TL;DR A prestige closure in clear acrylic is a brand surface before it is a component, so a fine white network on the surface sets off alarms even when the closure still works. That visual defect is usually stress crazing: stress crazing is the industry term for how a glassy polymer behaves under combined mechanical stress and solvent exposure. PMMA is a recorded material, the cap family records 7 catalog models, the caps and collars category records metallic finishes on 20/410 to 28/415 fitments, and alcohol resistance testing is a recorded testing capability. The boundaries are explicit: PMMA stress-crazing resistance is not a recorded property; crazing behavior is established per project, and annealing is not a recorded process; residual-stress relief is a project step. Crazing thresholds, solvent concentrations, haze percentages, and annealing conditions are not published by the program; configured and confirmed per project.
Why
The closure is the object a hand touches first, and in prestige lines it is usually the second-most-expensive piece of the package. The recorded catalog shows metallic finishing across the caps and collars range — a good proxy for how much visual capital rides on these components. Place a clear acrylic closure next to a formula rich in essential oils or alcohol, and a new question appears: will the surface stay clear?
The failure people mean is not a crack that breaks the part. It is a frost of fine, flat, whitish marks — crazes — typically around thread roots, snap seats, and press-fit junctions, exactly where a closure carries mechanical stress in assembly and use. Under magnification the marks are crack-like zones bridged by strands of polymer, and light scattering off those strands is what the eye reads as haze. Cosmetic first, functional only in the extreme — and cosmetics is precisely this industry’s medium.
The record is compact and honest: PMMA is a recorded material, the cap family records 7 catalog models, the recorded PCR grades cover PP, PET, PETG, ABS and AS — PMMA not among them — and alcohol resistance testing is recorded as a testing capability. PMMA stress-crazing resistance is not a recorded property; crazing behavior is established per project. This article walks the mechanism end to end so the conversation is about physics, not folklore.
How
1. The material position: PMMA and the family it belongs to
Poly(methyl methacrylate) — PMMA — is the glassy, transparent polymer behind clear acrylic packaging parts: lenses, jars, and closures that read as glass without the weight; in this program it is a recorded material. The cap family records 7 catalog models, and the caps and collars category records metallic finishes on 20/410 to 28/415 fitments — the appearance surface is routinely modified before meeting a formula. Not recorded: PMMA among the recycled grades — the PCR set is PP, PET, PETG, ABS and AS — so a recycled-content acrylic claim has no catalog basis.
2. The crazing mechanism: a glassy polymer under stress and solvent
Stress crazing is the industry term for how a glassy polymer behaves under combined mechanical stress and solvent exposure. It is the glassy counterpart of polyolefin environmental stress cracking, discussed in the fluorinated HDPE article — cousins rather than twins. Where a polyolefin part cracks under stress and chemical attack, a glassy polymer crazes: planar zones where the polymer stretches into a network of fibrils between two faces, perpendicular to the applied tensile stress. The fibrils keep the zone connected — a craze is not yet a crack — but they scatter light, which reads as the characteristic whitish haze.
The solvent’s role is to lower the barrier to craze initiation. Small organic molecules — ethanol, terpene alcohols, aromatic fractions of essential oils — swell and plasticize the polymer surface, and a plasticized surface crazes under less stress. The exact stress level, solvent concentration, and haze percentage are not published here; none are recorded. The mechanism is unambiguous: stress plus solvent move the material toward crazing sooner than either alone.
3. Where the stress and the solvent meet in a closure
A closure carries several stress sources at once. Molded-in residual stress is the quiet one: the cooling part skims the mold while the core is still hot, freezing in a surface tensile field. Assembly adds more: snap features carry standing stress, press-fit inserts and collars do the same, and everyday closing generates torque and bending in threads and shell. The regions most likely to craze — thread crests, snap shoulders, insert seats — are the most stressed in normal use, which is why crazing appears at specific sites, not uniformly.
The solvents are ordinary ingredients: ethanol in perfumes and alcohol-based deodorants; essential oils — the character fractions of citrus, spice, woods, herbs — are terpenoid-rich, with several components small enough to interact with the polymer surface. The interaction is formula-specific: two similar formulations can behave differently against the same closure, which is the argument for testing rather than guessing.
4. Establishing crazing behavior per project
Because PMMA stress-crazing resistance is not a recorded property; crazing behavior is established per project. A project answers four questions in order: the part — which model from the 7-model cap family with which finish; the stress state — how it is assembled and torqued, since testing must reproduce the assembled condition; the chemistry — the actual formula or a worst-case placeholder, with exposure, temperature, and contact defined per project; and the judgment — an agreed optical and functional criterion, where alcohol resistance testing is the on-record capability and the twelve-test lab matrix frames the wider failure-mode picture; quality control criteria are configured per project.
5. Mitigation: the project steps that change the outcome
If crazing risk is confirmed, the levers are design-side and process-side. Generous radii reduce stress concentration; uniform wall thickness reduces differential cooling; assembly method — insertion force, snap geometry, torque targets — sets the standing stress the finished cap lives with. The classic process-side intervention is annealing: heating the molded part below its softening range to let residual stress relax. The boundary is explicit — annealing is not a recorded process; residual-stress relief is a project step — so its temperature, time, and conditions are never quoted from program data; they are configured and confirmed per project.
<pre class="language-mermaid"><code>flowchart LR
A[Luxury closure from the<br/>recorded 7-model cap family] –> B[Solvent exposure:<br/>essential oils, alcohol]
B –> C[Stress crazing mechanics]
C –> D[Crazing behavior established<br/>per project: part + stress + formula]
D –> E[Recorded alcohol resistance<br/>testing capability]
D –> F[Residual-stress relief<br/>as a project step]
E –> G[Configuration confirmed per project]
F –> G
</code></pre>
The flow keeps the layers distinct: family and finishes are recorded facts; everything between is a project matter.
What
Everything in this section is recorded program data:
- PMMA is a recorded material.
- The recorded PCR grade set covers PP, PET, PETG, ABS and AS — PMMA is not among them.
- The cap family records 7 catalog models.
- The caps and collars category records metallic finishes on 20/410 to 28/415 fitments: electroplated, anodized, sprayed, hot-stamped.
- Alcohol resistance testing is a recorded testing capability.
- PMMA stress-crazing resistance is not a recorded property; crazing behavior is established per project.
- Annealing is not a recorded process; residual-stress relief is a project step.
Gotchas
- Crazing is not cracking. Filaments bridge a craze, so a closure can craze cosmetically and still function structurally. Judge by the agreed criterion, optical first and functional second, and do not assume a hazy surface means a broken part.
- No threshold exists to quote. Crazing thresholds, solvent concentrations, and haze percentages are neither recorded nor published; any numeric claim in a sales conversation is a project-generated figure, not program data.
- Bench parts lie. A free-standing closure can look perfect while the assembled state — snap engagement, press-fit insert, applied torque — crazes. Always evaluate the assembled, torqued configuration; the low-profile dropper collar and diaphragm is the same geometry in a different family, where assembly generates the stress.
- Annealing is a project step, not a recorded process. Its conditions must not be quoted from catalog data; they are configured and confirmed per project. The magnetic closure adds its own assembly stress sources for the same per-part evaluation.
- The recorded PCR grades do not include PMMA. Copy pairing “recycled content” with clear acrylic closures cannot draw on the catalog; the recycled-grade set is PP, PET, PETG, ABS and AS.
- Finishes are recorded; their behavior is not. The metallic finishes recorded on 20/410 to 28/415 fitments say nothing about coating adhesion or crazing resistance under solvent contact — a per-project subject, established on the project’s samples.