A Tale of Two Amies

I’d like to introduce you to Amy.

Here she is:

Sometimes she likes to join up with her Amy friends in single lines like this:

And sometimes she likes to join up with them in branched lines like this:

Now before you start calling for the little men in their white coats to take me away, let me explain what I’m going on about. I’m not a chemist or a biotechnologist (or even an artist, as you can tell from my doodles above!), but when I read Rose’s description of her recent tests to improve the performance of unbleached flour without resorting to nuking it in the microwave, I wondered why potato starch was so much more effective than cornstarch at eliminating dipping in cakes baked with unbleached all-purpose flour.

The key seemed to be in starch gelatinization.

Starch gelatinization is important to the structure of all cakes. When starch gelatinizes, its structure breaks down so that the granules dissolve in liquid. This creates a kind of gel that sets as it cooks and provides structural support to the cake.

When you put a cake in the oven, the starch granules begin to swell and absorb the liquid in the batter. However, gluten proteins in the batter are also absorbing this liquid. This means that there is not enough liquid available for the starch granules to fully gelatinize at this stage.

As the gluten proteins absorb liquid, they are able to stretch around the expanding gas bubbles that are sticking to the starch molecules in the creamed fat. The gluten structure stretches and stretches until it pops and becomes semi-rigid. When this happens, the liquid that had been absorbed by the gluten proteins is released into the batter. The starch granules are now able to absorb this liquid until they themselves gelatinize and the batter takes on the shape of the final baked cake.

Starch gelatinization requires more than just water, however. It also requires heat. The starch granules begin swelling at around 50 degrees C (120 degrees F) but gelatinization is usually not complete until the temperature reaches 95 degrees C (200 degrees F), and only then if there is enough time and available water. It follows that a cake’s structure sets earlier when the starch has a lower gelatinization temperature than when the starch has a higher gelatinization temperature.

A scientific paper in the Journal of European Food Research and Technology describes how cakes formulated with cornstarch collapsed because of insufficient starch gelatinization. This supports an earlier observation that the gummy centres and uncooked appearances of cakes prepared with a cornstarch batter were due to incomplete starch gelatinization resulting from a failure to reach setting temperature in the baking time.

Now, here’s something interesting.

Cake flour is more acidic than normal flour. Since acids promote faster setting, this means the starch will gelatinize sooner in the oven, reducing baking time and keeping the cake moister.

As Rose explained, potato starch gelatinizes at a lower temperature than cornstarch. So … I now wondered what exactly it was about potato starch that allowed it to gelatinize at this lower temperature and give the improved results that Rose observed when she used it in her unbleached flour cakes. Perhaps other sources of starch might have this same property …?

This is the part where Amy makes an appearance. Say “Hello,” Amy 🙂

Plants store energy as starch in the form of particles, or granules. These starch granules differ between plants in their size and shape. For example, some starch granules may be large and oval whilst others may be polygonal or elongated. Each starch granule is made up of two different types of molecules (polysaccharides) that consist of repeating units of glucose.

Now, imagine our friend Amy is a glucose unit. If you remember, sometimes she likes to link up with her Amy (glucose) friends in a single line. When she does this, the resulting molecule is called ‘amylose’.

At other times, Amy (glucose) likes to link up in branched lines. When she does this, the resulting molecule is called ‘amylopectin’.

As well as differing between plants in terms of their size and shape, starch granules also differ between plants in terms of the relative amounts of amylose and amylopectin their granules contain:

  • wheat starch, rice starch and cornstarch typically contain 25% amylose;
  • sorghum starch contains 24% amylose;
  • cassava flour contains 20-22% amylose;
  • potato starch contains 20% amylose;
  • arrowroot starch contains 18-20% amylose;
  • tapioca starch contains 15-18% amylose.

You probably remember doing the classic potato and iodine experiment at school -the one where you always end up staining your fingers deep purple. If you didn’t have that pleasure, then just rest assured that iodine is used by chemists as an indicator of starch. This photo shows how iodine stains the starch in potato cells a typically dark blue. This is because amylose has a high iodine-binding capacity.

However, there is a group of plants whose starch does not stain such a dark, deep purple when it comes into contact with iodine. These plants are referred to as ‘waxy’ or low-amylose plants. Iodine merely stains the ‘waxy’ starch of these plants a red or a brown colour. This is because their starch consists almost entirely of amylopectin, which has a low iodine-binding capacity. Glutinous rice is an example of such a waxy plant. The starch of glutinous rice has no or negligible amounts of amylose.

So, what does all this mean for starch gelatinization?

Well, it appears that amylose content is linked to the temperature of gelatinization in different sources of starch. Essentially, the higher the amylose content, the higher the gelatinization temperature. This is because it takes more energy to break the bonds between amylose molecules than it does between amylopectin molecules.

It follows that potato starch gelatinizes at a lower temperature than cornstarch, that tapioca starch gelatinizes at a lower temperature than sorghum starch, and so on. Because the amylose content is so reduced, glutinous rice starch gelatinizes at a much lower temperature than most alternative natural sources of starch.

If a cake’s structure sets earlier when the starch has a lower gelatinization temperature and if this is a key factor in reducing dipping and maintaining moisture in cakes baked with unbleached flour, then it could be predicted that starches with a lower amylose content will be more effective than starches with a higher amylose content as a substitution for part of the flour in a recipe.

I was curious to discover whether these predictions would translate into tangible results when applied to the question of improving the quality of cakes made with unbleached flour. I decided to bake five separate cakes to test the differences between four different sources of starch as a substitution for 15% of the flour in Rose’s Yellow Butter Cake recipe from The Cake Bible:

  1. 100% unbleached McDougall’s ’00’ plain flour
  2. 85% McD’s flour + 15% cornflour (cornstarch)
  3. 85% McD’s flour + 15% potato starch
  4. 85% McD’s flour + 15% arrowroot starch
  5. 85% McD’s flour + 15% glutinous rice starch

I made my own glutinous rice starch by whizzing some Thai Glutinous Rice (found in a local Continental Foodstore) in a food processor and sieving it multiple times until I obtained a fine powder. I did discover an online source of ready-milled glutinous rice flour, but the shipping costs seemed a bit extravagant for the minute quantity I needed!

The glutionous rice starch certainly behaved differently from the other starches. Perhaps unsurprisingly (it is called ‘sticky rice’, after all), it readily absorbed the liquid in the recipe and left me with a dense lump of dough after the first stage of mixing. To produce a cake with this flour, I increased the amount of milk to achieve a more regular batter consistency – completely unscientific, but I wanted to have my cake and eat it too! The extra-milk-glutinous-starch cake was so successful in terms of crumb and volume that I went ahead and baked a sixth cake to see what would happen when I used this thirsty starch in conjunction with heat-treated, kate flour.

Here are the cakes laid out neatly for you on my dissection table …

… and here are their vital statistics:

The starches with a progressively lower amylose content produced cakes with progressively more volume and a correspondingly softer crumb. Cakes made either entirely with unbleached flour or with unbleached flour + cornflour dipped in the centre. Cakes made with starches that had an amylose content of 20% or lower did not dip in the centre.

I didn’t have a long-enough ruler to show you the measurements of all cakes in one go, so I borrowed a ruler from L and shot a couple of close-ups for you. The letters refer to the same flours and starches as before.

Cakes made with potato starch and arrowroot had no discernible (to me and my little testers, at any rate!) off-flavour. The cakes made with glutinous rice tasted distinctly ‘ricey’, but in quite a delicious way. In fact, my fussiest little T couldn’t yum up enough of these cakes whilst turning up his nose at the others. If my food-processor is up to the task (it takes quite a lot of effort to grind down that sticky rice), I can see that I’ll be making more of these cakes in the very near future.

Just for comparison, here are the two Sticky Rice Cakes: the cake on the left was made with untreated flour plus glutinous rice starch; the cake on the right was made with heat-treated flour plus glutinous rice starch.

The cake on the right is how T’s tummy looked after he’d eaten the cake on the left 😉 .

Previous Post
Leave a comment


  1. Susan

     /  March 12, 2010

    Kate, how interesting! … and I thought that was the end of the kate-flour science thingamajig. Whilst the glutinous rice task sounds too time-consuming for me, the option of arrowroot and potato is more appealing. I bet you are mighty pleased with these results and huge congrats are in order. Cheers for your hard labour! Susan

  2. Jeannette

     /  March 12, 2010

    Hi Kate! My, you have been busy ! I wonder if the glutinous rice starch is similar in properties to rice flour? I usually have this in the cupboard for making shortbread biscuits, it would be good to know if it could be used instead of the rice starch. Did you use xanthum gum as well in your tests? Jeannette

    • It’s only the waxy rice that is so low in amylose, Jeannette – regular rice flour has quite a high amylose content, so I wouldn’t expect the results to be the same. But you never know! I didn’t use xanthan gum in any of these cakes because I wanted to be able to see the effects of the starch component alone.

  3. Dad

     /  March 12, 2010

    Save some of those cakes Kate.
    Mum and I are on our way down tomorrow to sample a few morsels.

  4. Ohmygosh, I LOVE baking chemistry. This is so fascinating!! I bet the GF bakers would love to hear this information too.
    There’s a brand of sweet rice flour readily available here in Oregon, and to my understanding sweet rice flour is the same as glutinous flour. Do you think you could find it where you are? It is called Mochiko.
    Thanks so much for doing all this research and sharing it with us!

  5. I am loving this scientific flour research. I am still interested in what a heat treated flour and potato starch cake would look like.
    That is the one I will probably have to research for myself.
    Do you think it would be similar to the heat treated sticky rice cake results?
    This really is interesting to know.

  6. Julie

     /  March 13, 2010

    Kate, these are wonderful! Loved, loved, loved this post, I’m picking up some sticky rice soon. Thanks so much for sharing these test cakes.

  7. Cherry

     /  March 13, 2010

    Thanks for all your work on kateflour, which I now use regularly in all my cake baking (including Victoria sponge) with excellent results. You can get glutinous rice flour very inexpensively at any Chinese/Oriental supermarket. Any city or large town near you should have even a small shop stocking this.

  8. Matthew

     /  March 13, 2010

    Wow Kate, what a great experiment and write up. Have you thought of soaking the sticky rice over night? It will be easier to grind and you might not need the extra liquid for the batter. I know that is how people in south India make rice batters for dosas and idli.

    One other comment, I didn’t realize there was a distinction between cassava and tapioca. When I buy cassava starch it is also labeled tapioca.

    • Soaking the rice sounds like a great idea, Matthew – thanks!
      You’re right about the cassava starch – I should have written ‘cassava flour’. Tapioca is the extracted starch of the cassava root whereas cassava flour includes the whole of the root (inc. fibre etc). Thanks again – I’ll go correct it now 🙂

  9. Great information!

    Thank you for sharing all of this with us. I can’t wait to run some experiments myself!

  10. Vicki

     /  March 16, 2010

    I think you might have a career in text book writing. Had you written my college chemistry book, I may have had a chance of passing the class the first go around!

    • Lol – my old chemistry teacher would have a shock! The most memorable thing I ever achieved in school science was dropping and smashing a large storage bottle of copper II sulphate solution on the chemistry lab floor!

  11. Dad

     /  March 16, 2010

    Thanks Kate – your cakes were lovely with the Yorkshire rhubarb and yoghurt.

  12. This post has been an education. Thank you.

  13. Very very interesting! Had to admit I almost ran away when I saw the term amylose and amylopectin. I love baking but I suck at chemistry. Thank you for doing the testing and for sharing the pictures.

  14. Salma

     /  March 28, 2010

    Hi Kate. I’ve been following your experiments with tremendous interest, and am bowled over by your smarts and persistence.

    Recently, I jumped up for a moment while reading about Rose’s results using potato starch with unbleached flour, only to return to earth in the five seconds it took me to remember that along with bleached flour, potato starch is also unavailable here. Your findings on arrow root had me going up and down for quite a while:) Thank you- I can’t wait to try this!

    One clarification needed though, please: do excuse my denseness, attribute it to my inability to think clearly in my euphoria right now, but does this mean that untreated, unbleached all-purpose flour + arrow root starch may produce a cake that (while not comparable to heat-treated flour), is a decent eat?

    • If you can find some 00 grade flour then yes, I’ve had good results using Rose’s recipes with untreated flour + arrowroot. Bog-standard plain/all-purpose flour gives a much coarser crumb than untreated 00 grade flour, but people have eaten the cakes I’ve made with plain flour without complaining! Personally, if I’m using Rose’s recipes, I prefer heat-treated 00 grade flour with arrowroot and a pinch of xanthan gum added – it’s the closest way I can find of approximating the results that cake flour gives. However, if I’m making a good, old-fashioned Victoria Sponge from a tried-and-tested WI recipe, I use McDougall’s Supreme Sponge flour, which is self-raising (available in Sainsbury’s here in the UK).

      I think it largely comes down to personal taste. I know some people who prefer the coarser crumb of all-purpose flour over the lightness of cake flour.

      I hope that makes sense!

  15. roselevyb

     /  December 25, 2012

    i salute your genius scientific brain and incomparable teaching ability!


Leave a Reply to Matthew Cancel reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: