Elongational properties and proofing behaviour of wheat flour dough

Accepted Manuscript Elongational properties and proofing behaviour of wheat flour dough A. Turbin-Orger, A. Shehzad, L. Chaunier, H. Chiron, G. Della Valle PII: DOI: Reference:

S0260-8774(15)00338-6 http://dx.doi.org/10.1016/j.jfoodeng.2015.07.029 JFOE 8262

To appear in:

Journal of Food Engineering

Received Date: Revised Date: Accepted Date:

5 December 2014 17 July 2015 21 July 2015

Please cite this article as: Turbin-Orger, A., Shehzad, A., Chaunier, L., Chiron, H., Della Valle, G., Elongational properties and proofing behaviour of wheat flour dough, Journal of Food Engineering (2015), doi: http://dx.doi.org/ 10.1016/j.jfoodeng.2015.07.029

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Elongational properties and proofing behaviour of wheat flour dough A. Turbin-Orger 1, A. Shehzad1,2, L. Chaunier 1, H.Chiron1, G. Della Valle 1* 1

INRA, UR 1268 Biopolymères Interactions Assemblages (BIA), 44 316 Nantes, France

2

Present adress : National Institute of Food Science and Technology, University of

Agriculture, Faisalabad-38040, Pakistan *

Corresponding author:

Mail address: [email protected] Tel: 33(0)2 40 67 50 00

Abstract Two series of dough pieces were prepared: (a) at constant composition and different mixing conditions, and (b) by modifying only composition. For both, elongational properties were measured by lubricated squeezing flow (LSF) and proofing kinetics determined by 2D imaging. Elongational viscosity followed a power law in the strain interval [0.1, 1.25] and varied from 140 to 1400 kPa.s, for Hencky strain and strain rate values of 1 and 10-3s-1, respectively. Strain hardening index varied little, in the interval [1.2, 2]. Positive correlation between consistency K and flow index n (series a) suggested that mixing conditions modify gluten network crosslinking, whereas negative correlation between K and n (series b) showed that liquid fraction plasticized the network. Porosity kinetics were found to be mainly governed by gas production factors, rather than dough rheology. Finally, elongational viscosity contributed to limit stability loss during fermentation, which could be attributed to the resistance it imparted to bubbles coalescence.

Keywords : elongational viscosity; fermentation; gluten; porosity; stability.

-1-

Nomenclature

a

Approached value of the overall porosity increase of dough during proofing (-)

a'

Initial value of stability during dough proofing (-)

(a'-c') Overall loss of dough stability during proofing (-) b

Maximum volume expansion rate during dough proofing (min-1)

b'

Starting time of the stationary stability phase during dough proofing (min)

c

Time of the porosity inflection point during dough proofing (min)

c'

Asymptotic value of stability (t→+∞) during dough proofing (-)

d

Porosity value such as (a+d) = P(t→+∞), with d<
e

Neper number (≈ 2.72)

Es

Specific mechanical energy during mixing (kJ/kg)

E’max, E’min εb •

Maximum and minimum values, respectively, of dough storage modulus

Biaxial Hencky strain during LSF testing (-)

εb

Biaxial strain rate during LSF testing (s-1)

ηE

Bi-extensional, or elongational, viscosity determined by LSF (Pa.s)

H

Height of the sample during dough proofing (mm)

K

Consistency index determined for εb =1 (Pa.sn)

Φ vl

Dough volumic fraction of liquid (-)

Lmax

Maximum width of the sample during dough proofing (mm)

LSF

Lubricated Squeezing Flow

n

Flow index computed for εb =1 (-)

v

Compression speed for LSF testing (5, 10 or 100 mm/min)

P(t)

Evolution of porosity during proofing, modelled by a Gompertz function (-)

SHI

Strain Hardening Index determined for ε b =10-2s-1 (Pa)

σ

Extensional, or elongational, stress (Pa)

S(t)

Evolution of the dough shape ratio during proofing, fitted by an exponential decay (-)

Td

Temperature of dough at the end of mixing (°C)

V

Volume sample during dough proofing (mL)

•

-2-

1. Introduction Proofing is an important step of breadmaking, during which dough expands and bread cellular structure is developed. This operation can be assessed at macroscopic scale, by 2D imaging, to measure dough volume evolution, or porosity kinetics, which can be fitted by a Gompertz model (Romano et al. 2007). Shehzad et al. (2010) completed this approach, in the case of free standing dough, by fitting the dough shape-ratio evolution, so to estimate dough stability, by a simple exponential decay model. This basic imaging principle has been widely used to model the dough expansion at the macroscopic scale, under various conditions of proofing temperature, yeast content and leavening agents (Ktenioudaki et al., 2009; Bellido et al., 2009; Soleimani Pour-Daman et al., 2011). Recent developments using structured light might improve the method, by bringing directly 3D information on dough expansion during fermentation (Ivorra et al., 2014). Several mechanistic models of dough expansion during fermentation have been developed based on the growth of a gas bubble in a polymeric matrix, also called the Cell Model (Amon & Denson, 1984). In a first approach, this model can be envisioned as bubble growth resulting from a balance between gas pressure and viscous resistance. It has been completed by including mass transfer phenomena, visco-elastic, surface tension and coalescence effects, and bubble size distribution (de Cindio and Correra, 1995; Chiotellis & Campbell, 2003; Hailemariam et al., 2007; Bikard et al., 2008). However, these models can become very complex and sometimes lack of knowledge about material (dough) properties, rheological for instance, for the models inputs. More recently, by applying, to various dough compositions, the cell model with gas production dynamics for a single bubble proposed by Shah et al. (1998), Cordoba (2010) has identified a value of elongational viscosity matching with experimental value determined experimentally elsewhere. The poor sensitivity of the model to the elongational properties has also been underlined, and later confirmed by Stanke et al. (2014) for different yeast contents and proofing temperatures. These last results might appear surprising since viscosity, and particularly elongational, or extensional, viscosity is the property by which the dough resists to the growth of gas bubbles; moreover the importance of rheological properties, namely strain hardening, in the ability of dough to retain gas and expand has been recognized for long (van Vliet et al., 1992; Dobraszczyk & Roberts, 1994). Their role in the growth of bubbles during dough proofing has also been evidenced thanks to other imaging methods, like NMR and X-ray tomography (Rouille et al., 2005; Trinh et al., 2013). More recently, setting a balance of gas pressure during dough proofing, Grenier et al. (2010) have underlined the significance of elongational viscosity, depending on dough moisture. -3-

Dough elongational properties have currently been measured as they have been considered as an important tool for the assessment of breadmaking performances (van Vliet et al., 2008; Ktenioudaki et al., 2010). Their measurement is not easy because the solid-like viscoelastic behavior of dough shows transient, i.e. not steady-state, elongational viscosity. It is not a function of extension rate only, but also a function of the history of extension. This is well pointed out by Tanner et al. (2018) who proposed a model including a damage function of the strain. Basing on measurements, Launay and Michon (2008) have proposed a simpler model of elongational stress as function of strain and strain rate. Regarding measurements, although the bubble inflation technique has often been used (Tanner et al., 2008), one of the most current test for determining these properties is the uniaxial compression, or squeezing flow, under lubricated conditions (LSF), as first proposed by Chatraei et al. (1981). This method has been popularized by the research group at Wageningen University and the rheological basis well overviewed by Launay and Michon (2008). It has often been used at constant compression speed, rather than at constant strain rate as basic rheology principles would recommend. Nonetheless, Kouassi-Koffi et al. (2010) have shown the relevance of both approaches at least for comparing various wheat flour dough based products, and thus predicting their breadmaking performances. Recently, Mohamed et al. (2013) have evidenced the role of starch-gluten adhesion in dough by modelling its behaviour using LSF tests. However, in spite of the significance of the method, elongational properties of dough have never been measured as input for the mechanistic models of proofing mentioned before, nor directly determined for comparison with results from experiments of macroscopic follow-up of dough proofing. So, the aim of this work is to determine the elongational properties of wheat flour dough in a large range of composition and processing conditions, together with examining their possible links with their behaviour during proofing. In this purpose, two series of wheat flour dough pieces have been prepared: (a) at constant composition and different mixing conditions, thus mainly acting on gluten network, and (b) by changing dough formulation. For both series, elongational properties have been measured by LSF and their proofing behavior assessed according to the usual 2D imaging method.

-4-

2. Materials and methods Two series of experiments were carried out: a first one keeping same dough composition but changing the mixing conditions precisely assessed by Shehzad et al.(2012), and a second one with various compositions whilst keeping same mixing conditions, as described in detail by Turbin-Orger et al. (2012). It has been checked that all conditions further lead to acceptable breads, i.e. exhibiting an aerated crumb structure.

2.1. Dough processing 2.1.1 Raw materials Wheat flour T55 (Minoterie Giraudineau, F44-Saint Colomban), contained 11 % protein and 14% water; fresh yeast (Springer, TransGourmet, F44-Nantes) was stored at 4°C and renewed every 7 days after first use; sugar (Beghin-Say, Tereos, F-59 Lille) and fat (rapeseed oil, Rustica, Leclerc, F-94 Ivry-Sur-Seine) were purchased from local stores. 2.2.2 Mixing The dough was obtained by mixing flour with other ingredients in the same standard recipe, i.e by adding 62 % water, 2.5% fresh yeast, 2 % salt and 40 ppm ascorbic acid to flour basis, with a spiral mixer (Diosna SP12, Osnabrück, Germany). The mechanical power supplied to the dough and dough temperature were measured continuously. Detailed descriptions of the mixer and procedure are given by Shehzad et al. (2012). For the first series (a), ingredients were first mixed at low speed (100 rpm) for 240 s, before texturing under different mixing processes, with speed varying between 80 and 320 rpm, and durations from 180 to 660 s leading to 15 different samples (Table 1a). Under these conditions, specific mechanical energy Es, varied between 5 and 87 kJ/kg; because of viscous dissipation, it was correlated to the increase of dough temperature which led to temperature values at the end of mixing Td between 14 and 36°C (Shehzad et al., 2012). For the second series (b), the same first mixing step was applied, and texturing was performed at 200 rpm during 8 min. Only 2% fresh yeast, instead of 2.5%, were added, in order to reduce the gas production rate because we wanted to better observe the mechanisms of bubble growth by X-ray tomography in a connected study (Turbin-Orger et al., 2012). The dough temperature at the end of mixing was 25 +/- 1°C. The compositions have been selected in the following range (g/100g wheat flour) with 2% fresh yeast, 2 % salt and 40 ppm ascorbic acid: 60 ≤ Water ≤ 66, 0 ≤ Sugar ≤ 15, 0 ≤ Fat ≤ 10 (Table 1b). Assuming that sugar was solubilized in the water, these formula contained different volumic fractions of liquid. It is defined as the sum of volumic fraction of each component, apart from flour, over the total dough volume : -5-

Φ (mi/ρi) / (mT/ρT) vl = Σ

(1)

where mi was the mass fraction of each ingredients i (without flour), ρi their density, mT the mass fraction of dough, ρT its density, taking same values of density as Shehzad et al. (2010).

2.2. Dough rheology Rheological measurements were performed at large bi-extensional deformations by Lubrificated Squeezing Flow test (LSF). At the end of mixing, unyeasted dough samples (≈ 5g) were loaded in Teflon cylinders (h0=14 mm, Ø= 20 mm), lubricated with paraffin oil (110-230 mPa.s at 20°C) and kept at room temperature for 30 min. Sample dimensions were chosen to prevent the softer dough, containing a high quantity of liquid, from collapsing. The homogeneous samples were then removed from the cylinders and placed between two parallel plates of Teflon (Ø= 20 mm) lubricated with paraffin oil. The upper plate was attached with movable crosshead of a traction/compression machine (Instron type #1122) equipped with force sensor in the range [0; 100 N] (Instron Corporation, Canton, MA, USA). The cylindrical samples were compressed until a final height of 1 mm, at a constant speed v (5, 10 and 100 mm/min). Measurements were repeated 4 times at each compression speed, and 2 replicates were carried for all formulas. The force applied to samples F was recorded as a function of displacement (h0-h(t)), h(t) being the height of the sample at time t. Data were processed according to procedures described by Launay & Michon (2008) and van Vliet (2008). Stress σ, defined as the ratio of measured force F to plates area was plotted against biaxial deformation εb :

εb = -1/2 ln (h(t)/h0)

(2) •

For given deformations εb [0.1, 0.25, 0.5, 0.75, 1.0, 1.25], biaxial deformation rate ε b was calculated: •

εb = −

v 2h(t )

(3)

Then stress σwas plotted against biaxial deformation rate. The average slope of those curves defined n, the flow behaviour index. From the same plot, different values of Ln(σ) as function •

of εb may be derived for a given value of ε b . The slope of the curve of ln(σ) versus εb,

-6-

⎛ ∂ ln σ ⎜⎜ ⎝ ∂ε b

• ⎞ ⎟⎟ defines the strain hardening index SHI. Bi-extensional viscosity ηE = σ/ ε b , was ⎠ ε• b = cst •

calculated for each deformation εb and plotted against biaxial deformation rate ε b .

2.3. Dough proofing measurements Images of a rounded dough piece (25 g), during proofing in a controlled ambience (T=27°C, RH=75%), were acquired every 5 min with a digital camera for 240 min, according to the procedure described in detail by Shehzad et al. (2010). This procedure encompasses image analysis to determine the volume sample V, the height H and the maximum width Lmax of the dough sample, at a given time, assuming a cylindrical symmetry. The resulting volume V was converted into porosity, P(t), and the dough shape ratio was defined by S(t)=H/Lmax. The porosity curves can be modelled using a Gompertz function, as proposed by Romano et al. (2007): Pt  a. exp  exp 

 

   

(4)

where a is an approximation of the final porosity increase from the initial value, b is the maximum volume expansion growth rate, i.e., the slope at inflection point, c is the time for inflection point, or characteristic porosity time, and d is such as (a+d) = P(t→+ ∞) with d<
St  a c  . exp ⁄  c

(5)

where a' and c' being the stability at t=0 and t→+ ∞, respectively, (a'-c') is the overall loss of dough stability, and b' is the starting time of the stationary phase. To avoid any bias due to hand rounding of the dough prior to measurement, all curves were homothetically shifted to the same value of a', here 0.6. Measurements were triplicated.

-7-

3. Results & discussion 3.1 Elongational properties 3.1.1 Results of LSF tests For both series of samples, stress-strain curves for bi-extensional flow were obtained by LSF measurements (Fig.1a); the variations of bi-extensional viscosity, ηE, with strain rate, ε , could be represented for different strain values (Fig.1b), and fitted by a power law equation:

ηΕ = K. ( )n-1

for constant strain εb

(6)

where K (Pa.sn) is the consistency index and n (dimensionless) the flow behavior index, with n <1 reflecting a strain-thinning behavior of the dough. The numerical values of these indices computed at (εb =1) for viscosity, and at ( =10-2 s-1) for SHI were reported in Tables 1.a and 1.b. K values varied from 5 to 18.6 kPa.sn (#a13 and a3) and they varied in the same range when changing dough composition (series b); conversely, values of n for series a varied from 0.3 (#5, 7, 13 and 15) to 0.49 (#3), whereas values from series b where slightly higher, in the interval [0.4, 0.6] for # 5 and 7, respectively. Compared to consistency index, the strain hardening index SHI, varied in a narrower interval [1.2, 2], and [1.6, 1.9], for series a and b, respectively. Differences between repeated experiments led to an uncertainty on these values lower than 15%. These values were in the range of those encountered in the literature for different flours and processing conditions (Launay & Michon, 2008; van Vliet, 2008; Ktenioudaki et al., 2010; Kouassi-Koffi et al. 2010). Their variations showed a different trend according to the series: n increased with K, for dough prepared under different mixing conditions (series a), whereas n decreased when K increased when changing ingredients, series b (Fig.2). Assuming that elongational properties of dough are mainly due to its structuration by gluten network, these variations may reflect a different effect of the variable modified, mixing or composition, on gluten network. Indeed, the first trend reflects an increased stiffness of the dough, which might be attributed to a stronger crosslinking of the network; conversely, the second trend may be attributed to a larger mobility of the biopolymer chains, due to their plasticization by smaller molecules. 3.1.2 Main variables influencing elongational properties No straightforward relation appeared between elongational viscosity and mixing variable, like Es, Td. So, to ascertain the preceding interpretations, we have represented, for series (a), the variations of the consistency index with the ratio of maximum to minimum values of storage modulus measured by Dynamic Thermomechanical Analysis (DMA), E’max -8-

/E’min. This ratio has been considered as a possible indicator of gluten network structuration, or cross-linking (Angioloni and Dalla Rossa, 2005): the lower the value of the ratio, the less efficient was heat to promote thermal aggregation of proteins, suggesting that the network was already cross-linked before heating. These values have been reported by Shehzad et al. (2012), for the various mixing conditions. The negative correlation (Fig.3a), between measurements performed at low strain (DMA) and large strains (LSF), has never been established before, up to our knowledge, and it adds weight to the proposed interpretation. This interpretation has to be qualified because point #a12 did not belong to the correlation, and we believe that it was due to its atypical mixing conditions, at low temperature and high energy. Except for this particular point, results of rheological measurements on dough of series (a), and especially the negative correlation between consistency and storage modulus ratio, clearly show that the high values of elongational viscosity are concomitant to a high level of gluten network structuration, which is strongly influenced by the mixing conditions. However, for the same data set of mixing conditions and storage moduli ratio, Shehzad et al. (2012) did not find simple relation between Es and Td with dough rheological properties. Indeed, gluten structuring may result from complex interactions between variables such as mixing time, specific mechanical energy amount and strain mode (Chin & Campbell, 2005; Peressini et al., 2008; Kansou et al., 2013), and their determination is not in the scope of this study. Regarding series (b), the plasticization effect is evidenced by the negative correlation between consistency index and liquid volumic fraction of the dough (Fig.3b). In complement, no significant variations of storage modulus was observed when changing the amount of liquid fraction, in agreement with results from Rouillé et al. (2010), which confirmed that the structure of gluten network was not modified. Clearly, the liquid phase increases the mobility of the network. However, all ingredients may not have the same effect because dough #b6 having a high level of fat (10%), but low level of water (55%), displays a high consistency value (17.8 kPa.sn), whereas dough #b7, having low fat content (2%) but high amount of added sugar (15%), has much lower consistency value (14.2 kPa.sn). These results may help to predict the values of elongational properties, from their processing conditions and composition. One way to ascertain the preceding hypotheses would be to apply a full non-linear viscoelastic model like Tanner’s model to the present data (Tanner et al. , 2008), in order to improve the structural, or molecular, interpretation of the results. In this purpose, a damage function should be determined from relaxation tests, which would require complementary rheological experiments.

-9-

3.2 Porosity and stability of fermenting dough For all conditions, porosity kinetics displayed the classical sigmoid shape with an asymptotic value (a+d) ≈ 0.8 (Fig.4a and Tables 1a and b). This asymptotic value is in agreement with the value of relative volume increase (ΔV/V0 ≈ 3) found by Romano et al. (2007) for different yeast contents, ranging from 1 to 3%. Porosity curves were very close to each other except for the dough having a high liquid fraction, with a large sugar content; this result could be owing to the lower absorption of water, and consequently a higher viscosity of the dough, which slowed down the growth of gas bubbles, or to osmotic stress exerted on yeast, hence limiting gas production. Conversely, shape ratio curves all exhibited a continuous decrease in the time interval studied (Fig. 4b). All porosity and shape ratio kinetics could be very well fitted (r2 ≥ 0.99) by eq. (4) and (5), respectively; hence the coefficients characterising the dough during proofing behaviour (a, b, c, b’ and c’) could be determined accurately and their values are reported in Table 1a and b . The cumulated errors due to repetitions and image treatment assumptions led to an overall uncertainty on these values of 10%. Among these coefficients, the more significant variations were obtained for the characteristic time of porosity c, and the loss of stability (0.6-c’). Porosity characteristic time c’ varied from 0.5 to 51 min (series a) and from 34 to 88 min (series b); the case of dough #a15 (c = 0.5 min) is peculiar because dough temperature at the end of mixing was so high (Td=36°C), that maximum fermentation rate could have been reached before the proofing follow up experiment started. To illustrate the importance of dough temperature on the porosity kinetics, values of characteristic time c have been plotted together with Td, for series a (Fig.5a). The negative correlation between both variables, in this range of temperature, confirmed this observation: the higher the temperature of the dough at the end of mixing, the quickest the maximum fermentation, likely because of yeast activation and gas production. Conversely, the large time value observed for #b7 (c = 88min) may be attributed to the large sugar content of this dough (15% on flour basis) and this observation is confirmed by the correlation between c values and sugar content for series b (Fig.5b). This trend may clearly be attributed to the negative role of sugar, likely due to osmotic stress exerted on yeast. Both results confirmed the importance of factors linked to yeast activity on porosity kinetics (Romano et al., 2007). In the range of processing and compositions studied, these factors had a larger influence than those linked to the rheological properties of the dough. Regarding shape ratio kinetics, the loss of stability (0.6-c’) varied between 0.1 and 0.25 for series a, and from 0.13 to 0.25 for series b. Lower values of stability loss reflect the -10-

capacity of dough to maintain its shape; larger values reflect dough spreading during proofing, and these values were reached for lower values of consistency index K, like dough a5 and b8, for instance; this trend is illustrated by the variations of stability loss values with the consistency for both series (Fig.6). Although correlations may be considered only as fair, this graph suggests a trend for elongational properties to limit the loss of stability; the consistency index provided a larger range of variation than the SHI for this purpose. For same consistency values, the larger loss of stability of dough from series b might be attributed to their larger liquid fraction. Clearly other mechanisms than the resistance of dough to bubble growth are contributing to the reduction of stability loss. Recently, basing on observations of fermenting dough by X-ray tomography, we have suggested that shape ratio displayed a plateau and dough finally gained stability likely because gas bubbles were separated by liquid films, preventing them from coalescence (Turbin-Orger et al., 2015). We also proposed that the loss of stability was concomitant to a greater heterogeneity of cellular structure. Hence, the stability of fermented dough was not only due to the elongational properties of the starch gluten matrix but also to the surface tension of these liquid films and to the distribution of gas cell size. These different mechanisms are discussed in the following section.

3.3 About the link between rheological properties and proofing behavior The significance of gas production (or limitation) factors (dough temperature, and sugar content) on porosity kinetics is in line with several preceding findings, considering the modelling of dough expansion during proofing. Various numerical models have been established to describe the growth of a single bubble in a viscous medium (Amon and Denson, 1984); Cordoba (2010) has adapted it to a gas bubble in wheat flour dough and showed that the gas production rate is the most sensitive variable, which was later confirmed by Stanke et al. (2014). The decrease of characteristic time with increasing dough temperature reported here (Fig.5a) is in agreement with the increase of gas production, predicted by such models. Conversely, no results are available to compare the increase of porosity characteristic time obtained by increasing sugar content (Fig.5b); moreover, sugar may have various, and sometimes contradictory, influences, in dough proofing, by providing substrate to yeasts but also absorbing water and slowing gas diffusion, and finally, exerting osmotic stress on yeasts. Both modelling studies also agree that comparatively, changes of viscosity, even in a factor of 10, poorly affect the volume increase kinetics. Indeed, viscosity measured here, for a value of strain rate  =10-3s-1 currently encountered for bubble growth during proofing (Babin et al., 2006), leads to values ranging from 140kPa.s to 1440kPa.s; these values are 100 to 10 times lower than those derived in these models. So, it may be concluded that, in spite of important -11-

variations, either due to mixing, or to compositions changes, elongational properties poorly affect the gain of volume of dough during proofing. Clearly this phenomenon depends more on the amount of gas produced by yeast activity (Romano et al., 2013); however, the rheological properties of dough matrix may favor bubble growth by limiting collapsing. Indeed, at microscopic scale, bubble growth during dough proofing has been studied by high resolution imaging methods, like Ultrasonic probing, Magnetic Resonance Imaging and X-ray tomography (XRT) (Elmehdi et al., 2003; Rouillé et al., 2005; Turbin-Orger et al., 2012, 2015; Trinh et al.., 2013); it has been shown that, after disproportionation, until a characteristic time indicated by the inflection point on the porosity kinetics, bubbles grow freely, mainly due to yeast activity producing CO2; after this characteristic time, bubbles begin to connect each other and possibly coalesce. A good agreement has been found between the values of this characteristic time, determined at microscopic scale by XRT and at macroscopic scale by 2D imaging (Shehzad et al., 2010). Therefore, the mechanisms inferred for bubbles growth and connection may also be considered as responsible for dough expansion and loss of stability during proofing. During both stages, the cellular structure is modified as shown by the changes of distribution of gas cell size and cell wall thickness, which also evidence the increased heterogeneity of the cellular structure during proofing (Chiotellis & Campbell, 2003; Babin et al., 2006; Turbin-Orger et al., 2012). Bubbles coalescence likely contributes to this increased heterogeneity by creating larger bubbles, and, in concomitant way, to the loss of stability. This is illustrated by the difference of cellular structure between fermented dough reported by Turbin-Orger et al. (2012) at large porosity (P≈0.6), for different compositions, here # b8 and b10. The more heterogeneous the cellular structure, the less stable the dough during proofing. Elongational viscosity is the property by which the dough matrix resists to the stretching due to bubbles growth, hence to coalescence; therefore, this mechanism explains the negative trend between consistency and stability loss. The fact that the correlation is only partial suggests that other mechanisms are contributing to dough stability during fermentation, like the surface tension of the liquid films that may separate gas bubbles at larger porosity (Gan et al., 1995); the importance of this mechanism has been clearly evidenced by the addition of surface active components that increase dough volume, without changing its rheological properties (Sroan & MacRitchie, 2009). So it would be interesting to assess the balance between these two mechanisms, by computing a Capillary number that compares these two forces, as suggested by Della Valle et al (2014). The data on the elongational properties determined here will be useful in this purpose.

-12-

4. Conclusion Dough fermentation is an essential step for the creation of bread cellular structure, a step during which gas bubbles grow and become connected, whilst dough expands and may spread and lose stability. Dough elongational properties likely play an important part in these phenomena, but there is a lack of experimental data on these properties. So, we have determined these properties for a large range of dough mixing conditions (series a) and compositions (series b), by lubricated squeezing flow test (LSF). For all dough tested, elongational viscosity followed a power law, which allowed determining consistency and flow indices for a strain value (=1), relevant for proofing step. Strain hardening index values did not vary much, in the interval [1.2, 2], contrary to consistency index which varied in the interval [ 5, 18.6 kPa.sn]. Positive correlation between consistency and flow index reflected a strengthening of gluten network due to modifications of mixing conditions (series a); this interpretation is supported by the variations of consistency index with the ratio of storage moduli; conversely, the negative correlation of consistency with volumic liquid fraction (series b) suggested that the liquid fraction rather plasticized the network without modifying its structure (series b). Dough proofing kinetics were also determined by 2D imaging, according to a method previously validated. Porosity kinetics were found to be mainly governed by gas production factors (dough temperature for series a, and sugar content for series b), rather than dough rheology, which confirmed the results of dough proofing models. Finally, elongational viscosity contributed to reduce the stability loss during fermentation, which could be attributed to the resistance that it imparted to dough matrix against bubbles coalescence. Elongational properties would thus contribute to control the homogeneity of dough cellular structure. This interpretation could be ascertained by studies of bubbles growth in dough by Xray microtomography.

Acknowledgements Part of the results mentioned in this paper are issued from works which have received funding from the European Community's Seventh Framework Programme (FP7/ 2007-2013) under the grant agreement n°FP7-222 654, project : Design and development of REAlistic food Models with well characterized micro- and macro structure and composition (DREAM), and from the French National research agency (ANR-ALIA) through Incalin (Knowledge Integration in Food Processing) and Braise (Energy Efficiency for Sustainable Bakery)

-13-

projects in the ALIA program. Arnaud Turbin-Orger’s PhD Thesis was funded by the Région Pays de Loire.

-14-

References

Amon, M., & Denson, C.D., 1984. A study of the dynamics of foam growth: Analysis of the growth of closely spaced spherical bubbles. Pol. Eng. Sci. 24, 1026-1034. Angioloni, A., & Dalla Rossa, M., 2005. Dough thermo-mechanical properties: influence of sodium chloride, mixing time and equipment. J. Cereal Sci. 41, 327-331. Bikard, J., Coupez, T., Della Valle, G. & Vergnes B., 2008. Simulation of bread making process using a direct 3D numerical method at microscale. Part I: analysis of foaming phase during proofing. J. Food Eng. 85, 259–267 Chatraei, S.H., Macosko, C.W. & Winter, H.H., 1981. Lubricated squeezing flow: a new biaxial extensional rheometer. J. Rheol. 25, 433-443. Chin, N. L., & Campbell, G. M., 2005. Dough aeration and rheology: Part 2. Effects of flour type, mixing speed and total work input on aeration and rheology of bread dough. J. Sci. Food Agri., 85, 2194-2202. Chiotellis, E., & Campbell, G.M., 2003. Proving of bread dough I:Modelling the evolution of the bubble size distribution. Food Bioprod. Proc. 8,194−206 Córdoba, A., 2010. Quantitative fit of a model for proving of bread dough and determination of dough properties. J. Food Eng., 96, 440–448 Della Valle, G., Chiron, H., Cicerelli, L., Kansou K., Katina, K., Ndiaye, A. , Whitworth, M., Poutanen, K., 2014. Basic knowledge models for the design of bread texture. Trends Food Sci. Tech., 36, 5-14. Dobraszczyk, BJ ., Roberts, CA., 1994. Strain hardening and dough gas cell-wall failure in biaxial extension. J Cereal Sci. 20, 265-274. Elmehdi, H.M.,Page, J. H., Scanlon, M.G., 2003. Monitoring dough fermentation using acoustic waves. Trans. IChemE., 81,Part C, 217-223 Gan, Z., Ellis, P.R., & Schofield, J.D. 1995. Mini Review: Gas cell stabilisation and gas retention in wheat bread dough. . J. Cereal Sci. 21, 215-230. Grenier, D., Le Ray, D., Lucas, T., 2010. Measurement of local pressure during proving of bread dough sticks:Contribution of surface tension and dough viscosity to gas pressure in bubbles. . J. Cereal Sci., 52, 373-377 Hailemariam L., Okos M.,Campanella O., 2007. A mathematical model for the isothermal growth of bubbles in wheat dough. J. Food Eng., 82, 466–477 Ivorra E., Verdu Amat, S. , Sanchez, A. J., Barat, J. M. & R. Grau. 2014. Continuous monitoring of bread dough fermentation using a 3D vision Structured Light technique. J. Food Eng., 130, 8–13 -15-

Kansou, K., Chiron, H., Della Valle, G., Ndiaye, A., Roussel, P., Shehzad, A., 2013. Modelling wheat flour dough proofing behaviour: effects of mixing conditions on porosity and stability. Food Bioproc. Tech., 6, 2150-2164. Kouassi-Koffi, J. D., Launay, B., Davidou, S., Kouame L. P., & Michon, C., 2010. Lubricated squeezing flow of thin slabs of wheat flour dough: comparison of results at constant plate speed and constant extension rates. Rheol. Acta, 49, 275-283. Ktenioudaki, A., Butler, F., Gallagher, E., 2010. The effect of different mixing processes on dough extensional rheology and baked attributes. J. Sci. Food Agri.,90, 2098-2104. Ktenioudaki A., Butler F., Gonzales-Barron U., Mc Carthy U., Gallagher E., 2009. Monitoring the dynamic density of wheat dough during fermentation. J. Food Eng., 95, 332–338. Launay, B., & Michon, C., 2008. Biaxial extension of wheat flour dough: lubricated squeezing flow and stress relaxation properties. J. Texture Studies, 39, 496-529. Peressini, D., Peighambardoust, S. H., Hamer, R. J., Sensidoni, A., & van der Goot, A. J. 2008. Effect of shear rate on microstructure and rheological properties of sheared wheat doughs. J. Cereal Sci., 46, 426-438 Romano, A., Toraldo, G., Cavella, S., Masi, P. 2007. Description of leavening of bread dough with mathematical modelling. J. Food Eng.,, 83: 142–148. Romano,A., Cavella, S., Toraldo, G., Masi, P., 2013. 2D structural imaging study of bubble evolution during leavening. Food Res. Int., 50, 324–329 Rouillé, J., Bonny, J-M., Della Valle, G. Devaux, MF, Renou, JP. 2005.. Effect of flour minor components on bubble growth in bread dough during proofing assessed by Magnetic Resonance Imaging. J. Agric. Food Chem. 53, 3986-3994. Rouillé, J., Chiron, H., Colonna, P., Della Valle, G., & Lourdin, D. 2010. Dough/crumb transition during French bread baking. J. Cereal Sci. 52, 161-169. Shah, P., Campbell, G., McKee, S., Rielly, C. 1998. Proving of bread dough Modelling the growth of individual bubbles. Food Bioprod. Proc. 76, 73-79. Shehzad, A., Chiron, H., Della Valle, G., Kansou, K., Ndiaye, A., Réguerre, A. L., 2010. Porosity and stability of bread dough during proofing determined by video image analysis for different compositions and mixing conditions. Food Res. Int., 43, 19992005. Shehzad, A., Chiron, H., Della Valle, G., Lamrini, B., & Lourdin, D. 2012. Energetical and rheological approaches of wheat flour dough mixing with a spiral mixer. J. Food Eng., 110, 60–70.

-16-

Soleimani Pour-Daman, A.R., Jafary, A., Rafiee, Sh., 2011. Monitoring the dynamic density of dough during fermentation using digital imaging method. J. Food Eng., 107, 8–13 Sroan, B.S., MacRitchie, F., 2009. Mechanism of gas cell stabilization in breadmaking. II. The secondary liquid lamellae. J. Cereal Sci., 49, 41-46 Stanke, M., Zettel, V., Schütze, S., Hitzmann, B. 2014. Measurement and mathematical modeling of the relative volume of wheat dough during proofing. J. Food Eng., 131, 58– 64. Tanner, R. I., Dai, S. C., Qi, F.Z., 2008. Bread dough rheology in biaxial and step-shear deformations. Rheol. Acta, 47, 739-749. Trinh, L., Lowe, T., .Campbell, G.M., Withers, P.J., Martin, P.J. 2013. Bread dough aeration dynamics during pressure step-change mixing: Studies by X-raytomography, dough density and population balance modelling. Chem. Eng. Sci., 101, 470-477 Turbin-Orger, A., Boller, E., Chaunier, L., Chiron, H., Della Valle, G., Réguerre, A.-L. 2012. Kinetics of bubble growth in wheat flour dough during proofing studied by computed X-Ray micro-tomography. J. Cereal Sci. 56, 676–683. Turbin-Orger, A., Babin, P., Boller, E., Chaunier, L. Chiron, H., Della Valle, G., Dendievel, R., Réguerre, A.-L., Salvo, L. 2015. Growth and setting of gas bubbles in a viscoelastic matrix imaged by X-ray microtomography: the evolution of cellular structure in fermenting wheat flour dough. Soft Matter, 11, 3373-3384 van Vliet, T., Janssen, AM., Bloksma, AH, Walstra, P. 1992. Strain hardening of dough as a requirement for gas retention. J. Texture Studies 23, 439-460. van Vliet, T. 2008. Strain hardening as an indicator of bread-making performance: A review with discussion. J. Cereal Sci., 48, 1-9.

-17-

Table 1: Processing conditions and main results of proofing behaviour and elongational properties for the two series of dough processed (a) under various mixing conditions for same composition and (b) under same mixing conditions and different compositions. Maximum and minimum values are indicated in bold. Uncertainty is 10% and 15 % on kinetics and rheological coefficients, respectively. (a) Dou gh #

Mixi Mixi ng ng Tim spee e d

Td(° C)

Εs (kJ/ kg)

Porosity b

E’

Stability b

Elongational properties c

max a

/ E’ min

(min

(rpm

)

)

a

b

c

(a+

(0.

(mi

(mi

d)

6-

-1

n)

n )

b’(m K(Pa. in)

n

sn)

S HI

c’)

(P a)

1

7

200

23

24.4

2

7

80

19

5.1

3 4 5 6 7 8 9 10 11 12

7 11 3 4 4 10 10 10 11 7.5

320 200 200 116 284 116 140 140 200 284

28 28 19 19 23 21 20 26 29 22

56 47.4 9.9 7.1 22.9 18 18.3 17.8 49.4 47.8

16

0. 7

0.0 06

30

0.7 8

0.1

19

14.5

0. 41

1. 6

12. 7

0. 53

0.0 07

47

0.7 2

0.1 7

32.5

14.9

0. 45

1. 7

11. 8

0. 59

0.0 07

38

0.7 7

0.0 7

18.6

0. 49

1. 4

15. 5

0. 69

0.0 06

29

0.7 8

0.1 3

24

13.2

0. 39

2. 0

19. 7

0. 58

0.0 08

39

0.7 4

0.2 5

30

5.7

0. 3

1. 2

18. 4

0. 53

0.0 08

46

0.7

0.2 2

35

9.0

0. 31

1. 8

20. 1

0. 57

0.0 07

38

0.7 3

0.1 6

27

9.5

0. 3

1. 4

21. 8

0. 68

0.0 06

38

0.8 1

0.2 0

30

5.8

0. 32

1. 7

22. 1

0. 59

0.0 07

51

0.7 7

0.1 7

23

8.3

0. 36

1. 4

22. 6

0. 72

0.0 08

24

0.7 8

0.1 7

23

8.5

0. 33

1. 5

21. 2

0. 68

0.0 08

20

0.7 3

0.0 8

19

8.2

0. 35

1. 5

10. 3

0. 64

0.0 06

44

0.8

0.1 1

27

9.1

0. 31

1. 6

-18-

23

13 14 15

10 10 10

80 80 320

22 23 36

8.8

33. 7

0. 52

0.0 07

44

0.7 2

0.2 2

35

5.0

0. 3

1. 6

26

0. 55

0.0 08

37

0.7 1

0.2 2

31

6.7

0. 36

1. 3

0. 99

0.0 08

0.5

0.7 9

0.1

24

11.4

0. 3

2

8.7 87.4

17

(b) Doug

Wat

Sug

Fa

h#

er

ar

t

Porosityb

Φ vl

Stabilityb

Elongational propertiesc

a

b

c

(mi

(mi

n-1)

n)

(a+ d)

(0.6 -c’)

b’(mi

K(Pa.s

n)

n

n

)

SH I (Pa )

1

2

3

4

5

6

7

8

9

10

66

66

66

66

60

55

55

66

66

62

0

5

0

5

2

2

15

10

2

0

0

0

5

5

2

10

2

10

2

0

0.58

1.0

0.00

8

7

9

0.59

0.9

0.00

8

6

7

0.60

0.7

0.00

4

4

7

0.61

0.8

0.00

3

2

7

0.58

0.7

0.00

1

9

8

0.59

0.7

0.00

3

5

7

0.59

0.7

0.00

2

6

5

063

0.7

0.00

3

6

7

0.59

0.8

0.00

6

3

7

0.57

0.9

0.00

6

1

8

45

0.7

0.1

26

14.4

8 45

0.77

0.2

0.4

1.7

6 23

10.7

0.5

1.6

1 34

0.8

0.1

23

12.8

6 63

0.74

0.2

0.5

1.9

3 23

11.1

0.5

1.8

6 44

0.78

0.1

26

16.9

9 48

0.74

0.1

0.4

1.8

2 26

17.8

0.4

1.8

25

14.2

0.4

1.6

6 88

0.73

0.2 2

57

0.75

0.2

8 18

8.6

0.6

1.6

16

11.5

0.4

1.9

5 47

0.77

0.2 1

41

0.77

0.1 3

-19-

7 23

17.8

0.4 5

1.9

a

these values have been determined and reported in Shehzad et al. (2012). these coefficients are defined in eqs (4), (5). c these values have been determined from eq(1), (2, (3) and (6). K and n are computed for εb b

•

=1 and SHI for ε b =10-2s-1

Figure captions

Fig.1: Typical results of elongational properties measurements obtained for dough a1 by LSF (a) strain-stress recording at different compression speed v, and (b) elongational viscosity at different strain values εb (0.1,; 0.25,  ; 0.5, ; 0.75, ;1, O; 1.25, *). Straight lines result from linear fitting (r2 > 0.99). Fig. 2: Variations of flow index with consistency index of elongational viscosity for the two series of dough processed at constant composition (, r2=0.68) or under same mixing conditions (, r2=0.78). Fig.3: Variations of consistency index K with (a) ratio of storage moduli E’max/E’min for dough processed at constant composition (series a) and (b) with volumic liquid fraction Φ vl for dough mixed under same conditions (series b). Curves show best fitting (K=403.[ E’max/E’min]-1.28 in (a) and K= 111-163. Φ vl in (b)). It does not take into account point #a12 (O) in Fig.(a). Fig.4: Examples of kinetics of (a) porosity and (b) shape of dough #a4 (O, high Es and Td), a14 (, low Es and Td), b3 (, low Φ vl), b8 (, high Φ vl) during proofing. Fig.5: Variations of porosity characteristic time with (a) dough temperature at the end of mixing (series a, ) and (b) dough sugar content (series b, ) Fig.6: Variations of stability loss during proofing with consistency index for the two series of dough processed at constant composition (series a, ) or under same mixing conditions (series b, ).

-20-

Figure

1 1E+05 (a) V=10 mm/min 1E+04 Stress s (Pa)

V=100 mm/min V=5 mm/min

1E+03

1E+02

1E+01 0,0

0,2

0,4

0,6

0,8

1,0

1,2

1,4

1,6

Biaxial strain eb (-)

1,E+06

Extensional viscosity hE (Pa.s)

(b)

1,E+05

1,E+04

1,E+03 1,E-03

1,E-02

1,E-01

1,E+00

Biaxial strain rate e b (s-1)

Fig.1

dough proofing & viscosity

2 3 4

-1-

Turbin-Orger et al.

Figure

1 0,7 0,6

flow index (n)

0,5 0,4 0,3

0,2 0,1 2

4

6

8

10

12

14

16

18

consistency index K (Pa.s^n)

Fig.2 dough proofing & viscosity Turbin-orger et al. 2 3 4

-1-

20

Figure

1

20,0 (a)

K(kPa.sn)

16,0 12,0 8,0

4,0

r2=0.65

0,0

0

10

20 E'max/E'min

30

40

20

(b)

K(kPa.sn)

16 12 8 r2=0.7

4 0 0,56

Fig.3

0,58 0,6 0,62 Volumic Liquid Fraction Fvl

dough proofing & viscosity

2 3 4

-1-

0,64

Turbin-Orger et al.

Figure

1

0,9 (a) 0,8 0,7 Porosity

0,6 0,5 0,4 0,3 0,2 0,1 0,0 0

30

60

90 120 Proofing time (min)

150

180

0,65 (b)

Shape ratio

0,60

0,55 0,50 0,45 0,40

0,35 0,30

0

15

30

45

60

75

90

Proofing time (min) Fig.4

dough proofing & viscosity

2 3 4

-1-

Turbin-orger et al.

Figure

characterisic porosity time c (min)

characterisic porosity time c (min)

1

70

(a)

60

50 40 30 20 10 0

0

100

10 20 30 40 Dough temperature Td ( C)

(b)

90 80

R² = 0,82

70 60 50 40 30 20 0

Fig.5

R² = 0,81

5 10 15 Sugar content (% flour)

dough proofing & viscosity

2 3

-1-

20

Turbin-orger et al.

Figure

1

0,3 0,25 R² = 0,58

Stability Loss

0,2

0,15 0,1 R² = 0,46

0,05 0 0

Fig.6

5

10 15 Consistency Index K (Pa.sn)

dough proofing & viscosity

2

-1-

20

Turbin-orger et al.

Highlights

* Dough elongational viscosity follows a power law for a large range of formulations * Increase of liquid fraction decreases consistency by gluten network plasticization * Mixing conditions modify dough consistency by gluten network cross linking * Porosity kinetics are governed mainly by gas production rather than dough rheology * Elongational viscosity limits the loss of stability during fermentation

-21-