Beyond Lithium-Ion: New Battery Chemistries

a very brief introduction

If we’re going to replace most electricity and a big chunk of all energy use on Earth with fluctuating renewables like solar and wind, we’ll need a lot of batteries.

Like, 30-50 times as much battery capacity as exists today by the end of the decade, mostly in electric vehicle batteries and electrical grid storage.

A big question is, what kind of batteries will those be?

Will they be lithium-ion batteries like today’s electric-vehicle batteries, or will some new battery chemistry become dominant?

The main considerations for this question, to a first approximation, are energy density and cost.

Energy density (or specific energy), measured in watt-hours per kilogram, is particularly important for electric cars and for the future potential of electrifying trucks and aircraft.

A portable battery in an electric vehicle needs to be lightweight while still enabling the vehicle to run for a long time between charges.

Batteries with energy densities much lower than today’s lithium-ion batteries are basically non-starters for electric vehicles. Batteries with higher energy densities might gain market share, and will be absolutely necessary if we’re going to electrify long-haul trucks (which need greater driving range) or aircraft (which need even greater flying range and lighter batteries.)

One thing worth remembering is that the theoretical energy density of a battery material is a maximum, and the actual energy density is usually lower. For the purposes of this post, I only consider the “energy density” of a battery to be known when someone has built a cell and measured its output. The real-world energy density of a battery contains the weight of the whole cell — not just the cathode, anode, and electrode, but any packaging or other components. So we should expect energy densities measured in the lab to be overestimates, but also expect energy densities to increase somewhat as batteries are manufactured at scale and the design is tweaked.

Cost per kilowatt-hour is important for all battery use cases. Energy storage is basically a commodity; when it comes to grid-scale storage (where portability is not a concern), I expect cost to be the dominant consideration in which battery chemistries get adopted.

A given battery chemistry’s cost is lower-bounded by the cost of materials. We can expect manufacturing costs of any new type of battery to decline according to a power law as production scales up, in a Wright’s curve effect. The exact shape of that learning curve is hard to predict, though, so it’s a good first heuristic that batteries made from cheaper raw materials will have a lower long-run price (even if they start out expensive because they’re an early-stage technology.)

Batteries that don’t use lithium, for instance, are appealing especially in the context of current lithium shortages which have driven up the price of lithium ion batteries in 2022 after 3 decades of steadily falling prices.

Other considerations that might be relevant for battery chemistry adoption include:

• safety (lithium ion batteries are toxic and prone to catching fire)

• ethics of raw material extraction (leading lithium-ion chemistries include cobalt, which is often mined using child labor)

• cycle life (how many times a battery can be charged before needing to be replaced)

• charging speed (consumers may balk at overly slow-charging car batteries)

I’ve made a spreadsheet of various types of battery chemistry and their properties, which you can view here.

The High Energy Density Options

These are the battery chemistries that have sufficient energy density to be practical in electric vehicles.

The most common lithium-ion battery types used in electric vehicles today are NCA (lithium cobalt aluminum, used in Teslas), NMCO (lithium nickel cobalt oxide), LCO (lithium cobalt oxide), LFP (lithium iron phosphate), and LMO (lithium manganese oxide). These have energy densities around 150-220 Wh/kg.

Lithium-metal and lithium-air batteries (which have a pure lithium anode as opposed to the lithium compounds in the cathode of a lithium-ion battery) can have much higher energy densities, but they tend to have shorter cycle life and are prone to catching fire.

Rechargeable zinc-air batteries, similarly, have higher energy densities than lithium-ion, and have the added advantage of not including lithium (which is costly, toxic, and highly reactive). In fact, startup Zinc8 is claiming to offer prices as low as $60/kWh (less than half of the cost per kWh of lithium-ion batteries).

Lithium-silicon and lithium-sulfur batteries, which have been proposed for use in electric aircraft because of their high energy density, are also limited by low cycle life.

Lithium-ceramic batteries, or solid-state batteries, use a solid ceramic electrolyte rather than the liquid or gel electrolyte in other lithium-ion batteries. This increases the energy density and reduces flammability. It’s possible to combine a solid electrolyte with a variety of different anode and cathode materials — for example, QuantumScape uses a pure lithium anode and an NMCO cathode along with its flexible ceramic electrolyte.

Molten salt batteries use molten salts as an electrolyte and have high energy density and low cost. But the high temperatures they need to operate at make them impractical for electric vehicles; they’re more often considered for stationary energy storage.

Finally, sodium-ion, aluminum-ion, and zinc-ion batteries have energy densities comparable to some lithium-ion batteries used in electric vehicles, while being less flammable, cheaper, and in the case of aluminum-ion, much faster-charging. Sodium-ion batteries are the closest to mass-production of the three — battery giant CATL has started making them.

The Low Cost Options

The cheapest available battery chemistry is iron-air, which uses the oxidation (rusting) of iron to power the battery. The energy density is too low to be practical in electric vehicles, but in large grid-scale storage installations, iron-air batteries have potential. They are also safe and have a very long cycle life.

Molten-salt batteries are similarly inexpensive, but are limited by shorter cycle life and the need for high temperatures.

Sodium-ion and zinc-air batteries, which also have high energy densities, can be quite inexpensive due to being made of common raw materials.

I would tentatively predict that large-scale grid storage facilities over the next decade will use these low-cost newer technologies over lithium-ion batteries.

The first iron-air grid scale storage plan, a collaboration between Form Energy and Georgia Power, was announced this year.