A Farewell Sous Vide Cheesecake
June 30th, 2020
This past month, a dear co-worker of mine left to start a new job at a big biotech company. They had managed to secure a 2-month transition period to finish up their work, only to have that unfortunately get eaten up by Singapore’s “circuit breaker” (coronavirus partial closure) period. Fortunately, we got to be back in the lab together for 1 week before their last day. To adhere to the required social-distancing and rules of not sharing food, to say goodbye I made them a hermetically sealed version of one of their favourite desserts, cheesecake.
Last time I introduced the concept of gels as a class of material defined by an associative networked microstructure. The gel that I used in my coffee drink was a network where the connections between elements could break apart and reform, but this does not have to be the case. These connections—or crosslinks—can also be “permanent” in the sense that once broken, they do not reform. Eggs are common example of this type of gel; on heating, the proteins in an egg unfold and attach to each other to form a solid network. While different temperatures and cooking timescales give rise to a huge variety of mechanical properties one can obtain from an egg, if we take hard-boiled eggs as an example, once broken apart it cannot be reformed. Because these microstructures cannot reform and flow, they are not considered yield-stress fluids.
Cheesecake is an example of a mixture of milk or cream with egg. Some make a distinction between such mixtures that are cooked with or without continuous stirring into creams versus custards, in which case cheesecake belongs to the latter category. From a rheology or materials design perspective, we would conceptualize this distinction as one in which different processing of identical formulations (ingredients) can give rise to different material microstructures and vastly different mechanical properties.
Making the cheesecake
For this cheesecake, I followed a recipe for sous vide cheesecake in a jar. My wife got me a sous vide machine for my birthday this year, and this is the first time I’ve used it to make a dessert. It was quite simple to make; I blended everything together with an eggbeater since I don’t have a food processor. After straining the mixture, I cast it into a few jars that we had laying around. My wife elected to add a teaspoon of matcha powder to hers which worked pretty well and gave it a characteristic green tinge. Compared to a traditional cheesecake, this was much easier as there was no worrying about cracking due to rapid heating or cooling. Some people might be bummed out at the lack of a graham cracker crust, but my wife and I both think that’s the worst part of a normal cheesecake.
Since cheesecake is prepared as a custard, it falls into a bit of a gray area—rheologically speaking—that I struggled with in my work categorizing yield-stress fluids. Visually, cheesecake seems to be a yield-stress fluid; it certainly yields and seems to re-heal if you push it into itself. The gray area here is that just after baking cheesecake is not technically a yield-stress fluid; however, it can become one after it has initially yielded.
Here I’ll introduce the idea of “microgels”, particles that themselves are composed of a gel network. These micron-sized particles pack together and can fulfil the definition of a yield-stress fluid since they can rearrange and flow at a critical stress. Even though each particle is it’s own network, there is not an attractive network that spans across the entire material, and thus these materials are considered soft “glassy” materials rather than “gels”. Cheesecake initially is not a yield-stress fluid because it has a gel network that once broken cannot reform, in the same way that a boiled egg cannot reform its gel network. However, with enough agitation, the network can be broken apart to such an extent that you end up with gel microparticles that jam together. To demonstrate this, we can compare the behaviour of a fresh spoonful of cheesecake to the same material after I’ve vigorously stirred it. The rough surface texture characteristic of cheesecake completely disappears after stirring and it flows much more smoothly. In terms of the eating experience, the stirred cheesecake is a bit softer and creamier, which I happen to prefer.
Not wanting to open up my co-worker’s cheesecake to stir it, I ended up making her a separate batch and my wife nicely decorated the jar. Fortunately, my co-worker also hates the graham cracker crust and gave their review: “it’s too awesome”. Unfortunately, for some reason they didn’t think to vigorously stir the whole thing before eating it. Oh well, maybe next time they’ll give it a try so as to be able to eat a non-ambiguous yield-stress fluid.