Save-My-Bananas Loaf, and Sugar Chemistry

Everybody needs a go-to banana cake recipe. As soon as black-spotted squidgyness has compromised its eating appeal, a use must be found for the over-ripe banana which doesn't involve slicing it onto breakfast cereal. It's rare for me to ever throw away food, and the poor old banana has had quite a journey before it reaches my kitchen! Happily, it can languish in the fruit bowl getting blacker and blacker until I get around to making this cake. If fact, the blacker the better. If little bits of mould have started to appear, your banana has reached the pinnacle of cake-ready ripeness.

This is not one of those banana cakes of the dense, fruity kind. It has more of an open texture, albeit with a sturdy, satisfying crumb.  It's the kind of the banana cake which it was de-rigueur to eat with one's coffee when I lived in Sydney (maybe it still is?); it keeps well but after a few days it is a treat to heat a slice in the toaster and spread it with salty butter or even jam.

I have adapted it over time from a Hummingbird Bakery recipe. The original was a bit too sweet, and I always reduced the sugar (as well as substituting in a rich dark muscavado). Then recently I had something of a revelation. I was learning to make caramel. This turned out to be harder than I had imagined; my first attempts produced watery, burned disasters. On the excellent blog of David Lebovitz, I learned an interesting fact. I will quote it here:

"Because sugar is partially water, heat liquefies it. ... Sugar makes things moist. Remember that next time you're thinking about reducing sugar in a recipe."

Not all of this is true. Sugar is not partially water, though it does contain plenty of H and O. So why does it make cakes moist?

Aside: sugar chemistry

Is it to do with the heating process? There has been an interesting debate in the literature recently over the “melting” of sugar, that tricky caramel-making process. Normal kitchen sugar is mainly sucrose - a largish organic molecule. The sucrose molecule actually comprises two smaller, simpler sugar molecules: fructose and glucose. In 2011, researchers at the University of Illinois¹ reported that sucrose doesn't actually “melt” - instead it degrades into other substances (including fructose and glucose, but not water). They dubbed this process 'apparent melting' for obvious reasons. It makes sense when you think about it: if you cool caramel down again it doesn't turn back into white sugar; the actual melting point is dependent on the rate at which you heat it. Other scientists have taken issue with some technicalities, but the basic consensus is this: some actual melting does go on as well as the decomposition, but the decomposition products interfere, making it all a bit flukey. It goes some way to explaining the fickleness of that caramel process.

So heating sugar doesn't release water. But it turns out that sugar is a hygroscopic substance, which means it attracts water molecules. It is a humectant (opposite of a desiccant), which absorbs water molecules (desiccants adsorb them). The absorbed water still contributes to the cake's moisture but is unavailable for the businesses of evaporation and microbial action. My conclusion is that sugar doesn't “make” the cake moist, it just stops it from drying out  – while it’s cooking and afterwards. So in the case of my banana loaf, I have compensated for the reduced sugar by adding more moist, black bananas.

Save-My-Banana Loaf

180g sugar (I use half light and half dark muscavado)
2 eggs
280g very ripe banana (usually 3 bananas)
280g plain flour
1 tsp baking powder
1 tsp bicarbonate of soda
1 tsp ground ginger
1 tsp ground cinnamon
140g unsalted butter

1. Preheat the oven to 170°C. Grease a loaf tin. The cooking time is quite long - if your tin is thin you can line it with baking parchment to protect the cake.
2. Mash the bananas thoroughly with a fork. In a mixing bowl, beat together the sugar and the eggs, then stir in the bananas.
3. Mix in the flour, baking powder, bicarb and spices. Melt the butter (microwave is fine, and you can use the now-empty banana bowl to save washing up), and stir it into the mixture. Mix it until fully combined, no more.
4. Tip the whole lot into your prepared tin, and place in the oven. It will take about an hour; test by inserting a clean knife or skewer, which should emerge still shining.

¹Joo Won Lee, Leonard C. Thomas, Shelly J. Schmidt. Can the Thermodynamic Melting Temperature of Sucrose, Glucose, and Fructose Be Measured Using Rapid-Scanning Differential Scanning Calorimetry (DSC)? Journal of Agricultural and Food Chemistry, 2011; 59 (7): 3306 DOI: 10.1021/jf104852u