Yeast-Leavened Laminated Salty Baked Goods : Flour and Dough Properties and Their Relationship with Product Technological Quality

Puff and Danish pastries are among the most common baked products involving an interleaving of thin layers of fat and dough sheets in their preparation. Puff pastry is made by layering shortening with tough but pliable dough in a similar way to Danish pastry (1). Dough expansion depends largely on the ability of dough layers to remain separate and discrete from fat layers, although other factors also contribute, to some extent, to the product lift ing. In unleavened puff pastry, water is held in dough layers and, when converted to steam, is trapped in the melting fat between the dough layers, producing the lift ing. Consequently, in sweet Danish pastry, a very open network of crispy and fl aky layers is formed and the presence of yeast generates a relatively soft and porous structure in the baked dough layers (2).


In troduction
Puff and Danish pastries are among the most common baked products involving an interleaving of thin layers of fat and dough sheets in their preparation.Puff pastry is made by layering shortening with tough but pliable dough in a similar way to Danish pastry (1).Dough expansion depends largely on the ability of dough layers to remain separate and discrete from fat layers, although other factors also contribute, to some extent, to the product lift ing.In unleavened puff pastry, water is held in dough layers and, when converted to steam, is trapped in the melting fat between the dough layers, producing the lift ing.Consequently, in sweet Danish pastry, a very open network of crispy and fl aky layers is formed and the presence of yeast generates a relatively soft and porous structure in the baked dough layers (2).
The suitability of fl our to obtain puff and Danish pastries has been reported, e.g.Cauvain and Young (2) and Matzs (3) affi rmed that the use of too strong fl our can cause excessive shrinking of puff pastry products.Strong types of fl our require longer resting periods than weak fl our in order for the dough rheology to become optimised for sheeting and laminating.Matzs (3) found an improvement in Danish pastry quality when 15 or 20 % of hard fl our was replaced by soft wheat fl our.Davies et al. (4) studied the structure and functionality of proteins in pastry dough before and during the baking process and they found that high quality pastry fl our is able to form thin dough laminates (30 μm), while the fl our of inferior quality formed thicker and less well-defi ned sheets.In 1989, Zabik and Tipton (5) evaluated the infl uence of quantity and quality of the gluten of soft wheat fl our on textural characteristics of pastry.Positive correlations were found between gluten quantity and fl akiness and crust shrinkage, while surface blistering score and breaking strength decreased as the gluten quantity decreased.Hay (6) found that specifi c puff pastry height increased with increasing protein content.In a similar way, the specifi c puff pastry volume had positive correlations with protein content high molecular mass, low molecular mass and quantity of dough strength-related high-molecular mass glutenin subunit.
Although considerable eff orts have been made to establish fl our quality parameters that can be used to reliably predict puff and Danish pastry product quality, the objective has been partially fulfi lled.Geitt ner (7) recommended the Brabender extensogram as a quality parameter and proposed a range between 80 and 110 cm 2 .Hay (6) identifi ed dough properties (tension energy and water absorption) as the best predictors of specifi c puff pastry height and volume.On the other hand, Hay (6) also studied the relationships between pastry quality parameters and bread.However, bread baking performance cannot be used as an indicator of the baking quality of a pastry, and the models developed to predict loaf volume from protein components of fl our do not hold in this context.Morgenstern et al. (8) argued that pastry dough is diff erent from bread dough, not only in composition but also in the strains and strain rates that are applied during pastry making.Therefore, against some existing tedious methods untranslated into fundamental rheology properties, they proposed a quick and easy method to measure extensional properties of sheeted dough pieces.However, the issue of establishing fl our quality parameters to help predict the quality of laminated bakery products in order to obtain products with similar and uniform characteristics over time is still unresolved.Therefore, the study of yeast-leavened laminated salty products will contribute to the scientifi c knowledge about laminated systems and its industrialization in countries where laminated baked goods are among the most consumed products (9).
The yeast-leavened laminated salty products share some characteristics with puff and Danish pastries.However, the simultaneous presence of salt and yeast in a laminated system aff ects fl our quality requirements, dough behaviour during sheeting, fermentation and baking stages and fi nal quality of the product in diff erent ways.The yeast plays a signifi cant part in the aeration of the dough during fermentation and baking, and it also disrupts the integrity of the dough layers and fat layers (2).
In this context, the aim of our work is to establish fl our and dough quality parameters to help predict the quality of a yeast-leavened laminated salty product.The relationships between fl our characteristics, laminated dough pieces and baked products were studied.

Flour characterisation
Total proteins were determined following the AACC International method 46-10.01(10).Wet gluten was obtained by the hand washing method following the AACC International method 38-10.01(11).The glutenin macropolymer was isolated according to Don et al. (12) and its protein content was determined by the AACC International Kjeldahl method 46-13.01(13).Results were expressed in g of glutenin macropolymer per 100 g of fl our.
Solvent retention capacity (SRC) profi le was obtained according to the AACC International method 56-11.02(14).Flour samples (5 g) were suspended with 25 g of water, 50 % sucrose, 5 % sodium carbonate, and 5 % lactic acid.The samples were hydrated for 20 min and centrifuged at 1000×g for 15 min.Each obtained precipitate was weighed and the SRC of each sample was calculated according to the AACC International method (14).Sodium dodecyl sulphate sedimentation index was determined following the AACC International method 56-70.01(15), measuring the volume (cm 3 ) obtained from 1 g of fl our suspended with 12 mL of sodium dodecyl sulphate reagent and submitt ed to shaking and resting periods.Each test was performed at least in duplicate.

Elaboration of a yeast-leavened laminated salty product
The dough was prepared from 100 g of wheat fl our, 20 g of lard, 2.8 g of compressed yeast, 2.5 g of salt, 1.4 g of sugar and 50 mL of water.The ingredients were mixed for 3 min in a mixer (MPZ Pedro Zambón e hĳ os, Córdoba, Argentina), until the dough was obtained.A mass of 33.3 g of lard was envelope-folded into a dough sheet and then gauged to 60 mm thickness in six steps, using a sheeter (MA-AR Acrilic, S.R.L., Córdoba, Argentina).The dough was given a twofold turn and allowed to rest for 20 min at 23 °C; then it was gauged to 50 mm thickness in seven steps and given another twofold turn.It was then allowed to rest for 20 more min and gauged to a thickness of 50 mm.The dough was laminated via a two-fold turn and the fi nal gauging was to about 150 mm thickness.Circular perforations (diameter d=2 mm) of 1.6 cm each were done on the dough to avoid the complete separation of layers during baking.Square dough pieces (5 cm×5 cm×1.5 cm) were fermented at 35 °C and 80 % relative humidity, until they duplicated their height.They were baked at 175 °C for 27 min in a Beta 107 IPA convector oven (Pauna, Córdoba, Argentina).Three products of each sample were produced and the procedure was repeated at least twice.

Baked product quality parameters
The conformational change of the dough pieces during the production process was evaluated.The dimensions (height, width and length) of the dough pieces at the beginning of fermentation and of the baked products aft er cooling for 1 h were measured.The height and width ratios of baked vs. unfermented dough were calculated.The shape factor (SF) of the baked products was calculated as follows: The product volume was determined by rapeseed displacement aft er cooling for 1 h.The specifi c volume was expressed as the volume per mass ratio of the fi nal product.A piece of crumb of 20 mm thickness, previously cut in a longitudinal direction, was compressed to 40 % of its initial height using a cylindrical probe (d=2.5 cm) in the Instron 3342 (Norwood, MA, USA) texture analyser.Force deformation curves were obtained at a crosshead speed of 1 mm/s.Crumb fi rmness was defi ned as the maximum force registered and was expressed in Newton (N).Crust fi rmness was determined directly on the baked product, under the same conditions mentioned before and expressed in N. Three products of each sample were tested and each measurement was performed at least in duplicate.The determinations were carried out on baked products aft er cooling for 1 h.

Dough quality parameters
A stress relaxation test was done using the Instron 3342 texture analyzer (Norwood).Cylindrical pieces of nonfermented laminated dough prepared as described above were compressed with a cylindrical probe (d=5.0cm), at 0.5 mm/s speed up to 30 % of its initial height and maintained compressed for 2 min.The force was recorded as a function of time and the relaxation curves were fi tt ed with the following equations: /2/ /3/ This expression corresponds to a Maxwell element in parallel with a spring (16), where σ is the stress, E 1 is the elastic modulus, ε 0 is the equilibrium elastic modulus and t r is the relaxation time, which is defi ned by Eq. 3, where the viscosity (η) and E 1 are related.
A piece of nonfermented laminated dough prepared as described above was compressed up to 40 % of its ini-tial height, using a cylindrical probe (d=2.5 cm).Force deformation curves were obtained at a crosshead speed of 1 mm/s.Dough resistance to deformation was defi ned as the maximum force registered.Three dough pieces of each sample were tested and each test was performed at room temperature (25 °C) and at least in duplicate.

Statistical analysis
The experimental determinations were done at least in duplicate and compared by the analysis of variance (ANOVA), using the Di Rienzo, Guzmán and Casanoves test (DGC), where the relationship between the measured parameters was assessed by Pearson's test (signifi cant level at p≤0.05) (17).A multivariate analysis of variance (MANOVA; InfoStat statistical soft ware, Faculty of Agricultural Sciences, UNC, Córdoba, Argentina) was used in order to analyse global diff erences between the samples considering more than one variable.To compare the multivariable hypothesis, the Hotelling method was used (signifi cant level at p≤0.05).

Characteristics of fl our samples
Samples 1 and 2 of soft wheat fl our had the lowest protein values, while soft wheat fl our sample 3 and all samples of hard wheat fl our had higher protein content (Table 1).Sliwinsky et al. (18) related the diff erences in fl our baking performance with puff pastry dough rheological properties using the fl our with a protein content in the range of 10.3-13.5 %.Sample 1 of soft wheat fl our also had the lowest wet gluten content, followed by soft wheat fl our sample 2, and hard wheat fl our samples 6 and 8. Hard wheat fl our samples 4 and 9 had intermediate wet gluten percentages.Soft wheat fl our sample 3 and hard wheat fl our samples 5 and 7 had wet gluten values higher than 37 % (Table 1).Soft wheat fl our sample 2 had the lowest glutenin macropolymer mass fraction, followed by soft wheat fl our samples 1 and 3. Flour from hard wheat had higher glutenin macropolymer content than soft wheat fl our samples, with the highest mass fraction in hard wheat fl our sample 5 (Table 1).Don et al. (19) and Steff olani et al. (20) found similar glutenin macropolymer values in hard wheat fl our samples (0.5-3.6 and 2.0-3.3 %, respectively).
Lactic acid SRC values (Table 1) were in general higher than 115 % in hard wheat samples, except for hard wheat sample 6, while soft wheat fl our samples had lower lactic acid SRC values.This tendency allowed us to associate hard wheat fl our samples with a high glutenin network quality, and consequently with a high relative strength (21).Moiraghi et al. (22) reported a similar range of lactic acid SRC values in Argentinian hard wheat fl our (99.9-121.0%).Hard wheat fl our samples had sucrose SRC values greater than 90 %, while soft wheat fl our samples had lower percentages.These results suggest that hard wheat fl our samples had a high content of pentosan and gliadin ( ) tinian hard wheat and hard red winter wheat fl our from the USA, respectively.The sodium carbonate SRC values of soft wheat fl our were lower than of hard wheat samples, showing that hard wheat fl our had a higher level of damaged starch.This could be att ributed to the greater force applied to the grains of hard wheat during the milling process (24).The hard wheat samples had higher values of water SRC, indicating their great ability to hold water (24).Soft wheat fl our had similar water SRC values to those of Argentinian soft wheat samples determined by Moiraghi et al. (25).Hard wheat fl our had the highest sodium dodecyl sulphate sedimentation volume, which revealed strong capacity to form a protein network, necessary to retain the gas during fermentation.These results are in agreement with those of Colombo et al. (23) for Argentinian hard wheat samples (11.75-19.25 %).On the other hand, hard wheat fl our sample 6 and soft wheat fl our samples had the lowest sodium dodecyl sulphate sedimentation volume.There was a great variability among the studied samples which allowed us to evaluate the relationships among fl our characteristics, laminated dough pieces and baked products.

Dough properties
Dough samples made with hard wheat fl our showed diff erent stress relaxation characteristics from samples made with soft wheat fl our (Table 2).Dough samples 7, 8 and 9 made with hard wheat fl our had the highest elastic modulus (E 1 ) values, which could be associated with stiff dough samples (26).At the equilibrium, hard wheat dough samples had higher elastic modulus (ε 0 ) values than soft wheat dough samples.Characteristic relaxation time (t r ), considered a discriminator of strength (27), of dough samples made with hard wheat fl our and soft wheat fl our sample 3 was higher, revealing a stronger solid-like behaviour (16).Dough samples 1 and 2 made with soft wheat fl our had lower values of t r , related to a fast relaxation of the system.Li et al. (28) found that dough and gluten from English strong fl our had higher relaxation modulus and relaxation intensity than those with weak fl our over the whole relaxation time in a fundamental rheology test.They suggested that diff erences in the relaxation behaviour between fl our types with different baking quality were related to the gluten network structure.Dough and gluten from strong fl our had a strong network, which may be att ributed to the protein molecular entanglements and physical cross-links.The lower values of viscosity (η) of dough samples made with soft wheat fl our indicate a higher capacity to fl ow than the samples made with hard wheat fl our.
Samples of the dough made with soft wheat fl our had values of resistance to deformation lower than 10 N (Fig. 1), while samples of the dough made with hard wheat fl our were harder, which is associated with higher values of resistance to deformation.Positive and signifi cant correlation between resistance to deformation and η (R=0.72;p≤0.05) suggested that samples of hard dough were more viscous.

Properties of yeast-leavened laminated baked product
Yeast-leavened laminated baked products were prepared (Fig. 2) using soft and hard wheat fl our.The lateral view of the products made with hard wheat fl our revealed sheeted structure with thin layers aligned horizontally.On the other hand, products made with soft wheat fl our had a disrupted structure where layers seemed to be merged, forming a coarse and uneven strata.
In yeast-leavened laminated products the desired expansion is an upward growth, during which the dough maintains its symmetry without excessive lateral growth.In this study a shape factor, which refl ects the dimensions of a baked product, was considered.The shape factor values (Table 3) for products made with hard wheat fl our were higher, revealing a greater height and smaller width and length than the products made with soft wheat fl our.Soft wheat fl our products had higher values of width ratio than hard wheat fl our products, while products made with hard wheat fl our samples 4 and 5 had the highest height ratio (Table 3).Most samples did not exhibit a signifi cant diff erence in their specifi c volumes, except for the hard wheat fl our product 6, which had the highest value.Hay (6) found that a high quality puff pastry had a high specifi c volume.However, in yeast-leavened laminated products the specifi c volume was not a parameter that allowed diff erentiating the quality of the product.The baked products made with soft wheat fl our had lower values of crumb fi rmness than products with hard wheat fl our samples (Fig. 1), revealing a soft er structure.The same tendency was observed for the crust fi rmness (Fig. 1), where all the products of hard wheat samples, except sample 6, had tougher crusts.The baked products from soft wheat fl our seemed to have a more spongy structure, while hard wheat fl our products had a fi rm arrangement associated with a previously layered structuration of the dough sheets.

Relationships between fl our characteristics, dough and yeast-leavened laminated product properties and quality
Glutenin macropolymer content and dodecyl sulphate sedimentation volume of the fl our samples showed associations with dough viscosity (R=0.77and 0.79, respectively; p≤0.05) and resistance to deformation (R=0.70 and R=0.77, respectively; p≤0.05).The hard wheat fl our samples with a high quality gluten network produced hard sheeted dough with a more solid-like behaviour and less capable of fl owing.Protein content of the fl our was signifi cantly and negatively correlated with width ratio and the shape factor (R=-0.50 and -51, respectively; p≤0.05) of the baked products, while the wet gluten was only negatively associated with the width ratio (R=-0.47;p≤0.05).There were signifi cantly positive correlations between the glutenin macropolymer content and height ratio and the shape factor (R=0.55 and 0.78, respectively; p≤0.05), and a negative correlation with the width ratio (R=-0.69;p≤0.05).This indicated that the fl our high in glutenin macropolymer content produced laminated products which rise during the baking process and suff er less lateral expansions.Hay (6) found a positive correlation between fl our protein, high molecular mass glutenin subunit and low molecular mass glutenin subunit content and the specifi c pastry height in puff pastry products without yeast leavening.
Predictive tests showed signifi cant associations between the fl our quality and the dimensional parameters of the products.The lactic acid SRC had positive correlations with the height ratio and the shape factor (R=0.59 and 0.63, respectively; p≤0.05) and a negative correlation with the width ratio (R=-0.46,p≤0.05).The sodium dodecyl sulphate sedimentation volume had positive correlations with the height ratio and the shape factor (R=0.60 and 0.79, respectively; p≤0.05), and a negative correlation with the width ratio (R=-0.46,p≤0.05).Results show that the expansion seems to be more related to the quality of protein and gluten than to their quantity.
In dough with added fat, fat competes with the aqueous phase for the surface of fl our particles during dough mixing (29).The added shortening causes a lipid plasticisation of gluten and modifi es or competes with the interactions of endogenous components.In the dough with high fat content (more than 5 %), the added fat appears to saturate the system and only part of the added lipid can exert a plasticising eff ect (30).One portion of the incorporated water will be an integral part of the dough and another will act as a plasticiser.The extent to which the water fulfi lls either role will depend on the presence of certain water-absorbing components, such as proteins, arabinoxylans and damaged starch.When the dough is subjected to compression, the mobility of water molecules present in the interparticle space infl uences the relaxation of dough.In the same way, in a layered structure system with thin fat/dough layers the fat would have a strong infl uence on the stress relaxation.Sodium carbonate SRC and water SRC showed signifi cant associations with dough viscosity (R=0.71 and 0.80, respectively; p≤0.05) and resistance to deformation (R=0.77and 0.81, respectively; p≤0.05).These relationships suggest that in dough made with fl our high in hydrophilic components, water mobility is limited.Water exerts a plasticiser eff ect and dough presents a strong cohesive structure with a higher value of η (31).The correlations between dough resistance to deformation and the width ratio (R=-0.77;p≤0.05) and the shape factor (R=0.70; p≤0.05) of the yeast-leavened laminated products revealed that hard wheat dough samples may have a more rigid structure than soft wheat dough pieces.Therefore, dough samples made with hard wheat fl our have a great capacity to form layers that can be vertically expanded.Shape factor showed positive correlations with the fl our capacity to absorb water (R=0.83;p≤0.05), sucrose (R=0.83;p≤0.05) and sodium carbonate (R=0.83;p≤0.05) and with dough viscosity (R=0.73;p≤0.05).This revealed that the existence of a certain level of arabinoxylans and damaged starch generated viscous dough, which had a positive eff ect on the desirable expansion of the product.There were negative correlations between the width ratio and the sucrose SRC (R=-0.74;p≤0.05) and sodium carbonate SRC (R=-0.78;p≤0.05) of fl our and dough viscosity (R=-0.79;p≤0.05).These showed that undesirable increments of the width of the dough samples are related to less viscous dough samples, such as those made with soft wheat fl our, which had a low content of components capable of holding water.
The following variables were simultaneously analyzed by MANOVA (Table 4): glutenin macropolymer content, lactic acid SRC and sodium carbonate SRC, viscosity, resistance to deformation, width ratio and shape factor.Flour samples, dough samples and yeast-leavened laminated products from hard wheat were signifi cantly diff erent from those from soft wheat.Hard wheat fl our samples showed a good quality gluten network with high glutenin macropolymer content and lactic acid SRC values.Hard wheat fl our had higher sucrose SRC and sodium carbonate SRC values than soft wheat fl our, related to the greater amount of hydrophilic components.Dough samples showed a diff erent behaviour when subjected to compression and during the stress relaxation process.Dough samples from hard wheat fl our were more viscous and harder than those from soft wheat fl our.The diff erences in fl our and dough characteristics have an eff ect on the properties of yeast-leavened laminated products.Products from hard wheat fl our samples had a higher shape factor and lower width ratio than products from soft wheat fl our.These results are in agreement with Manley's (29) observations, since he mentioned that laminated dough relaxation periods are related to the length and shape of the baked product.The combined use of the selected variables related to fl our, dough and product characteristics allowed to diff erentiate among wheat fl our samples with diff erent suitability to produce yeast-leavened laminated products.

Conclusions
Flour quality and dough parameters that help to predict the quality of a yeast-leavened laminated salty product were established.A fl our sample with high glutenin macropolymer content, a strong glutenin network, and a certain level of hydrophilic components is suitable for the production of viscous dough with high resistance to deformation and, consequently, a laminated baked product with an optimum quality, which rises and does not lose the expected shape.
The components present in the fl our and later in the dough have an infl uence on the fi nal quality of the laminated baked products.Hydrophilic components which can hold water generated a rigid structure and viscous dough.The gluten network quality infl uenced, to a lesser extent, dough fi rmness.Dough samples with a more viscous behaviour suff ered a lift rather than a lateral expansion, without losing the expected shape, when exposed to heat during baking.Hard wheat fl our samples were signifi cantly diff erent from soft wheat fl our samples.Laminated baked products from hard wheat fl our had bett er quality properties than those from soft wheat fl our, revealing that fl our from hard wheat is more suitable for making this kind of product.
. The observed values are in agreement with those of Colombo et al. (23) and Xiao et al. (24), who registered sucrose SRC values greater than 95.11 % in Argen-

Fig. 1 .Fig. 2 .
Fig. 1.Textural parameters of the dough and yeast-leavened laminated salty product: a) dough resistance to deformation, b) baked product crumb fi rmness, c) baked product crust fi rmness.Determinations were done at least in duplicate.Columns with a diff erent lett er are signifi cantly diff erent (p≤0.05).Sw=soft wheat fl our samples, Hw=hard wheat fl our samples Wheat samples from the 2011 Argentinian bulk harvest were provided by the Experimental Research Station Marcos Juárez of the National Institute of Agricultural and Fishing Technology (INTA), Buenos Aires, Argentina.

Table 1 .
Flour quality parameters and predictive testsSw=soft wheat fl our, Hw=hard wheat fl our, WG=wet gluten, GMP=glutenin macropolymer, SRC=solvent retention capacity, V s =sodium dodecyl sulphate sedimentation volume.Values are on a 14 % moisture basis.Determinations were done at least in duplicate.Values followed by a diff erent lett er are signifi cantly diff erent (p≤0.05)

Table 2 .
Laminated dough stress relaxation parameters

Table 4 .
The results of MANOVA analysis Sw=soft wheat fl our sample, Hw=hard wheat fl our sample, GMP=glutenin macropolymer, SRC=solvent retention capacity, η=viscosity, DR=dough resistance to deformation, W=width ratio, SF=shape factor.Diff erent lett ers within a sample indicate signifi cant diff erences (p<0.05) between the groups of samples considering the eight variables at the same time

Table 3 .
Yeast-leavened laminated salty product quality parameters Sw=product made with soft wheat fl our, Hw=product made with hard wheat fl our, v=specifi c volume, W=width ratio, H=height ratio, SF=shape factor.Determinations were done at least in duplicate.Values followed by a diff erent lett er are signifi cantly diff erent (p≤0.05)