# Why did my glass pot explode when I boiled sugar in it and added cold milk?

I decided to try some cooking with sugar and put a glass bowl on the flame with half a cup of sugar in it and some water. After bubbles appeared I stirred the sugar for about a minute and then I decided to add milk, and as soon as I poured the milk, everything exploded and the glass pot shattered into little pieces rapidly.

What just happened? Why did my glass pot explode when I boiled sugar in it and added cold milk? And does this also happen with laboratory equipment?

• You have discoverd that cheap glassware does not like thermal stress. Congratulations. Good thing you did this with milk and sugar. – Karl Jul 23 '17 at 11:00
• Of course, immediately. And that was not an explosion. What on earth were you expecting from this experiment? – Karl Jul 23 '17 at 11:04
• @Karl Nougat,Honestly. – soundslikefiziks Jul 23 '17 at 11:07
• You said that you used a "glass bowl." You didn't specify if the bowl was Pyrex. But even Pyrex is susceptible to breaking with "great" thermal stress, especially if the glass has been weaken by scratches. I had a Pyrex beaker break when making dilute sulfuric acid. // You get a lot of thermal stress when only a part of the glass container is made much hotter or colder than the rest of the glass. Such as pouring a relatively cold liquid into a hot glass container. – MaxW Jul 23 '17 at 17:26
• @soundslikefiziks Look, if you cannot judge that by yourself, now that you have the info given, you should not be doing such experiments at all. You are endangering your health, especially you eyesight and fingers, and that of all people around you. Sorry. Stick to the procedures given your cookbook. – Karl Jul 23 '17 at 17:35

Assuming initially your glassware contained a cup of sugar and a little of water, you basically prepared caramel. Depending on the sugar employed, temperatures needed to trigger caramelization vary, but $\pu{160 ^\circ{}C}$ ($\pu{320 F}$) mentioned succrose, as example (which is the normal household sugar) is quite high. Most materials, including glass, dilate upon heating, which is fine as long as this is evenly done.

Now if you add cold milk, say of $\pu{20 ^\circ{}C}$, the glass aims to contract back to its initial state. Because of the quantity of milk added, and the large heat capacity of water as a major constituent of milk, milk served momentarily as considerable heat sink. Now taking into consideration that the glass ware's walls are unevenly heated -- at the outside still heated by gas or stove to more than $\pu{160 ^\circ{}C}$, and the inner of $\pu{20 ^\circ{}C}$, this suddenly generates a lot of mechanical strain and stress on the material. The more the walls of your glass ware are thick, the more easy these may then crack and shatter into pieces just by the sudden temperature change. (For the same token, you place hot glass ware on a plank of wood to allow slow cooling to room temperature.)

In addition, if the glass ware were closed tightly just after addition of the cold milk, the hot ($\pu{160 ^\circ{}C}$) is able to boil off the water in the milk, too; generating steam that likes to expand, or -- if confined in volume -- will build up pressure. Under normal circumstances, for each litre of (liquid) water, up to $\pu{1.7 m^3}$ of steam may be generated (at normal pressure). This represents an additional stress for the material, and standard kitchen glass ware is not designed to withstand such pressures.

• What if i had boiled the milk separately before i added it ? would that be an example of "evenly done" ? – soundslikefiziks Jul 23 '17 at 11:35
• @soundslikefiziks Your right, gentle addition of milk already boiling to the caramel (which often need not be at $\pu{160 ^\circ{}C}$ for mixing) lowers the temperature change, hence lowers the mechanical stress the glass ware will experience and consequently will lower the chance that the glass ware will shatter (and getting the content spilled all over ...). Beside looking glass ware dedicated for cooking, why not simply switching altogether to a sauce pan, made of metal? (So far, I still did not see a sauce pan crack...) – Buttonwood Jul 23 '17 at 11:50
• I will definitely follow that advice and switch to metal made cookware. i was just surprised by this explosion. – soundslikefiziks Jul 23 '17 at 12:01
• " if the glass ware were closed tightly just after addition " That is impossible. – Karl Jul 23 '17 at 17:30
• While certainly not as tight as a pressure cooker, we all experience once upon a time a lid accidentally crammed to the pan, for example. Then, the comment above, "but also quasi instantly evaporated. 18 gramms of water are >25 liters of gas. I'm sure this looked a bit like an explosion." has "Karl" as signature tag, too. – Buttonwood Jul 23 '17 at 20:20

Whilst it's almost certainly down to thermal stress and/or the milk effectively "exploding", there's another possibility that is interesting enough to note: "glass cancer".

When glass is made there are often impurities in it, including nickel sulphide crystals. These can exist in two forms, alpha and beta, depending on temperature. When glass is toughened it is heated up and then cooled rapidly - the heating required is coincidentally high enough to change the NiS inclusions into the more compact form, and when the glass is chilled the NiS is trapped in this form. Over time they will revert to the larger form, potentially shattering the glass in the process. Even a UV photon can trigger the process, I've seen a pint glass sat on a table in a beer garden spontaneously explode, it's quite impressive.

The heat shock could be enough to trigger this even if it's not enough to mechanically shatter the glass. The fact that you say it shattered into "small pieces" (relatively regular, quasi-cubical pieces?) is typical of toughened glass failing in this manner.

• Yes, it is interesting and well worth noting. i also wonder how much of these impurities are needed in order for this to happen. – soundslikefiziks Jul 23 '17 at 17:17
• This does not happen with pyrex glass, only standard safety glass. The former does not contain these impurities, and the latter is unsuitable for heating anyway. And the story about the UV photon is, well, less than scientific. – Karl Jul 23 '17 at 17:44
• I have already experienced the sudden explosion of two glasses. Temperature unchanged. In one case I did spoon an aspirine tablet, glass went into lots of small cubes (I do not know how to translate, perhaps safety glass but this could be a glass with metal grid or polymer films. In the second case the glass exploded in two tre pieces and was even noiser. I think these case are related to annealing induced tensions, isn't like that? – Alchimista Jul 23 '17 at 20:29

Older Pyrex is true borosilicate. to save money, some of the newer pyrex is tempered glass. I think the true borosilicate glass has pyrex in lower case and the tempered glass has PYREX in upper case.