the conductor, and so forth, about two or two and a half inches a minute, not very fast. In actually moving through the wire, it collides with other things, impurities.

The crystalline structure is often imperfect and almost always is imperfect. Lots of fractures and so forth. It'll bump into an atom, dislodge one of the electrons that are there and take its place.

Well, but that frees up an electron. That is kind of the function of resistance. And that's what causes heat under current flow.

And higher current flow, higher heat is the actual physical motion of the electrons through the conductor. But the effects are a charge effect that's universally felt throughout the conductor instantaneously. If I put one electron in one end, I've increased the pressure at every point, including the other end, by exactly one electron.

And so when we hook up our charger here, and I say we pull an electron off here, they usually depict this as an electron traveling over here. Instantly, the charge on the whole anode, or the whole cathode, and the carbon that is mixed into the cathode material is instantly decreased. It becomes more positive.

And instantly, the anode, current collector, and the carbon that's connected to that becomes one electron more negative. Now the actual route of the electron through the wire could take seconds, minutes, days, I don't know how long is it and how big a round is it and so forth. But the charge is felt immediately.

And that's not very important, but I'd like you to keep it in mind. So we're gonna take electrons off of this end. Instantly, our cathode current collector and our cathode material becomes more positive.

And in becoming more positive, it, to some degree, repels a positive ion. And we have two to pick from, lithium and iron. And lithium, as I said, is the party girl.

So it exits this material. This diagram actually shows a little SEI layer on the cathode. That's true, but it's such a little bit of material and so inconsequential to battery operation, you can ignore it.

That lithium ion is gonna come out into our electrolyte. Now the electrolyte is the same thing. It's made up of a lithium hexafluorophosphate, which is a lithium salt, the short form is LiPF6.

And that is positive lithium ions throughout the electrolyte. And of course, much like the negative charges of the electrons, they seek to repel each other. And so they have a pressure in the electrolyte.

So when we say that lithium ions travel from the cathode to the anode, that's true, they do. But what really happens is it comes off the cathode here, increases the pressure everywhere, and our differential here at the anode goes up. The carbon on the anode is more negative, and the positive pressure from the electrolyte and the electrolytes necessary to conduct this goes up more positive.

And so that's what causes intercalation at the anode side. Let's drop that for a minute. And so when the lithium ion intercalates in this crystalline matrix on the cathode side, on the anode side, that pairs with the negative charge of the electrons.

That we put there and kind of neutralizes it to the point that we can store more electrons on that current collector because we keep absorbing that charge with our positive lithium ions. Or counteracting it to some degree. The lithium ion is actually not molecularly part of the carbon.

It's held between graphene sheets. Each carbon sheet is made up of a hexagonal or a six atom ring. And then the sheet beneath it is likewise.

The lithium ion is actually held between the sheets, exactly halfway between the sheets, and exactly centered in that six atom ring. Picture the metal ball levitation thing on YouTube where you've got two magnets and then this ball and they can hold it in position with the magnetic field. This is an electrostatic field, but same concept.

And it's held in that position between the layers of a carbon crystal sheet. Let's talk a little bit more about something we haven't covered very well because it's a little bit complex. And that is our cathode material.

I'm gonna put this up there. This shows here a grain, a single grain of lithium ion phosphate. Lithium, we know, is an element.

Iron is pretty common. Phosphate is a combination of phosphorus and oxygen. And in fact, iron phosphate is a very common dirt cheap fertilizer.

And so with a very small amount of lithium and some fertilizer, we can make a battery. It's also totally non-toxic in any respect. These are so much more benign than a lead acid battery.

Lead is actually a horrible poison. Iron phosphate is a garden supplement. It's dirt cheap and it's entirely benign.

We mix lithium in with it. It forms what's called a polyanion and more commonly a olivine crystal structure. And that's a little bit complicated.

We have the phosphate, would be the larger blue sections here. The orange sections would be our iron that's bound to that. And between that is some channels, a zigzag path really of space-time.

There's a lot of space left over in this matrix. And the lithium ion, if that structure is sufficiently negative, will be held in place for that same electrostatic charge. But there is a difference on this end.

These channels are very narrow. And so we think of that as, they call it a domino effect. If you add an electron in here at the bottom, anywhere in here, it has to move up.

All the electrons in front of it, lithium ions in front of it, have to move ahead one position. And they are positions, kind of like dimples in a Chinese checkers board, BBs in the little toy that you have to put all the BBs in the right place. They lock into these positions.

And then if you shove another one in behind, it'll move the ones ahead up. And so they actually have to tunnel in there. And that diffusion is what limits the power output of lithium ion phosphate and causes the diffusion delay.

As well as on the anode side. You can see over here on the grain, a very small nanoparticle of lithium ion phosphate, that it works its way in from the out to the in. And that's what they're trying to depict here, is the core and then the outer area as it fills in with this intercalation.

I've got another diagram that shows this a little better, a little larger, but again, you can see the phosphate and the iron and the red balls are lithium ions that have winded their way through this tunnel structure to intercalate there. Let me put up another diagram. This is from the thesis of Mr. Jens Grut, Division of Electric Power Engineering, Department of Energy and Environment at Chalmers University of Technology in Goteborg, Sweden.

And he did a 150 page analysis of cycle testing in lithium we're going to talk some more about because I like it, it's a little scattered. This diagram shows you some of the breakdowns that are common in lithium ion cells. You have your current collector corrosion, micro cracking of the graphene layers, and here is your SEI layer.

And they're talking about SEI dissolution, SEI reformation and growth and so forth. Let's talk about the solid electrolyte interface layer because it's kind of important to Boeing Dreamliners. When you first make a cell, there are no lithium ions in the cathode.

There's lithium mixed in. And what we do in our first charge, it causes the electrons to be ripped away from the lithium and we then have a lithium ion. And that migrates through the electrolyte and the electrolyte has lithium ions in it, as I said, the pressure builds and it intercalates into the anode side.

But along the way, there are some interactions. Occasionally, a lithium ion, for whatever reason, will collide with some of the materials in the organic solvents that are used to suspend the lithium salt, the LiPF6, or lithium hexafluorophosphate in the electrolyte. The solvents commonly used are an ethylene carbonate, EC, dimethylene carbonate, DMC, diethylene carbonate, DEC, and occasionally some polyvinyl carbonate, PC, that's used as a stabilizer.

And that's the liquid that make up the electrolyte. Now, why do we do that with organic solvents? Water undergoes hydrolysis at a little over two volts. And so an aqueous solution, since our cell, the difference in potential in the cell is the lithium on the cathode is about 3 1⁄2 volts.

The potential of the anode is typically 100 to 150 millivolts the other direction. And when we algebraically sum that positive and negative, that's where we get our 3.4 volts open circuit voltage on a lithium iron phosphate cell, which is quite different from the other lithium chemistries. They'll be at different redox potentials for the anode.

Cathode may exhibit a different open circuit voltage. As we go across here to the anode, we might collide with certain impurities, the most horrendous one being water. But these organic solvents do not hydrolyze at the 3.6 volts.

Water would, and that would essentially destroy our electrolyte, give off hydrogen and oxygen and blow up our battery. So at these higher voltages of 3.6 volts, we have to use non-aqueous, a non-water bearing, organic solvents, and EC, DEC, DMC, and PC are the normal ones. Everybody's got a different magic sauce of those electrolytes, thinking it'll make their battery better.

And some do. Unfortunately, the lithium ion, if it comes in contact with certain elements of these larger organic molecules, can have a reaction. The first thing I want you to remember is that reaction always results in heat.

Now, if it's an occasional bump in the night, it's a little bit of heat. And it causes it to rain a precipitant that is kind of a polyvinyl. And that migrates to the surface of the anode and coats the grains of the carbon.

And that, in the formative stage of the battery, and certainly the first couple of charges, they tend to charge it very slowly at low current levels. And this SEI layer, or solid electrolyte interphase layer, builds up kind of a rubbery skin on the surface of our carbon anode. That's kind of important in this case.

And here's why. We can store a lot of lithium ions in this carbon. And to give you an order of magnitude there, the iron phosphate can typically store about 170 milliamp hours of cations per gram.

The carbon is actually much more capable at about 372 milliamp hours per gram. So it's over twice the thing. Now, they'll adjust that by having more of the lithium iron phosphate and a thinner layer of the carbon.

But it's still normally, you'll have more storage here than you have over here. As this thing charges the first couple of times, this SEI layer forms. Well, it's got some unique properties.

It kind of glues all the carbon together, which the carbon does get about a 10% volumetric expansion and contraction during intercalation and deintercalation. But more importantly, the organic molecules in the electrolyte cannot pass through the SEI layer at all. They just can't do it.

But the lithium ions can. Now, they have to kind of tunnel through there. And so the SEI layer becomes kind of a function of our DC impedance or resistance of the battery itself.

Because lithium ions have to migrate through that SEI layer. Interestingly, in an over-discharge situation, or in trying to charge in very, very cold weather, really for lithium iron phosphate, anything below freezing, the lithium can't migrate through the SEI layer then. And it forms a lithium plating on top of the SEI layer, which is not good.

But normally, whenever we're charging the cell, lithium ions migrate through the SEI layer and are stored between the graphene sheets on the anode. And these graphene sheets, as this picture depicts, are not how you normally see them with, you know, just teeth sticking out at regular intervals from the current collector. They're actually a jumble of sheets.

And so it's not homogenous. Some places accept lithium ions more readily than others, because the graphene sheets are turned in a certain way or at a certain polarity that's more receptive to the lithium ions. So the more homogenous we can get this, the smoother our SEI layer will be.

But this is at a molecular level. So once the lithium ion is in there, it's safe. Because the electrolyte, the solvents, cannot pass through the SEI layer.

This lithium ion over here in the carbon cannot come in contact with the solvent and cannot have that reaction. And that's a good thing, because over time, we're going to bring lithium ions through here and store them over here, and they're going to build up. And so that's an important concept, is that that SEI layer insulates the lithium ions from reacting with the solvents.

And the solvents, it takes a certain contact with the lithium ions to have these reactions. So many lithium ions come through and don't have that reaction at all and get through the SEI layer and are safe in the carbon. Anode.

Now, let's charge the battery completely. And we have taken all of the lithium out of the cathode, and we have migrated it through the electrolyte, and we have stored it in the anode. But we still have a charger hooked up here.

That's going to do a couple of things. The remaining lithium ions in our electrolyte that are always there begin to migrate out into the anode, but that drops the positive level of the electrolyte. And if you drop the positive level, you're becoming more negative.

You have to have a difference of potential here of at least 50 millivolts, better 100, to get across here. And so what happens is that potential drops, and we start to have lithium ions collect on the surface of the SEI, and it can't get through. And as another lithium ion comes along, they combine into a crystalline structure.

And we build up lithium plating on the SEI layer. Some of it does get through the SEI layer from that crystalline structure. And so we have kind of a wound in our SEI layer that is a function of the size of that lithium crystal, and it tends to grow outwards as a dendrite.

Some describe it as moss. More commonly, you'll hear it termed a dendrite, a little needle that comes out and starts poking into our separator. If it goes through the separator, we short the cell and we have a fire.

But we don't have to do that to get a fire. Well, over here, we're still taking electrons, we're still charging. Where can we get a cation from this now delithiated iron phosphate? Well, from the iron.

It's a cation too, an ion, a positive ion. And so we start to leach iron that is released from our cathode side into the electrolyte. This has a couple of effects.

One, we're seriously degrading the structural integrity of our matrix at that point. Remember, the lithium ion just kind of tunneled in, but the iron was kind of bound to the phosphate. As we start to pull more negative charge off of this, making a more positive charge, the iron positive cation comes out.

And it starts to migrate through the electrolyte and it deposits on the SEI layer. And it will form dendrites. And it'll even form compounds with the lithium.

And so now we have wounds and ulcer here. When we do that, we're opening a hole, a crack, a fissure in the SEI layer. And that lets these organic solvents get to some of the nearby lithium ions that are held in the carbon.

And that gives off heat. And that kind of worsens the problem. Softening the SEI layer around this ulcer of dendrites of iron or lithium.

And lets in more organic solvents. And obviously if you have a point and it's growing, the circumference of the ring grows too. And so there's a greater area and exposing new areas of carbon with lithium tightly intercalated in there to the organic solvents.

And they react with the organic solvents. They're held in position now. They're going to come in contact.

And that gives off heat. Between 110 and 130 degrees centigrade, the SEI layer starts to come unglued. It starts to melt.

It's coming apart. And as it comes apart, now those solvents, here is this huge treasure trove of lithium ions suddenly exposed, fairly suddenly exposed to these solvents and reacting with it and generating heat. And we go into thermal runaway.

Now before that even happens, at about 90 degrees centigrade, one of our organic solvents, dimethyl carbonate, I believe DMC, has a boiling point of 90 degrees centigrade. And that starts to give off gas which causes our cell to swell up. And there is no scenario where a cell simply swells on its own.

There's no magic here. Swelling is telling, it's damage. And it's caused by the gasification usually of DMC.

And DMC also gives it kind of that sweet pear smell when it vents. And you can smell a battery going bad. It's kind of a sweet pear smell.

But in any event, this is going into thermal runaway. And as it gets hotter, the remaining SEI dissolves. And now all of our lithium ions can react with our solvents and the temperature just keeps going up.

Over here on the cathode, we have some iron remaining, some phosphorus, and some oxygen. And at about 250 degrees centigrade, up to about 400, this starts to break down and release free oxygen. Now free oxygen can combine with the electrolytes and with the lithium ions.

And we have a fire burning out of control that you can't put out because it makes its own oxygen. Why do I like lithium ion phosphate so well? Well, because it's at 250 degrees centigrade and up to about 400 degrees centigrade. And it's not really a lot of oxygen that it releases.

In the case of lithium cobalt oxide, it's a lot of oxygen. And it starts to go at 130 degrees centigrade, which is back here where we're melting our SEI layer at between 130 and 150 degrees centigrade. So much lower temperature will cause free oxygen in a lithium cobalt oxide cell.

And that's why they burn. They don't burn because they're lithium cobalt oxide. There is no magic.

They do not spontaneously go into ignition. If they burn, it's because somebody overcharged them. Now the voltage on lithium cobalt oxide is 3.72. There is somehow gotten to be a belief system, usually among the same guys that are trying to top balance these batteries, that they can float that cell at four volts and they will not overcharge it.

I have no idea where this comes from. It's prevalent. It runs through the whole thing.

And I think it goes back to lead acid. The open circuit voltage of these cells is a function of the lithium compound, the redox electric potential of the lithium compound in the cathode, algebraically summed with the same thing in the carbon anode. And whatever that comes out to, you, me, and Jesus Christ can't change.

It just is what it is. If you apply any potential to these metal current collectors above that, you are charging the cell. That you're not charging it very hard, nobody, including most of all the cell, cares.

You're still charging the cell. You're still migrating lithium and perhaps iron from the cathode to the anode. That you're doing it slowly or fastly is not the point.

The damage is really pretty much the same. It may be slower coming up to it until this goes into cascade and you actually melt the SEI layer and dump. You could kind of sneak up on it for a while and do a lot of damage without actually getting into thermal because it's more gradual.

But it's eventually gonna go. And when it does, you go into thermal runaway. It's better to have a lithium iron phosphate cell than it is a lithium cobalt simply by temperature.

But either one of them, neither one of them will spontaneously and if by magic burst into flame, no matter how roughly you handle them, external temperatures can bring these up to where all this happens anyway. Above 110 degrees centigrade, the SEI layer's gonna start to melt if the battery's not even working. Above 150 in lithium cobalt, it's going to start giving off free oxygen.

Everybody worries about the electrolytes. It's true they are flammable. And it's true if they vent, they will gasify into a very fine vapor that theoretically could be explosive.

But what these are are methanols and ethanols. It's like whiskey. You can pour 100 proof whiskey in your hand and light it with a match and it won't burn your hand while it's burning.

It's escape velocity is faster than the flame. I've done it. Your hand doesn't even get warm as long as it's going up.

So while the solvents are flammable, that's no part of this really. It's not a problem. It doesn't cause anything.

It's a side effect. The dimethyl carbonate, DMC, has the lowest boiling point at about 90 degrees centigrade and it turns to a gas and swells up the battery and vents the battery. And that's really kind of disconnected.

It's an early sign there's a problem, but it's not what causes the fire. And so the real problem is the gradual and then accelerating breakdown of the solid electrolyte interphase layer, which in reacting with the lithium ions, it gives off an increasing level of heat and goes into a thermal runaway situation. If that catches up the cathode, it's kind of all bets are off.

You will not get it to stop burning until it runs out of oxygen. And halon and so forth on top of it doesn't have any effect because it's getting the oxygen from the material itself. I mentioned, so I hope that that shows you what happens when you overcharge the cell.

Very quickly and because it's not terribly interesting, but you saw in my deconstructively over discharged cell that below a certain potential, the copper on the current collector starts to lose physical integrity and it leaches through the SEI and into the cell and it can be deposited anywhere. And it does form dendrites and it does short your remaining cells, which have no charge on them anyway, but this is why you can't put a charge on them when you get done, you're shorted out by dissolved copper. So that's your two failure modes.

Is over discharge and overcharge and overcharge being the more likely to cause a fire. Over discharge can, and that is the other mode of overcharge. If a dendrite pierces the insulator and shorts to the next cell, you will get current flow and it will cause heat.

Well, now we got the same problem. That heat is gonna break down the SEI layer and melt it. Again, between 110, 130 degrees centigrade, this comes apart.

It doesn't care where the 110 or 130 come from, it's still gonna come apart and go into thermal runaway. So dendrite formation to the point where it pierces the separator and actually shorts an anode to a cathode, you'll get instant current flow there. And in the case of over discharge, you really don't, but the next time you charge it, you will.

And so if you have a battery that's gone flat, hook it up to the charger and run a few amps into it. What you'll normally see is it comes up to about a volt, a volt and a half, and then it reverses and starts to go down. You're done.

If it doesn't continue on up, you will not recover that cell. It'll normally come up, I've seen them to a volt, a volt and a half, and they're coming up pretty good and then they reverse and start going down. You're headed for a short circuit and it will cause a fire.

If you continue to charge, just disconnect the charger and walk away. You're done. You cannot get that one back.

Let's take a short break. I mentioned that I got some of these diagrams from a guy in Sweden, who's doing his thesis for his licentiate from an engineering school. He's combined a lot of too much into one 150 page document, but buried in there is some very interesting information about cycle life, cycle life testing, and the effects of various things on the longevity of your cells.

Stay with me. We're back at the big green wall, and we're going to talk now a little bit about what causes a deterioration in your battery cells. I have a relatively new document, 2012, which is a thesis for the degree of licentiate of engineering titled State of Health Estimation of Lithium-Ion Batteries Cycle Life Test Methods by Jens Grut.

He's at the Division of Electric Power Engineering Department of Energy and Environment, Chalmers University of Technology of Gutberg, Sweden. And he did this in 2012. Like many young people, he bit off a bit more than he could chew.

He's combined a whole lot of things into one massive test. Impressive, not only in its length, but its breadth, most of which serves to confuse a few basic issues. But he did get some very interesting things in there and did, in his conclusions, dig them out fairly properly.

I think this is our document. What we're showing here, this is not a good graphic. It's not going to be easy for you to read, but I'll point some things out to you.

Our axis, vertical axis here, is percent of the original capacity of the cell with 100% at the top and 0% at the bottom. He calls it capacity throughput across the x-axis. This is actually the number of cycles out to 12,000.

He actually performed 9,000 cycles on a whole series of cells under different conditions. And we have to talk about that a little bit. The purple and the kind of dirty green graphs here are both cycle A. Cycle A is a very convoluted thing.

It's basically a simulation of a city bus route by a hybrid bus in Gothenburg, Sweden. Now, I know you all thought that Batman was a U.S. figure. He's actually apparently Swedish, and as we all know, Gothenburg is his, or is that Gotham City? In any event, Gothenburg has a hybrid bus.

They took 40,000 measurements during a typical day route through this city and reduced it to 2,000 seconds at different levels, and so ran this as a cyclic test. They did 9,000 of these cycles, and if you notice, there's kind of a line. That's the filtered output.

The individual marks are where they did a repetitive performance test on the battery that was completely different. That measured the capacity, the power output, the internal resistance, and so forth of the cell. And so they plotted it by the number of cycles, of course.

Interesting to note of cycle A, and I've got it over here on this table, but cycle A only varied from 50% state of charge down to 22.6% state of charge, but included accelerations of this bus and regenerative braking. And the peak charge current was 17.3 C, and the peak discharge current was 22.3 C, which is a lot of current and a lot of strain on a cell. You will see also a cycle C, and it's over here in a blue and a kind of a teal that you can barely tell apart.

It just looks like a dark blue line. And that is cycle C, which is from about 11.4% state of charge to 100% state of charge, and at a constant current of charge and discharge of 3.76 C. Now this is how you normally see a cycle testing net, is 100% down to some low value or even 0% at a fixed rate. This is kind of a high rate at 3.76 C. You'll often see this done at one C, but as you can see, the two C cycles both reach 80% of capacity right about 2000 cycles.

And these are indeed LiPo4 cells. They're not the ones we use. They're not A123, and they're not CalBs, but very similar cells in all respect.

And so when we say you get 2000 cycles, this is the testing that does this. Now here's one of the reasons I'm kind of keen on these batteries. Cycle A is more like you're driving a car.

Actually, it's more like you're driving a bus, and it's more like you're driving a bus in Gotham City, Sweden, but it's a real life thing where you accelerate and decelerate and do this. It's actually 2000 little iterations over 2000 seconds, draining us from 50% state of charge down to about 22.6%. Notice that we're not going to fully discharge, and we're not going to fully charge. This starts, each cycle starts at 50% state of charge.

He only fully charges it on the dots here, circles and triangles, the green and the purple, where he stops and measures capacity of the cells. At 9,000 cycles, we're still about 85% on the green one, and we cross 80% right at about 9,000 cycles on the purple one. Let's talk about that for a minute.

The purple one is at 23 degrees centigrade. The green one is in a climate box at 35 degrees centigrade. He wanted to see what the effects of temperature was on life of the batteries, and stated early in his assumptions that of course it would cause earlier failure.

Arrhenius law, our equation, and unfortunately his data supported in my view that for some reason these cells, everything in them works better at higher temperatures up to a point, and I think that point is about 45 degrees centigrade. Doing this at 35 degrees centigrade, what's that, a little over 100 degrees Fahrenheit, the cell lasted noticeably, demonstrably, and measurably longer than it did at 23 degrees centigrade, and normally you'll see our specs of cycle life and our curves at 25C, but for some reason he used 23. If you look at cycle C, of course, with the 2,000 cycles, that's actually, both of them, there's no difference, and so they draw one kind of blue line.

There's actually two mapped there, and one of them is pluses and one of them is squares, and it goes on this extended cycle map at 12,000 cycles, it looks very sharp, but this is our normal cycle life depiction of 80% to 2,000 cycles, but that's charged to 100% and discharged to 11% in each cycle, and there really isn't any difference between 23 and 35 degrees centigrade, which could lead some to believe, because lead acid cells fail earlier at higher temperatures, that that would be true of lithium cells. I've known for some time that it isn't. I've actually read some papers recently where they prove that they were.

They were really bad papers, and so I haven't cited them. This young man's student's work I find much more persuasive. Because of the number of cycles, the very careful climate control of the 35 degree test, and at every step on cycle A, the real driving cycle, that you can see a clear differentiation between the green and the purple, with the purple failing first, and the green never really quite reaching 80% during the test

It never did get down to 80% capacity on any of his interim, what do you call them, RPTs, reference power tests. And so there's a couple of things that jump out of this at me, and kind of tend to refute a lot of what is being talked about about these cells. First, elevated temperature, at least at a 35 degree centigrade rate, does not decrease cell life, it appears to increase it.

Second, note that this Gotham City driving cycle is at 17 and 22 C, very high charge and discharge current rates, and clearly it's not a problem. And by doing the actual driving, when we say that, the other thing that jumps right out at me is here is our curve of 2,000 cycles, 200% charge to 11% state of charge, and simply repeated at a continuous rate over and over again by a very stupid MAKO test computer. That's what we're used to seeing, and that's what we're used to seeing about LiPo4 cells at 2,000 cycles, sometimes 3,000 cycles.

They say that if you limit it to 20% depth of discharge, you can go 3,000 cycles. No one sits around and runs their batteries this way. You're going to drive a car like cycle A. And the good news is that in this test, you get 9,000 or 10,000 cycles and you're still not at 80%.

And that is what's going on with LiPo4 cells. Now, I will say that there is an extreme element in here and that his maximum state of charge is 50%. And so it would appear that the dangers to cycle life are not temperature and they're not the amount of current that you draw out of them.

It is a function of the extreme ends of the charge and discharge cycle. I've said this for years. If you could charge it to 51% state of charge, discharge it to 49% and then charge it again to 51% and stay in that range, they would last forever.

According to Mr. Groot's study, I would say probably three and a half forever. And again, the trick is not to over-discharge and not to over-charge. We bottom balance the cells to prevent damage in over-discharge and we under-charge the cells slightly.

I would say if cycle life is your main criteria, and it's certainly up there for me. I don't need the range. I need the batteries to last a long time.

That it would be perfectly permissible to drop your charge voltage, not even to jack 3.55, but let's say 3.4, the open circuit voltage of the cells, and charge them to 60% or 70% and then stop them in all cases before they get below 20%, these cells could literally do 10,000 cycles. If you do 300 cycles a year, that's 30 years that the batteries will last. And I believe it's entirely possible.

The functions of aging, he cites primarily as a depletion of the recyclable lithium. As you use these cells more, very gradually, the SEI phase formed very quickly at the beginning and then much more slowly, but it does continue to form for the entire life of the cell. It thickens and it takes more lithium out of the game.

You also have fractures in the carbon and fractures in the cathode material that eliminate recyclable lithium. And so he blames most of the aging on depletion of the available lithium ions. That opens up a curious situation.

I could pull that vent cap and with a needle, stick it down in there and poke a bunch of lithium hexafluorophosphate in, LIPF6, and replenish the amount of available recyclable lithium. Now I can't restore the areas that are blocked off by lithium plating or iron plating or so forth, but if you'll avoid fully charging them and avoid fully discharging them, you shouldn't have any of that. And so we can see a huge extension of life cycle way beyond what we had assumed.

And this was alluded to by Professor Jay Whitaker at Carnegie Mellon after some tests of A123 cells that 7,000, 10,000, 15,000 cycles are not out of the question in a certain restricted operating zone. And this then appears to be it. So we can take a lot of current out of the cells.

I had suspected that too high a current level might damage the cells. And of course, I've been kind of the lonely voice on that temperature, high temperatures were not as damaging to cycle life as we had thought. In this, it's absolutely definitive and with a huge number of cycles and a clear delineation of temperature from 23 to 35 degrees and a big separation in the results that a higher temperature up to a point actually extends the life of the cells, increases its performance, increases its capacity.

And it isn't always a good thing for lithium ion iron phosphate cells. And so the heavy current loads and temperature do not appear to be players in shortening the life, but fully charging the cell and fully discharging the cell does. Here's 2,000 cycles, here's 9,000 cycles.

The 9,000 come as the results of charge and discharges, a map, a program derived, they took 40,000 data points from an actual bus drive and filtered them down to 2,000 seconds with a different value each second and ran that as a program. And that was the cycle they performed there and had five times longer cell life with that as long as they started at 50% state of charge and didn't go below 22% that it was fine. And so the biggest determinant of cycle life I would draw from this young man's work is the state of charge, how close, how far you push the limits, the top and the bottom.

There's no reason to be there. Fully charging a cell, what if you gave up a mile, two miles, five miles and just drove it less? It's a $10,000 pack of batteries and it'll last 20 years now instead of five years. Is that worth it? I don't know, the batteries are gonna change, maybe not.

The amount of current you take out of them doesn't appear to be a problem and the heating does not appear to be a problem. You certainly wanna keep it below the area of 60 to 90 degrees centigrade. But we don't have a problem with that anyway.

Our cells will generate 10 degrees centigrade under extreme abuse, current wise, under 3C charges. And so the high temperature, the necessity for cooling them, I simply do not see it. We do not cool cells.

In fact, we're leaning more toward heating cells all the time, certainly if you live in a cold weather climate. Lithium plating is very real when charging below zero degrees centigrade and it's quite extremely real at 20 degrees centigrade below. So if you're gonna do anything, heat your batteries, don't cool them.

And I'm gonna be a little less worried about how much current we take out of these cells. 10C from these calves, I don't think it's a problem. We'll use 60, and what that means, I can use 60 amp hour cells, take 600 amps out of them, and it's probably not going to do anything to my cell life as what I'm getting from this paper.

And finally, and to reiterate, very carefully, very fully charging your pack is a loser. Let it get up to 3.45, 3.55 volts per cell. If you have 10 cells, that's 35 volts, cut it off.

Take the penalty. It's not worth damaging the cells to get another mile of range, or even two miles. And what some of these top balancing guys were doing was getting 300 yards, endangering the cells continuously for absolutely nothing.

Not enough range to fool with. And so that's my advice on life cycle extension, and this paper has been pretty definitive in what that is. Not heat, not current abuse, stay to charge.

Avoid the two N's, and your batteries will last a long time. And the good news is, in a real driving cycle, your cycle A, you're automatically gonna get a lot more cycles than what's on the spec sheet, cycle C, at 2,000 or 3,000. So all around good news, I think.

And so I thought I'd throw that in with our discussion of how to burn up a battery. It's also how to make one last forever. And that appears to be the state, I would say, the ideal is 80% state of charge to 20% state of charge.

Coincidentally, that's about what General Motors is doing with the Chevy Volt. Who would've thunk it? Stay with us.