I’ve posted on the intricacies of batters before. The suggestions there work, and provide the light coating-style batters desired for fish for example. I have tried many variations in the meantime, working on the scientific premise: never leave well enough alone. However, none have improved on the original so far.
Tempura is another matter. The aim for tempura is almost the exact opposite for coating batters and creates some interesting challenges.
Tempura is synonymous with Japan, however its origins are European. The Spanish and Portuguese were the first Europeans to land in Japan – they already dominated the Asian spice trade and were expanding their influence in the region.
By the late 16th century they had established Jesuit missions in southern Japan that introduced European ideas including battered food deep-fried in oil. The Japanese proceeded to modify the technique, which evolved (by the 18th century) into what we now know as tempura. The Japanese didn’t just adopt deep-frying; they acquired an understanding of the process and integrated it into the philosophy of their cuisine.
Coating batters based on wheat flour hold onto the food largely because of gluten. The trick is to control the amount of gluten. Too much makes a tough and greasy coating (gluten bonds both to water and oil), too little and the batter will slide off the food when fried.
Wheat flour is not particularly high in gluten and gluten is not the main wheat protein. There are two main wheat proteins, glutenin and gliadin. In the presence of oxygen and/or water the glutenin proteins start to attach themselves to each other end to end, making long coiled strands that are somewhat elastic. It is these strands that are known as gluten. The gluten strands weakly attach to the smaller gliadin molecules which are like ball bearings and allow gluten to slide against other glutens. It is the manipulation of this process that contributes to the difference between bread, cakes, cookies, pancakes and crepes, for example.
In general; fat, salt and sugar disrupt gluten formation, whereas oxidising agents (bread improvers), kneading/mixing and air encourage it. The amount of water determines how densely-packed the glutens are and how well they can stick to each other. Egg proteins in some batters coagulate during cooking and strengthen the batter, whereas fats and emulsifiers in the yolk weaken gluten formation. Leavening agents (yeast, calcium bicarbonate) add gas and lighten the batter or dough. The gas doesn’t form bubbles itself though; it diffuses into bubbles that already exist and expands them. This is why air is often incorporated into doughs and batters, for example during kneading or by creaming butter and sugar. It is the same process that underlies foaming milk for a cappuccino.
Even coating batters benefit from gluten reduction, which is why my recommended batter adds some potato or corn starch. However tempura tries to take this further. Minimal mixing and very cold water allow a little gluten formation but keep it to a minimum. An egg adds body. The traditional mixing method is to hold ~4 chopsticks and use them to stab at the batter, rather than by stirring. The batter is used immediately, as gluten continues to form and strengthen with time (even without beating or kneading).
Proteins, while important for gluten and therefore many batters, are a minor component of flours – at least 70-80% of flour consists of starch granules. Again there are two main forms – amylose and amylopectin. I have elaborated on these in a previous post (on potatoes). Mostly, amylose is a gelling agent while amylopectin is a thickening agent. During cooking, these starches absorb water, swell in size, firm up and set the batter or dough.
Which brings me to a rice strain known as glutinous rice. The term glutinous is highly misleading, as glutinous rice does not contain the proteins that make gluten (like all rice, it contains neither glutenin nor gliadin, but does contain other proteins). The name is meant to refer to glueyness. It was selectively bred from a single mutation originating somewhere in South East Asia more than 2,000 years ago (as confirmed by recent genetic studies). Its main characteristic is that it contains negligible amylose. This means that the principal component of glutinous rice is amylopectin – and that makes it sticky and gluey.
Ferran Adria realised that the stickiness of glutinous rice could replace the gluiness of gluten-based batters and provide a new tempura. The absence of gluten means less greasiness. All that needs to be done is to mix glutinous rice flour and water and the tempura batter is made. There is no concern about continued development of gluten and it can be kept refrigerated until needed.
His actual recipe (for tempura avocado) adds wasabi flavour: 30g glutinous rice flour; 10g powdered wasabi; 65g water; mix; refrigerate; proceed as normal.
Pictured are three small pieces of avocado given this treatment, each dusted with lime zest and garnished (left to right) with coriander, mint and a sweetened white rum and lime gel. The rum/lime gel won.
In case you might be wondering how sticky glutinous rice can be, it seems that ancient Chinese stonemasons added it to their mortar. Archeological investigation has confirmed that glutinous rice porridge is part of the mortar in the ancient city walls of Xi’an (1368-1644), the provincial capital of Shaanxi Province. It is thought that it was also added to the mortar used in the Great Wall. Its effectiveness may have contributed to the survival of many ancient Chinese structures.