Crust Formation and Its Role During Bread Baking

Trends in Food Science & Technology 20 (2009) 333e343

Review

* Corresponding author.

0924-2244/$ - see front matter � 2009 Elsevier Ltd. All rights reserved.doi:10.1016/j.tifs.2009.04.001

Crust formation and

its role during bread

baking

F.M. Vanina,b, T. Lucasa,b,* and
G. Trystramc
aCEMAGREF, Food Process Engineering Research Unit,

Rennes, France (Tel.: D33 (0)2 23 48 21 77; fax: D33

(0)2 23 48 21 15; e-mail: [email protected])bUniversite Europeenne de Bretagne,

F-35000 Rennes, FrancecAgroParisTech, Joint Research Unit Food Process

Engineering MASSY, France

The final properties of the crumb and crust differ according to

their heatemoisture dynamics. Compilations of heating and

drying rates reported in the literature are discussed and will

serve to validate future models of baking. Their impact on

the structural elements in dough films and the porous network

are discussed, highlighting the lack of data and the need to re-

produce these dynamics inside the instrument of analysis.

Some roles of the crust setting during the whole baking process

are also presented, suggesting further research in this area.

Finally, as the region covered by the crust should be defined

as a starting point to future studies, definitions proposed in

the literature are discussed.

IntroductionDuring baking, heat (HT) and moisture (MT) transport

take place in the dough simultaneously and interdepen-dently, and involve three major changes e (for furtherdetails see recent review by Mondal & Datta, 2008):

1. Water vaporises at the cell/dough interface, and gasesaccumulated during fermentation (CO2, ethanol) orgenerated by chemical raising agents are also vapor-ised: the cell volume increases provided that the doughfilm retains gases and is deformable.

2. Starch gelatinisation and protein coagulation transformthe viscous dough into a mainly elastic crumb; theserheological changes limit the cell growth described in1 and enhance pressure build-up.

3. The structure with gas cells separated by films is trans-formed into a porous structure with inter-connectedpores. In theory ruptured films limit cell growth (1.):gas molecules are exchanged between adjacent opencells, and finally transported out of the dough. Doughfilms rupture when they can no longer withstand overpressure. Film rupture is often associated with the onsettemperature of starch gelatinisation (Bloksma, 1990) ora higher temperature. This is also likely to happen if pres-sure is low, but the dough film presents poor mechanicalresistance, as when the protein content in the flour is low,the water content in dough is high, or if the dough is underor over-kneaded (Dobraszczyk & Salmanowicz, 2008).

The crust and crumb come from the same original dough,but their final properties differ according to a distinct localheatemoisture treatment. As soon as the dough is placed inthe oven, water evaporates very fast from the surface layers,resulting in a much lower water content (<20% wet basis)than at the core. The total water loss (WL) from the crust isobviously of economic significance (weight loss). Waterdistribution between the crust and crumb also contributessubstantially to the organoleptic perception of the final prod-uct. As the water activity (Aw) of sponge cake increased fromw0 to 0.75, both initial modulus and critical stress fell by anorder of magnitude with both correlating (r> 0.90) witha trained panel’s assessment of sensory harness (Attenburrow,Goodband, Taylor, & Lillford, 1989). A crispy texture is alsoassociated with low moisture content and water activity, whenstarch and gluten matrix are in a glassy state making cell wallsmore prone to fracture (Stokes & Donald, 2000). Low watercontent in the crust will also affect the rheological changesin the dough films and ultimately the cell growth during bak-ing. Starch gelatinisation and protein denaturation are limitedwhen there is restricted access to water. This has an effect an-tagonistic to the one of the decreasing water content and mayprolong the deformability of dough films. The enhancedescape of gases due to the proximity of the boundary to theoven may explain the smaller-sized cells which are typicalof the crust. These obviously contribute to the mechanicalproperties of the crust (resistance to rupture during baking,texture in mouth) in addition to the low water content.

mailto:[email protected]

334 F.M. Vanin et al. / Trends in Food Science & Technology 20 (2009) 333e343

As the water content decreases in the crust, the temper-ature can exceed 100 �C, which supports other reactionssuch as the Maillard reaction responsible for the develop-ment of colour and the release of flavours, and also theproduction of toxic compounds with safety implications(Ahrne, Andersson, Floberg, Rosen, & Lingnert, 2007).

Reactions specific to the crust also jeopardize the nutri-tional value of the baking process. On one hand, the Maillardreaction decreases the protein digestibility and the lysinebioavailability which is the limiting amino-acid in cerealproducts (O’Brien & Morrissey, 1989). On the other hand,gelatinised starch can be degraded by amylases from the sa-liva whereas ungelatinised fraction undergoes a much slowermetabolism and can remain undigested. Conversely, healthbenefits (obesity risk reduction) are claimed since bread rollswith a higher proportion of crust have been reported to raisecapillary blood glucose more slowly than a correspondingloaf (Glatzel & Rettenmaier, 1962).

All these microstructural changes in the cell size anddough films also contribute to the structural differentiationof the crust. Most previous studies have focused on thecrust properties at the end of baking or their changes duringstorage (Luyten, Pluter, & van Vliet, 2004), but the settingof these properties during the baking and post-chilling pro-cesses have been little studied to date.

The first aim of this review is therefore to present how thewell-known final properties (once cooled) are created dynam-ically during baking and how the underlying transformationsdiffer from those in the crumb (section Mechanisms specificto the bread surface and contributions to crust formation).The crust properties differ from a bread making technologyto another (pan bread versus hearth bread for instance), butalso from one face to the other, depending on the extent ofheat and mass transfer at the bottom and top surfaces for in-stance. Given the very few data available in the literature onthe topic, elaborating a typology from this review is out ofscope; however, the variety present in the literature will helpto identify key factors in the crust setting process.

Improvements in the observation techniques will behighlighted in section Mechanisms specific to the breadsurface and contributions to crust formation and summedup in the conclusion.

As the crust cannot be considered separately from therest of the bread, but interacts with other mechanisms,some of its role during baking already identified in theliterature will be presented in section How the crust affectsother mechanisms during baking.

As changes in bread are gradual, proposing a simple defini-tion of the crust is a difficult although necessary starting pointfor future studies, a second focus of the concluding section.

Mechanisms specific to the bread surface andcontributions to crust formationTemperature

Surface temperature quickly reaches 100 �C and thenapproaches more slowly the oven air temperature. This

rapid increase in surface temperature is enhanced (i) bythe low thermal conductivity of the aerated dough limitsHT at core; (ii) by high radiation originating from thewarm oven walls (from 66.2% to 81.5% of overall HT)(Baik, Marcotte, & Castaigne, 2000). Surface temperaturesexceeding 100 �C are explained by the difference in watercontent between surface and core: water evaporates fromthe surfaces more quickly than it can be transported fromthe core; in addition, due to the evaporationecondensa-tionediffusion mechanism (Wagner, Lucas, Le Ray, &Trystram, 2007), water content remains almost constant atthe core.

Table 1 summarizes the heating rates observed at thesurface compared to the centre, as approximated by theslope of the linear segment between the initial temperatureand 100 �C. The heating rate was high in the crust (up to14.4 �C/min) while not exceeding 3 �C/min at core. Notethat time derivation of temperature kinetics gave evenhigher values (20e35 �C/min) at the onset of baking.

Temperature was measured using thermocouples (Dogan,2002; Lostie, Peczalski, Andrieu, & Laurent, 2002a; Zanoni& Peri, 1993) or optic fibres (Wagner, Loubat, et al., 2008)placed at different locations inside the loaf, usually onlynear the top surface and in the centre. However, this intrusivetechnique introduces biases, including delicate positioning(1e3 mm of accuracy), possible heat conduction along themetallic wire of the thermocouple and constraint to oven-rise. To minimize such bias, surface temperature can bemonitored by infrared sensors (Lostie, Peczalski, Andrieu,& Laurent, 2002b; Primo-Martin, van Nieuwenhuijzen,Hamer, & van Vliet, 2007).

Water contentThe water content in the crumb remains almost constant

compared to the initial value while considerable dehydrationoccurs in the crust (Fig. 1). Water activity can also be used todescribe the result of WL and its effects on water availability,reaction kinetics and mechanical properties (Bassal, Vasseur,& Lebert, 1993; Czuchajowska, Pomeranz, & Jeffers, 1989;Dogan, 2002; Lind & Rask, 1991; Van Nieuwenhuijzen,Tromp, Hamer, & Van Vliet, 2007). Given the low final watercontent in the crust (Fig. 1), a moderate variation in watercontent can affect Aw considerably (Lind & Rask, 1991).Aw is also involved in the calculation of water transfer duringbaking, since the driving force was generally assumed to bethe difference in partial vapour pressure between the productsurface and the oven atmosphere. Aw was measured at ambi-ent temperatures, and very little information is available fortemperatures observed in the crust during baking. Studieshave been carried out at 100e150 �C for cake dough (Bassalet al., 1993), and at 80e120 �C for wet and dried crumb (Jury,Monteau, Comiti, & Le-Bail, 2007). Additionally, theassumption of thermodynamic equilibrium between liquidand gaseous phases on which the measurement of Aw is basedhave also been questioned, given the high heating rates

Table 1. Heating rates for different cereal products and baking conditions at bread surface and core

Product Mass of dough/batter (g)

Surfaces exposedto heat transfer

Position of temperaturein dough/batter

Oven airtemperature (�C)

Heating rate(�C/min)

Reference

Bread 200 Top bottom Top surface 185 7.7(Wagner, Loubat,et al., 2008)

Bottom surface 2.6Centre 2.4

Bread 341 All Top surface 203 11.3 (Zanoni et al., 1993)Bottom surfaceat 1 cm beneath

9.5

The top surface 2.6Bread e All Top surface 210 10.3* (Thorvaldsson, &

Skjoldebrand, 1996)Mid-width 3.9*Centre 1.8*

Pan bread e e Surface 220 8.8 (Dogan, 2002)Francala e e 14.4Bread 1.5 L All Top surface 225 6.7 (Thorvaldsson, &

Skjoldebrand, 1998)Bottom surface 4.4Bread 760 All Side surface 235 7.4 (Marston & Wannan, 1976)

Centre 2.8Biscuit e All Centre 300 75.0**

(400 g/kg dry air)(Chevallier et al., 2002)

300 61.7**(20 g/kg dry air)

240 44.7**(200 g/kg dry air)

180 19.6**(20 g/kg dry air)

Sponge cake 400e600 Top Top surface 200 0.5 (Lostie et al., 2002b)

‘‘e’’ Refers to missing information.*Data from reheated samples.**With different relative air humidity.Bold and italic values emphasizes high surfaces values.

335F.M. Vanin et al. / Trends in Food Science & Technology 20 (2009) 333e343

involved during baking (Zhang & Datta, 2006). This ap-proach was therefore not included in this review.

HT has been considered the main driving force for watervaporisation; in other words, MT from the vaporisation frontto the loaf outside, proceeding mainly by convection, was notconsidered to be a limiting factor (Lostie, Peczalski, &Andrieu, 2004; Zhang, Doursat, Flick, & Lucas, 2008). HTcan be enhanced by increasing the oven air temperature,and eventually the air renewal at the product surface; thesevariables have consistently been experimentally correlatedto WL (Wahlby & Skjoldebrand, 2002). The addition of

0

10

20

30

40

50

60

70

80

0 10 20 30 40Time (min)

Water co

nten

t

(g

o

f w

ater/ 100g

o

f d

ou

gh

)

Wagner et al (2007)Zanoni et al (1993)Czuchajowska et al (1998)Westerlund et al (1989)Marston et al (1976)

Fig. 1. Water content in crust (hollow symbols) and crumb (filled sym-bols) during baking from different literature sources.

steam increased the relative humidity of the oven air to0.7e0.9 (Wiggins, 1998). As the dough surface was coldcompared to the surrounding air, moisture condensed ontoit from the air, up to 1% of loaf weight. Steam injectionthus reduced WL at the onset of baking. However, as the sur-face reached the dew point temperature, steam acceleratedtemperature rise and WL for long baking times (Eliasson &Larsson, 1993). Chevallier, Della Valle, Colonna, Broyart,and Trystram (2002) reported greater WL with higher airhumidity in the oven (20e400 g/kg of dry air) for biscuits.Internal resistance to HT also affects WL; because this in-volves interlinked mechanisms, it will be discussed in sectionHow the crust affects other mechanisms during baking.

Water content was generally obtained by weighing theproduct after stopping the baking process at different times,which is extremely time consuming. Moreover, assessingprofiles requires cutting the loaf into samples while stillhot, and also deformable. Vaporisation of water from cutsurfaces and squeezing crumb samples are thus sources ofbias. Another difficulty of the sampling method wasseparating the crust from the crumb. Thorvaldsson andSkjoldebrand (1996) developed an infrared method for con-tinuous measurement of local water content, but the effectof crumb density on the infrared signal was not corrected,generating bias during the oven-rise period.

Validating the thorough description of the mechanismsof water transport (Lostie et al., 2002a; Lucas, Wagner,

0

30

60

90

120

150

180

210

0 20 40 60 80

Water content (wet basis)

Eliasson et al 1980Chamepnois et al 1995Burt et al 1983Cuq et al 2001 (gluten GT)Toufeili et al 2003 (gluten GT)Le Meste et al 1992 (bread GT)

2nd endotherm

1st endotherm

gelatinisation

gluten and breadglass transition

3rd endotherm

Tem

peratu

re (°C

)

Fig. 2. Dependence on water content of gelatinisation temperatures ofwheat starchewater mixtures and GT of gluten and bread as demon-

strated by DSC from different literature sources.


Doursat, Flick, & Trystram, 2009; Zanoni et al., 1993)would obviously require higher spatial and temporal resolu-tion for profiles of water content; this difficulty has beendemonstrated for the crumb (Lucas et al., 2009) and is validfor the crust with even greater acquisition constraints(spatial domain of a few hundred to a few thousandmicrometers). Note that experimental evidence hardly sup-ports a modelling approach based on uniform temperatureand water content as suggested by Fan, Mitchell, andBlanshard (1999).

Biochemical reactionsStarch gelatinisation

The degree of starch gelatinisation (DSG) can beobserved on final cereal products (once completely cooled)by the loss of birefringence of starch granules or the disap-pearance of the ‘‘Maltese crosses’’ observed in polarizedlight (Eliasson & Larsson, 1993; Jenkins & Donald, 1998).DSG has been shown to be directly affected by water(Burt & Russell, 1983; Cuq, Abecassis, & Guilbert, 2003;Fessas & Schiraldi, 2000; Jenkins & Donald, 1998; LeMeste, Huang, Panama, Anderson, & Lentz, 1992) and theheating rate (Bloksma, 1980; Donovan, 1979), which againboth differ between the crust and crumb. If DSG consistentlydiffered between cereal products because of their differentcompositions and baking times (Eliasson & Larsson,1993), it also varied within the same product: the greaterthe distance to the surface, the greater the starch gelatinisa-tion (Luyten et al., 2004). Up to 40% of the starch in the crustdid not gelatinise during bread baking (Primo-Martin et al.,2007). This was consistent with DSC measurements carriedout with open pans (Fessas & Schiraldi, 2000): for a globalwater content decreasing from 44.2% to 30.7% (wb), DSGwas about 44% e instead of 66% measured for sealedpans at initial water content of 44.2%.

Heatemoisture treatment of starch is known to affect the ge-latinisation temperature. Differential scanning calorimeter(DSC) thermograms obtained from samples of high water con-tent (>90% wet basis (wb)) exhibited a single endotherm peak(Fig. 2). Beyond this point and at water contents abovew60%wb, the endotherm presented three major endotherms(Burt & Russell, 1983; Champenois, Colonna, Buleon, Valle,& Renault, 1995; Eliasson, 1980), the first peak correspondingto cooperative, water-mediated melting of starch crystallites,the second peak to the melting of the remaining crystallites,and the third to amyloseelipid complex melting transition(Biliaderis, Maurice, & Vose, 1980; Burt & Russell, 1983; Che-vallier, Colonna, & Lourdin, 2000; Fessas & Schiraldi, 2000).As water content progressively decreased below this point,the intensity of the first endotherm was reduced, but remainedat almost the same temperature (Fig. 2), until its total disappear-ance below to 35% of water (Burt & Russell, 1983; Champenoiset al., 1995; Eliasson, 1980). Conversely, Sopade, Halley, andJunming (2004) reported an increase in the gelatinisation tem-perature of 1.1 �C for a reduction in water content from 50%wbto 40%wb. At the same time, when water content was lower

than 60% the second and third endotherms moved to highertemperatures but at different rates (Fig. 2) and finally mergedinto a single endotherm above 35%wb (Burt & Russell, 1983;Champenois et al., 1995). In fact, the starch granules and thegluten will compete for water, which will result in an additionalincrease in the gelatinisation temperature (Eliasson & Larsson,1993; Wang, Choi, & Kerr, 2004). The loss of birefringence ob-served for very low water content (<25%) and at high temper-atures was related to the second endotherm (Burt & Russell,1983). A peak at 70 �C was observed in the thermogram ofa freshly sampled crust, corresponding to the starch crystalsthat did not gelatinise during baking, in addition to the peakcorresponding to the amyloseelipid complex (Primo-Martinet al., 2007).

Some studies have also shown that the gelatinisationtemperature was affected by the heating rate. Althoughapplicable to potato starch (Donovan, 1979), observedthat the greater the heating rate (2e10 �C/min), the higherthe temperature of the endotherm peak (þ2 �C). Likewise,Patel and Seetharaman (2006) observed a shift in the swell-ing of wheat starch granules at higher temperatures whenthe heating rate increased from 5 to 25 �C/min. It mustbe mentioned here that changes in dough viscosity werealso delayed by increasing the heating rate from 3 to9 �C/min (Bloksma, 1980).

To conclude, it can be expected that starch granulespresent in the superficial layers only partially gelatinise, be-cause of the strong reduction in water content and the morerapid heating rates, which both increase the starch meltingtemperature. This will influence the rheological propertiesof dough films during baking and possibly cell growth inthe crust. As far as the crust is concerned, it would be in-teresting to characterise this reaction during simultaneousheating and drying.

Gluten coagulationThe lower the water content, the higher the denaturation

temperature of proteins (Eliasson & Larsson, 1993). Foru-gliadin aqueous solutions for instance, it increased from


117 to 157 �C by lowering the water content from 20%wb to3%wb (Noel, Parker, Ring, & Tatham, 1995). Denaturationis followed by aggregation and, for some proteins, gel for-mation (Eliasson & Larsson, 1993). In contrast to the firststage, this stage is a kinetics-dependent exothermic process:the lower the heating rate, the more advanced the process ofaggregation (Myers, 1990).

As a consequence of these combined effects, the breadcrust showed a minimal decrease in water-extractible pro-teins, meaning that they were not as aggregated and/orcross-linked as in the crumb (Westerlund, Theander, &Aman, 1989). This would undoubtedly affect the mechani-cal modifications of the gluten network in the superficiallayers upon heating. The disappearance of soluble proteinwas almost complete in wheat gluten and glutenin at25e30%wb of water content heated at very high rates(25 �C/s from 40 to 170 �C) (Strecker, Cavalieri, Zollars,& Pomeranz, 1995); to the light of this last study, resultsfrom Westerlund et al. (1989) could be explained to thevery low water content encountered in bread crust.

The amount of water to be released was also expected to belower, with possible limitations on other biochemical reac-tions including starch gelatinisation (Marston & Wannan,1976; Wang et al., 2004) and/or to water transport (Mondal& Datta, 2008; Thorvaldsson & Skjoldebrand, 1998).

Glass transition (GT) is also an issue given the low watercontents associated with high temperatures (Huang,Haynes, Levine, & Slade, 1996) (Fig. 2). However, it iscommonly accepted that GT usually occurs upon cooling(Cuq et al., 2003) e also see Fig. 2, which is outside thescope of this review and is not discussed here. Nevertheless,local heatemoisture treatment during baking (sectionsTemperature and Water content) undoubtedly affect theconsecutive process of GT during post-chilling.

Maillard reactionHighly reactive compounds are initially produced from

this reaction; afterwards, their polymerization yields prod-ucts of brown colour (Martins, Jongen, & van Boekel,2001) observable at around 105e115 �C (Broyart, Trystram,& Duquenoy, 1998; Wahlby et al., 2002; Zanoni, Peri, &Bruno, 1995a). Starch is also consumed during the reaction,leading to a reduction in the starch concentration in the crustcompared with the crumb (Cauvain, 1998). Finally, some ofthe desired flavours and aromas of bread, among which2-acetyl-1-pyrroline, 4-hydroxy-2,5-dimethyl-3(2H)-fura-none, 2,3-butanedione, methional, (E)-2-nonenal, methyl-propanal, and 2- and 3-methylbutanal for baguette crust,are also produced, mainly by the Maillard reaction (Poinotet al., 2008; Zehentbauer & Grosch, 1998a, 1998b).

Low water activity (optimal between 0.4 and 0.8) (Charissou,Ait-Ameur, & Birlouez-Aragon, 2007) and high temperatures(commonly above 105 �C as mentioned above) acceleratedthe Maillard reaction. WL and browning were linearly corre-lated (Purlis & Salvadori, 2007; Wahlby et al., 2002). Indepen-dently of the HT mode (natural or forced convection), the

lightness of the bread surface decreased as the oven temperatureincreased (180e220 �C); at even oven air temperatures, forcedconvection caused higher browning than natural convection(Purlis et al., 2007).

Different models based on first order kinetics were pro-posed to predict the kinetics of colour development. Theywere dependent on surface temperature alone (Zanoniet al., 1995a) or combined to water content, yielding errorsof prediction less than 10% for WL higher than 10% (Purliset al., 2007) or between 1% and 24% on biscuits at the endof baking (Broyart et al., 1998).

The Maillard reaction may also result in toxic com-pounds, such as acrylamide. Several recent reviews havebeen focalized in the study of acrylamide formation incereal products (Claus, Carle, & Schieber, 2008; Konings,Ashby, Hamlet, & Thompson, 2007; Sadd & Hamlet,2005) (for further details see the cited reviews).

Previous studies correlated its formation to the oven airtemperature, although the key variable is the exact localtemperature at the dough surface which depends on manyvariables other than the oven air temperature. The amountof acrylamide in the crust of bread, flat bread, dry starchsystem and dried rye-based flat bread rose with longerbaking times and higher oven temperatures (Brathen &Knutsen, 2005). A similar effect was reported by Ahrneet al. (2007).

Rheological evolutionThe rheological properties of dough change significantly

during baking, mainly as the result of the temperature-in-duced changes to the molecular structures. Between 26 and60 �C, the viscosity of dough decreased by a factor of 5(Bloksma, 1990), supporting the deformation of dough films.The water content is a key factor since a variation of 4% in-duced a variation in viscosity by a factor of 2 (Bloksma,1990). Above 60 �C, viscosity increased by a factor of 10,and the starch granules inflated and the gluten moleculeswere polymerized, damaging dough extensibility (Bloksma,1990). Moreover, the higher the heating rate, the lower theminimum viscosity and the higher the temperature wherethis minimum viscosity occurred (Bloksma, 1980).

Although starch gelatinisation and protein denaturation areresponsible for the loss of extensibility of the crumb, thisshould not apply to the crust (section Starch gelatinisation):starch gelatinisation is expected to be delayed by the highheating rate and the low water content encountered in the crustand this would favour low levels of viscosity. It is thus mostprobable that the loss of extensibility of superficial layers dur-ing baking is due to dehydration, although this point is notclear in the literature.

The rheological properties of bread dough have beenwidely characterised experimentally but generally not in con-ditions realistically representing the baking process,especially regarding the crust. Firstly, viscosity and eventu-ally elasticity have been thoroughly measured at ambientand/or constant temperatures, but rarely during dynamic

Fig. 3. Microstructure of bread baked in microwaveeinfrared combi-nation oven observed by SEM at �50 for (a) crumb, (b) crust. Bars rep-resent 300 mm. Source: Datta et al. (2006). Reprinted from Journal ofFood Engineering, 79, Datta, A.K., Sahin, S., Sumnu, G., & Keskin,S.O., Porous media characterisation of breads baked using novel heat-ing modes. 106e116, Copyright (2006), with permission from Elsevier.


heating. When applied, heating rates have been appropriate tothe bread core (Collar, Bollain, & Rosell, 2007), and rarely tothe crust, the highest comparable values being 9e11 �C/min(Bloksma, 1980; Singh & Bhattacharya, 2005). Secondly,bread dough has been characterised at very low uniaxial shearrates (between 10�2 and 10�3 s�1) (Bloksma, 1990) and morerecently under biaxial extension at rates of 5� 10�2 s�1

(Rouille, Della Valle, Lefebvre, Sliwinski, & vanVliet,2005) and 10�1 s�1 (Dobraszczyk & Salmanowicz, 2008).Most of these strain rates failed to reproduce those encoun-tered during baking (1e2� 10�3 s�1) (Bloksma, 1990).With the exception of biaxial extension (Dobraszczyk &Morgenstern, 2003), other tests used in combination withdynamic heating only approximately reproduce the expecteddeformation of dough films over a gas bubble during baking.Finally, rheological measurements applicable to the crustshould ideally be conducted under dynamic drying; againmost of the data available were obtained at a constant watercontent, with levels relevant to the dough.

In the current mathematical models of baking, themechanics are usually simplified and the dough consideredas viscous only (Lostie et al., 2002b; Lucas et al., 2009).Viscosity was not dependent on water content and was atbest temperature-dependent. Nevertheless, fair agreementbetween simulation and experiment was usually obtaineddue to the numerous parameters to be adjusted and becausethe validation was based on overall measurement of theoven-rise (Lostie et al., 2002a; Zhang & Datta, 2006),and rarely on local measurements (Lucas et al., 2009).The simplified rheological behaviour of dough (viscousrather than viscoelastic) used in these models could be re-sponsible for the discrepancies in total height (Lostieet al., 2002a) or in the intensity of local compression inthe crumb (Lucas et al., 2009).

Cell growthBecause of the proximity of the boundary with the oven

air, gases easily escape the crust and the pressure remainsclose to the atmospheric pressure (see pressure profilescalculated by Lucas et al., 2009 and Zhang & Datta,2006). Additionally, the loss of extensibility of dough filmsis expected to be more rapid, presumably because of thegreat reduction in water content. Thus, one reason to ex-plain the smaller size of cells in the crust than in the crumb(Fig. 3) is that the forces favouring expansion are not suf-ficient for cell growth, although the full mechanism under-lying cell growth warrants in depth study.

Moreover, the surface layers cannot deform fast enoughto accommodate the core expansion. This leads to a domedshaped top surface, and also to similar curvature of theinternal structures, i.e. elliptical shapes of cells observedduring baking (Whitworth & Alava, 2004). In fact, whensetting of the surface layers occurred early compared tothe heating at core (Fig. 4f), two opposite forces wereexerted on the intermediate regions, which tended to becompressed (Fig. 4d), the extent of compression being

related to the time elapsed between stiffening of doughfilms and their rupture (Jefferson, Lacey, & Sadd, 2006;Lucas et al., 2009).

The dimension, shape and number of gas cells in thecrumb have been studied by various imaging techniques;comparatively the crust was subject of few studies (Datta,Sahin, Sumnu, & Keskin, 2006; Primo-Martin et al., 2007;Zhang & Datta, 2006). It must be emphasized that the major-ity of microscopy techniques are destructive and do not al-low dynamic observation of a process such as baking.Other techniques providing dynamic measurements oftenaverage the bubble size distribution over the whole doughsample, which is relevant in the case of a quite homogeneousalveolar structure as expected during proving (Leroy et al.,2008). Another requirement for imaging is high spatial res-olution, given that the mean cell size in the crust is around100 mm. Among the techniques available, X-ray synchrotron

Fig. 4. Porosity profile for different final total heights (45 (a), 50 (b) and 55 (c) mm) after 45 min of baking. Porosity (d), water content (e) and tem-perature (f) profiles at different baking times for a final height of 60 mm. Adapted from Lucas et al. (2009) and Zhang et al. (2007).


seems to be the most appropriate. A quantitative, dynamicstudy of the alveolar structure in the crumb during bakingwas recently reported (Babin et al., 2006). The feasibilityfor the crust still needs to be tested. It would require warmingthe sample at higher rates and using higher frequencies of ac-quisition than those used for the crumb, although the signalto noise ratio would be lower due to low water content andlow levels of signal accumulation.

How the crust affects other mechanisms duringbakingMechanical constraint to crumb expansion, bothlocally and overallLoss of extensibility

The cessation of overall expansion was experimentallyrelated to the occurrence of a dry surface during spongecake (Lostie et al., 2002b) and bread (Zanoni et al.,1993) baking. Similarly, protecting dough samples withparaffin oil to prevent dehydration allowed continuousexpansion until 100 �C, whereas it ceased between 36 and62 �C (Le Meste et al., 1992).

Using magnetic resonance imaging (MRI), Wagner,Quellec, Trystram, and Lucas (2008) recently reportedthat if the crust was set early, and even if total expansionhad ceased, local expansion could continue (Fig. 4a-c),especially in the colder areas which had not alreadyexpanded (Fig. 4f). These areas expanded to the detrimentof others which were then compressed (Fig. 4d). Such com-pression was accompanied by the disappearance of the larg-est bubbles, suggesting that it involved the weakeststructures from a mechanical view-point (Zhang, Lucas,Doursat, Flick, & Wagner, 2007) developed a device with

a fabric lid which made it possible to stop the oven-riseat different heights. Using MRI they showed that the laterthe oven-rise was stopped, the deeper the location of thesqueezed crumb. Such a trend could be reproduced witha model of baking (Fig. 4a-c) which in turn made it possibleto relate the occurrence of compression to the mechanicalproperties of dough films (already ruptured but not alreadystiffened). Similarly, in the theoretical approach developedby Jefferson et al. (2006), the more a cell is set when itsfilm fractured, the less squashing took place and the lowerthe final density in that part of the bread. Likewise, Hayman(1998) suggested that crust formation resulted in an in-crease in the internal pressure in the unbaked portion ofthe dough, imposing additional stress on the dough films,and ultimately leading to bubble coalescence and toa coarser crumb structure. If no crust was formed, a finecrumb structure was expected, which was evidenced fromcross sections of loaves produced with their electrical resis-tance oven (ERO) for a wide range of wheat flours.

Additionally, MRI and X-ray images showed how crustrupture allowed prolongation of the local expansion andalso contributed to strong spatial heterogeneity in crumbexpansion (Wagner, Quellec, et al., 2008; Whitworth &Alava, 2004).

Crust setting has rarely been taken into account in themodels of baking. To do so, the rheological properties ofdough should be dependent on both temperature and watercontent (section Rheological evolution). Moreover, a properdescription would be two-dimensional (Zhang & Datta,2006). In current models of unidirectional transport andexpansion, the cessation of oven-rise was ‘‘successfully’’reproduced by artificially imposing nil displacement of


the outer boundary after a certain baking time e adjustedon experimental data e.g. Lucas et al. (2009), or by usinghigh values of viscosity at high temperatures e.g. Lostieet al. (2002a).

Lower permeability to gasesThe constraints to crumb expansion may also be gener-

ated by low permeability of the crust to gases, through thepressure build-up. Almost no pressure build-up wasobtained from baking models when high permeabilityvalues with uniform distribution through the dough wereused (Zhang & Datta, 2006). The authors finally retainedin their simulations lower permeability for the ‘‘crust’’(the region of interest was not defined) than for the crumb.The effect of low permeability of the surface layers on thepressure build-up was recently evidenced by Grenier, LeRay, and Lucas (2009).

It is well known that the permeability of a porous medium isaffected by the pore fraction and structure. Very little informa-tion is available in the literature for the dough/crumb, and someof it is controversial. Baker (1939) related the gas flow to cellsize in commercial white bread, but no permeability values areavailable in this study. Dough permeability was directly relatedto porosity, whatever the water (15e50%wb) and fat (2e8%)contents (Goedeken, 1993). Based on this study, the perme-ability of typical crumb (75% of porosity) was estimated ataround 2.2� 10�11 m2, and of typical crust (40% of porosity)around 10�12 m2. These orders of magnitude were used forcrumb (10�11 m2) and crust (2.5� 10�12 m2) in the bakingmodel developed by Zhang & Datta (2006). All other bakingmodels considered uniform permeability through the product(Lostie et al., 2002a; Lucas et al., 2009), which makes theirsimulations of CO2 release and oven-rise questionable.

Secondary effect on the gas released (throughpressure build-up)

Zhang et al. (2007) monitored the CO2 released duringbaking and observed an induction period followed by a sud-den, linear increase in CO2 release, ending with a plateauzone. They also found that the earlier the oven-rise wasstopped, the shorter the induction period. Similar trendscould be reproduced by a mathematical model of baking,which showed in addition that film rupture was necessarybut not sufficient to demonstrate CO2 release and that itmust be accompanied by a pressure build-up similar tothat stemming from an early cessation of the oven-rise.Similarly, Lucas, Le Ray, Peu, Wagner, and Picard (2007)found lower CO2 release for lower oven air temperatures,which delayed the formation of the crust.

CO2 was considered to be the main gas responsible foroven-rise (Bloksma, 1990). Leakage of CO2 obviouslyimplies a decreasing amount of gas in cells and presumablya lower potential for cell growth. Interpreting the CO2

release is however a little more complicated since it alsoreveals a build-up in internal pressure, which is the drivingforce for cell growth provided that the dough films are still

deformable. The net balance between these two antagonis-tic mechanisms has not been assessed to date. CO2 is alsoregarded as a vector of transport of the aromatic com-pounds (Eliasson & Larsson, 1993). For all these reasons,the mechanisms governing CO2 release warrant further in-vestigation for a thorough understanding and a clear identi-fication of the key factors.

A barrier to heat and water transportWahlby et al. (2002) compared the WL of buns reheated

with and without crust. Whatever the heating time, WLwithout crust was three times greater than with crust.This was attributed to the differences in crust and crumbporosity. Breads baked in an impingement or hybrid oven(forced convection combined with microwave) lose lesswater than conventionally baked breads (Patel, Waniska,& Seetharaman, 2005). This was attributed to early crustformation in the former case. Although this effect on WLis commonly accepted (Eliasson & Larsson, 1993), the un-derlying mechanisms have only recently been studied withthe aid of models (Lostie et al., 2004; Zhang et al., 2008).

As reported in section Water content, HT governed theprogress of the vaporisation front in the surface layers andthus WL during baking. As the front went deeper, the resis-tance to HT increased and slowed down the WL. Similarly,high porosity in the surface layers increased this resistance(Vanin et al., 2009) since the same amount of water was dis-tributed over a thicker area. Changes in thermal conductivitycould also affect the resistance to HT in the surface layersand in turn the WL (Vanin et al., 2009). Thermal conductiv-ity of the dough/crumb decreased with decreasing watercontent and increasing porosity (Jury et al., 2007; Rask,1989; Zanoni, Peri, & Gianotti, 1995b). However, conductiv-ity measured on crust and crumb samples has been found tobe very close (Jury et al., 2007), meaning that the effects ofporosity and water content may compensate for each other.

Water transport to the core by the evaporationeconden-sationediffusion mechanism also contributed to the dehy-dration of superficial layers, but to a lower extent; it wastheoretically favoured by higher porosity in the crumb(Zhang et al., 2008).

It must be remembered here that the porosity profilewhich affects the crust thickening through the mechanismsdescribed above is itself greatly affected by fairly earlysetting of the crust (see section Loss of extensibility, andalso Fig. 4a-c). A thorough understanding of these stronglyinterlinked mechanisms is necessary for fine control ofcrust formation, WL and local density together (Vaninet al., 2009).

ConclusionCan a definition of the ‘‘crust’’ region be proposed?

An accurate definition of the spatial domain comprisingthe crust remains necessary for any further study. Changesin its properties are believed to be gradual, making it diffi-cult to define a boundary with the crumb beneath.


A crust is commonly referred to as a dry, hard, dense,coloured zone, following exposure to high temperatures:‘‘. a hard, vitreous surface layer formed of collapsedcrumb pore walls’’ (Eliasson & Larsson, 1993); ‘‘.adenser, darker surface parts bread, near to its surface’’ (Jef-ferson et al., 2006). It must be remembered that low watercontent and high temperature are determinants (but not theonly) of many of the other features (colour, porosity, molec-ular structures, hardness.).

Only a few experimental studies gave their criterion sep-arating the crust from the crumb. The crust sample was oftenunique, distinguishable from the crumb sample(s) by itsdarker colour (visual criterion, e.g. Lind & Rask, 1991) orits more cohesive, harder structure (mechanical propertye.g. Westerlund et al., 1989). Many factors, including lowerwater content, higher density, smaller and more elongatedcells, vitreous transition, etc. may contribute together orseparately to a distinct mechanical behaviour. Mechanicalseparation may also originate from constraints created dur-ing cooling (e.g. Le Bail, Monteau, Lucas, Chargelegue, &Reverdy, 2005), in the case of part-baked breads Zanoniet al. (1993) separated the crust by quick freezing whichcaused rupture between the ‘‘crust’’ and ‘‘crumb’’. All thismakes the mechanical criterion not very reliable whencomparing results from various studies. The same appliesto colour, with the effect of small carbohydrates.

Surprisingly, despite these raw criteria, the propertiesmeasured on crust samples are quite consistent betweenstudies (Fig. 1). Likewise, thickness of the crust has beenreported to increase linearly with baking time (Wiggins,1998; Zanoni et al., 1993). However, the number of studiesis not sufficient to conclude at this stage.

From a theoretical point-of-view, the crust boundary canbe refined by using a threshold value applied to a criterionsuch as density or water content. Jefferson et al. (2006)characterised crust thickness as the distance from the sur-face where the final density was the same as the originaldough density and Zhang et al. (2008) from the top surfaceto the point where the water flowed most rapidly, i.e. the va-porisation front. A criterion with physical relevance to thethreshold should represent a sharp change between the sur-face and the core, as water content or temperature do(Fig. 4e, f). Porosity presents either no sharp change(Fig. 4a) or many sharp changes (Fig. 4c) between the sur-face and core, which makes this criterion hardly reliable.Corresponding experimental profiles of water content andtemperature are few and thus it is a definition that couldnot be applied at the experimental level to date. All thissupports the need to develop a non-invasive, dynamic tech-nique to study the specific behaviour of surface layers, withthe special requirement of high spatial resolution.

Need for further researchSuperficial bread layers differ from core layers of the

crumb mainly in their dynamics during baking. Lack ofpublished data and information on prevailing mechanisms

in the superficial layers of bread have prevented modellingstudies from considering crust setting and its effects onheat, mass transport and cell growth in the crumb, as wellas the full understanding and control of baking.

On one hand, this review has shown that the knowledgeacquired regarding the crumb cannot be extrapolated tosuperficial layers with confidence, especially because themechanisms are highly interrelated, often working antago-nistically and because they are highly sensitive to tempera-ture and water content. The identification of prevailingmechanisms thus remains an issue for future studies.

On the other hand, we have attempted to fill the gap inthe published data through the different enclosed charts.We have also shown that methods of measurement for mon-itoring the dynamics (temperature, water content) in thesuperficial layers present limitations in terms of conve-nience and accuracy, and there is a real need to develop,or sometimes simply implement, non-intrusive, continuoustechniques of measurement. The dynamics impact on themicrostructures and hence the main properties of the crust.Understanding of the underlying reactions has beeninferred from measurements performed on samples ofdifferent water content, but very few studies have relatedheating to drying. There is therefore a real challenge toreproduce these dynamics with miniaturisation inside theinstrument of analysis (rheometer or X-ray synchrotronfor instance). Aguilera (2005) recently emphasized the pos-sibilities offered by the recent advances in imaging todynamically study of microstructures during processing.All these developments will contribute to providing infor-mation of higher spatial resolution and to a clearer defini-tion of the crust region, a prerequisite for further studiesin this area.

AcknowledgmentsThis study was carried out with financial support from

the Council of Brittany (France) and the European Commu-nity Commission (FP6, Thematic Area ‘‘Food quality andsafety’’, FOOD-2006-36302 EU-FRESHBAKE). It doesnot necessarily reflect the Commission’s views and in noway anticipates the Commission’s future policy in this area.

References

Aguilera, J. M. (2005). Why food microstructure? Journal FoodEngineering, 67(1e2), 3e11.

Ahrne, L., Andersson, C. G., Floberg, P., Rosen, J., & Lingnert, H.(2007). Effect of crust temperature and water content on acrylam-ide formation during baking of white bread: steam and fallingtemperature baking. Lwt-Food Science and Technology, 40(10),1708e1715.

Attenburrow, G. E., Goodband, R. M., Taylor, L. J., & Lillford, P. J.(1989). Structure, mechanics and texture of a food sponge. Journalof Cereal Science, 9(1), 61e70.

Babin, P., Della Valle, G., Chiron, H., Cloetens, P., Hoszowska, J.,Pernot, P., et al. (2006). Fast X-ray tomography analysis of bubblegrowth and foam setting during breadmaking. Journal of CerealScience, 43(3), 393e397.


Baik, O. D., Marcotte, M., & Castaigne, F. (2000). Cake baking intunnel type multi-zone industrial ovens e Part I. Characterizationof baking conditions. Food Research International, 33(7),587e598.

Baker, J. C. (1939). The permeability of bread by air. Cereal Chemistry,16, 730.

Bassal, A., Vasseur, J., & Lebert, A. (1993). Measurement of wateractivity above 100 �C. Journal of Food Science, 58(2), 449e452.

Biliaderis, C. G., Maurice, T. J., & Vose, J. R. (1980). Starch gelatini-sation phenomena studied by differential scanning calorimetry.Journal of Food Science, 45, 1669e1674.

Bloksma, A. H. (1980). Effect of heating rate on viscosity of wheat flourdoughs. Journal of Texture Studies, 10(3), 261e269.

Bloksma, A. H. (1990). Rheology of the breadmaking process. CerealFoods World, 35(2). 228e236 and 959e960.

Brathen, E., & Knutsen, S. H. (2005). Effect of temperature and time onthe formation of acrylamide in starch-based and cereal modelsystems, flat breads and bread. Food Chemistry, 92(4), 693e700.

Broyart, B., Trystram, G., & Duquenoy, A. (1998). Predicting colourkinetics during cracker baking. Journal of Food Engineering, 35(3),351e368.

Burt, D. J., & Russell, P. L. (1983). Gelatinization of low water contentwheat starch e water mixtures (A combined study by differentialscanning calorimetry and light microscopy). Starch/Starke, 35(10),354e360.

Cauvain, S. P. (1998). Bread e the product. In S. P. Cauvain, &L. S. Young (Eds.), Technology of breadmaking (pp. 1e17). London,UK: Blackie Academic & Professional.

Champenois, Y., Colonna, P., Buleon, A., Valle, G. D., & Renault, A.(1995). Starch gelatinisation and gelation in white pan bread.Sciences des Aliments, 15, 593e614.

Charissou, A., Ait-Ameur, L., & Birlouez-Aragon, I. (2007). Kinetics offormation of three indicators of the Maillard reaction in modelcookies: influence of baking temperature and type of sugar. Journalof Agricultural and Food Chemistry, 55(11), 4532e4539.

Chevallier, S., Colonna, P., & Lourdin, D. (2000). Contribution of majoringredients during baking of biscuit dough systems. Journal ofCereal Science, 31(3), 241e252.

Chevallier, S., Della Valle, G., Colonna, P., Broyart, B., & Trystram, G.(2002). Structural and chemical modifications of short dough dur-ing baking. Journal of Cereal Science, 35(1), 1e10.

Claus, A., Carle, R., & Schieber, A. (2008). Acrylamide in cerealproducts: a review. Journal of Cereal Science, 47(2), 118e133.

Collar, C., Bollain, C., & Rosell, C. M. (2007). Rheological behaviourof formulated bread doughs during mixing and heating. FoodScience and Technology International, 13(2), 99e107.

Cuq, B., Abecassis, J., & Guilbert, S. (2003). State diagrams to helpdescribe wheat bread processing. International Journal of FoodScience and Technology, 38(7), 759e766.

Czuchajowska, Z., Pomeranz, Y., & Jeffers, H. C. (1989). Water activityand moisture content of dough and bread. Cereal Chemistry, 66(2),128e132.

Datta, A. K., Sahin, S., Sumnu, G., & Keskin, S. O. (2006). Porousmedia characterisation of breads baked using novel heating modes.Journal of Food Engineering, 79, 106e116.

Dobraszczyk, B. J., & Morgenstern, M. P. (2003). Rheology and thebreakmaking process. Journal of Cereal Science, 38, 229e245.

Dobraszczyk, B. J., & Salmanowicz, B. P. (2008). Comparison of pre-dictions of baking volume using large deformation rheologicalproperties. Journal of Cereal Science, 47(2), 292e301.

Dogan, I. S. (2002). Dynamic rheological properties of dough as af-fected by amylases from various sources. Nahrung-Food, 46(6),399e403.

Donovan, J. W. (1979). Phase transitions of the starch-water system.Bio-polymers, 18, 263e275.

Eliasson, A. C. (1980). Effect of water content on the gelatinization ofwheat starch. Starch/Starke, 32(8), 270e272.

Eliasson, A. C., & Larsson, K. (1993). Cereals in breadmaking. Amolecular colloidal approach. New York: Marcel Dekker Inc..

Fan, J., Mitchell, J. R., & Blanshard, J. M. V. (1999). A model for theoven rise of dough during baking. Journal of Food Engineering, 41,69e77.

Fessas, D., & Schiraldi, A. (2000). Starch gelatinization kinetics inbread dough e DSC investigations on ‘simulated’ baking pro-cesses. Journal of Thermal Analysis and Calorimetry, 61(2),411e423.

Glatzel, H., & Rettenmaier, G. (1962). Nutritional comparison of breadproducts. IV. Postprandial hyperglycemia. Nutritio et Dieta: Euro-pean Review of Nutrition and Dietetics, 4, 283e296, (in German).

Goedeken, D. L. (1993). Permeability measurements of porous foodmaterials. Journal of Food Science, 58(6), 1329e1331.

Grenier, D., Le Ray, D., & Lucas, T. (2009). Combining local pressureand temperature measurements during bread baking: insights intocrust properties and alveolar structure of crumb. Submitted toJournal of Cereal Science.

Hayman, D. A. (1998). Effect of pressure (crust formation) on breadcrumb grain development. Cereal Chemistry, 75(5), 581e584.

Huang, V. T., Haynes, L., Levine, H., & Slade, L. (1996). Glass transi-tions in starch, gluten and bread as measured dielectric spectros-copy and TMA methods. Journal of Thermal Analysis, 47,1289e1298.

Jefferson, D. R., Lacey, A. A., & Sadd, P. A. (2006). Understanding crustformation during baking. Journal of Food Engineering, 75(4),515e521.

Jenkins, P. J., & Donald, A. M. (1998). Gelatinisation of starch:a combined SAXS/WAXS/DSC and SANS study. CarbohydrateResearch, 308, 133e147.

Jury, V., Monteau, J. Y., Comiti, J., & Le-Bail, A. (2007). Determinationand prediction of thermal conductivity of frozen part baked breadduring thawing and baking. Food Research International, 40(7),874e882.

Konings, E. J. M., Ashby, P., Hamlet, C. G., & Thompson, G. A. K.(2007). Acrylamide in cereal and cereal products: a review onprogress in level reduction. Food Additives and Contaminants, 24,47e59.

Le Bail, A., Monteau, J. Y., Lucas, T., Chargelegue, A., & Reverdy, Y.(2005). Impact of selected process parameters on crust flaking offrozen part-baked bread. Journal of Food Engineering, 69(4),503e509.

Le Meste, M., Huang, V. T., Panama, J., Anderson, G., & Lentz, R.(1992). Glass-transition of bread. Cereal Foods World, 37(3),264e267.

Leroy, V., Fan, Y., Strybulevych, A. L., Belido, G. C., Page, J. H., &Scanlon, M. G. (2008). Investigating the bubble size distribution indough using ultrasound. In G. M. Campbell, M. G. Scanlon, &D. L. Pyle (Eds.), Bubbles in food 2 (pp. 51e60). St Paul: EaganPress.

Lind, I., & Rask, C. (1991). Sorption isotherms of mixed minced meat,dough and bread crust. Journal of Food Engineering, 14, 303e315.

Lostie, M., Peczalski, R., & Andrieu, J. (2004). Lumped model forsponge cake baking during the ‘‘crust and crumb’’ period. Journalof Food Engineering, 65(2), 281e286.

Lostie, M., Peczalski, R., Andrieu, J., & Laurent, M. (2002a). Study ofsponge cake batter baking process. II: Modeling and parameterestimation. Journal of Food Engineering, 55(4), 349e357.

Lostie, M., Peczalski, R., Andrieu, J., & Laurent, M. (2002b). Study ofsponge cake batter baking process. I: Experimental data. Journal ofFood Engineering, 51(2), 131e137.

Lucas, T., Le Ray, D., Peu, P., Wagner, M., & Picard, S. (2007). A newmethod for continuous assessment of CO2 released from doughbaked in ventilated ovens. Journal of Food Engineering, 81(1),1e11.

Lucas, T., Wagner, M., Doursat, C., Flick, D., & Trystram, G. (2009).Heat and mass transport and expansion during bread baking.


II. Simulation and experimental verification. Submitted to AIChEJournal.

Luyten, A., Pluter, J. J., & van Vliet, T. (2004). Crispy/crunchy crusts ofcellular solid foods: a literature review with discussion. Journal ofTexture Studies, 35(5), 445e492.

Marston, P. E., & Wannan, T. L. (August 1976). Bread baking: thetransformation from dough to bread. The Bakers Digest, 1976,24e28.

Martins, S. I. F. S., Jongen, W. M. F., & van Boekel, M. A. J. S. (2001). Areview of Maillard reaction in food and implications to kineticmodelling. Trends in Food Science & Technology, 11, 364e373.

Mondal, A., & Datta, A. K. (2008). Bread baking e a review. Journal ofFood Engineering, 86(4), 465e474.

Myers, C. D. (1990). Study of thermodynamics and kinetics of proteinstability by thermal analysis. In V. R. Harwalkar, & C. Y. Ma (Eds.),Thermal analysis of food (pp. 16e50). USA: Elsevier science pub-lishers LTD.

Noel, T. R., Parker, R., Ring, S. G., & Tatham, A. S. (1995). The glass-transition behavior of wheat gluten proteins. International Journalof Biological Macromolecules, 17(2), 81e85.

O’Brien, J., & Morrissey, P. A. (1989). Nutritional and toxicologicalaspects of the Maillard browning reaction in foods. Critical Re-views in Food Science and Nutrition, 28(3), 211e248.

Patel, B. K., & Seetharaman, K. (2006). Effect of heating rate on starchgranule morphology and size. Carbohydrate Polymers, 65(3),381e385.

Patel, B. K., Waniska, R. D., & Seetharaman, K. (2005). Impact ofdifferent baking processes on bread firmness and starch propertiesin breadcrumb. Journal of Cereal Science, 42(2), 173e184.

Poinot, P., Arvisenet, G., Grua-Priol, J., Colas, D., Fillonneau, C., LeBail, A., et al. (2008). Influence of formulation and process on thearomatic profile and physical characteristics of bread. Journal ofCereal Science, 48(3), 686e697.

Primo-Martin, C., van Nieuwenhuijzen, N. H., Hamer, R. J., & vanVliet, T. (2007). Crystallinity changes in wheat starch during thebread-making process: starch crystallinity in the bread crust. Jour-nal of Cereal Science, 45(2), 219e226.

Purlis, E., & Salvadori, V. O. (2007). Bread browning kinetics duringbaking. Journal of Food Engineering, 80(4), 1107e1115.

Rask, C. (1989). Thermal properties of dough and bakery products:a review of published data. Journal of Food Engineering, 9,167e193.

Rouille, J., Della Valle, G., Lefebvre, J., Sliwinski, E., & vanVliet, T.(2005). Shear and extensional properties of bread doughs affectedby their minor components. Journal of Cereal Science, 42(1),45e57.

Sadd, P., & Hamlet, C. (2005). The formation of acrylamide in UKcereal products. Chemistry and Safety of Acrylamide in Food, 561,415e429.

Singh, A. P., & Bhattacharya, M. (2005). Development of dynamicmodulus and cell opening of dough during baking. Journal ofTexture Studies, 36(1), 44e67.

Sopade, P. A., Halley, P. J., & Junming, L. L. (2004). Gelatinisation ofstarch in mixtures of sugars. II. Application of differential scanningcalorimetry. Carbohydrate Polymers, 58(3), 311e321.

Stokes, D. J., & Donald, A. M. (2000). In situ mechanical testing of dryand hydrated breadcrumb in the environmental scanning electronmicroscope (ESEM). Journal of Materials Science, 35(3), 599e607.

Strecker, T. D., Cavalieri, R. P., Zollars, R. L., & Pomeranz, Y. (1995).Polymerization and mechanical degradation kinetics of gluten andglutenin at extruder melt-section temperatures and shear rates.Journal of Food Science, 60(3), 532e537.

Thorvaldsson, K., & Skjoldebrand, C. (1996). Method and instrumentfor measuring local water content inside food. Journal of FoodEngineering, 29, 1e11.

Thorvaldsson, K., & Skjoldebrand, C. (1998). Water diffusion in breadduring baking. Lebensmittel Wissenschaft und Technologie, 31,658e663.

Van Nieuwenhuijzen, N. H., Tromp, R. H., Hamer, R. J., & VanVliet, T. (2007). Oscillatory water sorption test for determiningwater uptake behavior in bread crust. Journal of Agricultural andFood Chemistry, 55(7), 2611e2618.

Vanin, F. M., Grenier, D., Doursat, C., Flick, D., Trystram, G., & Lucas, T.(2009). Water loss and crust formation during bread baking. II.Technological insights from a sensitivity analysis. Submitted toJournal of Food Engineering.

Wagner, M. J., Loubat, M., Sommier, A., Le Ray, D., Collewet, G.,Broyart, B., et al. (2008). MRI study of bread baking: experimentaldevice and MRI signal analysis. International Journal of Food Sci-ence and Technology, 43(6), 1129e1139.

Wagner, M. J., Lucas, T., Le Ray, D., & Trystram, G. (2007). Watertransport in bread during baking. Journal of Food Engineering,78(4), 1167e1173.

Wagner, M., Quellec, S., Trystram, G., & Lucas, T. (2008). MRI eval-uation of local expansion in bread crumb during baking. Journal ofCereal Science, 48(1), 213e223.

Wahlby, U., & Skjoldebrand, C. (2002). Reheating characteristics ofcrust formed on buns, and crust formation. Journal of Food Engi-neering, 53(2), 177e184.

Wang, X., Choi, S. G., & Kerr, W. L. (2004). Effect of gluten content onrecrystallisation kinetics and water mobility in wheat starch gels.Journal of the Science of Food and Agriculture, 84(4), 371e379.

Westerlund, E., Theander, O., & Aman, P. (1989). Effects of baking onprotein and aqueous ethanol-extractable carbohydrate in whitebread fractions. Journal of Cereal Science, 10, 139e147.

Whitworth, M. B., & Alava, J. M. (2004). Non-destructive imaging ofbread and cake structure during baking. In 12th ICC Cereal &Bread Congress, Harrogate, UK.

Wiggins, C. (1998). Proving, baking and cooling. In S. P. Cauvain, &L. S. Young (Eds.), Technology of breadmaking (pp. 120e148).London: Blackie Academic & Professional.

Zanoni, B., & Peri, C. (1993). A study of the bread-baking process. I: Aphenomenological model. Journal of Food Engineering, 19(4),389e398.

Zanoni, B., Peri, C., & Bruno, D. (1995a). Modelling of browningkinetics of bread crust during baking. Lebensmittel Wissenschaft &Technologie, 28(6), 604e609.

Zanoni, B., Peri, C., & Gianotti, R. (1995b). Determination of thethermal diffusivity of bread as a function of porosity. Journal ofFood Engineering, 26(4), 497e510.

Zehentbauer, G., & Grosch, W. (1998a). Crust aroma of baguettes e I.Key odorants of baguettes prepared in two different ways. Journalof Cereal Science, 28(1), 81e92.

Zehentbauer, G., & Grosch, W. (1998b). Crust aroma of baguettes II.Dependence of the concentrations of key odorants on yeast leveland dough processing. Journal of Cereal Science, 28(1), 93e96.

Zhang, J., & Datta, A. K. (2006). Mathematical modeling of breadbaking process. Journal of Food Engineering, 75(1), 78e89.

Zhang, L., Doursat, C., Flick, D., & Lucas, T. (2009). Water loss andcrust formation during bread baking. I. Model validation and roleof local porosity. Submitted to Journal Food Engineering.

Zhang, L., Lucas, T., Doursat, C., Flick, D., & Wagner, M. (2007).Effects of crust constraints on bread expansion and CO2 release.Journal of Food Engineering, 80(4), 1302e1311.

Documents

Crust Formation and Its Role During Bread Baking