Silicone sensory stones: curing system, smell and the tear at the edge
Two silicone stones that feel the same can differ by a curing system and a post-cure. One passes a smell test and one does not.

Platinum and peroxide cure are different products
Silicone for a sensory stone is either platinum-catalysed or peroxide-cured, and the two differ in what they leave behind. Platinum cure leaves almost nothing; peroxide cure leaves volatile by-products that have to be driven off in a post-cure.
The difference matters for a toy because the residue is what a smell test finds and what a migration test measures. A peroxide-cured stone without a proper post-cure smells of ketone or alcohol on opening and continues to smell for weeks.
Platinum cure costs more per kilogram and needs cleaner handling, because the catalyst is poisoned by sulphur, tin and some pigments. A contaminated batch cures poorly and produces a tacky surface that no post-cure fixes.
The specification should name the curing system, because a drawing that says silicone has named a family and not a material. The curing system explains the smell, the tack and the tear.
Shore hardness and the squeeze that a child expects
A sensory stone is squeezed, rolled and mouthed, and its hardness sets all three. A stone at Shore A 30 is soft and collapses in a hand; one at 60 is firm and resists a small child's grip.
Hardness also controls the stone's recovery, which is what makes the object satisfying. A soft stone recovers slowly and feels doughy; a hard one snaps back and feels like rubber.
The hardness is set by the formulation and by the wall thickness, and a stone with an uneven wall feels different in different places. A wall that varies by thirty per cent is a stone that feels wrong at one corner.
The specification should give Shore A with its test method and a wall thickness band, because hardness alone does not describe the feel of a hollow object.
Post-cure: the step that decides the smell
A silicone part is post-cured to complete the crosslinking and to drive off volatiles, and the schedule is a time and a temperature. A short post-cure leaves volatiles inside a thick section that release over weeks.
The schedule is set by the section thickness, because the volatiles have to diffuse out. A stone at fifteen millimetres needs a longer post-cure than a mat at three millimetres, and a schedule copied from a mat will fail on a stone.
The smell test that catches it is a sealed jar at forty degrees for forty-eight hours, opened and assessed. A stone that smells on opening has an incomplete post-cure or a contaminated batch.
The specification should state the post-cure time and temperature against the maximum section thickness. A post-cure stated as a line speed is a post-cure that changes with the oven's load.
Tear strength and the edge that splits
A hollow silicone stone fails at the edge where the wall turns, because that is where the stress concentrates when the stone is squeezed. The failure is a split that grows across the wall.
Tear strength in silicone is lower than in TPR, and it falls as hardness falls. The softest stones are the most tear-prone, which is the trade a designer makes for a soft feel.
The geometry helps more than the material: a rounded edge spreads the stress and a sharp edge concentrates it. A stone with a two-millimetre radius at the turn will outlast the same material with a sharp turn.
The test is a trouser tear on a moulded specimen plus a squeeze test at the edge on ten parts, looking for a white line that becomes a split. The specimen predicts the material and the squeeze predicts the geometry.
Pigment, catalyst poisoning and the batch that will not cure
Silicone pigments and additives can poison a platinum catalyst, and the result is a batch that cures slowly, cures tacky, or does not cure at all. The poisoning is invisible until the parts come out of the oven.
The common culprits are sulphur-containing pigments, tin-based additives and some organic rubbers that contaminate the mixing equipment. A batch mixed in a vessel previously used for a sulphur compound will fail.
The consequence is a sticky stone that collects dust and cannot be packed. The batch is usually scrap, and the loss is the whole mix rather than the parts.
The control is a dedicated mixing line or a strict cleaning protocol, plus a cure check on every batch. A cure check is a small cup left at room temperature, inspected for tack after the stated time.
Food contact, saliva and the migration that matters
A sensory stone is mouthed, so it is assessed as a mouthable article. The migration test uses saliva simulant or a substitute, at a stated time and temperature, and the result is a figure per unit area.
Silicone is generally a good material for this, but the additives are not: a plasticiser or a processing aid that is fine in an industrial part can fail a saliva test. The migration is usually from the additive rather than the polymer.
The test should be on the finished, post-cured part, because the post-cure changes what is available to migrate. A test on an uncured sample measures a different material.
The file should keep the result per colour and per batch, because a pigment change is an additive change. A file organised by shape rather than by colour will miss the one colour that fails.
Texture and the sensory claim
A sensory stone is sold on its texture, and the texture is produced by the tool or by a surface treatment. A tool texture is durable and a coating is not, because a coating on silicone adheres poorly and peels at the edges.
The texture should be specified in a standard scale with a depth, the same as a TPR tool, and the draft has to suit it. A deep texture with no draft is a part that tears on ejection, which is the same failure as on a soft TPR.
The texture also collects residue, and a deeply textured stone is harder to clean. A texture that is too deep for a mouthed article is a hygiene problem disguised as a sensory feature.
The specification should name the texture scale, the depth and the cleaning instruction. A texture without a cleaning instruction is a texture that will be cleaned wrongly.
Hollow moulding and the wall that collapses
A hollow stone is made by slush moulding, by rotational moulding or by a closed mould with a core, and each method gives a different wall uniformity. A slush-moulded stone has a wall that is thick at the bottom and thin at the top.
The wall uniformity decides where it collapses, and a stone that collapses at one point feels like a defect even when it is within tolerance. The collapse is a design feature when it is uniform and a complaint when it is local.
The check is a section through the stone at four points, with the wall measured. A wall that varies by more than the stated band is a stone that will feel wrong at one place.
The specification should name the moulding method and the wall band, because the method sets the band. A band without a method is a band that no process can hold.
Cleaning, sterilising and the claim on the pack
A mouthed silicone article is cleaned by boiling, by steam or by a dishwasher, and the claims have to be tested on the finished part. Boiling is a hundred degrees for a stated time; a dishwasher is alkaline detergent at sixty-five.
Silicone tolerates boiling well and tolerates alkaline detergent less well, because the detergent attacks the surface and produces a chalky film. A stone sold as dishwasher safe without a detergent test is a stone sold on an assumption.
The claim should be tested with the detergent of the target market and for a stated number of cycles, checking for chalk, tack and colour shift. A stone that passes twenty cycles and chalks at fifty has been qualified for the wrong life.
The pack should state the cleaning method it was tested for, not a generic food-safe line. A generic line is a claim that no test supports.
Eight lines for a silicone stone
Eight lines carry a silicone sensory stone: curing system, Shore A with its test method, wall band with the moulding method, post-cure time and temperature against the maximum section, tear strength plus an edge squeeze test, finished-part saliva migration per colour, texture scale and depth with a cleaning instruction, and a cleaning claim tested with the market's detergent.
The lines are unusual because four of the eight describe the cure and the wall rather than the material. That reflects what a silicone part is: a cured section whose feel is a schedule.
The two lines that decide the smell and the split are the post-cure and the edge radius, because those are the two that a visual inspection cannot see. Both are measurable before shipment.
Write the eight, keep a retained part per batch, and re-qualify when the curing system, the pigment or the moulding method changes. Those three move all eight.
The stone whose post-cure schedule is stated against its maximum section is the silicone sensory stone, and the section is what decides whether it smells on opening.
- 01Curing system named: platinum or peroxide, not simply “silicone”.
- 02Shore A with its test method, plus a wall thickness band.
- 03Moulding method named, with the wall band that method can actually hold.
- 04Post-cure time and temperature stated against the maximum section thickness.
- 05Sealed jar at forty degrees for forty-eight hours, opened and smelled.
- 06Tear strength on a specimen plus a squeeze test at the edge on ten parts.
- 07Saliva migration on the finished post-cured part, per colour.
- 08Cleaning claim tested with the target market's detergent for a stated cycle count.