Several families of molecules can store energy for weeks, months, or even years, each with different trade-offs in density, cost, and practicality. The simplest is hydrogen gas itself, but ammonia, methanol, synthetic methane, and more exotic photoswitchable compounds all serve as chemical vessels for long-term energy. The question is less “which molecule works” and more which molecule fits a given storage duration, scale, and end use, because no single chemistry dominates every scenario.
Hydrogen, the Simplest Starting Point
Hydrogen sits at the center of nearly every molecular energy-storage conversation. Water electrolysis powered by wind or solar splits H₂O into hydrogen and oxygen, and that hydrogen can later be burned or fed into a fuel cell to regenerate electricity. The appeal is straightforward: hydrogen carries a lot of energy per kilogram, produces only water when consumed, and can be made from nothing more than water and renewable electricity. Interest in green hydrogen production through electrolysis has grown sharply as countries pursue large-scale decarbonization of power grids, industry, and transport.1Elsevier / Energy Reports. An overview of water electrolysis technologies for green hydrogen production
The catch is that hydrogen is a gas at ordinary temperatures and pressures, so storing large quantities of it takes serious engineering. Above-ground compressed tanks work for short cycles and have the highest round-trip storage efficiency, around 92%, but their costs climb steeply for seasonal storage. Projections show the levelized cost of above-ground compressed storage jumping from roughly $0.33 per kilogram for daily cycling to about $25 per kilogram for a four-month cycle.2Energy Conversion and Management. Projecting the levelized cost of large scale hydrogen storage for stationary applications That enormous gap is the economic reason researchers keep looking for alternative molecules that are easier and cheaper to store over long periods.
Underground Salt Caverns for Bulk Hydrogen
One way to tame hydrogen’s storage cost is to go underground. Salt caverns, massive voids dissolved out of geological salt formations, can hold hydrogen at high pressure for months. They already store natural gas in many countries, and recent research suggests hydrogen behaves well in the same structures. A study of the Rüdersdorf salt cavern in Germany found no fundamental barriers to hydrogen storage there: the cavern maintained its structural integrity, and oxygen and hydrocarbon contamination stayed negligible after storage. The main impurities that crept in were halogenated compounds and water, which would need to be scrubbed out before the hydrogen could be used in a fuel cell.3International Journal of Hydrogen Energy. Assessment of the feasibility of hydrogen storage in salt caverns: Implications for purity and contamination pathways for fuel cell mobility
Separate laboratory tests reinforce those results. When halite and anhydrite, the minerals that line salt cavern walls, were exposed to hydrogen at high pressure and temperature for 90 days, essentially nothing happened. No meaningful changes in mineral composition or pore structure were detected, confirming that the rock itself does not react with hydrogen on the timescales relevant to seasonal storage.4Fuel. Geochemical reactions and pore structure analysis of anhydrite/gypsum/halite bearing reservoirs relevant to subsurface hydrogen storage in salt caverns Salt cavern storage also posts the lowest levelized cost among the major options, at roughly $0.14 per kilogram for daily cycling.2Energy Conversion and Management. Projecting the levelized cost of large scale hydrogen storage for stationary applications The limitation is geographic: you need suitable salt geology, and not every region has it.
Solid-State Hydrogen Storage
Instead of compressing hydrogen into a tank or a cavern, you can lock it into a solid material. Metal hydrides and complex hydrides form chemical bonds with hydrogen, absorbing it into their crystal structure at modest pressures and releasing it when heated.5PubMed Central. Solid-state hydrogen storage materials The advantage is density: a metal hydride bed can hold more hydrogen per liter than a compressed-gas tank, and there is no risk of a pressurized gas leak. The disadvantages include weight, the energy needed to heat the material for hydrogen release, and the fact that many promising hydrides are still too slow at absorbing and releasing hydrogen for practical cycling. This area remains an active research frontier rather than a commercially mature storage method, but for applications where safety and compactness matter more than speed, solid-state approaches are compelling.
Ammonia as a Chemical Battery
Ammonia, NH₃, is one of the most widely discussed alternatives to storing raw hydrogen. The idea is to combine hydrogen with nitrogen from the air using the well-established Haber-Bosch process, creating a liquid that is far easier to ship and store than compressed gas. Ammonia is already produced and transported on a massive industrial scale, so the logistics infrastructure largely exists. Its volumetric energy density is attractive, at about 3,530 watt-hours per liter when liquefied at moderate conditions.6Next Energy. Round-trip efficiencies of green ammonia and green hydrogen
The trouble with ammonia is getting the energy back out. Converting ammonia back into electricity involves either cracking it back into hydrogen and feeding a fuel cell, or burning it in a turbine. Both routes lose a lot of energy. Round-trip efficiencies for conventional green ammonia power cycles range from about 23% using a steam turbine up to roughly 34% with a gas turbine combined cycle, and recent improvements in electrolysis push the best-case scenario to around 42%.6Next Energy. Round-trip efficiencies of green ammonia and green hydrogen That is well below what you would get from compressed hydrogen or lithium-ion batteries. A more ambitious design, coupling ammonia synthesis and decomposition with a pressurized reversible solid-oxide fuel cell and a steam power cycle, has been modeled at up to 72% round-trip efficiency for a 100-megawatt system, though that figure comes from an idealized process model rather than a built plant.7AIChE Journal. Conceptual design of ammonia‐based energy storage system: System design and time‐invariant performance
The levelized cost of storing hydrogen in the form of ammonia is also the highest among major options, at roughly $3.51 per kilogram of hydrogen equivalent, and projections suggest only a 20 to 25% cost reduction by 2050.2Energy Conversion and Management. Projecting the levelized cost of large scale hydrogen storage for stationary applications Ammonia still makes sense in contexts where you need to move energy across oceans, because a tanker full of liquid ammonia beats a tanker full of compressed hydrogen on volume and handling. But for stationary grid storage, the efficiency and cost penalties are steep.
Methanol and Synthetic Methane
Moving up in molecular complexity, methanol (CH₃OH) and synthetic methane (CH₄) offer something hydrogen and ammonia do not: compatibility with liquid-fuel infrastructure. Methanol is a liquid at room temperature with an energy density of about 15.8 megajoules per liter, which beats both liquid hydrogen (around 8.5 MJ/L) and liquid ammonia (about 12.7 MJ/L), though it falls short of gasoline at roughly 32 MJ/L.8Elsevier / ScienceDirect. A new transportation energy review: methanol catalytic synthesis from CO2 green hydrogenation You can pour it into a tank, truck it anywhere, and use it in engines, fuel cells, or chemical plants.
Green methanol is made by reacting green hydrogen with captured carbon dioxide. The CO₂ can come from industrial exhaust or direct air capture, making the cycle carbon-neutral in principle. Synthetic methane follows a similar logic: the Sabatier process combines CO₂ or CO with hydrogen to produce methane, which can then be injected directly into existing natural gas pipelines.9Energy Conversion and Management. Power-to-Gas: Process analysis and control strategies for dynamic catalytic methanation system That pipeline compatibility is a big practical advantage, because it means seasonal energy storage could piggyback on infrastructure that already exists across Europe and North America.
Both methanol and methane carry a round-trip efficiency penalty. Methanol storage efficiency sits at roughly 50%, and each conversion step from electricity to hydrogen to fuel and back again shaves off energy.2Energy Conversion and Management. Projecting the levelized cost of large scale hydrogen storage for stationary applications But when you are storing energy for months and need it in a form that can be shipped globally, the convenience of a liquid fuel can outweigh the efficiency loss.
Liquid Organic Hydrogen Carriers
Liquid organic hydrogen carriers, or LOHCs, are an elegant middle path. These are stable organic liquids that can be loaded with hydrogen through a chemical reaction (hydrogenation) and later unloaded (dehydrogenation) to release the hydrogen on demand. The carrier molecule itself is not consumed; it cycles back and forth between its hydrogen-rich and hydrogen-lean forms.
The leading candidates are mixtures based on benzyltoluene and dibenzyltoluene. They store about 6.2% hydrogen by weight, are nontoxic, have low melting points, and remain thermally stable across a wide temperature range.10International Journal of Hydrogen Energy. Compressed liquid density and thermodynamic modeling for the promising liquid organic hydrogen carrier Benzyltoluene/Dibenzyltoluene Because the loaded carrier looks and handles like diesel fuel, it can be pumped, stored in ordinary tanks, and shipped in conventional tankers with no special pressurization or cryogenic cooling. That makes LOHCs well suited for transporting renewable energy from sun-rich regions to energy-hungry ones. The trade-off is the heat needed for dehydrogenation, which reduces overall round-trip efficiency. If waste heat from an industrial process is available, the economics improve considerably.
Formic Acid as a Compact Hydrogen Vessel
Formic acid, the simplest carboxylic acid, is a less discussed but surprisingly capable hydrogen carrier. A liter of formic acid contains about 53 grams of accessible hydrogen, giving it an energy density of approximately 1.77 kilowatt-hours per liter.11PubMed Central. Highly Efficient Additive-Free Dehydrogenation of Neat Formic Acid It is a liquid at room temperature, relatively safe to handle, and decomposes into hydrogen and CO₂ when passed over the right catalyst.
Catalyst development has been a major focus. Researchers have demonstrated highly active iron-based catalysts that liberate hydrogen from formic acid efficiently, avoiding the need for expensive precious metals.12PubMed. Efficient dehydrogenation of formic acid using an iron catalyst Others have developed gold-palladium nanoparticles immobilized in porous frameworks that achieve complete conversion of formic acid to high-quality hydrogen at convenient temperatures.13PubMed. Synergistic catalysis of metal-organic framework-immobilized Au-Pd nanoparticles in dehydrogenation of formic acid for chemical hydrogen storage The CO₂ released during dehydrogenation can, in theory, be recaptured and rehydrogenated back to formic acid, closing the loop. Formic acid is unlikely to be the backbone of a national grid storage system, but for distributed or portable applications where liquid handling and moderate energy density matter, it is a serious contender.
Molecular Solar Thermal Fuels
Most of the molecules above store energy that originally came from electricity. Molecular solar thermal (MOST) fuels take a different approach: they absorb sunlight directly and store it as chemical strain energy within their bonds. When triggered, the strained molecule snaps back to its relaxed form and releases that stored energy as heat.
The most studied system is the norbornadiene-to-quadricyclane pair. Norbornadiene absorbs a photon and rearranges into quadricyclane, a highly strained molecule that can sit indefinitely at room temperature without reverting. When you want the energy back, a catalyst or a small heat pulse triggers the reverse reaction, and the stored heat pours out. The system has high energy storage density and the potential to hold energy for very long periods.14PubMed Central. Engineering of Norbornadiene/Quadricyclane Photoswitches for Molecular Solar Thermal Energy Storage Applications Recent work on triplet-sensitized switching has expanded the range of light these molecules can capture into the visible spectrum, improving their practical usefulness.15PubMed Central. Triplet-Sensitized Switching of High-Energy-Density Norbornadienes for Molecular Solar Thermal Energy Storage with Visible Light
Azobenzene-based systems represent another branch of the same idea. Carbon nanotubes functionalized with azobenzene molecules have demonstrated volumetric energy densities comparable to lithium-ion batteries while functioning as solar thermal fuels.16PubMed. Azobenzene-functionalized carbon nanotubes as high-energy density solar thermal fuels These materials are still firmly in the laboratory stage, but the concept is appealing for applications where you want heat storage without electrochemistry, such as heating buildings or running industrial processes that need moderate temperatures.
Redox-Active Organic Molecules for Flow Batteries
Flow batteries store energy in liquid electrolytes pumped through an electrochemical cell. Traditional versions use vanadium salts, but a growing body of research focuses on organic molecules that can accept and donate electrons reversibly. Quinones, viologens, aza-aromatics, iron coordination complexes, and nitroxide radicals have all been tested as redox-active species in aqueous organic flow batteries.17PubMed. Electrolyte Lifetime in Aqueous Organic Redox Flow Batteries: A Critical Review
The attraction of organic molecules is cost. Vanadium is expensive and mined in a handful of countries, whereas organic molecules can in principle be synthesized from abundant feedstocks. The challenge is longevity. Capacity fade in these systems tends to be time-based rather than cycle-based, meaning the electrolyte slowly degrades even when the battery is sitting idle. Degradation pathways include reactions where molecules rearrange, pair up, or undergo irreversible side reactions that produce species that can no longer store charge.18PubMed Central. Molecular Design and Redox Chemistries for Aqueous Organic Redox Flow Batteries (AORFBs) The best organic electrolytes now achieve fade rates below 0.02% per day, which sounds small but compounds over months of seasonal storage.17PubMed. Electrolyte Lifetime in Aqueous Organic Redox Flow Batteries: A Critical Review Solving that stability problem is the central engineering challenge, and recent molecular design strategies, including tweaking ring structures and functional groups to block the most common degradation routes, are making progress.
Iron-Air Batteries
Iron-air batteries take yet another molecular angle. The anode is iron, one of the cheapest and most abundant metals on Earth. During discharge, iron reacts with oxygen from the air; during charge, the reaction reverses and regenerates metallic iron. The chemistry is simple and the raw materials are dirt cheap. Iron-air systems have emerged as a potentially transformative long-duration storage option because of their ultra-low projected cost, inherent safety (iron is not flammable the way lithium can be), and high theoretical energy density.19Advanced Sustainable Systems. Sustainable Development of Iron–Air Batteries as Long‐Duration Energy Storage Systems: The Key to Improving the Redox Kinetics of Iron‐Based Anode The main obstacle is kinetics: the iron reactions at the anode tend to be sluggish, limiting how quickly the battery can charge and discharge. Improving those reaction rates without sacrificing the cost advantage is where most of the current research is focused.
Metal Oxides for Thermochemical Heat Storage
If the end goal is storing heat rather than electricity, metal oxide redox reactions offer some of the highest energy densities available. The concept is straightforward: heat a metal oxide until it releases oxygen (reduction), then store the reduced material. When you need the heat back, reintroduce air and the material re-oxidizes, releasing the stored thermal energy. Because the reactant is air, there is no need for expensive pressurized gas storage. These systems achieve higher energy storage densities than conventional sensible-heat approaches, where you simply heat up a solid or liquid and let it cool.20Advanced Energy and Sustainability Research. Recent Progress on Redox Materials for High‐Temperature Thermochemical Heat Storage
Closed thermochemical storage systems, where the released gas is captured and stored for later recombination, have been modeled with overall energy efficiencies around 50%.21Energy. Closed and open thermochemical energy storage: Energy- and exergy-based comparisons That is modest compared to electrochemical storage, but the advantage is the ability to store energy indefinitely without self-discharge. A reduced metal oxide sitting in a container does not slowly lose its charge the way a battery does. For concentrated solar power plants that need to dispatch heat after sunset or through cloudy weeks, metal oxide thermochemical storage is a natural fit.
Microbial Electrosynthesis and Bio-Inspired Routes
Biology already solves the long-term energy storage problem every day: plants convert CO₂ and sunlight into sugars and fats that remain chemically stable for years. Researchers are trying to borrow from that playbook. Microbial electrosynthesis uses living microbes on an electrode to convert CO₂ and electricity into useful chemicals. In one system, a mixed microbial community on a carbon-based cathode produced short-chain fatty acids, primarily acetate, by reducing CO₂ through the same metabolic pathway used by acetogenic bacteria in nature.22Accounts of Chemical Research. Microbiome for the Electrosynthesis of Chemicals from Carbon Dioxide Those fatty acids can serve as feedstocks for fuels or as energy carriers themselves.
Artificial photosynthesis research pursues a similar vision with synthetic materials rather than living cells, aiming to convert solar energy directly into storable chemical bonds using photoelectrochemical cells and engineered catalysts.23PubMed Central. Artificial Photosynthesis: Current Advancements and Future Prospects Both bio-inspired approaches are early-stage, but they hint at a future where energy storage and carbon recycling are part of the same process.
The Hydrogen Leakage Problem
Any conversation about molecular energy storage that centers on hydrogen should account for what happens when some of it escapes. Hydrogen is an indirect greenhouse gas: once in the atmosphere, it extends the lifetime of methane and promotes ozone formation. A multi-model assessment estimates hydrogen’s 100-year global warming potential at about 11.6, meaning each kilogram of leaked hydrogen has roughly the same warming effect as 11.6 kilograms of CO₂ over a century. The biggest contributors to that warming are changes in methane persistence (about 44% of the effect) and ozone production (about 38%).24Communications Earth & Environment. Effect of hydrogen leakage on the life cycle climate impacts of hydrogen supply chains
At a 2% leakage rate across the supply chain, the additional warming amounts to less than 0.5 kilograms of CO₂-equivalent per kilogram of hydrogen produced, which increases the overall greenhouse gas intensity by less than 15% for most production pathways. The carbon intensity of the electricity used to make the hydrogen still matters far more than the leakage itself.24Communications Earth & Environment. Effect of hydrogen leakage on the life cycle climate impacts of hydrogen supply chains Still, minimizing leakage is important, especially at scale, and this is one reason that converting hydrogen into a more stable carrier molecule before shipping it long distances can be environmentally as well as logistically preferable.
Compressed Carbon Dioxide Energy Storage
CO₂ itself is being explored not as a fuel but as a working fluid for energy storage. In a compressed carbon dioxide energy storage system, surplus electricity compresses CO₂ into a dense state; when power is needed, the CO₂ is expanded through a turbine to generate electricity. One analysis found that a system storing CO₂ as a high-pressure liquid on both sides could achieve a round-trip efficiency of about 56% with an energy storage density of roughly 16 kilowatt-hours per cubic meter.25Applied Thermal Engineering. Thermodynamic and economic analysis of compressed carbon dioxide energy storage systems based on different storage modes That is not spectacular, but the system uses an abundant, nonflammable, nontoxic molecule and does not require geological formations like salt caverns. It occupies a niche somewhere between compressed air storage and chemical storage, with the advantage that CO₂ is denser and more compressible than air, so the tanks can be smaller.
Why No Single Molecule Wins
If one molecule clearly outperformed all others, the research landscape would be far narrower than it is. The reality is that each storage scenario has different priorities. A grid operator smoothing out day-to-day solar variability cares most about round-trip efficiency, which favors compressed hydrogen or flow batteries. A shipping company moving renewable energy from Australia to Japan cares about volumetric density and handling safety, which favors ammonia or LOHCs. A concentrated solar power plant storing heat for nighttime dispatch needs thermochemical cycles, not electrochemistry. A remote village wanting portable fuel from a local solar array might benefit most from methanol or formic acid.
The molecules covered here span a range of technological readiness, from ammonia and methane, which can lean on existing global infrastructure, to norbornadiene photoswitches that are still being optimized in the lab. Over the next few decades, the winners will likely be determined less by pure chemistry than by economics, geography, and the infrastructure each region already has in the ground.