1. What a permanent magnet is (the short version)

• Performance is read from three quantities: Br (how strongly magnetization remains), HcJ (how much reverse field the magnet withstands) and (BH)max (a measure of the energy it can deliver outside itself). Br is set almost entirely by the main phase, whereas HcJ is set by the material’s internal microstructureSourced [Sources 3, 4, 10]
• Strong, but weak to heat: neodymium magnets are the strongest, but their coercivity falls sharply as temperature rises. Drive motors in hybrid vehicles are said to reach about 200 °C, and the coercivity needed at that temperature has been supplied by dysprosium and terbiumSourced [Sources 3, 4]
• The supply chain is long: mining, refining (separation and smelting) and alloy and magnet manufacture are each concentrated in different countries, and China’s export controls move through announcement, entry into force and suspension. At the date checked (10 October 2026), the expanded measures announced in October 2025 are suspended until 10 November 2026Sourced [Sources 15, 17, 24]

2. Reading performance — the demagnetization curve and Br, HcJ, (BH)max

A magnet’s performance is read from its demagnetization curve — the second quadrant of the magnetization curve, traced as a reverse field is applied to a magnetized magnet. The same curve defines the following quantitiesSourced [Source 10].

All Sourced (Arnold Magnetic Technologies technical note TN 1501 [Source 10]).

When a magnet is used in a motor or similar device, its working point is where the load line, set by the magnet’s shape and the magnetic circuit, meets the B curve. In a design that minimizes magnet volume, the working point sits at the (BH)max point (roughly H = Br/2), so at the operating temperature the intrinsic coercivity, expressed as a flux density (μ0HcJ), must exceed half of Br. Hirosawa et al. explain this condition as a reason for using heavy rare earthsSourced [Source 3].

3. Four families — sintered neodymium, samarium-cobalt, ferrite, bonded magnets

The permanent magnets in use today can be sorted by main phase and by how they are made (sintered or bonded). Coey puts production of rare-earth magnets at about 1.4×10⁵ t a year and ferrite magnets at about 1×10⁶ t a year (as stated in a paper of around 2018)Sourced [Source 4]. The table below sets manufacturers’ stated values and a paper’s values side by side with their conditions. Because the conditions differ, it does not rank the rows.

All Sourced (Arnold catalog Rev. 210607 [Source 7] and Recoma table Rev. 160205a [Source 9], Eclipse Magnetics datasheet [Source 11], Qin and He (2013) [Source 12]). Maximum use temperatures are manufacturers’ stated values, and the manufacturer itself notes that under strong demagnetizing fields or a low load line they can be considerably lower [Source 9]. Curie temperature: Sagawa et al. (585 K) and Coey (315 °C) [Sources 1, 4].

Even with the same main phase, how the powder is consolidated changes how strong the result can be. Coey sets out the ideal maximum energy products for Nd₂Fe₁₄B (Ms = 1.28 MA/m)Sourced [Source 4].

All Sourced (Coey’s Table 1 [Source 4]; ideal upper limits, not measured values).

What the table says is that both alignment and fill factor bear on (BH)max. A bonded magnet has a lower share of magnetic powder because of the resin, and an isotropic one cannot use the advantage of alignment. Even so, Coey notes that bonded magnets are used where shape freedom, dimensional accuracy and magnetization in any desired direction matter, as in the spindle motors of hard disk drives [Source 4].

4. How a neodymium magnet is made, and where coercivity is decided

Neodymium magnets were reported side by side in J. Appl. Phys. in 1984 by Sagawa et al. and Croat et al. (Coey dates the discovery to 1982)Sourced [Sources 1, 2, 4]. Sagawa et al. found a tetragonal compound of Nd, Fe and B (about 1 wt%) with lattice constants a = 0.880 nm and c = 1.221 nm, a Curie temperature of 585 K, an anisotropy constant of about 3.5 MJ/m³ and a saturation magnetization of 1.35 T (the value reported at the time). For a sintered magnet of composition Nd₁₅B₈Fe₇₇ they give Br = 1.23 T, HcI = 960 kA/m, (BH)max = 290 kJ/m³ and a Br temperature coefficient of −1260 ppm/KSourced [Source 1].

The main points of the process are as followsSourced [Source 5].

• Strip casting: molten alloy is poured onto a water-cooled copper wheel and cooled fast. This suppresses dendrites of α-Fe and improves the distribution of the Nd-rich phase (α-Fe is hard to powder and hinders alignment)
• Hydrogen decrepitation (HD): hydrogen is absorbed first by the grain-boundary phase, which swells and cracks intergranularly; the main phase then absorbs hydrogen and cracks, leaving a coarse powder. The reaction gives off heat and the powder reacts readily with oxygen, so it is handled under inert atmosphere
• Jet milling: particles are collided in gas jets at 5–10 bar, giving a narrow size distribution and low oxygen content
• Pressing in a field: the easy axes of all particles are aligned with the final magnetization direction and the powder is compacted. The green compact is about 60% of theoretical density
• Sintering and heat treatment: in inert gas or vacuum, aiming for over 95% of theoretical density. A post-sinter heat treatment can heal defects in the grain-boundary phase, make the Nd-rich phase more continuous, and raise coercivity

The place where coercivity is decided is not the “inside” of a crystal grain but the grain boundaries and surfaces. The Hirosawa et al. review (NIMS) sets this out as followsSourced [Source 3].

• Grains are micrometre-sized, far larger than the domain-wall width (about 4 nm) or the exchange length (about 1.2 nm in Nd₂Fe₁₄B). Even so, coercivity is governed by nanometre-scale local properties
• The grain-boundary phase between grains (about 2 nm thick) was found by three-dimensional atom probe to contain about 80 at.% or more Fe and Co, and three methods showed it to be ferromagnetic
• Nd ions and some Fe atoms at the interface have in-plane (negative) anisotropy, which lowers coercivity

Coey adds that reverse magnetization tends to nucleate at irregularities on the surface at edges and corners of the magnet, because the demagnetizing field is largest there in cylinders and blocks [Source 4].

5. Using less heavy rare earth — grain boundary diffusion, finer grains, light rare-earth substitution

According to Coey, heavy rare earths (Dy and Tb) are added because they raise uniaxial anisotropy through spin-orbit interaction. In high-temperature grades of around 2011, Dy and Tb were about 5% by weight, the Dy/Nd ratio was about 0.16, and raw-material cost was about double that of a magnet with no heavy rare earths. That situation spurred research into magnets with less or no heavy rare earthSourced [Source 4]. There is a price to pay. Adding heavy rare earths raises anisotropy but weakens magnetization (Coey) [Source 4]. In effect, strength is given up to buy coercivity.

There are three main ways to use less heavy rare earth.

1. Grain boundary diffusion (GBD): reverse magnetization tends to nucleate at irregular spots on surfaces and grain boundaries, so only those spots are strengthened with Dy or Tb. After sintering, heavy rare earth is introduced from the surface and diffuses along the grain boundariesSourced [Sources 4, 6, 8]
2. Finer grains: reducing powder particle size to the 1 µm range has cut the heavy rare earth neededSourced [Source 4]. Hirosawa et al. explain that with smaller grains the effective demagnetizing field produced by reversed surface grains is smaller [Source 3]
3. Substituting light rare earths: replacing part of the Nd with the more abundant, cheaper La and CeSourced [Sources 4, 14] (rare-earth-free magnets are covered in our explainer on rare-earth-free magnets)

View the original SVG schematic and exact annotations

The effect of grain boundary diffusion can be seen in Arnold’s catalogs. The company says its “GBD Neo” magnets enhance coercivity while using less Dy and Tb and limiting the loss of remanence, and that they suit thin magnets (under 6 mm) or laminated materialSourced [Source 8]. Set side by side at the same coercivity class, standard and grain-boundary-diffused grades compare as follows.

Sourced (Arnold catalog Rev. 210607 [Sources 7, 8]; typical Br and (BH)max, minimum HcJ). The difference column is Our calculation (diffused grade divided by standard grade, minus 1). This is a comparison within one maker and does not show other makers’ values. Diffused grades suit thin magnets, so shape and size conditions differ.

Besides grain boundary diffusion, there are routes that avoid heavy rare earths altogether. The next two items are company announcements and do not carry over directly to other companies or applications.

• Hot-deformed magnets: on 12 July 2016, Honda and Daido Steel announced that they had become the first in the world to put into practical use, in a hybrid vehicle’s drive motor, a hot-deformed neodymium magnet containing no heavy rare earth. The hot-deformation method is said to give crystal grains about one tenth the size of those in a sintered magnet, and Honda states that, with the magnet and rotor shapes revised, the motor showed torque, output and heat resistance equivalent to a motor using a conventional magnetSourced (company announcement) [Source 13].
• Lanthanum-cerium substitution: on 20 February 2018, Toyota announced a magnet that replaces part of the Nd with La and Ce, cuts Nd use by up to 50%, uses no Tb or Dy, and has heat resistance equivalent to earlier neodymium magnets. The company said it would advance mass-production technology and aim for early adoption in automobiles and roboticsSourced (company announcement) [Source 14]. Its status of practical use was not checked here Not yet confirmed.

6. The temperature constraint — usable temperature is set by coercivity

A magnet’s properties change with temperature. Coey divides the losses into three kindsSourced [Source 4].

• Reversible loss: Br and HcJ fall as temperature rises, because Ms and K₁ depend on temperature, and recover on returning to room temperature. It cannot be avoided
• Irreversible loss: from changes in the metastable domain structure of a fully magnetized magnet at high temperature. Remagnetizing at room temperature restores it
• Irredeemable loss: from changes in composition or microstructure, such as oxidation or volatilization of rare earths

Coey also says that suppliers of magnets for high-temperature use sometimes “condition” the magnet at a temperature about 50 K above the working range, so that irreversible losses are worked out in advance [Source 4].

Temperature has other sides tooSourced [Source 10].

• Low temperature: neodymium undergoes a “spin reorientation” at about 140 K, where the easy axis of magnetization becomes a cone and magnetic properties fall markedly. Replacing 80% or more of the Nd with Pr removes this reorientation
• Ferrite is the opposite: coercivity rises with temperature. In exchange, coercivity becomes small at low temperature; the practical lower limit is −40 °C for standard ferrite and about −60 °C for La-Co ferrite
• Maximum use temperature depends on conditions: the maximum use temperature a manufacturer gives can fall substantially under strong demagnetizing fields or a low load line [Source 9]

7. Applications — EV motors and wind power

USGS estimates that the largest use of rare earths worldwide is magnets (in the United States, catalysts)Sourced [Source 15]. Hirosawa et al. list, in order of production volume, automobiles (other than electric vehicles), disc drives, HEVs and EVs, electric bicycles, and transducers (including loudspeakers), and name wind generators and other power-generation and energy-storage equipment as near-future fields that will use large volumes of high-performance magnetsSourced [Source 3].

Most drive motors in EVs and HEVs are IPM (interior permanent magnet) types, with magnets buried in thin slots in the rotor. In this type the magnets receive the strong demagnetizing field produced by the stator windings, so high coercivity is a prerequisite. At high speed the magnets’ flux induces a large voltage in the windings, which must be cancelled by “flux weakening”, and that adds losses, according to Hirosawa et al. [Source 3].

The IEA sees rare earths as a small share of the final product. Its Global Critical Minerals Outlook 2026 says rare earths account for about 40% of permanent magnet costs but less than 1% of a vehicle’s valueSourced [Source 17]. This is stated in the context that, even when prices rise, the effect on the final product can be limited; it is not a price outlook.

The mass of magnet, and the rare-earth content, per wind generator or EV motor varies widely with design and capacity. This article could not confirm a single representative figure from public bodies or peer-reviewed papers, so it gives no number (Chapter 10).

8. Rare-earth supply — mining, refining, and China’s export controls (stages and dates)

(1) Where it is mined, and where it is refined

According to the USGS Mineral Commodity Summaries 2026 (February 2026), world rare-earth mine production in 2025 (estimate) was about 390,000 t (oxide equivalent): China 270,000 t, the United States 51,000 t, Australia 29,000 t and Myanmar 22,000 tSourced [Source 15]. Our calculation China’s share is about 69% (270,000 / 390,000). Reserves are 44 Mt in China and over 75 Mt worldwide (USGS), so China’s share is under 60% (44 / over 75).

Refining is more concentrated than mining. The IEA states that for gallium, graphite, manganese and rare earths the top refiner, China, accounts for over 90% of global supply. It also says rare-earth refining was an exception, with concentration falling slightly in 2023–2025 thanks to new projects in the United States and higher output in MalaysiaSourced [Sources 17, 18].

For heavy rare earths (Tb, Dy and others), USGS writes that the United States mined heavy-rare-earth minerals in 2025 but no company produced heavy-rare-earth compounds or metals at sustained commercial scale. US import sources for 2021–2024 were China for 100% of terbium and 100% of holmium (USGS)Sourced [Source 16]. It also records government-linked financing to support separation facilities [Source 16].

(2) The IEA’s outlook (scenarios)

In IEA scenarios, existing and announced projects would give, by 2035 outside China, mining capacity of about 50 kt of rare earths, planned refining and separation capacity of under 40 kt (mainly Malaysia and the United States), and downstream capacity for metals, alloys and magnets of only about 18 kt (rare-earth content)Not yet confirmed [Source 18]. The same report says diversified refining capacity would equal about two thirds of 2035 mined supply while planned magnet production capacity is only about one third, and it estimates that diversifying magnet rare-earth supply chains would need about USD 60 billion of investment over the next decadeNot yet confirmed [Source 17].

(3) China’s export controls — keeping announced, in force and suspended apart

Stages and dates were checked against official announcements of China’s Ministry of Commerce (MOFCOM) and the General Administration of Customs. The date checked is 10 October 2026; later changes are not reflected.

All Sourced (original texts of announcements by China’s Ministry of Commerce and Customs [Sources 19-25]; USGS [Source 15]). Nos. 55 and 58 concern superhard materials and battery materials, not rare earths, so they are not treated in this table. English renderings of the announcements were made by this article.

9. Policy — the EU Critical Raw Materials Act and Japan’s specified critical materials

European Union: Regulation (EU) 2024/1252 (the Critical Raw Materials Act) lists “rare earth elements for permanent magnets (Nd, Pr, Tb, Dy, Gd, Sm, and Ce)” among its strategic raw materials. Its 2030 benchmarks (Article 5) are: extraction in the EU of at least 10% of annual consumption, processing of at least 40%, recycling of at least 25%, and no single third country supplying more than 65%Sourced [Source 26].

Permanent magnets have dedicated provisions (Articles 28 and 29). For products including MRI devices, wind generators, industrial robots, motor vehicles, heat pumps and electric motors, they require a label showing whether permanent magnets are present and of which type (neodymium-iron-boron, samarium-cobalt, aluminium-nickel-cobalt, ferrite), and a data carrier giving the weight, location and chemical composition of the magnets and how to remove them. The Commission is to adopt the implementing act by 24 November 2026, and the requirements apply two years after that act enters into force (products designed primarily for defence or space are outside Article 28). Disclosure of the recycled-content share (Nd, Dy, Pr, Tb, B, Sm, Ni, Co) is due by 24 May 2027 or two years after the delegated act enters into force, whichever is laterSourced [Source 26]. Whether the implementing and delegated acts have been adopted was not checked here Not yet confirmed.

Japan: according to the Cabinet Office, in December 2022 (Reiwa 4) a Cabinet Order designated 11 items, including permanent magnets, as specified critical materials under the Economic Security Promotion ActSourced [Source 27]. The individual certifications and support measures were not checked here.

Recycling today: USGS writes that only limited quantities of rare earths are recovered from batteries, permanent magnets and fluorescent lamps in the United States [Source 15]. The IEA says magnet recycling could benefit as end-of-life volumes grow from early generations of EVs and wind turbinesNot yet confirmed [Source 17]. Recycling technology is not covered in this article (Chapter 10).

10. Limits, and what could not be confirmed

(1) Manufacturers’ numbers

Arnold’s typical and minimum values are stated values for its standard grades, not other makers’ values. The diffused grades target magnets under 6 mm thick. The Eclipse Magnetics figures come from a distributor datasheet, not a maker’s specification. The TDK ferrite datasheet could not be reached from this site and was not checked. The bonded-magnet values are one example from one paper (MQP-B powder, epoxy, press molding).

(2) The amount of heavy rare earth saved

The percentage (by weight) of Dy and Tb saved by grain boundary diffusion or hot deformation could not be confirmed from primary sources and is not given. The Honda and Daido Steel and Toyota announcements are company announcements and cannot be generalized to other applications or other makers’ magnets. The status of practical use of Toyota’s magnet was not checked.

(3) Typical usage figures

The mass of magnet, or rare-earth content, per wind generator or EV motor could not be confirmed as a single representative value from public bodies or peer-reviewed papers, so it is not given.

(4) Supply and policy

The suspension under China’s Announcement No. 70 runs to 10 November 2026 (about a month after the date checked), and the position after that could not be confirmed. Whether the EU’s implementing and delegated acts on permanent magnets have been adopted, and the content of individual certifications and support under Japan’s specified critical material “permanent magnets”, were not checked. The effect of No. 1 of 2026 on civilian use cannot be read from the text of the announcement.

(5) Out of scope

Alnico magnets, recycling technology (such as hydrogen processing), prices and price forecasts, corrosion and surface treatment of magnets, and safety (handling strong magnetic fields) are not covered. Rare-earth-free magnets are in our explainer on rare-earth-free magnets, magnetocaloric materials in our explainer on magnetocaloric materials and magnetic refrigeration, and magnetic thin-film materials in our explainer on spintronic materials.

11. How to read evaluations and standards

A permanent magnet’s performance is measured by tracing the demagnetization curve and reading Br, HcB, HcJ and (BH)max from it. The usual methods are a closed-circuit hysteresigraph (electromagnet method), in which the sample is clamped between electromagnet poles, and an open-circuit method for very-high-coercivity magnets (for example a vibrating-sample magnetometer with a superconducting magnet). Results depend on sample shape, temperature and the degree of magnetization [Sources 10, 30, 32]. Standards divide into test methods and specifications of minimum values for each material.

Standard numbers, titles and editions in each row are Sourced (checked on IEC, JSA, BSI and AFNOR catalogue pages and in the references of an Arnold technical note). Arranging the table is an editorial task of Our calculation. For ASTM standards, the ASTM site was not opened; only numbers and titles confirmed from non-ASTM public sources are given, and no test conditions, procedures or acceptance criteria are described. The texts of the standards are paid and were not checked.

How the positions differ: IEC 60404 is an international standard (TC 68, magnetic alloys and steels). JIS C 2501 and C 2502 are national standards that technically modify the 2015 editions of IEC 60404-5 and 60404-8-1 (MOD). JIS C 2500 was established in 2022 with reference to IEC TR 63304:2021. IEC has IEC 60404-18:2025, covering the same kind of open-circuit method with a superconducting magnet (SCM-VSM and SCM extraction methods). ASTM standards are US industry standards used alongside IEC and JIS. JIS C 2502 is based on the 2015 edition of IEC 60404-8-1, and whether JIS has taken in the additions of the 2023 edition (hot-deformed Nd-Fe-B and others) was not checked here Not yet confirmed.

Cautions when reading standards

• A standard is a “test method”; it does not guarantee performance. The minimum values in IEC 60404-8-1 and JIS C 2502 are guaranteed-side values for a grade and differ from the “typical values” in catalogs
• The same “HcJ” can change with how it is measured. The closed-circuit electromagnet method is prone to error for high-HcJ magnets, and the open-circuit superconducting-magnet method needs a demagnetizing-field correction. Check which method, and at what temperature, a value was measured
• Read a maker’s value together with the standard, edition, measurement temperature, and whether it is typical or minimum. A maximum use temperature is not a quantity in a standard; it reads as a value the maker gives under conditions such as a load line

12. Glossary

13. References (primary sources)

1. Sagawa M. et al. “New material for permanent magnets on a base of Nd and Fe (invited)”, J. Appl. Phys. 55, 2083–2087 (1984), DOI 10.1063/1.333572 — doi.org/10.1063/1.333572
2. Croat J. J. et al. “Pr-Fe and Nd-Fe-based materials: A new class of high-performance permanent magnets (invited)”, J. Appl. Phys. 55, 2078–2082 (1984), DOI 10.1063/1.333571 — doi.org/10.1063/1.333571
3. Hirosawa S., Nishino M., Miyashita S. (NIMS) “Perspectives for high-performance permanent magnets: applications, coercivity, and new materials”, Adv. Nat. Sci.: Nanosci. Nanotechnol. 8, 013002 (2017), DOI 10.1088/2043-6254/aa597c (open access) — doi.org/10.1088/2043-6254/aa597c
4. Coey J. M. D. “Perspective and Prospects for Rare Earth Permanent Magnets”, Engineering 6, 119–131 (2020), DOI 10.1016/j.eng.2018.11.034 (open access) — doi.org/10.1016/j.eng.2018.11.034
5. Herraiz Lalana E. “Permanent magnets and its production by powder metallurgy”, Rev. Metal. 54(2), e121 (2018), DOI 10.3989/revmetalm.121 (open access) — doi.org/10.3989/revmetalm.121
6. Hirota K. et al. “Coercivity Enhancement by the Grain Boundary Diffusion Process to Nd–Fe–B Sintered Magnets”, IEEE Trans. Magn. 42, 2909–2911 (2006), DOI 10.1109/TMAG.2006.879906 (bibliographic record only checked) — doi.org/10.1109/TMAG.2006.879906
7. Arnold Magnetic Technologies “Neodymium Iron Boron Magnet Catalog” (Rev. 210607) — arnoldmagnetics.com/…/Arnold-Neo-Catalog.pdf
8. Arnold Magnetic Technologies “Grain Boundary Diffused Neodymium Iron Boron Magnet Catalog” (Rev. 210607) — arnoldmagnetics.com/…/Arnold-Grain-Boundary-Diffused-Neo-Catalog.pdf
9. Arnold Magnetic Technologies “RECOMA Summary of main grades of SmCo5 and Sm2Co17” (table Rev. 160205a, Recoma Combined) — arnoldmagnetics.com/…/Recoma-Combined-160301.pdf
10. Arnold Magnetic Technologies “What is the Magnetic Hysteresis Loop” (TECHNote TN 1501 rev.2015a) — arnoldmagnetics.com/…/Tech-Notes_Hysteresis-Loop.pdf
11. Eclipse Magnetics “Ferrite and ceramic magnets datasheet” (distributor datasheet) — eclipsemagnetics.com/…/ferrite_and_ceramic_datasheet.pdf
12. Qin W., He J. “Research on Composite Powder and Magnet Properties of Bonded NdFeB Magnets Prepared by Press Molding”, Appl. Mech. Mater. 345, 218–222 (2013), DOI 10.4028/www.scientific.net/AMM.345.218 (abstract checked) — scientific.net/AMM.345.218
13. Honda and Daido Steel “Daido Steel and Honda Adopt World’s First Hybrid Vehicle Motor Magnet Free of Heavy Rare Earth Elements” (12 July 2016, news release) — global.honda/en/newsroom/news/2016/4160712eng.html
14. Toyota Motor Corporation “Toyota Develops New Magnet for Electric Motors Aiming to Reduce Use of Critical Rare-Earth Element by up to 50%” (20 February 2018, news release) — global.toyota/en/newsroom/corporate/21139684.html
15. U.S. Geological Survey (USGS) “Mineral Commodity Summaries 2026: Rare Earths” (February 2026) — pubs.usgs.gov/periodicals/mcs2026/mcs2026-rare-earths.pdf
16. U.S. Geological Survey (USGS) “Mineral Commodity Summaries 2026: Rare Earths (Heavy)” — pubs.usgs.gov/periodicals/mcs2026/mcs2026-rare-earths-heavy.pdf
17. International Energy Agency (IEA) “Global Critical Minerals Outlook 2026”, Executive summary — iea.org/reports/global-critical-minerals-outlook-2026/executive-summary
18. International Energy Agency (IEA) “Global Critical Minerals Outlook 2026”, Outlook — iea.org/reports/global-critical-minerals-outlook-2026/outlook
19. China, Ministry of Commerce and General Administration of Customs Announcement No. 18 of 2025, decision on export controls on certain medium and heavy rare-earth items, 4 April 2025 (in Chinese) — mofcom.gov.cn/…/art_5e663d0b…
20. China, Ministry of Commerce and General Administration of Customs Announcement No. 57 of 2025, decision on export controls on certain medium and heavy rare-earth items, 9 October 2025 (in Chinese) — mofcom.gov.cn/…/art_9510f263…
21. China, Ministry of Commerce and General Administration of Customs Announcement No. 56 of 2025, decision on export controls on certain rare-earth equipment and raw and auxiliary materials, 9 October 2025 (in Chinese) — mofcom.gov.cn/…/art_22129555…
22. China, Ministry of Commerce Announcement No. 61 of 2025, decision on export controls on overseas rare-earth items, 9 October 2025 (in Chinese) — mofcom.gov.cn/…/art_e2587592…
23. China, Ministry of Commerce Announcement No. 62 of 2025, decision on export controls on rare-earth technologies, 9 October 2025 (in Chinese) — mofcom.gov.cn/…/art_04a2df12…
24. China, Ministry of Commerce and General Administration of Customs Announcement No. 70 of 2025, decision to suspend implementation of Announcements Nos. 55, 56, 57, 58, 61 and 62, 7 November 2025 (in Chinese) — mofcom.gov.cn/…/art_da147202…
25. China, Ministry of Commerce Announcement No. 1 of 2026, announcement on strengthening export controls on dual-use items to Japan, 6 January 2026 (in Chinese) — mofcom.gov.cn/…/art_5acdf514…
26. European Parliament and Council “Regulation (EU) 2024/1252 establishing a framework for ensuring a secure and sustainable supply of critical raw materials” (11 April 2024; Official Journal 3 May 2024) — eur-lex.europa.eu/…CELEX:32024R1252
27. Cabinet Office of Japan Supply chain resilience initiatives (system for securing a stable supply of critical materials) (in Japanese) — cao.go.jp/keizai_anzen_hosho/suishinhou/supply_chain/supply_chain.html
28. IEC “IEC 60404-5:2015” product page (IEC Webstore) — webstore.iec.ch/en/publication/22142
29. IEC “IEC 60404-8-1:2023” product page (IEC Webstore) — webstore.iec.ch/en/publication/68440
30. IEC “IEC 60404-18:2025” product page (IEC Webstore) — webstore.iec.ch/en/publication/76495
31. Japanese Standards Association (JSA) JIS C 2501:2019 “Methods of test for permanent magnet”, viewing preview (in Japanese) — webdesk.jsa.or.jp/preview/pre_jis_c_02501_000_000_2019_j_ed10_ch.pdf
32. Japanese Standards Association (JSA) JIS C 2500:2022 “Methods of measurement of the magnetic properties of permanent magnet (magnetically hard) materials in an open magnetic circuit using a superconducting magnet”, viewing preview (in Japanese) — webdesk.jsa.or.jp/preview/pre_jis_c_02500_000_000_2022_j_ed10_ch.pdf
33. Japanese Standards Association (JSA) JIS C 2502:2019 “Materials for permanent magnet”, viewing preview (in Japanese) — webdesk.jsa.or.jp/preview/pre_jis_c_02502_000_000_2019_j_ed10_ch.pdf
34. British Standards Institution (BSI) “ASTM A1101-23” catalogue page (BSI Knowledge; not an ASTM site) — knowledge.bsigroup.com/…/ndfeb-permanent-magnets-1
35. AFNOR “ASTM A1054-16” catalogue page (AFNOR Editions; not an ASTM site) — boutique.afnor.org/…/am95890/275610

14. Claim-to-source audit

Last updated 10 October 2026. Sources are limited to primary material (peer-reviewed papers, magnet makers’ stated values, USGS and IEA reports, announcements of China’s Ministry of Commerce, an EU regulation, and standards-body catalogue pages). The texts of standards are paid and were not checked. The ASTM International site was not opened, because that body states that AI use of the site is not permitted. Magnet numbers are makers’ stated values and papers’ values, given with their conditions (temperature, sintered or bonded, typical or minimum). Because the article includes structural readings and materials-design interpretations, those are marked as Commentary and kept separate from sourced fact. Usage per wind generator or EV, the reduction in heavy rare earth, the position after the end of the suspension under Announcement No. 70, the adoption status of EU implementing acts and the practical use of Toyota’s magnet could not be confirmed and are not stated here. All figures are explanatory concept graphics. Figs. 2 to 6, 8 and 9 are vector drawings; Figs. 1 and 7 are AI-generated images, and none of them shows a real micrograph or physical product.