Zirconium

What is Zirconium?

21 CFR 700.16 deems any aerosol cosmetic product containing zirconium adulterated under section 601(a) of the Federal Food, Drug, and Cosmetic Act. That rule was published at 42 FR 41376 on August 16, 1977, and it reaches products introduced into interstate commerce after September 15, 1977. Read the scope carefully, because it is narrow and it is the most misquoted fact about this element. The regulation covers aerosol cosmetics. It is not a ban on zirconium, it does not touch stick or roll-on antiperspirants, and it says nothing about food or supplements.

Zirconium is element 40, atomic weight 91.22, a hard silvery transition metal with a specific gravity of 6.51 and a melting point the NIOSH Pocket Guide gives as 3,375 degrees Fahrenheit (about 1,855 degrees Celsius), insoluble in water. Martin Heinrich Klaproth identified it in 1789 in the gemstone zircon and named it for the mineral. Commonly cited reference tables put its crustal abundance near 130 milligrams per kilogram, more abundant in rock than copper, zinc or lead. It occurs in more than 140 minerals, but commerce runs almost entirely on one: zircon, zirconium silicate, ZrSiO4.

The US Geological Survey's Mineral Commodity Summaries of February 2026 reports world zircon mine production of about 1,200 thousand metric tons gross weight in 2025, down from 1,300 in 2024, led by Australia at 400, South Africa at 270 and Mozambique at 160. None of it is produced for human consumption. It goes into ceramics, foundry sand, refractories, zirconium chemicals, nuclear fuel cladding, dental crowns and orthopedic bearing surfaces. Zirconium nonetheless shows up on supplement labels, inside liquid "colloidal mineral" and fulvic blends, which is why this issue exists.

Common Names

Zirconium, Zr, element 40. Product and compound names a buyer may encounter include zirconium dioxide, zirconia, yttria stabilized tetragonal zirconia (Y-TZP), zirconium silicate, zircon, zirconium oxychloride, zirconium tetrachloride, zirconium sulfate, sodium zirconium lactate, the eight aluminum zirconium chlorohydrate salts in antiperspirants, oxidized zirconium (the Zr-2.5Nb implant alloy marketed as OXINIUM), Zircaloy, and sodium zirconium cyclosilicate, the prescription drug LOKELMA. Supplement panels usually print the bare word "Zirconium" with no compound named.

Primary Active Compounds

• Zirconium dioxide, ZrO2, zirconia. The implant and dental ceramic, chemically inert in bulk, and the form most people carry in their bodies on purpose.

• Zirconium silicate, ZrSiO4, zircon. CAS 10101-52-4. Glaze opacifier, foundry sand, refractory. The dominant occupational dust.

• Zirconium tetrachloride, ZrCl4, CAS 10026-11-6. Water reactive, hydrolyzes to hydrochloric acid, NFPA health rating 3. Carved out of the NIOSH exposure limit.

• Soluble salts: zirconium oxychloride, nitrate, sulfate and acetate. The forms behind the only oral toxicity numbers any US agency has derived.

• Aluminum zirconium chlorohydrate complexes, eight named in 21 CFR 350.10, permitted up to 20 percent in nonaerosol antiperspirants only.

• Sodium zirconium cyclosilicate, Na1.5H0.5ZrSi3O9 times 2 to 3 H2O. A prescription potassium binder engineered not to be absorbed.

Key Note

Zirconium has no Dietary Reference Intake, no Recommended Dietary Allowance, no Tolerable Upper Intake Level and no Daily Value. The entire regulatory apparatus that exists for it governs air you breathe, skin you apply it to, and devices implanted in you. The European Union prohibits zirconium and its compounds in cosmetics at Annex II entry 391 of Regulation (EC) No 1223/2009, with two carve-outs: the antiperspirant complexes in Annex III, and zirconium lakes, pigments or salts of colorants listed in Annex IV. The United States sets an occupational air limit of 5 milligrams per cubic meter. Nobody has set a number for swallowing it voluntarily, because nobody anticipated that anyone would. That gap, a dense occupational and dermal record beside an almost empty oral one, structures everything below.

Primary Functions & Benefits

Zirconium performs no function in human physiology, and the chemistry explains why better than any slogan. Zr4+ is a small, highly charged, strictly tetravalent ion with no accessible redox couple under physiological conditions. Biology uses metals that cycle electrons (iron, copper, manganese, molybdenum) or carry charge across membranes (sodium, potassium, calcium, magnesium). Zirconium does neither. What it does instead is hydrolyze: in water near neutral pH, Zr4+ forms polynuclear hydroxo complexes and then drops out as insoluble hydrated oxide. There is no zirconium transporter, no zirconium-dependent enzyme, and no zirconium binding protein identified in any organism.

Ghosh, Sharma and Talukder reviewed the question in Biological Trace Element Research in 1992, volume 35, pages 247 to 271, under the title "Zirconium. An abnormal trace element in biology." Their conclusion was that zirconium is present and retained in relatively large quantities yet has not been associated with any specific metabolic function. That assessment has not been overturned in the three decades since.

Forms & Standardization

For zirconium this section is the toxicology of each chemical form, and it is the most useful part of this issue, because the form determines nearly everything. Two materials with identical zirconium content by mass can be a dental crown and a corrosive industrial reagent. Standardization here means specification compliance, never potency.

Zirconium metal and alloys

Bulk zirconium metal is unreactive and is not the hazard. Powder is. The NIOSH Pocket Guide classifies zirconium metal as combustible and difficult to ignite as a solid, but states that the powder may ignite spontaneously and can keep burning under water, which is why fine powder is stored completely immersed. Incompatibilities are potassium nitrate and oxidizers. The UK National Poisons Information Service monograph gives an explosive limit in air of 0.16 grams per liter. US Department of Transportation numbers split the material by condition: 1358 for wet powder under Guide 170, 2008 for dry powder and 1932 for scrap, both under Guide 135. Zircaloy and the Zr-2.5Nb alloy in orthopedic bearings are solid metal and present no inhalation route once fabricated.

Zirconium dioxide, zirconia

This is the form in the largest number of human bodies, and the one with the most conservative engineering behind it. Surgical grade material is yttria stabilized tetragonal zirconia polycrystal, governed by ISO 13356, whose test battery names the failure mode. Specimens are aged at 134 degrees Celsius in saturated water vapor for 5 hours, then re-measured for monoclinic phase fraction by X-ray diffraction and for four point flexural strength on at least 30 specimens. Cyclic fatigue runs per ISO 22214 on at least 5 specimens in 37 degree Celsius saline at a minimum load of 320 megapascals, 20 hertz, one million cycles. The standard also carries a radioactivity clause, because zirconium ore carries uranium and thorium.

The reason for all of that is a real clinical failure. On August 10, 2001, the French health products agency AFSSAPS (now ANSM) suspended use of Prozyr zirconia ceramic hip prosthesis heads from TH batches made by Saint Gobain Desmarquest, citing 56 materiovigilance reports of ruptures. An independent laboratory found deterioration of the zirconia as porosity and decohesion of the ceramic grains, and concluded it existed before implantation. Five lots were named, following a June 2001 recall of lot TH 93038. The regulator did not state a fracture rate, and no denominator of units implanted was published, so any percentage quoted for this episode is somebody's reconstruction.

Zirconia is not inert in every presentation. The UK National Poisons Information Service monograph records a granulomatous skin reaction in a 15 year old girl after a 4 percent zirconium oxide cream for poison ivy dermatitis, citing Baler in Archives of Dermatology 1965, volume 91, pages 145 to 148, who noted that reactions to insoluble forms were rarer but more persistent once started. I did not read that paper, so treat it as the monograph's summary.

Zirconium silicate, zircon

This is the industrial dust, and where the human lung evidence lives. Liippo and colleagues reported in the American Review of Respiratory Disease in 1993, volume 148, issue 4 part 1, pages 1089 to 1092, on a nonsmoking ceramic tile worker 25 years of age with worsening dry cough and dyspnea after 3.5 years sorting and glazing tiles. Open lung biopsy showed intense granulomatous interstitial pneumonia with mild fibrosis compatible with hypersensitivity pneumonitis, plus numerous very small birefringent crystals around the terminal airways and occasionally inside granulomas. Particle analysis put the inhaled dust burden at nearly 100 times normal background, mainly clay minerals and zirconium silicate. She died one week after the biopsy, which the UK poisons monograph attributes to pneumothorax, acute right heart failure and splenic rupture. The authors' point was specific: a sarcoid granulomatosis pattern was already known after long term zirconium exposure, and this case showed an acute, fulminant allergic alveolitis-like reaction was also possible.

Zircon also carries natural radioactivity, a separate hazard from the chemistry. A Spanish Nuclear Safety Council survey of ceramic industries by Serradell and colleagues at the Universidad Politecnica de Valencia, published in the mid 2000s, measured uranium-238 at 2,681 to 3,615 becquerels per kilogram in zircon sand and 3,134 in zircon flour, with thorium-232 at 597 to 714 in sand and 592 in flour. Modeling 2,000 working hours per year, annual internal dose exceeded 1 millisievert only in some areas of the zircon milling plants, with a raw material store at 3.32 millisieverts external. That is a milling and glaze plant exposure, not a dental crown exposure, and not a reason to worry about a ceramic mug.

Zirconium tetrachloride

This compound is handled separately by the regulators for good reason. CAMEO Chemicals describes it as reacting vigorously with water to form hydrochloric acid with evolution of heat, and corrosive to metals in the presence of moisture. NFPA ratings are health 3, flammability 0, instability 2, with the W special notation for water reactivity. Inhalation irritates the upper airways, probably through released hydrochloric acid, and ingestion is described as causing burning of the mouth and throat, vomiting, bloody diarrhea, collapse and convulsions. The UK monograph notes that pulmonary edema after exposure may be delayed as much as 36 hours, the detail that matters most clinically. Department of Energy Protective Action Criteria for zirconium chloride are 26, 220 and 1,300 milligrams per cubic meter for PAC-1, PAC-2 and PAC-3. The NIOSH recommended exposure limit for zirconium compounds explicitly does not cover this one.

Soluble zirconium salts

Zirconium oxychloride, nitrate, sulfate and acetate are the forms behind EPA's only oral numbers. The 2012 Provisional Peer Reviewed Toxicity Value assessment covers soluble zirconium compounds specifically, so its numbers should never be applied to zirconia ceramic or to zircon. Both key studies came from Schroeder's laboratory and both gave 5 parts per million zirconium sulfate in drinking water (about 1.28 parts per million as elemental zirconium) plus feed containing 2.66 micrograms per gram. In Long-Evans rats, 56 males and 58 females over a lifetime, the lowest observed adverse effect level was 0.79 milligrams per kilogram per day in males and 0.89 in females, with increased glycosuria, higher fasting glucose in females and higher cholesterol in males. In CD-1 mice, 54 of each sex over a lifetime, it was 1.71 in males with body weight reduction above 10 percent, and EPA logged 1.75 in females as a frank effect level because survival fell. Neither study had a second dose group, so neither produced a no-effect level, and the rat study was disturbed by a pneumonia outbreak.

Aluminum zirconium chlorohydrate complexes

These are the forms a consumer actually buys, and the specification detail is unusually tight. 21 CFR 350.10 paragraphs (k) through (r) list eight: aluminum zirconium octachlorohydrate, octachlorohydrex gly, pentachlorohydrate, pentachlorohydrex gly, tetrachlorohydrate, tetrachlorohydrex gly, trichlorohydrate and trichlorohydrex gly. All eight are capped at up to 20 percent calculated on an anhydrous basis excluding buffer, and all eight are restricted to nonaerosol dosage forms. Where applicable the material must meet the aluminum to chloride, aluminum to zirconium, and aluminum plus zirconium to chloride atomic ratios in the United States Pharmacopeia and National Formulary. The source citation on that section is 68 FR 34291, June 9, 2003.

The European Union writes the same chemistry with a second ceiling. Annex III entry 50 of Regulation (EC) No 1223/2009 permits aluminum zirconium chloride hydroxide complexes and their glycine complexes in antiperspirants at a maximum of 20 percent as the anhydrous complex and 5.4 percent expressed as zirconium, requires the ratio of aluminum atoms to zirconium atoms to fall between 2 and 10 and the ratio of aluminum plus zirconium atoms to chlorine atoms to fall between 0.9 and 2.1, bars use in aerosol dispensers, and mandates the warning "Do not apply to irritated or damaged skin."

These complexes can sensitize. Skelton and colleagues reported in the Journal of the American Academy of Dermatology in 1993, volume 28, issue 5 part 2, pages 874 to 876, the first recognized hypersensitivity granuloma caused by an aluminum zirconium complex, in an adult woman's axilla. Montemarano, Sau, Johnson and James reported cutaneous granulomas from an aluminum zirconium antiperspirant complex in the same journal in 1997, volume 37, issue 3 part 1, pages 496 to 498, again in an adult woman. The published case literature on these complexes is small against a user population in the hundreds of millions. It is also not empty.

Sodium zirconium lactate

This compound started the story and it left deodorants decades ago. Prior, Rustad and Cronk published experimental pathology on deodorant preparations containing sodium zirconium lactate in the Journal of Investigative Dermatology in 1957, volume 29, issue 6, pages 449 to 463, and The Lancet carried a note titled "Zirconium granulomas" on May 31, 1958, volume 1, issue 7031, pages 1164 to 1165. The UK monograph states that Shelley and Hurley, in the British Journal of Dermatology 1958, volume 70, pages 75 to 101, reported 70 cases of axillary granuloma between 1956 and 1958. I did not read that paper, so the count is the monograph's.

Sodium zirconium cyclosilicate

This form inverts the usual problem. LOKELMA is a prescription drug, FDA approved May 18, 2018 after two complete response letters in May 2016 and March 2017. Its formula is Na1.5H0.5ZrSi3O9 times 2 to 3 H2O, a crystalline inorganic cation exchanger that traps potassium in the gut lumen and is engineered to stay there. The label states it is not absorbed systemically after oral administration, is not subject to enzymatic metabolism, and that in a rat mass balance study it was recovered in feces with no evidence of systemic absorption. Section 12.3 reports urine and blood zirconium similar between treated and untreated patients, either undetectable or near the assay's lower limit of quantification. That needs care: a value reported as below the limit of quantification is not a measurement of zero, and the label does not claim it is. The one zirconium compound a regulator has reviewed for oral human use was built on the premise that none of the zirconium should get in.

Its prescribed schedule is 10 grams of the salt three times daily for up to 48 hours, then 10 grams once daily, titrated in 5 gram steps weekly or slower within a range of 5 grams every other day to 15 grams daily. Each 5 grams carries 400 milligrams of sodium. Those figures describe a supervised prescription for diagnosed hyperkalemia under potassium monitoring. They are drug facts, not transferable to anything bought off a shelf.

Zirconium phosphate and hydrous zirconium oxide

These are the sorbent materials in regenerative dialysis cartridges, where acid zirconium phosphate exchanges ammonium and cations and hydrous zirconium oxide handles anions. The patient's exposure is to dialysate that has passed through zirconium media, which puts the safety question on device leachate limits.

Food Sources

Zirconium is not something to source from food, so what follows inventories how it actually reaches people, ranked by how much arrives and how deep it gets.

Diet and water

The dietary figures in circulation trace back to one era of work. Schroeder and Balassa published "Abnormal trace metals in man: zirconium" in the Journal of Chronic Diseases in May 1966, volume 19, issue 5, pages 573 to 586, and the commonly quoted estimates of roughly 4.15 milligrams per day (about 3.5 from food and 0.65 from water) and a total body content near 250 milligrams descend from it. The 1992 Ghosh review reports daily human uptake "known to be as high as 125 milligrams," a thirtyfold spread from the usual figure. Read that range as a sign the dietary dataset is sixty years old and was built with sixty year old analytical chemistry. EPA's 2012 assessment states outright that data on the differential bioavailability of zirconium from drinking water versus diet were not available.

Inhalation

This is the route the hazard record is built on. Occupational dust in zircon milling, foundry sand handling, ceramic glaze and frit production, refractory manufacture and abrasive work produced both the pulmonary case reports and the exposure limits. The UK monograph records coal miners with up to 14 parts per million zirconium in heavily pigmented lung sections, 22 parts per million in pulmonary lymph nodes, blood up to 100 micrograms per liter and urine up to 12 micrograms per liter.

Skin

Antiperspirants are the dominant consumer dermal route, at up to 20 percent aluminum zirconium complex in nonaerosol products in both the US and the EU. Absorption through intact skin is poor. The UK monograph notes that granulomas occur mainly where skin is damaged, which is why both the US monograph and the EU annex carry a damaged-skin warning.

Implants and devices

Zirconia crowns, zirconia and oxidized zirconium orthopedic bearing surfaces, and zirconium phosphate dialysis sorbents put the element inside the body by design, in solid or immobilized form. For a crown or a femoral head the exposure is a surface, not a dose.

Prescription medicine

Sodium zirconium cyclosilicate is the one oral zirconium exposure anyone has studied properly, and the point of its design is that the zirconium does not leave the gut.

Supplements

This is the route with no justification at all. The honest summary is not "eat the food instead." It is that there is nothing here to seek, in food or in a bottle.

Who Should Take Zirconium

Zirconium belongs in a reactor cladding tube, a foundry mold and a dental crown, and there is no person, age, symptom or diagnosis for which taking it by mouth is the right decision. What this section can cover is who already carries an exposure they did not choose, and who has reason to ask for a measurement.

Unavoidably exposed

• Zircon milling and grinding plant workers, who face both the dust and the uranium-238 and thorium-232 in the ore, with modeled internal dose above 1 millisievert per year in some plant areas.

• Ceramic tile sorters and glazers, the occupation in the 1993 hypersensitivity pneumonitis case, plus foundry workers handling zircon sand, refractory installers and frit producers.

• Zirconium and hafnium metal production workers, including anyone handling zirconium tetrachloride, and nuclear fuel fabrication workers machining Zircaloy.

• Dental laboratory technicians milling zirconia copings, and anyone carrying a zirconia crown, bridge, implant abutment or joint bearing.

• Hyperkalemia patients on prescription sodium zirconium cyclosilicate.

• Everyone using a nonaerosol antiperspirant with an aluminum zirconium active, which is a very large fraction of the adult population.

Who should consider testing

NIOSH references medical test guidance under its Medical Tests entry 0247, and blood and urine zirconium can be measured by inductively coupled plasma mass spectrometry. The problem is interpretation: no clinical action threshold, no reference interval tied to outcomes, no treatment that follows a result. A worker with suspected zirconium lung disease needs imaging, pulmonary function testing and tissue, not a blood level. A patient with a persistent axillary papular eruption needs a biopsy and an ingredient review, not a metal panel.

Who Should AVOID or Use Caution

Contraindications

• Any voluntary oral zirconium product, for everyone. No indication, no reviewed product, no dose.

• Application of any zirconium-containing antiperspirant or cream to broken, irritated or damaged skin. The US OTC antiperspirant monograph requires "Do not use on broken skin." The EU annex entry requires "Do not apply to irritated or damaged skin." Both exist because damaged skin is where the granulomas formed.

• Continued use of a zirconium antiperspirant by anyone who has previously developed a granulomatous axillary reaction to one. Delayed hypersensitivity does not resolve with re-exposure.

• Zirconium metal powder without engineering controls. It may ignite spontaneously and continues burning under water.

Use Caution

• Kidney disease. The OTC antiperspirant monograph requires "Ask a doctor before use if you have kidney disease," the only systemic caution in that monograph.

• Occupational exposure above the OSHA eight hour limit of 5 milligrams per cubic meter, or above the 10 milligrams per cubic meter short term figure NIOSH and ACGIH set for a 15 minute period.

• Zirconium tetrachloride handling. Pulmonary edema can be delayed up to 36 hours, so a worker who feels fine at the end of a shift is not cleared.

• Patients on sodium zirconium cyclosilicate with heart failure or renal disease, because of the dose dependent edema signal in the label.

Critical Safety Point

The dermal and pulmonary record for zirconium compounds is a delayed hypersensitivity record, so the dangerous property is latency rather than acute potency. Axillary granulomas appeared days to weeks after contact, per the UK poisons monograph. Lung granulomas, in FDA's own words in the 1977 rule, may not be detected until they are advanced or permanent. A substance that does nothing on day one and something irreversible in year five is exactly the kind for which "I have taken it and I feel fine" carries no information.

Recommended Dosages

Nothing in this section is a dose, and no number below describes an amount of zirconium to take. Every figure here is a regulatory ceiling on an exposure people cannot avoid.

Occupational air, United States

The OSHA permissible exposure limit for zirconium compounds as Zr is 5 milligrams per cubic meter as an eight hour time weighted average, set in 29 CFR 1910.1000 Table Z-1, with no short term limit and no ceiling. The NIOSH recommended exposure limit is 5 milligrams per cubic meter over up to 10 hours with a 10 milligrams per cubic meter short term exposure limit, and expressly does not cover zirconium tetrachloride. The ACGIH threshold limit value matches at 5 time weighted average and 10 short term, with the A4 notation, not classifiable as a human carcinogen. The NIOSH immediately dangerous to life or health value is 25 milligrams per cubic meter as Zr. Sampling runs by NIOSH methods 7300, 7301 and 9102 or OSHA method ID121.

Occupational air, for contrast

Hafnium, the element directly below zirconium and its unavoidable geochemical companion, carries an OSHA permissible exposure limit of 0.5 milligrams per cubic meter, ten times stricter. Commercial zirconium metal contains 1 to 3 percent hafnium, and the two are separated only when the application demands it.

Oral, United States

EPA's December 27, 2012 assessment derived a screening chronic provisional reference dose for soluble zirconium compounds of 8 times 10 to the minus 5 milligrams per kilogram per day, from the 0.79 milligram per kilogram per day rat lowest observed adverse effect level with a composite uncertainty factor of 10,000. The same value was adopted as the screening subchronic value because no subchronic oral study exists. EPA declined to derive a full reference dose at all, citing database limitations, and declined to derive any inhalation reference concentration. A screening value carrying a 10,000-fold uncertainty factor states that the data are inadequate, not that an intake is safe.

Topical, United States and European Union

Up to 20 percent aluminum zirconium complex on an anhydrous basis in nonaerosol antiperspirants under 21 CFR 350.10, and zero in aerosol cosmetics under 21 CFR 700.16. The EU sets the same 20 percent plus a second ceiling of 5.4 percent expressed as zirconium, with atomic ratio requirements and an aerosol prohibition.

Food contact

Annex II of Regulation (EU) No 10/2011, as amended through Regulation (EU) 2020/1245, sets specific migration limits for 19 metals into food from plastics, from iron at 48 and copper at 5 milligrams per kilogram of food down to nickel at 0.02 and lead, mercury and arsenic at not detectable. Zirconium is not on that list. The absence is not a clearance; it reflects that zirconium was not identified as a migrating metal needing a limit.

The one oral regimen that exists

Sodium zirconium cyclosilicate has a prescribed regimen. Its numbers sit in the Forms section above, because they describe a supervised drug schedule rather than an intake anyone should copy.

Duration

For occupational exposure the relevant duration is a career, which is why the limits are written as working lifetime averages. For voluntary oral intake there is no duration, because there is no appropriate first dose.

Timing & Administration

There is no timing or administration guidance for zirconium, because there is no intake to time. The only administration schedule that exists belongs to sodium zirconium cyclosilicate, and it matters for one reason: other oral medications with pH-dependent solubility should be given at least 2 hours before or 2 hours after a dose, because the binder transiently raises gastric pH. Drugs without pH-dependent solubility need no spacing. That is a prescribing instruction, not a supplement protocol.

Timeline of Effects

The only timelines worth publishing for zirconium are the latencies of harm, measured in weeks to decades rather than hours.

Delayed pulmonary edema from zirconium tetrachloride: up to 36 hours after exposure, per the UK poisons monograph, the shortest clinically important interval in the record.

Axillary granuloma after dermal contact: onset delayed days to weeks from exposure to visible red-brown firm papules, histologically resembling sarcoidosis.

Granuloma resolution after exposure stops: complete recovery within 6 to 12 months once exposure ceases, longer with insoluble forms. Steroids may hasten resolution. Ending exposure is the treatment.

Occupational lung disease: the 1993 hypersensitivity pneumonitis case developed after 3.5 years of tile sorting and glazing in a 25 year old nonsmoker. Chronic sarcoid-type granulomatosis in the older literature followed exposure measured in years to decades.

Sodium zirconium cyclosilicate: 48 hours to the primary endpoint in the phase 3 trial, with maintenance assessed through day 12. That is a pharmacodynamic timeline for a prescription potassium binder, and the only properly measured oral zirconium timeline in existence.

There is no fair trial period for zirconium supplementation, because there is no endpoint to measure and no hypothesis under test.

Benefits of Taking Zirconium

Taking zirconium has produced no demonstrated benefit in any species, for any indication, at any dose, and no trial in the literature or the registry set out to look for one. What follows is the history, told as history.

The antiperspirant era, 1950s. Sodium zirconium lactate entered deodorant sticks because zirconium salts are astringent and occlude sweat ducts effectively. Within about two years the dermatology literature had described the granulomas, and the compound left the category.

The aerosol era, 1960s and 1970s. Aluminum zirconium complexes were formulated into aerosol antiperspirants, documented in a run of patents through the period. FDA closed that route with 21 CFR 700.16 in 1977 on the stated basis of human skin granulomas plus animal lung and organ toxicity, and the deep lung reach of aerosolized particles. Nonaerosol aluminum zirconium antiperspirants were not restricted and remain the dominant category today.

The trace mineral era. Schroeder and Balassa's 1966 paper established that zirconium is present in human tissue at measurable levels. That descriptive finding got absorbed into the "every element in the body must be needed" logic that drives colloidal and plant-derived mineral marketing. Schroeder's own later animal work, the drinking water studies EPA used in 2012, found adverse effects rather than benefits.

The one real therapeutic use. Sodium zirconium cyclosilicate works, and it works because it holds potassium and leaves in the stool. FDA approved it on May 18, 2018 after rejecting the application twice. That is a genuine achievement, and an argument against oral zirconium exposure rather than for it.

Potential Negatives & Side Effects

The adverse effects of zirconium compounds sort cleanly by route, and mixing the routes is how this subject gets misreported.

Dermal. Granulomatous hypersensitivity reaction, documented with sodium zirconium lactate in the 1950s and with aluminum zirconium complexes in the 1993 Skelton and 1997 Montemarano reports. Lesions are firm shiny erythematous papules, usually axillary, histologically sarcoid-like.

Pulmonary. Granulomatous interstitial pneumonia and hypersensitivity pneumonitis in the 1993 ceramic tile worker case, and chronic sarcoid-type granulomatosis in long exposed workers in the older literature. The UK monograph notes that a causal link between lung granulomas and delayed hypersensitivity remains inconclusive, a real limitation and not a reason to discount the cases. Several epidemiological series found no significant lung impairment: 32 metal handfinishers, 22 reduction plant workers, and 178 men mostly below 10 milligrams per cubic meter with a maximum of 30.0.

Corrosive. Zirconium tetrachloride and soluble salts can cause burns of the mouth and throat, vomiting, bloody diarrhea, collapse and convulsions on ingestion, and airway irritation with delayed pulmonary edema on inhalation. Ocular contact produces irritation, lacrimation, blurred vision and conjunctivitis, with no case series behind that description.

Oral, prescription. In the sodium zirconium cyclosilicate placebo controlled trials up to 28 days, edema occurred in 4.4 percent at 5 grams daily, 5.9 percent at 10 grams and 16.1 percent at 15 grams, against 2.4 percent on placebo, and in 8 to 11 percent in longer uncontrolled trials. Hypokalemia below 3.5 milliequivalents per liter occurred in 4.1 percent of non-dialysis patients. The label ties its edema warning to the 400 milligrams of sodium in each 5 gram dose, not to the zirconium.

Oral, supplemental. No adverse event dataset exists, because there are no trials and no required reporting of the amount in the product.

Deficiency Symptoms

What zirconium addresses

Nothing. Low zirconium is not a clinical state, and the concept fails at the first step, because a deficiency requires a function that stops working when the element is withdrawn.

Compare a genuine ultratrace element. Molybdenum has an adult Recommended Dietary Allowance of 45 micrograms per day, a Tolerable Upper Intake Level of 2,000 micrograms per day, a defined cofactor, and enzymes that require it including sulfite oxidase, xanthine oxidase and aldehyde oxidase. None of that exists for zirconium.

Bottom line

No test result, symptom, or diet pattern establishes a zirconium shortfall, and no supplement corrects one.

Toxicity Symptoms

Zirconium toxicity is an exposure story rather than an overdose story, and the signs below belong to specific compounds and specific routes.

At high intake

• Inhaled dust above the 5 milligrams per cubic meter occupational limit over years: cough, dyspnea, and on biopsy granulomatous interstitial pneumonia. The NIOSH Pocket Guide lists skin and lung granulomas as the human effects, with skin and respiratory system as target organs.

• Zirconium tetrachloride inhalation: upper airway irritation, then pulmonary edema that can be delayed as much as 36 hours.

• Soluble salt ingestion in quantity: corrosive injury to mouth, throat and gut. The UK monograph states that no reports of zirconium intoxication from ordinary ingestion exist, which records an absence of case reports rather than evidence of safety.

• Intravenous administration, historical: two people developed vertigo after 50 milligrams of zirconium malate intravenously. No current product uses this route.

• Dust at or above 25 milligrams per cubic meter: the NIOSH immediately dangerous to life or health threshold.

Signs to reduce or stop

For an exposed worker: new or worsening dry cough, exertional breathlessness, or any papular eruption at a contact site. For anyone using a zirconium antiperspirant: persistent firm papules in the axilla that do not behave like ordinary irritation, which warrants a biopsy and a product change rather than a stronger steroid cream. For anyone taking a supplement that lists zirconium, the signal to stop is the ingredient itself, not a symptom.

General note

Zirconium is not acutely potent by mouth the way a heavy metal is, which is precisely what makes the hazard easy to understate. The documented harms are granulomatous, latent and in the lung potentially permanent. Low acute toxicity and low long term risk are different claims, and only the first is supported here.

How Zirconium Works

Zirconium does not work in the sense this section usually means, so what follows is how its compounds behave once they reach tissue.

Hydrolysis and insolubility. Zr4+ is strictly tetravalent with a high charge to radius ratio, and near physiological pH it forms polynuclear hydroxo species and precipitates as hydrated oxide. That one property explains the poor gastrointestinal absorption, the poor penetration of intact skin, and the persistence of deposited particles.

Distribution and elimination. The UK monograph reports that in four autopsies of people who were not poisoned, zirconium was widely distributed with the highest concentrations in liver and body fat, and that most absorbed zirconium leaves in bile with small amounts in urine. The 1992 Ghosh review describes initial retention in soft tissue followed by slow accumulation in bone, and reports passage across the blood brain barrier and the placental barrier into milk, from animal work.

Delayed hypersensitivity. The granuloma is an immune response, not a chemical burn. Kang, Bice and Salvaggio tested delayed hypersensitivity to zirconium compounds alongside beryllium sulfate in rabbits in the Medical Journal of Osaka University in 1976, volume 26, pages 131 to 145. The human picture, a latency of days to weeks with sarcoid-like histology, fits a cell mediated mechanism.

Particle burden. For the lung the operative variable is retained particulate mass, which is why the 1993 tile worker case was quantified as a dust burden near 100 times background rather than as a blood concentration. Zirconium tetrachloride was negative in Salmonella strains TA98, TA100, TA102, TA1537 and TA2637, and zirconium oxychloride in TA97, TA98, TA100, TA1535 and TA1537; nothing is reported for zircon or zirconia.

Synergistic Supplements

None. No combination, cofactor, carrier or "enhanced absorption" pairing makes zirconium intake worthwhile, and the only pairing question with clinical content is the two hour dose separation around prescription sodium zirconium cyclosilicate.

Interactions & What NOT to Take

No oral zirconium supplement has a documented interaction profile, because no one has studied one. The interactions that exist belong to the prescription compound and to the physical chemistry.

Prescription sodium zirconium cyclosilicate. The label directs that other oral medications be given at least 2 hours before or 2 hours after a dose when they have pH-dependent solubility, since the binder transiently raises gastric pH. In the label's pharmacology section, systemic exposure increased for the weak acids furosemide and atorvastatin, decreased for the weak base dabigatran, and did not change for losartan, glipizide or levothyroxine. Drugs without pH-dependent solubility require no spacing.

Industrial handling. Zirconium tetrachloride must be kept from water, which is why it ships and stores under moisture exclusion. Zirconium metal powder is incompatible with potassium nitrate and oxidizers.

What not to take. Any product that lists zirconium as a mineral ingredient. These products are almost always blends, so the decision is to decline the whole bottle rather than subtract one line from the panel.

Quality, Testing & Adulteration

The quality problem with zirconium is not contamination of a legitimate ingredient. It is that the ingredient is on labels at all, with no number attached to it.

What the labels actually say. Searching the NIH Dietary Supplement Label Database for zirconium returns 105 labels out of roughly 215,000. Of those, 58 name zirconium as an ingredient, and of the 56 whose Supplement Facts rows I could parse, every one lists it inside a proprietary blend row with no quantity. Named examples include Buried Treasure Pure Colloidal Minerals, Drucker Labs intraMIN Organic Trace Minerals, Healthy Natural Systems Liquid Colloidal Minerals, Tropical Oasis Mega Multiple Vitamin Mineral and Sunwarrior Silver Strength plus Gold. Eighteen of the 60 labels in the set are flagged off market, including one of two Drucker Labs intraKID listings.

What third-party testing covers and does not. USP and NSF certification verifies identity, declared content, dissolution and contaminant limits against a monograph. There is no dietary supplement monograph for zirconium, so there is nothing to verify against. The standard elemental impurity work in the compendia addresses arsenic, cadmium, mercury and lead. Zirconium is not in that panel, so a clean heavy metals certificate says nothing about it. Before concluding that any material fails any specification, confirm that the assay and the specification cover the same elements, because here they frequently do not.

Where real specifications do exist. For antiperspirant actives, the United States Pharmacopeia and National Formulary sets the atomic ratios 21 CFR 350.10 requires. For surgical ceramics, ISO 13356 sets phase stability, strength, fatigue and radioactivity requirements. Both are enforceable and both govern products other than supplements.

What a buyer can settle without a laboratory. Look for an amount next to the word zirconium. If there is no amount, the manufacturer has committed to nothing, and no certificate of analysis will be anchored to anything. Then look for the compound name. A panel that cannot say whether it holds an oxide, a silicate or a soluble salt is describing an element rather than an ingredient, and the three differ by orders of magnitude in how they behave.

Special Considerations

Pregnancy and lactation. No human data. The 1992 Ghosh review reports placental transfer and appearance in milk from animal work. With no benefit to weigh against that, the decision is trivial.

Children. No data, no indication, and at least one zirconium-listing children's product recorded in the label database. Absent pediatric data is not reassurance.

Kidney disease. The only systemic warning on US antiperspirant labels concerns kidney disease, and the prescription compound's edema signal is most relevant in renal disease and heart failure.

Dental and orthopedic patients. A zirconia crown or bearing surface is a solid ceramic held to ISO 13356, and nothing here argues against it. The 2001 Prozyr suspension was a manufacturing and phase stability failure in specific lots, not a verdict on the material class, and the standard's accelerated aging test exists because of it.

Occupational surveillance. The practical content of surveillance is respiratory: symptom history, spirometry and imaging. A urine or blood zirconium level has no action threshold attached. The 1993 index case was a nonsmoker, so smoking is not required for this pathology, and anyone with existing interstitial disease working in zircon dust has a worse margin than the limits assume.

Research Status & Evidence Quality

Tier 1, strong human evidence. Only one body of work qualifies, and it concerns the prescription compound. Packham and colleagues randomized 753 patients with hyperkalemia across four doses of sodium zirconium cyclosilicate and placebo in the New England Journal of Medicine, volume 372, pages 222 to 231. Mean serum potassium fell 0.7 millimoles per liter at 48 hours in the 10 gram arm against 0.3 on placebo, and normokalemia held through 12 days of once daily maintenance. Adverse event rates were similar between arms, 12.9 percent against 10.8 percent in the initial phase, diarrhea most common in both.

Tier 2, moderate evidence. The dermal record. Two well characterized case reports of aluminum zirconium granuloma in 1993 and 1997, a 1950s experimental pathology literature on sodium zirconium lactate, and a regulatory finding of human skin granulomas sufficient for FDA to act in 1977.

Tier 3, weak or mixed. The pulmonary record. One detailed 1993 hypersensitivity pneumonitis case with particle analysis, older sarcoid-type granulomatosis reports, and several small negative epidemiological series. Causality for the lung granulomas is called inconclusive in the poisons literature.

Tier 4, no evidence. Oral zirconium as a supplement. A ClinicalTrials.gov search for zirconium returns the prescription potassium binder, zirconia dental and joint implant studies, and zirconium-89 PET imaging agents. Supplementation is not among them, and PubMed turns up no trial or case series either. IARC has not reviewed zirconium. It is not in the National Toxicology Program Report on Carcinogens. ATSDR has published no toxicological profile for it. EPA classified it as having inadequate information to assess carcinogenic potential.

Research Limitations

The oral database is two lifetime rodent studies from one laboratory, each with a single dose level, so neither produces a no-effect level, and the rat study was confounded by a pneumonia outbreak. EPA's 10,000-fold uncertainty factor quantifies how little confidence that supports. The dietary intake figures date to 1966 and span 4.15 to 125 milligrams per day depending on the review. Human absorption rests on nothing EPA could cite: its 2012 assessment reports no gastrointestinal absorption percentage and names no study on the question. The pulmonary cases lack exposure reconstruction good enough for dose response. And the one well studied product is well studied precisely because it does not get absorbed, so it says nothing about what absorbed zirconium does.

Summary & Key Takeaways

Bottom Line

Zirconium is not a nutrient, has no function in human biology, and belongs in ceramics, foundries, reactors and dental laboratories rather than in anything swallowed on purpose. The federal record consists of one cosmetic rule from 1977, an occupational air limit of 5 milligrams per cubic meter, a 20 percent cap on nonaerosol antiperspirant actives, and a screening oral value EPA built with a 10,000-fold uncertainty factor and declined to finalize. A Supplement Facts panel listing zirconium is sufficient reason to put the bottle down.

Key Safety Points

• 21 CFR 700.16 makes any aerosol cosmetic containing zirconium adulterated. It is not a general ban, and nonaerosol aluminum zirconium antiperspirants remain legal at up to 20 percent.

• The documented human harms are granulomas of skin and lung, delayed by weeks to years, and lung granulomas may be advanced before anyone notices.

• Zirconium tetrachloride is a different hazard class from the rest: water reactive, corrosive, excluded from the NIOSH limit, with pulmonary edema that can arrive 36 hours late.

• Keep zirconium products off broken or irritated skin. Both the US monograph and the EU annex require that warning.

• Zirconium, hafnium, zircon the mineral and zirconia the implant ceramic are four different things toxicologically. Hafnium's occupational limit is ten times stricter at 0.5 milligrams per cubic meter.

• The occupational and dermal evidence does not transfer to oral intake, and the oral evidence is two confounded rodent studies.

Special Note

The one zirconium compound with an approved oral indication, sodium zirconium cyclosilicate, was engineered so the zirconium stays in the gut and leaves in the stool, and the label reports urine and blood zirconium at or near the assay's detection floor in treated patients. When the only oral zirconium product a regulator ever cleared is built around keeping zirconium out of the bloodstream, a mineral blend that puts an unstated quantity of it in your mouth has the logic exactly backward.

Researched and drafted with AI assistance. Every claim verified against primary sources and human-reviewed before publication. Last reviewed: October 15, 2026. Corrections: reply to any issue.

This issue is educational and is not medical advice. Zirconium is covered here as a toxicology and regulatory subject, not as a supplement worth considering. If you take prescription medication, have a diagnosed kidney, heart or lung condition, or have an occupational exposure you are concerned about, talk to a clinician before changing anything.