Baking question about lemon cake and curdling

January 11, 2024 7:16 PM

Recently I made Nagi’s lemon cake and it was lovely, but the step where you add the heated milk to the lemon juice resulted in a curdled mess. Future me would like to avoid this, for aesthetic reasons.

At the time I proceeded anyway as this is basically like baking with buttermilk, and it worked out fine, but as curdling produces precipitated protein chunks, the final cake had this brown flecked appearance. I also assume that it means the milk is performing differently in the batter given it basically turns into curds and whey. But when I think about it, I don’t see how this step can’t result in curdling (acid + hot dairy = protein precipitation, after all). When you look at the recipe pictures, there is zero curdling, and only one person in the ~300 comments mentions this happening to them. Nagi is typically very detailed in her instructions, so if it was supposed to happen I believe she would mention it. I want to make it again but without the curdling, as it seems this is possible.

Does anyone have any suggestions I could try? I’d be tempted to just add the lemon juice or the milk after the rest of the liquid ingredients are incorporated into the batter, but I think the idea is to mix it as little as possible after the first step to retain volume in the batter.

Thanks!

posted by BeeJiddy to Food & Drink (4 answers total)

Best answer: The recipe text at the bottom seems to indicate that the warm milk + melted butter get poured beside the lemon juice + zest, but not actively mixed until the "lightening" when you whisk them both together with 1.5c of the egg batter. Is it possible that this phase is supposed to work more like a whisking of both lemon and milk/butter into the egg batter, so that both get sufficiently diluted by the batter before the juice has the chance to curdle the milk? Did you instead mix the milk/butter directly into the lemon juice before you added the batter?

Alternately, could your milk have been too hot? The recipe says not to boil it, just to heat until the butter melts.
posted by Bardolph at 7:33 PM on January 11, 2024 [1 favorite]

Best answer: Cooks Illustrated’s short-lived science spinoff had an article about making clarified milk punch (404ed, but here’s a mirror) that talked about the difference of adding citrus to milk or adding milk to citrus:

Adding milk to the punch invariably resulted in a twisted mass of curds suspended irregularly within the punch. This happens because the milk comes in contact with the highly acidic punch and coagulates on impact. The result is that only a portion of the punch actually gets clarified, giving you a colorful (in a bad way!), cloudy drink. Adding the punch to milk might not seem all that different at first glance, but it’s all about rate of acidification. As the (acidic) punch streams into the milk, it slowly (relative to the milk-into-punch method) drops the pH of the milk. Once all the punch is in the milk, the pH is low enough (lower than 4.6, the pH at which casein proteins precipitate, or fall out of solution) for it to curdle. At that point the milk and punch are evenly mixed, so when the milk curdles it’s able to trap impurities from the entire mass.

Sounds like your recipe should have specified (slowly) adding the lemon juice to the milk mixture and not the other way around. Depending on the amount of lemon juice it might not curdle at all because the pH never gets low enough.
posted by fedward at 7:38 PM on January 11, 2024 [7 favorites]

Best answer: I dug out my On Food and Cooking because I remembered McGee had a good section on milk curdling... a couple ideas that popped out:

• He advises to always add hot to cold, to prevent flash coagulation of a small amount of cold liquid as it enters the heat.
• Milk proteins unfold at 160 degrees F (70 C), at which point they will start to bond to each other. I don't think you want to go any higher than that.
• Did you use skim milk? Fat globules will grab on to casein proteins as they unfold, which prevents the protein from sticking to itself and creating curds. A whole-milk and lemon mixture will resist curdling much better than a skim one. (He notes that anything over 25% fat will bind to all the casein and eliminate curdling entirely, even with a lot of acid, which is why you can boil a cream sauce)
• Any chance the milk was a little less fresh? He notes that the pH of milk naturally drops as it ages, which takes away some of your anti-curdling insurance.
• If nothing else works, McGee also notes that a little bit of starch in the mix can prevent most curdling - a tablespoon of flour, or 2 teaspoons of cornstarch of arrowroot, per cup of milk - because the starches gelatinize at 175F (77C), right as the proteins are opening up and looking for something to grab on to.

Good luck! The recipe looks delicious.
posted by graphweaver at 8:56 PM on January 11, 2024 [7 favorites]

Response by poster: Thanks so much, everyone. These are all excellent ideas/suggestions, and I'll incorporate aspects of each next time I make this.

There is definitely a chance my milk was a tad hot the first time I made it so I will be especially careful there. I think the idea of adding a little starch or batter could be a good idea either way and shouldn't negatively affect it.
posted by BeeJiddy at 11:16 PM on January 11, 2024