Cooking with Cheese

The melting page establishes the mechanism: whether a cheese flows, crumbles or holds its shape is a structural question about calcium cross-links, and the answer was decided months earlier by pH. This page is what to do about that at the stove, where you cannot change any of it and have to cook the cheese you actually have.

⭐ Three ways to stop a cheese sauce breaking, and only one addresses the cause

A cheese sauce breaks when the protein network contracts and squeezes out the fat, leaving you with oily strings in a watery pool. Three fixes are in common use, and it is worth knowing they are not three versions of one idea.

1. Starch — a roux, or flour, or cornflour. The classic béchamel route. The starch thickens the water phase and physically gets between the fat droplets so they cannot coalesce. It works, it is what almost every mac and cheese recipe in the collection does, and it changes the texture of the finished sauce — which is sometimes what you want and sometimes the reason a homemade sauce tastes floury.

2. Evaporated milk. Extra casein, already partly denatured by the in-can sterilisation, brought in to help hold the emulsion. Cheap, effective, and it is why a certain kind of stovetop mac has a suspiciously silky texture.

3. Sodium citrate. 🔴 And this one is not an emulsifier, whatever the packet implies. It binds calcium, which pulls the calcium out of the casein cross-links — so the casein itself is freed to do the emulsifying. See Cheese Sauce, Sodium Citrate.

⭐⭐ Notice what that means: the first two fixes work around the problem and the third dismantles it. Starch and evaporated milk manage a network that is fighting you. Citrate goes and removes the staples. It is the same mechanism the melting page describes, run deliberately backwards — and it is why a citrate sauce can be made from a two-year cheddar that would otherwise be hopeless, with nothing in it but cheese, water and a pinch of salt.

The three things you control at the stove

Heat — lower than you think, and add the cheese last. The protein network contracts as it gets hotter, and past a point it wrings the fat out. Get the liquid hot, take the pan off the heat, then stir the cheese in. A sauce that broke almost always broke because it boiled.

Age — younger than you think. This is the counter-intuitive one, and it costs people money. As a cheese ages it loses moisture and its protein is progressively broken down, so an old cheese is drier and its network is more brittle: it melts less smoothly, not more. The instinct to reach for the best, oldest cheese in the fridge for a sauce is backwards. A six-month cheddar makes a better sauce than a two-year one. Save the two-year one for the board, where the crystals are an asset rather than grit.

Surface area — grate, do not cube. Grated cheese melts through before the outside overheats. Cubes require you to hold the pan hot long enough to melt their centres, which is exactly the condition that breaks the sauce.

🔴 Pre-shredded cheese, and the wrong lesson everybody took from 2016

Bagged shredded cheese is coated with an anti-caking agent — commonly powdered or microcrystalline cellulose, sometimes potato starch. In 2016 this became a scandal: there is wood pulp in your Parmesan.

The scandal was real and it was about something else. The FDA investigation that triggered the coverage concerned Castle Cheese Inc., and the violation was economic adulteration — product sold as “100% Parmesan” that did not meet the standard of identity for Parmesan, which is 21 CFR 133.165. That is a fraud and labelling problem. Cellulose as an anti-caking agent is separately permitted and GRAS, and it is a purified plant fibre, not sawdust.

⚠ So “there’s wood in your cheese” is the wrong lesson from a story whose real lesson is that “100%” on a grated-cheese tub is a claim somebody has to be checking — which is the labeling wing’s argument, not a chemistry one.

The actual kitchen objection to pre-shredded cheese is functional and much duller. The coating exists to stop the shreds sticking to each other. In a sauce, that is precisely the job you do not want done: you are asking the shreds to fuse, and something has been added to prevent it. The shreds are also drier than freshly cut cheese, having sat with a desiccant against a large surface area.

🔴 I want to be honest about the strength of this claim. It follows from the mechanism and it matches what I and every cook I know report at the pan — but the one peer-reviewed study I found on the question (Consumer perceptions of anticake agents on shredded Cheddar cheese, Journal of Dairy Science 2021, doi 10.3168/jds.2020-19052) is paywalled and I have not read it, so I am not resting anything on it. Grate your own for a sauce; use the bag for a topping, where none of this matters.

Which cheese for which job

Straight from the melting spectrum, translated into shopping:

• Flows and stretches: young mozzarella, pasta filata generally, young cheddar, gouda, Monterey Jack, the alpine cheeses. These are your sauce and pizza cheeses.
• Browns instead of flowing: halloumi, paneer, juustoleipä — the high-pH, high-calcium end. Grill or fry them. Trying to melt one is not a technique failure, it is a category error.
• Crumbles and will not save a sauce: feta, most fresh cheeses, and the very old grating cheeses unless you bring citrate to the fight.
• Flavour rather than texture: Parmesan, Pecorino and the rest of the grating cheeses. Their job is seasoning, not structure — add them at the end.

Two more things worth knowing

Cheese is a seasoning, so season last. Almost every cheese carries serious salt — that is what salting is for. Salt the dish before the cheese goes in and you will have salted it twice.

A cheese that will not melt will brown beautifully, and that is a feature. A hard cheese laid straight on a hot pan gives you a frico — a lacy crisp — because the fat renders out and the protein and residual sugars brown. The same property that makes halloumi useless in a sauce makes it the best thing on the grill.

Where this goes next in the collection

Three pages take these principles somewhere specific: mac and cheese, which is the sauce argument at full length and with real recipes attached; fondue, raclette and queso, which are three cultures solving the same emulsion problem three ways; and rinds and leftovers, which is what to do with the parts of the cheese this page has been quietly assuming you threw away.

Sources

• Melting mechanism, calcium cross-linking and the pH connection: /cheese/science/melting/ and /cheese/science/salt-and-ph/, which carry the underlying references.
• Sodium citrate as a calcium sequestrant rather than an emulsifier — see Cheese Sauce, Sodium Citrate.
• 2016 grated-Parmesan coverage: the underlying FDA action concerned economic adulteration against the 21 CFR 133.165 standard of identity, not the safety of cellulose as an anti-caking agent.
• Consumer perceptions of anticake agents on shredded Cheddar cheese, Journal of Dairy Science (2021), doi 10.3168/jds.2020-19052 — cited but not read; paywalled. No claim here depends on it.