FeMoco, Nitrogenase and Fertilizer: Quantum Computing's Famous Test Case
Why fertilizer is on a quantum list
Modern farming leans on ammonia fertilizer, made industrially by the Haber-Bosch process, which needs high heat and pressure. Certain soil microbes do the same job at ordinary temperatures using an enzyme called nitrogenase. At its center sits a metal cluster, FeMoco. If we understood exactly how it works, we might design catalysts that need less energy. One trade source cited for this page puts Haber-Bosch at roughly 1 to 3 percent of global energy use, and sources disagree on the exact share, so treat it as "a big slice."
Why FeMoco is quantum-hard
FeMoco is a textbook strongly correlated system: several iron atoms and a molybdenum atom with electrons that refuse to behave independently. That makes accurate classical calculation difficult, which is the pattern explained in why chemistry is quantum-hard.
The 2016 landmark and the shrinking estimates
In 2016 Markus Reiher, Nathan Wiebe, Krysta Svore, Dave Wecker and Matthias Troyer posted a paper (later in PNAS, 2017) arguing that nitrogenase chemistry could run in reasonable time on a small, error-corrected quantum computer. Their estimates were large by modern standards. A summary of that work reports on the order of 10 to the 15th power T-gates for one serial approach.
Then the algorithm researchers went to work. A 2020 paper by Joonho Lee, Dominic Berry, Craig Gidney, Ryan Babbush and colleagues, published in PRX Quantum in 2021, used a technique called tensor hypercontraction and reported that FeMoco would need about four million physical qubits and under four days of runtime, assuming 1 microsecond cycle times and gate error rates no worse than 0.1 percent. One industry analysis says estimates for this problem fell by about five orders of magnitude between 2017 and 2025, with later work adding further speedups, and quotes the Lee et al. estimate as roughly 2,142 logical qubits. A 2025 paper on modular processors, meanwhile, cites estimates in the range of 1,137 logical qubits with far fewer gates than the 2016 approach. Notice that qubit counts and gate counts moved in different directions across methods, so the numbers are not directly comparable. That is itself a lesson: this is an active research area, and every figure comes with assumptions.
Where things stand in 2026
- No quantum computer has done FeMoco. The machines needed (thousands of logical qubits) do not exist. See error correction in 2026.
- Roadmaps point toward the right neighborhood. IBM targets a fault-tolerant system called Starling by 2029 (IBM's roadmap), and Quantinuum Ltd names a fault-tolerant system called Apollo for 2029. These are targets, not deliveries, and early fault-tolerant machines are expected to be smaller than a full FeMoco run needs.
- One analysis projects 2033 to 2036 for the first industrially relevant FeMoco simulations. That is a single author's projection.
The honest debate
Two cautions deserve space. First, classical methods may close the gap. A January 2026 preprint from Garnet Chan's group (arXiv 2601.04621) claims a classical calculation of the ground state energy of a standard FeMoco model to chemical accuracy. It covers the resting state of the enzyme, not the reaction pathway, and we have not checked whether it has been peer reviewed, so treat it as a serious claim that is still being examined. It shows classical researchers are actively attacking this exact benchmark. Second, even a perfect answer is not a fertilizer. One analysis says the business value is mainly mechanistic insight that must be translated into catalyst design, not a direct product path. And the same analysis says the quantum edge for one nitrogen-fixation calculation is mainly accuracy rather than speed.
Claims you may see online, such as quantum computers already finding a cheaper fertilizer catalyst, were not confirmed in this research. The 2026 examples we found were podcast summaries that contradicted each other and cited no paper. Be skeptical of any specific percentage savings until a peer-reviewed source appears.
Why it is still an exciting target
A benchmark that shrinks by orders of magnitude through clever math, before the hardware even arrives, is a good sign for the field. Think of it as the Apollo program of computational chemistry: a clear, concrete moon. See also honest timelines.
Sources and further reading
- Reiher et al., Elucidating Reaction Mechanisms on Quantum Computers (arXiv 1605.03590)
- Lee et al., tensor hypercontraction for chemistry (arXiv 2011.03494)
- Efficient algorithms for quantum chemistry on modular quantum processors (arXiv 2506.13332)
- Zhai et al., classical simulation of the FeMo-cofactor model (arXiv 2601.04621, January 2026)
- PostQuantum: quantum chemistry utility map
- IBM quantum roadmap
Reported as of 2026-10-09. Resource estimates and timelines change as algorithms improve, and company statements are not independent verification. Check the primary papers. Nothing here is financial advice, and the QNT memecoin is independent of Quantinuum Ltd and of every company named on this page.
Frequently asked questions
What is FeMoco?
The iron-molybdenum cofactor at the active site of nitrogenase, the enzyme that converts nitrogen from air into ammonia in nature.
Has a quantum computer simulated FeMoco?
No. Reported resource estimates need thousands of logical qubits, which no machine has today.
How big a machine does FeMoco need?
It depends on the algorithm. One 2021 paper reports about four million physical qubits and under four days. Other methods trade qubits against gates, so figures are not directly comparable.
Will quantum computing cut fertilizer energy use?
It might help scientists understand catalysts, but no verified result exists yet and the path from insight to product is long.
Keep reading
- Why Chemistry Is the Problem Quantum Computers Were Made For
A plain English look at why molecules are hard for ordinary computers, which chemistry problems are truly quantum-hard, and why most of chemistry stays classical. - Quantum Chemistry Timelines: An Honest, Optimistic Guide
When might quantum computers do useful chemistry? A grounded look at published projections from the early 2030s to the 2040s and what could change them. - Quantum Error Correction Explained
Qubits are fragile, so quantum computers need error correction. Learn how logical qubits are built and why this is the key challenge. - IBM's Quantum Roadmap: From Nighthawk to Starling in 2029
IBM's published plan for fault tolerant quantum computing: processors per year, the Starling system, and what to watch in 2026 and 2027.
All Quantum computing guides | Back to top | Search the site
Main pages: Quantum computing explained | Quantum and crypto | Companies | Quantum news | Glossary