Is Pyrex, Pyrex?

A reader writes:

When I was visiting my mother the other day, I dropped her glass casserole baking dish... thing... (I'm not much of a cook), and it broke, and so of course I said I'd get her a new one. The old one was "Pyrex" brand, but she told me I should just buy whatever similar sized glass dish is cheapest, because, and I quote "Pyrex isn't made from Pyrex any more".

The philosophical implications of that statement aside, were Pyrex products made from special glass, and now they're not? All I know about Pyrex is that I've seen that word written on laboratory glassware.


In the olden days, the "Pyrex" brand, wherever you saw it, meant borosilicate glass. Borosilicate glass doesn't change size much in response to temperature (it has a low "coefficient of thermal expansion"), so if you heat or cool it suddenly, it's unlikely to shatter.

("Pyrex" wasn't actually the first borosilicate glass; Otto Schott invented it, and the Schott company still sells it under the "Duran" brand. But Pyrex became the genericised trademark for borosilicate glass. Lab glassware that's intended to be used on heat is pretty much all borosilicate, under one name or another.)

Ordinary "soda-lime" glass expands and contracts more with temperature. So if, for instance, you suddenly cool a hot plain-glass baking dish by putting it the sink and turning on the tap, the inside surface of the dish contracts as it cools, the outside surface stays expanded, and stress between the two encourages the glass to break.

This can also happen when a glass object is originally manufactured. After forming the object, if you don't "anneal" the glass by slowly cooling it (a special kiln for doing this to glass that's been made somewhere else is called a lehr), a brand-new glass object can break spontaneously as it cools, or be right on the edge of breaking from the slightest shock.

There are numerous tricky ways to make glass objects more sturdy, the most common of which takes advantage of soda-lime glass's thermal expansion and contraction, to "temper", or "toughen", the glass and force the outside of the glass object to be under great compressive stress, which glass tolerates very well.

The simplest way of tempering glass is by rapidly cooling the outside of molten glass, so it solidifies and contracts quickly, and is then pulled into compression when the core of the glass cools later. Now, any insult suffered by the object will have to overcome the compression built into the outer layers before it can get the glass into tension and get a crack going. And if a crack does start, the whole glass object will collapse into zillions of distinctive little lumps of glass with quite safe large-angled edges, rather than dagger-like shards.

The forces involved in tempering glass are the same as the forces that make unevenly-cooled, unannealed glassware fragile; they're just tightly marshalled to make the material more durable, in the same way that prestressing "tendons" can make concrete far stronger.

(The most extreme version of the tempering process is Prince Rupert's Drops...

...which you can make at home, while wearing suitable protective clothing, by dripping molten glass into a bucket of water. Internal tensions make the body of each drop amazingly strong, but if you snap the thread-like tail - which is also very strong, but so thin that it can easily be bent or sheared past its limits - the whole drop instantly explodes into tiny particles.)

(Oh, and again, if you'd like to have the above explained much more clearly, try J.E. Gordon's classic "The New Science of Strong Materials, or Why You Don't Fall through the Floor", which is one of my favourite books, along with "Structures, Or Why Things Don't Fall Down".)

A fancier kind of tempered glass is "Corelle", which is laminated tempered glass, but doesn't look or feel much like glass at all. This is partly because it's opaque (though I don't think there's anything about the manufacturing process that says it has to be), and partly because it's so strong that plates and bowls made from the stuff can be very thin and lightweight.

Which brings us back to Pyrex, because the Pyrex and Corelle brands are now both held by World Kitchen, LLC. World Kitchen would really like people to stop saying that Pyrex kitchenware isn't made from borosilicate glass any more, because although this statement is actually correct, it wasn't World Kitchen that changed it. World Kitchen say the change happened "more than 60 years" ago; other sources can't put an exact figure on it, but it seems pretty clear that it's not a recent development.

In any case, what World Kitchen sell today as "Pyrex" bakeware isn't plain soda-lime glass, but "heat strengthened" soda-lime, which presumably means the usual kind of tempered glass. Tempered glass resists breaking from temperature changes pretty well, and resists breaking from mechanical insults very well, so it's a good choice for bakeware, which is bumped by other bakeware much more often than it has to tolerate large temperature shocks.

Well, it's a good choice for bakeware as long as your oven doesn't get hot enough to anneal the glass, which I think it definitely doesn't.

This is despite the additives in soda-lime glass, which are there to make the stuff melt at a reasonable temperature. Silica, also known as quartz, makes up the bulk of all normal glass compositions, and could be used to do anything ordinary glass does. But quartz's melting point is way up around 1700 degrees Centigrade. This is higher than the melting point of iron, and makes quartz unreasonably difficult to use for glassware, unless you're making furnace windows or something.

To make soda-lime glass from scratch you need a furnace that burns hot enough to melt silica - which is why recycling glass is so popular - but once the ingredients are mixed, the melting point of the mixed material plummets to less than 600°C.

Annealing happens significantly below the melting point, but you still need a temperature of more than 500°C to anneal soda-lime glass, even if you're willing to wait for hours, and no household oven goes that high. Actually, I don't think any food oven goes that high. The hottest are probably coal-fired pizza ovens (the great problem of making "authentic" pizza at home is getting the oven hot enough); I think those top out at around the 1000°F/540°C mark, but they usually run rather cooler.

I'm sure there are many companies that make tempered, or toughened, glass kitchenware, and I'm also sure that other companies again make plain soda-glass kitchenware, which may not even be properly annealed, much less properly tempered. So your mum may be right that Pyrex-brand glassware is not particularly good - but you also shouldn't buy the cheapest glass casserole dish you can find, unless you've good reason to believe it's made from tempered glass. Which may or may not be clearly, or honestly, indicated on the box.

I think the best way to authoritatively tell the difference is by bopping any dish you're planning to buy with a ball-peen hammer. I leave the formulation of techniques by which one could get away with this as an exercise for the reader.

Psycho Science is a regular feature here. Ask me your science questions, and I'll answer them. Probably.

And then commenters will, I hope, correct at least the most obvious flaws in my answer.

8 Responses to “Is Pyrex, Pyrex?”

  1. rob Says:

    Great stuff, now I understand why the cool Bodum Pavina glasses I bought lasted a couple of weeks each - since they're made from borosilicate glass all it took was for them to fall over on the dish drainer and they were bin-fodder. But as a decorative item they did look amazing.

  2. dan Says:

    I think you can toughen borosilicate glass as well. You can't do it by thermal tempering, but an alternative way of making toughened glass is by submerging the object in a bath of molten potassium nitrate. This swaps some potassium for sodium in the outermost layer of the glass, the bigger atoms wanting to take up more space. That might work on borosilicate glass, since boron has an even lower atomic number than sodium, but I'm only guessing.

    You'd think fancy brand-name glass products would have this sort of treatment. Maybe they've cut back on quality after their product-placement on Battlestar Galactica :-).

  3. rippo Says:

    I remember an article, which my google skills are unable to locate now, about the impact of the change to the makeup of Pyrex had on the operation of clandestine drug labs. Requiring them to acquire lab grad glassware rather than commodity kitchenware.

  4. greyhairedoldbadger Says:

    The wikipedia article suggests that Borosilicate glass "tends to crack into large pieces rather than shattering (it will snap rather than splinter)."

    Having broken more than my fair share of unreasonably expensive laboratory glassware over the years, I can confirm that this is indeed the case.

  5. Shivoa Says:

    In a 'things change depending on where you live' expansion to the Pyrex story:

    "France-based cookware maker Arc International acquired Newell's European business in early 2006.[6], and currently owns rights to the brand in Europe, the Middle East and Africa."

    "Arc International's Pyrex products, and Pyrex laboratory glassware is made of borosilicate glass." (both from Wikipedia)

    So anyone buying kitchenware with the Pyrex brand in the EMEA region can be assured that their cookware is unlikely to break due to thermal issues but also that dropping anything will cause large sharp pieces of glass all over the hard tiled floor of their kitchen. I have personally verified this with stock brought in the last 18 months here in the UK (due to incompetence rather than a need to verify the borosilicate nature of my purchase through destructive testing, luckily the kitchenware quality Pyrex is actually really cheap and so this mistake only cost me one of three casserole dishes purchased together for £15).

  6. evilspoons Says:

    Your hammer-based testing method reminds me of the Calvin & Hobbes comic where after seeing a sign for bridge weight limit, Calvin asks his dad how they know what the maximum size truck is for a bridge they are driving across.

    Calvin's dad replies that they build the bridge, drive larger and larger trucks across until it collapses, then build another bridge and put up a sign indicating one size smaller than the last truck.

  7. Pacomius Says:

    It may be of interest to people to note that the characteristics of the Prince Rupert's Drop were a key plot point in Peter Carey's novel Oscar and Lucinda in which they were likened "to the human heart".

  8. cr Says:

    Delighted to see you're familiar with J E Gordon's excellent books on materials technology. I always found lectures on 'strength of materials' totally boring and darn near incomprehensible. Until I came across Gordon's books. He not only explains it clearly, but in a way which makes you want to read more, like a good novel. It's a gift not many writers have. And he puts it all in historical perspective, from longbows to sailing ship design, to bridges to cathedrals - 'materials' underlies them all. I'd warmly recommend them to anybody interested in the history of - civilisation, really. (Click on the links in Dan's article above to find them on Amazon).

Leave a Reply