In the lab, Stanford makes hydrogen and graphite 10x more efficiently

At 1,000°C, a methane-pyrolysis reactor becomes an exercise in getting heat to its middle. Stanford Engineering researchers Henry Moise, Sebastian Moll and Matteo Cargnello have developed an internal-heating method that, in their Science study, produces hydrogen and solid graphite with a roughly 10-fold efficiency gain over conventional external heating.

Hydrogen is essential to major industrial and agricultural operations, but most hydrogen made today is produced through processes that release large amounts of CO2. Methane pyrolysis takes a different route: it breaks methane into hydrogen gas and solid carbon, rather than the gaseous CO2 released by conventional production. The process has been studied for years, but removing the carbon and supplying enough heat have kept it from moving beyond the lab.

The Stanford team tackled the heat problem by putting the burner inside the reactor. Rather than burning natural gas, the burner combusts a portion of the hydrogen produced by the reaction, generating the high heat needed to drive the process while emitting mostly water vapor. Henry Moise says the result is a roughly 10-fold efficiency gain: using the same amount of energy, one reactor can produce the same amount of material as 10 other reactors heated conventionally from the outside.

The researchers also found that the solid carbon was high-quality graphite, a material used in batteries, electrodes and other technologies. That could offer the U.S. a path toward domestic production of a material it currently imports. The limit is clear: the graphite is not yet pure enough for demanding battery applications, and the researchers say more work is needed.

Hydrogen is used in fertilizer, refining and methanol production; the method still needs to be scaled and verified for mainstream hydrogen production. For now, the work is a laboratory result, not a mainstream supply plant. Scaling up the reactors and verifying their use for mainstream hydrogen production are the next steps, with no deployment timetable given.