Fresh ideas about additives

by Emily Wilson
Share This:

Flour additives are important for millers to understand because bakers often ask millers about ways to improve the shelf life of their products. Millers need to know which additives are most effective and which ones could have negative effects on softness and shelf life. An independent knowledge of shelf-life improving additives allows millers to ask suppliers specifically for certain additives, instead of buying only what is offered.

Millers can also use this knowledge to create value-added products, save money by omitting less effective additives and impart a small amount of independence from the supplier.

Depending on the region and the size of the bakery, some customers may request high-quality, value-added products, while others request cheaper, yet consistent flour. In any case, an understanding of enzymes can help millers create the desired product.


RULES OF FRESHNESS. The freshness of stored, wrapped baked goods is determined by appearance, taste, flavor, crust crispiness and primarily crumb softness. Since freshness correlates well with the compressibility of the crumb, it can be measured by rather simple mechanical devices. The lower the resistance to compression, the higher the freshness will be rated.

Crumb softness depends on the volume of the baked good. The higher the volume, the less material is available to form the cell walls of the pores, hence the resistance to compression decreases.

Furthermore, structural changes taking place during storage result in an increased crumb firmness. The structural changes are caused by retrogradation, or re-crystallization, of the starch, mainly of its amylopectin moiety, that previously has been gelatinized upon baking.

While the smaller amylose molecules migrate from the starch granule into the environment and recrystallize soon after baking, the large amylopectin molecules remain in their amorphous gelatinized form for a longer time. In the amorphous form, amylose and the linear parts of amylopectin form a-helices that are able to bind non-polar molecules. It has been proven with iodine tests that the inside of the helix is non-polar (lipophilic).

Some authors have stated that protein may also play a role in bread staling, but such claims have not been confirmed.

The role of water in bread staling is not yet exactly understood. According to one model, the water interferes with the formation of starch-gluten complexes, which contribute to crumb hardening. In analogy to emulsifiers, water probably attaches with its hydrophilic end to the protein, showing the more lipophilic ends to the outside. This creates a lipophilic surface coat on the protein and hence prevents interaction with the hydrophilic starch molecules.

Although drying out of bread is not the cause for staling, breads containing higher levels of water stale at a slower rate (Bechtel et al., 1953). Moisture migration plays an important role, as wrapped bread without a dry crust keeps fresh longer. Several authors have stated that water may be transported from starch to protein, or vice versa, but there is no conclusive evidence as yet.

Consumers perceive these chemical phenomena as the staling of baked goods when they notice physical changes, such as shrinkage and formation of wrinkles, loss of crustiness, loss of crumb softness, increased opacity of crumb, increased crumbliness, reduced flavor and a change in the product’s feel in the mouth — a firmer, dryer, shorter bite.

Flour additives can reduce the rate of staling by degradation of or interaction with starch fractions that otherwise would recrystallize, or increasing the availability of water.

There are many tools that millers can use to improve crumb softness and shelf life of bread, including enzymes, emulsifiers, oxidizing agents and hydrocolloids.


ENZYMES. a-Amylase. One option to retard the staling of baked goods is to apply enzymes that affect the starch fraction responsible for the bread staling. If the crystalline regions of the amylopectin can be broken down or the amorphous amylopectin can be prevented in any other way from re-crystallization after baking, staling can be decelerated.

The enzyme a-Amylase is able to attack amylose and amylopectin in the middle of the molecule, breaking them down first to smaller fractions and finally to short dextrins and branched limit-dextrin. The current theory of amylase anti-staling effects states that the smaller fraction of amylose and amylopectin have a lower tendency to crystallize and that short dextrins interfere with other regions of the starch molecule in such a way that the crystallization is prevented to a certain extent.

Native starch present in flour only hydrates to a limited extent during dough preparation since most of it is included in compact starch granules. Only the starch in granules damaged by the milling process is accessible by water and hence can swell in the relatively cold environment of a dough system. Since most amylases can only act on hydrated starch, the effect of these enzymes during dough preparation and fermentation is rather limited, and so is any effect on staling.

Fungal a-amylase. Gelatinization of wheat starch occurs between 62° and 75°C (144° to 167°F) (Pomeranz, 1984), provided that sufficient water is available. Gelatinization is characterized by an opening of the dense starch granules followed by water uptake, which makes starch a good substrate for amylases. Unfortunately, standard fungal a-amylases do not survive the increasing temperature for a sufficient long time to be effective in hydrolyzing gelatinized starch. Their stability curve shows a steep decline when temperatures exceed 55°C/131°F (Figure 1), even though starch as the substrate has a stabilizing function on the enzyme.

A new alternative is available in a novel amylase from a non-GMO fungus, a Rhizopus oryzae strain.

In contrast to other fungal amylase, this enzyme (labeled Softase Six in Figure 2) is able to partially hydrolyze even non-gelatinized starch and thus reduce the tendency of re-crystallization. The effect on crumb softness ranges between that of one of the best monoglycerides for this purpose (glycerol monostearate) and a very efficient GMO amylase.

Cereal a-amylase. Cereal grains naturally contain two different amylases, a-amylase and ß-amylase. The a-amylase contributes to anti-staling, because it exerts a higher heat stability than fungal amylases (Figure 3). Hence, it still will be active when at least parts of the starch are already gelatinized. Nevertheless, final baking temperature will be sufficient to completely inactivate the enzyme.

Cereal ß-amylase. According to Würsch and Gumy (1994), ß-amylase has some potency to inhibit amylopectin retrogradation. The enzyme reduces the length of external chains (also called A chains) within the molecules that are responsible for the formation of crystalline structures, together with parts of the B chains.

Bacterial a-amylase. Not only are Bacillus species able to produce bacterial amylases with high heat resistance for the cereal industry, but they are the most commonly used microorganisms for this purpose.

After 20 min at 95°C / 203°F (about the maximum bread core temperature during baking) there will still be some remaining activity (Figure 4). This results in a desired anti-staling effect by degradation of starch molecules. But because the bacterial amylase keeps on breaking down the starch, excessive softening or even liquefaction of the crumb upon storage can occur.

Nonetheless, bacterial a-amylases of similar types are still being used to improve crumb softness, but only combined with fungal a-amylases and at very low concentrations to avoid crumb liquefaction.

Amylases from genetically modified organisms. Most suitable for prolongation of crumb softness are amylases of intermediate heat stability. They can be obtained by genetic modification of conventional microorganisms (e.g. Diderichsen and Christiansen, 1986).

The ability of some of these enzymes to create short-chain maltodextrins seems to be advantageous (Min et al., 1998) and could be due to the same mechanism as for ß-amylases. Kragh and co-authors (1999) claim that also a non-maltogenic exo-amylase is capable to retard retrogradation, which also could be due to the removal of A chains from the B chains.

Hemicellulases. Hemicellulases are able to break down the pentosans in flour. There are two fractions of pentosans, one is water soluble and the other is soluble only in mild alkaline.

Conversion of the insoluble fraction results in soluble pentosan able to form gels with high water-binding capacity. Hydrolysis of the soluble fraction releases water.

While increasing the water binding capacity results in dryer dough, reducing it improves the formation of the gluten network due to the availability of water. Furthermore, pentosans are able to form covalent linkages to the gluten protein, thus increasing the strength of the gluten

With pentosanases (belonging to the family of hemicellulases), the rigidity of this network can be optimized for a maximum volume yield.

Today, hemicellulases are probably the most important baking enzymes to obtain a large volume yield. Originally only an unnoticed side-activity of certain enzyme preparations, they can now be obtained commercially in a purified and standardized form.

Generally speaking, bread with large volume also has a softer crumb structure. Thus, any staling process starts already at a higher softness level. Even if the staling rate (the slope of the softness curve) is not affected, the final softness will be superior to bread with lower volume.

Furthermore, some hemicellulases result in a finer crumb structure due to the optimization of the gluten structure. More pores per volume represent a large surface area. Hence, the cell walls have to be thinner, resulting in a lower resistance to compression — a better softness.

Yin and Walker (1992) reported a significant decrease in the crumb firming rate during storage after the addition of pentosans, obtained from commercial gluten separation. Later, van Eijk and Hille (1995) found in bread a correlation of pentosanase treatment with free pentosan, probably released from the pentosan-gluten network. Therefore, an effect of pentosanases on crumb softness appears likely.

So far, there seem to be no reports on the contribution of the altered water absorption capacity of pentosan to shelf life. Although, at least from the theory, an increased availability of water due to the action of the enzymes should also decrease the staling rate.

Lipases. Lipase converts lipids already present or those added to flour into free fatty acids and di- or monoglycerides. Fatty acids, especially long chain saturated fatty acids, are able to interfere with the starch molecule and to retard staling, as do mono- and diglycerides.

Unfortunately, the use of lipases is not without risk. Some free fatty acids exert a strong flavor, especially butter fat (a constituent of milk powder), which can result in an unpleasant off-flavor of the baked good. As the mill rarely has control of the end-use of the flour, the application of lipase as flour treatment enzyme cannot be recommended.

Lipoxygenases. Improved crumb softness has also been reported with lipoxygenases, produced from soybeans. A complex model from Chung and Pomeranz (1977) suggests the release of free lipids from gluten-lipid complexes, which then could interfere with starch. In another approach, the crumb softening is explained by the oxidation of lipids, as saturated lipids have improved anti-staling properties.

Other enzymes. Many other enzymes have been tested for their effect on bread staling. Phospholipase has been said to reduce the staling tendency by formation of lyso-phospholipids from endogenous or added phospholipids.

Pullulanase, a debranching enzyme belonging to the amylase family, attacks the a-1,6-branching sites in amylopectin (which a- and ß-amylase cannot). This is also said to reduce the re-crystallization potential, while some found no positive effect (e.g. Si, 1995). On the other hand, a branching enzyme — an enzyme capable to build up larger, branched molecules by attaching glucose subunits to an existing fragment — could also be able to alter the structure of starch in such a way that organized re-crystallization becomes impossible.

To my knowledge, none of these enzymes is actually being used in anti-staling enzyme compounds.


EMULSIFIERS. Emulsifiers modify the gelatinization and re-crystallization behavior of starch due to their interaction with amylose and amylopectin. They inhibit the re-crystallization of amylopectin and the formation of an inter-granular starch matrix by amylose.

Their efficiency depends on the type of emulsifier. Molecules containing a long and straight non-polar tail are considered to be most suitable (Stauffer, 2000). For flour treatment, only powdered emulsifiers with good free-flowing properties can be used.

Lecithin. Two fatty acids and one phosphoric acid group are linked to a glycerol backbone. The phosphoric acid group can carry different functional residues, predominantly choline, serine, inositol, ethanolamine or hydrogen.

The fatty acids represent the lipophilic end of the bipolar molecule, with the phosphatidyl group at the hydrophilic end. Although the amylose-complexing capacity of native lecithin is inferior to that of some synthetic emulsifiers, including SSL, CSL and monoglycerides (in their alpha-form), this natural emulsifier is recommended as anti-staling agent because of its positive overall contribution to the baking process.

Lysolecithin. Phospholipase A2 removes a fatty acid from the second carbon atom of the glycerol backbone of phospholipids, while phospolipase A1 takes away the fatty acid from the first carbon. The resulting hydrolysed phospholipids are more polar (improved water solubility) than the original molecule, and also more effective in complexing starch (Figure 5).

For sterical reasons, lysophospholipids with a fatty acid next to the phosphatidyl group are less effective than those with the hydrophobic end farther away from the hydrophilic end, and hence the latter show superior anti-staling properties. The anti-staling potency increases with increasing degree of hydrolysis (Van Nieuwenhuyzen, 1999). Sprayed on a carrier or in a de-oiled form, lysolecithin is also suitable for the milling industry.

Hydroxylated lecithin. Chemical modification by hydroxylation also improves the complexation capability of lecithin. It is not clear whether the effect is due to the hydroxyl group added to the molecule or to partial hydrolysis under the applied alkaline conditions (Van Nieuwenhuyzen, 2001).

Mono- and diglycerides. Fat molecules can be converted into emulsifiers by removal of one or two fatty acids from the glycerol backbone. Distilled monoglycerides are widely used because of their anti-staling effect. Unfortunately, for full function they need to be in their hydrated alpha-form.

Only limited spontaneous formation of the alpha form occurs when monoglyceride powders are present during dough preparation. Hence, pastes containing monoglycerides are more effective than powders. However, combination with co-emulsifiers such as lecithin results in more complete formation of the alpha-form even from the powdered state. Spray-dried compounds of monoglycerides and lecithin have been used in the milling industry since several decades.

SSL and CSL. Sodium-stearoyl-2-lactylate (SSL) and calcium-stearoyl-2-lactylate are among the emulsifiers with the best amylose complexing properties, only outperformed by some distilled, hydrogenated monoglycerides (e.g. Krog, 1971). They are widely used in bread improver compounds, especially for soft buns, since they not only impart a long-lasting crumb softness but also a soft crust. Their use in the milling industry is not very common, partly due to their difficult handling properties (a tendency to form lumps).

Datem. Diacetyl tartaric acid esters from mono- and diglycerides, mainly known for their volume and dough tolerance increasing effect, are also able to complex amylose (Krog, 1971). Since they cause a large increase in volume, it is difficult to say which portion of the improved softness is due to the larger volume and which to the inhibition of retrogradation. For similar reasons as SSL and CSL, Datem is being used as a bread improver rather than as a flour treatment agent.


OXIDIZING AGENTS. Oxidizing agents are hardly mentioned in the concern of bread staling. It is not known whether there has been any investigation of the effect of these substances on staling. Nevertheless, there are some simple considerations that can be made.

Potassium bromate, azodicarbonamide, ascorbic acid and the like do have the potential, provided a proper use, to increase the volume of a baked good. For reasons mentioned earlier, this improves the crumb softness at the beginning of storage. Even with the staling rate unchanged, a softer impression would result at the end of the storage period.

Furthermore, oxidative gelation of pentosans (Geissmann and Neukom, 1973) and oxidative formation of gluten-pentosan complexes (Hoseney and Faubion, 1981) have been shown, both with the potential to increase the uptake of water, a prerequisite for a good shelf-life. Reducing agents as cysteine or sulfite, could in contrast be able to reduce the shelf-life.


HYDROCOLLOIDS. Substances with a large water binding capacity are used in the milling industry to increase the amount of water that can be added to the flour without getting sticky doughs. They are also applied in bread improver compounds for particular applications such as for deep frozen dough to absorb water released from cold damaged protein or starch.

As mentioned before, bread with higher water content keeps its softness longer. But there are limitations. In some countries, the moisture content of bread is restricted and there is an optimum level of water for a long-lasting freshness (Kulp, 1979).

Guar gum, locust bean gum, pre-gelatinized waxy starches (rich in amylopectin), and cellulose derivatives, e.g. carboxymethyl cellulose, have been used for some decades already. Pectin, agar agar and alginates as well as hydrocolloids from microbial sources such as xanthan, dextran, curdlan, or gellan also do have some promising properties but are not being used in flour mostly for cost reasons, as their performance/price ratios are lower than the thickening agents listed above .


Dr. Lutz Popper is head of research and development in the baking laboratory at Mühlenchemie GmbH, Ahrensburg, Germany. Dr. Popper thanks Willem van Nieuwenhuyzen of Eridania Béghin Say, for valuable contributions on emulsifiers.