Why recycling entered the sourcing conversation
Recycled content used to arrive as a question from a sustainability team and leave without consequence. It now arrives from three directions at once, and at least one of them carries a compliance deadline.
The supply argument is the one that should interest a category manager most. End-of-life magnets sit inside equipment already in your region — hard drives, motors, speakers, wind generators — so recovering them creates material whose origin is not subject to the same export-licensing exposure discussed in the supply risk guide. That is a structurally different kind of diversification from qualifying another supplier in the same region.
The energy argument is well supported: life cycle assessments of magnet-to-magnet recycling, where end-of-life magnets are broken down to powder and re-sintered, have reported energy reductions of more than 45% against production from primary material. That is a genuine footprint difference rather than a rounding.
The processing routes below are proven and several are running commercially. What does not yet exist is volume. Dedicated recycling plants currently operate at capacities measured in the low hundreds of tonnes per year, against a global magnet market measured in hundreds of thousands. Any sourcing plan that assumes recycled material is broadly available at scale today will not survive contact with a quotation.
The three routes, and what each produces
Recycling NdFeB is not one process. The routes differ in what feedstock they tolerate, what they output, and how much of the magnet's value they preserve — and those differences determine which one is relevant to you.
| Route | Feedstock needed | Output | Value retained | Status |
|---|---|---|---|---|
| Direct / magnet-to-magnet (hydrogen) | Clean, identified, low-oxidation magnets; ideally disassembled | NdFeB alloy powder, ready to re-press and sinter | Highest — alloy chemistry is preserved | Commercial at small scale |
| Hydrometallurgical | Tolerant — mixed, coated, oxidised, swarf | Separated rare earth oxides or salts | Elements only; alloying and processing must be redone | Established; the workhorse route |
| Pyrometallurgical | Tolerant of mixed feed | Master alloy or metal | Moderate; energy-intensive | Limited commercial use for magnets |
How the hydrogen route works
Hydrogen processing of magnet scrap exploits a straightforward property of sintered NdFeB: exposed to hydrogen at close to room temperature and atmospheric pressure, the neodymium-rich grain boundary phase absorbs hydrogen and expands, and the magnet breaks apart into a friable, demagnetised powder without any additional energy input. The powder separates readily from nickel coatings and from the surrounding assembly, which is why the route can pull magnets out of hard drives and motors without extensive disassembly.
The technique was developed at the University of Birmingham and is the basis of the plants now operating in the UK and Germany, with further capacity in development in the United States. Reported property recovery can exceed 90% where the input is clean, unoxidised and properly separated — and that qualification is doing a great deal of work, as the next section explains.
If the question is recycled content for disclosure, the hydrometallurgical route is more likely to be the source, because the recovered oxide re-enters the ordinary alloy supply chain and gets used everywhere. If the question is a physically traceable recycled magnet, the direct route is what you are asking for, and availability is far more limited. These are different products and should be specified differently.
The feedstock problem
Recycling capacity is easier to build than feedstock is to collect. This is the binding constraint on the whole sector, and it explains why announced capacity and actual output diverge.
Pre-consumer versus post-consumer
| Pre-consumer | Post-consumer | |
|---|---|---|
| Source | Machining swarf, sintering rejects, off-spec production | End-of-life motors, drives, speakers, generators |
| Volume | Substantial — machining loss on sintered magnets is significant | Growing but dispersed across the installed base |
| Composition | Known exactly; single grade per batch | Mixed and often unrecorded |
| Condition | Swarf oxidises rapidly; rejects are clean | Variable, frequently corroded or thermally aged |
| Collection | Straightforward — it is already inside a magnet factory | The hard part; requires reverse logistics and disassembly |
| Regulatory credit | Not currently counted by the EU rules; proposed amendments would change that | What the current EU disclosure obligation is written against |
That final row is a live policy question rather than a settled one. The EU disclosure obligation as enacted concerns material recovered from post-consumer waste. Amendments proposed by the Commission in early 2026, on which the Council adopted a position in March 2026, would take pre-consumer waste into account as well and would extend magnet circularity provisions more broadly. The file sits with the European Parliament and the outcome is not yet fixed — treat any supplier claim that pre-consumer swarf already counts toward the obligation as unverified.
Why post-consumer collection is genuinely hard
Magnets are small, bonded into assemblies, and rarely labelled. Recovering them requires knowing which products contain them, getting those products into a recycling stream, and separating the magnets without destroying them — and shredding, the default in electronics recycling, disperses magnet material into a mixed fraction from which direct recycling is no longer possible. This is why the EU proposals include labelling requirements identifying the presence and composition of magnets in more product categories: the labelling exists to make the recycling possible.
Magnet-bearing waste is starting to be treated as a strategic material rather than as waste. The European Commission has signalled restrictions on exports of recyclable rare earth waste, and related battery scrap classifications were scheduled to tighten during 2026. If your reverse logistics currently sends magnet-bearing scrap across borders, that route may not stay open. Verify current status before relying on it.
What recycled content does to the magnet
This is the section engineers care about and commercial teams tend to skip. Recycled input is not neutral with respect to magnetic performance, and the constraint has a specific mechanism.
Oxygen is the limiting variable
Rare earth metals oxidise readily, and every handling step between the end-of-life magnet and the new sintered part adds oxygen. Oxygen consumes neodymium into oxide phases that contribute nothing magnetically and disrupt the grain boundary structure that coercivity depends on. Recycled powder consequently arrives with a higher oxygen content than virgin powder, and that shows up as reduced coercivity and, at the margin, reduced remanence.
Producers compensate by blending recycled powder with virgin material, by adding rare earth metal to restore the grain boundary phase, and by grain boundary diffusion. All of these work, and all of them consume some of the cost and footprint advantage that motivated recycling in the first place.
What this means for grade selection
The practical guidance is to specify recycled content on parts running standard grades with reasonable coercivity margin, and to leave your high-temperature and safety-critical parts on virgin material until a supplier has demonstrated otherwise on your specific part. Check the working point using the demagnetization and permeance calculator before assuming a grade substitution is harmless — the margin you have determines how much coercivity variation you can absorb.
Recycled-content material from a new process route is a material change, and it deserves the full first-article and PPAP treatment rather than a paper substitution. Include a demagnetization curve at temperature, not just room-temperature remanence, because that is where the difference appears.
The regulatory picture, as it stands
The obligations below are current as of mid-2026 and several are actively moving. Verify status before building a compliance programme on them, and treat this as orientation rather than as legal advice.
| Obligation | Substance | Timing |
|---|---|---|
| Recycled content disclosure | Publish the share of Nd, Dy, Pr, Tb, B, Sm, Ni and Co recovered from post-consumer waste in magnets, on a free-access website | By 24 May 2027, or two years after the enabling delegated act enters into force, whichever is later |
| Threshold | Applies where total covered magnet weight in the product exceeds 0.2 kg | As above |
| Calculation rules | Delegated act establishing how the share is calculated and verified | Was due 24 May 2026; confirm current status |
| Minimum shares | Commission may set minimum recycled-content shares by delegated act, with transitional periods | After the calculation act; in any event by 31 December 2031 |
| Magnet labelling | Identify presence and composition of magnets to support removal and recycling; proposals extend this to more product categories | Implementing act on label format expected; obligations follow two years later |
| Proposed amendments | Count pre-consumer waste, strengthen magnet recyclability provisions, extend product scope | Council position adopted March 2026; Parliament stage ongoing |
Two structural points matter more than the individual dates. First, the obligation lands on whoever places the finished product on the EU market — which may well be your customer rather than you, but the data has to come up the chain from the magnet supplier, so you sit in the middle of it either way. Second, the operative details depend on secondary legislation that in several cases is not yet finalised. Building a rigid system against assumed calculation rules is a way to build it twice.
Whatever the final rules say, they will require you to know which parts contain magnets, what those magnets weigh, and what their composition is. That inventory is useful regardless and is the slow part. Build it now; wire it into a reporting system once the calculation methodology is settled.
Writing recycled content into a specification
A recycled-content requirement that says only "shall contain recycled material" buys nothing. It is unverifiable, and the supplier will meet it in whatever way is cheapest — which may be a mass-balance allocation against material that never went near your part.
Decide which claim you are buying
| Model | What it means | Verifiable by | Use when |
|---|---|---|---|
| Physically segregated | The material in your part is recycled material, kept separate throughout | Batch records traceable to feedstock | You need a defensible physical claim; expect limited availability and a premium |
| Mass balance | Recycled input is accounted across a production system and allocated to output | Audited accounting system, certified scheme | Reporting a share across a portfolio; widely used and legitimate if audited |
| Attributed / book-and-claim | The claim is decoupled from physical flow entirely | Registry only | Rarely appropriate for a materials requirement; check whether it is even accepted |
Mass balance is the model most likely to be practical, and it is not a lesser claim provided the accounting is audited to a recognised scheme. What matters is that you know which model you bought and can say so, because a regulator or customer asking about recycled content will ask exactly that question.
What to demand as evidence
The percentage basis catches people out. A magnet described as containing 30% recycled content may mean 30% of the rare earth elements or 30% of total magnet mass — and since rare earths are roughly a third of NdFeB by mass, those two readings differ by a factor of about three. Fix the basis in the specification.
Whatever you agree, require it to appear on the certificate of conformance for each lot rather than in a one-time letter. A claim that lives in a sales document and not in the lot certification will not survive an audit and will not be there when your customer asks for it in three years.
What is realistically available today
An honest summary, so that a strategy built on this does not overpromise.
The reasonable position for most buyers in 2026 is to treat recycled content as a qualification project rather than a sourcing switch. Identify two or three parts running standard grades with comfortable coercivity margin and undemanding thermal duty. Qualify recycled-content material on those properly. Build the magnet inventory data your future disclosure obligation will need. And do not commit a recycled-content percentage to a customer or a regulator before you have a supplier who can actually deliver against it at your volume.
It is also worth being clear about what recycling does and does not solve. It genuinely diversifies origin and genuinely reduces embodied energy. It does not, at current scale, materially reduce your exposure to primary supply disruption, and it does not remove the need for the ordinary defences — a qualified second source, sized inventory, and design flexibility on grade. Those remain the load-bearing parts of a magnet supply strategy, and the rare-earth-free and reduced-dysprosium options guide covers the design-side lever that sits alongside them.
The EU legislative file amending the Critical Raw Materials Act was in progress through 2026 and the delegated acts governing calculation, verification and any minimum shares were still being finalised. Recycling capacity is also expanding from a small base, so availability statements here date quickly. Treat this page as a quarterly-review item and confirm current requirements with your own compliance advisors before acting on them.