Fermentation in Bakery Products PDF

18 Fermentation N. Therdthai Department of Product Development, Faculty of Agro-Industry, Kasetsart University, Bangkok, Thailand Introduction 325 Change in starch digestibility and Microorganisms and fermentation...

18 Fermentation N. Therdthai Department of Product Development, Faculty of Agro-Industry, Kasetsart University, Bangkok, Thailand Introduction 325 Change in starch digestibility and Microorganisms and fermentation availability of minerals in bakery products 326 and vitamins 331 Baker’s yeast fermentation 326 Synthesis of antifungal compounds 331 Lactic acid bacteria fermentation 326 Fermented dough making Sourdough 327 processes and monitoring systems 332 Interaction between yeast and lactic Straight dough processes 332 acid bacteria in sourdough 328 ‘No-time’ dough process 332 Mechanisms during fermentation 328 Sponge and dough process 332 Generation of carbon dioxide gas 328 Monitoring systems during Development of exopolysaccharide 329 fermentation 332 Proteolysis 329 Conclusion 333 Synthesis of volatile compounds 330 References 333 Introduction in fermented bakery products depends on the flour variety, ingredients, and fermentation tech- Around 3000 BC, fermented bakery products nology used. In 19th century, however, industrial were first developed by ancient Egyptians. The scale baking was developed, and the most com- major microorganisms involved in fermentation mon leavening agent was changed to baker’s are yeast and lactic acid bacteria. In Europe, lac- yeast – Saccharomyces cerevisiae (Corsetti and tic acid fermented dough (sourdough) has been Settanni 2007). The dough making process may a part of diets for 5000 years. Sourdough has have one or two fermentation stages (also known also been traditionally used as a leavening agent as proofing). To reduce quality variation and cost in various wheat and rye flour based products for in industrial production, however, the straight a long time (Ganzle and others 2008). The variation dough process has always been used. Bakery Products Science and Technology, Second Edition. Edited by W. Zhou, Y. H. Hui, I. De Leyn, M. A. Pagani, C. M. Rosell, J. D. Selman, and N. Therdthai. © 2014 John Wiley & Sons, Ltd. Published 2014 by John Wiley & Sons, Ltd. 326 CH 18 FERMENTATION Currently, there is a major worldwide interest the packaging may be designed to further reduce in healthier products, so there has been renewed the effect of oxygen. This type of yeast is called significant interest in sourdough bakery prod- protected active dry yeast (Kulp 2003). ucts, which have health benefits and unique flavors (Ganzle and others 2008). This has encouraged researchers to try to understand Lactic acid bacteria fermentation more deeply the mechanisms of fermentation Lactic acid bacteria used in traditional sour- and their impact on the quality of these fer- dough are employed variously to develop a mented bakery products. This chapter discusses diversity of flavor and texture in bakery products. the fermentative microorganisms, the fermenta- Generally, these bacteria can be classified into tion mechanisms and their impacts on product three groups including obligate homofermenta- quality, and how to monitor change during tive bacteria (such as Lactobacillus amylovorus, fermentation. La. acidophilus, La. farciminis, La. mindensis, La. crispatus, La. johnsonii, and La. amylolyticus); obligately heterofermentative bacteria (such Microorganisms and as La. acidifarinae, La. brevis, La. fermentum, La. reuteri, La. rossiae, La. frumenti, and La. zymae); fermentation in bakery and facultative heterofermentive bacteria (such products as La. plantarum, La. pentosus, La. casei, La. paralimentarius, and La. alimentarius). Baker’s yeast fermentation Robert and others (2009) reported that the In the late 19th century, commercial yeast for dominant Lactobacillus species in sourdoughs dough making was introduced. Yeast was grown from wheat flour were La. plantarum, La. curva- in aseptic systems to ensure only suitable strains tus, La. paracasei, La. sanfranciscensis, La. pento- could be selected. Today, baker’s yeast is regarded sus, La. paraplantarum, La. sakei, and La. brevis. as Saccharomyces cerevisiae. Media for the inoc- In rye flour sourdoughs, La. reuteri was found to ula should be composed of sugars, minerals, vita- be the dominant and stable microorganism, pos- mins and salts, and the number of yeast cells can sibly due to production of the antibiotic reuteri- be effectively increased with intensive aeration. cyclin. Enterococcus, Lactococcus, Leuconostoc, In addition, the nutrient feed, pH regulation, and Pediococcus, Streptococcus, and Weissella species temperature monitoring should be carefully con- are less frequently used (De Vuyst and others trolled. Bakers’ yeast is produced in both wet and 2009). In Amaranth flour sourdoughs, La. sakei, dry forms. The wet form (also known as cream La. plantarum and Pediococcus pentosaceus were yeast) contains around 82% moisture. Shelf-life reported as the dominant microorganisms (Sterr at 2–4.5 °C is around 3–4 weeks. In contrast, the and others 2009). Competitiveness of microor- dry form can be kept for a longer period at ambi- ganisms in sourdough mainly depends on the ent temperature. A tunnel or conveyor belt dryer type of substrate, fermentation temperature, is used to remove moisture. The dry yeast is sub- acidity, and interaction among the microorgan- sequently rehydrated before the dough making isms (De Vuyst and others 2009). process. The lag phase can be extended until Lactic acid fermentation starts with glucose. the yeast adjusts to the new environment. To By using homofermentative lactic acid bacteria, eliminate this step of rehydration, a fluidized- glucose is converted into lactic acid through bed dryer is recommended to increase particle glycolysis, resulting in two molecules of adeno- porosity, resulting in so-called instant dry yeast. sine triphosphate (ATP). Hexose fermentation Therefore, the instant dry yeast can be mixed through glycolysis can be also conducted by fac- directly with other ingredients. In addition, some ultative heterofermentative lactic acid bacteria, antioxidant agents may be added, and of course because they contain fructose-1,6-diphosphate  MICROORGANISMS AND FERMENTATION IN BAKERY PRODUCTS 327 aldolase, which is a key enzyme for glycolysis. and, therefore, facultative heterofermentative In heterofermentation, not only is lactic acid lactic acid bacteria can ferment pentose in the produced, but also carbon dioxide gas, acetic acid, same way as the obligate heterofermentative lac- and ethanol through the 6-phosphogluconate/ tic acid bacteria. Pentose fermentation results in phosphoketolase (6-PG/PK) pathway. Redox the equal production of lactic acid and acetic acid potential of the fermentation system regulates without formation of carbon dioxide gas (Corsetti the ratio of acetic acid to ethanol, and this path- and Settanni 2007). way yields one molecule of ATP. Nonetheless, two molecules of ATP are possi- bly obtained when fermentation starts at pentose. Sourdough The obligate heterofermentative lactic acid bac- Sourdough is a mixture of flour (usually wheat teria should be used because they contain phos- flour or rye flour) and water, fermented with phoketolase, which is a key enzyme in the 6-PG/ lactic acid bacteria with or without the addition PK pathway. Pentose is phosphorylated and of yeast. Depending on the preparation process converted to ribulose-5-phosphate or xylulose- and the metabolic activity of the main lactic acid 5-phosphate which enters the lower half of 6-PG/ bacteria, sourdough can be classified into three PK pathway. Pentose phosphate is then cleaved types (Decock and Cappelle, 2005) as shown in into two compounds – glyceraldehyde-3- Table 18.1. phosphate (GAP) and acetyl phosphate – to To make bakery products using type III produce lactic acid and ethanol, respectively. sourdough, dried sourdough is mixed with flour Two molecules of ATP can be obtained by phos- and other dry ingredients prior to dough making. phorylation of GAP. In facultative heterofermen- The typically recommended ratio of dried sour- tation, phosphoketolase can react with pentose dough to flour is 1 : 9. Firstly, dough is kept at Table 18.1 Classification of sourdough Sourdough Description Microorganisms Type I sourdough A mixture of flour and water that is incubated at low La. brevis, La. plantarum, temperature (< 30 °C) with a continuous back- La. paralimentarius, La. rossiae slopping by using the mother sponge taken from La. sanfranciscensis the previous fermentation, in order to keep the metabolic activity of microorganisms at high rate. Type II sourdough A mixture of flour and water that is incubated at high Heat and acid tolerant microorganisms temperature (> 30 °C) for long fermentation time such as: La. amylovorus, (up to 5 days) with addition of leavening agent. La. fermentum, La. pontis, La. reuteri. This sourdough can be liquid and is widely used Leavening agent: Saccharomyces as an acidifying agent and an aroma carrier. cerevisiae Type III sourdough A dried sourdough which includes lactic acid bacteria Heat tolerant microorganisms and leavening agent. Microorganisms in an active Leavening agent: Saccharomyces state are first transformed to a latent state by cerevisiae drying and then reactivated later. The number of the revitalized microorganisms is dependent on the drying conditions. A dried sourdough can be used as a starter in premixes. 328 CH 18 FERMENTATION approximately 40 °C for revitalization of the to their maltose-negative and acid-tolerant char- selected microorganisms. During this period, acterization, Kazachstania exigua and Candida other contaminating microorganisms can also be humilis should be fermented together with revitalized, resulting in a reduction of the selected maltose-positive lactic acid bacteria such as microorganisms and spoilage of the dough cul- La. sanfranciscensis. During fermentation, La. ture. After revitalization, the temperature should sanfranciscensis hydrolyses maltose through the be reduced to approximately 30 °C for normal activity of an intracellular maltose phosphory- fermentation of the selected microorganisms. lase without ATP utilization, resulting in unphos- phorylated glucose plus glucose-1-phosphate. The unphosphorylated glucose is then excreted Interaction between yeast and outside the cells and used by the maltose- negative yeasts, that is, Kazachstania exigua and lactic acid bacteria in sourdough Candida humilis. Without the maltose-negative In the fermentation system that includes yeast yeasts, the unphosphorylated glucose may be and lactic acid bacteria, the maximum growth excreted; as a result, glucose repression is induced rate of yeast is reduced due to the presence of the in maltose-positive yeasts (De Vuyst and others lactic acid bacteria. Ethanol production by yeast 2009). Therefore, interaction between yeasts and is also therefore decreased; however, the final lactic acid bacteria significantly affects the stabil- cell counts of yeast are not affected. On the other ity of the sourdough fermentation. hand, yeast fermentation enhances the produc- tion of mannitol and acetic acid by the lactic acid bacteria (Paramithiotis and others 2006). Stability of the lactic acid bacteria population Mechanisms during can be determined by the rate and intensity of fermentation acidification. Normally, stability of the lactic Generation of carbon dioxide gas acid bacteria should be observed every few days (De Vuyst and others 2009). However, acidifica- Carbon dioxide gas is generated during yeast tion may vary depending on environmental fermentation, according to Equation 18.1. The parameters including temperature, time, and generation rate is related to the number of yeast water content. For example, incubation at 25 °C cells and rate of metabolism of yeast. for 4 h slightly decreases pH, compared to incu- bation of the same duration at 37 °C (Gaggiano C 6 H 12 O6 + yeast → 2CH 3CH 2 OH (18.1) and others 2007). To ensure a high degree of + 2CO2 + yeast acidification and flavor development, microor- ganisms should be selected from their transition After carbon dioxide gas is generated, it migrates and stationary phases for sourdough making. towards the initial nuclei of air bubbles (which Saccharomyces cerevisiae is one of the most are formed during mixing), to develop a foam- frequently found yeasts in the sourdough fer- like structure with pores. As a result, dough mentation, and this may be because baker’s volume is expanded. The curve of dough volume yeast is used so commonly in bakeries that it expansion during the period of proofing can be can effectively become a contaminant. In the characterized in three phases including lag, presence of lactose and whey protein in dough, growth, and stationary phases. The lag phase is however, Cabellero and others (1995) reported the time taken for the initial yeast fermentation superior proofing activity of Kluyveromyces and for the carbon dioxide gas produced to marxianus compared to Saccharomyces cerevisiae. diffuse towards the air nuclei. The subsequent Kazachstania exigua and Candida humilis are phase is where the dough expands to reach the other yeasts frequently found in sourdough. Due maximum volume. In the last phase, the gas  MECHANISMS DURING FERMENTATION 329 generation rate is balanced by the rate of gas of monosaccharides such as glucose, galactose, leaving the dough. Therefore, the volume expan- and rhamnose) (Bounaix and others 2010). sion remains constant. The optimum proofing Exopolysaccharides (such as glucans, fructans, time can be defined as the time required for glucooligosaccharides, and fructooligosaccha- dough expansion to reach the maximum possible rides) have been claimed as a gut health pro- volume. Dough growth in terms of the volume moter (Poutanen and others 2009). Lactobacillus expansion ratio (y(t)) can be simulated by the sanfranciscensis could produce fructans that modified Gompertz equation (Equation 18.2) stimulate bifidobacterial growth, thus acting as a (Romano and others 2007): prebiotic. These substances developed during fermentation were recognized as safe, resistant to ⎛ ⎛ μe ⎞⎞ high acidity, and resistant to high temperature y ( t ) = α exp ⎜ − exp ⎜ ⎝α ( tlag − t ) + 1 ⎟ ⎟ ⎠⎠ (18.2) during baking (Corsetti and Settanni 2007). ⎝ With the addition of sucrose (glycosyl donor), where α is the maximum relative volume expan- homopolysaccharides can be synthesized by extra- sion ratio, μ is the maximum specific volume cellular glucans and glycosyltransferase. Hetero- growth rate, ttag is time during lag phase and e is polysaccharides are synthesized intracellularly the Neper number. from sugar nucleotides by glycosyltransferase. In In addition, dough volume expansion is also addition to synthesis of exopolysaccharides, acid accelerated by increasing proofing temperature is formed and may affect both the volume and because a higher temperature speeds up yeast taste of bakery products (Corsetti and Settanni activity and decreases gas solubility (Chiotellis 2007). The development of heteropolysaccharides and Campbell 2003). Relative humidity during reduces resistance to deformation and elasticity proofing is another parameter affecting dough of sorghum sourdough. As a result, specific volume properties and thereby gas retention (Therdthai of bread can be increased, whilst hardness of and others 2007). The number of pores and the bread crumb can be decreased. However, the effect dough volume can be analyzed using the mag- of the heteropolysaccharides on the viscoelastic netic resonance microscopy technique. During properties of wheat sourdough is not significant proofing, the number of pores in dough is dra- due to the strong gluten network in wheat dough matically increased and is described by the (Galle and others 2011). Boltzmann sigmoidal function (Equation 18.3): Proteolysis A YF ( t ) = +B (18.3) During wheat dough fermentation, major pro- 1 + exp(− ( t − t0 ) /Δ teins including glutenins (both high molecular weight (HMW) and low molecular weight where A is amplitude, B is offset, t0 is time lag and (LMW)) and gliadins are degraded through Δ is the transition interval. proteolysis. This involves proteolytic enzymes including the proteinase group and peptidase group. The proteinase can firstly catalyze protein Development of exopolysaccharide degradation to produce a small fraction of pep- During lactic acid fermentation, exopolysaccha- tides; the peptides are then hydrolyzed and bro- ride is possibly developed, resulting in the change ken down by peptidase, resulting in amino acids. of viscoelastic properties of dough and thereby Fermentation of wheat germ by La. plantarum bread itself. Exopolysaccharides include homo- LB1 and La. rossiae LB5 could increase the polysaccharides (containing one type of mono- concentration of free amino acids by 50%, par- saccharide, either glucose or fructose) and ticularly leucine, lysine, phenylalanine, valine, heteropolysaccharides (containing a combination histidine, alanine and methionine. In vitro protein 330 CH 18 FERMENTATION digestibility was also significantly increased fresh taste. Major volatile compounds strongly (p ≤ 0.05) (Rizzello and others 2010). The hydrol- influencing the bakery flavor include organic ysis of peptides is enhanced by acidic conditions, acids, alcohols, esters, aldehydes, and carbonyls. and so the production of lactic acid through lactic They can be synthesized during both fermenta- acid fermentation, which can reduce pH to tion and baking. Baking generates crust flavors around 3.5–4.0, can result in optimum conditions through Maillard reactions and caramelization for the activity of proteinase. However, these depending on the degree of baking and related conditions would not enhance the degradation of process conditions. The resulting flavor com- HMW proteins. In addition to acidification, La. prises roasty, malty, and sweet notes. Poinot and pontis can reduce disulfide bonds in gluten pro- others (2008) reported that a decrease in yeast teins. As a result, the solubility of dough proteins content may affect the quantity of the flavor, can be increased and become more accessible to because yeast contributes to the formation of the activity of enzymes (Ganzle and others 2008). Maillard reaction precursors. The HMW glutenins are responsible for dough Fermentation is mainly responsible for crumb strength and elasticity, while the LMW glutenins flavor depending on the types of microorganisms are responsible for dough viscous properties. The used, the material substrates, and fermentation dough viscous properties are also affected by conditions. Proteolysis during lactic acid fermen- gliadins which are monomeric alcohol soluble tation yields amino acids which are converted to proteins. Therefore, the degradation of glutenins aldehydes and alcohols via the Ehrlich pathway and gliadins causes modification of the viscoelas- (Corsetti and Settanni 2007). Poinot and others tic properties of dough. In bread, the extent of (2008) reported that volatile compounds in proteolysis should be limited, although some fermented bread included ethyl acetate, ethanol, proteolysis is desirable for the production of 2,3-butanedione, 1-propanal, 2-methyl-1-propanol, amino acids, which may be converted by brown- 2-methyl-1-butanol, 3-methyl-1-butanol, 3-hydroxy- ing reactions to flavor molecules during baking. 2-butanone, acetic acid, propionic acid, benzalde- In contrast, extensive protein degradation might hyde, isobutyric acid, butyric acid, isovaleric acid, be required in some products such as soda hexanoic acid, and phenylethyl alcohol. Rizzello crackers. Moreover, proteolysis has been recom- and others (2010) reported alkanes and alkenes mended as an alternative process for the as major volatile compounds arising from the production of bakery products for persons with fermentation of wheat germ by La. plantarum celiac disease (Poutanen and others 2009). LB1 and La. rossiae LB5. Generally, proteolysis is controlled by enzyme Homofermentative lactic acid bacteria tend to reactions and by the solubility of glutens. To produce a high amount of aldehyde groups. increase the amount of proteolysis, a combined Flavor compounds such as ethyl acetate and technique can be used involving the addition of hexyl acetate are generated from the fermenta- protease, chemical acidification to optimize the tion of heterofermentative lactic acid bacteria. pH for enzymatic reaction, and the addition of a As heterofermentative lactic acid bacteria pro- reducing agent to reduce intermolecular and duce both lactic acid and acetic acid, the ratio of intramolecular disulfide bonds, thereby increas- lactic acid to acetic acid is another important fac- ing the solubility of proteins and their accessibil- tor affecting aromatic profiles in heterofermen- ity to enzymes. tation. To obtain enough volatile compounds, lactic acid fermentation requires 12–24 h. However, flavor generation time can be short- Synthesis of volatile compounds ened to a few hours in the presence of yeast The concentration of volatile compounds in fermentation. dough is very low, but it plays an important role When yeast fermentation is used alone, flavor in sensory quality and particularly regarding a generation time is longer, compared to a  MECHANISMS DURING FERMENTATION 331 combination of yeast and lactic acid fermenta- not be sufficient to reduce the amount of phytate tion. The intensity of flavor compounds can be (Corsetti and Settanni 2007). Yeast and lactic acid increased by adding some sugars. Torner and oth- bacteria also contain phytases, which can dephos- ers (1992) observed the high intensity of flavor phorylate phytate and form free inorganic phos- compounds from yeast fermentation in dough phate and inositol phosphate esters which have when either glucose or sucrose was added. less influence on mineral solubility and bioavaila- However, addition of maltose did not show sig- bility than phytate. It was found that the reduction nificant flavor improvement. During storage, the of pH to 5.5 through sourdough fermenta- volatile compounds in yeast leavening products tion, resulted in the amount of phytate being reduce quickly, whereas in general the flavor of decreased by 70% (Poutanen and others 2009). bakery products can be maintained for a longer Cereal products are a good source of vitamins period with the addition of sourdough. Moreover, such as folate, thiamine, and vitamin E; for exam- addition of kefir into the sourdough bread can ple, 45–70% of the folate in bread is from yeast slow the rate of decrease of the concentration of (Patring and others 2009). During yeast fermen- total volatile compounds during a 5-day storage. tation, the folate content in both wheat and rye As a result, freshness of the bread was main- dough can be increased three-fold in the best tained for longer, compared to bread without case. Similarly, thiamine and riboflavin content kefir (Plessas and others 2011). could be increased during a long period of yeast fermentation. The increase in folate, thiamine, and riboflavin in dough is also mainly from yeast Change in starch digestibility fermentation, rather than lactic acid fermenta- and availability of minerals tion. In addition, no synergistic effect from the yeast and lactic acid bacteria on the contents of and vitamins vitamin B group was observed. During sour- Changes in pH can selectively enhance the per- dough fermentation, some losses of vitamin E formance of some enzymes. In turn, the activity were observed (Poutanen and others 2009). of enzymes and microbial metabolites affects the nutritional availability of, for example, starch, vitamins, and minerals (Poutanen and others Synthesis of antifungal 2009). Due to acidification from sourdough fer- mentation, it was observed that pH was decreased compounds and starch digestibility was retarded. Moreover, Bakery products, particularly yeast leavening prod- the level of soluble carbohydrates (glucose, fruc- ucts, are often spoiled by fungi such as Aspergillus, tose, sucrose, maltose, and raffinose) were Cladosporium, Endomyces, Fusarium, Monilia, decreased (Rizzello and others 2010). A decrease Mucor, Penicillium, and Rhizopus. Such spoilage in pH to 4.1–4.5 could also slow down digestion can be prevented naturally by sourdough addition. rate. Acetic acid and propionic acid formed dur- During heterofermentative fermentation, acetic ing heterofermentation could prolong gastric acid and lactic acid are produced and these act emptying rate. Therefore the glycemic index may as antifungal compounds. The fungistatic effect be decreased (Poutanen and others 2009). This is of acetic acid is more significant than that of lactic due to the effect of the lactic acid fermentation, acid. Among the heterofermentative lactic acid rather than the activity of yeast. bacteria, La. sanfranciscensis was claimed to have In the presence of phytate (myo-inositol the most effective antifungal activity due to the hexaphosphate), the bioavailability of minerals production of a mixture of acetic, caproic, formic, may be limited. Therefore, phytase should be propionic, butyric, and n-valeric acids. La. plan- added to increase the bioavailability of minerals. tarum, one of the most widely used strains in sour- Flour contains some phytase, but the level might dough, also provided long-term effective antifungal 332 CH 18 FERMENTATION activity even after baking, due to the production of In the straight dough process, the required level phenyl-lactic acid and 4-hydroxy-phenyl-lactic acid of compressed baker’s yeast should be around (Corsetti and Settanni 2007). 2.5%. However, with the ‘no-time’ dough process, the level of baker’s yeast should be increased to 3.5%. This reduces fermentation time at 30–35 °C Fermented dough making and 85% relative humidity by 2 h. The quality of processes and monitoring the dough developed through using the ‘no-time’ process may be poorer than that developed systems by the straight dough process, due to insufficient In dough fermentation, either a natural fermen- fermentation time. Therefore, some reducing tation or a starter induced fermentation can be agents such as l-cysteine (40 ppm) may be added used. The use of natural fermentation may result to enhance gluten maturity (Kulp 2003). in variation of the quality of the fermented dough due to variable types and amounts of microor- Sponge and dough process ganisms in the system. Fermentation using a starter proceeds with higher numbers of microor- Unlike the straight dough process, ingredients ganisms and less variation of microbial counts; it for the sponge and dough process are divided is easier to control the fermentation and hence into two lots. The first lot, including flour (around the consistency of the fermentation process. 50–80% of total), yeast, and other ingredients are Therefore, industrial bakers prefer fermentations mixed with water, to obtain fairly stiff dough using starters. called ‘sponge’. The sponge is fermented for 3–5 h to develop a web-like structure which enables gas retention. After 60–70% of the fermentation Straight dough processes time, the dough may collapse. Then the second The straight dough process is simple and quick. All lot of ingredients is mixed with the sponge using ingredients are mixed in a single stage and allowed high-speed mixing, and this mix is allowed to to bulk ferment for 3–4 h, therefore quality of the ferment for another 15–30 min to complete the fermented dough may be quite variable. To required volume expansion. The sponge and improve dough quality, knocking-back (punching dough process is widely used in bread making in down the dough) can be integrated into the origi- the United States. In the United Kingdom, fer- nal fermentation process, to partially degas the mentation time for the development of sponge is dough and to redistribute the yeast and substrates very long (about 12 h) because only one-third of after an initial fermentation for 1–2.5 h. Therefore, total flour is used. To speed up this UK approach, gas production rate is kept high and dough volume a greater quantity of yeast is used, and this modi- is improved and controlled. Moreover, salt and fat fication is called ‘flying sponges’ (Kulp 2003). may be added at the knocking-back step, instead of making the additions at the beginning with the Monitoring systems during other ingredients. The delay of salt addition can allow a rapid dough development at the beginning, fermentation while the delay of fat addition can make dough As dough evolution is one of the most important more tolerant to mechanical forces (Kulp 2003). steps affecting the final quality of bakery products, various methods have been developed for moni- toring dough expansion during fermentation ‘No-time’ dough process (Table 18.2). The monitoring system should be The ‘no-time’ dough process involves a very short non-destructive and enable on-line measure- fermentation time which is achieved by increas- ment, in order not to interfere with dough evo- ing the number of fermentative microorganisms. lution. Measurement level may vary from  REFERENCES 333 Table 18.2 Monitoring system for dough expansion during fermentation Measurement Method Reference Dynamic dough density Digital image analysis Soleimani Pour-Damanab and others (2011) Macroscopic structure: porosity and shape Video image analysis Shehzad and others (2010) ratio of dough Microscopic structure: pore distribution Magnetic resonance imaging (MRI) Bajd and Serša (2011) 3D cellular structure and porosity evolution X-ray tomography (XRT) Babin and others (2006) Kinetics of dough fermentation by monitoring Acoustic waves using low frequency Skaf and others (2009) the change in ultrasonic velocity ultrasound macroscopic to microscopic structures. Recently, acids, which promote yeast growth and enable image analysis has been developed to monitor the development of Maillard flavor compounds. macroscopic structures such as dynamic dough Therefore the fermentation process affects the density and shape ratio. Porosity can be moni- structure, texture, flavor, and shelf-life of bakery tored by video image analysis and X-ray tomog- products. To control the consistency of fermenta- raphy. The distribution of pores, which are the tion, starters (such as baker’s yeast, type II sour- microscopic structures, can be monitored by dough in liquid form, and type III sourdough in magnetic resonance imaging. In addition, the dry form) are preferred in industrial scale baking detection of variation in ultrasonic velocity rather than relying on natural fermentation. The can be used to monitor the kinetics of dough types of fermentation process include the straight fermentation. dough process, and the sponge and dough process. To monitor the parameters of dough expansion during fermentation, non-destructive measure- ment using video image analysis, MRI, XRT, and Conclusion acoustic waves has been developed. Fermentation is one of the most important steps in bakery product production after mixing. It is mainly related to yeast (particularly baker’s References yeast, Saccharomyces cerevisiae) and lactic acid Bajd F and Serša I. 2011. Continuous monitoring of bacteria (heterofermentative microorganisms dough fermentation and bread baking by magnetic are preferred). The metabolic activities of fer- resonance microscopy. Magn Reson Imaging mentative microorganisms are responsible for 29:434–42. the quality of dough and hence the quality of the Babin P, Della Valle G, Chiron H, Cloetens P, final bakery products. Yeast fermentation with Hoszowska J, Pernot P, Réguerre AL, Salvo L, sugar as the substrate generates carbon dioxide Dendievel R. 2006. Fast X-ray tomography analysis gas which later migrates toward the centre of of bubble growth and foam setting during bread- bubble nuclei, resulting in an increase in dough making. J Cereal Sci 43:393–7. volume during fermentation. Fermentation with Bounaix MS, Gabriel V, Robert H, Morel S, Remaud- Siméon M, Gabriel B, Fontagné-Faucher C. 2010. heterofermentative lactic acid bacteria yields lac- Characterization of glucan-producing Leuconostoc tic acid (which decreases pH), acetic acid (which strains isolated from sourdough. Int J Food acts as an antifungal agent), exopolysaccharides Microbiol 144:1–9. (which can act as a gut health promoter), and Caballero R, Olguín P, Cruz-Guerrero A, Gallardo F, volatile compounds (which act as a flavoring García-Garibay M, Gómez-Ruiz L. 1995. Evaluation agent). In addition, protein is degraded during of Kluyveromyces marxianus as baker’s yeast. Food fermentation through proteolysis to yield amino Res Int 28:37–41. 334 CH 18 FERMENTATION Chiotellis E, Cambell GM 2003. Proving of bread Poinot P, Arvisenet G, Grua-Priol J, Colas D, Fillonneau C, dough I: Modelling the evolution of the bubble size Bail AL, Prost C. 2008. Influence of formulation and distribution. Food Bioprod Proc 81:194–206. process on the aromatic profile and physical charac- Corsetti A, Settanni L. 2007. Lactobacilli in sourdough teristics of bread. J Cereal Sci 48:686–97. fermentation. Food Res Int 40:539–48. Poutanen K, Flander L, Katina K. 2009. Sourdough Decock P, Cappelle S. 2005. Bread technology and and cereal fermentation in a nutritional perspective. sourdough technology. Trends Food Sci Technol Food Microbiol 26:693–9. 16:113–20. Rizzello CG, Nionelli L, Coda R, De Angelis M, De Vuyst L, Vrancken, G, Ravyts, F, Rimaux T, Weckx S. Gobbetti M. 2010. Effect of sourdough fermentation 2009. Biodiversity, ecological determinants, and on stabilisation, and chemical and nutritional char- metabolic exploitation of sourdough microbiota. acteristics of wheat germ. Food Chem 119:1079–89. Food Microbiol 26:666–75. Robert H, Gabriel V, Fontagné-Faucher C. 2009. Gaggiano M, Cagno RD, De Angelis M, Arnault P, Biodiversity of lactic acid bacteria in French wheat Tossut P, Fox PF, Gobbetti M. 2007. Defined multi- sourdough as determined by molecular characteri- species semi-liquid ready-to-use sourdough starter. zation using species-specific PCR. Int J Food Food Microbiol 24:15–24. Microbiol 135:53–9. Galle S, Schwab C, Arendt EK, Gänzle MG. 2011. Romano A, Toraldo G Cavella S, Masi P. 2007. Structural and rheological characterisation of heter- Description of leavening of bread dough with opolysaccharides produced by lactic acid bacteria mathematical modelling. J Food Eng 83:142–8. in wheat and sorghum sourdough. Food Microbiol Shehzad A, Chiron H, Della Valle G, Kansou K, Ndiaye A, 28: 547–553. Réguerre AL. 2010. Porosity and stability of bread Gänzle MG, Loponen J, Gobbetti M 2008. Proteolysis dough during proofing determined by video image in sourdough fermentations: mechanisms and analysis for different compositions and mixing con- potential for improved bread quality. Trends Food ditions. Food Res Int 43:1999–2005. Sci Technol 19:513–21. Skaf A, Nassar G, Lefebvre F and Nongaillard B 2009. A Kulp K. 2003. Baker’s yeast and sourdough techno- new acoustic technique to monitor bread dough dur- logies in the production of U.S. bread products. In: ing the fermentation phase. J Food Eng 93: 365–378. Kulp K, Lorenz K, editors. Handbook of dough fer- Soleimani Pour-Damanab AR, Jafary A, Rafiee SH. mentations. New York: Marcel Dekker. p 97–143. 2011. Monitoring the dynamic density of dough dur- Paramithiotis S, Gioulatos S, Tsakalidou E, ing fermentation using digital imaging method. Kalantzopoulos G. 2006. Interactions between J Food Eng 107:8–13. Saccharomyces cerevisiae and lactic acid bacteria Sterr Y, Weiss A, Schmidt H. 2009. Evaluation of lactic in sourdough. Proc Biochem 41:2429–33. acid bacteria for sourdough fermentation of ama- Patring J, Wandel M, Jägerstad M, Frølich W. 2009 ranth. Int J Food Microbiol 136:75–82. Folate content of Norwegian and Swedish flours and Therdthai N, Zhou W, Jangchud K. 2007. Modelling of bread analysed by use of liquid chromatography– the effect of relative humidity and temperature on mass spectrometry. J Food Comp Anal 22:649–56. proving rate of rice-flour-based dough. Lebens Wiss Plessas S, Alexopoulos A, Bekatorou A, Mantzourani I, Technol 40:1036–40. Koutinas AA, Bezirtzoglou E. 2011. Examination of Torner MJ, Martínez-Anaya MA, Antuña B, de freshness degradation of sourdough bread made Barber CB. 1992. Headspace flavor compounds with kefir through monitoring the aroma volatile produced by yeasts and lactobacilli during fermen- composition during storage. Food Chem 124: tation of preferments and bread doughs. Int J Food 627–33. Microbiol 15:145–52. 19 Baking Tiphaine Lucas IRSTEA Food Process Engineering Research Unit, Rennes Cedex, France Introduction 336 Structural changes in starch and protein Heat and water transport during baking 336 during baking and subsequent Modes of heat transfer 336 changes in dough/crumb Modes of heat transport into mechanical properties 345 the dough 337 Basic mechanisms of starch Measurement of temperature gelatinization 345 during baking 338 Factors affecting starch gelatinization Spatial distribution of water and extent of starch gelatinization in bakery products 338 in bakery products 345 Measurements of water content Protein coagulation 346 in bakery products 339 Impact on dough viscosity during Cell expansion and squeezing, crumb baking 347 formation, and overall collapse 339 Impact on mechanical properties Growth from gas nuclei 339 of crumb 347 Mechanisms favoring volume Crust formation 348 increase in cell 340 Dehydration and its effect on Mechanisms limiting volume reactions in and mechanical increase in cell 340 properties of the crust 348 Starch gelatinization 340 Role of steam injection on the crust Ruptures in cell walls 341 formation 349 Setting of the crust and its effect Cell expansion in the crust; of cell collapse or coalescence 343 effect on permeability of crust Collapse of products 344 to gases 350 Total water loss 344 Factors affecting the crispiness Measurements of water content of bakery products, and in bakery products 344 bread crust 350 Spatial distribution of water content The end point of baking 351 in different bakery products 344 References 352 Bakery Products Science and Technology, Second Edition. Edited by W. Zhou, Y. H. Hui, I. De Leyn, M. A. Pagani, C. M. Rosell, J. D. Selman, and N. Therdthai. © 2014 John Wiley & Sons, Ltd. Published 2014 by John Wiley & Sons, Ltd. 336 CH 19 BAKING Introduction Oven-rise (the change of dough dimensions during baking, look at a video at http://www. During baking, heat and mass transfer takes weekendbakery.com/posts/baking-bread- the- place in the dough simultaneously and interde- movie/) together with dough/crumb transition pendently, and involves four major changes: and crispiness are the most outstanding macro- scopic features of baking, with relevance to 1. Gases are vaporized as temperature texture. Reactions are another essential feature, increases; the gas cell increases in volume related to taste (production of flavour), digestibil- provided that the cell wall retains gases and ity (starch gelatinization), and safety (production is deformable. of acrylamides). 2. Starch gelatinizes upon the increase of tem- These different changes do not occur sequen- perature to an extent depending on the local tially but do overlap. Roughly in the first stage of availability of water, and proteins coagulate. baking, cell expansion and water loss begin; one These changes limit dough extensibility. It of these can, however, be delayed during this may limit the cell expansion described in (1), stage. During stage 2, both the total height of and hence be a local driving force for pres- the product and the rate of water loss reach sure build-up and possibly fracturing of the their maximum. Color develops at the surface cell wall (3). during the last stage 3, concomitantly to a possi- 3. The initial structure with closed gas cells ble collapse of the porous structure and a separated by dough walls is transformed decrease in the rate of water loss. These changes into a porous structure with interconnected are detailed in the following sections. pores, known as the dough-crumb transition. Dough membranes rupture when they can no longer withstand the gas pressure, and Heat and water transport changes in (2) are a possible cause. In the- ory, ruptured walls limit the cell expansion during baking described in (1); gas molecules are Temperature dominates product quality because exchanged between adjacent open cells, it affects enzymatic reaction, volume expansion, and finally transported out of the dough. starch gelatinization, protein denaturation, non- Pressure forces in such regions may then enzymatic browning reactions, and water migra- decrease, a possible explanation for the col- tion, as will be detailed later. lapse of cells locally. Time-course changes in temperature and its 4. Under the action of high temperatures at distribution through the dough during baking are boundaries, water vaporizes in the oven determined by multiple factors, among them: the atmosphere. Depending on the product size of the product (biscuit versus bread loaf); thickness, but also on baking conditions, this the ratio between top, side and bottom heat flows; supports the formation of a dry, hard crust in the expansion of cells which reduces thermal con- dessert or bread dough, and may also lead to ductivity, but likewise favors the transport of water a complete drying in biscuits and cookies. vapor evaporated at the surface layers to the core, This is accompanied by specific chemical where it condenses and releases energy. reactions (in particular Maillard reactions), responsible for the browning of the crust and the development of flavour. Early setting Modes of heat transfer of the crust restricts total volume and modi- Heat is transferred through the combination of fies the balance in pressure forces between conduction, convection and radiation. Radiation the inner and outer dough and hence cell comes from the heating elements (burner flames) expansion (1) locally. and all hot metal parts (walls) in the oven. It is  HEAT AND WATER TRANSPORT DURING BAKING 337 responsive to changes in the absorptive capacity Under the dough surface, where there is still of the dough and there is a tendency for color liquid water, water evaporates to saturate the changes upon heating to accelerate after the first partial water vapor pressure in the pores. The browning has occurred (Therdthai and Zhou water vapor pressure reaches a maximum at a 2003). Contribution of radiation to heat transfer location from which the vapor can diffuse in at the top surface can be estimated by h-monitor two directions, towards the dough surface and and was found to be much higher in direct ovens, towards the core. The driving force for the trans- 70–80% (Baik and others 1999, 2000a) than in port of water vapor from this location to the sur- indirect ovens, 45–60% (Fahloul and others face has been detailed earlier. On the other hand, 1995). The contributions of the different heat the temperature gradient induces a water vapor sources were calculated through modeling for pressure gradient between this location and the tunnel type multizone industrial ovens; for bis- core, which produces a diffusive flow of vapor in cuit, heat transferred by conduction through the the same direction. Because the water migrates band was minimal (10%), radiant and convective counter to the thermal gradient, it is accompa- heat being of the same order (Fahloul and others nied by condensation, release of its latent heat 1995). Conduction represents more than half of and an increase in liquid water at the core. This is the total heat transferred to the product during why this mechanism is also called transport by tin baking, possibly reaching two-thirds in certain “evaporation-condensation”. This mechanism conditions (Baik and others 2000a). Conduction takes place as long as a temperature gradient mode is also predominant for products deposited remains (and water remains!) and the core tem- on the hearth of the oven, like Indian flat bread perature has not reached 100 °C. Such transport (chapatti) (Gupta 2001) or some French breads is promoted by the opening of the porous struc- (Sommier and others 2005). ture which occurs during baking; nevertheless it also proceeds in a cellular structure (closed cells) Modes of heat transport with four distinct phases (De Vries and others into the dough 1989): As soon as the product is introduced in the oven, 1. evaporation proceeds on the warmer side of heat transfer results in superficial water evapora- the gas cell, tion and heating up of the product. In the oven, 2. water migrates through the gas phase, where the temperature is high, the partial water induced by the vapor concentration gradient vapor pressure is far from saturation. Hence, inside the cell, water vapor in the superficial layers of the prod- 3. water condenses on the colder side of the gas uct diffuses into the oven atmosphere. As the dif- cell, fusive flow of liquid water from the core is less 4. liquid water is transported by diffusion than the flow of evaporated water at the surface, through the dough wall to the warmer side of the product dries at the surface, leading to a the next gas cell, where the series of processes very low water activity, and the water activity starts again. Thorvaldsson and Skjöldebrand decreases rapidly with water contents below (1998) suggested, however, that as the con- 0.15 kg/kg. Temperature near the product surface densed water is likely to be absorbed by does not exhibit a marked plateau at 100 °C (the starch gelatinization, transport of water by temperature of water at the pressure close to that this mechanism to and through the ungelati- of the atmosphere, and water activity close to nized closed region is unlikely to occur. that of pure water). Above 100 °C, the tempera- ture rises in thermodynamic equilibrium The transfer of water vapor with vaporization with the locally decreasing water content and near the surface and condensation at the core rises towards that of the oven temperature. contributes to heat transfer, and was proposed by 338 CH 19 BAKING Sluimer (cited by De Vries and others 1989) to and others (2005) reported higher heating rates explain rapid heat transport during baking. De at the core for small bread loaves, of 15–17 °C/min. Vries and others (1989) recorded the tempera- Higher heating rates can also be generated by ture in the centre of gas-free and fermented steam condensation onto the dough/batter sur- doughs during baking with the same external face subsequently to its injection into the oven. dimensions, and showed that heat transport in Sommier and others (2005) and Chevallier and fermented dough was much faster than in gas- others (2002) reported heating rates of 25 °C/min free dough. If conduction was the only mecha- for bread (baguette) and 37–150 °C/min for short nism, heat transport in a gas-free dough should dough. In a tunnel type multizone industrial oven, have been faster than in fermented dough pieces heating rates in the range of 18–50 °C/min were because of cells acting as isolating elements. Such reported for cakes (Baik and others 2000a). results were explained by the contribution of the evaporation-condensation-diffusion mechanism Spatial distribution of water to heat transport. The location where the partial water vapor in bakery products pressure is at maximum progressively moves Whatever the type of bakery product it is impor- inside the product, making the dry area grow tant to realize that moisture gradients exist in the thicker. Progressively, the superficial dry layer product at the end of the baking process. Such constitutes a major barrier to heat transport distribution in water content through the prod- inside the product, hence temperature and water uct governs textural properties such as the con- content gradients at the core progressively trast between the softness of crumb and the vanish. Heat transport is no longer intensified crispiness of the crust in high-moisture bakery by the diffusion of water vapor to the core. In products. It governs more generally the storage high-moisture products, temperature at the properties of the product such as the loss in crisp- core stabilizes at 100 °C; this plateau lasts as iness in high-moisture products or the checking long as its water content remains similar to that of biscuits (Ahmad and others 2001; Saleem and of the dough. others 2003). Heat transport towards the centre during baking induces water transport and there- Measurement of temperature fore a change in the water distribution (refer to the heat transport subsections). Water also plays during baking an important role in the major structural changes Reproducible measurements of the temperature that take place during bread baking, at the micro- inside dough during baking are difficult owing to scopic (for example, water vaporization in the the oven-rise and to the difficulty of accurately cell expansion) and molecular scale (for exam- locating the thermocouples. Temperature gradi- ple, starch gelatinization). ents are considerably increased in thin dough The initial water content typically ranges from pieces (about 15–50 °C in 2 mm thick biscuits), 11 to 30% for biscuit and crackers (Wade 1988), amplifying the inaccuracy in positioning the ther- 32–38% for cake batter (Baik and others 2000b) mocouple. Thermal conduction along the ther- and 38–48% for bread dough. mocouple wire is an additional source of bias; The final moisture of a freshly baked biscuit is however, this can be overcome by using optic fib- usually within the range of 1–5% depending on ers of less accuracy. Nevertheless, some general the type of product (Wade 1988). Water distribu- trends can be drawn for the core. tion is very difficult to measure in the thickness Heating rate in bread is close to 3–6 °C/min of a biscuit; it can be however assessed by simula- in the absence of steam injection (Therdthai tion or experimentally along the diameter of a and Zhou 2003; Sommier and others 2005; Patel disk-shaped biscuit, with differences up to 1.7% and others 2005; Wagner and others 2007). Patel (Ahmad and others 2001). The final water  CELL EXPANSION AND SQUEEZING, CRUMB FORMATION, AND OVERALL COLLAPSE 339 content in cakes varies between 17 and 24% (Hayakawa and Hwang 1981), 1.4 × 10–3 to (Baik and others 2000b). 9.4 × 10–4 kg kg/kg for yellow and chocolate cakes As bread loaves are larger pieces, reports of of various sizes (Baik and others 2000b), final water content are scarce for the loaf as a 2.5 × 10–4 kg kg/kg for sponge cake (Lostie and whole, with most data concerning the distribu- others 2002) and 2.8 × 10–5 to 2.4 × 10–4 kg kg/kg tion of moisture through the loaf. Total water (Hasatani and others 1992) and 2–5 × 10–4 kg kg/kg loss can be between 10 and 20%; it is enhanced (Sommier and others 2005) for bread. by higher baking temperatures (Hasatani and others 1992). The lowest water content at the end of baking is obviously encountered in the Measurements of water content bread crust, being 2–15% depending on the per- in bakery products meability of crust (Hirte and others 2010) or the In practical terms, although the overall water loss oven temperature (Ahrne and others 2007). is easy to assess, the determination of water con- Bread crumb located beneath the crust exhibits tent profiles in crumb during baking requires the intermediate water contents, from 2– to 12% setting up of a method for accurate cutting of the lower than that of dough (Thorvaldsson and loaf into samples while still hot and deformable, Skjöldebrand 1998) or from 3–9% lower than and the minimization of water losses by evapora- that of dough (Wagner and others 2007). The tion from cut surfaces. Water content is also sen- crumb at the core retained its initial moisture sitive to the time elapsed from the end of baking content or even increased (Wagner and others and the relative humidity in the atmosphere of 2007). When sampling took place just after storage; these important factors are rarely removal of bread loaves from the oven, an reported and make experimental results almost increase in water content by 1 to 3% was reported impossible to compare between studies. at the core (De Vries and others 1989; Zanoni and Peri 1993; Wagner and others 2007). In addi- tion to the conventional, destructive method of sampling combined with oven-drying for water Cell expansion and squeezing, content determination, a continuous method crumb formation, and overall based on near infrared reflectance (NIR) was proposed by Thorvaldsson &and Skjöldebrand collapse (1998). Light was sent into the sample through an During baking, total volume may increase by optic fibre and the intensity of the reflected light 16–100% (volume unit per 100 volume units of was measured. The technique is known to be the initial dough) for bread (Zanoni and Peri sensitive to the structure and temperature of 1993; Sommier and others 2005; Wagner and the sample. If dependence on temperature was others 2008), 50–80% for sponge cake (Lostie incorporated into the calibration, structural and others 2002) and 100–300% for biscuit dependence made quantitative measurements (Wade 1988; Chevallier and others 2002). in the sample difficult during the oven-rise. Thorvaldsson &and Skjöldebrand (1998) showed an increase in water content in the core immedi- Growth from gas nuclei ately after the local temperature had reached The presence of gas nuclei is a prerequisite for 70 ± 5 °C °C (i.e. after cell expansion was more or inflation (Baker and Mize 1937). The aeration less complete), which reached +8% at the end of of dough or batter is achieved during mixing baking (35 min at 225 °C °C) relative to that of or whipping for sponge cake. For bread, dough dough initially. porosity after mixing is in the range 3–15% Magnitudes of the overall drying rate were depending on the mixer type and mixer speed 1.8 × 10–3 kg kg water/kg dry matter for cookies (Cauvain and others 1999). The inflation of cells 340 CH 19 BAKING may be started before baking by fermentation, macroscopic scale (for example, pores connected while sheeting – involved in the production of to the exterior of the dough or setting of the outer cream crackers for instance – largely eliminates layers of dough). Few studies have encompassed the gas produced from previous steps of process- all these different scales in attempting to provide ing within the dough (Wade 1988). a clear explanation for controversial data and theories in the literature. In this context, the rela- tive contribution of the different mechanisms to Mechanisms favoring volume the restrictions of expansion will not be discussed increase in cell further as more multiscale research is required. Thermal expansion of gases initially occluded in the bubble is responsible for 10–15% of cell infla- Starch gelatinization Starch gelatinization and/ tion. Further inflation is hence supported by the or protein aggregation contribute to the stiffen- vaporization of dough components at the cell– ing of dough (Figure 19.1) and reduce the capac- dough interface; vaporization is hastened by tem- ity of cell walls to elongate and the gas cell to perature increase during baking. The solubility of increase in volume. Some evidence has been carbon dioxide accumulated during fermenta- obtained when studying the early interruption of tion decreases during baking and the dissolved the baking process, the most demonstrative case gases are vaporized. The vapor pressure of water being the growth of columnar cells in crumpets increases rapidly upon heating; the same applies where heat transfers unidirectionally (Pyle 2005). to ethanol accumulated during fermentation, As a consequence, the onset temperature of although this gas has received little attention in these reactions (in the vicinity of 60 °C for starch past studies. It is important to understand that gelatinization and 70 °C for protein aggregation) water vapor is the predominant gas for cell infla- is often related to a build-up in gas pressure and tion in specific bakery products like sponge cake the restriction of, and ultimately the cessation of, (Lostie and others 2002) or non-yeasted puff cell growth. This molecular mechanism would pastries. Finally, most biscuits, cookies, and cakes favor the influence of gases with low tempera- are chemically aerated, the most commonly used tures of vaporization on cell expansion, high- reagents being sodium bicarbonate, alone or with lighting a contrasting effect of temperature rise an acidulant, and ammonium bicarbonate. With which favors gas production on the one hand but sodium bicarbonate, for instance, the hydrogen limits dough extension on the other hand. ion reacts with the bicarbonate ion to produce However, pressure measurements recently per- water and carbon dioxide, the latter supporting formed at different locations in bread dough with cell inflation. The decrease in H+ concentration is miniaturized sensors showed no temperature- accompanied by an increase in pH in the batter. dependence of the pressure build-up during This is slightly counterbalanced at the end of baking (Grenier and others 2010); this result baking by the decrease in overall water content questions the assumed relationship between (Baik and others 2000b). dough stiffening and pressure build-up, but could be explained if ruptures in the cell walls occur simultaneously or earlier (see later). In addition, Mechanisms limiting volume neither of these explanations seems to apply to biscuit or cookie baking. Chevallier and others increase in cell (2002) observed that the cessation of the oven- Mechanisms resisting or limiting the expansion rise in biscuit, which occurred in the temperature are multiscale, from molecular scale (for exam- range of protein aggregation, was usually ple, stiffening subsequent to starch gelatiniza- followed by collapse, hence questioning the rigid tion), microscopic scale (for example, ruptures in character of the dough matrix produced by cell walls and opening of the porous structure) to protein aggregation.  CELL EXPANSION AND SQUEEZING, CRUMB FORMATION, AND OVERALL COLLAPSE 341 106 1 2 3 5°C/min-tensile-44% water (Vanin and others 2010) 15°C/min-tensile-44% water (Vanin and others 2010) 105 Viscosity (Pa.s) 4°C/min-squeezing-32% water (Lassoued-Qualdi 2005) 2.5°C/min-tensile-45% water (Dresse and others 1988) 3°C/min-squeezing-?% water 104 (Bloksma 1985) 6°C/min-squeezing-?% water (Bloksma 1985) 9°C/min-squeezing-?% water (Bloksma 1985) 103 20 40 60 80 100 120 Temperature (°C) Figure 19.1 Wheat dough viscosity as a function of temperature, a compilation of literature data, where ‘squeezing’ means the use of the testing mode employing biaxial deformation (lubricated squeezing mode), and ‘tensile’ means the testing mode employing uniaxial deformation. Refer to the text for more details about stages 1, 2, 3. For color detail, please see color plate section. Ruptures in cell walls Ruptures in cell walls and protein coagulation (70–90 °C in the presence the formation of an open porous structure are of sugar). For optimal oven-rise, Eliasson and evidenced by observation of a product section Larsson (1993) also recommended adjusting the with the naked eye or with microscopy (Lostie thermal gradient inside the bread dough to the and others 2002), as well as by the uniformity strength of the gluten film forming the interface in pressure as reported at the end of baking or of cells. during post-chilling (Grenier and others 2010). Rupture is more generally determined by the Nevertheless, closed pores would still remain at extent and rate of deformation in cell walls; resist- the end of baking, occupying a still uncertain vol- ance to rupture can hence be modulated by dough ume fraction, of 1–20% (Datta and others 2007). formulation (high protein content in the flour, Rupture in cell walls is a prerequisite for limiting low water content in dough, addition of emulsi- cell shrinkage upon cooling and for preserving fiers) and/or the mastering of the mixing step in the oven-rise. Drastic shrinkage of bread loaves terms of under- or over-kneading (Dobraszczyk formulated with tapioca starch was related to the and Salmanowicz 2008). Consistently, dynamic absence of ruptures in cell walls (Kusunose and monitoring by MRI showed that for low interfer- others 1999). ence of the crust setting, the expansion of bread Pressure gradients between neighboring cells dough ceased locally at temperatures lower than tend to promote rupture. In the literature, rup- the onset temperature of starch gelatinization, a ture in cell walls is very often assumed to be due result which was associated with early ruptures in to the pressure build-up subsequent to dough cell walls (Wagner and others 2008). In this study, stiffening. Mizukoshi and others (1980) working an influx of gas higher than the outflux was sup- on biscuit dough heated in a liquid paraffin bath posed to maintain the open porous structure observed the formation of bubbles on the dough and avoid its collapse until its setting by starch surface and the massive emission of gas towards gelatinization. the exterior of the bath as soon as the internal The net balance between influx and outfluxes temperature of the sample exceeded that of is crucial for evaluating the impact of the pore  (a) Region 4 Region 3 Region 2 Region 1 100 110 Total thickness 100 Region 1 Region 2 90 90 Region 3 Thickness variation (%) 80 Region 4 70 Large bubble 80 Area (mm2) in region 4 60 50 70 40 30 60 20 10 50 0 –10 40 0 2 4 6 8 10 12 14 16 18 20 Time (min) (b) 110 Total thickness 100 Region 1 90 Region 2 Thickness variation (%) Region 3 80 Region 4 70 60 50 40 30 20 10 0 –10 0 2 4 6 8 10 12 14 16 18 20 Time (min) Figure 19.2 Changes in local porosity as monitored by MRI during bread pan-baking (Wagner and others 2008. Reproduced with permission of Elsevier). Dough regions are separated by one or two rows of oil microcapsules of which a high MRI signal (bright spots) permit the tracking of dough displacement subsequent to bubble inflation. Heat is transferred from bottom and top faces only. Associated MRI images are displayed at the top of each sub-Figure 19.for selected baking times; high signal intensity is related either to oil microcapsules (bright spots) or to dense regions of crumb especially present during the second part of baking. The tracking of microcapsules in case b was stopped when crumb structure became too dense  CELL EXPANSION AND SQUEEZING, CRUMB FORMATION, AND OVERALL COLLAPSE 343 in their vicinity (see last image). (a) case of late crust setting (oven temperature 130 °C): the different regions stop expanding all together; final porosity is unevenly distributed because of the limited mass transfer at the bottom (next to the baking tin); standard deviation from three repetitions was less than 3% in thickness variation (b) case of early crust setting (oven temperature 180 °C): inflation at the core (region 2) occurs after the cessation of the overall expansion and proceeds at the expense of the other regions which get squeezed. For color detail, please see color plate section. opening on cell expansion: for tin baking, mass pushing of the inner part of the dough will have exchange between the bottom region and the no squeezing effect; if a cell wall is still partly exterior is limited by the tin walls and also the flexible, it will be at least partly squeezed. The dough core structure becomes porous later, limit- relationship between dough stiffening and rup- ing exchange of the bottom region with the top ture is hence crucial; the greater the separation region and the oven; the bottom is hence likely to between the two events, the less the degree of retain its gases despite ruptures in cell walls and, squeezing (Jefferson and others 2007). Since as a result, the porosity will expand proportion- these events are often associated with threshold ally more than in other regions, by a factor of two values of temperature, and the dough is heated to four (Wagner and others 2008). from the outside mainly by conduction, the outer crumb (beneath the crust) is more prone to Setting of the crust and its effect of cell collapse or squeezing. Zhang and others (2007) developed coalescence Early crust setting is an additional, a device with a fabric lid placed above the top macroscopic determinant of cell expansion. There is surface of a bread pan and which made it pos- minimal interference by the crust with local cell sible to stop the oven-rise at different heights. inflation if crust setting occurs (just) after the Using MRI they showed that the later the oven- dough at the core has expanded (Wagner and oth- rise was stopped, the deeper the location of the ers 2008) (Figure 19.2a). The reverse is, however, squeezed crumb. Consistently with the above more frequently encountered in realistic baking mechanistic sequence, this result was explained conditions (Figure 19.2b). The intensity of the crust by the isotherm of wall rupturing moving deeper interference depends on the dynamics of dough into the product as more time was allowed for stiffening at the product surfaces and heat trans- heat transport into the dough. A wider tempera- port to the dough core, supporting cell inflation. ture gradient would be expected to develop in Crust setting results in a constant total volume; larger cells, which are more prone to squeezing if the crust sets early, it is accompanied by the (Jefferson and others 2007). The growth of a building of pressure at the core, resulting from large cell followed by its squeezing upon the the vaporization of gases under limited increase expansion of the dough core is shown in in cell volume (Grenier and others 2010). The Figure 19.2a; a cell formed in the bread dough at pressure build-up particularly concerns the the end of fermentation might no longer exist unbaked region with closed pores. The force in the final baked structure of the loaf. Because exerted by the core on other dough regions of these force balances developing in the dough, becomes higher as baking progresses; thus, the changes in local porosity are not monotonic core inflates at the expense of other dough during baking, and phases of compression and regions in which gas gets compressed, resulting in inflation occur alternately. Thus, the outer layers a reduction in volume as visualized by MRI (see of bread may be 2–3 times denser than the initial Figure 19.2b), and/or an increase in pressure dough (Jefferson and others 2007). (Grenier and others 2010). The amount by which The increase in pressure generated by the crust a cell collapses is determined by the extent to setting also imposes additional stress on the cell which the cell wall is set when it ruptures walls, possibly leading to bubble coalescence in (Jefferson and others 2007). If the cell wall is the still unbaked part of the product and finally to completely set when it fractures, the outward coarse texture of the crumb. Hayman and others 344 CH 19 BAKING (1998) mimicked the crust setting by applying to permeability (Chevallier and others 2002); (ii) the top of the bread pan various levels of over- the late shrinkage of the crust subsequent to pressure and at different baking times. They found cross-linking reactions such as extensive protein that an over-pressure of 0.3 kPa applied when the coagulation and Maillard browning reactions temperature at the dough core was between 60 (Sommier and others 2005). and 70 °C accentuated the coalescence between neighboring cells, as shown by the more open- grain crumb structure of the final bread slices. When pressure was continuously monitored Total water loss inside bread dough during baking, sudden falls in Whatever the type of bakery product, it is impor- pressure were associated with bubble coalescence, tant to realize that moisture gradients exist in the which was frequently observed at the dough core product at the end of the baking process. The and less frequently elsewhere (Grenier and oth- distribution in water content through the product ers 2010). These pressure drops started at 40 °C governs thesection). Water also plays an important and stopped at 70 °C; dough stiffening that pro- role in the major structural changes that take place ceeds beyond this temperature may render the during bread baking, at the microscopic (e.g. water coalescence more difficult. Pressure drops were vaporization in the cell expansion) and molecular also more frequent when higher pressure gradi- scale (for example, starch gelatinization). ents were developed such as for short-proved doughs (Grenier and others 2010). The increase in pressure generated by the crust Measurements of water content setting may also lead to its own rupture, allowing in bakery products prolongation of the oven-rise, but also contribut- In practical terms, although the overall water loss ing to strong spatial heterogeneity in crumb is easy to assess, the determination of water con- expansion as monitored by MRI for bread tent profiles in crumb during baking requires a (Wagner and others 2008) or X-rays for cake method to be set up for the accurate cutting of (Whitworth 2008). loaf samples while still hot and deformable, and Finally, pressure build-up in bread is accompa- the minimization of water losses by evaporation nied by a release of CO2 (Zhang and others from cut surfaces. Water content is also sensitive 2007). The net balance of the two mechanisms on to the time elapsed from the end of baking and cell expansion has not been assessed to date; the relative humidity of the storage atmosphere; despite the release of CO2, pressure build-up these important factors are rarely reported and might be favorable to cell expansion. On the make experimental results between studies almost other hand, CO2 release may be desired since it is impossible to compare. assumed to be a vector of aroma compounds (Eliasson and Larsson 1993). Spatial distribution of water Collapse of products The collapse of products is content in different bakery often reported in the literature, and to significant extents. For example: −13% (Hayman and others products 1998) and −30% (Zanoni and Peri 1993) for The initial water content typically ranges from bread, −20 to −25% for sponge cake (Lostie and 11 to 30% for biscuit and crackers (Wade 1988), others 2002), and −100 to −150% for biscuit 32–38% for cake batter (Baik and others 2000b), (Chevallier and others 2002). Nevertheless, this and 38–48% for bread dough. phenomenon has received little attention so far. The final moisture of a freshly baked biscuit is Two main hypotheses have been proposed to usually within the range 1–5%, depending on the explain it: (i) the insufficient rigidity of the matrix type of product (Wade 1988). Water distribution is to overcome the effect of gas loss or increase in very difficult to measure in the thickness of a  STRUCTURAL CHANGES IN STARCH AND PROTEIN DURING BAKING 345 biscuit; it can, however, beBread crumb located by the starch granules and the increase in volume beneath the crust exhibits an intermediate water of these starch granules; (ii) hydration-facilitated content, reported as being 2–12% lower than that melting whereby the crystalline structure of the of dough which was at 42% (Thorvaldsson and granules is irreversibly lost, and (iii) disruption of Skjöldebrand 1998) or 3–9% lower than that of starch granules owing to further hydration of the dough which was at 49% (Wagner and others disordered polymer chains, followed by the partial 2007) or at 3% itself (Marston and Wannan 1976). dispersion of amylose and amylopectin. Amylose The crumb at the core retained its initial moisture association increases as the concentration of solu- content or it even increased (Wagner and others ble amylose increases and when the temperature 2007). When sampling took place just after removal is less than the melting temperature of amylose of bread loaves from the oven, an increase in water crystallites (Morris 1990). Granular swelling content by 1–3% was reported at the core (De together with exudation of amylose from gran- Vries and others 1989; Zanoni and Peri 1993; ules, are responsible for the increase in viscosity. Wagner and others 2007). In addition to the Starch gelatinization is an endothermic pro- conventional, destructive method of sampling cess, and the number of endothermic transitions combined with oven-drying for water content (one to four) depends on dough composition, determination, a continuous method based on near mainly water content (Figure 19.3). The first and infrared reflectance (NIR) was proposed by second endothermic transitions reflect the disor- Thorvaldsson and Skjöldebrand (1998). Light was ganization (melting) of starch crystallites; third sent into the sample through an optic fiberThor- and fourth endothermic transitions (endotherms) valdsson and Skjöldebrand (1998) showed an are attributed to the dissociation of amylose–lipid increase in water content in the core immediately complexes that formed in the previous process after the local temperature had reached 70 ± 5 °C steps. Attainment of the first two endotherms is (that is, after cell expansion was more or less com- accompanied by the loss of birefringence, which plete), which reached +8% at the end of baking is characteristic of native starch (Burt and Russell (35 min at 225 °C) relative to that of dough initially. 1983). The extent of starch gelatinization during Magnitudes of the overall drying rate were baking can be estimated by the quantity of heat 1.8 × 10–3 kg water/kg dry matter for cookies uptake as assessed during calorimetric analysis (Hayakawa and Hwang 1981), 1.4 × 10–3 to relative to a reference; the reaction follows first- 9.4 × 10–4 kg/kg for yellow and chocolate cakes of order kinetics (Marcotte and others 2004). various sizes (Baik and others 2000b), 2.5 × 10–4 kg/kg for sponge cake (Lostie and others 2002), and 2.8 × 10–5 to 2.4 × 10–4 kg/kg (Hasatani Factors affecting starch and others 1992) and 2–5 × 10–4 kg/kg (Sommier gelatinization and extent of starch and others 2005) for bread. gelatinization in bakery products In the presence of sodium chloride or sucrose Structural changes in starch at low concentrations, gelatinization occurs at and protein during baking and slightly higher temperatures, and is less complete as evidenced by lower enthalpies of melting subsequent changes in dough/ (Biliaderis 1991; Ghiasi and others 1983). Marcotte crumb mechanical properties and others (2004) reported a peak gelatinization Basic mechanisms of starch temperature in the presence of cake ingredients in the range 94–97 °C. Fats and emulsifiers also gen- gelatinization erally delay gelatinization (Richardson and others Starch gelatinization comprises a series of 2003). In the temperature range 60–70 °C, a high processes on a molecular scale, which overlap in level of activity of α-amylase in doughs causes time: (i) swelling, that is the absorption of water enzymatic breakdown of gelatinized starch, which 346 CH 19 BAKING 200 Toven = 100°C (Vanin and others 2012) Toven = 180°C (Wagner and others 2007) 180 Toven = 205°C (Zanoni & Peri 1993) Toven = 235°C (Marston & Wannan 1976) 160 Toven = 250°C (Vanin and others 2012) 140 Starch gelatinization and melting Temperature (°C) 120 3rd endotherm 100 Tg 80 (gluten and bread) 2nd endotherm 60 1st endotherm 40 20 - 10 20 30 40 50 Water content (% wet basis) Figure 19.3 Heat-moisture pathways of crust (filled symbols) and crumb (open symbols) from the literature, as represented in the state diagram of starch and gluten, and where Toven is the oven temperature. in turn attenuates the increase in viscosity nor- leaching, phase separation from amylopectin and mally associated with gelatinization. As a result of network formation, possibly accounting for the the limiting amount of water and/or the presence differences in gel properties and baked product of co-solutes, the starch is only partly gelatinized texture (Patel and others 2005). in many baked products. The degree of gelatiniza- tion is low in dry products, such as biscuits, where a considerable degree of birefringence can still Protein coagulation be observed after baking (Wade 1988). Up to 70% Heating above 60 °C also gradually modifies the of the starch granules gelatinize at the center of a protein network, with the polymerization of bread loaf (Biliaderis 1991); some starch granules glutenins and the changes from gluten gel to lose their birefringence but still appear intact, coagel. In dough, water does not migrate rapidly; whereas other are disrupted. The partially swollen thus, every starch granule gelatinizes by using the granules usually become stretched into elongated water in or near the granule. As there is insuffi- forms enabling gas cells to expand. cient water in the dough to gelatinize completely, The swelling of starch granules as observed by some of the water will transfer from the protein light