Submitted to: Cereal Chemistry
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 12/21/2005
Publication Date: 5/1/2005
Citation: Wilson, J.D., Bechtel, D.B., Todd, T., Seib, P.A. 2005. Measurement of wheat starch granule size distribution using image analysis and laser diffraction technology. Cereal Chem. 83:259-268. Interpretive Summary: Starch constitutes the greatest weight portion of the wheat endosperm (65-75%) and contributes its own unique functional qualities such as texture, volume, consistency, aesthetics, moisture and shelf stability to various baked products. Wheat gluten proteins have received the greatest amount of attention due to their unique properties of extension and elasticity, which gives them their unique dough forming properties, and are what allows wheat to be such a unique and versatile raw material for so many food products. Cereal starches have been well studied in dilute aqueous systems, the functionality of starch in concentrated water-limiting systems such as that in dough and breads is far from understood. Particle size and shape has long been recognized as an important variable in the efficiency of a range of processes including predicting rheology and flow behavior. While it is known that particle geometry, as well as size distribution, affects the characteristics and behavior of particulate materials, this parameter has always been difficult to characterize in a simple way. Wheat starch is multimodal in distribution and has a unique oblate spheroid (lens) shape making it difficult to accurately size with current methodologies. The present study reports on two methodologies; digital image analysis and laser diffraction sizing. Digital image analysis, performed as modified in this study, is the accurate reference method. This work also shows some inherent differences of digital image analysis compared to laser diffraction sizing. A method to correct these differences in laser diffraction is given so one can use both methods to accurately evaluate particle size distributions. Digital image analysis was evaluated as to measurement error and correction of starch granules touching the edge of field of view. Digital image analysis data were also converted into volume data so it could be compared to distributions measured on four different laser diffraction sizing instruments. An adjustment was developed and applied to laser diffraction data to more closely approximate diameter-volume percent based on oblate spheroid volumes obtained with image analysis for these four wheat classes. The application of digital image analysis and laser diffraction particle sizing for accurately and quickly measuring wheat starch size distribution may be an important development in routine and consistent analysis of starch granule size distributions.
Technical Abstract: Starch was isolated from flour of four wheats representing hard red winter (Karl), hard red spring (Gunner), durum (Belfield 3), and spelt (WK 86035-8) wheat classes. Digital image analysis (IA) coupled to light microscopy was used to determine starch size distributions where the volume of granules were calculated as spherical particles or oblate spheroids. Starch granules were classified into 3 size ranges, A-type granules (>15 um), B-type granules (5-15 um)and C-type granules (<5 um). An error was noted in using digital image analysis because the perimeter of some granules touch the edge (PTE) of the field being analyzed. PTE granules are traditionally treated in IA by eliminating them, and ignoring the errors, or by eliminating half the PTE particles in the calculations. The error is highest for the largest granules and the distribution is skewed towards the smaller sized granules. To correct for this error the PTE granules were manually replaced into the field by measuring their diameters and entering them into the database. The results showed differences in the starch-size distributions between the classes of wheat evaluated, as well as the method of analysis. Another factor found to affect the distribution data was the total number of granules counted per analysis, the "concentration" effect. In general IA of 5000 vs. 1000 granules increased the proportion of A-type granules in a distribution. Four laser diffraction sizing instruments were used to measure granule distributions of the four classes of wheat. The minimum sized granules detected in volume% distribution curves were between 0.4 to 1.2 um, while the largest granules detected were 44 to 62 um, dependent on the instrument and variety of wheat starch. Laser diffraction sizing compared to IA resulted in a ~40% underestimation of the A-type granule diameter and ~50% underestimation of the B-type granule diameter. A correction factor (adjustment) was developed from IA data to correct laser diffraction size analysis. Laser diffraction data correlations before adjustments to image analysis data, ranged from R squared = 0.02 (power of ns) to 0.55 (power of ***). After adjustment, these correlations improved to a range of R squared = 0.72 (power ***) to 0.93 (power of ***) depending on the class of wheat starch evaluated.