Cereal flours are the structure-forming materials in extruded products, comprising a dispersed protein phase, usually in a fibrous form aligned with the extrusion direction, in a continuous starch phase. The extent of fragmentation during extru- sion processing influences the structure formation of the extruded products. In the extruder, cereal flours are subjected to thermomechanical stress that may lead to the depolymerization of biopolymer components, depending on the severity of the extrusion conditions. The review by Porter and Casale (88) on the stress- induced degradation of synthetic polymers emphasizes that as the molecular weight of polymers increases, mechanical energy is stored in the molecule rather than dissipated as heat. The concentration of mechanical energy into a smaller number of bonds results in bond rupture (89). Several investigators report that extruded starch has lower average molecular weights and significantly different molecular weight distribution in comparison to unextruded starch (37, 63, 90–
99). Fragmentation is more significant at high temperatures, high screw speeds, and low extrusion moisture. Colonna et al. (63) compared the weight-average molecular weights (Mw) of pre- and postprocessed starch exposed to extrusion processing and drum drying and reported a decrease in the molecular weight of extruded samples. However, because the temperature during extrusion processing can be as high as 180°C, depolymerization of starch during extrusion processing may occur due to thermal degradation as well as mechanical degradation. To- masik et al. (100) noted that depolymerization occurs in dry starch below 300°C,
resulting in dextrin formation. However, the time required for thermal dextriniza- tion of starch is several order of magnitues longer than the residence time of starch in the extruder during processing.
Macromolecular degradation is also reported for cereal flours during extru- sion processing by several investigators (10, 40, 101–104). Kaletuncá and Bres- lauer (10, 40) monitored the extrusion processing–induced fragmentation by reductions in the extrudate glass transition temperature relative to the glass transi- tion temperature of preextruded flour. These investigators designed their experi- ments so that extrusion-induced fragmentation would be the primary cause ofTg
reduction. The general trend observed was a reduction ofTgas a function of SME for high-amylose and high-amylopectin corn flours and wheat flour (Fig. 6).
Figure 6shows the range of SME values (200–1400 kJ/kg) achieved in a pilot-plant-size twin-screw extruder and the correspondingTgvalues for freeze- dried extrudates of corn and wheat flour (10, 40). Kaletuncá and Breslauer (40) also demonstrated that for wheat flour extrudates the weight-average molecular weight decreases in a trend similar to the decrease in Tg value as a function of SME. Figure 7shows the dependence of the Mw of wheat flour extrudates (determined by gel permeation chromatography) (38) as well as the calorimetri- cally determinedTgon SME (40).
A similar observation of molecular weight reduction with increasing SME for extruded wheat starch was reported by Meuser et al. (96). Using their data,
Figure 6 SME dependence of glass transition temperature for䊉wheat flour,䊐high- amylose corn flour,䊊high-amylopectin corn flour. (Combined from Refs. 10 and 40.)
Figure 7 SME dependence of glass transition temperature and molecular weight for wheat flour. (From Ref. 40.)
these investigators developed a mathematical model predicting the molecular breakdown of wheat starch, quantified as mean molecular weight, as a function of SME. Kaletuncá and Breslauer (40) observed thatTgvalues of extrudates with similar SME values (416 and 432 kJ/kg) but different extrusion processing con- ditions were indistinguishable. Mw for these two extrudates were virtually iden- tical, 3.34⫻ 106 for 432 kJ/kg and 3.41⫻ 106 for 416 kJ/kg (38). These in- vestigators concluded that for a given flour, the extent of extrusion-induced fragmentation is related to the mechanical energy generated in the extruder as quantified by SME. A similar conclusion was reached by Meuser et al. (96); that is, the energy history, primarily SME, causes a significant change in the molecular structure of a given material, in turn defining the characteristic attributes of the final product.
Figure 6also shows theTgand molecular weight values of freeze-dried, unextruded cereal flours corresponding to a zero SME value. Note thatTgof the extrudates is less than that of the corresponding flour, even for the lowest SME value of 236 kJ/kg, indicating fragmentation in the extruder. It was shown in a previous study on molten maize starch (94) that macromolecular degradation first appears when the mechanical energy reaches about 107J/m3(⬃10 kJ/kg). The lowest SME of 236 kJ/kg was reported by Kaletuncá and Breslauer, corresponding to⬃108J/m3, which is above the minimum mechanical energy required to induce fragmentation. At higher SME values, the difference between theTgvalues of wheat flour and its extrudate becomes larger. Input of mechanical energy also
causes damage to starch granules. Significant granule damage in starch is reported when SME is greater than 500–600 kJ/kg, while even below 250–360 kJ/kg, a large fraction of flattened and sheared granules were reported (105, 106).
Starch degradation is investigated using viscometry, gel permeation chro- matography (GPC), and light-scattering techniques. Although the fragmentation of starch molecules during extrusion processing is well established, the extent of molecular fragmentation and the involvement of amylose and amylopectin in fragmentation are not well characterized. While Davidson et al. (92) proposed that molecular degradation in starch is due to debranching of amylopectin mole- cules, Colonna et al. (107), using iodine binding and hydrolysis by β-amylase, determined that the percentage ofα(1–6) linkages did not change as a result of extrusion. Therefore, Colonna et al. (107) concluded that depolymerization of starch molecules is by random chain scission and that amylose and amylopectin have the same susceptibility to degradation. Later, Sagar and Merrill (98) used GPC coupled with light scattering to study the molecular fragmentation of starch.
Their results indicated that amylose also loses branches during extrusion pro- cessing. Several investigators report that depolymerization did not produce sugar monomers, but the end products were composed of shorter-length macromole- cules (63, 67, 103). Viscometry and light scattering are more sensitive to the higher molecular fractions, so contributions from small-molecular-weight frac- tions and from sugar monomers, if any exist, will not be reflected in the average molecular weight. Because the molecular weight resolution by GPC increases at high elution volumes, this technique is not sensitive to very large or very small molecules. Sagar and Merrill (98) discussed extensively the limitations of both GPC and viscometric methods in determining the molecular weight distributions of complex polymer systems, such as starch.
Kaletuncá and Breslauer (10, 40) reported that the reduction ofTgwith in- creasing SME, as assessed by∆Tg/∆SME, differed significantly for various types of flours. They reported∆Tg/∆SME values of 0.05°C/(kJ/kg) for high-amylopec- tin corn flour, 0.09°C/(kJ/kg) for high-amylose corn flour, and 0.06°C/(kJ/kg) for wheat flour. The amylose–amylopectin ratios of these flours were 0, 2.3, and 0.35 respectively. These investigators noted the increased sensitivity ofTgto SME with increasing amylose–amylopectin ratio. They proposed that this observation may indicate a greater stress-induced reduction in Mw values due to greater frag- mentation in high amylose flours than in high-amylopectin flours. They further hypothesized that the breakage of linear amylose chains would produce rela- tively large fragments, which may decrease drastically the Mw of the starch. The breakage in amylopectin may occur at branch points, and the breakage of small- molecular-weight branches may not affect the overall molecular weight of amylo- pectin. This hypothesis is in agreement with the findings of Porter and Casale (88), who reported that the extent of stress-induced reaction depends on the im-
posed strain, chain topology, cross-link density, and chemical composition and also that the junctions of long branches of branched polymers are most susceptible to stress reactions.
As discussed by Kaletuncá and Breslauer (10, 40), SME can be utilized to quantify the extent of applied mechanical stress during extrusion processing.
SME is a processing parameter that combines the effects of equipment (screw speed), process (mass flow rate), and material (viscosity) properties. The viscosity of the flour–water mixture in the extruder is affected by the applied shear, temper- ature, and moisture content. Thus, SME reflects the collective influences of vari- ous operating conditions (e.g., temperature, screw speed, mass flow rate, moisture content, and additives) on the extent of fragmentation. In direct expanded cereals, mechanical energy conversion is important in creating the product structure, be- cause the mean residence time is on the order of seconds. The influence of extru- sion operating variables on the extent of fragmentation is indirect. Each operating variable affects SME (76), and their combined effect, in terms of SME, defines the extent of fragmentation.