Anyone who has watched Rose’s presentation on flour for the NYU Experimental Cuisine Collaborative may have noticed a point during the event where Rose explained to Woody that some kate flour would be coming his way from the UK. Well, it arrived … and from there we started an exciting series of cross-Atlantic flour experiments.

A few weeks ago, Woody made 2 yellow butter cakes – one with bleached cake flour and the second with kate flour. On the same day, I also made 2 yellow butter cakes – one with kate flour from the same batch as the flour I mailed to Woody and the second with a freshly-made batch of kate flour. My cakes were pretty much the same as each other, which showed that kate flour stores … and, more importantly in terms of the testing, that Woody could be fairly sure that the cake he was making was representative of cakes made with kate flour in the UK. From Woody’s cakes therefore, we were able to directly compare the baking performance of side-by-side examples of cake and kate flour.

We saw that cakes made with kate flour are yellower and have a coarser grain than those made with cake flour. Woody also noticed a taste he described as “slightly popcorny” in the kate flour cake, which we ascribed to the flour’s heat-treatment in the microwave. Additionally, the cake-flour cake was higher at both the sides and in the centre than the kate-flour cake. Here’s one of Woody’s photos (with my grateful thanks to him for recording his testing in such detail 🙂 ):

As a result of our cross-Atlantic flour experiments, I also had an opportunity to view and directly compare the batters produced by the two flours. Just as the finished crumb of the cake was coarser when using kate flour, so the batter appeared to be grainier and less gloopy than that of cake flour.

Cake flour batter …

… and kate flour batter …

Hmmm.

This was very interesting.

Both chlorination and heat-treatment of flour cause hydrophobicity of starch granules, which improves their oil binding ability and increases the stability of air bubbles in cake batter. It appears that chlorination induces this hydrophobicity through chemical changes, and heat-treatment through conformational changes to proteins on the surface of the starch granules (see this article by Masaharu Seguchi). However, it is likely that starch may be damaged during the microwave treatment of flour.

I discovered from the Food Industries Manual that cake flours need to be low in starch damage. This is because damaged starch granules have a greater affinity for water and can absorb more than twice their weight. Consequently, these starch granules have a lower capacity to bind water in a cake batter, which results in a lower viscosity.

The viscosity of batter is extremely important in defining the overall volume and texture of a cake. Heat from convection moves the batter and this flow has an effect on how the air bubbles incorporated in the batter disperse during baking. In batters with a high viscosity, the rate of gas diffusion is slowed so that air bubbles are encapsulated and retained. When hydrophobic starch granules are bound to the surface of these air bubbles through either chlorination or heat treatment, the bubbles are stabilized and expand uniformly. Cakes therefore have a high volume and fine, regular crumb grain.

In batters with a low viscosity, the convection flow is increased. Air bubbles rise quickly to the surface where they escape from the batter and are lost. Cakes therefore have a low volume and closed, irregular crumb grain.

Significantly, I also read in the Physico-chemical Aspects of Food Processing that it is the increased viscosity of batter made with treated flour that enables a lower amount of flour to be used in relation to sugar without risk of the cake collapsing. In other words, a higher viscosity is crucial to the success of high-ratio recipes. A reduction in flour is desirable because it decreases the starch content level of the cake and causes the crumb to have a softer texture than the crumb of a cake with a higher starch content.

It follows that any damage to starch during the microwave treatment of flour leads to a lowering in the viscosity of a cake batter made with this flour. Could this explain the differences I observed in the batters of cake and kate flour? Armed with a new set of key words, I headed once again for Google.

It turns out that in 1995, a group of cereal chemists from Kansas showed that the addition of xanthan gum to heat-treated flour increased the viscosity of batter and gave rise to an improvement in the volume and texture of cakes made with this flour. This is supported by findings two years earlier suggesting that the insensitivity of xanthan gum to temperature allowed batters to remain highly viscous for longer during baking. As a result, the batter could expand more before the structure of the cake set.

I also discovered a patent in which advantages were claimed for the addition of a non-flour fibre to heat-treated flour:

The addition of a non-flour fiber to the high ratio baking composition of the present invention is essential for obtaining baked volume, improved organoleptic properties and uniform cell structure.

The inventors used oat fibre. However, it is striking that xanthan gum is also a fibre source.

Xanthan gum is widely used as a substitute for gluten in gluten-free baking. I found a bottle of it in the ‘Free-From’ aisle of my local Tesco supermarket. Apparently, it is effective in relatively small amounts. The Kansas chemists recommended adding 0.12% of the weight of the batter in xanthan gum to the dry ingredients before mixing.

I discovered something further from this most recent research. As we had come to suspect, it seems that the crucial factor in heat treatment is the structural change to the starch granules, rather than the reduction in moisture levels per se. It is therefore both possible and beneficial to re-hydrate the flour after heat treatment:

After the heating is completed, the soft flour may be tempered to replace the moisture in the soft flour. Tempering is a process where the soft flour is spread out in a thin layer in a cabinet with humidity control and the flour absorbs moisture from the humidified air until the flour reaches the equilibrium moisture level.

Okay … I could do that!

This is what happened. I microwaved the flour in the usual way, being careful not to increase the temperature so quickly that the flour browned. In practice, this meant that I microwaved the flour in short bursts of no more than 30 seconds at a time. I then spread it out on a baking tray and popped it on the middle shelf in the (cold) oven. I boiled the kettle and filled a second baking tray on the bottom shelf with boiling water. I shut the oven door and waited while the water cooled. I then emptied the water-tray and refilled it a second time with boiling water. Door shut – flour in steam bath. When the flour had cooled, I sieved it. I used 4 1/4 oz of this flour + 1 oz cornflour + 1/4 tsp xanthan gum (kate x flour!) as a replacement for cake flour to make a 9″ x 1 1/2″ yellow cake.

Here is the batter:

And here is the cake:

Closer views of the crumb of cakes made with cake flour …

… kate flour …

… and kate x flour …

… suggest that we are heading in the right direction. Is xanathan gum the missing x-factor?