Case Study 1 — Danny Reyes-Park's Aquafaba Pavlova for His Vegan Cousin

The phone call came in late September. Danny Reyes-Park's cousin Camila was getting married in November, and Camila — vegan since college — had asked Danny if he would make her wedding dessert.

"It has to be a pavlova," Camila said. "The one that grandma always made."

"Tía Sofia's pavlova."

"Yes. But vegan."

Danny held the phone to his ear in the small apartment kitchen above his roommate Tomas's bedroom and looked, for a long moment, at the ceiling. Tía Sofia's pavlova. The dessert that had appeared at every Christmas, every Easter, every quinceañera, every wedding for as long as Danny could remember. A vast oval mountain of meringue, baked until it was crisp on the outside and marshmallow-soft inside, topped with whipped cream and a riot of fresh fruit — passionfruit, mango, kiwi, strawberry, depending on the season. The pavlova that Sofia, his Korean great-aunt who had married into the Mexican side of the family in the 1970s, had perfected over forty years and that nobody else in the family had ever quite been able to replicate.

The pavlova was egg whites, sugar, cornstarch, vinegar, vanilla. It was a foam in its purest form. And Camila wanted a vegan version.

"How long do I have?"

"Eight weeks."

"And you trust me with this?"

"Danny."

"Yeah. Okay. I'll figure it out."

He spent the next two weeks reading. The first thing he learned was that vegan meringues were a real thing — aquafaba meringues, made from the protein-rich liquid drained from cans of cooked chickpeas. The technique had been documented in 2014 by a French amateur cook named Joël Roessel, who'd noticed that the foam from chickpea cooking water was unusually stable, and the term aquafaba (Latin for "bean water") had been coined by an American baker named Goose Wohlt shortly after. Within two years, vegan baking communities had built an entire genre of aquafaba meringues, mousses, marshmallows, even macarons. The chemistry was real.

The chemistry, when Danny tracked it down in a 2019 paper from the Journal of the Science of Food and Agriculture, was satisfying. Aquafaba is roughly 1-2% protein by weight (about a tenth of egg white's protein content), but the proteins are predominantly legumin and vicilin — globulin storage proteins from the chickpea seed. These proteins, like ovalbumin in egg whites, can partially denature at the air-water interface and form a stabilizing film around bubbles. The foam is not as dense as egg-white foam, but it can hold soft and stiff peaks, and with sugar added it can survive in the oven long enough to form a meringue shell.

Two further chemistries help: aquafaba contains saponins (from the chickpea coat) which are themselves surfactants and assist the proteins in stabilizing the foam, and starch from the cooked chickpea, which raises bulk viscosity and slows drainage. The combination — proteins plus saponins plus starch — produces a foam meaningfully more stable than any single-component vegan substitute.

But there were problems.

Danny's first attempt was a textbook failure. He drained the liquid from a 540 mL can of organic chickpeas (yielding about 250 mL of aquafaba), poured it into the stand mixer with cream of tartar, started the whisk on medium, and waited. The foam built up slowly, taking about 8 minutes to reach soft peaks (compared to 2 minutes for egg whites). He added sugar gradually. The foam climbed to stiff peaks in another 6 minutes. So far, so good — though slower than egg whites.

He scooped the foam onto parchment in a wide oval, used the back of a spoon to make a shallow well in the center for the eventual cream and fruit, and put it in a 110°C / 225°F oven. Two hours later he checked. The pavlova had collapsed almost completely, leaving a sticky, gummy, half-inflated mound that was browned on the outside and damp inside. The structure had not held.

Second attempt: more cream of tartar, longer whipping, slightly higher sugar ratio. Same result — collapsed, gummy. The foam had risen, but the matrix had not solidified into the crisp shell that a pavlova requires.

Danny called his roommate Tomas, who was a microbiology grad student. "I think I'm under-dehydrating."

"Or over-dehydrating," Tomas said. "Or the protein content is too low. Or the proteins are wrong."

"Which is it?"

"Run more experiments."

So Danny ran more experiments. He tried reducing the aquafaba by half on the stove before whipping, concentrating the proteins. (Better stability, but the residual chickpea flavor became too prominent.) He tried adding cream of tartar plus a touch of guar gum for additional bulk viscosity. (Improved structure, but the texture was rubbery.) He tried baking at much lower temperatures (95°C / 200°F) for much longer times (4 hours). (Improved drying, but flavor still off.) He tried adding cornstarch to the meringue itself — which is the trick that makes a classic egg-white pavlova have its marshmallow center, since the starch absorbs water and creates a soft interior. (Big improvement.) He tried adding a small amount of vinegar — same as Tía Sofia's recipe — to the foam. (Surprising improvement; the lower pH seemed to help the legumin proteins associate.)

By his fifth attempt, Danny had a recipe that worked. The protocol:

1. Drain the liquid from two cans of unsalted organic chickpeas (about 500 mL).
2. Reduce the aquafaba by 30% on the stove (concentrating the proteins) — about 12 minutes at a low simmer. Cool to room temperature.
3. Whip with ¾ teaspoon of cream of tartar at medium-high speed for 8 minutes to soft peaks.
4. Add 200 g of sugar, one tablespoon at a time, while whipping. (The slow sugar addition is critical — too much too fast collapses the developing protein film.)
5. Add 1 tablespoon of cornstarch slurry (1 tablespoon cornstarch in 2 tablespoons cold water) and 1 teaspoon of white vinegar. Whip another 2 minutes to fully combine.
6. Spread on parchment in the desired shape, well in center.
7. Bake at 95°C / 200°F for 3 hours, then turn the oven off and leave the pavlova inside for another 1 hour to dry slowly.
8. The shell should be crisp, the interior softly marshmallow-textured.

The ninth attempt at the recipe was for the test run, three weeks before Camila's wedding. Danny made the pavlova in the morning, topped it in the afternoon with whipped coconut cream (full-fat coconut milk chilled overnight, the solid layer scooped off and whipped with a touch of powdered sugar), and arranged passionfruit, sliced mango, kiwi, and strawberries across the top in roughly the same composition that Tía Sofia would have used for a fall wedding.

Camila came over in the evening to taste it.

She took one bite, paused, took another, looked at Danny across the kitchen table and said: "How did you do this."

Danny told her about the aquafaba, the protein chemistry, the cornstarch interior, the bean-cooking-water foam stability research from 2017. He talked for about ten minutes before he realized she was crying.

"It tastes like Tía Sofia's pavlova," Camila said. "Exactly like it. The crisp shell, the soft inside. The fruit. I haven't had this since I went vegan."

The wedding pavlova was a triumph. Camila served it to a room of 80 guests, more than half of whom had no idea it was vegan. Several people — including Danny's father, who had grown up on Tía Sofia's pavlovas — could not tell the difference. Tía Sofia herself, then 82 years old, took a bite and looked at Danny with something close to satisfaction.

"You used the chickpea water," she said. (Her English had grown more limited in recent years, but her knowledge of food chemistry had not.) "I read about this on the Korean baking forums two years ago. I always wondered if anyone would try. You did good."

What Danny had actually learned

Looking back on the project, Danny made several entries in his lab notebook:

1. Aquafaba foam is real and reproducible, but the protein content (~1-2%) is much lower than egg white (~10-12%). The foam therefore takes longer to whip, builds slower, and is more sensitive to mistakes. The trade-off for Camila — and for any home cook with vegan friends or egg allergies — is reasonable but not free.

2. Reducing the aquafaba concentrates the proteins and meaningfully improves foam stability. This is a small modification with disproportionate payoff. Most published recipes don't include this step; Danny added it after his second failure.

3. The starch addition (cornstarch slurry) does the same job as in classical pavlova — absorbs water, creates the marshmallow interior, contributes to bulk viscosity that slows drainage during the long bake.

4. The acid addition (vinegar) helps. Aquafaba's legumin proteins are more soluble at slightly acidic pH, similar to but distinct from egg-white proteins' isoelectric-point chemistry. The same logic, different proteins.

5. The bake temperature must be lower and longer than for an egg-white pavlova. Aquafaba foam is more delicate; high heat too soon causes the foam to collapse before the matrix can set. The 95°C / 200°F slow bake is critical.

6. Aquafaba's residual chickpea flavor is detectable in plain meringue but masked by sugar, vanilla, and acid. Above a certain sugar:aquafaba ratio (about 4:1 by weight), the chickpea flavor disappears. Pavlova's high sugar content is fortunate for this purpose.

7. Cultural inheritance can be carried by chemistry. Tía Sofia's pavlova was a specific texture-and-flavor combination that Camila had grown up with; Danny did not need to use exactly the same ingredients to reproduce that experience. He needed to reproduce the chemistry — the foam, the matrix, the contrast between crisp shell and soft interior, the dairy-cream-and-fruit topping. With aquafaba in place of egg white and coconut cream in place of dairy, the dessert was molecularly different but sensorially identical. This is a remarkable thing about food: the experience can be detached from the specific molecules that produce it, if the cook understands which molecules are doing the work.

The wedding pavlova has, since Camila's wedding, become a small Reyes-Park family thing. Danny has made it three more times for various vegan and egg-free family events. Tía Sofia, who has not stopped making her own pavlova at her kitchen counter into her ninth decade, has asked Danny to write the recipe down so she can teach the chickpea version to a younger cousin.

"For the future," she said. "When the family changes."

This is theme #4 of the book — food traditions are accumulated scientific knowledge — meeting theme #2 — understanding why gives you power. Tía Sofia knew her pavlova chemistry as well as anyone alive, even without the chemistry vocabulary. Danny inherited the knowledge and adapted it for a different ingredient set. The next generation will inherit the adaptation. And the dessert, which is a chemistry of foams from the very first whisk, persists.

Analyze This

1. The protein concentration problem. Aquafaba is roughly 1-2% protein by weight; egg white is roughly 10-12%. How does this lower protein concentration affect the kinetics of foam formation? What are two specific ways Danny compensated for the low concentration?

2. The role of saponins. Aquafaba contains saponins — surface-active compounds from the chickpea coat. How might saponins assist the legumin and vicilin proteins in stabilizing the foam? What does this suggest about the difference between a "pure" protein foam and a "mixed-surfactant" foam?

3. Why cornstarch matters in pavlova. Tía Sofia's recipe (and Danny's adaptation) both include cornstarch. What is the cornstarch doing during the meringue stage? What is it doing during the bake? Why does it produce the characteristic marshmallow center?

4. Cultural transmission via chemistry. Tía Sofia learned about aquafaba "on the Korean baking forums" (the vegan and food-allergy adaptation of traditional foods is widely discussed in Korean cooking communities). What does this tell you about how food traditions adapt to changing dietary practices? Could a tradition lose its identity if its ingredients change too much, or does the chemistry preserve identity even as ingredients change?

5. The chickpea-flavor masking problem. Plain aquafaba meringue has a detectable chickpea flavor; sweetened aquafaba meringue does not. Above a sugar:aquafaba ratio of roughly 4:1 by weight, the chickpea flavor disappears. Reason about why high sugar content masks legume flavor. (Hint: this is partly about taste-receptor competition and partly about the chemistry of aroma compound retention by sugar matrices.)