Case Study: Temporary Crowns for Molars — A 10-Week Provisional on Tooth #30 Through Root Canal and Crown Lengthening

Here's what he came in with:

• An old MOD composite from maybe eight years back, cusps undermined, the kind of restoration you look at and already know how it ends.
• Sharp pain on biting, only on one spot, only sometimes. Classic cracked tooth story.
• No history of root canal work on that tooth. Vital, but asking questions.
• Soft tissue sat a bit high on the buccal — the dentist flagged biologic width would be tight once he reduced down to sound structure.
• The guy flies for work. Two long-haul trips booked inside the window. Barometric load matters here, and we'll come back to that.

He wasn't our patient. We never saw him. The dentist sent us scans, photos, a written specification, and a timeline. That's what we build to.

What the dentist asked us for

The specification came in on a Tuesday. Short, direct, no fluff:

"Need a provisional for #30 that lasts the full sequence. Endo first, then crown lengthening, then four weeks for the tissue to settle before final impression. Count on ten weeks in the mouth, maybe eleven. He grinds a little, no diagnosed bruxism. Margin needs to sit above tissue even after lengthening. Make it bulky where it has to be. Don't make me adjust it chairside — I've got a packed morning when he comes back."

Unpacked, that's six asks:

1. Service life ten to eleven weeks. Not two to three. Material and finish have to match that.
2. Margin position planned for after the surgery, not before. Supragingival on the lengthened crown-root interface, with room for the tissue to settle.
3. Occlusion dialed in on the model. He wanted light centric contacts, zero lateral interference, and no high spot that would need chair time to grind off.
4. Minimum bulk on the occlusal. Thin provisionals wear through on molars inside a month. He wanted it built like it meant it.
5. Access-friendly if endo had to be redone. If the pulp didn't settle, the endodontist needed to be able to open through the provisional without the whole thing coming apart.
6. Shippable to a US zip code and in his hands before the first appointment Monday. We had five working days.

That last line is the one most specs leave out, and it's the one that decides whether the lab is actually helpful or just technically correct.

The technical problem, written down

Before we start a case like this, we write down what the real risk is. Not what the risk sounds like in a textbook — what it is for this tooth, this timeline, this operator.

Here's what we wrote:

The point of the table isn't that we're the only ones who thought about this. The point is: when a provisional on a molar has to last ten weeks, every one of these failure modes is live, and the right material is the one that survives all of them, not the one that scores well on any single test.

How we made this crown

Eight steps, in order. Nothing special about any single one — the point is that none of them got skipped, and that each one had a written output the next step could check.

Step 1 — case intake and scan review

STL files came in from an iTero scan. First thing we did was open them against the dentist's spec sheet and see whether what he asked for was actually buildable with what he sent.

Three things we checked before any design software opened:

• Preparation reduction on the occlusal, measured against the opposing arch — was there room for the 1.5 mm bulk he wanted?
• Finish line visibility all the way around — any blurry spots on the mesial that would come back to bite us later?
• Bite registration quality — the buccal scan lined up with max intercuspation, no interference ghosts.

Everything checked. If one of those had failed, our case coordinator would have called the dentist that same hour, before anything else moved. On this case nothing had to.

Step 2 — CAD design

A senior CAD designer took the case with a senior technician sitting next to her. Three decisions locked in on this step, in this order:

• Material: milled high-molecular-weight PMMA. For a ten-plus week service life on a lower first molar, this is the one with published long-service data behind it. A prospective series of milled PMMA long-term provisionals showed 90.4% survival at sixteen months and PMMA sits in a usable flexural range for the posterior. We weren't going to get ten weeks out of a chairside bis-acrylate on this bite.
• Occlusal thickness: 1.5 mm minimum at the functional cusp, 1.3 mm at the central fossa. Below those numbers this crown would wear through inside six weeks on a mild grinder. We wrote the number into the CAD and set a check for it on the next step.
• Margin: 0.5 mm short of the planned post-lengthening crown-root line. Not short of today's finish line — short of where the surgeon was going to put it. If we hugged today's line, we'd be compressing new tissue by week six. Sitting short, the margin stays supragingival all the way through and tissue has room to settle.

The design file got saved with those three numbers on a sticky layer, so the milling tech couldn't open it without seeing them.

Step 3 — milling

Mill ran on a standard PMMA blank. Internal relief set at 40 µm die spacing — enough that the crown would seat, not so much that retention would fall onto the cement, which on a molar is the weakest link in the assembly. In published same-mouth comparisons, milled provisionals have been measured at around 27–30 µm marginal gap on buccal and lingual. That's roughly where ours lands when the preparation is clean, which this one was.

Step 4 — hot-water post-polymerization

Everything coming off the mill goes into a hot-water bath before it touches a polishing wheel. Not optional on anything running over three weeks of service. In testing this step has been shown to cut residual MMA elution by up to 80% and raise microhardness by up to 50%. It's invisible on the finished part. The patient would never see it, the dentist would never see it. We do it anyway.

Step 5 — finishing and polishing

Two things got done here, and the order matters:

• Occlusal shaping. Rounded cusps, shallower than we'd put on the final zirconia crown. A provisional with sharp cusps wears into chaos over ten weeks. A provisional with rounded cusps wears into a flat plane you can live with.
• Polish down to below Ra 0.2 µm on all exposed surfaces. Above that threshold plaque accumulation rises, and across ten weeks on a lower molar that difference stops being theoretical. The finishing tech ran a surface check with a profilometer at two spots, buccal cervical and distal contact, before signing it off.

Step 6 — articulator check for occlusion

Mounted on a semi-adjustable articulator against the opposing cast the dentist sent. Four checks, each with a pass/fail that got written on the case sheet:

• Centric: light, even contacts at three points, no fulcrum. Pass.
• Lateral excursion, working side: no interference. Pass.
• Lateral excursion, non-working side: cleared. Pass.
• Protrusion: cleared. Pass.

If any of those had come back as a fail, the crown would go back to Step 5 for adjustment, not out the door. On this case no step-back was needed.

Step 7 — QC, shade and thickness logging

Final inspection before packing. Our QC tech photographed the crown from four angles, logged the shade reading (A3 — matched the dentist's specification, verified against a Vita 3D-Master guide), and took thickness measurements with a digital gauge at six points: central fossa, mesio-buccal functional cusp, disto-buccal, mesio-lingual, disto-lingual, and the mesial contact area. All six points came back at or above the number from Step 2. The sheet got taped to the shipping box.

Step 8 — packing and shipping

Packed in a dentist-labelled individual case, with the QC sheet inside and a one-page cementation note on top: temporary cement recommended, non-eugenol if the final cementation of the zirconia crown was going to be bonded rather than conventionally cemented — a precaution supported by published work showing eugenol-containing temporaries and delayed dentine sealing can reduce the final bond strength (p = 0.002).

Scan arrived on a Tuesday morning Shenzhen time. The box was on the dentist's front desk the following Monday morning Pacific. No rush fee.

Try-in and feedback

The crown stayed in service for the full ten weeks. The dentist logged notes at four touch points and the patient was asked for a reaction at seating and again at the final impression. Summarised below, with what each observation told us about the build.

From the dentist

• At seating. The crown dropped onto the preparation without resistance and needed one localized spot of adjustment on the disto-buccal — the area where a model-driven occlusion and an in-mouth bite most often disagree. Two minutes of polishing resolved it; nothing had to be ground through. Contacts tested cleanly on floss. The crown was cemented with a non-eugenol temporary cement and the appointment closed inside ten minutes of chair time.
• Three weeks in. No contact needed. The patient flew long-haul in that window — the first of two trips flagged on the spec sheet — and the retention held through the pressure change. Barometric events are documented in the published literature as a plausible dislodging mechanism, which is why travel sat at the top of the pre-build risk table; the geometry carried the load rather than the cement.
• Six-week check, two days post crown lengthening. The margin sat where it was designed to sit. Soft tissue around the preparation was not inflamed, and no chairside modification of the provisional was needed to accommodate the new tissue level. The 0.5 mm short we built into the margin on Step 2 was aimed at exactly this appointment; a margin that hugged the pre-surgical finish line would have required a reline here, possibly a remake.
• Final impression at ten weeks. Full retention. Marginal seal intact under direct inspection. Occlusal wear limited to a single light facet on the mesio-buccal cusp, consistent with mild parafunction over the service window and well short of a vertical collapse. The provisional came off intact so the preparation could be impressioned clean.

In a follow-up note after the case closed, the dentist described the specification work on the front end as the main reason his chair time stayed short across all four appointments, and asked us to run the final zirconia crown through the same workflow.

From the patient

• At seating. Reported the crown felt bulkier than expected but was not uncomfortable. For a ten-week posterior service life, that is the right initial reaction — thickness is what prevents a wear-through call later in the window, and comfort within a few hours of seating is the acceptance criterion.
• At ten weeks. Reported being largely unaware of the crown through the service period, maintained flossing as instructed, and noted that the pre-treatment biting discomfort on the tooth resolved from day one of wear. Unawareness is the outcome we aim at on a long-service posterior provisional; the crown did its job without pulling attention toward itself.

The combined picture across both perspectives: a crown that seated with minimal adjustment, held retention through a travel event and a surgical event, maintained its margin position after the gingival line shifted, wore within a predictable envelope, and left no clinical reason to intervene before the planned final impression.

FAQ

How long can temporary crowns for molars reasonably stay in service?

Two to three weeks is the design assumption behind most of the systems you'll find on a bench. Ten to twelve weeks is doable with the right material and the right spec — this case ran ten and finished clean. Longer than that is a case-by-case call, and it stops being free around the point when longer provisional phases start showing a measurable link to pulpal problems in large-sample work (p < 0.001).

Why milled PMMA for a ten-week service on a molar, not a bis-acrylate?

Bis-acrylate is a fine single-unit, short-service material and we use it where it belongs. For a ten-week posterior service life on a molar, we want the documented long-service performance — a prospective series on milled PMMA logged 90.4% survival at sixteen months — and we want the marginal fit a mill delivers consistently rather than the one a chairside hand delivers once.

What's the real reason temporary crowns on molars come off?

Almost never the crown breaking. The cement interface releases first. Temporary cements run around 20–68 N tensile while the crown itself needs over a thousand newtons to fracture. The crown loosens, then cracks. Sticky food is more of a problem than hard food. Pressure change from flights or diving has been reported as a dislodging event too — which is why we flagged travel in this case.

Do you make the final crown too, or just the provisional?

Both. The case above moved into a zirconia final at week twelve. Our workflow runs the provisional and the final out of the same facility, from the same scan set, with the same CAD team following the case. That continuity isn't the main reason it works, but it does cut the number of handoffs that go wrong.

What do you need from the dentist to build one of these?

A full-arch scan of the prepared arch, a scan or impression of the opposing arch, a bite registration in max intercuspation, a buccal scan if the registration is tight, and a written note on three things: expected service length, planned surgical interventions, and margin preference (supragingival, equigingival, or how far subgingival). On this case we also had the dentist's note about the patient's travel, which changed how we spec'd retention.

How quickly can you turn one around for a US-based dentist?

Five working days door to door is our standard for a case like this, with no rush fee. Faster is possible on single units where there's no multi-stage sequence. Faster isn't always better — we'd rather send a part that doesn't need chair time than send a part that technically arrived early.

Working with us on a case like this

A long-service posterior provisional through a staged surgical sequence is the kind of case we'd rather talk through on the spec sheet before anything gets scanned. If you're running one — endo and lengthening on a molar, or anything else where the provisional needs to sit quiet for weeks on end — send the scan set and the timeline through istardentallab.com. We'll come back with questions before we come back with a quote.

References

1. Field J, Wassell R. Provisional restorations (Part 1). Br Dent J. 2023;234(11):805–809. PubMed 37291303
2. Rego MRM, Santiago LC. Tensile bond strength of temporary cements. J Appl Oral Sci. 2004;12(3):209–212. PubMed 21049254
3. Sari T, et al. Fracture resistance of provisional crown materials. Clin Oral Investig. 2020;24(11):4061–4068. PubMed 32583239
4. Nilsson G, et al. Provisional phase length and loss of vitality (n = 15,879). J Prosthet Dent. 2023;130(6):833–839. PubMed 35105459
5. Huettig F, et al. Milled PMMA long-term provisionals: a prospective series. Clin Oral Investig. 2016;20(1):161–168. PubMed 25898895
6. Díez-Quijano C, et al. Colour stability of long-term provisional materials. J Prosthet Dent. 2020;124(3):351–356. PubMed 31810613
7. Takamizawa T, et al. Wear and flexural properties of provisional materials. Oper Dent. 2015;40(6):603–613. PubMed 25405905
8. Sarfaraz H, et al. Clinical comparison of provisional fabrication methods. BMC Oral Health. 2026. PubMed 42177549
9. Lee SY, et al. Post-polymerisation and residual monomer elution. Eur J Oral Sci. 2002;110(2):179–183. PubMed 12013564
10. Bollen CM, et al. Surface roughness threshold for plaque accumulation. Dent Mater. 1997;13(4):258–269. PubMed 11696906
11. Gulve MN, Gulve ND. Dislodgement of a provisional restoration under barometric change. Case Rep Dent. 2013;749142. PubMed 23984113