The crack that stops itself in a zirconia crown

Zirconia is used in the body because of a phase change that expands where a crack is starting — a mechanism with a documented reverse gear, and a rare earth holding it in place.

Reviewed by Peter Uppal

A single small white ceramic crown with contoured cusps and grooves resting alone on a plain grey surface, a soft shadow beneath it. — Illustrative artwork: a ceramic dental crown of the kind produced using engineered zirconia. Not a facility, equipment or material connected to this project. · Illustration · Osmond Hub

The short version

Zirconia used in the body is not plain zirconium oxide. It is yttria-stabilised tetragonal zirconia, held in a crystal phase it would not otherwise keep at body temperature, and its toughness comes from letting that phase change happen exactly where a crack is trying to start. The Sheffield account describes the tetragonal-to-monoclinic transformation as producing a 2–5% volume expansion and 16% shear strain(opens in a new tab), which squeezes the crack shut behind its own tip. The stabiliser doing the holding, yttrium oxide, is itself a rare earth.

A ceramic that responds to being damaged

Ceramics fail by cracking, and a crack in a brittle material normally runs. Zirconia is unusual because it carries a stored structural change that the crack itself triggers.

At the temperatures a body operates at, the tetragonal phase of zirconia is not the stable one — the monoclinic phase is. Adding yttria holds the tetragonal grains in place anyway, in a metastable state. The dental review describes the classic 3 mol% Y2O3 formulation as maintaining a fully tetragonal, fine-grained structure(opens in a new tab), and the Sheffield thesis describes a 'Partially stabilized zirconia powder with uniform dispersion of 3 mol % Yttria'(opens in a new tab).

What happens next is the mechanism. The review describes grains that, under stress, locally transform to the monoclinic phase, producing a volumetric expansion that impedes crack propagation(opens in a new tab). The thesis names the consequence: the expansion associated with the transformed zone in the wake of the crack results in closure of the crack, referred to as 'crack shielding'(opens in a new tab). The material gets in the crack's way by swelling into it.

The numbers do not agree, and that is worth saying

How much it swells depends on which paper you read. The thesis gives 2–5 % volume expansion(opens in a new tab). The 2025 dental review gives a 4–5% volumetric expansion in one passage(opens in a new tab) and, elsewhere in the same paper, a 3–5% expansion(opens in a new tab). None of these is a point value, and no source read for this article settles them into one.

The strength figures behave similarly. The review reports that 3Y-TZP typically exhibits a flexural strength of 900–1300 MPa(opens in a new tab) while 5Y-PSZ usually has a flexural strength of about 500–800 MPa(opens in a new tab) — the more translucent formulation trading strength for appearance. These are ranges with their own hedges, and they are reproduced with them.

The material's advantage is that it responds to damage. Its weakness is that it can make the same move for no good reason, slowly, in the wet.

The reverse gear

If a stress-triggered phase change is the strength, an untriggered one is the problem. The review describes low-temperature degradation, in which water molecules penetrate the zirconia lattice and facilitate spontaneous tetragonal-to-monoclinic conversion at the surface, causing microcracks, surface uplift, and gradual strength loss over time(opens in a new tab). The transformation still happens; nothing is holding a crack shut, and the material simply ages.

This is not theoretical. The Sheffield thesis records that the most high-profile zirconia failure occurred in 2001, when more than 350 Y-TZP implanted hip prostheses failed requiring revision surgery(opens in a new tab). What followed is on the record here as a regulator's act rather than a voluntary one by the manufacturer: the republished French regulatory decision of 10 August 2001 counts a different thing again — 56 reports of incidents reporting rupture of heads from the affected batches(opens in a new tab) — and the two figures are not the same quantity. Both are given here; neither is merged into the other.

That episode is why the standards layer exists in the form it does. ISO 13356:2015 specifies 'the requirements and corresponding test methods for a biocompatible and biostable ceramic bone-substitute material based on yttria-stabilized tetragonal zirconia (yttria tetragonal zirconia polycrystal, Y-TZP) for use as a material for surgical implants'(opens in a new tab) — a standard written for one material in one clinical role.

Back to the mineral

The chain to a mineral sands deposit is short and worth stating precisely. The Zircon Industry Association describes zirconia being made by heating zircon with coke past its fusion point in an electric arc furnace, where it dissociates to zirconium oxide and fumed silica(opens in a new tab) — one of several routes it names.

Osmond's published work stops well short of that. Its preliminary metallurgy produced a zircon concentrate grading 50.2% ZrO₂ at approximately 70% recovery, against a stated premium target of ≥66%(opens in a new tab). Premium-grade zircon is a chemical specification for a mineral concentrate, not a medical one, and nothing in the record addresses whether material from this deposit could serve an implant chain. It is a long way from a concentrate grade to a standard written about crowns.

Exploration results and mineralogical estimates only. Orión has no JORC-compliant Mineral Resource or Reserve; maiden MRE and Scoping Study pending, targeted Q3 CY26.

Related