NEW Potential Breakthrough using Tesla's Dry Electrode Coating Process

Welcome back everyone! I m Jordan Giesige and this is The Limiting Factor.

27% of the energy in nickel-based lithium ion batteries isn t accessible, which could

provide many EVs with 38% more range. That s because nickel-based lithium ion

batteries are capable of operating at up to 4.7 volts or more, but EV makers restrict

the voltage with battery management software to around only 4.2 volts.

They do that because there s greater reactivity and therefore faster

degradation at higher voltages, which reduces the number of charge-discharge

cycles that a battery can last for. Battery scientists and engineers have

tried to solve the degradation problem with chemical and materials engineering,

but most solutions have had limited success. However, a new paper from The University of

Chicago has been able to achieve a 21.8% increase in energy, with long cycle life,

through a small change to the cathode design. The catch is that Tesla is

currently the only company in the world with the technology stack to make it work at scale,

which is their dry electrode coating process. So today I ll walk you through the cathode design,

why it works, and in the long term what it could mean for Tesla.

Before we begin, a special thanks to my Patreon supporters, YouTube Members, and

Twitter subscribers as well as Rebellionaire.com. They specialize in helping investors manage

concentrated positions. Rebellionaire can help with covered calls, risk management,

and creating a money masterplan from your financial first principles.

Let s start with how high voltages cause degradation in the cathode of a lithium

ion battery cell. To be clear, in this video I m going to be focused on the nickel-based

cathode materials that are used in longer range EVs rather than the iron based LFP batteries

that are commonly used in shorter range vehicles. I ll explain why in a moment.

The cathode film of a lithium ion battery cell contains three materials. First, the cathode

particles, which hold the lithium in a ceramic oxide crystal framework that s usually about

80-90% Nickel and 10-20% manganese, aluminum, and cobalt. Second, conductive carbon particles

form an electronic matrix to get electricity to and from the cathode particles. And third,

a polymer binder that holds the cathode film together like glue. 1h1, 1h4, 1h5, 1h6

As the battery charges, the voltage of the cathode increases from about 3 volts at 0% state

of charge to 4.2 volts at typical 100% state of charge. The higher the voltage, the more lithium

ions are forced out of the cathode and are pushed to the anode. At 3 volts,

the cathode structure is packed with lithium ions, and at 4.2 volts, about 73% of the lithium in the

cathode has been transferred to the anode. The exception is LFP batteries. That s because

compared to a nickel-based cathode about a third more of their crystal structure is filled by

structural atoms instead of lithium atoms, and so almost all of the lithium can be removed and they

still maintain structural and chemical stability. That means there's minimal headroom to increase

the voltage further by the time most of the lithium has been pushed out at around 3.6 volts,

which is usually considered 100% state of charge for LFP batteries. 1g, 2a, 2a1

Getting back on track, between 4.2 volts to about 4.5 volts, the nickel-based cathode particles and

conductive carbon particles start reacting with the electrolyte to produce a variety of solids,

liquids, and gases that reduce the cycle life of the battery.

Then, above 4.5 volts, the degradation becomes more aggressive. The nickel-based framework

of the cathode crystal starts dissolving into the electrolyte,

which poisons the electrolyte solution, and the cathode crystal starts cracking and collapsing,

which further accelerates the dissolution. In the past, battery scientists and engineers

tried to prevent those degradation mechanisms with improved electrolytes and better cathode designs.

Those solutions have provided great improvements in cycle life and small increases in energy

density for some high nickel cells, but for the most part, 27% of the lithium and

therefore energy still remains out of reach. Next, let s do a quick refresher on the basics of

a typical wet slurry coating process versus the dry coating process that Tesla uses.

That s because it s critical to understanding the University of Chicago research paper.

A typical wet coating process involves mixing cathode powder, carbon powder, and a dissolvable

binder powder with a liquid to form a slurry. That slurry is coated to the cathode foil and

dried in an oven. As it dries the binder comes out of solution and coats all of the particles

to hold them together, forming a cathode film. Tesla s dry process also starts by mixing cathode

powder, carbon powder, and a binder powder, but the binder used is PTFE, also known as Teflon.

When PTFE is exposed to shear forces and heat, it fibrillizes like bubble-gum into spider silk like

filaments that turn the powder mixture into a kind of dough, which is heated and compressed

to form a cathode film no drying required. With that refresher in place, let s look at the

paper mentioned in the introduction by Zhang, et al titled, Dry Electrode Architecture Design for

Pushing Energy Density Limits at Cell Level. The paper starts by describing mechanisms that

could be contributing to the high voltage degradation problem.

Figure 1a shows a 125 micron thick cathode that was formed using a wet process.

Figure 1b shows a cross-section of that cathode using x-ray diffraction and there are four spots

highlighted and numbered, which are pinpointing specific cathode particles. The image is a point

in time snapshot of a series that was captured using video, and the different colors indicate

changes in the crystal structure that occur as lithium ions leave the cathode during charging.

That time series data is graphed out in Figure 1c, where the progression of the crystal structure

changes during charging can be seen more clearly. Below 4.2 volts, the crystal structures of the

particles progress in unison. Then, above 4.2 volts, particles 1&2 diverge from particles 3&4,

showing differences in the crystal structure, indicating differences in lithium content.

Differences in lithium content mean the particles are losing lithium at different rates,

which means the particles are experiencing different voltages. So what the graphs are

showing is that above 4.2 volts, different regions within the broader cathode film are

experiencing different voltages, which in turn means different degradation rates.

So what s going on here? As I said a moment ago, in a wet-slurry based coating process the

binder is dissolved into solution and coats all of the particles as the solvent dries,

holding the cathode film together. What I didn t mention is that the binder is an

electronic insulator which can interrupt the flow of electrons through the cathode film. So

different regions and therefore particles within the cathode film have greater or lesser access to

the electrons that drive chemical reactions. That is, the poor electronic network created

by the wet process is the core problem and weakest link in the electrochemical chain,

so it s no wonder there s been limited success at solving high-voltage degradation through improved

electrolytes and better cathode designs. While those solutions might improve durability up to,

for example 4.5 volts, the reality is that in a cathode running at 4.5 volts, some of the

individual cathode particles might actually be spiking up to 4.8 volts or more leading to rapid

degradation despite the improved durability. That in turn means the process and engineering

involved in manufacturing the cathode needs to be fixed before chemical or

materials based solutions are applied. If the cathode film could be manufactured so

that the insulative binder didn t get deposited between the particles and block electron flow,

it would lead to more uniform reactions, eliminate pockets of extreme voltage, and reduce degradation

in the cathode as a whole at higher voltages. However, some of you may have noticed that there

s a few missing pieces here. For example, in figure 1C, why do the differences in

electrochemical activity only show up above 4.2 volts? And, even if the conductive carbon matrix

could be improved to prevent cathode particles from degrading, it doesn t solve for the fact

that above 4.2 volts, the carbon itself would still react with the electrolyte.

To resolve those questions, let s take a look at the solution proposed by Zhang,

et al, and it s performance vs a conventional slurry coated cathode.

The paper proposes replacing the carbon particles in the conductive

carbon matrix with a more string-like carbon material, then coating the binder over those

strings like electrical sheathing. That would solve three problems at once:

First, the binder sheath would insulate the carbon from reacting with the electrolyte.

Second, a string-like carbon material would have lower surface area than carbon particles.

Third, by applying the binder around the carbon strings rather than between carbon particles,

the binder would no longer create electronic resistance in the conductive carbon matrix.

When I first read these design principles, my gut instinct was that the solution they were

about to propose would either require a tedious manufacturing process or expensive materials. But

shockingly, their solution requires neither: First, use the dry electrode process that

Tesla claims they ve now solved. Second, replace the carbon particles

with Vapor Grown Carbon Fiber, or VGCF. At a high level, that s it. Pretty simple. Let

s take a look at the cost and how it works. At a molecular level, vapor grown carbon fibre

looks like a series of nested waffle cones, and in terms of manufacturing cost and complexity,

it s between carbon nanotubes and carbon fibre. At scale it would probably cost around

$100 per kilogram and would probably only make up around 2-4% of the cathode by weight.

If that s correct, it would only add about a $1-2 per kilowatt hour, or a 1-2% cost premium over

conventional battery cells. However, factoring in the 21.8% energy density boost it could provide,

the battery cells would end up being 20% cheaper. But, we re getting ahead of ourselves. How does

the proposed solution actually work? When vapor grown carbon fibres are used

in the dry electrode process in place of carbon particles, the shear forces of the mixing cause

the long PTFE filaments, shown here in yellow, to get twisted up around the carbon fibres,

which are shown in black. Then, the carbon fibres naturally form a chemical bond with the PTFE. That

results in a binder sheathed carbon fibre that s insulated from reacting with the electrolyte.

It s not completely insulated because the binder doesn t perfectly cover the carbon, but as we ll

see in a moment, it helps tremendously. As a side note, I imagine carbon nanotubes

could also be used, and they might actually be cheaper. That s because although they cost 10

times more and have greater surface area than vapor grown carbon fibre, they could be used at

dosages of maybe 20x times less. However, I don t know if they d entangle or bond with the PTFE

as effectively as vapor grown carbon fibre. Even if not, they would likely still improve

high-voltage performance in wet- or dry-coated cathodes by creating

a more continuous electronic network with less reactive surface area than carbon particles.

Next, let s check out the test results of the proposed dry coated cathode versus a

wet coated cathode. The images on screen are colorized cross sections of cathode material.

On the left are Zhang et al s proposed dry coated cathode design, and on the right

are a comparable wet coated cathode design. Images A and B show cathode particles in red,

the carbon-binder matrix in dark blue, and pores in light blue.

As you can see there s less porosity in image A, indicating better physical connectivity.

Images C and D highlight that connectivity more explicitly. The cathode particles are in

gray and you ll notice multi-coloured pixels between them. All pixels of the same color

share an electronic connectivity network. Image C for dry coating is mostly pink pixels,

means electrons can travel almost anywhere in the electrode through the conductive

network without having to pass through poorly conductive cathode particles.

Image D for wet coating shows many colors and large black voids,

meaning it s difficult for electrons to percolate through the cathode film.

Image E and F are where it all comes together. The colors indicate current density,

ranging from low current density in blue to high current density in orange and red.

Image E for dry coating shows that most of the cathode particles are blue and

therefore barely used as a current pathway. Meanwhile, the conductive carbon-binder matrix

is mostly shades of orange, showing that it s the preferred electronic pathway.

Image F for wet coating shows cathode particles that are mostly light blue

to yellow, indicating heavier use of the cathode particles as an electronic pathway. Additionally,

the carbon-binder matrix contains more yellow and green than image E,

so it s carrying less current. What all this means is that the

dry-coated cathode directs most current through the preferred low resistance carbon-binder matrix,

and any current that flows through the cathode particles does so relatively evenly.

That s as opposed to the wet coated cathode where the carbon matrix is electronically

fragmented, forcing uneven and heavy use of the inefficient cathode particles as a current path,

which of course leads to the variations in voltage and therefore degradation

that we saw earlier in the video. But why do the voltage variations

begin to appear above 4.2 volts in the wet coated cathode? At higher voltages

the cathode particles which are being used as an electrical pathway, but shouldn t be,

begin reacting with the electrolyte and form a thick cathode electrolyte interphase, or CEI,

that s an electrical insulator. That isolates some of the weakly connected

cathode particles and forces more current through the well-connected particles, which

increases voltage differences between particles and causes them to degrade more quickly.

Meanwhile, the current moving through the exposed carbon particles in the wet-coated

cathode reacts with the electrolyte above 4.2 volts, driving further degradation.

For clarity, on screen is the high-voltage degradation pathway of a wet coated cathode.

As for the proposed dry-coated cathode design, the carbon-binder matrix is highly conductive

and well-connected, which avoids using the cathode particles as a current pathway. That

in turn prevents thickening of the cathode electrolyte interphase, which means reactions

remain uniform throughout the cathode and avoids voltage spikes. Furthermore,

the carbon-binder matrix functions like insulated electrical wire, which prevents the carbon from

reacting with the electrolyte. This combination dramatically reduces high voltage degradation.

Again, for clarity, on screen is a summary of how the dry coated cathode design works.

So, how well does the proposed dry-coated cathode perform? Pouch cells using a

high-nickel NMC 811 cathode and graphite anode were charged to 4.55 V and retained

78% of their capacity after 1,000 cycles. That cycle life is roughly in line with typical

high nickel battery cells. However, at 4.55 volts, the dry-coated cell would

provide about 21.8% more energy density and therefore EV range than a battery cell

that reaches 4.2 volts. And, it equates 317 Wh/kg of useable energy density compared to

the 260 Wh/kg of the current Tesla 4680 cell. A 21.8% energy density increase is of course well

short of the theoretical 38% possible at higher voltages, but that may never be possible because

cathode dissolution accelerates above 4.5 volts. With that said, the 21.8% energy density increase

was reached with a basic cathode and anode chemistry, NMC811 and graphite.

Even a small amount of silicon, which is standard in many high nickel battery cells,

would add about 10% to the energy density. The next question is, will Tesla adopt the

cathode design proposed by the team at The University of Chicago? There s no way to know,

but I can provide some further context. First, Tesla has probably known about the

design for quite some time. That s because there s deep cross-pollination between Tesla

s Maxwell acquisition and Shirley Meng s lab at UC San Diego going back at least 8 years.

UC San Diego was one of the teams involved in the paper and Shirley Meng was listed as an author.

That lead time increases the odds we could see the high-voltage cathode design in the next 3-5 years,

because it often takes over a decade for these innovations to reach commercialization.

Second, Tesla hasn t yet started using silicon doping in the 4680 to increase its energy density.

That s likely because it would add complication during the critical early ramp phase of 4680

production. But, it could also indicate that they plan to take another pathway to higher energy

density, such as by increasing the voltage. Third, on that note, given the fully dry 4680

ramp has just begun, I m not expecting major chemistry changes in the next year or so. After

that, when or if they do decide to use a high voltage cathode design, they would likely take

an incremental approach to mitigate safety and performance risks. So we wouldn t see a 21.8%