GROWING METAL

What if some of tomorrow’s nickel came from fields, not deeper holes?

A Fieldbook for founders and investors | 10 September 2026
An opportunity narrative built from an independently audited research dossier.

I am excited by a strange biological fact: certain plants pull nickel from rock-rich soil and store it in their leaves. The science is real. The company is not obvious.

01 · The possibility

WHAT IF A FIELD COULD MAKE METAL?

Nickel is normally won by moving rock: dig it, crush it, concentrate it, heat it, and refine it. But a small group of plants can do the first act of concentration themselves. Their roots take up nickel dissolved in soil water. Their tissues tolerate what would poison most plants. At harvest, some of the metal that was scattered through the ground is now gathered into biomass we can cut and carry.

This is called phytomining, or more precisely nickel agromining when the crop is grown deliberately to produce nickel. The possibility is bigger than a novelty crop. In the right places, a field might produce a metal-bearing harvest while keeping land under vegetation and building an entirely different supply chain from mine to material.

I can imagine farms supplying metal processors, degraded nickel-rich land gaining a productive use, and plant breeders improving a new kind of crop over decades. But imagination is the easy part. The first company must survive a much less romantic equation: how many kilograms of saleable nickel leave one hectare, what does it cost to make them, and who will buy the intermediate material?

02 · The biological machine

ROOTS DO THE SORTING

The plants we care about are hyperaccumulators. That long word means they can collect unusually high concentrations of a metal in their living tissue without dying. Nickel is not sitting underground as convenient beads. Weathering releases a small share into soil water; roots take up some of those nickel ions; the plant moves them upward and stores much of them in leaves.

That mechanism creates three separate numbers. Biomass is the dry mass of the harvested crop. Nickel concentration is the share of that dry crop that is nickel. Multiply them and we get contained nickel:

dry harvested biomass x whole-crop nickel concentration = contained nickel

Contained nickel is still not product. Harvesting, drying, burning or other thermal treatment, chemical processing, and product purification can each lose material. A leaf assay is not a whole-field yield. Nickel in ash is not recovered nickel. Recovered nickel is not necessarily material a buyer will pay for.

Keeping those denominators separate is the first discipline of this industry.

03 · What the field has proved

REAL BIOLOGY, SMALL PLOTS

The cleanest long-running field evidence I found comes from Albania. Researchers grew the nickel hyperaccumulator Alyssum murale for five years. Their best reported treatment produced 9.0 tonnes of dry biomass per hectare and 105 kilograms of contained nickel per hectare when the small-plot result was scaled to a hectare. Bani et al., 2015

That is meaningful. It shows a cultivated crop can repeatedly remove measurable nickel under field conditions. It is also easy to oversell. The experimental plots were 18 square metres, not commercial hectares. The 105 kilograms were nickel contained in harvested biomass, not recovered saleable product. The result does not tell us what a large farm, a processor, or a buyer will deliver together.

A 2026 specialist review reports European field results around 105 to 106.3 kilograms of nickel per hectare, while noting one improved cultivar in Oregon reached a reported 400 kilograms through the combination of 20 tonnes of biomass and 2% nickel. The same review warns that startup claims above roughly 500 kilograms per hectare should be treated with doubt until independently demonstrated. van der Ent et al., 2026

My conclusion is neither “it works” nor “it does not.” The biology has crossed the proof line. The audited public evidence in this Fieldbook does not independently establish repeatable commercial performance, even though companies report commercial activity.

04 · The part hidden by the plant

THE HARVEST IS NOT THE PRODUCT

Once the crop is cut, phytomining stops looking like unusual agriculture and starts looking like a difficult materials business.

The wet crop must be dried or handled quickly. Thermal treatment can shrink a large volume of biomass into a much smaller mineral-rich ash, sometimes called a bio-ore. That ash is an intermediate, not pure nickel. It can still contain other metals and minerals. Turning it into a consistent nickel salt or another buyer-ready product requires processing, quality control, waste handling, and a customer specification.

There is no universal rule that every conventional processor will reject it. A specialist 2026 assessment says some electric-arc smelters may accept biomass and that a stainless-steel route is emerging. It also argues that building a dedicated smelter, likely costing tens of millions of dollars, is not a sensible starting point. Processor fit is site- and material-specific, and public evidence on commercial terms remains thin. van der Ent et al., 2026

This changes the founder question. I am not building a clever farm. I am qualifying a new feedstock across an entire chain: field, harvest, concentration, processing, product, and buyer.

05 · The arithmetic nobody should skip

THE MIDDLE CASE LOSES MONEY

Here is a deliberately plain one-hectare scenario. These are author assumptions, not observed commercial economics.

At those assumptions, break-even requires about 131 kilograms of recovered nickel, or 154 kilograms contained in the harvested crop, per hectare. With nine tonnes of dry biomass and 85% whole-chain recovery, the crop would need to average about 1.71% nickel across everything harvested.

The model includes no remediation fee, carbon credit, “green nickel” premium, or by-product revenue. Those may exist in a particular contract. Until a payer signs one, they are upside cases, not the base business.

This is why a beautiful leaf assay can coexist with a bad company.

06 · The land can push back

A CROP STILL NEEDS A FARM

Hyperaccumulators tolerate nickel. They do not escape ordinary agriculture. They still need water, nutrients, workable soil, labour, harvest equipment, and seasons.

A two-year field experiment in Sabah, Malaysia, found negative calcium, phosphorus, and potassium budgets without fertilisation, while nickel in fallen litter cycled slowly back into the soil. That does not prove every phytomining field will follow a universal depletion timetable. It does prove that nickel yield cannot be managed alone: nutrient replacement and soil health belong in the operating model. Tisserand et al., 2024

The strongest bear case is therefore ordinary and powerful. The plant works, but acreage scales slowly. Inputs recur. Harvest quality varies. The processor is too far away or refuses the material. Payability is low. Each season takes a year to teach us something. A landowner likes the story but has no budget for it.

If those conditions hold, better genetics will not rescue the business quickly enough.

07 · Who is building

THE RACE HAS STARTED, QUIETLY

Genomines announced a $45 million Series A in September 2025 to develop plant-based nickel extraction. The company says its breeding and biotechnology have doubled yields, can lower costs by 40-50%, and can produce battery-grade material. Those are company claims, not independent field results, but the financing shows serious capital is entering the category. Genomines announcement

The audited public evidence is better at describing a research and practitioner ecosystem than a mature competitive market. The specialist literature spans plant breeding, agronomy, field projects, thermal treatment, chemical recovery, and possible stainless-steel offtake. That tells me the industry has several possible control points: genetics, project development, biomass logistics, processing recipes, and buyer qualification. It does not tell me which one already supports a durable venture-scale business.

I do not see a monopoly forming. I see an infant industry whose commercial map is still poorly measured.

The dangerous instinct is to claim all of them at once.

08 · The smallest company worth testing

SELL THE QUALIFICATION, NOT THE DREAM

My proposed wedge is a paid field-to-buyer qualification service for a funded agromining developer or land operator on previously cultivated, nickel-rich land in the Balkans.

Before anyone plants, the project must verify land access and permissions, choose a crop that is locally appropriate, assess nearby habitat and the risk of an introduced plant escaping, and reject any site that depends on converting habitat or accepting unresolved mining-tailings liabilities. Previously cultivated land is a filter, not a permission slip.

The service would answer one expensive question: can this specific field, crop, process route, and buyer specification form a chain worth scaling? It would combine a local farming and research partner, whole-harvest sampling, independent assays, a documented mass balance, processor sample tests, and a buyer-ready data package.

This is an unvalidated hypothesis, not evidence that a service market exists. The first kill test is paid demand. In the first three months I would interview ten landowners or developers and three processors, but conversation is not success. Success is one customer paying for a qualification pilot, plus a processor providing written sample requirements and a conditional route to acceptance and pricing.

If nobody pays, or no processor will define the path to acceptance, I stop. I do not hide the failure inside a larger “platform” story.

09 · The 24-month attack

EARN ONE COMPLETE MASS BALANCE

In months zero to three, I would identify one field, one crop and agronomy partner, one processor route, and one potential buyer. Before planting, I would document land rights and permissions, habitat and introduced-species checks, the crop’s local suitability, and the absence of unresolved tailings liabilities. Any required habitat conversion is a stop. I would also agree in writing on sampling, acceptable impurities, product form, and who pays for what.

In months three to nine, timed to the actual growing season, I would establish the pilot and measure inputs, survival, dry biomass, and nickel concentration across the whole harvested crop. A rough first-year planning envelope of $150,000 to $300,000 is an author assumption for designing the test, not a vendor quote or industry benchmark.

In months nine to eighteen, I would harvest, process a representative batch, and reconcile every denominator: wet biomass, dry biomass, contained nickel, ash, recovered nickel, final product, and buyer payability. A high ash-leaching percentage would not count as high whole-chain recovery.

In months eighteen to twenty-four, I would repeat the crop cycle if the season allows, or extend the programme rather than pretending annual repeatability. The go decision requires a paid customer, a workable processor route, independently measured yield and recovery, and a contribution model that can plausibly turn positive without imaginary revenue.

The company earns the right to own more of the chain only after that chain works once.

10 · The founder decision

BUILD THE BRIDGE, NOT THE MYTH

I would work on phytomining because the underlying phenomenon is wonderful, the supply chain could become genuinely different, and the industry is early enough for a small team to discover an important control point.

I would not start by promising a green nickel mine made of plants. The best public field evidence available to this audit is still small, processor economics are opaque, and this proposed project does not yet have an independently verified commercial mass balance. Reported commercial activity elsewhere does not remove the need to prove one here. A serious founder should make those weaknesses the product roadmap.

I would pursue the qualification wedge only if a real customer pays to reduce this uncertainty and a processor engages before the crop goes into the ground. I would stop if whole-field yield, whole-chain recovery, nutrient needs, logistics, or buyer payability keep the economics below break-even.

Two ideas remain interesting research hypotheses: using engineered root-associated fungi to extend metal collection, and using low-voltage electric fields to move nickel ions toward roots. Neither belongs in the base case; both need evidence that they improve net recovered nickel without unacceptable ecological or energy costs.

The opportunity is not that plants magically make metal. It is that biology may become one useful machine inside a new metal supply chain. The founder’s job is to prove the rest.

Research receipts

• Bani et al., five-year Albania field experiment
• van der Ent et al., 2026 specialist review
• Tisserand et al., Sabah nutrient and litter study
• Kidd et al., agromining chain and processing review
• Genomines Series A announcement