"Pt. referred for crown on #27. RCT 3 wk ago after cracked DL cusp. Pt. chewing right side only x12 mo. Wife reports grinding. Would like crown ASAP, prefers shade close to #26, no splint yet."
What that looked like in practice:
- A woman in her early fifties. The upper left second molar had failed on her about a year earlier, when a cusp cracked off while she was eating. She had been avoiding that side ever since — chewing on the right, which the attending clinician could see in the wear pattern on her lower right molars.
- A root canal procedure was completed three weeks before the scan reached us. Access cavity sealed with a bonded composite core. No post placed. Pulp was non-vital, so sensitivity was not a complaint.
- The remaining tooth was short. Not short the way a prepped first molar is short — short in the sense that the cracked distolingual cusp had taken most of that wall with it, and the preparation that remained was asymmetric. The lingual was the problem corner.
- Ferrule circumferentially 1.4 mm — below the 1.5–2 mm usually wanted on a root-canalled posterior tooth. On the distolingual it was closer to 1.0 mm.
- Margin sat 0.4 mm subgingival distally, where the fracture line had dropped below the gum; the rest was equigingival, clean, dry.
- Antagonist was #37, natural enamel, with visible wear facets on the buccal cusps — consistent with the grinding history her husband had been complaining about.
- No night guard in use. The attending clinician had recommended one twice and the patient had not yet gone to the appointment to have one made.
- Shade memory. The patient wanted the new crown to look like #26 next door. The dentist's photograph showed #26 at roughly A3. The stump under #27, however, scanned at A3.5 — darker than the adjacent crown's final shade. The gap between those two numbers was going to decide how this crown looked.
Here is what the scan actually measured once it was open on our CAD station, next to the targets we hold against on a zirconia crown for posterior teeth:
Parameter | Measured | Target | Status |
|---|---|---|---|
Axial wall — buccal | 3.4 mm | ≥ 4.0 mm | Below |
Axial wall — lingual | 2.6 mm | ≥ 4.0 mm | Below |
Axial wall — mesial | 2.2 mm | ≥ 4.0 mm | Below |
Axial wall — distal | 2.9 mm | ≥ 4.0 mm | Below |
Total occlusal convergence (TOC) | 22° | 10–20° | Over |
Occluso-cervical to faciolingual ratio | 0.33 | ≥ 0.40 | Below |
Occlusal clearance (thinnest point, mesial) | 0.9 mm | ≥ 1.0 mm | Below |
Occlusal clearance (central groove) | 1.3 mm | ≥ 1.0 mm | OK |
Margin readability on the scan | 360° readable, distal slightly obscured | Readable | Partial |
Stump shade | A3.5 | — | Dark background |
Antagonist | #37 natural enamel, wear facets | — | — |
Bruxism (reported) + splint | Yes / No | — | Risk |
The short version: three of the four retention numbers were out, and the fourth (gap and clearance) would be at risk the moment a crown had to sit there with any kind of anatomy on it.
2. What the attending clinician asked for
The prescription itself, trimmed:
- Material: monolithic zirconia, "high strength so it won't break."
- Shade: A2.
- Finish: glazed, matched to #26.
- Cementation plan: conventional glass ionomer.
- Timeline: "as soon as possible."
Three of those lines disagreed with the tooth, and we said so in writing before design started.
- A2 on an A3.5 stump through a thin 3Y wall. Zirconia gets more translucent as it gets thinner — in-vitro work on monolithic zirconia across thicknesses shows the translucency parameter roughly doubling from 1.5 mm to 0.5 mm. A dark background bleeds through at the thin end. The A2 target was not going to hold.
- "Won't break" as the reason to pick zirconia. Fracture was not the risk here. Retention was. The published pattern on short-walled second molars is a debonding one, not a cracking one.
- Glass ionomer on a 2.2–3.4 mm preparation. That cement choice assumes mechanical retention. This tooth did not have mechanical retention.
The attending clinician read the write-up, agreed, and we revised the plan together before CAD opened the design file.
3. The four numbers that decided this crown
A zirconia crown for posterior teeth on this kind of tooth stands on four numbers. Three came from the clinician's side of the chair. One is a parameter we set in CAD before the mill runs.
# | Number | On this case |
|---|---|---|
1 | Axial wall height | Shortest 2.2 mm, longest 3.4 mm — all four below the 4 mm molar target in the Goodacre preparation geometry work |
2 | TOC | 22° — past the point where cutting grooves recovers resistance form |
3 | Occlusal clearance | 0.9 mm at the thinnest point, 1.3 mm at the central groove |
4 | Internal gap / edge gap | Our CAD setting, pulled to 60 µm / 25 µm on this case, tighter than our 80 µm default, following retention data on all-zirconia crowns |
Numbers 1–3 were the attending clinician's. Number 4 was ours. We did not spend our budget on anatomy, strength or shade — we spent it on retention and seating.
4. How we built the crown
Seven decisions, tied to this specific tooth.
4.1 Asked the clinician to re-prep the cervical third
A TOC of 22° cannot be rescued by sticking a groove on it. Grooves respond to displacement, not to taper. We asked the clinician to re-prep the gingival third to bring convergence back to roughly 16–18°, which took about 0.3 mm of tooth off the proximal walls. She did that at a short follow-up visit the patient was already scheduled for, and the second scan showed TOC at 17°.
4.2 Added a distal box, not a buccolingual groove
Once the taper was workable, we designed a shallow distal box, not a groove. On this preparation the risk was rotational and buccolingual, which is what a box resists. A buccolingual groove resists mesiodistal movement, a different force vector that was not the issue here.
4.3 Switched to adhesive seating
Even after the re-prep, the shortest wall was still 2.4 mm (lingual re-measured). We did not pretend that was 4 mm. The seating moved to the interface — air-abrasion + MDP + resin cement, which is the current consensus for bonding to zirconia:
- 50 µm alumina airborne-particle abrasion at ~2 bar
- 10-MDP primer on the clean, matte intaglio
- Dual-polymerising resin cement, matched to the primer
We also noted on the dispatch sheet, in bold: "Clean the intaglio after try-in with a zirconia cleaner; do not rely on alcohol." Chairside contamination can cost up to 50% of bond strength. On a tooth we were already bonding, we did not want that half back.
Hydrofluoric acid did not come into any of this. Zirconia is a polycrystalline oxide with no glass phase, so HF does not etch it the way it etches lithium disilicate.
4.4 3Y zirconia, not 4Y
Nocturnal grinding plus 0.9 mm of space on the mesial. Dropping to 4Y buys translucency and loses strength. There was no aesthetic budget on this tooth to spend on translucency. 3Y zirconia keeps flexural strength in the 1,000–1,500 MPa range reported in the state-of-the-art review of dental zirconia and is the standard pick for a monolithic case like this one.
4.5 Asymmetric anatomy, declared in writing
Something had to give at 0.9 mm. We held the functional cusp at 1.0 mm minimum thickness and dropped the mesial marginal ridge — the result is a crown that does not mirror #26's anatomy at that ridge. We said so in writing, with a side-by-side CAD view attached, before the mill ran. The clinician signed off.
4.6 Shade: A3 with reduced translucency, not A2
A3.5 stump + thin 3Y wall cannot read as A2 without an opaque layer that would show as a white patch next to the A3 neighbour. We offered A3 with reduced translucency, held characterisation subtle, and asked the clinician to confirm with the patient at try-in. Background structure affects posterior zirconia colour perceptibly — the in-vivo colour work on zirconia-based all-ceramic crowns is clear on that.
4.7 Polished, not glazed
Mechanical polishing with a zirconia-specific sequence. No glaze layer, despite the prescription asking for glazed. Polished monolithic zirconia produces antagonist enamel wear at or below the rate of natural enamel, and about half the wear of glazed zirconia in direct comparison. A glaze layer eventually wears through and exposes a rougher substrate. For an opposing arch that already had wear facets, polished was the right finish.
Istar's full workflow for this case — case receiving → CAD design review → model print → CAM milling → sintering → finishing/polishing → three QC gates → packing — all ran inside our Shenzhen facility. Nothing on this case was sub-contracted.
5. What left the bench
Measured on the day of dispatch, logged against the unique case ID:
Item | Set | Measured on dispatch |
|---|---|---|
Marginal gap | 25 µm | 41 µm |
Internal gap | 60 µm | 58 µm |
Occlusal contacts at maximum intercuspation | 3 | 3, verified after confirmed seating on the die |
Proximal mesial contact | Light | Light; floss passes with resistance |
Proximal distal contact | Light–medium | Light–medium; floss passes with resistance |
Intaglio | Matte, uniform | Matte, uniform under 10× magnification |
Shade | A3, reduced translucency | A3 reduced translucency, photo on file |
Two things worth saying out loud about that table.
- The margin came in at 41 µm, not 25. Milling plus sintering adds real-world variance to any CAD setting. 41 µm sits comfortably inside the clinically acceptable band — a 2025 systematic review of finish-line effects on fit places chamfer, rounded shoulder and shoulder designs between roughly 55 and 73 µm, with no statistically significant difference between finish-line designs. We log the measured value because that is what the clinician will encounter.
- Occlusion was checked only after seating was confirmed. On a short-wall case, a crown that reads as "high" is often a crown that is not fully seated. Chasing occlusion first is how a technician ends up adjusting a problem that does not exist.
The case shipped on day 4 after receipt, inside Istar's standard 2–3 day design plus 2–3 day production window for a single posterior unit.
6. At chairside try-in — what the dentist and the patient said
The attending clinician sent her try-in notes the same day. In her own words:
Dry fit at 2 min. Full seat on first attempt, no spring-back. Mesial contact light but closes with floss resistance. Occlusion checked after seat confirmed — three MI contacts, no excursive interference. Showed Pt. the shade result in the mirror. Pt. said it was 'slightly darker than #26 but closer than she expected,' and that she liked that it 'did not look fake white.' Reminded her we had agreed this in writing before milling. She was fine. Will proceed to cement."
A few things that matter in those lines, from a lab point of view:
- Full seat on first attempt means the 60 µm internal setting and the 41 µm margin were working together. On short-wall cases we would rather see this result than a theoretical 25 µm margin that will not seat.
- "Slightly darker than #26 but closer than she expected" is what A3 reduced translucency over an A3.5 stump actually looks like — the opposite result of what she would have got if we had milled A2 on a dark preparation. On a thin 3Y wall, dark stump colour pulls everything toward itself. Meeting the stump in the middle is the honest option in CAD.
- "Did not look fake white" is a patient-side observation that quietly validates the polished, un-glazed finish. A chased mirror glaze would have added brightness this patient was already uncertain about.
- No excursive interference is the output of the polishing-only finish plus the asymmetric mesial ridge. A full mirror of #26's anatomy would almost certainly have added a working-side contact that was not meant to be there.
The attending clinician cemented the same visit with the adhesive protocol we had specified. She confirmed by WhatsApp that the Ivoclean step and the MDP primer were on hand before cementation started.
7. If your next posterior zirconia case looks like this one
Short-walled second molar. Dark stump under a bright-shade request. A root-canalled molar with a modest ferrule. A grinder who has been avoiding one side for a year. These cases come in regularly, and most of the thinking is reusable.
Send the scan and the ten prescription fields below. Istar's case coordinators will come back inside one working hour (9:30–18:00 China time) with the wall heights, convergence and ratio we measure, anything that falls outside threshold, and the design options we see — before anything is designed.
# | Field | Why we ask | Required |
|---|---|---|---|
1 | Tooth position (arch + number) | Load, access, likely wall height | ✅ |
2 | Remaining axial wall height (mm) | First retention variable; we re-measure on the scan | ✅ |
3 | Occlusal clearance (mm, thinnest point) | Below 1.0 mm the anatomy has to give | ✅ |
4 | Stump shade | Drives the masking decision | ✅ |
5 | Target shade | Only meaningful alongside #4 | ✅ |
6 | Antagonist (material + wear status) | Predicts long-term risk | ✅ |
7 | Bruxism / parafunction + splint status | Changes generation choice and thickness | — |
8 | Pulp status, post, ferrule height | Whether the geometry can be recovered at all | — |
9 | Margin location (supra- / equi- / subgingival mm) | Predicts whether the scan is readable | — |
10 | Material generation preference (3Y / 4Y / multilayer) | Not interchangeable at the same thickness | — |
Istar accepts STL, PLY, OBJ and DCM out of 3Shape TRIOS, Medit, iTero, CEREC, Shining3D, Dexis, Primescan, Panda and Planmeca — any intraoral scanner that exports an open-standard file. Physical impressions ship to our Shenzhen facility; we issue the inbound label.
One hard line on scans: if the margin is not continuously readable around 360°, we will say so and stop. We will not estimate a line we cannot see. A re-scan is faster than a remake — the published work on intraoral scanning confirms this is not a stylistic preference, subgingival accuracy drops measurably when the sulcus is not open and dry.
8. Turnaround, remakes, traceability
- Design turnaround. 2–3 days from receipt of a complete case.
- Production. Scheduled from the day a case is complete. Weekends and Chinese public holidays do not count toward the committed window, and time spent waiting on data or clarifications does not either — the clock restarts when the case is complete.
- Remakes. If a remake is on us, it is free, prioritised, and usually finished in 3–5 working days. If a re-prep is on the clinician, new material and shipping may apply and we discuss it before starting.
- Traceability. Every case gets a unique ID that binds the responsible technician, the equipment used, the material batch, and the QC record. Records are kept on file for three months by default and longer on request.
- In one place. Case receiving, CAD, model printing, CAM milling, zirconia processing, layering/glazing where applicable, polishing and QC all happen inside Istar's Shenzhen and Huizhou facilities. Cases are not sub-contracted.
9. FAQ
Why do zirconia crowns on second molars keep coming off?
Almost always geometry, not material. A 2–3 mm axial wall cannot generate the friction a crown needs, and once TOC is past 20° or so, adding grooves does not recover resistance — the classic preparation geometry work is still the clearest reference on this. Where the geometry cannot be recovered, adhesive seating through alumina abrasion and a 10-MDP primer is how a zirconia crown for posterior teeth stays on.
What is the first thing to check when a zirconia crown debonds?
In order: Was the intaglio abraded, or does it still look mirror-bright? Was it cleaned after try-in — because contamination can halve the bond strength? What was the internal gap? Was the preparation retentive? Was the cement system matched to the retention available? A cement choice on its own is rarely the whole story.
Does stump shade really matter on a back tooth?
More than it looks. Posterior zirconia is thinner than people expect at the margins, and translucency rises as thickness falls — the TP of a typical zirconia roughly doubles going from 1.5 mm to 0.5 mm. A dark stump bleeds through at the thin end. Record the stump shade whenever the restoration is high-translucency, the walls are thin, the stump is dark or sits over a metal post, or the adjacent teeth are unusually light.
Does a zirconia crown wear down the opposing tooth?
A polished monolithic surface wears opposing enamel at a rate at or below natural enamel. A glazed surface produces roughly twice the wear of a polished one because glaze wears through and exposes a rougher substrate. Finishing an occlusal adjustment with a zirconia polishing kit, not a glaze layer, is why antagonist wear is usually a non-issue on a properly finished crown.
Monolithic or layered zirconia for a back tooth?
Monolithic by default. Not because it is stronger in the abstract, but because it removes the zirconia–porcelain interface and a whole class of chipping failures with it. In implant-supported single crowns, annual chipping runs about 1.65% for veneered zirconia versus 0.39% for monolithic. Layered is a reasonable call when clearance is generous and aesthetic demand is high — but it should be a decision, not a default.
How fast can Istar turn a posterior zirconia case around?
Design in 2–3 days, production from receipt of a complete case. Istar runs a Shenzhen + Huizhou facility that produces around 8,000 zirconia single units a month in low season and up to 15,000 in peak. Volume capacity is not the constraint — a readable scan and the ten prescription fields above are.
10. Send us your next posterior zirconia case
If the case on your schedule looks anything like Case 27 — a short-wall second molar, a short-ferrule root-canalled tooth, a dark stump under a bright target shade, or a crown that came back on you inside a year — send the scan with the ten fields above.
We will come back with the wall heights, convergence and ratio we measure, anything that falls outside threshold, and the design options we see — before any design starts.
Contact our case review desk — Istar's case coordinators reply inside one working hour during business hours (9:30 am to 6:00 pm China time), and within 12 hours outside them.
References
- Goodacre et al., Tooth preparations for complete crowns: an art form based on scientific principles. PubMed
- Posterior Full-Contour Zirconia Crowns: Preparation Design. Spear Education
- Restoring a Tooth With a Short Clinical Crown. Inside Dentistry
- State of the Art of Different Zirconia Materials and Their Indications. PMC
- Comparative Analysis of Translucency in Different Thicknesses of Monolithic Zirconia Crowns. Clinical and Experimental Dental Research
- The Effect of Die Spacer Thickness on Retentive Strength of All-Zirconium Crowns. IJMRHS
- Changes in Occlusal Contacts upon the Cementation of Zirconia Crowns with Different Cement Spacers. Dentistry Journal
- Adhesion to Zirconia: A Systematic Review of Surface Pretreatments and Resin Cements. PMC
- Effect of different cleansing agents and adhesive resins on bond strength of contaminated zirconia. ScienceDirect
- An evaluation of antagonist enamel wear opposing full-coverage zirconia crowns — umbrella review. PMC
- Natural teeth wear opposite to glazed and polished ceramic crowns: a systematic review. PMC
- A systematic review and meta-analysis of veneered and monolithic implant-supported single crowns. PubMed
- Effect of background structure on the color of posterior zirconia-based all-ceramic crowns. PMC
- Effect of finish line designs on the dimensional accuracy of monolithic zirconia crowns. PubMed
- Influence of tooth preparation design on margin discrepancy — systematic review of in-vitro studies. ScienceDirect
- Accuracy of intraoral scanners for full crown tooth preparations with subgingival margins. PubMed