Thursday, November 24, 2016

Substitutions by Macronutrients / Vegan croissants and eggwash

Summary: 
  • Soy milk 3x concentrated (ex: 3 cups heated on the stove until 1 cup remains) is an effective, glossy and great-looking vegan eggwash
  • Croissants made with peanut butter rather than butter can produce laminations, but taste a bit savory, more like peanut bread than a croissant, due to sodium content.
  • Future directions for peanut butter-based laminated dough may be as kouign amanns, where the extra caramelized sugar on the outside may balance the peanut butter savoriness
Peanut butter-based croissants
As you may have guessed from some of my previous posts, I enjoy experimenting with laminated doughs, and particularly with croissants.  Similar to the last post of scallion pancake croissants, which used a butter-sesame oil mixture to laminate dough layers, I wanted to investigate if butter could be substituted entirely in a croissant recipe.

Left: ~9 tbsp of peanut butter used to laminate normal croissant dough. Center: Triangles cut from laminated dough. Right: Resting and proofing peanut butter-based croissants


I believe I have seen recipes that make vegan croissants using vegan butter-mimicking products, but I was curious if there were any possible substitutes that could lend some more character and flavor to the finished product.  To investigate this, I did my usual macronutrient-based analysis - comparing substitute ingredient fat, protein, carbohydrate, and water content to the original ingredient.  I found that butter and peanut butter had similar macronutrient content (peanut butter: 50% fat, 18.75% carbohydrate, 25% protein, 6.25% water by weight, butter: 84.5% fat, 0% carbohydrate, 0.7% protein, 14.8% water) in the two macronutrients that likely mattered the most, fat and water content, since protein and carb content of butter is negligible.  A direct substitution led to an interesting baked good - the dough handled similar to croissant dough, and had visible laminations, however, the taste was very savory, rough, and bread-like, and the layer separation was quite small.  The lower fat content of peanut butter could result in less tenderizing of the flour.  The higher protein content likely resulted in a somewhat dryer product, and the lower water content may have resulted in less lift.  Small adjustments in each may be necessary in future iterations (adding a few tablespoons of water or oil).  Additionally, I neglected the sodium content of each ingredient - typically I make croissants with unsalted butter, with 11 mg of sodium per 100g while peanut butter has ~30mg of sodium per 100g.  As a result, the final product was not over-salty, but definitely a savory-type bread.  By dusting the surface with sugar, the flavor was more balanced, suggesting that peanut butter-laminated doughs may be more useful for something like a kouign amann.

Finished product, with the center croissants using the vegan eggwash substitute described below

Layers visible in croissant cross-section

Since peanut butter is a vegan product, I thought I should investigate whether I could use a similar approach to identify effective vegan eggwash substitutes.  After some brief googling, I did not find many satisfactory images, but a few suggestions of ingredients to consider.  In the graph below, you can see the conditions I decided to test:



Again, I saw that the largest components of eggwashes by mass are water and fat, so I chose a concentration factor for soymilk and unsweetened almond milk that would result in similar water and fat content.

From left to right: untreated, soy milk, 3x concentrated soy milk, unsweetened almond milk, 10x concentrated almond milk on scraps of peanut butter croissants before (left) and after (right) baking at 425 F for 18-25 minutes

As you can see, the concentrated soymilk created a very nice finish on the baked goods in the above image as well as the croissants seen further above.  The high sugar content may contribute to higher browning than even in the regular eggwash, and the protein content appears to have given a nice glossy finish.  Despite having similar protein content, the concentrated almond milk did not perform well - this could be because manufacturers are often allowed to round nutrient contents below 1g up, so the actual macronutrient content may be much lower.