The role of sugar and fat in sugar-snap cookies: Structural and textural properties

Journal of Food Engineering 90 (2009) 400–408 Contents lists available at ScienceDirect Journal of Food Engineering journal homepage: www.elsevier.com/locate/jfoodeng The role of sugar and fat in sugar-snap cookies: Structural and textural properties Bram Pareyt a,*, Faisal Talhaoui a, Greet Kerckhofs b, Kristof Brijs a, Hans Goesaert a, Martine Wevers b, Jan A. Delcour a a Laboratory of Food Chemistry and Biochemistry and Leuven Food Science and Nutrition Research Centre (LFoRCe), Katholieke Universiteit Leuven, Kasteelpark Arenberg 20, B-3001 Leuven, Belgium b Department of Metallurgy and Materials Engineering, Katholieke Universiteit Leuven, Kasteelpark Arenberg 44, B-3001 Leuven, Belgium a r t i c l e i n f o Article history: Received 18 March 2008 Received in revised form 7 July 2008 Accepted 10 July 2008 Available online 19 July 2008 Keywords: Sugar-snap cookies Fat Sugar X-ray microfocus computed tomography Structure Texture a b s t r a c t The impact of sugar (17.6–31.2%) and fat (8.7–15.8%) levels on cookie structure was studied. Cookie diameter increased and its height decreased with increasing sugar or fat levels. X-ray microfocus computed tomography porosities and cell sizes increased with fat level, but cell size distribution, cell wall thickness and distribution were not affected by fat level, indicating that fat primarily incorporates air. In contrast, the sugar level influenced porosity, cell size, cell wall thickness and their relative distributions. Thus, the sucrose level, probably by affecting dough viscosity during baking, largely influences the baked cookie structure. Cell and cell wall anisotropy measurements indicated that the inner orientation of cells and cell walls probably depends on the horizontal spread behaviour, rather than on the maximum cookie height and collapse. Finally, the surface cracking pattern was determined by sugar level, rather than by structural collapse at the end of baking. Ó 2008 Elsevier Ltd. All rights reserved. 1. Introduction Cookies have high sugar and fat levels and, at the same time, low water levels (1–5%). Cookie dough constituents largely influence dough making and handling, cookie baking, and the quality of the resulting product (Pareyt and Delcour, 2008). Sucrose is the most important sugar in cookie making. It delivers sweetness, influences the structural and textural properties of cookies, and is presumed to incorporate air into the fat during cookie dough preparation. Furthermore, in cookies, sucrose decreases dough viscosity (Maache-Rezzoug et al., 1998). During baking, the undissolved sugar progressively dissolves, and hence contributes to cookie spread (Hoseney, 1994). Other parameters that are influenced by the recipe’s sugar (level) include cookie hardness, crispness, colour, and volume. Finally, Hoseney (1994) believes that, during baking, sucrose recrystallization at the cookie surface causes the typical surface cracking pattern. However, in contrast, others (Slade et al., 1993) relate the cracking pattern to the degree of collapse at the end of baking. A second crucial cookie dough ingredient is fat. It imparts shortening, richness and tenderness, and improves mouth feel, flavour (intensity) and perception (Zoulias et al., 2002). Fat contributes to cookie spread and to general product appearance, it enhances aeration for leavening and volume and makes the cookies more * Corresponding author. Tel.: + 32 0 16321649; fax: + 32 0 16321997. E-mail address: bram.pareyt@biw.kuleuven.be (B. Pareyt). 0260-8774/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.jfoodeng.2008.07.010 easily breakable (Maache-Rezzoug et al., 1998). For further aspects of the functionality of sugar as well as fat during cookie making, the interested reader is referred to Pareyt and Delcour (2008), and references cited therein. A present day dietary concern is the consumption of high quantities of fat and sugar. Fat consumption is associated with various health disorders such as obesity, cancer, high blood cholesterol, and coronary heart disease, whereas high sucrose intake is considered to be one of the factors which cause obesity and dental illness. In this context, there have been efforts to produce cookies with reduced fat and/or sugar levels (Zoulias et al., 2000; Sudha et al., 2007). This presents challenges for the baking industry, since fat and sugar cannot be easily replaced, especially in a complex food system such as cookies (Zoulias et al., 2000). From the above, and as already reported (Zoulias et al., 2000), it is clear that reduction in and/or replacement of sugar and fat largely influence the texture of the final cookie. Most published work dealing with cookie texture describes the variation in cookie break strength under different conditions, such as the use of different fat types (Jacob and Leelavathi, 2007), different fat (Zoulias et al., 2000, 2002; Sudha et al., 2007) or sugar replacers (Zoulias et al., 2000; Gallagher et al., 2003) and varying sucrose levels (Manohar and Rao, 1997). Cookie structure itself has been described as a cellular solid, with a porous inner texture (Pareyt and Delcour (2008) and references cited therein). However, to the best of our knowledge, no papers describe the effects of different sugar and/or fat levels on cookie (dough) B. Pareyt et al. / Journal of Food Engineering 90 (2009) 400–408 structure or relate the observed changes in cookie texture to its structure. In addition, despite much research, there seems to be no clear view on the factors that determine the cookie surface cracking pattern. We therefore set out to monitor and understand the changes in cookie structure and texture that take place when adapting (reducing) the recipe’s sugar and fat levels. Cookie structure was investigated for the first time using X-ray microfocus computed tomography (lCT). For a description of the principle of the tomography technique, we refer to earlier published work (Sasov, 1987). The technique is based on the difference in material density and atomic number, which themselves are reflected in the different degrees of X-ray absorption (Babin et al., 2007) or attenuation. This non-destructive technique allows obtaining three-dimensional (3D) images of the bulk of heterogeneous materials at the scale of their cellular structure (Maire et al., 2003; Babin et al., 2007). In X-ray lCT, the focal spot of the X-ray source is decreased in order to create images with higher spatial resolution than in medical CT. The technique allows accurate reconstructing the internal microstructure (Kerckhofs et al., 2008). In contrast to other microscopy techniques, this non-invasive technique requires no sample preparation and yields highly contrasted images of the full 3D foam structure of the material (Babin et al., 2006). Recently, Lim and Barigou (2004) used X-ray lCT to study the 3D structure of cellular foods such as aerated chocolate, mousse, marshmallows and muffins. Other authors used the technique to study bread crumb (Falcone et al., 2005), bubble growth and foam setting during bread making (Babin et al., 2006) or to elucidate the relationship between texture, mechanical properties and structure of cornflakes (Chaunier et al., 2007), or extruded starches (Babin et al., 2007). There seems to be a consensus that X-ray lCT, coupled with image analysis, is an important tool to study the structure of cellular food products. We here report on the macro- and microscopic changes occurring when sugar or fat levels of the sugar-snap cookie recipe are altered. The macroscopic changes concern the dimensions of the baked cookies, i.e. their diameter and height, their texture, i.e. the surface cracking pattern, and their mechanical properties, i.e. break strength, whereas the microscopic changes deal with the internal structure of the cookies. Parameters used to describe the internal structure include porosity, mean cell size, mean cell wall thickness and their distributions, and the degree of anisotropy. Furthermore, the relationship between the changes in these parameters and the varying mechanical properties, i.e. cookie break strength, was studied. 2. Materials and methods ipe to 8.7%. These changes implied an increase in sugar level from 31.2% to 34.4%. We chose this approach to keep the flour/sugar ratio and dough water level constant. Thus, while the ingredients used in a standard cookie dough recipe were flour (43.3%, containing 14.0% moisture), sugar (31.2%), fat (15.8%), 5.2% water, and 0.9% sodium bicarbonate, those in the dough recipe with the least fat were flour (47.7%), sugar (34.4%), fat (8.7%), 6.2% water, and 1.0% sodium bicarbonate. In a parallel experiment, sugar level was decreased from 31.2% (dough base) for the standard recipe, to 17.6%, which increased the fat level from 15.8% to 19.3%. Margarine and sugar were creamed in a Kitchen Aid Professional KPM5 mixer (Kitchen Aid, St. Joseph, MI, USA) for 3 min with scraping down every min. Then, deionized water was added to adjust dough moisture content to 15.0%. Mixing continued for 2 min with intermediate scraping. At the end, the cream was scraped down. Finally, flour and sodium bicarbonate (2.0%, flour base) were added, followed by mixing for 2 min, with scraping down every 30 s. Dough pieces were slightly flattened with the palm of the hand and laminated with a National Mfg. (Lincoln, NE, USA) sheeter (gap width 6.35 mm). Cookie dough was cut with a circular cookie cutter (inside diameter 63.5 mm). Dough pieces were weighed and immediately baked for 14 min at 185 °C in an electrically heated rotary oven (National Mfg.) on a stamped steel baking tray with baking paper. At least three pans of four cookies were baked per test and these were all used for further quality assessment. Baked cookies were removed from the oven and their diameter was measured (AACC, 1983) after cooling for 30 min. 2.3. Time-lapse photography During baking, time-lapse photography was conducted as described earlier (Pareyt et al., in press). Spread onset time, set time, i.e. the time at which the cookie dough stops spreading, and spread rate were determined directly by plotting the cookie diameter as a function of baking time. Spread rate was the slope of the linear regression equation of the linear portion of the graph (i.e. between the spread onset time and the set time). In addition, cookie height was monitored during baking and made it possible to quantify the collapse. The percentage collapse was calculated as follows: Collapseð%Þ ¼ 2.1. Materials Commercial cookie flour [moisture content: 13.7%, protein content (N x 5.7): 10.7% (db, dry base), ash content 0.66% (db)] was obtained from Meneba NV (Hasselt, Belgium). Commercial sugar (average crystal size: 470 lm) was from Iscal Sugar NV (Moerbeke–Waas, Belgium) and the margarine [moisture content: 18.9%, solid fat contents of 31% and 20% at 20 °C and 25 °C respectively] from Vandemoortele NV (Izegem, Belgium). Sodium bicarbonate (BICARÒ) was from Solvay Chemicals International NV (Brussels, Belgium). Moisture, ash, and protein contents were respectively determined with Approved methods 44–19 and 08– 01 (AACC, 1983) and the Dumas method (Pareyt et al., in press). All reagents were of analytical grade and from Sigma-Aldrich (Steinheim, Germany), unless specified otherwise. 2.2. Cooking making Cookies were made according to Pareyt et al. (in press), with a decrease in fat level from 15.8% (dough base) for the standard rec- 401 maximum dough height ðmmÞ during baking-cookie height ðmmÞ  100 maximum dough height ðmmÞ during baking 2.4. Instrumental texture determination Dough hardness was determined according to Pareyt et al. (in press) by using a Texture Analyser TAXT2i (Stable Micro Systems Ltd., Surrey, UK) equipped with a 5 kg load cell in compression mode with a cylindrical probe (25 mm diameter). Pre and post-test speeds were 2.0 mm/s, while test speed was 1.0 mm/s. Dough pieces, prepared as above and thus sheeted at 6.35 mm and cut with a circular cookie cutter of inside diameter 63.5 mm, were precisely centered and compressed 50%. The probe was held for 10 s at maximum compression. Dough hardness was derived from the positive peak value obtained. Cookie break strength was measured using the three point bending test (Pareyt et al., in press). Pre-test, test and post-test speeds were respectively 2.5 mm/s, 2.0 mm/s and 10.0 mm/s. The peak force, an index of cookie break strength, was measured for at least eight cookies per batch. Standard deviation of cookie dough hardness and cookie break strength did not exceed 10%. 402 B. Pareyt et al. / Journal of Food Engineering 90 (2009) 400–408 2.5. X-ray lCT The system setup (Fig. 1) consisted of an X-ray source, a rotation stage and a radioscopic detector (Maire et al., 2003). When targeting the specimen with an X-ray beam, the X-rays that are not or only partially absorbed by the specimen fall on a specially designed X-ray detector (Babin et al., 2007), resulting in a projection or radiograph, which displays the differences in attenuation at each of the detector-elements of a charge-coupled device (CCD) camera. A complete 3D scan, i.e. a tomographic scan, is made by acquiring a large number of two-dimensional (2D) X-ray absorption radiographs of the specimen over 180° or 360°. This is achieved by rotating the stage on which the specimen is fixed in small steps of angular increments. The rotation has its axis perpendicular to the X-ray beam (Maire et al., 2003; Babin et al., 2007). An appropriate mathematical algorithm (Feldkamp et al., 1984) allows to combine the series of 2D radiographs, and to reconstruct a 3D dataset consisting of a stack of 2D cross-sectional images with a thickness of 1 voxel (Babin et al., 2007). The 3D map of the local absorption coefficients in the specimen results, indirectly, in images of the structure at the scale of the resolution of the setup used (Maire et al., 2003). In this study, we used an in-house assembled X-ray lCT device. A Philips HOMX 161 (Philips, Eindhoven, The Netherlands) X-ray source was used and operated at a voltage of 68 kV and a current of 0.51 mA. To reduce noise, a frame averaging of 32 was applied. The exiting radiation was detected by a 10241024 Adimec MX12P (Adimec, Eindhoven, The Netherlands) CCD camera. A HOMX 161 object manipulator was used to position the sample in the field of view. Cookies were cut with a circular cutter with inner diameter of 45.5 mm to obtain a cylindrical sample, which was then positioned in a plastic cylindrical support device. The samples were scanned over an interval of 0–180° with a 0.5° angular increment, which took about 15 min and comprised 374 2D radiographs. NRecon reconstruction software (Skyscan NV, Kontich, Belgium) was used to reconstruct the lCT images from the radiographs. The program uses a modified Feldkamp algorithm with automatic adaptation to the scan geometry in each microCT scanner. The resulting spatial resolution of the lCT images, defined as twice the voxel size, was 91 lm. 2.6. Image processing The 3 D set of digital images (256 grey levels) were analyzed with ImageJ freeware (Chaunier et al., 2007) to obtain quantitative data. Segmentation, i.e. the conversion of the grey-scale images into black and white images by determining the population assignment (void space or solid material) for each voxel in the image (Mendoza et al., 2007), was performed with Otsu thresholding (Otsu, 1979). A rectangular region of interest of each cookie sample was extracted from the middle part of the scanned region to eliminate any edge effects, artefacts and damage from sample cutting (Babin et al., 2007) and subjected to further image analysis. The CTAn analysis software (Skyscan NV) was used to perform a 3D granulometric analysis and to obtain volumetric mean cell size, mean cell wall thickness and their distributions. Cell sizes and wall thicknesses were calculated as true 3D averages of the local thickness at each voxel respectively representing air or solid (Ulrich et al., 1999). Furthermore, the degree of anisotropy of cells and cell walls, a measure of the 3D asymmetry, i.e. their preferential orientation along a certain axis (Falcone et al., 2005), was calculated based on mean intercept length and Eigen analysis (Odgaard, 1997). Finally, the percentages of closed porosity were determined. The coefficients of variation for mean cell size, mean cell wall thickness as well as for cell and cell wall anisotropy did not exceed 5%. 2.7. Statistical analysis Significant differences (a < 0.05) for several variables, based on at least three individual measurements, were determined by the ANOVA procedure. Pearson’s correlation coefficients (P < 0.05) were calculated with the Statistical Analysis System software 8.1 (SAS Institute, Cary, NC, USA). 3. Results and discussion 3.1. Influence of different sugar and fat levels on cookie dough properties Table 1 lists the dough parameters with varying levels of fat or sugar in the dough recipe. Generally, a decrease in dough piece weight, a measure for its density, was observed when more fat was present in the recipe. Chevallier et al. (2000), hypothesized that fat stabilizes the cells in cookie dough. As a consequence, the observed decrease in density with increasing fat levels is probably the effect of higher levels of air incorporation during dough making when more fat is present. Furthermore, dough piece hardness, possibly due to the same effect, decreased linearly (R2 = 0.92) with increasing fat level in the recipe. The present observations are in line with results of Maache-Rezzoug et al. (1998), and Sudha et al. (2007), who also noticed that cookie dough becomes more firm when its formulation contains lower fat levels. Dough recipe’s sugar level, on the other hand, did not significantly alter dough piece weight, but it did change its hardness, which decreased linearly (R2 = 0.89) with increasing sucrose levels. Several authors (Manohar and Rao, 1997; Maache-Rezzoug et al., 1998) also observed a decrease in cookie dough hardness when more sugar was added. This can be related to an increase in the liquid-like character of the dough when more sucrose is present, since each gram of sucrose dissolved per gram water increases the solvent volume with ca. 0.66 ml (Hoseney, 1994). 3.2. Effect of sugar and fat level on the changes taking place during cookie baking Fig. 1. General setup for X-ray microfocus computed tomography, showing the Xray source (left), the rotation stage with rotation axis perpendicular to the X-ray beam (middle) and the detector (in this case an image intensifier is added). Picture taken from Hirakimoto (2001) and adapted. During baking, an increase in fat level was generally associated with higher spread rate (R2 = 0.90) and decreased duration of the spread (R2 = 0.83) (Table 1). The higher spread rate can probably be related to the increased mobility in the system when the fat melts during baking. Higher fat levels lead to more oil phase in baking. This increases mobility in the system, and, as a consequence, results in a higher spread rate. No clear differences were found in spread onset time or the time at which maximum height was reached during baking. In contrast to the present results, 403 B. Pareyt et al. / Journal of Food Engineering 90 (2009) 400–408 Table 1 Dough parameters and baking characteristics with varying fat (upper part of Table) and sugar (lower part of Table) levels Dough fat level (%) Dough total sugar / total moisture ratio Dough piece weight (g) Dough piece hardness (N) Spread onset time (min) Set time (min) Spread rate (cm/min) Max. height during baking (mm) Time at max. height (min) Collapse (%) 15.8a 2.1 2.1 13.8 2.1 13.1 2.2 12.4 2.2 11.7 2.2 11.0 2.2 10.2 2.2 9.5 2.3 8.7 2.3 7.6 (0.1) 6.6 (0.1) 6.2 (0.4) 6.8 (0.4) 5.9 (0.5) 7.4 (0.9) 6.9 (0.5) 7.5 (0.1) 7.0 (0.1) 7.3 (0.2) 7.4 (0.2) 0.399 (0.004) 0.481 (0.003) 0.451 (0.002) 0.402 (0.025) 0.463 (0.004) 0.341 (0.037) 0.335 (0.004) 0.300 (0.006) 0.316 (0.033) 0.255 (0.007) 0.216 (0.009) nd 14.5 2.8 (0.1) 3.3 (0.3) 2.9 (0.4) 2.9 (0.5) 3.4 (0.6) 3.9 (0.4) 3.5 (0.1) 3.2 (0.1) 3.4 (0.4) 3.6 (0.3) 2.5 (0.2) nd 2.1 11.5 (0.4) 16.5 (1.1) 13.8 (0.8) 18.0 (1.2) 20.1 (0.8) 21.9 (1.0) 27.7 (0.8) 34.7 (1.6) 38.2 (2.3) nd nd 15.1 26.8 (0.7) 26.7 (0.9) 26.5 (0.9) 26.6 (0.8) 26.3 (0.8) 26.8 (0.4) 27.2 (0.6) 27.2 (0.5) 27.9 (0.4) 28.1 (0.7) 28.4 (0.5) 17.0 (0.1) 17.0 (0.1) 17.5 (0.5) 17.5 (0.5) 17.3 (0.2) 17.8 (0.3) 20.3 (0.2) 19.5 (0.5) 20.0 (1.0) 21.5 (0.5) 7.8 (0.5) 8.4 (0.1) 8.3 (0.1) 6.7 (0.1) 7.4 (0.5) 7.9 (0.4) 8.0 (0.4) 7.6 (1.1) 8.2 (0.3) 8.0 (0.3) 57.1 (0.1) 56.2 (0.1) 54.0 (1.3) 52.9 (1.3) 50.9 (0.7) 53.4 (0.1) 57.2 (0.5) 54.3 (1.2) 54.0 (2.3) 53.9 (1.1) Dough sugar level (%) 31.2a 2.1 1.9 28.0 1.9 26.9 1.8 25.7 1.7 24.5 1.6 23.2 1.6 21.9 1.5 19.7 1.4 17.6 1.3 7.6 (0.1) 7.9 (0.2) 7.5 (0.2) 7.2 (0.3) 6.8 (0.6) 7.1 (0.2) 6.6 (0.1) 5.9 (0.1) 5.6 (0.2) 5.7 (0.0) 5.1 (0.3) 0.399 (0.004) 0.434 (0.002) 0.448 (0.026) 0.410 (0.003) 0.434 (0.046) 0.399 (0.021) 0.374 (0.004) 0.412 (0.009) 0.350 (0.005) 0.321 (0.009) 0.310 (0.003) nd 29.1 2.8 (0.1) 3.7 (0.3) 3.7 (0.2) 3.7 (0.5) 3.5 (0.4) 3.8 (0.5) 3.4 (0.4) 3.3 (0.2) 3.1 (0.1) 3.8 (0.3) 3.8 (0.5) nd 2.0 11.5 (0.4) 17.9 (0.7) 31.3 (1.1) 25.2 (1.1) 33.6 (2.1) 44.2 (3.0) 46.2 (3.2) 46.5 (3.0) 45.2 (4.1) 49.3 (3.6) 73.1 (5.4) nd 30.2 26.8 (0.7) 26.7 (0.6) 26.2 (0.5) 26.6 (0.5) 26.8 (0.7) 26.7 (0.6) 26.3 (0.5) 26.5 (0.6) 26.3 (0.4) 26.7 (0.5) 27.2 (0.5) 17.0 (0.1) 17.5 (0.5) 18.0 (0.1) 17.0 (0.1) 18.5 (0.1) 17.8 (0.8) 16.5 (0.5) 16.5 (0.5) 15.8 (0.2) 15.5 (0.1) 7.8 (0.1) 8.2 (0.2) 7.7 (0.6) 7.1 (0.1) 8.0 (0.1) 7.3 (0.4) 6.6 (0.6) 6.2 (0.5) 6.4 (0.4) 6.4 (0.5) 54.7 (0.1) 55.4 (1.3) 55.6 (0.1) 51.2 (0.1) 56.8 (0.1) 50.3 (2.1) 44.2 (1.7) 42.4 (1.7) 35.9 (1.0) 31.6 (0.1) a nd Standard recipe. Standard deviation between brackets. nd = not determined. Chevallier et al. (2000) stated that fat influences the rise during baking by delaying the chemical leavening action. Higher sucrose levels in the cookie dough recipe lead to increased sucrose dissolution during baking. This results in higher quantities of solvent phase, and, as a consequence, in higher spread rates. In the present case, the sugar level not only was linearly correlated with the spread rate (R2 = 0.86), but also with the set time (R2 = 0.93), and, as a consequence, with the difference between set time and spread onset time (R2 = 0.97), i.e. the duration of the spreading during baking. This indicates that the phenomena that induce cookie dough setting, whether or not determined by glass transition phenomena (Slade et al., 1989; Miller et al., 1996), are postponed to higher temperatures when more sugar is present. Furthermore, spread onset time generally increased with sugar levels between 21.9 and 30.2%. This can probably be related to the increasing amount of undissolved sugar crystals delaying the spread onset. Table 1 shows that an increase in sugar level generally leads to higher and longer oven rise during baking, followed by a more pronounced collapse. This would mean that sugar not only delays the action of the chemical leavening (Chevallier et al., 2000), but also influences the degree of vertical expansion during baking. 3.3. Macroscopic and textural changes due to different sugar or fat levels Table 2 shows that cookie weight, much as dough weight (Table 1), decreased linearly with increased fat level in the recipe. Furthermore, increasing fat contents correlated linearly (R2 = 0.98) with increasing cookie diameter, and, as a consequence, with decreasing cookie height. The present observations are in agreement with data obtained by Sudha et al. (2007), but are in contrast to findings by Finney et al. (1950), who noticed no substantial differences in cookie diameter. However, the observed increase in diameter with higher fat levels may probably be explained by the influence of the fat on air incorporation during dough making. The amount of air incorporated influences the viscosity of the system (Jacob and Leelavathi, 2007), which itself affects the cookie spread. Thus, higher levels of fat and the resultant lower dough viscosities logically result in larger cookie diameter. Table 2 shows that final cookie moisture content decreased with fat levels between 8.7% and 13.1% and then remained fairly constant. Cookie break strength increased linearly (R2 = 0.97) when the fat level in the dough recipe decreased. This was in agreement with earlier 404 B. Pareyt et al. / Journal of Food Engineering 90 (2009) 400–408 Table 2 Cookie parameters with varying fat and sugar levels Dough fat level (%) Cookie weight (g) Cookie diameter (mm) Cookie height (mm) Cookie moisture (%) Cookie break strength (N) Density corrected break strength (N.cm3/g) 15.8a 15.1 14.5 13.8 13.1 12.4 11.7 11.0 10.2 9.5 8.7 Dough sugar level (%) 31.2a 30.2 29.1 28.0 26.9 25.7 24.5 23.2 21.9 19.7 17.6 23.5cde 23.2cde 23.0de 23.3cde 22.8e 23.4cde 23.8bc 23.7dc 24.6ab 24.8a 25.2a 90.1a 89.4ab 88.4b 86.5c 86.8c 85.3d 84.7d 84.4d 82.8e 81.5f 79.6g 7.1g 7.3fg 7.4f 8.1e 8.2e 8.5de 8.4de 8.7cd 8.9bc 9.2b 9.9a 2.7b 1.8d 2.2cd 2.8b 2.1cd 2.4bc 2.8b 2.3bc 3.5a 3.5a 3.9a 20.5h 23.8gh 25.6gh 25.3gh 27.4fg 33.1ef 36.2de 39.5cd 45.3bc 45.7ab 51.3a 39.5h 47.0g 50.6f 51.7f 58.3e 68.7d 72.0d 81.1c 88.2b 88.4b 100.3a 23.5ab 23.4b 22.9b 23.3b 23.4b 23.2b 23.0b 23.2b 23.0b 23.5ab 24.2a 90.1a 87.6b 86.0c 85.0d 83.9e 83.1e 80.8f 79.1g 76.7h 73.1i 70.5j 7.1i 7.7h 7.8h 8.0g 8.3f 8.0g 8.8e 9.2d 9.5c 10.1b 10.6a 2.7cd 2.9c 2.8c 3.0c 2.7de 2.2e 2.4de 3.0c 3.0c 3.8b 4.6a 20.5ab 21.1ab 17.9bc 19.0bc 18.9bc 22.9a 20.6ab 19.0bc 19.6abc 17.5bc 16.1c 39.5abc 42.0ab 35.2cde 37.0bcde 36.9cde 43.1a 40.5abc 37.0bcde 37.5abcd 31.5ef 27.4f a Standard recipe. Values with the same letter are not significantly different (a < 0.05). results by Sudha et al. (2007). Correcting the cookie break strength for the density of the material showed the same trend, and indicated that the tenderizing effect exerted by the fat was not only due to a lower density. Table 2 lists the same parameters for cookies baked with different levels of sucrose. Generally, an increase in sugar level went hand in hand with increased spread (Finney et al., 1950; Maache-Rezzoug et al., 1998). Our data showed that cookie diameter increased, and cookie height decreased with the sugar level, both to a larger extent than with the recipe’s fat level. Cookie moisture content decreased from 4.6% in cookies baked from dough with 17.6% sugar to 2.2% in cookies produced from dough contain- ing 25.7% sugar. At all higher sucrose levels, cookie moisture was relatively constant at 2.8%. Cookie break strength (and its density corrected derivative) increased linearly from 17.6% up to 25.7% sucrose (Table 2). Similar results were obtained by Gallagher et al. (2003). 3.4. Effect of different levels of sugar and fat on the surface cracking pattern of cookies The cookie surfaces (Fig. 2) showed more pronounced cracking when less fat, and thus relatively more sugar, was present in the recipes. The same effect was noticed by Finney et al. (1950), who Fig. 2. Photographs of reduced fat cookies showing different surface cracking pattern. Reducing the amount of fat resulted in cookies with more pronounced cracks. The dough fat level is indicated in the photographs. 405 B. Pareyt et al. / Journal of Food Engineering 90 (2009) 400–408 found that increasing width of the cracks went hand in hand with increasing sugar level, and, later, by Sudha et al. (2007), who stated that the surface becomes uneven and shrinks with reduced fat levels. This was related to the recrystallization of sucrose on the cookie surface during baking (Hoseney, 1994), or to the degree of collapse occurring at the end of baking (Slade et al., 1993). However, the cookies in Fig. 2 varied widely in surface cracking pattern, while their collapse was very similar (Table 1). Based on these data, we conclude that the collapse phenomenon is not responsible for cracks in the cookie surface. As the relative sucrose level increased slightly and cracks became more pronounced with decreasing fat levels, this can indicate that, indeed, the recrystallization of sucrose is the main cause for the surface cracking pattern. Therefore, an additional baking experiment of cookies with reduced fat levels but with constant sugar levels was performed. Comparison of the latter cookies (Fig. 3) with those with reduced fat and higher sucrose levels (because of the constant flour/sugar ratio) (Fig. 2) revealed that the latter had a more pronounced surface cracking pattern, which confirmed our earlier findings. Fig. 4 (upper left photograph) shows a typical cookie surface cracking pattern with the highest sucrose level added (31.2%). Furthermore, reducing the dough sugar level resulted in cookies with a very smooth surface (Fig. 4). 3.5. Structural observation of reduced fat and sugar cookies with X-ray lCT Table 3 X-ray microfocus tomography parameters Dough fat level (%) Porosity (%) Mean cell size (lm) Cell anisotropy (%) Mean cell wall thickness (lm) Cell wall anisotropy (%) 15.8a 15.1 14.5 13.8 13.1 12.4 11.7 11.0 10.2 9.5 8.7 Dough sugar level (%) 31.2a 30.2 29.1 28.0 26.9 25.7 24.5 23.2 21.9 19.7 17.6 46.4 42.5 43.2 42.7 44.3 42.4 44.3 41.9 40.7 39.2 40.8 719 646 741 726 695 692 694 699 682 686 643 65.7 66.4 65.6 66.4 66.9 68.7 70.5 70.1 72.3 74.7 75.3 535 534 597 594 538 555 546 605 653 578 555 66.2 66.1 65.6 66.6 68.2 69.8 71.3 71.3 72.5 76.0 75.3 46.4 46.2 43.2 41.6 41.8 41.4 43.4 44.9 42.8 38.1 37.3 719 711 682 611 648 586 607 559 502 454 399 65.7 66.9 67.2 68.0 68.4 69.3 70.4 73.4 78.6 80.0 84.6 535 545 564 547 529 488 462 438 425 422 419 66.2 67.2 67.6 68.6 69.3 70.5 71.2 74.1 79.3 81.0 86.1 a The effects of sugar and fat levels on cookie porosity, mean cell size, mean cell wall thickness, cell and cell wall anisotropy were studied with X-ray lCT (Table 3). Cookie porosity increased (R2 = 0.59) when more fat was present in the dough recipe. This explained the decrease in cookie weight when more fat was used. No influence was found on the closed porosity, i.e. the number of pores that were not interconnected with other pores, which was overall very low (<0.2%) and probably at the border of the detection Standard recipe level. limit. However, this did not impact the conclusions, but showed clearly and, for the first time, that sugar-snap cookies have very open structures. Cookie mean cell size increased linearly (R2 = 0.78) from 8.7% to 14.5% fat added in the dough recipe, which showed that the fat level also affected the swelling of the cells during baking. However, no correlation was found between mean cell Fig. 3. Photographs of cookies with different fat levels and constant sugar levels (31.2% on a dough base). Fig. 4. Photographs of cookies with different sugar levels as indicated in the photographs. Addition of sugar turns the smooth surface into a surface with more pronounced cracks. 406 B. Pareyt et al. / Journal of Food Engineering 90 (2009) 400–408 since the cookie break strength corrected for density was correlated with fat level, the latter influenced the cell wall strength as well. Fig. 5a shows that the relative cell size distribution was not affected by the amount of fat added. The same was noticed for the cell wall thickness distributions (data not shown). These results indicate that the fat captures the air during cookie dough mixing, and that the amount of air incorporated is determined by its level. The combined effects of fat on the number of cells in the dough and the cell size of the baked cookie demonstrate the influence of fat on final porosity. Cell and cell wall anisotropy, both of which are a quantitative indication for the orientation of the structure, both increased linearly (R2 = 0.93) with decreasing fat levels. The cells and cell walls are hence presumably oriented more vertically, since reduced fat levels went hand in hand with less spread. To describe this phenomenon, a conceptual drawing of two dough pieces during baking is presented (Fig. 6a). Based on our observations, the left part of the drawing represents a high fat cookie, whereas the right part represents a reduced fat cookie. Before the baking stage (rectangles 1 in Fig. 6a), size and porosity. From literature it is known that, during baking, once the aqueous phase is saturated, the leavening gas diffuses to pre-existing gas cells, because it cannot create new ones (Hoseney, 1994). Bubble mechanics indeed show that the pressure (P) in a bubble is related to the radius of the bubble (r) and the interfacial tension (c) by the following relationship: P ¼ 2 c=r: Thus, in a system where the interfacial tension c does not change, if r approaches zero, then the pressure P required to start a new bubble is infinite. It follows that a single carbon dioxide molecule cannot create a gas bubble, and that, once the dough aqueous phase is saturated, it must diffuse to a pre-existing gas cell or the atmosphere surrounding the dough piece (Hoseney, 1994). This, combined with the lack of correlation between cell size and porosity, indicates that the total number of cells beaten into the system during dough making also depends on the fat level. No influence was found on mean cell wall thickness. However, as mentioned above, 14 12 frequency (%) 10 8 6 4 2 10 01 .4 1 11 83 .4 9 13 65 .5 6 15 47 .6 4 17 29 .7 1 19 11 .7 8 20 93 .8 6 81 9. 34 63 7. 26 45 5. 19 27 3. 11 91 .0 4 0 cell size (µm) 15.8 % 15.1 % 14.5 % 13.8 % 11.0 % 10.2 % 9.5 % 8.7 % 13.1 % 12.4 % 11.7 % 30.0 25.0 frequency (%) 20.0 15.0 10.0 5.0 20 93 .8 6 19 11 .7 8 17 29 .7 1 15 47 .6 4 13 65 .5 6 11 83 .4 9 10 01 .4 1 81 9. 34 63 7. 26 45 5. 19 27 3. 11 91 .0 4 0.0 cell size (µm) 31.2 % 30.2 % 29.1 % 28.0 % 23.2 % 21.9 % 19.7 % 17.6 % 26.9 % 25.7 % 24.5 % Fig. 5. (a) cell size distributions with varying dough recipe fat level. Dough fat levels are indicated as percentage below the graph, (b) cell size distributions with varying dough recipe sugar level, with dough sugar levels indicated as percentage below the graph. B. Pareyt et al. / Journal of Food Engineering 90 (2009) 400–408 407 Fig. 6. Conceptual drawing showing the cell size orientation in (a) high and reduced fat and (b) high and reduced sugar cookies. For the different cookie systems (1) presents the dough structure, (2) the baking dough, and (3) the final cookie structure. Cookies with low levels of the component under consideration have larger anisotropy values, indicating the upward orientation of the cells. the dough pieces contain considerably round-shaped and relatively homogeneously distributed cells. A smaller number of cells are shown in the reduced fat dough piece. When these dough pieces rise during baking, one can imagine the cells being stretched in a vertical direction, which is illustrated in Fig. 6a (rectangles 2). At the end of baking, cookies with a characteristic final structure (Fig. 6a, rectangles 3) are obtained. Here, a distinction should be made between high and reduced fat cookies. High fat cookies (the left part of Fig. 6a) result in larger spread and, at the same time, lower cookie height. Our data evidently suggest that, during baking, the cells (and cell walls) were first oriented in a rather vertical way. Then, as a result of the spread, some of the cells were elongated horizontally, which resulted in a lower degree of anisotropy, since part of the cells were still oriented upward, another part horizontally, while some intermediately oriented cells (angles of their longitudinal axis between 0° and 90°) were also present. In that manner, a relatively homogeneous structure is obtained. In contrast, low fat cookies (right part of Fig. 6a) showed smaller diameter and larger height after baking, implying that the re-orientation in a horizontal way was far less pronounced than with high fat cookies. As a consequence, this then resulted in higher degrees of anisotropy, as suggested by our data. Adjusting the sugar level in the ingredient bill also influenced the structure of the baked product. Increasing sugar levels resulted in linear increases in porosity (R2 = 0.59), mean cell size (R2 = 0.96), and mean cell wall thickness (R2 = 0.86). Mean cell wall thickness was strongly associated (R2 = 0.72) with the break strength corrected for density for cookies with sugar levels between 17.6% and 25.7%. Fig. 5b shows that sugar level, in contrast to fat level, largely influenced the cell size distribution in cookies. Again, similar results were obtained for cell wall thickness distributions (data not shown). Reducing the recipe’s sugar level did not only decrease the mean cell size, but also led to more homogeneous cell size distributions. As discussed above, sugar, in contrast to fat, did not significantly alter the amount of air incorporated during dough mixing. As no changes in dough density were noted, sugar does not seem to influence the number of cells incorporated in the fat during dough mixing (Fig. 6b, rectangles 1). However, by affecting dough viscosity, sugar can probably influence the cell size distribution during cookie dough making. During baking, sugar progressively dissolves (Hoseney, 1994) and affects dough viscosity to a larger extent than during dough making. The dough viscosity itself determines the cookie spread, and, at the same time, the differences in cell size and cell wall distributions. Lower sugar levels then lead to less rise, and consequently to smaller mean cell size. Cell and cell wall anisotropy increased when sugar levels decreased, or, stated differently, with reduced cookie spread. This confirmed our hypothetical model discussed above (Fig. 6). In conclusion, the data showed that both dough recipe fat and sucrose levels largely influence the macroscopic and microscopic properties of the cookies. The different fat levels used showed the fat to have an effect on the (amount of) air incorporation during cookie dough making. Furthermore, higher air incorporation was probably related to lower dough viscosity, and, as a result, to larger cookie diameters. In addition, higher fat levels reduced cookie break strength, which was not only due to a higher porosity, indicating that fat influences the strength of the cell walls (but not their thickness). X-ray lCT showed that the levels of added fat influenced the final cell size, but had no effect on the distribution of cell sizes and cell wall thicknesses. This indicated that fat served as the target material for the incorporation of air cells. Finally, relating the visual observation of the surface and collapse of the cookies led us to conclude that the cracking pattern is not determined by the degree of structural collapse, but that it is rather a consequence of the cookie’s sucrose level. Additional baking of cookies with reduced fat levels and constant sugar levels confirmed these results, because the surface cracking pattern of these cookies was less pronounced than that of the cookies with more sugar and comparable fat levels in the ingredient bill. The sucrose level had no influence on dough piece density or hardness. During baking, higher sucrose levels resulted in more spread and reduced cookie height. X-ray lCT showed that the cell size and cell wall thickness, and their distributions, were influenced by the level of sugar added, probably because of its effect on the dough viscosity. Higher sugar levels in the recipe resulted in more and longer oven rise. Furthermore, the mean cell wall thickness correlated strongly with the (density corrected) break strength for cookies with sugar levels varying between 17.6% and 25.7%. Finally, cell and cell wall orientation, measured by the degree of anisotropy, were shown to be rather vertically oriented and largely influenced by the final cookie diameter. We proposed a mechanism in which the cells first are oriented vertically due to the leavening action, and then, as a consequence of the final cookie spread, are more or less stretched and become more horizontally orientated. 408 B. Pareyt et al. / Journal of Food Engineering 90 (2009) 400–408 In that manner, high fat/sugar cookies yield lower degrees of anisotropy than reduced fat/sugar cookies. Acknowledgements The authors wish to acknowledge Flanders’ FOOD (Brussels, Belgium) and the Industrial Research Fund (K.U.Leuven, Leuven, Belgium) for financial support. This research is part of the Methusalem programme Food for the Future (2007–2014). Fruitful discussions with Dr R.C. Hoseney (R&R Research Services Inc., Manhattan, Kansas, USA) are highly appreciated. References AACC, 1983. Approved methods of the AACC (8th ed.). AACC International, St. Paul, MN, USA. Babin, P., Della Valle, G., Chiron, H., Cloetens, P., Hoszowska, J., Pernot, P., Reguerre, A.L., Salvo, L., Dendievel, R., 2006. Fast X-ray tomography analysis of bubble growth and foam setting during bread making. Journal of Cereal Science 43 (3), 393–397. Babin, P., Della Valle, G., Dendievel, R., Lourdin, D., Salvo, L., 2007. X-ray tomography study of the cellular structure of extruded starches and its relations with expansion phenomenon and foam mechanical properties. Carbohydrate Polymers 68 (2), 329–340. Chaunier, L., Della Valle, G., Lourdin, D., 2007. Relationships between texture, mechanical properties and structure of cornflakes. Food Research International 40 (4), 493–503. Chevallier, S., Colonna, P., Lourdin, D., 2000. Contribution of major ingredients during baking of biscuit dough systems. Journal of Cereal Science 31 (3), 241– 252. Falcone, P.M., Baiano, A., Zanini, F., Mancini, L., Tromba, G., Dreossi, D., Montanari, F., Scuor, N., Del Nobile, M.A., 2005. Three-dimensional quantitative analysis of bread crumb by X-ray microtomography. Journal of Food Science 70 (4), E265– E272. Feldkamp, L.A., Davis, L.C., Kress, J.W., 1984. Practical cone-beam algorithm. Journal of the Optical Society of America A: Optics Image Science and Vision 1 (6), 612– 619. Finney, K.F., Yamazaki, W.T., Morris, V.H., 1950. Effects of varying quantities of sugar, shortening, and ammonium bicarbonate on the spreading and top grain of sugar-snap cookies. Cereal Chemistry 27, 30–41. Gallagher, E., O’Brien, C.M., Scannell, A.G.M., Arendt, E.K., 2003. Evaluation of sugar replacers in short dough biscuit production. Journal of Food Engineering 56 (2– 3), 261–263. Hirakimoto, A., 2001. Microfocus X-ray computed tomography and its industrial applications. Analytical Sciences 17, i123–i125. Hoseney, R.C., 1994. Principles of Cereal Science and Technology, second ed. AACC International, St. Paul, MN, USA. Jacob, J., Leelavathi, K., 2007. Effect of fat-type on cookie dough and cookie quality. Journal of Food Engineering 79 (1), 299–305. Kerckhofs, G., Schrooten, J., Van Cleynenbreugel, T., Lomov, S.V., Wevers, M., 2008. Validation of x-ray microfocus computed tomography as an imaging tool for porous structures. Review of Scientific Instruments 79 (1), 013711–013719. Lim, K.S., Barigou, M., 2004. X-ray micro-computed tomography of cellular food products. Food Research International 37 (10), 1001–1012. Maache-Rezzoug, Z., Bouvier, J.M., Allaf, K., Patras, C., 1998. Effect of principal ingredients on rheological behaviour of biscuit dough and on quality of biscuits. Journal of Food Engineering 35 (1), 23–42. Maire, E., Fazekas, A., Salvo, L., Dendievel, R., Youssef, S., Cloetens, P., Letang, J.M., 2003. X-ray tomography applied to the characterization of cellular materials. Related finite element modeling problems. Composites Science and Technology 63 (16), 2431–2443. Manohar, R.S., Rao, P.H., 1997. Effect of sugars on the rheological characteristics of biscuit dough and quality of biscuits. Journal of the Science of Food and Agriculture 75 (3), 383–390. Mendoza, F., Verboven, P., Mebatsion, H.K., Kerckhofs, G., Wevers, M., Nicolai, B., 2007. Three-dimensional pore space quantification of apple tissue using X-ray computed microtomography. Planta 226 (3), 559–570. Miller, R.A., Mathew, R., Hoseney, R.C., 1996. Use of a thermomechanical analyzer – Study of an apparent glass transition in cookie dough. Journal of Thermal Analysis 47 (5), 1329–1338. Odgaard, A., 1997. Three-dimensional methods for quantification of cancellous bone architecture. Bone 20 (4), 315–328. Otsu, N., 1979. Threshold selection method from gray-level histograms. IEEE Transactions on Systems Man and Cybernetics 9 (1), 62–66. Pareyt, B., Delcour, J.A., 2008. The role of wheat flour constituents, sugar and fat in low moisture cereal based products: a review on sugar-snap cookies. Critical Reviews in Food Science and Nutrition, in press. Pareyt, B., Wilderjans, E., Goesaert, H., Brijs, K., Delcour, J.A., 2008. The role of gluten in a sugar-snap cookie system: a model approach based on gluten-starch blends. Journal of Cereal Science, in press. Sasov, A.Y., 1987. Microtomography 1 methods and equipment. Journal of Microscopy-Oxford 147, 169–178. Slade, L., Levine, H., Finley, J.W., 1989. Protein–water interactions: water as a plasticizer of gluten and other protein polymers. In: Phillips, R.D., Finley, J.W. (Eds.), Protein Quality and the Effects of Processing. Marcel Dekker, New York and Basel (pp. 9–124). Slade, L., Levine, H., Ievolella, J., Wang, M., 1993. The glassy state phenomenon in applications for the food-industry–Application of the food polymer science approach to structure-function-relationships of sucrose in cookie and cracker systems. Journal of the Science of Food and Agriculture 63 (2), 133–176. Sudha, M.L., Srivastava, A.K., Vetrimani, R., Leelavathi, K., 2007. Fat replacement in soft dough biscuits: Its implications on dough rheology and biscuit quality. Journal of Food Engineering 80 (3), 922–930. Ulrich, D., Van Rietbergen, B., Laib, B., Rüegsegger, P., 1999. The ability of threedimensional structural indices to reflect mechanical aspects of trabecular bone. Bone 25 (1), 55–60. Zoulias, E.I., Oreopoulou, V., Tzia, C., 2002. Textural properties of low-fat cookies containing carbohydrate- or protein-based fat replacers. Journal of Food Engineering 55 (4), 337–342. Zoulias, E.I., Piknis, S., Oreopoulou, V., 2000. Effect of sugar replacement by polyols and acesulfame-K on properties of low-fat cookies. Journal of the Science of Food and Agriculture 80 (14), 2049–2056.