Wheat (Triticum aestivum L.) flour gluten protein and starch functionality during fermented pastry making

Fermented pastry products, such as croissant and Danish pastry, are fat-rich bakery goods produced from laminated dough. Such dough consists of alternating layers of dough and fat. The latter can be butter, shortening or margarine. To produce laminated dough, a fat sheet is folded into a so-called ‘predough’, which is essentially a rich bread dough. Alternately sheeting and folding this dough yields a complex multi-layered system. Subsequent fermentation and baking of the dough, during which steam entrapment in the layered structure is supposedly responsible for dough lift, results in pastry products with a typical flaky texture. An acceptable pastry product has a honeycomb-type crumb structure with large, horizontally elongated gas cells.

The main ingredient for pastry is wheat flour. The functionality of its main constituents, gluten proteins and the glucose polymers of starch, amylose and amylopectin, have been researched in products such as bread, cake, cookies and pasta, but have hardly been examined in the context of fermented pastry making. Nevertheless, an expanded knowledge on the role of these ingredients specifically during the different steps in the production of fermented pastry would allow for industrially feasible solutions to produce pastry products of high and consistent quality. Furthermore, it would allow better control of product properties (a.o. product volume and crumb structure), which would enable (frozen) pastry manufacturers to meet specific customer needs. Against this background, the aim of the present doctoral dissertation was to study the functionality of wheat flour gluten proteins and starch during the entire manufacturing process of fermented pastry, i.e. during dough making, laminating, fermentation, baking and storage.

To do so, gluten protein or starch structure, and thus functionality, were selectively modified. The underlying hypothesis was that study of the impact of these alterations on (intermediate) product properties would provide insight into their functionality.

The role of gluten proteins was examined by including redox agents in the ingredient bill of a croissant-type laboratory scale pastry product. ‘Pastry burst rig’ texture measurements showed that dough strength increases during lamination up to a certain extent but decreases after continued sheeting and folding. Dough most likely develops further during the first lamination steps, but destruction of layer integrity then probably causes a decrease in dough strength. Microscopic images of cryo-sections of pastry dough showed improved layering for samples containing oxidizing agents. A strong gluten network is thus beneficial for better keeping the layered structure intact during sheeting and folding. Addition of oxidizing and reducing agents respectively increased and decreased initial laminated dough strength immediately after the final lamination step. However, all samples showed a similar decrease in dough strength during subsequent resting. Redox agents strongly influenced elastic recoil, i.e. dough contraction after lamination. This impacted the dimensions of the fermented products. Moreover, elastic recoil consistently occurred to a greater extent in the final direction of sheeting, most likely due to alignment and strengthening of the gluten network in this direction. This resulted in oval-shaped products. Based on these results and on Size Exclusion High Performance Liquid Chromatography (SE-HPLC) experiments in which the levels of protein extractable in sodium dodecyl sulfate containing medium (SDS-EP) was determined, a model was proposed for pastry dough elastic recoil behaviour. In this model, disulfide bridge formation determines initial dough strength, but secondary interactions, particularly hydrogen bond (re)formation are, along with an entropic component, the driving forces dictating recoil behaviour.

Addition of reducing and oxidizing agents respectively decreased and increased dough height during fermentation. Gas bubbles were shown to be folded in at the sheeting stage and were largely contained within the gluten network in individual dough layers. However, some gas cells were confined by the margarine layers as well. During baking, steam entrapment was of lesser importance for fermented pastry products, as maximal dough height was already reached before a temperature was reached at which significant steam production would occur. However, it is the most important mechanism for dough lift in non-yeasted pastry making. Control samples lacked the typical honeycomb crumb structure and to some degree showed a bread type crumb structure. This was even more pronounced when including a reducing agent. The use of oxidizing agents yielded larger products, with a desirable crumb structure with large pores. However, when large concentrations of a fast working oxidant were used, products tended to collapse during the later baking stage as dough layers tended to ‘slide’ apart. Most likely, no free thiol groups (SH) of glutenin were available to partake in thiol-disulfide interchange reactions during the late baking phase. Gluten oxidation within each individual dough layer presumably occurred to such extent during the dough stage that an insufficient number of SH groups were available for forming dough layer interconnections. Indeed, SE-HPLC also showed less incorporation of gliadin in the gluten network for these samples.

In conclusion, the key roles for gluten proteins during the production of fermented pastry are as determinants of dough strength and recoil behaviour, dough lift and crumb structure.

The role of starch during fermented pastry making was examined by changing starch structure in two ways. Firstly, by increasing the level of damaged starch (DS) in wheat flour by ball-milling. This increases the flour’s water absorption capacity. Secondly, amylases were included in the recipe. Both approaches were also combined, as DS is susceptible to amylase action already at the dough stage. Increased DS levels increased laminated dough strength presumably by making less water available for the gluten. This effect was partly overcome by the use of amylase. During baking, a lower dough viscosity as a result of enzymatic starch hydrolysis was responsible for increased pastry lift and improved crumb structure. Gelatinization of intact starch indeed limits dough lift and expansion. Differential scanning calorimetry and low resolution 1H nuclear magnetic resonance experiments indicated that increased levels of DS and amylase use impact the amylose network in the final product. However, neither these differences nor those in dough strength reported for the DS enriched samples had a notable impact on the final product. This suggests a significant role for gluten in pastry product structure formation.

In conclusion, starch’s key roles during pastry production are that it is a factor determining water absorption during dough making and that it limits dough lift due to the viscosity increase as a result of its gelatinization.

Overall, this doctoral dissertation has significantly expanded the knowledge on gluten protein and starch functionality during pastry making. The results obtained provide a basis for targeted selection of additives and/or well justified adaptations of processing steps to improve production efficiency and better control product properties.