Acid-Coagulated Tofu
October 31st, 2020
This month I was inspired to try and produce one of the staple foods here in Singapore. My wife often makes soymilk, but despite her uncle running a tofu shop in China, neither she nor I have ever made our own soy curd i.e. tofu. I’m fairly ambivalent towards soymilk but have really enjoyed some tofu dishes in Singapore, so learning how to make tofu was pretty interesting and exciting even though the final product ended up being less than delicious.
What is Tofu?
I will confess that for much of my life, I did not know what tofu was made of other than it was clearly something vegetable related. As I mentioned above, tofu is the curd of cooked soymilk. I briefly discussed “curdling” in the context or ricotta cheese when I prepared lasagna. Like cheeses, there are various ways to curdle soymilk, by which I mean inducing a separation of the components of the fluid. By adding a salt or acid to the soymilk, the repulsive charges on dissolved proteins are either screened or lost. This allows the proteins to coagulate with each other, aggregating to form solid networks that trap water and soybean oil. My understanding is that some of the methods of curdling and coagulation have some correlation with regional variations of tofu; gypsum salts were commonly used in China, the use of Nigari (a magnesium salt material) originates in Japan, American-made tofu reportedly use calcium chloride most commonly, and so on.
From a rheology and material property perspective, what is most interesting to me about tofu is the existence of multiple “grades” denoted apparently by the stiffness. Tofu commonly range from extra-soft to extra-firm, as well as the silken variety. The final properties depend on the type and amount of curdling agent that is used, as well as the strength and duration of pressing that is applied to remove water from the curd.
Making the Tofu
Recipes for tofu are generally simple: blend and sieve softened soybeans to get soymilk which is then coagulated. Due to my difficulty obtaining food-grade varieties of the salt-based curdling agents, I chose to use vinegar and roughly followed this recipe for “firm” tofu . This is a variety of tofu that calls for the curdled mixture to be pressed in a box that allowed for water to drain out.
To soften the soybeans, I soaked them in warm water for about 3 hours during which time they swelled significantly. After transferring the beans to fresh water I blended them for about 5 minutes. This resulted in a dilute slurry of soybean particles that I passed through a fine mesh sieve to get soymilk. Filtering out the soybean particles produced a wet granular yield-stress fluid that could be molded like Playdoh or wet sand, though it tended to crumble easily. As I brought the soymilk to a boil under constant stirring, I had to regularly remove foam from the surface. Soy proteins are known to concentrate at a surface during heating, and like many surface-active proteins they presumably coat the surface of and stabilize air bubbles. Mechanically scooping air bubbles off the surface prevented air from getting trapped inside the final tofu.
After briefly cooling the cooked soymilk, I mixed in the vinegar to form the curds. The process was quite rapid as soft solid curds separated from the liquid but remained suspended and relatively dilute. I poured the curd suspension into a tofu box I had made from some food delivery containers and lined with a cheesecloth. I then pressed the lid down onto the suspension and applied a heavy weight for about 15 minutes. Unsurprisingly, given how dilute the curd suspension seemed to be, a significant amount of water was pressed out and I was left with a fairly thin slab of tofu. The thickness of about 1 cm is much thinner than what you’d find in a restaurant, but was actually reasonably suited to what I wanted to do: mechanical testing.
Given that this was intended to be a “firm” tofu, I was curious what that would mean in terms of the measurable stiffness of the material. In rheology, we will often measure a type of “shear modulus”, though that is a bit tricky for me to do at my kitchen table. Instead, I attempted to measure the elastic modulus in uniaxial compression. Cutting a 20-by-20 mm section of tofu, I placed it on my kitchen balance with a stiff card on top to evenly distribute weight. The elastic modulus can be determined by applying a known stress to a material and measuring the amount of relative deformation. In my experiment, I applied 230 grams to the top of the card, and using image analysis determined that the tofu deformed from 9.6mm to 8.3mm. Here, stress is equal to the mass multiplied by the gravitational constant divided by the cross-sectional area of the tofu. Relative deformation is the change in the height divided by the initial height. Thus, the elastic modulus of this material is approximately 42 kPa, which seems reasonable to me since it is about 40% of the reported elastic modulus of Jello.
Unfortunately, the experience of eating this tofu was less than good, with my wife saying the slab didn’t deserve to be called “tofu”. Despite the stiffness, texture-wise the material was fairly crumbly when chewed; this could potentially be addressed by increasing the pressing pressure and duration. Unfortunately, neither of our palates found the vinegary taste to be pleasant. Some time in the future I’d like to take a look at comparing the effect of different curdling agents on the final stiffness and texture. I’ve not yet given up on making a delicious tofu.