Kitchen science and chemical leaveners

One of the biggest challenges in reconstructing historical baking has been the old sorts of chemical leaveners and the lack of reliable information about cooking with them. Without knowing how, say, saleratus actually works, it’s impossible to know what a recipe is supposed to do (or whether it actually does). So stand back, folks… we’re going to try science.

Three kinds of alkaline salts have been used to leaven baked goods since the end of the 18th century. Each reacts in different ways during the baking process to produce carbon dioxide, which is what leavens the bread, cake, or cookies. Here, in a nutshell, is the history and chemistry. (The stories behind all of it are the really interesting part, but for them, you’ll have to wait for the book!)

1. Pearlash was the first chemical leavener to appear, about 1760. By the 1790s it was commonly used around New York, then spread to New England and (as white settlement spread) the upper Midwest; it caught on less in the South. Pearlash, potassium carbonate (K₂CO₃) is a refined form of industrial potash. It can be reacted directly with an acid such as vinegar, molasses, or buttermilk, but it can be used without an acid as well. When dissolved in water, pearlash decomposes to potassium and carbonate ions; the carbonate becomes carbonic acid, which bubbles out as carbon dioxide. The potassium recombines with the water’s hydroxide ions to produce potassium hydroxide (KOH), which is also alkaline. In order to prevent it giving a bitter or soapy taste to the food, you have to add an acid to neutralize the potassium hydroxide, but you can get all of the leavening power without acidic ingredients.
2. Next came saleratus, potassium bicarbonate (KHCO₃), manufactured by bubbling carbon dioxide through pearlash. Saleratus has twice the carbonate of pearlash and thus twice the leavening power, but it does not decompose in aqueous solution; you must react it with acidic ingredients to use it as a leavener. That meant completely rewriting existing recipes, and the confusion seems to have slowed its adoption. Although first manufactured in the 1780s, it didn’t come into common use in the kitchen until the 1830s, and even then never entirely replaced pearlash.
3. Baking soda, sodium bicarbonate (NaHCO₃), was first marketed for baking in the 1840s — also under the name “saleratus,” which was and is confusing. But the new “soda saleratus” seems to have replaced both the potassium kind and the older pearlash within a few years. Baking soda has as much leavening power as saleratus, but it also undergoes heat decomposition at oven temperatures: in baking, that is, it decomposes to sodium carbonate, releasing half its carbonate ion and producing carbon dioxide. The remaining sodium carbonate must be reacted with an acidic ingredient to release its CO₂; otherwise, you run the risk of a bitter or soapy taste.

That’s good enough for a quick reference. But I wanted to confirm the actual leavening power of the three salts. In addition, “soda saleratus” was marketed as being a more effective leavener than potassium saleratus, and I didn’t entirely believe it. And even though my kitchen tests confirmed the general points above, I wanted a little data.

Thanks to Trish Lemm of the North Carolina Museum of Life and Science in Durham for helping me with these experiments and letting me use her equipment! Here’s what we did:

For each leavener, we dissolved 0.1cc of the substance in 20mL distilled vinegar (5% acidity) in a 1000mL Erlenmeyer flask and measured the resulting concentration of CO₂ in parts per million (ppm) using a Vernier CO₂ gas sensor. The results are shown in the table below. To calculate the weight of CO₂ produced, I subtracted the baseline reading of 1,300 ppm CO₂ and used this converter (and the fact that a teaspoon is 4.93cc). That conversion is problematic, because it depends on temperature, but the actual quantity of gas produced wasn’t my main concern; I was mainly interested in the relative quantities. The last column shows the quantity of available CO₂ compared to baking soda and is a gauge of actual leavening power.

Measuring carbon dioxide concentrations, equipment and photo courtesy of Trish Lemm. (full size)

This more or less conforms to my expectations, which we’ll call “theoretical predictions.” Potassium bicarbonate and sodium bicarbonate have essentially the same chemical structure; potassium and sodium are neighbors in the periodic table, with sodium just upstairs from potassium. Since sodium is a bit lighter than potassium, a given weight of sodium bicarbonate has a bit more carbonate ion — and therefore a bit more available carbon dioxide — than the same weight of potassium bicarbonate. “A bit” is good enough for the purposes of real-world baking, and in fact in my kitchen experiments, I couldn’t easily detect a difference between cookies leavened with baking soda and those leavened with saleratus. (I didn’t think it was worth setting up a blind-tasting panel.)

Oddly, although pearlash has half the carbonate and thus half the available CO₂ of saleratus by weight, in the test, it released only about 43% as much CO₂ as saleratus. The reason, I think, is that we measured the substances by volume, not by weight. The saleratus and baking soda were both finely powdered, but the pearlash was visibly granular, with round particles about the size of table salt crystals. A given weight of pearlash would take up a bit more space than the same weight of saleratus, and so an identical volume ought to release a bit less than half as much CO₂.

Why not measure the leaveners by weight? Measuring by weight would heave been more accurate, but that accuracy would be unrealistic. Nobody measures chemical leaveners by weight; they’re always measured by volume. In the 19th century they’d have been measured in whatever teaspoons one had around — not measured teaspoons but literally spoons for stirring sugar into tea — but always with the same spoon, presumably, so when someone switched from pearlash to saleratus or saleratus to baking soda, she’d have been using identical volumes.

The upshot of all this is that you can’t use the three substances interchangeably — a fact that created a lot of confusion in the first half of the nineteenth century.

Trish and I also ran some tests on different brands of molasses, but I’ll save those results for a later post.

Comments are closed.