Improving thermal conductivity of RTV silicone for hard candy

[westrock]

I use RTV silicone rubber to make moulds for hard candy. Hard candy is basically boiled sugar that is poured into moulds at just under 300 degrees F. The castings of hard candy develop bubbles on the surface because the silicone does not dissipate the heat, unlike traditional metal moulds that dissipate the heat quickly. I've tried many different matrix formulations using silicone mixed with silica, and zinc oxide; results improve with silica but not enough to eliminate all bubbles. I know that thermal conductive silicone is used in various industrial applications, I just don't know how they are formulated.

 

[Compositepro]

It is more likely that the silicone is off-gassing slightly due to the sudden heating. Try baking your mold at 300F just prior to molding to make sure it is completely dry. Try this first.

 

[westrock]

The moulds are all baked to finish off the curing process, so I'm pretty sure the material is not off-gassing. I asked the manufacturer that specific question, and they said there would be no gasses coming from the cured silicone.

 

[IRstuff]

The only way to improve the thermal conductivity is to impregnate with metal or ceramic.

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[westrock]

What type of ceramic material would typically be used to improve thermal conductivity?

 

[IRstuff]

alumina or any of the nitrides would be an improvement over the silicone itself.

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[Compositepro]

All polymers absorb moisture when exposed to humidity. Dry your mold.

What is your hypothesized reason that thermal conductivity is causing your problem?

Yes, metal molds have better conductivity. They also do not absorb moisture. I have worked many void problems, and moisture is the most common cause at 300F with no pressure.

 

[btrueblood]

C'Pro, I think the boiled sugar has bubbles in it from the start; the idea is that rapid cooling freezes the bubbles in the sugar in a more well-dispersed way, while slow cooling allows the bubbles to rise to the surface and form a less attractive "scum". OP - do I have this right?

 

[Compositepro]

Btrue, that is a plausible hypothesis. I do not think it is correct, but that is not important. When troubleshooting a problem you try the easy things first. How hard is it to dry a mold compared to inventing a new mold material?

 

[ornerynorsk]

You also need to be careful in your material selection if this is a commercial operation. Anything in direct contact with a foodstuff needs to be on the approved material list of FDA/NSF.

It is better to have enough ideas for some of them to be wrong, than to be always right by having no ideas at all.

 

[westrock]

Thermal conductivity was assumed the problem because the bubble problem in the castings improves as the temperature of the boiled sugar decreases. However I did not realize that silicone absorbed water, so I did a test by heating the moulds for 15 minutes at 200 degrees F. to dissipate any moisture. The moulds were poured immediately and the results appeared to be an improvement. However, I forgot to use a control mould so I'll do another test. Any idea on the rate of moisture absorption into silicone?

 

[TomDOT]

Depends on temperature, humidity and the exact composition of the silicone you use, though I don't have specifics on hand.

 

[Compositepro]

If your sugar is at 300F you should dry at 300F. 250F is a minimum drying temp. At 200F the drying rate of polymers is very slow because the vapor pressure of water is less than the air pressure surrounding it. It's like having a sealed can of water with a tiny pin hole in the lid. At 200F there is still air in the can and water will only move out of the can by diffusion though the hole.It may take days for the water to evaporate. At 212F and above you will have a steam jet coming out of the hole. You can dry things in minutes. But it still takes time for heat to soak into your mold and for water to leave.

The bubbles you see in the sugar are mainly superheated steam and will collapse when cooled. The temperature you cook sugar to is a measure of the residual water content. If you try to melt sugar without mixing with water the sugar will burn (decompose from the high temperature).

 

[westrock]

I recast using three mould samples:
1 regular silicone without drying
2 regular silicone dried in an oven at 300 degrees F. for about 20 minutes
3 Silicone matrix consisting of 50% silicone and 50% silica (sand) dried at 300 degree F. for 20 minutes.

The first sample returned the usual result full of bubbles on the casting surface (see pic 1)
The second sample was a big improvement, the dried silicone cut the bubbles dramatically (see second pic)
The third sample was even better; the casting was almost perfect; it's hard to tell, but most of the bubbles are on the back side of the casting.

Obviously moisture plays a part here, Compositepro I thank you for that. (I struggled with this for weeks) Not quite sure why the matrix with silica showed the best results unless thermal conduction is playing a part.

 

[NickE]

What about cooling rate?

If the heated molds did better, by being dryer, there was also a cooling rate difference. Does the mold with the 50% silica in the matrix cool slower than the one without the 50% silica? or faster? What is the thermal mass effect of replacing RTV Silicone with Silica?

I have a feeling that there is a critical cooling rate that will allow bubbles to nucleate and grow in the sugar/water solution. If you cool to below where the partial pressure of water vapor in sugar is low quickly, then slowly cool through the solidification of the sugar, the bubbles should rise out. Opened back molds correct?



Nick

 

[Compositepro]

Sand absorbs far less moisture than silicone. As for cooling rate there is an effect. Once the sugar at the mold surface freezes solid it can no longer have a bubble form in its surface. So faster cooling is generally better. Cooling too fast may prevent sugar from filling fine detail before it freezes.

Thanks for the feedback, westrock. That helps us all learn.

 

[westrock]

The molds are open faced or "open backed". I haven't measured the cooling rates of the matrix of silicone and silica Although there are several papers that discuss the various fillers used in Silicone that affect thermal conductivity. Check this example for one such paper...Link

In my application it appears that the critical factor affecting casting quality is moisture absorption into the mold material. I think Compositepro is on the right track by stating that the silica in the matrix is allowing less moisture into the mold material, (basically by displacing silicone). If I can eliminate as much silicone bulk from the side walls and other parts of the mold, cut silicone by 50% by substituting silica (which is basically inert) and dry the molds properly, the castings should be close to perfect. At 50% the silica does not seem to affect the qualities of silicone that make it a great mold material, although shear strength, and durability my be compromised in the long run.

 

[ornerynorsk]

The metal mold is sounding like a pretty good option. I guess it hasn't even been discussed here if the silicone mold is simply a tryout tool with the eventual production tooling to be done out of metal, or if this is simply a home confection project.

If this is a commercial venture, the problems associated with temporary tooling seem like a moot point. If it is extremely critical to provide aesthetically perfect samples, presumably for sales or tradeshow samples, perhaps the metal tooling is not a bad investment.

I did have one other thought on this. Years ago I was involved in a urethane casting project in which we had to put all material into a bell jar and apply high vacuum to de-bubblize the pour. Our molds were, in fact, metal, and the purpose of vacuuming was solely to address porosity in the finished casting. My memory fails, but it seems to me we had to exceed 28" in order to get a clean pour. With the shorter work time of a hardball or hardcrack sugar mixture, this idea may prove not to be viable, but thought I'd mention it.

Good luck!

It is better to have enough ideas for some of them to be wrong, than to be always right by having no ideas at all.

 

[westrock]

Metal molds are the traditional method for casting this type of confection, commonly called clear toy candy. Ye old craftsman would carve the models and then the molds would be cast in metal or pounded out of copper. Clear hard candy is making a resurgence, and the old methods are not commercially viable, so candy makers like me have to learn cnc milling for model making, and new economic mold materials.

I did try molten sugar in a vacuum; that was scary Molten sugar reacts violently in a vacuum chamber.

 

[FACS]

This was a great thread, the way they should be.
Thanks guys; things were learned, results were posted... very cool.


Charlie

http://www.facsco.com