Week 4: Angel Food Cake

Last week, I settled on a base recipe to play around with. If there's anything I learned in the past few weeks, it is that angel food cake is very technique-dependent. One of the most hotly debated topics in angel food cake recipes is meringue stiffness: soft peaks, full peaks, stiff peaks! To better understand this, let's dig into the science behind meringue and put it to the test!

Meringue Science

Meringue is a foam where small air bubbles are suspended in egg whites. I like to think of foams similarly to an emulsion (see lemon curd week 1), but instead of one liquid suspended in another, we have air bubbles suspended in a liquid. The act of whipping adds air bubbles and, much like an emulsion, breaks up larger air bubbles into smaller bubbles. To make a stable foam, we want to prevent the air bubbles from coalescing.

When you whip egg whites, you also mechanically unfold or denature the egg white proteins. Some of the exposed components (or amino acids) in the denatured egg white proteins are hydrophobic and thus move away from water. Egg whites are ~90% water, so the main place these hydrophobic amino acids can escape the water is at the surface of the air bubbles. The egg proteins thus gather on the surface of the air bubbles (much like an emulsifier) and form bonds, making more stable walls around the air pockets.

However, this network of protein-coated air bubbles is not strong enough to hold for long periods of time, which results in the water separating from the foam. By adding sugar, we dissolve sugar molecules in the water from the egg whites, creating a sugar syrup. This syrup is much thicker, or viscous, than the water alone, so it makes it more difficult for the air bubbles to move through the liquid and coalesce. This is similar to how coagulation thickened the liquid in lemon curd, preventing separation. Thus, the addition of sugar results in a more stable meringue. This is also why some recipes call for “aging” egg whites in the fridge. Aging egg whites allows some of the water to evaporate from the eggs, giving us a thicker liquid to begin with and thus a more stable meringue.

Acids can also denature proteins (see lemon curd). In meringues, adding an acid aids in denaturing the proteins in the egg whites, which helps the meringue whip up more quickly. Some types of bonds between amino acids on the protein chains are too tight and pull the proteins together, collapsing the foam. The most common of these is a Sulfur-Sulfur bond, which is formed when Sulfur-Hydrogen bonds give up their Hydrogen and nearby Sulfurs bond together. By adding an acid, we also add a higher concentration of Hydrogen ions, which makes it much more difficult for Sulfur-Hydrogen bonds to lose a Hydrogen. This means that acids also help make the meringue more stable. Vinegar, lemon juice, and cream of tartar are some common acids added to meringues. I like cream of tartar for angel food cake because it is a strong acid that doesn’t add any additional water (like lemon juice and vinegar do), and it has a fairly neutral flavor in the quantities we need for angel food cake.

The Great Meringue Debate

Now that we understand the basics of meringue, we can look at the hotly debated topic of meringue stiffness. The longer you whisk a meringue, the more air you will incorporate, but more whisking also makes more protein bonds, forming a tighter network. Like many physicists, I love to approximate things as a spring, so when I imagine this process, I think about stretching a spring out. Whipping more adds additional air, but it also stretches the spring out more. When we are at stiff peaks, we have essentially stretched the springs to their fullest extent, and whisking more will only cause them to pull back and contract, causing a gritty, broken meringue. At soft and firm peaks, the springs aren't fully stretched.

For angel food cake, the stiff peak crowd argue that more air equals more rise, so the best rise comes from incorporating the maximum amount of air before baking through whipping to stiff peaks. Folks on the soft/firm peak side argue that if you stretch the springs out to their fullest extent, they can't stretch more during baking as the air trapped in the bubbles expands in the hot oven, leading to less rise. They also argue that a stiffer meringue is likely to get more air knocked out as you mix in the flour and remaining sugar because it is more difficult to mix.

Both arguments are compelling, but let's put this to rest with a side by side comparison. I will try a few different whipping points to compare: soft peaks, just to firm peaks, firm peaks throughout, and stiff peaks. The meringue around the whisk is usually stiffer than the meringue on the edges of the bowl. Here "just to firm" means that the meringue around the whisk is at firm peaks while some of the meringue on the edges has softer peaks. Firm peaks throughout indicates that the meringue around the edges is also firm. I will use the same technique to make my meringue in each case. I'll start by whipping on medium-low until I see some bubbles, add my salt and cream of tartar, and whisk on medium-high until I have a foam close to soft peaks. Next I will add the sugar in slowly on medium (~30 s).  Adding the sugar too early can inhibit the formation of foam, and adding it too quickly can collapse air bubbles under the weight of the sugar. Finally, I will whip on medium-high until I reach my desired consistency.

Soft peaks are droopy.

Firm peaks stand taller but still have a droopy tip.

As you continue to beat the meringue, the firm peaks get taller and less droopy.

At stiff peaks, the meringue stands straight up.

Since I am using the same amount of batter, the height of the final product gives a good quantitative measure to compare the rise. The soft peak batch was 6.7 cm high, and the batch I whipped just to firm was 6.5 cm high. I would say that this difference in height is roughly within the uncertainty of my measurements as thickness can vary slightly across the surface of the cake. These two batches had indistinguishable texture. The batch I whipped to firm peaks throughout was also 6.5 cm tall. While it was still very tender, its texture was slightly tougher than the other two batches. The stiff peaks batch only rose to 6.0 cm, which is a significant difference in rise, and it was the toughest batch. Additionally, the flour and remaining sugar were also easier to incorporate into the soft/firm peak batches. I interpret these results to mean that any stage before stiff peaks will give you about the same rise, but going to soft/just firm peaks gives you a slightly more tender texture. The soft and just firm batches were virtually indistinguishable, so anywhere in this range will give the best results. Next week, I'll look into the debate between granulated and powdered sugar.

From left to right: soft peaks, just to firm, firm throughout, and stiff peaks. You can see that the top of the cake is flat in the soft and just to firm cakes. As you go to firm throughout and stiff peaks, you start getting a higher edge and slope across the top that is more pronounced as the meringue gets stiffer. This could indicate that some of the structure is collapsing as the cake bakes, which would be consistent with the proteins being overextended.