There's nothing inherently wrong with using self-rising flour in a chocolate cake, but when I tried one of the few recipes that met my 'self-rising flour only' requirement, this one from King Arthur Flour, I found it to be a bit dense, quite crumbly and a little dry. There was a bit of an incident that could have been the root of the dryness... the oven got switched off (ah!). I was standing right there to turn it back on, but as I found that I had to bake the cake much longer than I should have, it is possible that it lost more heat than I had realized. It was not until later, however, that I noticed a review of the recipe mentioning the difficulty of cutting the cake without ripping it apart, and this leads me to suspect that the crumbling was not the fault of the oven.
I don't know what the source of the crumbling was, but I'm somewhat inclined to think that it might have been related to the use of self-rising flour with no added baking soda or baking powder. Cocoa, it turns out, is acidic. Who knew? (Ok, many people I'm sure. Joy the Baker did. And now I do - and you!) Both baking soda and baking powder, a derivative of baking soda, react chemically when mixed with an acid, and one by-product of this reaction is carbon monoxide that is released when exposed to the heat of the oven; it is this chemical reaction that serves to leaven the cake. Another by-product of the reaction is the neutralization of the acid. This, it seems to me (bearing in mind that I could be considered a chemist by no stretch of the imagination), is why many cake recipes call for both baking soda and baking powder. As Craftsy explains it, this is 'a system of checks and balances. If a recipe contains baking soda to neutralize the acid in a recipe, it might not be enough leavening to give a decent lift to the baked good, but adding more baking soda might be too much in relation to the acid. Here’s where the baking powder comes in, to save the day by giving the recipe a little extra lift while maintaining a pleasant flavor.' What they don't mention, however, is that the high pH of baking soda also enhances the Maillard reaction responsible for browning. Huffington Post writer Amanda Greene explains the chemistry in some more detail and reports that cookies baked with baking powder substituted for baking soda, that is, cookies which did not have the Maillard reaction-enhancing high pH, tasted of their raw ingredients.
Because cocoa is an acid with a low pH, and there was no extra baking soda to neutralize it, it is possible that the Maillard reaction was affected. I am not sure (again, barely an armchair chemist here) how this could have impacted the texture of the cake and caused it to become crumbly, but I certainly am curious! Similarly, though the cake did rise without any additional leavener I wonder if that extra boost would have produced a fluffier result.
It's too bad experimenting with chocolate cake is so hard. Better get back to the mixing bowl.
