I like big batts
...and I cannot lie 🍑 : Intro to energy storage and semi-deep dive into batteries
Before we start…
How many types of moving parts do you have in your smartphone?1 Let’s take a peek:
Apart from the obvious ones (push buttons and vibration motors) and the maybe not-so-obvious ones (speaker diaphragms and accelerometer masses), there’s always the overlooked battery with the most number of moving parts. And no, I’m not being annoyingly technical. The lithium ions shuttling between the electrodes in your battery are >12,000 times the mass of electrons and each ion travels farther than the entire accelerometer mass oscillates. The technical term for this ion movement is “mass transport”, and it’s a physical phenomenon all batteries need to be properly designed for. Not to mention the natural swelling/contracting of inner active materials and outer cell dimensions upon charge and discharge, the battery is very much a moving part.
Whereas CPUs, OLED screens, and flash memory are all wonders of “solid-state” semiconductor engineering that continue to wow the family, batteries are like the fifth-year problem child that gets the same amount of love as your internet service provider—unappreciated when it’s working, never good enough when you notice it, and we just don’t have that many options. But batteries are hard because they are moving parts, and they’re a completely different beast compared to semiconductor chips. So don’t even try to compare batteries with Moore’s Law—use Murphy’s Law instead. (Btw, for those of you wondering, “solid-state” batteries still require mass transport and are nothing like “solid-state” semiconductors—we’ll talk about this later).
But today, we’ll deep dive into batteries and I promise you’ll love big batts too. First, we’ll talk about energy storage in general and how it relates to the grid infrastructure we talked about in the last post. We will then take a first-principles look at what “batteries” even are and the various types of electrochemical energy storage systems available. Finally, we’ll understand “conventional” batteries and how we ended up with the lithium-based systems we have today that are taking the world by force. In the interest of article length, we’ll deep dive into Li-ion batteries and beyond in the next post (coming next week), as they deserve special attention. Shortly thereafter, we’ll deep dive into “molten-salt thermal batteries” and another subset of electrochemical energy storage known as “flow batteries”. But today, we will cultivate enough big pic energy to understand foundational differences between all of them.
A quick note about hype: So much of our path to an emissions-free future hangs on our ability to store tremendous amounts of energy that batteries rightfully deserve the attention they are getting, and we should accelerate the herculean scaling efforts of mature technologies already underway. But while developing new batteries is very hard for reasons we’ll discuss below, the biggest challenge in the industry today is hype. “Battery breakthroughs” are touted on a near-weekly-basis, and inaccurate/overly optimistic reporting from tech media ultimately serves no one. The general public, including investors ranging from VCs to Robinhood users, can get lured into thinking we are closer to solving critical problems than we really are. Technical folks are too often then placed in the position to either play the responsible party-pooper or, more dangerously, inflate claims and/or partake in outright fraud, ultimately hurting everyone’s credibility. Incredible progress has already been made in energy storage, but we risk overlooking and underinvesting in the actual suite of technologies able to address our immediate climate crisis if we keep up with the froth. From now on, I vote the phrase “holy grail” to be *bleeped* out when talking about batteries. It just doesn’t exist. Here, we’ll take an honest look at these technologies, what makes them hard, and appreciate the marvel of what’s already in our hands.
So…let’s get t’werk 🍑 .
A giant spring for a giant parade
Recall our Mad Maxian parade analogy from the last post for the electric grid, and how we concluded by looking at what our future grid could look like. Here it is again:
Remember our parade needs to move at a constant speed of 60 mph (our analogy of 60 Hz grid frequency) in order to not break stuff. In today’s grid, this just means that our cadre of enlisted tow-vehicles (generating plants) always has to maintain the same speed and keep the tow-lines taught at any given moment, even if there is significantly more horsepower available for cheap. If only we could somehow build in some elasticity to our parade…like installing ginormous springs. Our tow-vehicles could then run loose and operate at their maximum efficiencies, storing excess energy in the springs to be used later only when needed. This, of course, is energy storage.
So…what’s so hard about a giant spring?
As we discussed in our first post, energy is the fabric from which everything else materializes, and it’s stored everywhere across various length scales—sub-atomic forces between nucleons (nuclear), traveling photons (electromagnetic radiation), electron interactions between atoms (chemical), vibrations of molecules (thermal), or macroscopic properties of matter and objects (mechanical). So why is energy storage such a hard problem?
The hard part isn’t being able to find stuff to extract energy from. The hard part is being able to put things back into their original state so that we can store and extract energy over and over again in a controlled fashion. Recall that energy is like a rambunctious child (second law of thermodynamics)—it’s hard to contain once we unleash it. Not only can it be hard to harness the unleashed energy to do useful work, but it is even harder to convert it back into a stored, useful form…not to mention the myriad of other requirements we want. Basically, we want our ideal giant spring to be able to perform the following list of demands:
As you can see, building our giant spring is no easy task, and the hard part is the ability to satisfy all of the demands. In this case, excelling at a few of the demands but failing miserably in others is a non-starter. We’d rather have something that is mediocre across all demands. And to that end, the fact that we already have devices that can satisfy these requirements to a reasonable degree is an engineering marvel. The types of devices that are leading the pack are called rechargeable electrochemical batteries, and I hope you feel bad now for ever cursing your battery. It’s trying its best 🥺 .
In order to truly understand them, we will start by establishing a foundational understanding of electrochemical energy systems and devices. We’ll also keep referring back to our list of demands, because we’ll see that batteries are a world of tradeoffs with no silver bullet and certainly no “h**y gr**l”. Excuse me.
A first-principles understanding
In perhaps an odd twist of things, the best way to really understand electrochemical energy storage devices is to start by looking at how fossil fuels work.
Fossil Fuels
The primary way we use energy stored in chemical bonds of fossil fuels is through combustion, which just means a reaction with oxygen molecules in the air to produce heat (and light). Most times, we want to harness the energy to produce either mechanical or electrical work, so we use a “heat engine” to convert the heat into mechanical motion. If we want electricity, we need to subsequently convert the mechanical motion into electrical energy using generators, and this entire process is pretty inefficient (see the first post).
Combustion is a classic example of a “redox reaction”—short for a reduction-oxidation reaction—which just means that there is a transfer of electrons in the chemical process (other chemical reactions, like acid-base reactions, transfer protons instead). The hydrocarbons (a.k.a. the fossil fuels) are “oxidized” while the oxygen in the air is “reduced”. Assuming there is sufficient oxygen around, the exhaust byproducts will be water and carbon dioxide (along with some other species like sulfur and nitrous oxides from other crap in the fuel).
If we were to consider fossil fuels as a candidate for our giant energy-storing spring, then combustion would just be the “discharge” process. How would we “charge” it? Well, the reaction byproducts (water and CO2) need to be converted back into hydrocarbons. To do this, a special protein known as RuBisCo is used to reduce (or “fix”) CO2 from the air to produce glucose and oxygen. Of course, we don’t have the capability to do this—only plants do, and they do so via photosynthesis. Eventually some of the glucose becomes cellulose and other plant matter, which then takes another couple million years to decompose into the fossil fuels we can readily combust again. Talk about fast-charge. Obviously not a very useful spring (fails miserably at demand #2). Here’s what that looks like:
Note that the heat engine is the power source “device” here, but it is merely a reaction chamber and contributes no inherent energy of its own. Fuel is fed into it and the byproducts are exhausted out of it. If your fuel tank were infinitely large, we could theoretically run the heat engine as long as we want.
Hydrogen Fuel
In a similar fashion, hydrogen fuel can also be combusted using heat engines if mechanical energy is what we need. Here we just combust (oxidize) hydrogen, forming water in our “discharge” process. While water is strictly the only byproduct in combustion with oxygen, air is only made up of 21% oxygen with most of the remainder (78%) being nitrogen. The high combustion temperatures of hydrogen fuel in air therefore also generates harmful nitrous oxides (NOx) that we don’t like—something we need to be cognizant about. To “recharge” our fuel, we can zap water with electricity to recreate hydrogen and oxygen gas via a process known as electrolysis. This is generally quite expensive due to costs of operating electrolyzers, so hydrogen is usually produced in another way known as steam methane reformation—though it feels less like “recharging” (demand #2) since we’re now using other feedstocks. Here, natural gas (a.k.a. methane) is combined with super hot steam to produce hydrogen fuel along with carbon monoxide.
The energy content on a per-kilogram basis of hydrogen fuel is very high, which is why there is interest in using hydrogen fuel as a potentially “clean” alternative for jet fuel, among other things. But it’s hard to compress hydrogen into small volumes, and handling hydrogen can generally be a pain in the batt (ha, see what I did there). We’ll talk about all of this in a later post deep diving into hydrogen. For now, here’s what the hydrogen fuel “discharge/charge” process looks like:
Electrochemical vs. Chemical Reactions
As we’ve seen, chemical combustion reactions require the use of heat engines and generators to produce electricity in a rather roundabout way with low efficiency (20-40%), and the “recharging” process isn’t always electrically compatible. This makes it difficult to integrate with the electric grid system and function like the ideal spring we want. This is where electrochemical reactions are clutch. Whereas chemical combustion reactions can be like bartering chickens for wheat just so you can buy some milk, electrochemical reactions let us use the unified, liquid currency of electrons.
Electrochemical reactions are always redox reactions, but they are special in that the transfer of electrons happens in a way that we can interact with directly as electricity, and very efficiently. These reactions can either generate or consume electricity, and it can be easy to control the direction of the reactions by simply extracting or supplying electrical energy. There are several types of electrochemical energy systems we’ll talk about below, but the easiest extension from combustion is the “fuel cell”.
Fuel Cells
Fuel cells have basically the same reactions as the combustion examples we discussed above and generate the same byproducts for exhaust. But instead of using a heat engine to blow shit up (a lil’ primitive), a fuel cell elegantly removes the electrons from the fuel to oxidize it, passes the moving electrons through conductors in an external circuit, and sticks them into the oxygen molecules on the other side to reduce them. This controlled electron transfer via a conductor is the electricity we can use to power things. Simultaneously, when the neutral starting molecules have electrons removed or added, they become “ions” that are charged. While the electrons are being lured through an external circuit for humans to exploit, some of the bigger/heavier ions must travel through a membrane inside the fuel cell to meet with the other species and form the final reaction byproducts. Again, this ion movement is known as “mass transport”; it’s slower and much harder to design for compared to electron transport in conductors. Here, the hydrogen or hydrocarbons are oxidized at the “anode” electrode, and the oxygen is reduced at the “cathode” electrode.
Here’s what that looks like. As you can see, a fuel cell is just a “fuel” passed through, well, a “cell”.
The fuel can be a gas (like hydrogen or methane) or a liquid (like methanol), and the oxygen is pulled from the air as an “air-breathing device”, just like your combustion engine. But we need the electrons to move through solid conductors since that’s the only way we know how to use electricity properly. So how do we get electrons into and out of gases via a solid? This is where we need some assistance from things called “catalysts”. Catalysts are like matchmakers that help shy reactants hook up without ultimately being part of the relationship (because that would be weird). Unfortunately, these catalyst materials can be expensive materials like platinum which aren’t helpful to cost performance (demand #10).
Fuel cells are fuel-consuming devices—a one-way street for the fuel’s journey to being an exhaust byproduct. So the “recharge” process is performed separately, similar to the chemical fuel examples we discussed above. Although there is some development of rechargeable fuel cells that could reverse reactions within the same device, they are still very much in the early stages of development.
We will deep dive into fuel cells in a later post when we talk about hydrogen.
Metal-air Batteries
Now let’s meet our first battery (about time!), but we start with a type that may be less known and certainly less common, only because it’s the next logical progression from our previous examples. It’s known as a “metal-air battery”, and also uses oxygen in the air.
Whereas fuel cells and combustion heat engines have their fuel stored elsewhere and piped into the device to react with the oxygen in air, metal-air batteries instead carry their own fuel within the device as a solid—a metal. The metal needs to be one that is capable of donating electrons to an external circuit, forming ions on one side of the battery (oxidation). These metal ions then need to react with oxygen ions which are reduced at the other side of a membrane, where the membrane is typically soaked in an electrolyte solution. Only the smaller of the two ions will make the trek across the membrane so the process won’t be too sluggish (oxygen ions do the moving in Zn-air batteries whereas lithium ions do the moving in Li-air batteries). In a metal-air battery, the metal “fuel” is the anode, the oxygen is the cathode, and they are another kind of air-breathing device. These batteries therefore cannot be hermetically sealed off from the environment like what you have in your smartphone—they need access to oxygen.
Since a metal-air battery effectively only carries half of its reactants and “breathes” the rest, the attainable energy densities are naturally very high—but only where oxygen is available of course (e.g. not in outer space). It’s kind of like considering the weight of a human to be “functional”. At sea level, it’s just the weight of the dude/dudette, but if on the moon, you have to add the weight of oxygen tanks and additional “packaging”. The zinc-air battery is the only commercialized metal-air battery available today (the kind used in hearing aids), and it has an even higher energy density compared to our beloved lithium-ion batteries. A kinda big caveat though…it’s not rechargeable. We’ll talk about why in just a sec. Metal-air batteries boast the highest theoretical energy densities for any kind of battery system (good for our demand #9), with lithium-air batteries being the highest of all. Unsurprisingly, they are a fascination for researchers, but other rechargeable metal contenders include zinc, aluminum, calcium, cadmium, and iron. Before we get too dreamy though, let’s get real and burst some bubbles2.
Reducing oxygen molecules from air is a sluggish process, which means the discharge rate capability of metal-air batteries are low (see our demand #5). So they are either only suitable for low-power devices (hearing aids) or you will need gigantic ones to do anything meaningful. But more importantly, it is very difficult to charge metal-air battery systems. Theoretically, if we reversed the electron flow by supplying a current to charge the cell, we would evolve/bubble-off oxygen on the cathode while converting metal ions back into their metallic state at the anode. Much easier said than done. In metal-air batteries, much more energy needs to be provided during charging to break metal-oxygen bonds and “exhale” oxygen than you can get back on discharging, which makes these devices typically terrible in energy efficiency (our demand #4)—not to mention being slow to charge (our demand #2). Some ways to reduce this energy inefficiency (or “hysteresis”) is to use catalysts again like platinum, gold, or other fancy materials to act as social lubricants for the reactions. But obviously this goes against our goal of being cheap (demand #10). Also, re-depositing certain metal ions back into their metallic states on charging often creates funky structures called “dendrites” that can end up destroying the cells and cause the cycle life to be garbage (demand #6)—we’ll talk more about dendrites later. Additionally, remember that metal-air batteries are “open” systems that freely exchange gas molecules with their surroundings, unlike the tightly-controlled, hermetically sealed cocoons of your typical batteries. This means all the other crap in the air (e.g. moisture, CO2, and impurities) must not ruin the cycle life of the cells, which is yet another challenge especially for non-water-based (a.k.a. non-aqueous) electrolytes used for metals like lithium or calcium. Remember, lithium + water = 💥 🔥 . Many labs working on metal-air batteries pipe in pure oxygen as they work on the more fundamental reactions in the cell just to get them to cycle. If and when they do make some progress, they’ll then need to learn how to work with “room air” that contains less than a quarter of oxygen by volume—which will inevitably affect performance.
But one important upside to consider for metal-air batteries is that you don’t need to “pay” for half of your battery (e.g. the cathode), just like how you don’t pay for the oxygen you breathe. You will still need a porous solid structure on the cathode side to act as your air-electrolyte-electrode (gas-liquid-solid) exchange “substrate”, but the goal would be to have it cost considerably less (if expensive catalysts can be avoided). Pair this with an abundant metal anode, like iron, and an aqueous electrolyte that is non-fancy, the costs of an eventual energy storage could be very low (good for demand #10). Discharge and charge power limitations would require a huge installation, which would be impractical for electric vehicles but not too crazy for a grid-scale storage facility.
As you can see, there are a host of scientific challenges that need to be solved before rechargeable metal-air batteries can be reliably deployed in commercial applications. The challenges above also serve as an initial glimpse into why batteries are hard. Don’t expect to see rechargeable lithium-air batteries playing a significant role in our decarbonization over the next decade. Lithium-air is far from replacing lithium-ion batteries for cars, and will unfortunately not be electrifying commercial aviation within the next two decades given their low technology readiness level (demand #11) and the long qualification times required for flying things that contain humans. The only types of metal-air batteries that may play a role in our decarbonization are likely to be a cheap/abundant-metal + aqueous electrolyte system for long-duration grid-scale storage. But that is, of course, pending some serious development around energy efficiency and creativity around charging and cycle life, which will inevitably take some time.
Let’s now turn to “conventional” batteries—the kinds you typically think of when you think of “batteries” and the ones we’re primarily here for today. Unfortunately, we don’t have a cooler name for them, but they are quite badass.
“Conventional” Batteries
Previous examples of chemical and electrochemical systems all benefited from using the oxygen in air, and therefore did not need to lug around their other half of reactants. Conventional batteries on the other hand are fully contained systems that bring everything they need along with them wherever they go. They’re like moms ❤️ .
Here, the anode “fuel” is typically a metal, like we’ve seen above, but does not always have to be in the pure metallic state. Materials capable of hosting or alloying with metals are very common (like in lithium-ion batteries) and can help avoid some of the challenges with pure metals we’ll discuss later. In other cases, the anode can be molten metal (liquid) using a molten-salt as the electrolyte that operates at toasty temperatures. These are known as “molten-salt thermal batteries”, not to be confused with the thermal batteries that work by just storing heat (both of which we will discuss in a future post). Finally, the anode can even be in the gaseous state (!), like hydrogen in our fuel cell example, but contained within the cell. Remember: it’s a mom.
The cathode is now just a species contained within the battery that is capable of being reduced upon discharge. The candidate materials here are more variable, but they are generally (not always) metal-, oxygen-, or sulfur-containing compounds that can be reduced, and, in the case of rechargeable batteries, re-oxidized.
Discharging and charging the battery now just entails passing electrons from one side to the other while having ion exchange through a membrane/electrolyte. How elegant. Here’s what that looks like:
Since everything is contained within the packaging of a conventional battery, the energy content is a fixed amount limited to the “stuff” inside. The power performance (the rate of discharge/charge) is also tied to how the battery is designed on the inside. Unfortunately, the design strategies to improve power performance are almost always at odds with the design strategies to improve energy density (it’s a world of tradeoffs). So all conventional batteries suffer from the classic energy-power tradeoff, and compromises must be made depending on which parameter is more important (e.g. fast-charge or mileage).
We’re going to discuss conventional batteries below, so we’ll come back to all of this shortly in much greater detail (don’t worry my nerds). Before then, let’s look at the last type of battery to complete our first-principles overview of electrochemical energy systems.
Flow Batteries
“Flow batteries” are similar to the conventional batteries in that they contain all their reactants within a closed-loop system, but they are also similar to fuel cells in that the reactants are contained in tanks that are separate from the “reactor cell” where the magic happens. Both the “anode” and the “cathode” are now pumpable liquids that are redox-active (e.g. can be reduced or oxidized), and they are technically called the “anolyte” and “catholyte”—fancy names.
To discharge, the anolyte and catholyte are pumped to the reactor cell where there is an ion exchange membrane to allow one to be oxidized and the other reduced. They are both then pumped back to their respective tanks from which they came. The relative concentrations of oxidized and reduced species in the anolyte and catholyte tanks, respectively, therefore determines how “discharged” the flow battery is.
To recharge the flow battery, the same reactants are pumped back into the reactor cell, but this time the electrochemical activity is reversed to return the fluids back into their original state. Here’s how all of that looks like:
One interesting feature about flow batteries is that if you were happy with the charge or discharge rate (demand #2 and 4) but wanted to store more energy (demand #1), all you need to do is increase the size of your tanks. Similarly, if you were content with the energy content (ha!) but wanted more charge/discharge power, you just need to increase the area of the reactor cells or add more of them without messing with your tanks. This “decoupling” of the energy and power performance allows more modularity and scalability for commercial deployment. Another advantage of having energy stored in the chemical state of liquids is that you don’t need to worry about the swelling/contracting of solid materials that inevitably degrades cycle life (demand #6). And if the anolyte or catholyte ever need “servicing”, one could always go in and top-off the tanks or “rejuvenate” them separately. This entire system is certainly going to be big and heavy (bad for demand #9), but the point isn’t to stick it into a vehicle and expect to have room for passengers or a roof. This type of electrochemical storage is really suited for longer-term grid-scale storage.
But as always, batteries are hard and there remains significant scientific barriers ahead. For one, the liquids can be corrosive (bad for demand #8), the operational temperature range can be limited (bad for demand #7), energy efficiency can be low (bad for demand #4), and some of the materials are expensive (bad for demand #10). Still, flow batteries are an important technology that could play an important role in our decarbonization, so we will do a deep dive into flow batteries in the post after Li-ion batteries.
Voilà, our 35,000 ft cheat sheet
Stepping back, you now have a decent first-principles understanding of the various types of chemical and electrochemical candidates for the giant springs we need. Only the last three rows of our cheat sheet fall into the category of “electrochemical energy storage”, a.k.a. “batteries”. With the full picture in mind, though, it really isn’t that complex. It should also make sense now that comparing batteries with gasoline itself is dumb, because that’s not a fair comparison. The fuel is only one part of the overall system.
Our next step is to zoom in on “conventional” batteries as there is a ton of stuff underneath that one category. Today, we’ll focus on the history of battery development and how we got to where we are today. In the next post, we’ll deep dive into lithium-ion and the technologies coming after, and subsequently close out the “conventional” battery segment with rather unconventional molten-salt thermal batteries. The next post after that will deep dive into flow batteries.
The long journey to lithium
Let’s try to zoom through 200+ years of history.
In 1799, an Italian dude named Alessandro Volta wanted to prove wrong his fellow-Italian and semi-mad scientist Luigi Galvani, who was performing experiments on twitching frog legs and claiming animals had inherent “animal electricity”. So Volta stacked together some copper and zinc discs separated by brine-soaked cardboard to create the first battery (a.k.a. the “Voltaic Pile”), and boom, the rest is history. Lol jk. Here’s what the rest actually looks like.
The zinc-copper Voltaic Pile was commercialized by lying it down horizontally in a wooden box and flooding it with a dilute sulfuric acid electrolyte, appropriately named as the “trough battery”. 50 years later, a French named Gaston Planté took the same design but swapped the zinc/copper plates for lead/lead-dioxide plates, which became the first rechargeable battery still used today—the lead-acid battery. This is basically the same design as the one you have under the hood of your car, but obviously not in a wooden box.
Around the same time, another French dude invented the “Leclanché cell” named after himself by replacing the copper cathode from the Voltaic Pile with manganese-dioxide, creating a longer-running battery, albeit still non-rechargeable and flooded. But this design quickly became a paste-like non-flooded “dry cell” with a healthy amount of carbon conductive (hence called “zinc-carbon”) that could be used in any orientation without spilling. This enabled the first portable electronic device—the iPhone. Jk it was the flashlight.
Then a Swede named Waldemar Jungner invented the rechargeable nickel-cadmium cell in 1899, which had higher energy density than lead-acid, and would become the only rechargeable battery for portable devices for a long while. It was meant to replace lead-acid due to lead’s toxicity, but it kinda was the kettle calling the pot black, as cadmium was found to be even more toxic. This was also the first battery to use an alkaline (e.g. basic) electrolyte instead of an acidic one. These alkaline electrolytes inspired the replacement of the acidic electrolytes in zinc-carbon dry cells, which gave birth to the non-rechargeable “alkaline” AA/AAA batteries we know today.
Separately, Thomas Edison (this guy’s everywhere!) took the Ni-Cd battery and swapped the expensive cadmium for iron to create nickel-iron batteries for his envisioned electric vehicle revolution. They had better energy density than the heavy lead-acid batteries, but they sucked in low-temperatures and self-discharged a bunch. Before these issues could be addressed though, Henry Ford beat him to making mass-produced vehicles powered by internal combustion engines, and nickel-iron batteries died out along with hopes of EVs for close to a century. But these nickel-based cathodes proved to be a good chemistry, and a unique nickel-hydrogen battery was developed in the 1970s for satellite applications (among others). This used hydrogen gas as the anode (like hydrogen fuel cells), and therefore needed platinum catalysts and was contained in a pressurized vessel at 1,200 psi—it was basically an R2-D2-looking bomb. It had remarkable cycle life and low-temperature performance—great for satellites in the cold of space, but obviously impractical, expensive, and dangerous for humans on earth. So a “metal-hydride” anode was invented to “lock-in” the hydrogen in a solid form to be used as a battery for consumer electronics. This became known as the nickel-metal-hydride battery, which outperformed Ni-Cd in almost every single way and became the incumbent technology for rechargeable portable electronics until the 1990s.
Then came the Lithium era…
(Quick aside on Lithium)
Lithium has the lowest standard potential of any material. This just means it has a very high tendency/willingness (electrochemical potential) to donate electrons (oxidize) if we let it, and batteries using lithium will have comparatively higher voltages. Lithium is also the lightest metal in the periodic table. All of this means that it will be a great anode material for high energy densities. But as with anything, it’s got some dirty secrets as well, and we’ll touch on two big ones today:
Lithium is very reactive, especially with water, so we can’t use aqueous electrolytes or else there will be 💥 🔥 . Instead, we use “organic, non-aqueous” electrolytes that tend to be volatile and flammable.
Like zinc and most other pure metal anodes, lithium metal is difficult to recharge without funky things happening. While lithium is happy to donate electrons and dissolve into solution upon discharge, it is much less cooperative when depositing back into metallic form upon charge. Instead of depositing (electroplating) uniformly across the entire surface, lithium tends to grow messily with a high degree of internal porosity, becoming thicker and uglier with each recharge. Most challenging, though, are the whisker-like structures known as “dendrites” that grow with each cycle. This also happens to be a positively-reinforced behavior where the sharp points of dendrites become the favorite spots for the next lithium ion to deposit. They will continue to grow and pierce through anything in their way (they’re stupidly strong and remarkably stubborn), including the separator, and eventually cause a direct short-circuit between the anode and cathode. This will result in cell failure…sometimes with a ka-boom because our organic electrolyte is flammable.
(Ok, back!)
The obvious next step we took from alkaline primary cells was to then swap the zinc with lithium, keep the manganese-dioxide cathode, and use a non-aqueous electrolyte, which gave birth to lithium-primary batteries. These had much higher voltages (3.2 volts) compared to alkaline batteries (1.5 volts), and therefore contained much more energy. But given the proliferation of alkaline batteries in the AA/AAA form factors that had only 1.5 volts, Energizer commercialized a lithium-iron-disulfide cell that swapped the manganese-dioxide cathode (not touched since the Leclanché cell) with iron-disulfide (a.k.a. “fool’s gold”, pyrite). This dropped the cell voltage back down to 1.5 volts so it didn’t fry any electronics, but contained significantly more capacity than typical alkaline batteries.
However, all of these lithium swaps still remained non-rechargeable, primarily because of the difficulties with recharging lithium metal (discussed above). On top of that, there were challenges with finding a suitable cathode that could be reversibly charged and discharged thousands of times with high energy density.
The birth of Li-ion
Then in the 1970s, Exxon (yes, the oil company) pioneered the development of batteries for electric vehicles due to the OPEC Oil Embargo (oh, how things have changed). There, a young Stanley Whittingham spearheaded the effort to understand the phenomenon of lithium ion “intercalation” into cathode candidate materials. Intercalation is just a fancy term describing the reversible insertion of a guest species into a host material, which usually has a layered, shelf-like crystal structure to hold things (if you’re wondering why it sounds so weird, “intercalation” actually originated from the insertion of February 29th into leap years). Anyway, he discovered titanium-disulfide (TiS2) as a cathode material that could intercalate lithium ions highly reversibly, along with several other “metal-disulfide” candidates. Spurring out of these cathode developments, a tiny startup in Canada called Moli Energy shattered all expectations and led the way in commercializing the first rechargeable lithium metal battery, using molybdenum-disulfide (MoS2) as the cathode and lithium metal as the anode. But on one August day in 1989, a cellphone in Japan using these batteries caught fire3. It was later attributed to the lithium dendrites (it’s always the damn dendrites) finding their way through the separator, shorting with the cathode, and going into “thermal runaway”—an uncontrolled release of energy that positively reinforces itself to burn all the way through. This led to a massive recall of batteries, and Moli Energy quickly went bankrupt. Hard lessons from commercializing untested technology too quickly.
Thankfully, other folks were hard at work in the meantime. John Goodenough discovered lithium cobalt oxide (LiCoO2, or LCO for short) in the early 1980s as a fantastic cathode material capable of intercalating lithium ions highly reversibly, and at a much higher voltage compared to metal-disulfides. Another unique feature of LCO that differed from all other cathode materials was that it already contained the “working” lithium inside of it. Whereas all other batteries we have met thus far are assembled in a default, charged state (where the anode “fuel” is ready to be oxidized), a battery using an LCO cathode material would actually be assembled in the discharged state. If LCO were paired with a lithium metal anode, for example, you would actually start by charging the cell, which entails pulling the “L” out of LCO and plating it onto the existing lithium metal. Here, the lithium metal would be extraneous. This was an important step needed for our final rockstar: Akira Yoshino, who designed the first actually rechargeable lithium-based cell that got around the dendrite problem.
Instead of using pure lithium metal as an anode, Yoshino instead used a carbon host material (heat-treated petroleum-coke). Carbon materials have a layered crystal structure (known as graphene layers) that can also reversibly intercalate lithium ions, just like LCO. By using carbon, this eliminated the need to ever deposit lithium in the metallic state and risking dendrite formation. Since the lithium in these new cells would always remain in “ion form”—shuttling back and forth between the cathode and anode “shelves”—these new batteries became known as lithium-ion batteries. All the useable lithium in these cells would come from the “lithiated” cathode since carbon obviously contains no lithium inherently. Shortly after Yoshino’s invention, the first Li-ion batteries were commercialized by Sony in 1991. The petroleum-coke eventually got replaced with graphite, and the LCO-graphite battery is still the chemistry you find in your smartphone today.
And with that, a star was born. Li-ion batteries have the highest energy densities out of rechargeable batteries (demand #9), the highest power densities for discharge (demand #5), very little self-discharge (demand #3), very high energy efficiencies >90% depending on rate (demand #4), and reasonably high cycle life (demand #6). In the next post, we’ll deep dive more into their shortcomings and how those are being addressed.
Whittingham, Goodenough, and Yoshino were awarded the Nobel prize in chemistry in 2019 for their work in pioneering the Li-ion battery revolution, but many other contributors also helped make it possible.
And boom…
…that’s the 200-year history of electrochemical batteries in a nutshell, and the story behind the two major types of batteries we use today: lead-acid and Li-ion, the latter taking a much more interesting journey. Despite the long journey, we have still only successfully and commercially mass-produced a handful of battery types. This stuff is hard, y’all. There are of course more types of off-shoot batteries that didn’t make it into our brief history lesson, but I made a (non-exhaustive) table below showing the various anode and cathode combinations, superlatives, and dirty secrets of each one.
Some takeaways from our history lesson
Battery people should not be allowed to name things. We have no consistent logic in naming batteries.
Battery development often entails hot-swapping one thing for another and calling it a day. You should notice a pattern with common materials used on either side.
Primary batteries can have way higher energy densities because reactants don’t need to be reversed, and can be super reactive to the point they’re explosive. We also don’t really care how they die. Rechargeability is the hard part.
Rechargeability is hard for pure metal anodes without dendrite growth. This is true for almost all metals (including cadmium and lead), but only notorious in lithium because of consequences with flammable organic electrolytes.
Now that we understand how things came to be, we are now primed to get into the nitty-gritty of Li-ion and what’s to come. We will do this in the next post, coming next week, so you better get ready and start stretching. Don’t forget your glutes 🍑 .
TL;DR
Battery development is hard, but the biggest challenge today is hype.
In our parade analogy for the grid, energy storage would just be a giant spring.
We have a long list of demands for any giant spring, which is what makes energy storage hard. Being mediocre at all is ok. Being great at some but failing miserably in others will not be ok.
To develop a first principles understanding, think of fuel as the anode in a battery and vice versa.
200+ years of battery development has produced only a handful of commercially viable rechargeable designs. Rechargeability is the hard part.
Li-ion batteries are an incredible innovation, and we are now ready to deep dive into them in the next post.
Thanks for sticking with me to the end, and I look forward to continuing our jam-sesh in the next deep dive. If you want to show some big pic energy, please share with your nerd friends to subscribe:)
Thanks and much love,
Adrian
Fun fact borrowed from “The Interchange” Podcast - “Are Batteries at a Turning Point?”
Li, Y., & Lu, J. (2017). Metal–Air Batteries: Will They Be the Future Electrochemical Energy Storage Device of Choice? ACS Energy Letters, 2(6), 1370–1377. https://doi.org/10.1021/acsenergylett.7b00119