Relationship Between Physical Properties of Dough and Expansion Ability During Bread- Making

Transcription

1 Food Sci. Technol. Res., +, (,), 3+ 3/,,**0 Relationship Between Physical Properties of Dough and Expansion Ability During Bread- Making Hideki KAWAI +, Fumitake TANAKA +, Hiroshi TAKAHASHI +, Naoto HASHIMOTO, and Hiroaki YAMAUCHI, + Department of Mechanical Systems Engineering, Muroran Institute of Technology,,1 + Mizumoto, Muroran, Hokkaido */* 2/2/, Japan, Department of Upland Agriculture, National Agricultural Research Center for Hokkaido Region (NARCH), Shinsei, Memuro, Hokkaido *2, **1+, Japan Received June 0,,**/; Accepted March,,,**0 We investigated whether the physical properties (rheological properties) of various bread doughs, obtained from a linear Kelvin four-element model of visco-elasticity, could be used to express the expansion ability of the dough during bread-making. The physical properties t * and t + (relaxation and retardation time, respectively) were determined from the creep. The stress relaxation curves derived from the t * values of various doughs, such as Victoria INTA (extra strong flour), Camellia (strong flour), and Hokushin (semi-strong flour), approximated the experimental results well. By analyzing the relationship between each coe$cient in this model and the expansion ability of bread dough (specific loaf volume and gas retention), it was proven that t * and t + were positively and negatively correlated, respectively, to the expansion of bread dough to a statistically significant extent. From these results, it was suggested that dough with more elastic properties shows greater expansion. It is known experientially that adequate dough elasticity is necessary for good expansion, and the results of this study support the previous data. Keywords : physical property, bread dough, expansion Introduction More than one million tons of flour is used to make bread every year in Japan. However, most of this flour is imported from foreign countries. Although some wheat for bread is produced domestically (less than +*,*** tons per year), much of it is subject to preharvest sprouting damage, which causes partial gluten decomposition by endo-protease, resulting in softening of the physical properties of the dough. Bread made from degraded flour is low in quality (Ichinose et al.,,**+). Yamauchi et al. (,**-) have reported that blending of domestic flours with ES (extra strong) flour such as Victoria INTA is useful for improving domestic flours of inferior quality. The physical properties of dough made from appropriate blends, such as breaking force and mixing peak time, are closer to those of the imported varieties Hard Red Winter Wheat (HRW) or No. + Canada Western Red Spring Wheat (+CW), which is used for making bread. In this way, domestic flour should to be able to enter into the bread market. Thus, the determination of physical properties by rheological analysis is important for understanding the bread-making quality of flour in detail. However, there has not been much progress in this area in recent decades, even in terms of the fundamental viscoelastic characteristics of homogeneous materials. Bloksma (+3/1) analyzed the dynamics of expansion in To whom correspondence should be addressed. hdkawai*@mmm.muroran-it.ac.jp bread dough by Alveograph, which is known as a method for measuring the expansion of dough, using a Maxwell model with a polar coordinate system, and calculated the stress behavior of bread dough using a computational treatment. He reported that stress relaxation time was the most important parameter in the expansion of the bread dough. Launay et al. (+311) also carried out expansion analysis of dough bubbles in detail using the equation proposed by Bloksma (+3/1). However, the relationship between the physical properties of dough and the quality of the resulting bread was not analyzed in above reports. Matsumoto et al. (+32+) studied the expansion of bread dough theoretically using the Maxwell model under conditions of constant cross-section (nominal strain). Although the analysis quantitatively explained stress behavior changes in terms of di#erences in initial stress, it was not applied to practical bread-making conditions. Shimiya and Yano (+322) analyzed the expansion mechanism of air bubbles in gel, showing that the viscoelastic behavior and instantaneous elasticity of bread dough is dependent on the value of setting stress. It can be seen that there has been little investigation of the practical application of bread-making. Materials and Methods Analytical model Figure + shows the Kelvin fourelement model, in which the parameters E *,E +,h +,and represent instantaneous elasticity, retardation elasticity, retardation coe$cient of viscosity, and regularity coe$-

2 92 H. KAWAI et al. Fig. +. Kelvin four-element model. cient of viscosity, respectively. These physical property coe$cients are derived approximately from the creep curve (Fig.,), which is measured as a time-dependent strain curve under conditions of constant stress. The relaxation time (t *) and retardation time (t +) are defined as t * /E * and t + h +/E + respectively. As t * approaches infinity and t + approaches *, the characteristics of elasticity are increased. Figure, is an example of the creep curve of Camellia (strong flour) under a constant stress of P *.33 Pa, in which the nominal strain (g) is approximated using the following equation based on the Kelvin four-element model: g E * E + + e t+ t Eq. (+) Here, P * is constant stress, and t denotes time. As shown in Fig.,, the creep curve is mainly classified in three regions: E * is the instantaneous elasticity region, E + the retardation elasticity region, and the regular viscosity region. The Kelvin four-element model is more accurate for describing these three regions because retardation elasticity cannot be dealt with under the Maxwell model, which results in overestimation of instantaneous elasticity. However, the Maxwell model is still useful and practical because of its simplicity, for example in cases in which retardation elasticity does not significantly a#ect the behavior of the whole. In Fig.,, a tangential line from the regular viscosity region is defined as y h in Eq.(,). t y h E * E + t Eq. (,) From Eq. (,), is obtained, and the term (y h g) is described as t y h g e t+ Eq. (-) E + E + and t + are calculated from Eq. (-) and, finally, E * and t * are derived from Eqs. (+) and (,). Theoretical solutions for stress relaxation When the compression speed and cross section are assumed to be constant with time, a theoretical solution for stress relaxation is derived from the Kelvin four-element model as follows: Here, Fig.,. Typical creep curve under constant stress. s E + h +l + A +e l+t E + h +l, A,e l,t Eq. (.) s t, s t+ e l, t, A + E + h +l + e l + t, e l, t, s t, s t+ e l + t, A, E + h +l, e l, t, e l + t, where s t+ and s t, are the stress values at t t + and t t, respectively, and l + and l, are the roots of the following Eq. (/): h + E * l, E*h+ E*hN E+hN E * l E+ * Eq. (/) Preparation and properties of trial flours Victoria INTA (extra strong wheat) was cultivated in the Memuro test field at the National Agricultural Research Center for the Hokkaido Region in,***. Flours of 0* extraction rate were obtained from the test mill of Bühler Inc., Uzwil, Switzerland. Commercial foreign hard flour, Camellia, and domestic semi-strong flour, Hokushin, were purchased from Nisshin Flour Milling Co., Ltd., Tokyo, and Ebetsu Flour Milling Co., Ltd., Ebetsu, respectively. The protein contents of the flours were measured using a nearinfrared reflectance instrument (Inframatic 2+,*, PerCon Co., Hamburg, Germany). The protein contents of Victoria INTA, Camellia and Hokushin flours were +..3, +,.+ and +*.., respectively. The absorbed water in the di#erent flour types was measured with a Farinograph (Brabender, Inc., Duisburg, Germany), which requires the use of small mixing bowls, according to the American Association of Cereal Chemists (AACC) method (+33+). The measurements for absorbed water in each type of flour Victoria INTA, Camellia and Hokushin were 0., 00 and /0./, respectively. Preparation of dough and tests of bread-making The tests for bread-making were carried out using standard

3 Physical Properties of Bread Dough 93 Table +. Baking conditions, physical property coe$cients of dough, SLV and GRD. white bread composition and the no-time method, modified partially by the method of Yamauchi et al. (,***), and the parameters shown in Table + were examined. The doughs were prepared using,** g of flour. The general dough formulation was as follows:,** g of flour, +* gof sugar, +* g of shortening,. g of salt,. g of yeast, and *.*, g of L-ascorbic acid (ASA) or glutathione (reduced form) (GSH). Doughs were mixed to just beyond peak development, as indicated by the current curve, on the modified mixer described by Takata et al. (,***). The dough was divided into two +** g portions, rounded, and rested for,* min in a fermentation cabinet. The portions were then placed into pans (height / cm width 3 cm length +-./ cm) and proofed at -* for 3* min, then baked at,** for,/ min to obtain open-top type bread. The specific loaf volume (SLV) was measured by the rapeseed replacement method, using bread left at room temperature for one hour after baking. The vacuum expansion of the dough (gas retention of dough, GRD) was measured by the method developed by Yamauchi et al. (,***), which required the use of,* g of dough following proofing. GRD was evaluated by measuring the maximum volume of dough in a cylinder under low pressure. It has been proven that there is the high correlation between GRD and SLV, and the vacuum expansion of dough is an index of bread expansion. The dough to be used for measurement of physical properties was prepared as for the bread dough, except that no yeast was added, because accurate measurement of physical properties becomes di$cult when leavened dough is used. Measuring condition of physical properties of dough Measurement of the physical properties of dough were conducted just after mixing. The dough was divided into -* g pieces, compressed to +./ cm thickness using a flat wooden box, and then kept in an incubator at -* for 0* min in order to ensure a flat surface. The physical properties of this dough were measured using a creep meter, Rheoner (Model RE --**/, Yamaden Co., Ltd., Tokyo). The dough sample was placed on a moving table and then vertically compressed at constant weight of.33 Pa, a compressing speed of +* mm/s and a hold time of + min. Each coe$cient of the Kelvin four-element model was determined using the automatic analytical software of the Rheoner. Simulation of creep and stress relaxation curve Creep simulation and stress relaxation simulation were carried out using Eq. (+) and Eq. (.), respectively. The physical properties were determined previously from the Kelvin four-element model. The boundary condition for creep and the initial conditions for stress relaxation are P *.33 Pa and s at t * s and / s, respectively. Statistical analysis All experimental data are shown as mean standard deviation (SD). Statistical data analysis was performed using Microsoft Excel,*** software. Results and Discussion Approximation of instantaneous elasticity (E *) Because the ideal instantaneous elasticity, which has no time dimension, cannot be obtained, it is usually approximated from the stress-strain curve (S-S curve) for a very short time. For this, the linearity between stress and strain should be checked for the measuring time, although previous data for instantaneous elasticity has not been clarified even in the commercial software services. Therefore, the relationship of the S-S curve was measured for each dough (Victoria INTA, Camellia and Hokushin) by operation of short-time compression, about *./ s, to the setting stress value (P *). The measuring conditions were same as those of the creep analysis described above. Although the data is not shown, the S-S curve for the first *./ s in all samples could be approximated as linear, and this relation gives a guarantee that E * is available as a physical property under these measuring conditions. Experimental and simulated creep and stress relaxation curve Figure - shows creep curves for dough made from Camellia (ASA). Taking account of the instantaneous elasticity determined in the previous process, the physical properties obtained from Eqs. (+), (,), and (-) approximate the creep curves comparatively well within the linear range. For the other samples and conditions, these linear

4 94 H. KAWAI et al. Fig. -. Comparison between experimental and simulated creep curve : : Simulation, : Experimental data. Fig... Comparison between experimental and simulated relaxation curve : : Simulation, : Experimental data. approximations were confirmed to be available. Figure. shows a comparison of the stress relaxation curves obtained experimentally and in simulation in the dough used for Fig. -. The curves were time-dependent under conditions of constant strain, and the coe$cients of physical properties, as well as the information obtained from Fig. -, were used for the simulation. The calculation was performed using Eq. (.) under the time boundary conditions of s t+ (at t + * [s]) and s t, (at t, / [s]). In general, initial conditions are chosen as the stress values and the first di#erential coe$cient at t *, but in this case, the di#erential coe$cient measured at t * was too large to be simulated with the linear model. In Fig.., the calculations were in good agreement with the experimental results in the linear region, and similar results were obtained in doughs made from di#erent flour samples and under di#erent conditions, which means that the values of physical properties determined from the experimental creep curve are well approximated in stress relaxation. Physical properties and GRD of various doughs and SLV of breads made from various doughs In doughs of original water absorption values, as shown in Table +, the E * of Camellia (ASA or GSH) was the highest, followed by Hokushin and Victoria INTA. The values of in these doughs were in the following order: Victoria INTA, Camellia (ASA), Hokushin, Camellia (GSH). For E + and h +, the di#erences among these doughs were not great, but E + and h + of Victoria INTA showed the largest and smallest values, respectively, and E + of Camellia (GSH) was the smallest. The t * values showed the same order as the values of, and the order of t + values was reversed for t *. From these results, as the gluten strength of the doughs increased, the doughs showed higher E +, lower h +, and higher values, and, as a result, higher t * and lower t + values. GSH (reducing reagent) did not have a large influence on E * and h + values, but did result in smaller values of E + and (especially ), weakening the dough. In both Camellia (ASA) and Hokushin doughs, lower values of water absorption resulted in increases in all coe$cients of physical properties except for t +; only t + decreased. In the relationship between SLV, GRD and each physical property coe$cient, SLV and GRD, as indexes of bread expansion, showed a basic tendency to increase as t * rose and t + declined. In doughs of original water absorption values, the SLV of Victoria INTA was the highest, followed by Camellia (ASA), Hokushin, and Camellia (GSH). This order corresponded to those of t *,E +,and, and inversely to that of t +. In the same doughs, the GRD of Victoria INTA was the highest, followed by Hokushin, Camellia (ASA), and Camellia (GSH). This order did not correspond to those of all physical property coe$cients. In the doughs of Camellia (ASA) and Hokushin in which the amounts of absorbed water were changed, the SLV and GRD of Camellia doughs were enhanced as water absorption was lowered, while, although Hokushin dough with high water absorption showed low SLV and GRD values, there were no great di#erences among the other doughs. However, the values for dough expansion (except for that of Camellia (GSH)) in final proofing were not significantly di#erent. For Camellia (GSH), the dough was very soft and the expansion of the dough in final proofing was somewhat small. Relationship between physical properties and the expansion ability of baked bread To clarify the relationship between the physical properties of dough and bread expansion, the Pearson s correlation coe$cients for SLV, GRD and each physical property coe$cient were calculated, as shown in Table,. SLV and GRD are correlated negatively with t + to a statistically significant extent. Although there was no statistical significance, t * was correlated positively with SLV and GRD. There was no significant correlation in the relationships of the other physical property coe$cients. For dough, a high t * and a

5 Physical Properties of Bread Dough 95 Table,. Pearson s correlation coe$cients between SLV or GRD and each physical property coe$cient. low t + mean that the dough has more elastic properties. From analysis using the Maxwell model and the Alveograph, Bloksma (+3/1) reported that the dough with a higher t * value showed larger expansion. It was also reported that the volume of bread made from dough mixed for various times increased with the relaxation time of the dough, which is nearly proportional to the stress relaxation time, t * (Frazier et al., +31/; Rao et al, +333), and that bread volume was weakly correlated with the physical properties of dough, such as dynamic modulus, tan d values and relaxation rate (Janssen et al., +33/; Rao et al., +333; Mani et al., +33,). The above papers show that bread of larger volume and dough with a higher gas retention value are obtained from dough with more elastic physical properties. The results of this study basically corresponded to those of the above-mentioned reports and may be summarized as follows: it was suggested that the physical properties of dough, such as SLV and GRD, greatly influence the expansion ability of bread, and that more elastic dough, with high t * and low t + values, underwent good expansion during baking. It has long been experientially known that adequate elasticity in dough is necessary to make the dough expand well, and the results of this study support the previous data. Acknowledgments This work was supported in part by Grantsin-Aid for Research and Development Programs for New Bioindustry Initiatives from Bio-oriented Technology Research Advancement Institution (BRAIN), Japan. References American Association of Cereal Chemists (+33+b). Approved Methods of the AACC, Method *2 *+, The Association, St. Paul, MN. Bloksma, A. H. (+3/1). A calculation of the shape of the Alveograms of some rheological model substances. Cereal Chem., -., +, Frazier, P. J., Daniels, N. W. R. and Eggitt, P. W. R. (+31/). Rheology and the continuous breadmaking process. Cereal Chem., /,, +*0r-+-*r. Ichinose, Y., Takata, K., Kuwabara, T., Iriki, N., Abiko, T. and Yamauchi, H. (,**+). E#ects of increase in a-amylase and endoprotease activities during germination on the bread making quality of wheat. Food Sci. Technol. Res., 1,,+.,+3. Janssen, A. M., van Vliet, T. and Vereijken, M. (+330a). Fundamental and empirical rheological behavior of wheat flour doughs and comparison with bread making performance. J. Cereal Sci.,,-,.- /.. Janssen, A. M., van Vliet, T. and Vereijken, M. (+330b).Rheological behavior of wheat glutens at small and large deformations. E#ect of gluten composition. J. Cereal Sci.,,-, --.,. Launay, B., Bure, J. and Praden, J. (+311). Use of the Chopin Alveographe as a rheological tool. I. Dough deformation measurements. Cereal Chem., /., +*., +*.2. Mani, K., Eliason, A., Lindahl, L. and Trägårdh, C. (+33,). Rheological properties and breadmaking quality. Cereal Chem., 03,,,,,,/. Matsumoto, H. (+32+). Various basic research on the expansion of bread. Cyourikagaku, +.,,+/,,+ (in Japanese). Rao, V. K. Mulvaney, S. J. and Dexter, J. E. (+333). Rheological characterization of long- and short-mixing flours based on stress-relaxation. J. Cereal Sci., -+, +/ Shimiya, Y. and Yano, T. (+322). Rates of shrinkage and growth of air bubbles entrained in wheat flour dough. Agric. Biol. Chem., /,,,213,22-. Takata, K., Yamauchi, H., Iriki, N. and Kuwabara, T. (,***). Relationship between seed storage protein and bread-making quality of wheat cultivar Haruyutaka. Ikushugaku Kenkyu,,, (in Japanese). Yamauchi, H., Noda, T., Matsuura-Endo, C., Nishio, Z., Takata, K., Tabiki, T., Saito, K., Oda, Y., Funatsuki, W. and Iriki, N. (,**-). Improving domestic flour for bread making by blending extra strong (ES) flour. Food Preservation Science,,3,,++,,*. Yamauchi. H., Ichinose, Y., Takata, K., Iriki, N. and Kuwabara, T. (,***). Simple estimation of bread-making quality of wheat flour by modified expansion test under reduced pressure. Nippon Shokuhin Kagaku Kogakkaishi,.1,.0.3 (in Japanese).