In Class Experiment – Emulsions (Roux, Mayonnaise, and Hollandaise)


In class last week, we started by making three different types of roux: light, medium, and dark.  A roux is a mixture of heated flour and fat that is then used to thicken sauces, soups, and stews.  In a roux, the starch granules (from flour) are coated and separated by the fat from butter and dispersed throughout the mixture. The roux is cooked before the addition of any liquid to eliminate the flour’s “cereal” taste. However, the longer that flour is cooked, the more it breaks down into short chains of starch.  The shorter the chains, the less efficient they are at forming a thickening network.  As a result, a darker roux should yield a thinner sauce.  A roux can be used individually, or several different colors and flavors can be combined, as they are in many Cajun recipes.  Once a roux is formed, a liquid is added to the liaison (the thickening mixture).  As a result, the starch granules begin to absorb water.  The original liquid thickens to a sauce-like consistency.

General Procedure to make a Roux: We used each roux to thicken a basic cheese Sauce Mornay.  In a Saucier (a tall pan with rounded edges), we combined 1 tablespoon of melted butter with 1 tablespoon (10 g) of flour.  We whisked this constantly and cooked it for varying times (depending on the intended result), added ½ cup of whole milk and boiled the milk mixture.  Once the roux was made, we slowly added 136 g of shredded Cabot Seriously Sharp cheddar cheese (which is itself an emulsion of protein, fat, and water). We made this recipe three times, each of which is described below.

Sauce Mornay with Dark roux

We cooked the roux on low to medium heat past the development of a buttery, popcorn aroma (indicating diacetyl molecules) at 4 minutes, all the way to a “meaty” smell (signifying the evaporation of milk proteins from the butter) at 8 minutes.

At left, the diacetyl molecule responsible for the aroma of buttered popcorn.

We added the cold milk at 9 minutes, which created brown, solid, grainy particles.  After the milk began boiling, we turned the heat to low and slowly added the cheese.  This yielded a creamy consistency by 15 minutes, and we stopped cooking the mixture at 16 minutes.  The result was a grainy, tan mixture with a browned, nutty, and meaty taste.  It tasted more like the roux than the cheese due to the darkened roux’s intense flavor.

Sauce Mornay with Medium roux

The medium roux was browned on medium low until it exuded a nutty aroma (about 6 minutes).

At this point, the heat was turned to low and the milk was added.  The grainy consistency occurred again, but the mixture was a lighter brown.  Once the added milk had heated up, the roux thickened almost immediately (at 7 minutes), and so, cheese was added incrementally.   The mixture was creamier after adding the cheese and was removed from the heat after a total of 11.5 minutes.  The sauce was much creamier, smoother, and paler than the first.  It was also thicker, stuck easily to the bread we served it on, and tasted more strongly of cheese than of the roux itself.

Sauce Mornay with Light roux

Lastly, the light roux was cooked like the other two for 3.5 minutes until a very faint nutty aroma was detected.

Then, the milk was added, and because the pan was cooler at the time, it did not begin to boil until 6.5 minutes.  The cheese was added at 7.5 minutes, and the pan was removed from heat at 9.5 minutes. The final result was very white, much thicker than the first two, and by far, the cheesiest in taste.

The darkest roux had cooked the longest, and so it provided the most time for the starch molecules to break down into new flavorful components (like diacetyl molecules, for example).  It also browned, undergoing the Maillard reaction between sugar molecules in the flour and amino acids in the milk proteins (from butter).  This developed a new, complex flavor that overpowered the melted cheese.  Additionally, it was the least efficient at thickening because the starch molecules broke down into shorter segments while cooking.  They were unable to form a large complex network to thicken the sauce.  Instead they thickened by forming short, less efficient chains.  In contrast, the other two batches increased in thickening ability and cheesy taste as the roux cook time decreased.


Next, we made two batches of mayonnaise, one by hand and the other in a food processor.

By Hand: We mixed 1 egg yolk, 1 tablespoon of lemon juice, 1 tablespoon of cider vinegar, ½ teaspoon of dry mustard, and ½ teaspoon of non-Kosher salt.  The dry mustard and the salt were used to physically grind up the components of the egg yolk resulting in thousands of separated protein and fat molecules.  After whisking this mixture for two minutes, 1 cup of vegetable oil was added in small increments (even drops at a time) for the next ten minutes.  Eventually, another tablespoon of lemon juice and another tablespoon of vinegar were added.  The whole mixture was whisked by hand for 25 minutes straight.  It was good thing we took turns!!  The result tasted much “fresher” than Miracle Whip, but it also had a much stronger lemon taste. This can be eliminated by using two tablespoons of water instead of two tablespoons of lemon juice.  However, if you choose to make this substitution, Beware!  The overly acidic environment created by the addition of lemon juice (pH 2-3) is inhospitable for Salmonella bacteria to live.  When using uncooked egg yolks (as we did in this recipe), a high acid content is the safest for killing microbes.

In the Food Processor: We then made mayonnaise in about 1 minute in a Cuisinart food processor.  The recipe was the same as the first one; however, we used a whole egg instead of an egg yolk.  The yolk contains all of the egg’s emulsifying components, and in the case of the “by-hand” preparation, the egg would provide too much extra water.  In the food processor, however, the machine can mix much faster.  As a result, the blades of the food processor can move to quickly and rip the lecithin coating from the yolk.  Keeping the white creates a “safety net” to protect the lecithin.  We were also able to add all of the ingredients (except for the oil) at once.  Then, we poured the oil through the top of the food processor while it ran.

Luckily, we did not make any of the following mistakes that Alton Brown advises against.

“Mayo Clinic” Maladies

Adding oil too quickly

Whisking too slowly

Adding too much oil after the emulsion forms

Using oil that is too warm

When these issues arise, the final product is a “broken” mayonnaise, which has separated into globs of fat floating on water.  It is relatively easy to fix a broken emulsion because the eggs are not cooked.  Because eggs yolks are emulsions of water and fat, the bad mayo can be incorporated back into a beaten egg yolk and dispersed into a proper consistency. Simply beat another egg yolk in a separate bowl and then add it incrementally to the broken mayonnaise.

Hollandaise Sauce

Finally, we made two batches of Hollandaise sauce (an emulsion I know best from Eggs Benedict).  Hollandaise literally means “from Holland,” in French, and it is a yellow sauce made with butter, egg yolks, and lemon juice or vinegar. The butter contributes milk proteins and some moisture to the sauce.  In this recipe, the mixture is heated very gently, to encourage the dairy proteins and the egg proteins to denature (unfold) without coagulating (clumping together).  When cooked properly, the proteins form a network that traps the moisture creating a rich, smooth Hollandaise.

1st Batch – Butter: Our first batch was made with 3 egg yolks and 3 tablespoons of lemon juice, which were whisked and heated on low heat for 4 minutes. At this time, 1 stick of melted butter was added and the mixture was stirred until 1 tablespoon of lemon juice was added at 7 minutes.  The sauce was stirred and heated gently for a final total of 11 minutes.  ¼ teaspoon of salt was added, and the resulting sauce was removed from the hot pan to prevent the eggs from scrambling.

2nd Batch – Margarine: Our second batch was the same, but we used 1 stick of margarine instead of butter.  The result was less buttery and more lemony in taste along with an overly “clumpy” appearance.  Because they are not present in the margarine, the observed “clumps” probably formed from curdled egg proteins within the Hollandaise.  We heated both batches in the same way, and so, this was probably due to the fact that margarine has no lactose, casein, or saturated fat.  Normally, these components block egg proteins from coagulating.


2 responses to “In Class Experiment – Emulsions (Roux, Mayonnaise, and Hollandaise)

  1. Great stuff guys… I’m trying to make Miracle Whip at home following the 1930s ingredients list, which consisted of simply: Oil, Vinegar, Sugar, Arrowroot, Cornstarch, salt & spices (in that order)… Now, I know cornstarch and arrowroot powder were used so that they needed less eggs which therefore cut costs, but my problem is that I’m not sure how to incorporate the cornstarches without mixing it with water and heating it (the ingredients don’t list it)… Do I mix them with the vinegar instead, and then heat as per usual, and mix it in with the mayo base? (There’s about 1/2 cup of vinegar to 2 cups of oil in my estimation…)

    Thanks for any help you can give.

  2. I think you got way to much vinegar for the mayo not to be runny, probably more like 1/4 cup at most, also for the starches you can cream them into the eggs like you would do sugar for a custard, then add the oil and finally thin and season with the vinegar and seasoing

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s