Recipes

    Melting

    Why does cheddar flow into a sauce while feta just sits there in crumbs, and halloumi neither flows nor crumbles but browns like a steak?

    The answer is one thing — calcium cross-links — and once you have it, every melting behaviour in the kitchen becomes predictable, including the ones that look like opposites.

    The mechanism

    The protein network in a cheese is held together partly by calcium bridges between casein strands. Calcium is divalent — it carries two charges, so a single calcium ion can bind two protein sites at once and staple them together. Sodium and potassium can’t do this: they’re monovalent and, as the literature puts it, cannot form crosslinks.

    Heat a cheese and two things compete. The fat softens and the protein wants to flow. But the calcium cross-links are holding the network rigid, and however hot it gets, a heavily cross-linked network will not flow — it will just sit there and eventually brown.

    So melting is a structural question, not a temperature one. And the amount of calcium cross-linking was decided months earlier, by pH.

    The pH connection, which ties this to everything else

    From salt and pH: lower pH strips calcium out of the casein network and into solution. There’s a positive correlation between pH and insoluble — that is, network-bound — calcium.

    That gives a straightforward spectrum:

    • High pH → lots of calcium still bound → heavily cross-linked → holds its shape
    • Middle → enough network to flow, few enough staples to let it → melts well
    • Low pH → most calcium gone, and often no proper rennet network to begin with → crumbles

    Which explains the three cases

    Cheddar, gouda, alpine, young mozzarella — melt well. A proper rennet network, moderate pH, moderate cross-linking. Heated, the network relaxes and flows.

    Halloumi, paneer, juustoleipä — don’t melt, and hold their shape. These sit at the high-pH, high-calcium end. Juustoleipä’s non-melting behaviour is attributed partly to added calcium chloride and a high pH of 6.1–6.6, and traditional paneer is described as lacking halloumi’s calcium cross-linking — which is why paneer softens somewhat more than halloumi does. Grill them and they brown and squeak instead of flowing. That squeak is the intact network resisting your teeth.

    Feta — doesn’t melt, and crumbles. The opposite failure. Feta’s low pH means it lost most of its colloidal calcium phosphate during cheesemaking. It has neither the intact cross-linked network of halloumi nor the balanced one of cheddar; heated, it softens and falls apart rather than flowing.

    So “doesn’t melt” describes two completely different materials at opposite ends of the same axis. I find that genuinely clarifying — it explains why substituting feta for halloumi in a grilling recipe fails, and fails differently from substituting it for cheddar in a sauce.

    Age changes it too

    Young cheese stretches; old cheese flows and then breaks.

    Proteolysis has been cutting the casein network into shorter and shorter pieces the whole time it’s been aging. Shorter strands can’t form the long elastic threads that give stretch — so a young mozzarella pulls into strings, and a two-year cheddar melts into a puddle and then separates into fat and grainy solids.

    This is the practical reason pizza uses young, low-moisture mozzarella rather than something aged and interesting. You want stretch, and stretch needs intact protein.

    It’s also why an aged cheese in a sauce is more likely to split: less intact network to hold the emulsion together.

    🔴 Sodium citrate, and the Kraft connection

    Here is the trick that closes a loop from the history wing.

    If melting is limited by calcium cross-links, then removing the calcium should make anything melt smoothly — and that is exactly what emulsifying salts do. Sodium citrate and sodium phosphates are calcium-sequestering salts: they bind the calcium in the casein micelles, and the sodium substitutes for the calcium that was holding the proteins together.

    The consequence, in the literature’s terms: the loss of calcium cross-links increases protein solubility, reduces protein–protein interactions, and increases the tendency to melt.

    That is why a little sodium citrate turns any cheese into a flawless, glossy, non-splitting sauce — it is genuinely the best macaroni-cheese technique available and it isn’t a cheat, it’s applied colloid chemistry.

    And it is the basis of processed cheese. Kraft’s 1916 patent was about producing cheese that would not disintegrate under heat, and modern processed cheese is built on emulsifying salts doing exactly what’s described above. American singles melt perfectly because their calcium cross-links have been deliberately dismantled.

    Which is worth being fair about. The thing this section spends most of its time celebrating — a living, particular, aged cheese — is precisely the thing that melts badly. Processed cheese is a well-engineered solution to a real problem. It just solved it by removing everything else.

    The kitchen summary

    WantUseBecause
    Stretch on a pizzaYoung low-moisture mozzarellaIntact long protein strands
    A smooth sauceModerate-age cheese + a pinch of sodium citrateSequestered calcium, no splitting
    Something to grill or fryHalloumi, paneer, juustoleipäHigh pH, high calcium, holds shape
    Flavour in a sauceAged cheese, but blended with a younger melterAged brings taste, young brings structure
    Not to be disappointedDon’t ask feta to meltIts calcium left during the make

    That’s the Science wing. Next: the families — a page each on the dozen kinds, and what defines them.

    Sources

    Comments (0)

    1. Loading…

    Comments are held for moderation — nothing appears until approved.