Between 55 and 60 million tonnes of starch are produced worldwide each year, of which the U.S. controls about 50%. However 90% of starch produced in the U.S. comes from corn. In Europe wheat, or more precisely wheat white flour, is the favored raw material.
Wheat starch production is unique in respect of the very high value of the major co-product, vital gluten. The economics of wheat starch production are thus sensitive not only to fluctuations in the prices of the raw material and of the starch itself, but also to fluctuations in the price of the co-product.
The importance of gluten price to the survival of the wheat starch industry can be clearly demonstrated. For three years beginning June 1, 1998, world gluten prices were sufficiently low after the U.S. government imposed a quota on imports of foreign wheat gluten. The quota was deemed inconsistent with World Trade Organization rules, and it promoted retaliatory action from the European Union.
In spite of representations from the U.S. gluten producers, the quota has not been continued. Instead, encouragement worth $40 million dollars has been made available for investment in technology aimed at novel value-added starch and gluten products.
Wheat and other closely related species, such as barley and rye, have as the main component of their endosperms, two types of starch granule, a large lens shaped form and a smaller, more nearly spherical form (Figure 1).
Starch consisting of large granules (up to 35µm) is conventionally termed A starch, denoting the prime product, and the smaller granule fraction is called B starch. The distinction between A and B granules occurs at 8-10µm but within each type there is a range of granule sizes.
About one third of starch mass is contributed by the small granules, so the maximum yield of prime product is 67% of starch mass (equivalent to about half the grain weight). The B fraction is less pure, partly because impurities are spread on the granules’ surfaces and the surface to volume ratio increases with a decrease in granule size. The main impurity on the surface of granules is protein. In a good A starch, a protein content as low as 0.2% can be expected.
The feedstock to separation processes used for wheat starch production is white flour rather than whole wheat. Milling allows the starch-bearing endosperm to be concentrated largely free of pericarp/testa and embryo particles, which would otherwise become enmeshed in the gluten and be difficult to remove without denaturing gluten.
A disadvantage of milling is that it introduces a degree of starch damage, a condition that is undesirable as it renders some of the granule components dispersible in water.
As water is used in all commercial starch extraction processes, the solubilized polymers enter the liquid phase and accumulate with soluble non-starch components during processing. In damaged starch the granule form is also changed from rounded and semi-rigid, to flat and flexible, even ‘floppy’ when wet. (Figure 2). In the damaged condition, starch can block screens and generally behave differently from undamaged granules.
The level of starch damage depends on two major factors: the hardness of the wheat endosperm and the conditions used in the milling process. The processor can minimize damage level by choosing soft wheats and by minimizing factors such as pressure and, more importantly, shear forces during the milling.
Starch in hard wheat suffers more damage because stresses imposed by roller milling are transmitted through the continuous matrix of hard endosperm to the embedded granules. In soft wheat, air spaces in the matrix allow a high proportion of the stresses to be dissipated without affecting the starch granules.
While many flours for starch production are produced by conventional roller milling, often with a ‘short flow,’ alternative systems such as hammer milling are also possibilities. One aspect of the suitability of the flour for use as a feedstock to the starch/gluten process concerns the amount of fibrous material present.
It is claimed that debranning applied before or as part of the milling system helps to economically achieve low fiber content. It has been adopted in some milling systems dedicated to production of flour for starch plants. The amount of specks of colored fiber coming from the bran can be detected and controlled by on-line image analysis facilities.
The quality of vital gluten is surprisingly insensitive to varietal differences. The results of gluten addition in baking depend more on the amount added.
The separation process
The high value of gluten (hydrated wheat protein) in baking depends upon its functional properties, whereby it improves loaf volume when included in bread flour. Ensuring that these properties are not destroyed dictates many aspects of the separation process by which wheat starch is extracted.
Ill-conceived drying regimes involving excessive temperature can have extremely deleterious effects on gluten efficacy, just as can other treatments that denature protein. The unique nature of wheat proteins also affects the separation process directly because, when wetted and physically manipulated, it gives rise to the rubbery, extensible network of gluten sheets and threads, which entrap many starch granules.
Like many technological processes, early industrial starch and gluten separation methods developed from a manual technique. Both hand and simple industrial processes exploit the phenomenon that, when wheat flour is hydrated, a dough forms which can be repeatedly or continuously washed with additional water while being manipulated, so that starch is progressively removed as an aqueous suspension, referred to as a starch milk.
Starch settles out of suspension under gravity or in centrifuges or hydro-cyclones, and it is thus easily concentrated. Alternatively it can be removed from suspension by sieving. The cohesive ball of gluten, which is eventually washed clean of most of the starch, requires only to be dried for storage and distribution.
The Martin industrial process to remove the starch follows the above principles closely, adding water at 60% of flour weight to form a dough. To conserve water during further washing, a counter-current principle is adopted whereby early stages are performed with recycled process water, and water of progressively greater purity is used to remove starch from increasingly concentrated gluten.
It is not only starch that is removed by the prolonged washing of gluten. Soluble proteins and salts and non-starch polysaccharides, some of which are soluble and some insoluble, also have to be removed from the starch.
More recently developed processes have been designed to use much less water than the Martin Process. They involve initial addition of more water (about 90% of flour weight) to form a slurry. As with dough formation, energy is required for producing a coherent gluten mass, in this case in the form of curds. It may be achieved with a mixer or by forcing the slurry through a small orifice.
Following this, a series of centrifugal devices are used, in which high gravitational forces separate components.
Decanters are capable of taking dense materials out of suspension and discharging them as a relatively dry
mass, as well as discharging suspended components through other outlets.
Some decanters can be used to separate three fractions, each with a high concentration of the respective types of component. Thus the lightest fraction, with the most water in it, contains soluble components such as non-starch polysaccharides, soluble proteins and salts. The intermediate fraction contains a concentration of gluten, fiber (branny materials) and B starch. The densest and driest fraction comprises mainly A starch.
The separation of the gluten from the sticky non-starch polysaccharides at an early stage in the process is regarded as a major advantage of the three-phase decanter system as, free from the viscous polysaccharides, gluten’s subsequent behaviour during further processing is much improved.
Because the separated fractions are concentrates rather than pure products, further treatment is needed. In the case of A starch and gluten, the objective of further treatment is purification; while in the case of the lightest fraction, the objective is recovery of starch and gluten. The last can be achieved by screening alone because only aggregates of gluten overtail the appropriate screens. Screening is also applied to the medium fraction, as a first stage, to remove B starch as a suspension or ‘milk,’ which passes through the screen while the gluten again overtails.
A gluten washer subsequently removes more starch in aqueous suspension.
Screening also plays a part as the initial stage in A starch purification. Process or fresh water is added, and washing can be performed in hydrocyclones or in centrifugal separators (Figure 3) that have stacks of smooth stainless steel cones rotating at high speed within the separating chamber.
At the top of the cones is a hole through which a starch suspension can be deposited as a thin layer on the surface of all the cones. Centrifugal force propels the dense starch granules towards the periphery, while water with non-starch solutes is discharged separately through another nozzle.
Several processing variables are critical in optimizing the process. They include resting time of the flour after milling, solids to water ratio, loading levels of each type of machine, and even water temperature.
Starch manufacturers and the machinery manufacturers continually strive to make the process more efficient and environmentally friendly.
The concept of a dry process is highly desirable because it would entail no expensive removal of water from the products and, perhaps more importantly, no disposal of liquid effluent.
In a recent European research and development project, A starch with a protein content as low as 2% was produced. The achievement depended on the use of flour from a specially bred, ‘supersoft’ wheat variety; further ‘milling’ using unconventional technology; and separation of starch and protein fractions by air classification.
Starch of this type may be suitable for some non-food applications but dry processing is unlikely to become commercially significant for some time yet.
Another possibility for the future is the introduction of wheat varieties with increased A starch content. If this can be achieved without reduction in total starch content, it will be a major boost for an industry whose path has not always been smooth.