Effects of fermentation in saltine cracker production

Cereal Chem. 66(l):610 Yeast was found to be necessary in cracker sponges. Stack weight During the 18-hr sponge fermentation, gas was produced, the pH declined, decreased with increased yeast fermentation because of the decreased and the flour proteins were enzymatically modified. With the total starter dough density. In addition, the cracker cell structure was finer and more system, cracker stack height and stack weight decreased as sponge uniform. The starter slurry inoculated the system with bacteria and was fermentation time increased. At the same time, the texture of the crackers required to decrease the pH. The lower pH allowed the flour proteases to changed from extremely tender and fragile to strong with increasing sponge modify the flour proteins. Stack height decreased when slurry was included fermentation time. The 4-hr rest period following dough-up was essentially in the cracker formula, but the textural strength of the cracker increased. a proof period and allowed equilibration of moisture. Saltine crackers are produced with a procedure that requires a total of approximately 24 hr of fermentation. Pizzinatto and Hoseney (1980a) showed that as fermentation time increased, the strength and the pH of the cracker sponges decreased. The role of a starter system, as developed for laboratory production of saltine crackers by Doescher and Hoseney (1985), is generally ignored in the literature. Piepner (1971) mentioned the use of a "buffer" or sponge, which could be added to cracker sponges to enhance fermentation. Most authors, however, simply mention the variability in sponge pH and cracker quality that arises from the fluctuation in material adhering to the troughs (Johnson and Bailey 1924; Micka 1955; Heppner 1959; Matz 1968, 1984; Smith 1972). The assumption can be made that adventitious bacteria play an important role in the fermentation process (Sugihara 1978). Pizzinatto and Hoseney (1980a) also suggested that the reduction of pH during fermentation brings the sponge to the pH optimum (approximately 4.1) of the native proteolytic enzymes of flour. Wu (1987) showed that the resulting enzymatic action is responsible for the rheological changes in cracker sponges. Salgo (1981) studied wheat proteases and found two enzymes with similar pH optima, 3.8 and 4.2, which were stable in the pH 2.5-5.0 range. Saltine crackers are unique baked products with a peculiar 'Contribution no. 88-163-J. Kansas Agricultural Experiment Station. Presented in part at the AACC 72nd Meeting, Nashville, TN, November 1987. 'Research assistant and professor, respectively, Department of Grain Science and Industry, Kansas State University, Manhattan 66506. This manuscript was prepared for electronic processing. ©1989 American Association of Cereal Chemists, Inc. 6 CEREAL CHEMISTRY texture. This texture, although readily recognizable to consumers, is difficult to describe or define objectively. Most reported instrumental texture evaluations of crackers, biscuits, or other pastry products use methods based on breaking strength. Samples are suspended across a bridge, and the force required to snap the test specimen is recorded (Swartz 1943, Stinson and Huck 1969, Bruns and Bourne 1975, Zabik et al 1979, Katz and Labuza 1981). Katz and Labuza (1981) examined crispness of four different snack foods, including saltine crackers. Samples were equilibrated at one of 10 relative humidities for three weeks before testing. A critical water activity, above which the product was unacceptable, was determined using sensory techniques. A snap test was used to measure cracker texture, with the initial slope of the force deformation curve taken to indicate crispness (a technique also used by Bruns and Bourne 1975). The objectives of this study were to determine the effects of yeast and starter on cracker quality and to determine what changes occur in the system during phases of fermentation. MATERIALS AND METHODS Materials Two commercial cracker flours provided by Lance Inc. and Dixie-Portland Flour Mills Inc., and one flour milled from soft wheat at Kansas State University were used in this study (Table I). These flours were selected to represent the range of proteins and flour qualities available from several commercial sources. Compressed yeast (Anheuser-Busch, St. Louis, MO, or Red Star, Universal Foods, Milwaukee, WI) was aged two to four weeks at 4°C before being used. Minor dry ingredients were supplied by Nabisco Brands, Inc. Hydrogenated vegetable shortening (Crisco, Proctor & Gamble, Cincinnati, OH) was used in the cracker baking. Absorption Determination Optimum absorption was determined by the mixograph procedure described by Rogers and Hoseney (1987). Cracker Baking Crackers were baked using a procedure slightly modified (Rogers and Hoseney 1987) from that developed by Pizzinatto and Hoseney (1 980b) and Doescher and Hoseney (1985). For the study on ingredient effects, the fermentation time remained constant. Sponges were set with flour and water plus yeast alone, yeast plus slurry (control), slurry alone, or no starter at all (flour and water). When "mock" systems were made, 5 ml of the water was held back and sponges were mixed for only 2 min. After the desired yeast fermentation, which varied from 1 to 8 hr, approximately 1.3 ml 85% lactic acid (quantity of acid needed varied slightly between flours) was diluted to 5 ml and mixed into the sponge for 1 min. This was followed by fermentation times varying from 8 to 14 hr. In other experiments, sponge dough-up fermentation times or doughup rest times were varied, holding ingredients and the other fermentation time constant. Gassing Power Samples of cracker sponge and dough were prepared as for baking. A 10-g sample of sponge or a 50-g sample of dough was placed into a half-pint reaction bottle in a gasograph (Coghill Corp., Hayden Lake, ID) (Rubenthaler et al 1980). Cracker Texture Analysis To adjust the water activity before texture analysis, crackers were placed on screens suspended over trays of sulfuric acid solutions diluted with water to produce an equilibrium relative humidity of 20%. The chambers were sealed and crackers were equilibrated for a minimum of five days. Water activity of the samples was measured with a thermocouple psychrometer (Decagon Devices Inc., Pullman, WA). The universal testing machine (UTM) (Instron model 1132) was used in the compression mode for instrumental texture analyses. For breaking strength determinations, the 2-kg load cell was used with a crosshead speed of 5 cm/min and a chart speed of 25 cm/ min. An individual cracker, top side down, was placed over a 2.5-cm bridge. A blunt chisel probe, placed parallel to the direction of the last sheeting, was lowered until the cracker broke. The peak force, or force required to break the cracker, was recorded. This test was designed to simulate the initial force required to bite, or break, a cracker. The 50-kg compression load cell was used for the crushing test, with a crosshead speed of 2.5 cm/min and a chart speed of 50 cm/min. A 1.27-cm diameter, flat, circular probe was centered over the middle docker hole of an individual cracker placed top side down on a flat plate. The cracker was measured for thickness and then compressed to 50% of its original thickness. The curves generated were examined for their general shape, including the angle to the first break, the degree of failure, the average force of the curve, and the final force. This test was designed to simulate the shattering of the cracker in the mouth during chewing with the molars. Organoleptic analysis was also carried out on most samples. RESULTS AND DISCUSSION Cracker Texture Analysis An untrained organoleptic panel was used to survey five brands TABLE I Analytical Data for Flours Used for Cracker Baking Flour % Protein % Ash of commercial crackers. Panel results revealed distinct differences in cracker texture. The first bite of cracker, with the incisors, indicated the hardness/tenderness of the cracker. Continued mastication with the molars indicated the friability of the cracker. To account for those differences, two separate but related cracker texture tests were performed. Typical UTM crushing curves are depicted in Figure 1. A tender, friable cracker (Fig. IA) requires a moderate amount of force to cause the first failure, yet the angle to the first break is steep. After the first failure, the cracker has many minor and major failures. The resistance to force remains fairly constant throughout the test period. A crackerjudged to be tough and pasty, shown in Figure I B, has a lower slope to the first break and continues to build resistance to force throughout the entire testing period. Few, if any, major failures are noted. Effect of Sponge (Starter) Ingredients The sponges made with no yeast or slurry produced doughs that were nonuniform or mottled in appearance and crackers that had uneven puffing. Because of the uneven puffing, the measured stack height (Table II) was not a good guide to the inner structure of the crackers. These crackers had excessive shelling (separation of external layers), poor lamination, poor cell structure, and were very tender. The extreme tenderness made it very difficult to obtain 10 unbroken cells from the sheet of 21 cells. When tested organoleptically, the initial bite was overly tender. However, there was a pastiness that remained around the molars during several compressions. The UTM compression procedure showed a building of resistance, with large variability within the treatment. The crackers baked from sponges containing flour, water, and yeast showed a slight decrease in stack height and a decrease in stack weight when compared with crackers made from sponges containing only flour and water (Table II). Increased gas production and/or increased gas retention would change the TABLE II Effect of Varying Starter on Baking Performancea Sponge Stack Htbc Stack Wtbcd Ht/Wt Starter pH (mm) (g) Ratio Non