Wednesday, December 12, 2012

Macaron Making Made Easy! / tips for reproducible and quantifiable macaronage

Hey guys, these past few weeks I decided to tackle another trendy food out there: macarons! A well made macaron has two meringue-based cookies that each have a crunchy exterior with a mildly sweet, soft almond interior, and a rich complementary filling.  Furthermore, the cookies should have a smooth, uncracked shell/crown, no air pockets inside, and feet (the rough ring below the smooth shell).  Fun facts:  Just like how you can order poutine at McDonald's in Canada, in France, McDonald's sells macarons at some branches.  McD's macarons are made by a company owned by Groupe Holder, who also happen to own LadurĂ©e (the original pastry shop that invented the double-decker macaron, now a chain of luxury pastry shops where you can buy macarons at a price of ~$3 a pop).


Left: Raspberry mascarpone macaron, Center: Banana custard macaron, Right: Dragonfruit French buttercream macaron

It is said that the making of macarons is the true test of a French pastry chef.  With only four ingredients, I knew that making this cookie would most likely involve some very specialized technique.  However, when researching the recipes available online, I was shocked by the wide range of methods used.  Some recommended steps, like leaving raw egg whites out for a day in advance, or needing to fold the batter exactly 40 times sounded like voodoo magic to me.  There had to be some way to standardize and quantify this otherwise elusive process.

After compiling about a dozen recipes online, I averaged the amount of each ingredient across the recipes.  Furthermore, I reasoned that even though each recipe had its own nuances, all the recipes must have the same bare essential steps in common.  These are: 1) Beating egg whites with sugar to make a stiff peak meringue, 2) Sifting confectioner’s sugar and almond flourinto the meringue, 3) Macaronage – or the “art” of folding the batter to just the right stage, 4) Piping the batter out and letting it rest, and 5) Baking of the macarons.

Step one: Making Meringue
The only fuss with meringues I saw were that some recipes called for making a simple syrup out of your sugar prior to adding it to your egg whites, and the majority of recipes saying to add sugar slowly during the beating process.  When the goal is to simply create a stiff-peak meringue, neither of these steps are necessary.  The rule ended up following with my meringue was – it’s better to incorporate too much air than too little.  I made my meringue by dumping in all the granulated sugar into the egg whites and mixing with a handmixer until I achieved stiff, glossy peaks (when you remove the handmixer from the meringue, two peaks the stand straight up are formed).  At this point, I added any desired coloring and mixed for an arbitrary additional minute.

Left: Meringue beaten to the point of stiff peaks, when removing handmixer from meringue, a good amount sticks to the beaters. Center: Adding food coloring. Right: Finished and colored meringue with stiff peaks

Step two: Combining Particulates and Meringue
Some recipes were more fussy about needing to sift your confectioner’s sugar and almond flour together 2-3 times prior to adding to the meringue, and ensuring that the two dry ingredients are definitely well-combined.  Some recipes specify that you must ensure that your confectioner’s sugar contains no corn starch, since (gasp!) some brands of confectioner’s sugar are in fact, cut with corn starch.  I took a look at my box of sugar, saw ‘corn starch’ listed as an ingredient, shrugged, and threw it in.  My thoughts were that they are both sugar-based particulates, so a small amount of corn starch probably wouldn’t be a big deal (spoiler: it wasn’t a big deal).  Lacking access to a sifter/food mill and lacking the patience to sit around sifting powder together, I took a simpler approach.  My thoughts were that since Step Three will involve lots of folding of the macaron batter, there shouldn’t be any need to painstakingly combine the sugar and almond flour. The sifting is actually important, since the folding of the batter will not be vigorous enough to break up large lumps of flour or sugar.  To add the particulates (almond flour + confectioner’s sugar) to the meringue, I dumped both together into a colander and shook it above my bowl of meringue until all of the dry ingredients that would pass through had passed through.

Left: Macaron ingredients - almond flour and confectioner's sugar in collander, sugar and egg whites in bowl. Center: my sifting contraption of a collander over the bowl of meringue, with a larger pot to catch spills. Right: Cool pattern of particulates generated from sifting
Step three: Macaronage
Yes, macaronage is the actual French word for the act of making macaron batter.  Every recipe agreed that this is by far the most important stage.  What happens when you combine the particulates and the meringue and do any additional mixing is the air previously incorporated in the meringue is forced out.  If too much air is kept in the batter, then your macarons may develop air pockets and your shells may crack.  If too much air is forced out of the batter, then your shells may collapse.  Recipes varied in their suggestions for reaching this all-important stage, from saying you should wait until it looks “like magma”, or if you drip a bit from your spatula and it re-incorporates in 10 seconds (some recipes said less, some said 20).  All these suggestions are trying to do is find a way of approximating the batter viscosity, so I figured there must be a better way to quantify when a macaron batter is ready.  Without any fancy lab equipment like a viscometer or rheometer, I tried to use some simple, reproducible ways of approximating viscosity and testing the batter at home. 

Left: My first attempt at a technique for measuring batter viscosity, dropping a mentos and timing descent. However, this did not yeild very precise results when dropped in water, honey, and oil.  Center: Another failed attempt at measuring viscosity, mainly since the rate of flow of batter out of the funnel depended in part on the height of the funnel (the batter flowed faster if held low, since a continuous stream was formed). Right: Using a straw to withdraw enough batter to make a single test macaron.
The most valuable test I came up with was to simply use a straw to suck up about ¾ straw’s worth of batter, squeeze the batter out onto a bit of parchment paper, and use that as a test cookie (rather than piping out the whole batch and finding out the batter wasn’t right). The techniques of dropping a small round object into the batter or running the batter through a funnel proved to be either totally not workable or not very precise. What I ended up using to approximate viscosity was a piece of paper (8.5” x 11”) placed on a board and leaned at an incline, producing a 10-5.5-8.35 right triangle.  I placed a tablespoon of the batter onto the paper between the 11” an 10” mark on the paper, raised the paper onto the incline, and timed 60 seconds.  After 60 seconds had elapsed, I lowered the paper and marked how far the batter had travelled.  This allowed me to compare viscosities during the macaronage folding process, and also between batches. 

Once I had found an ideal batter-travel-distance (BTD) corresponding to batter at the correct stage for making macarons, I tried a little experiment.  For my successful batch (the red macarons), I thought the macarons looked flatter than I would’ve liked.  When I looked at the cross section of a macaron with an air-pocket, I thought, “well, maybe the air pocket wouldn’t be there if there were more batter between the base and the shell” – aka, more particulates were needed.  For my subsequent batches, I tried using a quarter cup more of each almond flour and confectioner’s sugar, and again measured the BTD.

Left: A diagram of the tilted-paper batter-travel-distance (BTD) setup.  Right: The two hashmarks on the left were the ideal BTD for 0.75 cups of each almond flour and confectioner's sugar to 2 egg whites' worth of meringue.  The two hashmarks on the right and the batter pictured are the ideal BTD for 1 cup of each almond flour and confectioner's sugar to 2 egg whites' worth of meringue.  The ideal batter using more flour & sugar was significantly more viscous (2.375" inches travelled for low viscosity, 1.625" travelled for high viscosity).

The results – I found that with increased particulates in the batter, the batter was ready to bake at a higher viscosity (also meaning, less mixing was required and presumably more air was retained in the batter).  How could this be?

Left: High particulate concentration, primarily almond flour, prevents air bubble coalescence (purple spheres represent almond flour particles, yellow strands represent protein structure of the meringue, orange hexagons represent dissolved sugar from confectioner's sugar, blue bubbles represent air bubbles incorporated into the meringue, and the blue backdrop represents the water present in the batter from the meringue).  Right: By decreasing the (almond flour+confectioner's sugar) to meringue ratio, air bubbles are allowed to coalesce.  If allowed to bake, these large air bubbles will further coalesce and form air pockets in the macaron.

Well, macaron batter is basically a suspension of particles (almond flour and confectioner’s sugar – although I suppose the sugar may dissolve more than the almond flour) and small air bubbles in liquid (water from the egg whites).  What causes viscosity are the interactions and collisions between the particulate components, so having more almond flour + confectioner’s sugar means a more viscous batter.  Air pockets are formed when the small air bubbles in the batter coalesce.  Since large bubbles are more buoyant than small bubbles, these larger bubbles rise to the top during the baking process, resulting in a large pocket of air directly between the shell and the interior of the finished cookie. However, the viscous interaction between particles in the suspension prevent these bubbles from joining together.  What this translates to is – using a greater proportion of particulates (almond flour+confectioner’s sugar) to meringue means the batter can handle more air, will require less mixing, will be ready to bake at higher viscosity, and will result in taller macarons.  What this also implies is that any two recipes that differ even slightly in particulate-to-meringue proportion will have different ideal batter consistencies or, when two people talk about macaronage and the right consistency, there’s a good chance they’re not even talking about the same thing!

While the BTD test worked well for me, I’m hesitant to recommend it for all readers since I don’t know how reproducible it will be for everyone.  What I can recommend is for you to mix your batter to a stage where it flows off your spatula in continuous ribbons (err on the thicker viscosity side), use a straw to pull off one cookie’s worth of batter, and bake a test cookie.  If the shell cracks or air pockets form, give the batter about 20 more folds and try again.  You can set up your own BTD test once you have batter of the right consistency so that you’ll know about what viscosity to fold to every time.

Step four: Pipe and Rest
At this stage there was a good amount of consensus between recipes.  Use a piping bagto pipe out circles of batter about 1”-1.25” in diameter onto silpat or parchment paper.  Hit the pan flat on a surface a few times to force out any large air bubbles.  If your ideal batter viscosity is high, your batter circles may not flatten out very much on their own, and may have little points.  You can flatten these out by wetting your finger under the faucet and pressing the points down flat.  At this point, let the piped batter sit for upwards for 30 minutes.  This may take as long as over an hour, but I have found this step to be critical.  What you are looking for is for the batter to not stick to your finger or feel moist at all when touched lightly – it should almost have the feel of rubber.  When I ignore this step and put in circles of batter without any rest, even if the batter is the right consistency, I always get cracked or poorly formed shells.  I’ve heard that some others are able to avoid this step, it is possible that there is some combination of meringue composition and oven settings that make it so that the shell dries out on its own in the oven at just the right speed, but I haven’t found it.  

Left: Red macaron batter resting. Left Center: Red macarons after baking. Right Center: Yellow macaron batter resting.  Right: Yellow macarons after baking.  Note the larger height of the yellow macarons, which were made with more almond flour+confectioner's sugar, but with the same amount of meringue.
Step five: Bake
What worked for me was a little over 15 minutes at 300o F,  I’ve seen others use less time and higher temperatures, all the way up to 400o F, although I wouldn’t recommend it.

And you’re done!

Now the fun, and in my opinion, considerably easier part: picking a filling.  For my three, I wanted to have a little diversity and use things people maybe don’t see in macarons so much.  Some filling ideas:

  •         White chocolate mousse (can be flavored with any number of fruit syrups, ex: cherry, raspberry, etc. See my first post, http://mobiusbacon.blogspot.com/2012/10/the-greatest-cupcake-ever-made-honey.html)
  •         Mascarpone & raspberry puree (can be adapted to any fruit, or can use jam as a substitute, blend fruit in blender, sieve out seeds if desired or necessary, and then mix/blend with mascarpone, roughly 1 tbsp of puree to ½ cup mascarpone, with salt to taste)
  •         Chocolate ganache (melt chocolate, mix with butter and/or heavy cream)
  •         Banana Custard (blend banana, egg yolk corn starch, sugar, milk, vanilla, salt and heat on the stove till thickened) can be adapted for oranges, lemons, by instead simmering milk with ~one fruit’s worth of juice and 3-4 fruits’ worth of zest, whisking together eggs and all other ingredients, tempering the eggs with the milk, then adding the eggs to the milk and heating on stove till thickened.
  •        Straight up jam – really any kind of jam, just match the color and you’re good to go
  •         Some iteration of peanut butter – straight up peanut butter or made into a mousse, ganache, or just thick sauce
  •         Dragonfruit French buttercream – (as with mousse or mascarpone, just make the base buttercream and fold in fruit syrup).  With pink dragonfruits, heat pulp on stove, strain through coffee filter overnight, heat collected syrup on stove till reduced by half. Add to this volume the balance in recommended sugar in French buttercream recipe, and heat until dissolved and bubbling.  Temper eggs, then slowly whisk in the all of the syrup.  Let cool to room temperature, then whisk in butter 2 tbsp at a time.
  •         Classic American buttercream – (again, add whatever flavorings you want). Cream ~1 stick butter, add 1.5 cups confectioner’s sugar and whisk until combined, add vanilla, salt, and other flavorings to taste.
  •         Thai tea buttercream – make a concentrated batch of thai tea (bring 1 cup water to a boil with 2 pods star anise, add 4 bags black tea, let steep for 30 minutes so that flavor is extra concentrated, stir in 1 tbsp sugar).  Fold in 2-3 tbsp into a batch of either French or American buttercream as described below. Note: haven’t tried this, this is just speculative.

I bolded the ones that I actually used in my pictures.  Red had mascarpone+raspberry jam, yellow had banana custard, and purple had dragonfruit French buttercream.


Left: Egg yolks beaten till frothy/lightened in color. Left Center: Egg yolks beaten with dragonfruit syrup. Right Center: Softened butter beaten into yolk/syrup mixture. Right: Finished dragonfruit French buttercream.  Note: I thought the flavor was a bit strong, so I used a fair amount of vanilla extract and mixed just part of this with a batch of unflavored buttercream before spreading on my macarons.
Some helpful troubleshooting resources:
http://foodnouveau.com/2011/12/16/destinations/europe/france/a-macaron-troubleshooting-guide-useful-tips-and-advice-to-master-the-french-delicacy/

http://bravetart.com/blog/MacaronMyths

For the recipes, check out the links on the side of the page to learn how to make french macarons filled with raspberry mascarponebanana custard, or dragonfruit buttercream filling!

Monday, November 12, 2012

Two Hour Tonkotsu Ramen! (part 2) / emulsified tonkotsu broth

UPDATE (7/28/2015): I have finally gotten around to repeating this recipe (new article coming soon), and can confirm that an emulsifying agent is necessary, and that egg yolks can work (ensure that the broth is at least 160 F to ensure no risk of bacteria) - just temper the egg yolks first with hot broth, then add to broth and use an immersion blender.

Continuing from last time, more on how to make a rich, creamy tonkotsu ramen broth, and the finished assembled product.



Tonkotsu Ramen Broth
Tonkotsu ramen broth can be summed up as a richly pork-flavored broth, with high gelatin content (leaves a distinctly sticky sensation on your lips), emulsified pork fat, and smoky aromatics.  Traditionally, this broth is made by boiling pork feet and chicken bones for anywhere from 6 to 10 hours, in order to break down all the collagen in bone, marrow, and other connective tissue into gelatin (gelatin is denatured collagen, or collagen protein molecules that have become unfolded).  However, as with other cuts of meat high in connective tissue, this process can be expedited using a pressure cooker.

Pressure cookers accomplish this feat by allowing you to braise meats at higher temperatures than what is otherwise possible.  Water boils when the pressure within the water to enter the vapor phase exceeds the pressure exerted on it by the air.  Under normal atmospheric conditions, water boils when it is heated to a temperature of 100o C.  Any additional heat added to the water serves to push water into the vapor phase, rather than increasing the temperature.   However, a pressure cooker allows you to increase the pressure exerted on the water, meaning the contained water can be heated beyond 100o C.  Since temperature is a measure of available kinetic energy, this means that on a molecular level, there is more energy available to facilitate the breaking down of bonds within collagen molecules.  What this means for a piece of meat in a pressure cooker is that it will basically see more heat energy than it would under regular boiling conditions, speeding up cooking times.

While using a pressure cooker does let you break down collagen much faster and achieve a high-gelatin broth in about 2 hours, ­­­­J. Kenji Lopez-Alt noted that what comes out of your pressure cooker is by no means a tonkotsu ramen broth – there is strong pork flavor, but it also tastes oily, as there is a layer of separated pork fat on top.  The problem is that within a pressure cooker, the water is for the most part, not boiling.  If you were to take a cross section of a pressure cooker in action, you would see that the water is totally still.  However, using the traditional cooking techniques, the water is constantly bubbling.  Over the course of several hours, the constant agitation helps to emulsify the pork fat within the watery broth.  However, this is only one way to create an emulsion.

Left: Broth strained out of the pressure cooker after 2 hours cooking.  There is a unappetizing layer of fat on the top.  Center: The equipment necessary, an immersion blender and soy lecithin powder.  Right: The fat has been emulsified into the broth.  Some light foam remains at the top, which subsides after a minute or so.
As you may know, fats and water are basically immiscible, or they do not mix well.  You can see this in some salad dressings, where the oil will separate out into a layer on top of water-based components.  To remedy this, people will either shake up the dressing before using, or they will introduce an emulsifier, like egg yolk.  Emulsifiers are molecules that can form favorable interactions with both water and fat-based molecules.  These interactions support the formation of a stable, homogeneous mix of fats and water.   By introducing a small amount of soy lecithin powder (~1.5 tsp for ~5 quarts broth), and mixing using a immersion blender, I was able to achieve a lighter broth that did not have a separated oil layer and had a taste much closer to that of a true tonkotsu ramen broth.  Fyi, lecithin is one of the main emulsifying agents in egg yolks, so I imagine this could be done with regular egg yolks. However, you would need to do this at a low enough temperature that the egg yolk wouldn’t coagulate due to heat from the broth (thanks to my research into egg boiling, I can say that the broth would need to be <148o F).


For the condensed recipe (in part borrowed from J. Kenji Lopez-Alt, with modifications), check out my recipe pages for tonkotsu ramen brothmarinated soft boiled egg, and braised pork belly.

Pretty much every component can be made in advance and reheated as necessary.  The broth will solidify in the refrigerator (since it has high gelatin content), but can be microwaved or reheated on the stove.  

Two Hour Tonkotsu Ramen! (part 1) / eggs, noodles, pork belly


Sup suuupp so this article is written for all the ramen aficionados reading.  Anyone who’s been to enough good ramen places knows that tonkotsu broth is where all the flavor’s at.  My favorite ramen places, Ippudo NYC and Sapporo in Boston, have one thing in common: a ballin tonkotsu ramen.  The components are few: broth, noodles, pork belly, soft boiled egg, and misc other toppings (green onions, seaweed, maybe some pickled vegetables or buttery corn). When done right, you achieve a soup that is comforting and umami to the extreme.

Now, there are a few resources out there that describe fairly well how to go about making tonkotsu ramen at home (namely http://www.seriouseats.com/2012/02/how-to-make-tonkotsu-ramen-broth-at-home-recipe.html , which I borrowed from in this article).  However, most of what I’ve seen stick to the traditional & laborious 10-hour process.  With a bit of scientific understanding, I’ve crafted a few shortcuts that cut down the whole process to about 2 hours, take a look!

A sneak preview! Tonkotsu ramen broth with braised pork belly, marinated soft boiled egg, seaweed, crispy shallots and green onion
Soft boiled egg
The egg is possibly simplest component of the ramen that actually needs cooking.  Yet somehow, there’s all sorts of confusion about how to actually cook a soft-boiled egg.  I’ve heard techniques that involve putting eggs into water either before or after bringing water to a boil or simmer, then either killing the heat or maintaining a simmer or boil, lid on or off for 5-7 minutes.  
The problem with the techniques I’ve seen is that they are too vague or suffer from too much variability from kitchen to kitchen.  Since water simmers in a range of about 15 degrees, there’s not any good way of knowing what someone means by a “light simmer” unless you’re using a thermometer.  Removing pots from heat can result in totally different water temperatures depending on volume used and pot material.

My preferred technique: boil water (about 1 qt water: 2eggs, erring on the side of excess water), add ‘Large’ eggs straight from refrigerator, allow to boil for 6 minutes uncovered, remove eggs using tongs/ladle/etc and run under cold water for 60 seconds.  This technique is more reproducible between kitchens since most refrigerators are set to 35-38oF, so the eggs will always have roughly the same starting temperature.  ‘Large’ eggs are roughly the same size and mass due to careful legal regulation. Water always boils at one temperature at a given elevation (212o F for most of you).   Finally, using flowing water cools eggs more efficiently than eggs in an ice bath (convective and conductive transfer > conductive transfer only), keeping cooking times more consistent and precise.


Left: Large eggs boiled for 6 minutes, 5 minutes 45 seconds, and 5 minutes 30 seconds from left to right.  The 6 minute egg was significantly easier to de-shell. Right: The same eggs, halved.  The six minute egg had a more firm white, while still liquid yolk. 
Side note: I usually put a lot of trust in Serious Eats food lab articles. However, Serious Eats food lab writer J. Kenji Lopez-Alt recommends simmer eggs at 180 degrees for 6 minutes exactly.  Strangely though, in a picture in the same article, you can see that eggs BOILED for 5 minutes are underdone, and 7 minutes have set yolks.  Clearly, 6 minutes at 180 degrees is not enough.  Indeed, I tested this at home and ended up with raw egg all over my shirt after cracking the shell.  A nice model developed by Dr. Charles Williams (http://newton.ex.ac.uk/teaching/CDHW/egg/#accuracy) estimates that an average ‘Large’ egg from the refrigerator will be done at 5 minutes in boiling water (~57 gram egg with initial temperature of 4o C will reach a yolk temperature of 65o C, the temperature at which yolk will JUST begin to coagulate).  

A handy flash animation, courtesy of the University of Oslo, http://www.mn.uio.no/kjemi/tjenester/kunnskap/egg/..
A flash animation to help with the calculations on a variation of this formula can be found here: http://www.mn.uio.no/kjemi/tjenester/kunnskap/egg/.  However, one must keep in mind that the whites will be technically done at this point, but that only means they are opaque and slightly gelatinous, rather than firmly set.  In addition, the model makes the assumption that starting at t=0, the boundary condition of the egg is the same as the temperature of the boiling water (essentially neglecting any thermal properties of the shell).  To illustrate this inaccuracy, using the formula to calculate Tyolk at t=0 gives you a temperature of 27o C, while the overall initial egg temperature is supposed to be 4o C.   For the sake of getting slightly more firm whites (easier to peel, and just more appetizing to me), and to account for the need for the egg shell to heat up, I’d tack on at least an extra 30 seconds to the 5 minute estimate.  I would give a more precise correction, but couldn’t find any data on the thermal properties of egg shells.

Braised pork belly
Pork belly is notoriously difficult to cook well, in part due to its non-homogeneous composition of fat and meat and high connective tissue content.   As with any cut of meat with lots of connective tissue, the key is always “low and slow” cooking.  This would be another place where a pressure cooker would come in handy, but I’m assuming that most people out there only have one pressure cooker if any, and we’re going to need that for the broth.  Personally, I’ve found that my pork belly has had satisfactory texture with under 2 hours of simmering on the stove (1 hour, 45 minutes of simmering at 190o F in the braising liquid resulted in pork belly that was gelatinous at the edges with moist/tender meat in the center).  The recipe on Serious Eats recommends 3-4 hours, likely because the pork belly slabs pictured were roughly double the width of mine.  The energy required to heat a piece of meat to “done-ness” is proportional to volume, but heat transfer is proportional to surface area.  Therefore, a cylinder of pork belly that has twice the width of mine, doubling the volume, only has 1.5x the surface area.  By taking a ratio of the surface area-to-volume ratios, I estimate that the larger cylinder would take 4/3 as much time, or that my cylinders would be done in ~2.25 hours. Close enough.



What’s most important is to constantly monitor the internal temperature of the meat-side of the pork belly slab.  Thinner slabs, like mine, will cook faster, and once it hits ~155o F, it’s time to remove it from the pot.  The only disadvantages of thinner slabs is that they are a bit trickier to slice, and the ends may be slightly drier than the interior (however, these end pieces also tend to soak up more of the marinade, so no big deal).  I recommend letting your cooked pork belly rounds sit in the freezer for 15 minutes or more, and for very consistent slicing, use a mandolin.


Left: Approximately 1" pork belly slabs tied into cylinders.  Center: Uncooked pork belly rounds in marinade.  Right: Pork belly rounds after 1 hour of braising.

Ramen noodles
Ramen noodles are traditionally hand-pulled from dough that has a gluten structure that permits stretching without breaking.  I’ve given this a try for several days, and decided it’s ultimately not worth it.  Half the battle is finding the right dough formulation (lye, water, and protein content) that results in the right amount of gluten formation, and the other half is a long process of kneading, stretching and pulling.  While I’ve thought of shortcuts for the kneading and pulling process, I have no real way of knowing if I’m actually working with a proper dough formulation to begin with.  There doesn’t seem to be much in the way of improved flavor or texture through making them at home, certainly not enough to justify the days of practice it would take to master this technique.

Be sure to check out part 2 for the 2-hour tonkotsu broth!


For the condensed recipe (in part borrowed from J. Kenji Lopez-Alt, with modifications), check out my recipe pages for tonkotsu ramen brothmarinated soft boiled egg, and braised pork belly.

Pretty much every component can be made in advance and reheated as necessary.  The broth will solidify in the refrigerator (since it has high gelatin content), but can be microwaved or reheated on the stove.  

Tuesday, October 23, 2012

Infinite Bacon! / constructing the Mobius Bacon strip


To christen the birth of my new blog, it was only appropriate to construct the hallmark Mobius Bacon strip.  A Mobius strip is a surface with only one side, that is, if you were to trace your finger along one side of the loop, you would end up on the “other side” as you returned to the starting point.   Continuing to trace the loop would lead you back to the original starting point.   I’m sure this has some kind of actual food-based application somewhere.  Maybe if you fried bacon in space, the grease would never drip off and instead flow along the bacon forever, or something. Cool.


Transglutaminase-catalyzed reaction, drawn with the help of this site
There were a couple possible ways to build this, but I thought it would be a good time to use a new “toy” I’ve been meaning to try: transglutaminase (aka meat glue).  TG is an enzyme that catalyzes a bond formation between the amino acids lysine and glutamine (see figure above, c/o my favorite biochem textbook).  These amino acids are commonly present in meats (meat consists of muscle, which consists of protein, which consists of a wide range of amino acids).

Left: Two thick strips of bacon. Right: two strips of bacon glued together using transglutaminase
I assembled the Mobius bacon strip by gluing two strips together to form a loop with a half-turn in one strip.  About ¼ tsp of TG was dusted onto a square inch of bacon at each overlapping region and allowed to sit in the refrigerator for 2 hours (could be about 1 hour at room temperature, but this may be less sanitary).  The next problem was finding a way to stably cook the bacon while retaining a desirable 3D shape.  As with many bacon constructions (baconstructions?) such as the bacon weave, the bacon basket, the bacon vest (may have made up the last one), I opted to use the oven.

Above: Mobius bacon strip wrapped around two teacups and supported with chopsticks. A failure.
Bacon cooking approach #1 did not meet with much success.  After 20 minutes at 375oF, I opened the door to discover that my Mobius bacon strip had literally fallen apart at the seams.  The problem was that bacon shrinks as it cooks, so tightly wrapping around teacups applied too much force to the weak TG bonds (bacon is about half fat, which can’t contribute much to the gluing process).  I needed a way to keep the bacon strip around a mold that was somewhat malleable, and a cooking technique that was a bit faster.  20 minutes is a long time to wait for bacon.
It then occurred to me that the bacon’s thin and uniform composition would make it ideal for microwaving, and paper towels would serve a double purpose of cleaning up grease and shrinking as the bacon strips shrank.

Left: Two thick bacon strips glued together using TG, wrapped in towels with balls of paper towel used to hold shape. Right: SUCCESS!
As a bonus, this little project sold me on the idea that microwaving is the best way to cook bacon.  Due to some potential variation between home microwaves, you may need to make a quick bacon standard curve (try microwaving one strip for 30 seconds, another for 45, and another for 60 seconds) to determine what times correspond to what level of crispiness.  If you want to keep your bacon grease, simply microwave your strips in a setup that can collect drippings (in a colander over a bowl, or on a row of chopsticks).  If you want your bacon to cook in (and possibly soak up? I’m not positive on this) its own grease, simply microwave your strips in a microwave-safe bowl.  To save time on any kind of cleanup effort, wrap your strips up in paper towels like I did.  Booom. You’re welcome.

Left: Uncooked thick bacon strip.  Right: from top to bottom, thick bacon strips of increasing crispiness cooked for 30, 45, and 60 seconds in my home microwave.

The Greatest Cupcake Ever Made! / honey-roasting nuts




Take a good look dawgs, below is the greatest cupcake ever made! No joke, there is zero room for improving this baby*.  


The construction:

Devil’s food cake – when it comes to chocolate cake bases for cupcakes, nothing beats a devil's food cake.  Moist, rich, and medium-density.  A great complement to either richer or lighter frostings.

Butterscotch frosting – I’ve been wanting to learn how to make butterscotch for a while, the flavor is similar to some of my favorite things, like chai tea, thai tea…the list goes on-ish.

White chocolate cherry mousse – Adding an element of sweetness and tartness to counter the rich frosting and cake.  Whoever makes a cupcake without a filling is terribly misguided.  Filling fixes the common cupcake quandary of the cake/frosting ratio, such that there’s variety in every bite.  (Note: I’ve recently seen another clever solution offered here: http://www.buzzfeed.com/peggy/22-things-youre-doing-wrong #3).

Spicy honey roasted peanuts – Just like the benefits of adding a filling, “toppings” or garnishes can add more contrast and flavor elements to the cupcake.  One thing I never understand is when a cupcake is topped with the same flavoring used in the frosting. Why?? Blueberries on blueberry frosting, popcorn on popcorn frosting, or crumbs from the cake sprinkled on top all may add texture, but the flavor is redundant.  I decided to use nuts primarily for a) varied texture, b) complementary flavor to the other components, and c) seamless introduction of heat, adding a whole new dimension to the cupcake experience.  Just one of several possible ingredients for these purposes was peanuts.

Now each of these components could probably get its own article, but this week I wanted to investigate what it was about roasting nuts in honey that produced that crunchy, sweet-and-savory business.  Typically when I’ve been in the mood for honey roasted nuts, I’d just toss some nuts (usually walnuts) in a small amount of honey, sprinkle with salt, spread them out on foil, and bake for about 15 minutes at 400oF.  Other recipes I’ve seen online ask for lower temps, bake for roughly the same amount of time, and suggest varied techniques and ingredient proportions.  The results could range anywhere from nuts that were grossly burnt or were sticky rather than crunchy.  Clearly, the problem here is eliminating variability in order to get consistently solid honey-roasted nuts every time.

So what exactly is going on in the honey roasting process?  It turns out honey roasting works just like many other candy-making procedures.  Honey initially contains about 18% water by mass, with the rest being a mixture of sugars.  As the water evaporates in the oven, the sugar molecules can come closer together and form a crystal structure. 

Left: sugar solution with high relative water content (blue/red balls = water molecule, orange hexagons = sugar molecule). Right: sugar solution post-heating, water has evaporated allowing sugar molecules to begin to  bond with each other, or crystallize.
Given that honey basically only has two components (water, miscellaneous sugars), and that an oven is a well-controlled environment for heat exchange, I thought this would be an easy enough process to model. BUT the tricky thing about roasting honey is that as water evaporates, the chemical composition changes.  As the chemical composition changes, the boiling point changes.  Also, as the temperature of the honey rises, the heat flow from the honey diminishes.

A  MATLAB model was fitted to observed sugar stages of honey roasted at 400 degrees in order to estimate oven heat transfer coefficient.  This model was then used to predict behavior of honey roasting at 350 degrees.  If you want to look at the code or want more details on how I made the model, feel free to ask in the comments.  It is by no means a perfect model, btw.
Long story short, I put together a program that modeled the steady loss of water from roasted honey.  I first collected “data points” by roasting tablespoons of honey in the oven at 400o F for 3, 9, 12, and 15 minutes before checking which candy stages the honey “syrup” was at.  I fit my model to this data to find the heat transfer coefficient of my oven.  I then used this parameter to model honey roasting nuts at 350o F at 7, 10, 16 and 19 minutes (I observed that 400o F tends to over-roast the nuts).

7 minutes - the honey behaved like..warm honey (“thread stage”, still ~18% water).

10 minutes – the honey formed a firm ball when dropped into water (“firm ball stage”, 13% water)

16 minutes – the honey was beginning to darken, and when dipped in water could be pulled to form thin flexible threads (“soft crack stage”, 5% water)

19 minutes – the honey is still dark, but when dipped into water and stretched formed brittle pieces (“hard crack stage”, 1% water)

Left: honey at the “firm ball” stage.  Center: honey at the “soft crack” stage – a thin flexible thread can be pulled.  Right: honey at the “hard crack” stage – the honey can be pulled to form thin threads/points that break when bent.
With this model, I was able to derive some key lessons when honey roasting nuts:

  1. Using more honey doesn’t increase cooking time, so long as the exposed surface area scales by about the same factor.  Meaning, the recipe I give below can be scaled however you want.
  2. Decreasing the honey:nut ratio only matters to the extent when virtually all the honey is coating nuts – this results in a significant increase in surface area and decreased cook times.  The honey also acts as a buffer to the heat of the oven, keeping nuts from burning. Stick to the ratio I recommend, or err on the side of extra honey.
  3. Related to number 2, distribution of the nuts matters.  By spreading nuts out, you also increase the chance of the thin layer of honey flowing off the nut.  By keeping the nuts all together, there is better honey coverage for all.



A “loose” recipe:

Devil’s food cake

A solid recipe from the food network kitchens, http://www.foodnetwork.com/recipes/food-network-kitchens/devils-food-cake-recipe/index.html if you’re a die-hard from-scratch kind of person.  Otherwise, just work from a box.

Butterscotch frosting

A thorough recipe found here http://www.food.com/recipe/butterscotch-frosting-87307.  Following the technique closely is VERY important here.
 
White chocolate cherry mousse
  • 12 oz white chocolate
  • 1 can cherries (sweet or tart)
  • 1 cup heavy cream
  • 2 tbsp butter
  1. Heat white chocolate in the microwave using 10 second intervals, stirring in between, until fully melted
  2. While melted chocolate is still hot, stir in butter cut into 1tbsp pieces.
  3. Heat a ~1/4 cup of cherry juice from the can in a saucepan until you see bubbles stacking on top of bubbles.  The juice is now a syrup!
  4. Using a food processor, stand mixer, hand mixers, a whisk, or a fork (not recommended), whip the heavy cream until stiff peaks are visible.
  5. Stir one third of the whipped cream into melted chocolate to lighten.
  6. Stir in one tablespoon of the cherry syrup.
  7. Fold in the remainder of the whipped cream into the melted chocolate (be gentle, you don’t want to force air out of the mousse)
  8. Lightly fold in 1-2 tbsp of cherry syrup to create pink streaks in your mousse
  9. Refrigerate until use in cupcake filling

Spicy Honey Roasted Peanuts
  • 1/3 cup peanuts
  • 2 tbsp honey
  • 1 tsp salt (more to taste)
  • 3 tsp chili powder§
  1. Toss peanuts, and honey in a bowl until peanuts are well coated.
  2. Line a baking tray with foil, and then on top with a sheet of parchment paper (optional, but makes it possible to remove nuts when cooled).
  3. Dispense peanuts and honey into the center of the parchment paper (spread out so you have no nuts stacking on top of each other, but do not spread nuts apart from each other).
  4. Bake at 350o F for 20 minutes, stirring once halfway through with a wooden spoon to re-coat all nuts.
  5. Upon removing nuts, quickly stir the nuts one more time to re-coat
  6. Sprinkle salt, chili powder, and other flavorings while the nuts are still hot and sticky

*ok maybe the shape could be improved and the frosting job could be neater or fancier, but that’s just aesthetics
If you’re one of those people who only like to work from fresh ingredients, have fun pitting, juicing and straining about a half cup of cherries.
other types of nuts can be used, I like walnuts since their crevices tend to trap more honey
§ any kind of chili powder can be used.  If you want nuts with a lot of heat, consider mixing in some chili-infused oil with the honey instead.

Parameters: