"There's only one hard and fast rule in running: sometimes you have to run one hard and fast."

Tuesday, July 30, 2013

When Steve's Evil Kitchen Freezes Over

I decided to make vanilla ice cream. What I wanted was to make the eggy custard of French-style, but with the smoothness of the best Philadelphia-style. The secret to that smoothness is in having the smallest possible crystals, which is generally done by rapid freezing and stirring, causing nucleation to happen faster than crystal growth. Looking at recipes, I found Mark Bittman had one that fit my needs and had a step that read "heat to 175-180 degrees, do not boil." That's the magic number (range)! Making small sugar crystals to me means making fondant, which when heated to 175-180 degrees, solidifies when cooled; it's the way they make things like Peppermint Patties. That means there's going to be a waiting period, as the fondant ripens, which turns out to be just right for extracting the flavors from vanilla beans. It all works out nicely.

One way that ice cream can be rapidly cooled is to put it into liquid nitrogen; this is how they make Dippin Dots (or, at least did; they went out of business, sold the rights and are back on shelves again). That's a bit extreme - even for me - but I have access to dry ice and I thought of a clever way to utilize it. Just dumping the custard into dry ice won't work well, because you get stacked layers of frozen puddles. Instead, I used a trick that everyone in Minnesota knows from days when the temperature drops to -20F/-30C: if you boil water and then toss it into the cold air, it freezes before it hits the ground; you get a cloud of frozen vapor and some larger frozen droplets (and a few chunks). What's needed is a method of vaporizing the custard before freezing. That's why I put it all into a whipping siphon:

Then the contents are charged with nitrous oxide (carbon dioxide would freeze on contact with the dry ice - which is, after all, frozen carbon dioxide) and discharged into a metal container held in a bath of dry ice and denatured ethanol. You'd expect to get frozen whipped cream, but three things keep it from whipping: the solution is too warm, the fat percentage is too low (17.5%) and the cream is fresh.

Day one: Heat 2 cups half-and-half with opened vanilla bean, then put all in a refrigerator. Make fondant: 2 cups sugar, 3/4 c. water, 1/2 tsp. glycerin, heated to 240 degrees (F), poured into shallow pan and left undisturbed until it is below 46C - sorry, I use two different thermometers with different scales - then stirred until it solidifies; chunks are placed in a plastic bag, worked by hand until smooth and soft, then stored at "basement" temperature (preferably 55-60 degrees).

Day two

whisk 1 cup of fondant (you'll have extra) with 6 egg yolks (room temp.). Beat in 1/2 c. warm half-and-half, then add in 1 1/2 c. half-and-half. Cook on low heat to 175-180 degrees (F) [the fondant should be just melted], strain, pour in 1 c. heavy cream (and 1-2 tsp. vanilla extract, if you want) and pour into canister of whipping siphon. Charge with nitrous oxide and discharge contents onto the inner surface of a metal (preferably copper) container held in dry ice bath [Dry ice bath made by taking the Styrofoam container the dry ice comes in and adding ethanol]. The number of times you'll have to repeat the procedure depends upon the size of the siphon used.

Take metal container out of dry ice bath and (with oven mitts on, to protect from cold), scrape out ice cream. Serve when just about to start melting.


Robyn said...

My sister (engineer) has done it with liquid nitrogen and assures me it works well. The bonus there is that the LN2 evaporates rapidly, "churning" the liquid as it freezes. Might be simpler than your process, even when you figure in the difficulty of obtaining LN2. (However, because the LN2 is in direct contact with the food, you might want a food-grade source, which could be considerably more difficult.)

Chad Walstrom said...

That's some serious cookin!