I was waiting for someone to ask that question. Thanks, reader Emily! Dulce de leche recipes call for baking soda for one simple reason: to help the mixture brown. Those of you who’ve stuck with me through past posts about caramel know that it takes a fair amount of heat for sugars to start caramelizing and browning — in excess of 300 degrees Fahrenheit. Watery mixtures like milk can’t get that hot, since the boiling point of water is 212. Yet as you can see from the pictures down below, a sweetened milk mixture will start to brown after a few minutes of boiling as long as it has some baking soda mixed into it. What’s going on?
The answer: the Maillard reaction, a very complex chemical process that occurs when amino acids (the building blocks of proteins) react with sugars. The results of this interaction are hundreds of different types of molecules, few of which are very well understood. However what all food scientists can agree on is that they’re brown.
The Maillard reaction requires a few things to get going properly: protein, sugar, heat, relative dryness and a relatively high pH (i.e. an alkaline environment). It’s the alkalinity that really speeds the process along, which is why when bakers want browning in a hurry they raise the pH. It’s why bagels and pretzels are boiled in lye before they’re baked. Baking soda isn’t nearly as strong a base relative to lye, yet it still does a nice job of accelerating the Maillard reaction and turning a dulce de leche a nice tan color.
But Joe, you called dulce de leche a caramel, how can you say that when there isn’t any caramelization? Well, that isn’t strictly true. Late in the reduction process, as water boils out of the dulce de leche, the temperature goes up to the point where you do get some caramelization. However it is the Maillard reaction that’s responsible for the lion’s share of the browning. Seen in that light, I suppose dulce de leche isn’t a “true” caramel, though it sure tastes like one, and that makes it close enough for Joe.
UPDATE: Reader Chana comments:
You mention that the Maillard reaction needs a few things, one of which is “relative dryness.” At the risk of sounding like an idiot, I’m not sure what that means. (Where I come from, milk is pretty wet. Of course, I’m from New York City, so what do I know?)
Great point! I’m actually the idiot here, because I phrased that poorly. I’ll amend what I wrote by saying that better Maillard reactions tend to occur in drier environments. A good example is a sauté pan. If you crowd it with food none of the items in it will brown properly. The reason: because the escaping moisture inhibits the reactions. By sautéing with less food you get more evaporation and as a result better browning.
Which is an odd way of saying that while certain conditions are more conducive to Maillard reactions, they can and do happen under different circumstances.