Silica vs Quartz Fillers in RTV Molds: Shrink and Flow

Fumed silica and ground quartz are both silicon dioxide. They do not do the same job in a silicone mold compound.

Fumed (pyrogenic) silica is a high-surface amorphous powder, the classic reinforcing and thixotropic filler in platinum RTV, tin RTV, HCR, and many LSR grades. Ground quartz (silica flour, crystalline SiO2) is an extending filler: more volume for less viscosity penalty, more abrasion, and a shrink lever because mineral does not shrink with the polymer.

If you're buying a custom RTV mold, stirring “some silica” into a platinum jacket is not a spec. Untreated fume can gel a pour. Quartz can sand a master. Neither loading belongs as a shop-invented weight percent. Platinum RTV already sits near 0.1–0.3% linear shrink after a complete cure. If the casting shrinks, compensate the master for the resin, not with a fake mold-rubber percent.

LSR/HCR steel tools are where filler-driven shrink actually belongs in cavity scale — a different RFQ, cold-runner if LSR, never a LIM hot tool.

Two SiO2 powders, two mechanisms

Fumed silica is made in a flame. Particles are small, chained, and covered in silanol groups. In a polysiloxane they build a temporary network. That is why a little untreated fumed silica turns a thin RTV into a non-slump paste, and why a treated grade (HMDS or silane) can reinforce with a milder viscosity spike. Tear (ASTM D624) and green strength are why compounders use it. Clarity can survive better than with quartz flour.

Precipitated silica is also amorphous and reinforcing. Treat it as a cousin of fumed silica, not as quartz.

Quartz flour is crushed crystalline silica. Surface area is low compared with fume. You can put more quartz in before the mix refuses to pour. The cured stock is stiffer, heavier, more opaque, usually lower elongation, and harder on equipment. Respirable crystalline silica is a mill-room housekeeping issue. Do not bag them as one “silica.”

Fused silica (amorphous crushed glass) sometimes appears as an extender. If the TDS says only “silica,” ask which.

Common mix-up: quartz to “kill shrink” on a platinum jacket

Platinum RTV mold rubber is already in the near-zero shrink class after a proper cure. Quartz will mainly raise hardness, density, and abrasion, and it may ruin pourability. Tin-cure still loses condensate and keeps moving in storage; filler does not freeze that time-dependent shrink. A quartz-extended tin block is still a tin block.

ISO 3302-1 is a rubber tolerance class on the article drawing, not a silica percent. M2 is a common default.

What to ask when you're buying

• Named grade: reinforcing amorphous silica already in the RTV, or a quartz-extended heat-cured *part* compound?
• Pourable glove vs thixotropic brush-on?
• Casting shrink — do we scale the master?
• Tear vs stiffness: 10A–20A peel skin or a rigid jacket face?
• Who keeps the master if the first skin is too stiff to peel?

Worked example: 12-cavity bear, 18 mm, “add quartz so it doesn't shrink”

A confectionery RFQ asked to add quartz so pockets wouldn't drift. The platinum 15A jacket already sits ~0.1–0.3% shrink. Outer-row gummies running 3–6% heavy were from ~0.5–1.5 mm edge sag on an unsupported tray — a jacket problem. We kept treated-fume high-tear platinum, added the mother mold, and resized nothing with flour. Quartz would have killed peel on the ears.

Viscosity, abrasion, and which powder belongs

RTV pours live at low shear. Untreated fumed silica is a thixotrope: useful on a vertical master, hostile in a deep, fine cavity if yield stress keeps rubber off a letter. Surface-treated fume is how many pourable grades stay pourable and still tear. Datasheet cPs numbers are snapshots at one spindle speed. A fumed-silica paste and a quartz-extended paste can share a cup number and fill differently.

Quartz at extending levels makes a heavier slurry. Degassing is slower. Bubbles do not rise. Vacuum still works if pot life allows.

Quartz is hard and often angular. It wears mills, LSR screws, aluminum tryouts, and the master if you pour a heavily quartz-extended RTV jacket over a soft pattern. A quartz-filled mold rubber also changes the jacket: higher hardness, lower elongation, less forgiveness on undercuts. Wrong powder if you needed a soft platinum skin for a deep undercut.

Use fumed (or precipitated) silica when the compound must tear, hang on a vertical master, or stay closer to unfilled mechanicals. That is most quality RTV mold rubbers.

Use quartz flour when a compounder is extending a heat-cured part grade for hardness, lower shrink, or cost, and the process can take abrasion (mill or production LSR steel). Prototype printed tools are the wrong home.

Do not mix construction-grade silica sand into platinum RTV. Moisture, crystalline content, and inhibition are uncontrolled. Clarity, food, and medical are separate gates. FDA 21 CFR 177.2600 and ISO 10993 attach to the finished formulation, not to the word silica.

When this article is the wrong quote

LSR/HCR dimensional grades with quartz extenders belong on a parts OEM. A tiny jewelry mold doesn't need a mill-room silica lecture. Geometry can override: a rigid mother-mold face can be a stiffer RTV or a plaster/resin jacket, not a quartz dump.

Related reading: mold shrinkage, custom RTV molds, custom mold design, what tooling means.

What to send for an RTV mold quote

• Compound: named platinum RTV grade (filler as supplied)
• Do not add powder at the bench unless this is a documented let-down
• Targets: pourability, ASTM D624 tear, Shore, jacket
• Casting shrink: scale the master for the resin
• Ban shop percents
• Master, split, recast

Send whether the cavity must peel an undercut or act as a stiff block. We'll refuse “add silica to reduce shrink.”

We would not recommend adding quartz flour to platinum RTV to make the mold shrink less

We would not recommend raising Shore A first to fake a shrink fix. We would not dump construction sand into a 1:1 kit. We would not treat Shore A as a shrink calculator. Tooling charge ≠ a license to reformulate the drum.

RUUIPON Mold Structure Decision: sag is a jacket problem. Flat back? open-back. Cannot peel? two-part / core. Large wet casting or 10A–20A? mother mold — not quartz.

Typical shop values: fillers in RTV molds

| Item | Typical shop value | | --- | --- | | Platinum RTV shrink | ~0.1–0.3% after a complete cure | | Tin RTV | Keeps moving in storage; filler doesn't freeze it | | Mix | 1:1 or 10:1; vacuum degas until foam collapses | | Cavity Shore | 10A–20A high-tear platinum with jacket | | HCR / LSR part shrink | Often ~1.5–2.5% / ~2–3%; different object | | ISO 3302-1 | M2 common default; not a quartz percent |

Reference values only. Confirm material-specific limits against the supplier’s technical data and agree project acceptance criteria before production.

FAQ

Can I add quartz flour to platinum RTV to make the mold shrink less? Platinum RTV mold rubber is already in the near-zero shrink class after a proper cure. Quartz will mainly raise hardness, density, and abrasion, and it may ruin pourability. If the casting shrinks, resize the master for that resin. Do not invent a quartz percent to “correct” a platinum jacket.

Why does a little fumed silica thicken a mix more than a lot of quartz? Fumed silica builds a silanol particle network. Quartz is a low-surface extender; viscosity follows volume fraction and packing. Untreated fume is a thixotrope. Treated fume is milder. A cup cPs number can match while the two pastes fill a letter differently.

Is precipitated silica the same as quartz flour? No. Precipitated silica is amorphous and reinforcing, closer to fumed silica than to crystalline quartz. Quartz flour is crushed crystalline SiO2, used to extend, stiffen, and cut shrink at higher mineral volume in heat-cured grades. The mill-room dust controls are not the same.

Will quartz-filled LSR wreck an aluminum prototype tool? It can. Quartz is a hard, angular mineral. Production quartz-extended LSR belongs in steel. Printed resin tools and soft aluminum tryouts are the wrong cavity for that slurry — and the wrong quote if you needed a poured RTV mold.

Does more filler always mean less shrink on a tin-cure jacket? Not in a way you can CAD. Tin-cure keeps moving as condensate leaves. Mineral volume can change how a block feels on day two; it does not turn tin into platinum. For heat-cured LSR/HCR, extending fillers are a real shrink lever, and you still coupon the named grade.