close

Вход

Забыли?

вход по аккаунту

?

High efficiency planar microwave antennas assembled using millimetre thick micromachine polymer structures

код для вставкиСкачать
UNDERSTANDING AND CHARACTERIZING FLAKE POLYMORPHISMS AS A
QUALITY ATTRIBUTE FOR MICROWAVE POPCORN
by
Jess C. Sweley
A DISSERTATION
Presented to the Faculty of
The Graduate College at the University of Nebraska
In Partial Fulfillment of Requirements
For the Degree of Doctor of Philosophy
Major: Food Science and Technology
Under the Supervision of Professors David S. Jackson and Devin J. Rose
Lincoln, Nebraska
May, 2012
UMI Number: 3502358
All rights reserved
INFORMATION TO ALL USERS
The quality of this reproduction is dependent on the quality of the copy submitted.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.
UMI 3502358
Copyright 2012 by ProQuest LLC.
All rights reserved. This edition of the work is protected against
unauthorized copying under Title 17, United States Code.
ProQuest LLC.
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, MI 48106 - 1346
UNDERSTANDING AND CHARACTERIZING FLAKE POLYMORPHISMS AS A
QUALITY ATTRIBUTE FOR MICROWAVE POPCORN
Jess C. Sweley, Ph.D.
University of Nebraska, 2012
Advisers: David S. Jackson, Devin J. Rose
Popcorn (Zea mays everta) is a familiar snack food with sizable commercial
popularity. The shape of popped popcorn (i.e., flakes) is a quality factor that has not been
comprehensively defined or quantified. Thus, the objective of this work was to investigate
the importance of flake polymorphisms in microwave popcorn by (1) identifying and
characterizing differences between flake polymorphisms; (2) investigating factors influencing
formation of different flake shapes; and (3) determining how flake shape affects consumer
acceptability and relates to other quality attributes for popcorn, especially expansion volume.
Popcorn flake polymorphisms were identified by visual inspection based on
whether appendages were expanded unilaterally, bilaterally, or multilaterally. When
popcorn flakes were isolated after microwave popping, it was shown that unilaterallyexpanded flakes had the greatest amount of total fat and sodium and highest overall
product liking in consumer testing, while multilaterally-shaped flakes had the highest
quantities of popcorn-like aromatic pyrazines, lowest content of total fat and sodium, and
lowest overall product liking.
Formation of different flake polymorphisms is complex and depends on both
intrinsic and external factors, as well as interaction effects. Varying the relative proportion
of different flake polymorphs in microwave popping is theoretically possible by selecting
the optimum hybrids, growing location and environment, corn-oil ratio, and microwave
wattage.
Empirical modeling revealed both flake size and shape have a significant (p<0.05)
effect on expansion volume and popcorn packing characteristics. Bilateral and
multilaterally-shaped flakes produce greater expansion volumes than unilateral shapes.
In a standard movie-theatre style tub, small unilateral flakes pack together most tightly
and require 3x as many pieces to fill the package as large bilateral and multilateral flakes.
The results of this research indicate flake polymorphisms are an important
quality-attribute for popcorn that can be optimized to support new product development
and differentiated usage occasions. The development of new hybrids or techniques to
produce popcorn with the most desirable flake shapes can impact both consumer
satisfaction and economic profitability for vendors of popped popcorn.
iv
ACKNOWLEDGMENTS
I am indebted to many people for their support of this work. My sincerest
gratitude goes to Drs. David Jackson and Devin Rose for serving as co-advisors and
providing encouragement and guidance to me throughout this PhD program. Thanks to
Drs. Randy Wehling and Tom Hoegemeyer for serving as reading committee members
and providing council during my research. My appreciation also goes to Drs. Rolando
Flores and Gordon Smith for serving on my supervisory committee and nurturing the
industry relationships that allowed me to pursue these studies.
In-kind support was provided by ConAgra Foods to support this research, and I
am grateful to several colleagues whose support enabled this academic endeavor while
continuing my employment at ConAgra. I am particularly grateful to Mike Meyer for his
help with the experimental designs and statistical analysis. Additionally, I would like to
thank Mike Jensen, Mike Van Natta, Tabra Ward, and Indarpal Singh for sharing their
time and industrial knowledge of popcorn and related materials. A special note of
appreciation goes to Al Bolles, Corey Berends, and Rich McArdle who provided endless
support and afforded the flexibility required to complete this work.
Lastly, I would like to thank my wife and best friend, Autumn, and our four
children, Agnes, Jonah, Henry, and Greta, for their constant support, patience, and
personal sacrifice.
v
TABLE OF CONTENTS
CHAPTER 1.
LIST OF TABLES ........................................................................................................... viii
LIST OF FIGURES .............................................................................................................x
INTRODUCTION ...............................................................................................................1
Objectives and Hypotheses ..............................................................................................3
Organization .....................................................................................................................4
CHAPTER 1. QUALITY TRAITS AND POPPING PERFORMANCE
CONSIDERATIONS FOR POPCORN (ZEA MAYS EVERTA) ....................................... 6
1.1. Abstract .....................................................................................................................6
1.2. Introduction ...............................................................................................................6
1.3. Popcorn Kernels ........................................................................................................8
1.3.1. Kernel Botanical Traits ...................................................................................... 8
1.3.2. Kernel Composition ........................................................................................... 9
1.3.3. Kernel Classification ....................................................................................... 10
1.4. Popping Mechanics .................................................................................................11
1.4.1. Heating............................................................................................................. 11
1.4.2. Nucleation and Popping................................................................................... 13
1.4.3. Expansion ........................................................................................................ 13
1.4.4. Final Popped Structure .................................................................................... 15
1.5. Popcorn Quality Traits ............................................................................................16
1.5.1. Expansion Volume .......................................................................................... 16
1.5.2. Unpopped Kernels ........................................................................................... 18
1.5.3. Hull Dispersion ................................................................................................ 19
1.5.4. Shape of Popped Flakes ................................................................................... 20
1.5.5. Sensory and Nutritional Traits ......................................................................... 20
1.6. Factors Affecting Popping Performance .................................................................21
1.6.1. Hybrid, Environment, and Agronomic Inputs ................................................. 21
1.6.2. Conditioning and Kernel Moisture Content .................................................... 22
1.6.3. Kernel Physical Characteristics ....................................................................... 23
1.6.4. Pericarp Thickness and Kernel Damage .......................................................... 25
1.6.5. Kernel Composition ......................................................................................... 26
1.6.6. Heating Method ............................................................................................... 27
1.6.7. Ingredient Additions ........................................................................................ 28
1.6.8. Atmospheric Pressure ...................................................................................... 29
1.6.9. Microwave Power and Packaging ................................................................... 29
1.7. Conclusions .............................................................................................................30
1.8. References ...............................................................................................................31
vi
CHAPTER 2. COMPOSITION AND SENSORY EVALUATION OF POPCORN
FLAKE POLYMORPHISMS FOR A SELECT BUTTERFLY-TYPE HYBRID........... 47
2.1. Abstract ...................................................................................................................47
2.2. Introduction .............................................................................................................48
2.3. Materials and Methods ............................................................................................50
2.3.1. Materials .......................................................................................................... 50
2.3.2. Popping and Sorting ........................................................................................ 51
2.3.3. Compositional and Flavor Analysis of Popped Kernels .................................. 52
2.3.4. Sensory Protocol and Evaluation ..................................................................... 54
2.3.5. Data Analysis ................................................................................................... 55
2.4. Results and Discussion ...........................................................................................56
2.4.1. Popping Performance by Flake Morphology................................................... 56
2.4.2. Compositional and Flavor Analysis of Popped Kernels .................................. 57
2.4.3. Sensory Evaluation .......................................................................................... 59
2.4.4. Relationship between Composition and Sensory Measures ............................ 60
2.5. Conclusions .............................................................................................................61
2.6. Literature Cited .......................................................................................................61
CHAPTER 3. HYBRID AND ENVIRONMENT EFFECTS ON POPCORN KERNEL
PHYSIOCHEMICAL PROPERTIES AND THEIR RELATIONSHIP TO
MICROWAVE POPPING PERFORMANCE ................................................................. 77
3.1. Abstract ...................................................................................................................77
3.2. Introduction .............................................................................................................78
3.3. Experimental ...........................................................................................................79
3.3.1. Popcorn Samples ............................................................................................. 79
3.3.2. Composition and Physical Property Testing ................................................... 80
3.3.3. Pop Performance Testing................................................................................. 81
3.3.4. Statistical Analysis .......................................................................................... 84
3.4. Results and Discussion ...........................................................................................84
3.4.1. Kernel Physicochemical Parameters ............................................................... 84
3.4.2. Hybrid and Environment Influence on Kernel Properties ............................... 87
3.4.3. Microwave Popping Performance ................................................................... 89
3.5. Conclusions .............................................................................................................93
3.6. References ...............................................................................................................94
CHAPTER 4. EFFECTS OF HYBRID, ENVIRONMENT, OIL ADDITION, AND
MICROWAVE WATTAGE ON POPPED POPCORN MORPOLOGY ...................... 105
4.1. Abstract .................................................................................................................105
4.2. Introduction ...........................................................................................................106
4.3. Experimental .........................................................................................................107
4.3.1. Popcorn Samples ........................................................................................... 107
4.3.2. Test Design and Microwave Popping ............................................................ 108
4.3.3. Statistical Analysis ........................................................................................ 109
vii
4.4. Results and Discussion .........................................................................................111
4.4.1. Relationship of Kernel Properties to Popped Morphology ........................... 111
4.4.2. Predictive Models for Popped Morphology .................................................. 112
4.4.3. Relationship of shape to expansion volume .................................................. 115
4.5. Conclusions ...........................................................................................................116
4.5. References .............................................................................................................117
CHAPTER 5. PACKING CHARACTERISTICS OF POPPED POPCORN WITH
DIFFERENT FLAKE SHAPES AND SIZE .................................................................. 130
5.1. Abstract .................................................................................................................130
5.2. Introduction ...........................................................................................................130
5.3. Materials and Methods ..........................................................................................132
5.3.1. Popcorn Samples ........................................................................................... 132
5.3.2. Flake Size and Shape Determination ............................................................. 133
5.3.3. Test Design and Packing Experiments .......................................................... 134
5.3.4. Movie Theatre-Style Packaging Application ................................................ 136
5.3.5. Data Analysis ................................................................................................. 137
5.4. Results and Discussion .........................................................................................138
5.4.1. Flake Size and Shape ..................................................................................... 138
5.4.2. Displacement Testing .................................................................................... 138
5.4.3. Packing Performance ..................................................................................... 140
5.4.4. Relationship of Flake Properties to Packing Characteristics ......................... 141
5.4.5. Flake Size and Shape Effects on Expansion Volume and Packing Density .. 142
5.4.6. Packing Performance in Movie Theatre-Style Packaging ............................. 143
5.5. Conclusions ...........................................................................................................144
5.6. Literature Cited .....................................................................................................145
GENERAL CONCLUSIONS ..........................................................................................158
Conclusions ..................................................................................................................158
Significance of Findings and Suggestions for Future Research ..................................160
APPENDICES .................................................................................................................162
Appendix A. Sensory Testing Protocol and Ballot .....................................................163
Appendix B. Select SAS Code.....................................................................................165
B.1. Mean and Mean Difference Testing ................................................................ 165
B.2. Correlation Procedure ...................................................................................... 165
B.3. PCA of Flavor Volatiles................................................................................... 166
B.4. Regression Modeling for Shape Formation ..................................................... 167
B.5. Modeling Flake Size and Shape Effects on Expansion Volume and Packing
Density ..................................................................................................................... 169
viii
LIST OF TABLES
Table 1.1. Quality Traits for Popped Popcorn .................................................................. 44
Table 1.2. Inherent Kernel Factors Affecting Popping Performance of
Popcorn ............................................................................................................ 45
Table 1.3. Extrinsic Variables Affecting Popping Performance of Popcorn .................... 46
Table 2.1. Composition of Butterfly Popcorn Hybrid ...................................................... 68
Table 2.2. Microwave Popcorn Formulation .................................................................... 69
Table 2.3. Distribution and Pop Performance Measures of Three Popcorn
Flake Polymorphisms and Unpopped Kernels Produced After
Microwave Popping ......................................................................................... 70
Table 2.4. Chemical Composition of Three Popcorn Flake Polymorphisms
After Microwave Popping ............................................................................... 71
Table 2.5. Mass Balance Comparison of Raw Materials to Average Popcorn
Flake Composition ........................................................................................... 72
Table 2.6. Volatiles Obtained from Three Popcorn Flake Polymorphisms
After Microwave Popping ............................................................................... 73
Table 2.7. Consumer Sensory Mean Scores for Three Popcorn Flake
Polymorphisms ................................................................................................ 75
Table 2.8. Attribute Intensity Means Scores for Three Popcorn Flake
Polymorphisms ................................................................................................ 76
Table 3.1. Physical Properties of Three Popcorn Hybrids Grown in Three
Different Environments ................................................................................. 100
Table 3.2. Biochemical Composition of Three Popcorn Hybrids Grown in
Three Different Environments ....................................................................... 101
Table 3.3. Correlation coefficients between popcorn kernel physical and
biochemical properties ................................................................................... 102
Table 3.4. Pop Performance Measures for Three Popcorn Hybrid Grown in
Three Different Environments ....................................................................... 103
Table 3.5. Correlations Between Popcorn Kernel Physiochemical Parameters
and Unpopped Kernels and Expansion Volume ............................................ 104
Table 4.1. Table of Experimental Design ....................................................................... 123
ix
Table 4.2. Correlation Coefficients between Popcorn Kernel Physiochemical
Properties and Popped Flake Shape ............................................................... 124
Table 4.3. Estimated Coefficients of Fitted Equation for %Unilateral Shape ................ 125
Table 4.4. Analysis of Variance for Unilaterally-Expanded Popcorn Flakes ................. 126
Table 4.5. Estimated Coefficients of Fitted Equation for Bilateral and
Multilateral Flakes ......................................................................................... 127
Table 4.6. Analysis of Variance for Bilateral and Multilateral Flake Shapes ................ 128
Table 4.7. Estimated Coefficients in Universal Model for %Unilateral,
%Bilateral and %Multilateral Flake Shapes .................................................. 129
Table 5.1. Average Dimensions for Three Shapes of Popped Popcorn
Measured by Digital Calipers ........................................................................ 152
Table 5.2. Pop Volume and Packing Characteristics for Different Sizes and
Shapes of Popped Popcorn ............................................................................ 153
Table 5.3. Correlation Coefficients between Some Popcorn Flake Properties
and Packing Characteristics ........................................................................... 154
Table 5.4. Analysis of Variance for Expansion Volume and Packing Density .............. 155
Table 5.5. Parameter Estimates for Models of Expansion Volume and
Packing Density ............................................................................................. 156
Table 5.6. Popcorn Flake Packing in a Standard Movie Theatre-Style
Packaging Tub ............................................................................................... 157
x
LIST OF FIGURES
Figure 1.1. Sales of unpopped popcorn 1970 – 2009. ...................................................... 41
Figure 1.2. Various popcorn flake polymorphisms identified by qualitative
observation.. .................................................................................................... 42
Figure 1.3. Scanning electronic microscopic images of various popcorn flake
polymorphisms viewed at 200x and 500x resolution. .................................... 43
Figure 2.1. Popcorn flakes inspected and segregated into three shapes:
expanded unilaterally (A), bilaterally (B), or multilaterally (C). ................... 66
Figure 2.2. PCA of flavor volatiles measured by GC-MS grouped among
similar popcorn shapes and differentiation between different flake
polymorphs ..................................................................................................... 67
Figure 3.1. Oil amount lost to bag during microwave popping ........................................ 99
Figure 4.1. Categorization of popped popcorn shapes from butterfly-type
hybrids. ......................................................................................................... 120
Figure 4.2. Predictive distribution of popcorn shapes %unilateral, %bilateral,
and %multilateral polymorphs ...................................................................... 121
Figure 4.3. Correlation between expansion volume and %uniltateral,
%bilateral, and %multilateral flakes ............................................................. 122
Figure 5.1. Average flake size for 400 popped popcorn flakes.. .................................... 149
Figure 5.2. Popcorn flake shapes and sizes delineated for packing
experiments. .................................................................................................. 150
Figure 5.3. Visual depiction of apparatus used for measuring popcorn
expansion volume and packing characteristics (void volume,
packing density) using sand displacement. ................................................... 151
1
INTRODUCTION
Popcorn is a familiar snack food with a sizable commercial business. The most
studied and reported quality factor for popcorn is expansion volume, which is measured
and reported as the bulk volume of popped flakes produced by a known weight of
unpopped kernels or divided by the weight of popped flakes (Dofing et al 1990,
Mohamed et al 1993). Tender and fluffy popcorn associated with high expansion volume
has the dual benefit of being more desired by consumers (Lyerly 1942, Levy 1988,
Ceylan and Karababa 2001), and producing greater profits for commercial venues like
movie theatres where popcorn is purchased by weight and sold by bulk volume (Song et
al 1991, Hoseney et al 1983). Other quality factors often considered in popcorn include
the number or percentage of unpopped kernels, hull dispersion, and the size, shape, color,
and texture of the popcorn flake (Ziegler 2001).
The butterfly-type hybrids used for microwave popcorn have previously been
shown to produce several distinct shapes or polymorphisms when popped (Ziegler 2001).
The visual perceptions of a food product clearly influence the sensory properties of eating
(Hutchins 1977), and it is common for some individuals to scavenge through a bowl of
popcorn to handpick particular shaped flakes. Thus, the shape of popcorn flakes may be
a quality attribute of considerable interest, yet the importance of popcorn polymorphisms
have never been thoroughly investigated or characterized.
The need for elucidating innovative quality attributes for popcorn comes at an
important time. The US popcorn industry had enjoyed continual growth from 1950
through the early 1990s due primarily to marketing successes and the development of
pre-packaged microwavable popcorn that could be easily prepared (Ziegler 2001).
2
During the period of growth, popcorn research by academics, plant breeders, and industry
focused primarily on improving pop expansion and reducing unpopped kernels (Song et
al 1991). This resulted in dramatic improvements in popcorn quality, with the best
popcorn today popping nearly twice as large and with 75% less unpopped kernels than
just a half-century ago (Foer 2005). However, growth of the popcorn industry peaked in
1993, and consumption has been slowly eroding for the past two decades (The Popcorn
Board 2011).
At the same time, the snack foods industry as a whole is facing increased pressure
from industry and regulatory groups to deliver more wholesome snacks. For example,
the 2010 Dietary Guidelines for Americans released by the USDA and Department of
Health and Human Services identified a need for increased consumption of whole grain
as a key opportunity for dietary improvement in both healthy populations, as well as
individuals at risk for chronic disease, since diets high in whole grain have been found to
help with diabetes and weight management and may help reduce the risk of heart disease
and certain cancers. However, according to the 2010 Dietary Guidelines, less than 5% of
Americans consume the minimum recommended amount of whole grains, which for
many is about 3 servings (48 grams) per day (USDA 2010).
Increasing consumption of popcorn, especially as a percentage of snacking
occasions, seems to offer an opportunity to bridge the gap in whole grain consumption
since popcorn is a 100% whole grain food that is convenient, tasty, and economical.
Certainly, marketing efforts by industry to promote the inherent health aspects of popcorn
are part of the solution, although new research efforts to identify and characterize
innovative quality attributes for popcorn that are meaningful to consumers – such as the
3
shape of popped flakes – would provide insights that could be used to further improve the
acceptability of current products or enable new product development and usage occasions
for popcorn.
Objectives and Hypotheses
The overall objective of this research was to develop an understanding of flake
polymorphisms produced from butterfly-type popcorn hybrids and characterize their
importance as an end-use quality attribute. The central hypotheses were:
1. Popcorn produces distinct shapes, or polymorphisms, during popping that can
be identified, characterized, and differentiated.
2. Formation of popcorn flakes depends on both intrinsic and external factors, and a
wide distribution of flake polymorphs can be produced by selecting the optimum
hybrid, growing location and environment, corn-oil ratio, and microwave
wattage.
3. Popcorn flake shape influences sensory hedonics and consumer acceptability
of microwave popcorn, and flake shape affects popcorn packing properties
and expansion volume measurements.
Three specific objectives were developed pursuant to the goal of understanding
and characterizing the importance of flake polymorphisms:
1. To define shapes of popcorn flakes, using qualitative visual identification,
physiochemical and flavor analysis, and microscopic imaging.
2. To determine factors that influence formation of different flake polymorphisms
by developing statistical models that describe both hybrid and environment
4
effects and the role of external factors (microwave wattage, corn-oil ratio) on the
formation of flake shape.
3. To determine the importance of flake shape on consumer liking, expansion
volume and packing efficiency.
Organization
This dissertation is a compilation work that includes a literature review (Chapter
1) pertaining to the research described here as four journal articles (chapters 2-5),
followed by general conclusions. All chapters are presented using guidelines required for
submittal to various refereed, scientific journals, excepting that annotations used for
section headings, tables, and figures have been standardized in this document. Chapter 1
has been formatted using guidelines of Food Reviews International, Chapter 2 and
Chapter 5 for Cereal Chemistry, and Chapters 3-4 for Journal of Cereal Science.
References can be found at the end of each chapter, and follow the format of the
aforementioned journals. It is noted that the material presented in chapters 2 and 3 has
already been published, and other material is expected to be published in the near future.
References
Ceylan, M., and Karababa, E. 2001. Comparison of sensory properties of popcorn from
various types and sizes of kernel. J. Sci. Food Agric. 82:127-133.
Dofing, S. M., D‟Croz-Mason, N., and Buck, J. S. 1990. Inheritance of expansion
volume and yield in two popcorn X dent corn crosses. Crop Sci. 31:715.
Foer, J., 2005. Arming nature's grenade. Discover Magazine 26, 24-25.
5
Hoseney, R. C., Zeleznak, K., and Abdelrahman, A. 1983. Mechanism of popcorn
popping. J. Cereal Sci. 1:43-52.
Hutchins, J. B. 1977. The importance of visual appearance of foods to the processor and
the consumer. J. Food Quality. 1(3):267-278.
Levy, B. 1988. A new perspective on popcorn. Snack World. 45:24.
Lyerly, P. J. 1942. Some genetic and morphological characters affecting the popping
expansion of popcorn. J. Am. Soc. Agron. 34:986.
Mohamed, A. A., Ashman, R. B., and Kirleis, A. W. 1993. Pericarp thickness and other
kernel physical characteristics relate to microwave popping quality of popcorn. J.
Food Sci. 58(2):342-346.
Song, A., Eckhoff, S. R., Paulsen, M., and Litchfield, J. B. 1991. Effects of kernel size
and genotype on popcorn popping volumes and number of unpopped kernels. Cereal
Chem. 68(5):464-467.
The Popcorn Board, http://www.popcorn.org (accessed November 26, 2011).
US Department of Agriculture and US Department of Health and Human Services.
Dietary Guidelines for Americans, edition 7; US Government Printing Office:
Washington D.C., 2010.
Ziegler, K. E. 2001. Popcorn. Pages 199-234 in: Specialty Corn. A. Hallauer, ed. CRC
Press: Boca Raton, FL.
6
CHAPTER 1. QUALITY TRAITS AND POPPING PERFORMANCE
CONSIDERATIONS FOR POPCORN (ZEA MAYS EVERTA)
1.1. Abstract
Popcorn is a snack food with significant commercial popularity. Popcorn popping
mechanics can be described by a series of polymeric transformations. The most
important quality traits for popcorn are expansion volume and “eatability” factors
including unpopped kernels, hull dispersion, and the color, texture, and flavor of popped
flakes. Popcorn quality depends on both intrinsic factors – such as hybrid selection,
kernel conditioning, and kernel physiochemical attributes – and extrinsic variables
including popping method and ingredient additives. Developing new technologies and
establishing new quality attributes for popcorn may help to further increase consumer
liking and consumption.
1.2. Introduction
Popcorn (Zea mays Everta) has been enjoyed for centuries and has significant
commercial importance today. While both popcorn and conventional corn belong to the
same genus and species, Zea mays, the primary difference with popcorn is that upon
heating to sufficient temperature, the caryopsis expands with an explosive sound and a
large expansion of volume up to 30 times the volume of the original kernel.(1)
The use of expanded corn kernels is ancient.(2) Popcorn is among the earliest
cereal-based snack foods consumed by humans,(3) although limited information about
popcorn was written before 1880, and it is likely that popcorn was only grown as a
garden crop for home consumption until that time.(4) The development of popcorn into a
7
commercial industry in the United State (US) began with the introduction of popcorn as a
commercial crop in Iowa in 1885, and the subsequent unveiling of a mobile popcorn
popper by Charles Cretor at the World‟s Columbian Exposition in Chicago in 1893,
which enabled popcorn to be made and sold at any park, fair, or street corner.(5) During
the 1930s and 1940s, popcorn consumption tripled in the US because popcorn was both
inexpensive and abundantly available.(6)
The US popcorn industry enjoyed nearly continual growth from 1950 through the
early 1990s, due primarily to the successful marketing of popcorn as an accompaniment
to watching television at home, and by the eventual development in the mid-1980s of prepackaged microwavable popcorn that could be easily prepared.(4) During this period of
growth, popcorn research by academics, plant breeders, and industry focused efforts on
increasing expansion volume and reducing unpopped kernels, which has led to a doubling
in pop volume with 75% fewer unpopped kernels today compared with 50 years ago.(7)
Despite consistent growth of popcorn sales from 1970 to 1993, consumption has
slowly declined in recent years (Fig. 1.1). Research aimed at enhancing our
understanding of established popping performance measures and elucidating new quality
attributes for popcorn may support the development of improved or differentiated
products that can reverse this declining consumption trend. Thus, the purpose of this
review is to discuss current knowledge of popcorn quality and identify new research
areas that may contribute to a new “boom” in popcorn production and consumption.
8
1.3. Popcorn Kernels
Popcorn kernels are visually distinct and comparatively smaller than other types
of corn. In addition, popcorn kernels have several unique botanical and compositional
features.(8)
1.3.1. Kernel Botanical Traits
Popcorn kernels are the caryopses, or single-seeded indehiscent fruit of the
plant.(9) Popcorn kernels contain all three botanical parts of the caryopsis: germ,
endosperm, and an outer bran layer also called pericarp or hull.(10) Mature popcorn
kernels also contain a tip cap which attaches the kernel to the cob, but the tip cap may or
may not be retained during handling and is not considered an integral part of the
caryopsis. The germ is the botanical part of the popcorn kernel that becomes the plant
upon seed germination, and it is comprised of proteins, sugars, and is the primary source
of lipids in popcorn kernels. (10)
Popcorn endosperm contains both translucent (horny or “hard”) endosperm,
which lies around the crown and outer edges of the kernel, and opaque (floury or “soft”)
endosperm, which surrounds the germ at the base and center of the kernel.(10)
Translucent endosperm is composed of polygonal shaped starch granules with diameters
8-17μm embedded in a protein matrix, whereas opaque endosperm is comprised of
spherical starch granules of similar size, which are covered by protein films (matrix) and
separated by air spaces.(10) Popcorn contains a larger amount of translucent endosperm
and a higher protein:carbohydrate ratio than conventional dent corn.(11)
The pericarp of popcorn provides protection and containment of the endosperm,
and it acts as a pressure vessel during heating, which gives popcorn its distinct ability to
9
pop.(10,12) Popcorn pericarp varies in thickness from 40 to 120 μm depending on the
variety, location of kernel on the ear, and the position of the pericarp on the kernel.(13)
While the percentage of pericarp by weight is about the same in all types of corn,
popcorn contains the thickest pericarp among all corns, due to the smaller size of popcorn
kernels.(13,14)
1.3.2. Kernel Composition
Like all cereal grains, mature popcorn kernels are composed of protein, oils,
and carbohydrates. The actual percentages of these components will vary depending on
both the plant‟s genetics and its growing environment. Compared to normal dent corn,
popcorn tends to have similar starch and oil contents but higher protein content.(8) In
addition, the starch found in popcorn tends to have lower gelatinization temperatures,
peak temperatures, and pasting temperatures than normal dent corn starch.(8)
Park et al.(15) analyzed and reported extensive data on popcorn kernel composition
of six hybrids from Colorado and Nebraska. While popcorn hybrid was shown to be a
significant factor for all reported analytes, on average the popcorn hybrids contained
4.2% fat, 9.9% protein, and 64.0% starch with an amylose:amylopectin ratio of 27.5:72.5.
In addition, the most predominant fatty acids in the popcorn hybrids were linoleic and
oleic acids, accounting for an average of 58.4% and 25.5% of total fatty acids,
respectively.(15) Similar composition results were reported by Borras et al.(16) for
commercial popcorn kernels from Argentina.
Additional compositional analysis of several US commercial popcorn hybrids has
also shown that popcorn kernels contain 10.1–11.4% dietary fiber.(17) The fiber found in
10
popcorn is derived from the pericarp, which is composed of hemicelluloses (~69%),
cellulose (~21%), and lignin (~1%).(18)
1.3.3. Kernel Classification
Specific traits are used to classify different types of commercial popcorn,
including kernel size, shape, and color, as well as appearance after popping. While there
are no industry standards for kernel size, it is common to sort and classify kernel size
based on the number of kernels per 10 grams with “large” = 52-67 kernels, “medium” =
68-75, and “small” = 76-105 kernels.(19) Kernels can also be described by the shape of
the native (unpopped) kernel as rice or pearl types. The rice types have long kernels with
a sharp point at the top, while pearl types are more spherical and do not have sharp points
at the top.(20) Pearl-type popcorn kernels are the predominant types found in US
commercial production today.(4) The two main color types of commercial popcorn kernels
are yellow and white, although novelty colors found in conventional corn, such as blue
and red, are also available.(4)
A popped popcorn kernel is called a flake and can be characterized as either
mushroom or butterfly-type (Fig. 1.2),(19,21) and specific kernel hybrids have been
developed which produce flakes of each morphological type.(4) Butterfly-type popcorn is
characterized as having appendages spreading in various directions and are typically
preferred for home consumption because it has more tender texture and less prominent
hull pieces than mushroom-shaped popcorn.(4) Mushroom-type popcorn is
conventionally used for commercial pre-popped markets because the popped flakes are
less susceptible to breakage than butterfly-shaped popcorn during mixing, coating, and
packaging.(22)
11
1.4. Popping Mechanics
The mechanism describing popcorn popping has been investigated and reviewed
in detail by several authors.(10-11,23-25) Starch is the major polymer involved in popcorn
expansion. Although the absence of mechanical shear during popping results in some
unique properties, the thermal, physical, and chemical phase transitions of popcorn starch
during popping are similar to transitions experienced during extrusion of cereal foods.(25)
Thus, the mechanics of pop expansion can be understood by considering the general
mechanism described for vapor-induced puffing of other cereals(25) including orderdisorder transformation, nucleation, swell, and bubble growth and collapse.(26)
Correspondingly, the model for popcorn transformation from kernel to edible flake can
be contemplated as a series of polymer transitions consisting of several distinct steps
including heating, nucleation and popping, expansion, and final structure.
1.4.1. Heating
As heat is applied to popcorn, the pericarp acts as a pressure vessel which limits
kernel moisture loss and facilitates vapor pressure build-up within the kernels.(11) For
sound, undamaged popcorn, the outer pericarp is robust enough such that no more than
16-19% of the initial moisture escapes the kernel prior to popping.(11) DaSliva et al.(24)
studied the thermal diffusivity and conductivity of the pericarp using a photoacoustic
technique and reported that popcorn pericarp has 2.9x greater diffusivity and 2.0x greater
conductivity than conventional dent corn pericarp, which suggests the cell wall matrix of
popcorn is more ordered than conventional corn. Tandjung et al.(27) used x-ray
diffraction to show that arabinoxylan and cellulose, the main structural components of the
12
cell wall, undergo molecular changes when heated to form enhanced crystalline
structures. Enhanced crystallinity improves moisture retention in the kernel, which has
been shown to improve popping performance and decrease the prevalence of unpopped
kernels.(27)
As a result of increasing vapor pressure and temperature within the kernel, water
vapor penetrates the translucent endosperm, which softens and becomes a cohesive,
viscoelastic mass.(10) As this occurs, starch granules found in the translucent endosperm
undergo a process of gelatinization and crystalline melting. (10) These changes occur as
water is absorbed into the amorphous spaces of the starch and transfers heat on a
molecular basis to the crystalline regions of the starch, eventually causing the crystalline
structures to melt.(28-29) The kernel moisture content plays an important role in the extent
of starch gelatinization and melting that occurs prior to popping. Thermal analysis using
differential scanning calorimetry (DSC) of unpopped popcorn endosperm conditioned to
moisture contents ranging from 13.2-16.5% showed a glass transition between 55-61˚C
and an endothermic transition, likely to be starch melting between 174-183˚C.(30)
In contrast to the dynamics occurring in translucent endosperm during heating,
increasing vapor pressure does not apparently penetrate starch granules in the opaque
endosperm, but rather vapor pressure build-up occurs in the numerous open spaces
between starch granules.(8,31) As a result, the starch granules in the opaque endosperm do
not undergo gelatinization during the popping process.(8,31) Because differences in
gelatinization properties of starch between the translucent and opaque portions of
popcorn endosperm have only been inferred from post-popping analysis, however, further
13
study of the starch properties found in each endosperm fraction prior to popping using
DSC might be informative.
1.4.2. Nucleation and Popping
Nucleation occurs when small, gaseous embryos form and grow to sufficient size
to form permanent, stable bubbles called nuclei.(33) The hilum is a small microscopic
pore near the center of a starch granule which serves as the primary site for small bubble
nucleation. In addition, air bubbles can become entrapped within the endosperm voids
and provide an alternate site for nucleation of larger air cells.(10,32)
Just before popping, the pericarp swells slightly and small air bubbles appear in
the opaque endosperm and then merge to form a central pore with a diameter of
approximately 1 mm.(11) This occurs as vapor pressure builds up in the spaces between
starch granules of the opaque endosperm and forces them apart.(11) Pericarp ruptures
occurs when the vapor pressure of superheated steam inside the kernel exceeds the
combined burst pressure of the pericarp and atmospheric pressure. In sound grain at
appropriate moisture contents (~11-15%) this explosion typically takes place when the
internal kernel pressure is 760-930 kPa and internal kernel temperature is ~177-185°C.(1011,34)
1.4.3. Expansion
The endosperm plays a major role in formation of the flake after pericarp
rupture.(8) When the pericarp ruptures, superheated water vaporizes and rapidly diffuses
at the nucleation sites, providing the driving force for expansion.(10,11,34) It is believed
that the starch hilum provides the primary site for nucleation and vaporization of
14
superheated water in starch granules of the translucent endosperm.(10) Because pressure
being exerted by vaporizing moisture inside these starch granules is not counterbalanced,
significant expansion of individual starch granules in the translucent endosperm
occurs.(10-11) Expansion is ultimately governed by the multi-axial extension of individual
bubbles, which is driven by reduced vapor pressure exerted on the bubbles.(26) While
some localized rupturing of cell walls occurs during expansion, the majority of individual
starch granules are gelatinized but remain largely intact.(10,31) The resultant network of
expanded cells form a three dimensional foam structure.(10,31)
While starch granules in the translucent endosperm are highly expanded and
responsible for flake formation, the starch granules in the opaque endosperm appear to
undergo little change during popping other than moving apart.(11,31) Starch granules
obtained from opaque endosperm after popping have been shown to remain ungelatinized
and retain birefringence.(10,31) Because opaque endosperm is loosely packed, it is likely
that superheated water nucleates and vaporizes in the voids between starch granules
instead of within the individual starch granules.(10) This premise is supported by
microscopic analysis revealing large voids in the opaque endosperm portion after
popping, while also showing that the starch granules found in opaque endosperm do not
appear expanded like those in the translucent endosperm.(31)
The releasing steam has the dual purpose of providing pressure to drive expansion
and causing evaporative cooling which helps solidify the matrix.(25,35) Heat and mass
balances predict that temperature of popcorn immediately after popping is ~139˚C due to
temperature drop calculated from the moisture loss for popping and enthalpy loss.(25)
Expansion of the viscoelastic polymeric matrix ceases when the vapor pressure of
15
releasing endosperm becomes the same as the surrounding air pressure.(25) The entire
expansion mechanism from pericarp rupture to final flake takes less than 1/15 second to
complete.(11) Popping does not appear to substantially alter either the germ or pericarp,
other than fracturing the pericarp.
1.4.4. Final Popped Structure
To the naked eye, popped popcorn assumes a variety of distinct shapes ranging
from nearly spherical flakes produced by mushroom-type hybrids to an array of irregular
shapes with appendages spreading in various directions produced by butterfly-type
hybrids (Fig. 1.2). The microscopic structure of translucent endosperm from various
flake polymorphisms can be observed using scanning electron microscopy (Fig. 1.3).
These flake polymorphisms reveal a polygonal cellular structure with pore sizes typically
30-60 μm in diameter.(10,25,31) In comparison, pore diameters of dehulled kernels of
wheat, rice, and corn puffed in a puffing gun are typically 50-200 μm.(36)
Analysis of popcorn after popping by x-ray diffraction shows a diffraction pattern
that is typical of raw corn starch, but with reduced intensity.(30,37) This indicates
incomplete starch gelatinization during popping,(30) which is expected due to the partial
melting and gelatinization of starch discussed above. Residual crystallinity of starch from
popped popcorn depends on the moisture content of the kernel before popping. For
instance, the relative crystallinity of starch after popping increased from 24.5% to 48.4%
as the kernel moisture level increased from 10.0 to 16.5%.(30)
16
1.5. Popcorn Quality Traits
There are numerous popping performance traits that are important in high-quality
popcorn (Table 1.1). These can be broadly categorized as expansion volume and
“eatability,” which includes the number of unpopped kernels, hull dispersion, and the
shape, color, texture, and flavor of the popped flakes.(2) While expansion volume is
routinely quantified, most measures of “eatability” only can be defined qualitatively.(4)
1.5.1. Expansion Volume
Large expansion volume is generally desirable and associated with high popcorn
quality. In commercial venues like movie theatres, increased expansion volume
produces greater profits since popcorn is purchased by weight yet sold by bulk
volume.(10,39) Additionally, tender and fluffy popcorn associated with high expansion has
been shown to be more desired by consumers(40-41) and is associated with improved
tenderness and palatability.(42) As a result, popcorn with increased expansion volume has
the dual benefit of both increasing profits for commercial popcorn producers and
increasing enjoyment for the consumer.
The most widespread method for measuring and reporting expansion volume is by
measuring bulk volume of flakes produced by a known weight of unpopped popcorn
kernels. For oil-popping, a Metric Weight Volume Tester (MVTR) from C. Cretors and
Co. in Chicago, IL is the industry standard equipment for performing popcorn expansion
measurements.(4,39) For air and oil popped popcorn, expansion volume is reported as the
ratio of popped volume divided by the original sample weight (Eq. 1):
Expansion Volume =
(1)
17
For microwave popcorn, the conventional manner of reporting expansion volume
has also been to measure the volume of popped corn in a graduated cylinder and divide
by the original kernel weight as shown in Eq. 1.(43-44) The typical popcorn expansion
volume experienced by consumers using microwave popping and the conventional
method of determination is 36-55 cm3/g.(44-45)
Despite the prevalence of reporting expansion volume based on original kernel
weight, Pordesimo et al.(46) defined an alternate method for determining expansion
volume for microwave popcorn as the ratio of pop volume to the weight of popped flakes
only (Eq. 2). Pordesimo et al.(46) also defined a method for reporting the average flake
size of popped popcorn by dividing the bulk volume by the total number of popped flakes
(Eq. 3).
Expansion Volume =
(2)
Flake Size =
(3)
The primary advantage of the determination method recommended by Pordesimo
et al. (46) is that they eliminate the influence of loss factors such as unpopped kernels and
moisture loss during popping. While determining expansion volume based on original
kernel weight (Eq. 1) may be most appropriate for markets where popcorn is sold by
volume and loss factors have a large impact on profitability (e.g., movie theatre, sports
venues),(10) the methods of determination proposed by Pordesimo may be more
appropriate in markets where popcorn is prepackaged, labeled, and sold by weight.
Because expansion volume measures the bulk interaction between many pieces of
popped popcorn, the physical basis of expansion volume is actually comprised of many
18
components including flake size, flake density, the number of kernels that successfully
pop, and the void spaces between kernels.(41,47) The complexity of popcorn expansion
volume measurements parallels research of bulk density measurements in other systems.
Modeling of cereal grain packing efficiency has indicated bulk volume and density are
affected by 1) the combination and distribution of individual sizes and shapes in a
mixture, 2) the interaction effects between grain particles and adjacent particles, and 3)
the container boundaries.(48) Modeling with ellipsoids has shown that only ~55% of the
physical space is typically occupied by the actual particles in bulk systems.(49) While few
simulations have reported packing of irregular shaped objects like popcorn,(50) it is likely
that popcorn has a much lower pack efficiency than 55% due to the intrinsic
inefficiencies of packing irregular shapes. Research on the physical basis of bulk packing
of popped popcorn might lead to new perspectives on expansion volume, and would have
broad application to other fields of science interested in the bulk packing of irregularly
shaped particles.
1.5.2. Unpopped Kernels
Producers try to maximize expansion volume while minimizing the number of
unpopped kernels, which are considered a nuisance to consumers.(51) Unpopped kernels
can pass through a 7.14 mm square-hole screen(39) and are generally reported as a
percentage of total kernel weight before popping (Eq. 4). Consumers commonly
experience up to 10-12% unpopped kernels.(44,52)
% Unpopped Kernels =
(4)
19
Research has shown that most unpopped kernels are able to successfully pop if
heated a second time,(17) which suggests unpopped kernels do not arise because of
inherent inability to pop. In fact, the internal pressure and temperature requirement for
popping are known to vary from kernel to kernel,(35) and so achieving the thermodynamic
requirements of individual kernels is the most critical factor determining whether or not
an individual kernel will pop.(10)
1.5.3. Hull Dispersion
The hull is actually the pericarp of the popcorn kernel. Popcorn hulls are the part
of popcorn that sticks in consumers‟ teeth.(2) The most important feature of popcorn
which minimizes or avoids this undesirable nuisance is hull dispersion, which describes
the shattering of the pericarp during popping. Desirable hull dispersion is experienced
when the pericarp shatters into numerous small pieces during popping, whereas
undesirable hull dispersion is when the pericarp fractures into large pieces of intact
pericarp.(4,13) For example, popcorn breeders have crossed popcorn with dent corn to
produce higher yielding popcorn hybrids. These hybrids exhibit acceptable expansion
volume, but the pericarp of the resultant popcorn has very poor hull dispersion, often with
large half-pieces of pericarp.(4)
Hull dispersion is typically evaluated using visual, qualitative assessment, and
popcorn breeders are encouraged to evaluate the hull dispersion as part of hybrid
selection.(4) The development of a rapid method for quantifying hull dispersion would
offer utility for both popcorn breeders and manufacturers to make more consumerfriendly selections. One approach may be to pass a known weight of physically separated
hull material through a series of sieves with different mesh sizes and weigh the overs.(53)
20
1.5.4. Shape of Popped Flakes
As mentioned earlier, the shape of popped flakes has been long held as an
important quality factor for popcorn. Popcorn hybrids are commonly classified as being
either mushroom or butterfly-type depending on the type of flake produced upon
popping. However, even within a single butterfly-type popcorn hybrid, popped popcorn
flakes have been shown to assume a number of distinct polymorphisms, which can be
identified and characterized by whether the appendages are expanded unilaterally,
bilaterally, or multilaterally (Figure 1.2).(54) Data on the shape of popped flakes is not
routinely collected or reported,(8) although recent research indicates that these shapes
differ in nutrient composition after popping and are perceived differently by
consumers.(54) Thus, further research in this area is warranted.
1.5.5. Sensory and Nutritional Traits
Several additional factors are considered important quality traits for consumers,
including taste, color, texture, and nutritional attributes of popcorn.(4) Hoseney suggested
that popcorn‟s sustained popularity can be traced to its unique texture and delicate corn
flavor that makes it the ideal carrier for fat, salt, cheese, caramel and other flavors.(1)
Taste is closely associated with aroma, and numerous compounds have been identified
that contribute to popcorn aroma and flavor, including pyrazines, furans, pyrroles,
carbonyls, and substituted phenols,(55-56) with pyrazines playing a major role in imparting
characteristic popcorn flavor and aroma.(56-57) High-quality popcorn texture is defined as
crispy and melts in the mouth, versus low-quality being that which is chewy and adhesive
to the teeth.(4)
21
The inherent nutritional aspects of popcorn are a final quality factor that may also
be considered. Popcorn is a whole grain food that maintains all botanical parts (bran,
germ, endosperm) after popping.(10) This may be especially relevant since the 2010
Dietary Guidelines for Americans recommend increasing whole grain consumption due
to the relationship between whole grains and reduction in chronic health issues.(58)
Current whole grain consumption is well below recommended levels, and popcorn has
been recommended as a whole grain snack that can be used to increase whole grain and
dietary fiber in the diet.(58-59) For example, a 35 g serving of popcorn contains 4 g of dietary
fiber, or 16% of the US Daily Value.(6)
1.6. Factors Affecting Popping Performance
For most cereal grains, differences in grain quality characteristics arise due to a
variety of factors including genetics, growing environment, agronomic and postharvesting practices, and kernel composition, among others.(60) Similarly, a number of
studies investigating factors affecting quality attributes of popcorn have demonstrated
that popping performance is a complicated system that is dependent on a variety of
intrinsic and extrinsic factors (Table 1.2) .(46)
1.6.1. Hybrid, Environment, and Agronomic Inputs
While inter-annual variation and environmental changes play an important role in
product quality,(4) controllable factors such as hybrid selections and agronomic
procedures have also been shown to influence popcorn quality.(4,16,39-40,47,61-64) Popcorn
expansion volume has been shown to be a quantitative trait with high heritability
influenced by three to four major genes.(65)
22
Breeding methods used to improve dent corn are commonly used for popcorn,
except that the trait of emphasis in popcorn is popping performance, while dent corn
breeders usually focus on yield.(8) Much of these efforts have been based on the findings
by Dofing et al., which showed that while expansion volume tends to be negatively
correlated with yield, breeding methods that use additive genetic variation for expansion
volume and dominance variation for grain yield will likely result in simultaneous
improvements in producer yield and popping quality traits.(42)
1.6.2. Conditioning and Kernel Moisture Content
Physiological maturity for popcorn occurs at approximately 35% moisture, but
most popcorn acres are mechanically harvested with moistures 14-18%, and then the
elevator will dry the grain using low temperature air to the 13-14.5% moisture required
for optimum popping performance and storage stability.(8,19,21,66) Harvesting above 18%
moisture causes high shelling losses and kernel damage; below 14% moisture, kernels are
susceptible to impact damage, which ultimately decreases pop performance.(21)
Mature popcorn kernels are hygroscopic and gain or lose moisture depending on
temperature and humidity of the surrounding environment.(8) Conditioning popcorn
refers to achieving and maintaining the optimal moisture content to maintain or optimize
popping performance and minimize losses during storage.(8) Conditioning must be
completed quickly enough to prevent mold development but sufficiently slow to avoid
stress crack formation.(45) Typically, popcorn conditioning is conducted by equilibration
in a conditioning room with controlled temperature and relative humidity for a minimum
of 4 weeks before sealing in moisture proof containers.(4) Equilibrium isotherms have
been established for setting temperature and relative humidity to achieve desired kernel
23
moisture. For most popcorn, storage at 25˚C and 70% relative humidity will condition
popcorn to the desired moisture content for optimum popping.(67)
Since the model for popcorn popping relies on steam pressure building and
rupturing the pericarp upon heating the kernel, kernel moisture content is one of the most
important factors affecting popping performance.(10,45,66) Optimal pop performance
occurs between 13.0 and 14.5% kernel moisture,(8,45,66,68-69) with significant impairment
of popping performance occurring when the moisture content diverges beyond 1216%.(45) For example, Song and Eckhoff(69) showed popcorn expansion volume
decreased exponentially according to a second order polynomial (R2 = 0.86 – 0.99) as
kernel moisture content diverged from ideal. Having moisture content that is too low will
generate insufficient steam pressure during heating, while kernel moisture content that is
too high will cause a rubbery collapse of the pericarp at lower vapor pressure than
necessary to drive pop mechanics.(10-11,30)
1.6.3. Kernel Physical Characteristics
Physical properties of the popcorn kernel such as size, shape, and density that can
be measured nondestructively have been widely investigated for their relationship to
popcorn quality. The effect of kernel size on popping performance has been investigated
by several researchers. Willier and Brunson observed that expansion volume was highest
with smaller kernels, and they reported negative correlations between expansion volume
and kernel length, width, and thickness.(70) Lyerly also found a relationship between
expansion volume and kernel size, with medium-to short kernels showing the highest pop
volumes.(40) Song et al. reported kernel size had a significant effect on both expansion
volume and the number of unpopped kernels, with the middle-sized kernels (i.e., those
24
kernels retained on a sieve with 5.16 mm openings and passing through an opening of
5.95 mm) showing the most desirable performance.(39) By comparison, the smallest-sized
kernels (4.36 – 4.76 mm) showed the lowest expansion volume and the greatest number
of unpopped kernels and the largest kernels (>5.95 mm) had the highest percentage of
unpopped kernels by weight and expansion volume between the other sized kernels.(39)
The effect of kernel shape and density on popping performance has also been
investigated. Haugh et al. investigated the effect of test weight, specific gravity, and
kernel shape on expansion volume, and found that popcorn with increased test weight and
those kernels taken from the butt of the ear had greater expansion volume and were
shaped more approximately like a sphere.(71) Pordesimo et al. studied the effect of kernel
size, sphericity, and specific gravity on a number of popping performance measures
including expansion volume, flake size, and unpopped kernels. (46) Average kernel
sphericity was determined by calculating the ratio of geometric mean diameter (cubic
root of the multiplied length, width, and thickness) and average kernel length of fifty
kernels of each variety using calipers. Results agreed with previous findings that small to
medium-sized kernels had the highest expansion volume, and kernel size had a
significant effect on percentage of unpopped kernels. Mean sphericity ranged from 66 –
75%, and correlation analysis showed a positive relationship (r = 0.87) between
sphericity and expansion volume, with the greatest expansion volume occurring when
sphericity was greater than 70%. With the exception of one variety, the mean specific
gravity was not statistically different between popcorn varieties. However, when the
specific gravity of different kernels from a single variety was measured using the
flotation technique, specific gravity was shown to have a significant effect on expansion
25
volume within a given variety of popcorn, which supports the findings of Haugh et al.(71)
In addition, Pordesimo et al. observed that flake size increased up to specific gravity 1.35
but then leveled off.(46)
While these reports studied kernel physical properties on popping characteristics
in bulk, Tian et al. investigated the relationship between expansion volume and physical
properties including kernel size, sphericity, and density, on a single kernel basis.(47)
Expansion volume of individual popped popcorn flakes from three different commercial
popcorn hybrids was measured using sand displacement. Mean expansion volume for the
individual popped flakes ranged from 12.3 to 13.9 cm3/g, which corresponds to
individual popped flake density of 0.072-0.081 g/cm3. Analysis of variance showed that
hybrid, kernel sphericity, and kernel density all had significant effects (p<0.05) on
expansion volume.(47)
1.6.4. Pericarp Thickness and Kernel Damage
As discussed previously, popcorn pericarp is critical for enabling pop expansion
and distinguishes popcorn from other species of corn. Indeed, popcorn kernels with the
pericarp removed show a 90.4% reduction in pop expansion(24) and pericarp damage
reduces expansion volume since pressure is relieved through the damaged site prior to
popping.(40,72-73) Given this critical importance, Mohamed et al. tested 18 popcorn hybrids
from Nebraska in both a conventional Cretors popper and in a 700 W microwave oven.(44)
Pericarp thickness was directly proportional to both expansion volume (r=0.82, p<0.001)
and unpopped kernels (r=0.66, p<0.001). The researchers postulate that while a kernel
with thicker pericarp would take more time and energy to pop than kernels with thinner
26
pericarp, such kernels would give higher expansion volume.(10,44) Similar results were
reported by DaSliva et al.(24)
1.6.5. Kernel Composition
As discussed previously, the endosperm fraction of the kernel is primarily
responsible for formation of the flake.(10,23) The proportion of translucent and opaque
endosperm in the kernel has been shown to be closely related to expansion volume, and
altering the distribution of endosperm types will influence the popping quality.(10,31)
Popcorn expansion volume has been shown to increase as the amount of opaque
endosperm decreased.(27) In fact, having a higher ratio of opaque to translucent
endosperm is a major reason that grain sorghum and flint corn produce relatively modest
expansion volumes compared to popcorn.(70) Pordesimo et al. suggested that the need for
a small amount of opaque endosperm cannot be dismissed, since having a centralized
portion of such less densely-packed endosperm likely allows for greater steam pressure
buildup prior to popping, which results in greater expansion volume.(27)
Because the texture of endosperm is associated with the amount and types of
starch, protein, and lipids which are present in the matrix,(74,75) Borras et al. sought to
characterize how kernel composition and endosperm structural elements contribute to
popping performance using seven commercial popcorn hybrids from Argentina.(16)
Significant differences in expansion volume for the hybrids were reported, with a range
from 35.5 – 46.8 cm3/g. While no relationship was established between starch
composition and expansion volume for the popcorn hybrids studied, a significant positive
relationship was observed between kernel linoleic acid content and pop expansion
volume (p<0.02) and a significant negative correlation existed between expansion
27
volume and oleic acid concentration (p<0.005).(16) This was surprising, since most of the
lipids in popcorn are found in the germ fraction which is largely unchanged during
popping.(10) Further investigation designed to explain this phenomenon would be
valuable. Differentiation between lipids in the germ and lipids associated with the
endosperm may provide some insight.
Endosperm proteins may also affect pop volume.(16) Proteins extracted from
popcorn were separated using HPLC and a regression model was fitted to predict
expansion volume from principle component (PC) loadings of chromatograms. The
protein fractions that loaded onto the PC responsible for predicting expansion volume
corresponded to α-zein proteins, which have previously been associated with endosperm
hardness, and glutelin fractions. This suggests that increased levels of α-zein and glutelin
proteins have a positive effect on expansion volume, perhaps as a result of increased
endosperm hardness.(16)
1.6.6. Heating Method
The pop quality of popcorn is influenced by considerations besides inherent
kernel genotype, composition, and physical characteristics. In particular, extrinsic factors
such as the method of heating, atmospheric pressure, and the interaction with ingredients
added to the popcorn kernels during popping are important (Table 1.3).
Popcorn can be popped using heated air, heated oil, or microwave energy.
Gökmen studied the popping performance of five different popcorn genotypes at seven
different moisture contents, all tested using a number of different popping methods,
including microwave, hot-air, and cooking pan (both with and without salt added).(68)
Results showed that popping method had a significant effect on all popping performance
28
attributes. Hot air popping produced the highest expansion volume (32.0 cm3/g),
followed by cooking pan without oil and salt (31.7 cm3/g), then microwave popping (30.2
cm3/g). Popping with oil and salt resulted in significant decreases in both expansion
volume and flake size compared to other methods. Microwave popping had the highest
percentage of unpopped kernels (8.3%) compared to hot air (3.6%) and oil cooking in a
pan (4.0%), but microwave popping also produced the largest average flake size (4.2
cm3).(68) Several other studies have reported improved expansion volume with
conventional oil popping compared with 700-750 W microwave ovens.(43-44,64)
1.6.7. Ingredient Additions
Ingredient additions to popcorn can also impact popping performance. Lin and
Anatheswaran observed that adding 30% oil decreased expansion volume and reduced
unpopped kernels in microwave popping compared to lower oil additions.(73) This
relationship is especially noteworthy since 30% oil addition is the prevalent amount of oil
added to full-fat microwave popcorn products sold in the US market.(76) Singh and Singh
reported significant interactions among ingredients added to popcorn (77) They reported
increased expansion volume and greater unpopped kernels with increased butter levels,
while unpopped kernels were reported to decrease with increases in hydrogenated oil,
particularly in the presence of low levels of sodium chloride.(77) Similar results on
popping performance due to ingredient additions were reported by Ceylan and
Karababa.(78) These studies suggest complex interaction effects for different ingredients
with popcorn, which is to be expected since different formulations of the additive system
will have different dielectric constants and loss factors which will impact the
thermodynamic properties of heating.(77)
29
1.6.8. Atmospheric Pressure
As discussed previously, popcorn popping is governed by thermodynamic laws
and can be modeled as an adiabatic expansion which stops when the vapor pressure of
releasing endosperm becomes the same as surrounding air pressure.(25) Thus, it is
possible to improve popping performance by lowering the pressure surrounding
unpopped kernels,(79) which can be achieved by either popping at higher altitudes or
popping in modified atmospheric chambers. Quinn et al. designed a cooktop apparatus
equipped with vacuum pump, and they compared performance of commercial popcorn
hybrids popped under modified atmosphere to those heated on the stovetop without
modified atmosphere.(52) Expansion volume and flake size doubled and there was a
dramatic reduction in the percentage of unpopped kernels when popping under reduced
atmospheric pressure.(52)
1.6.9. Microwave Power and Packaging
For microwave popcorn, it is also important to consider the effects of microwave
power and packaging on popping performance. This is especially relevant since the
majority of retail popcorn is popped in a microwave,(61) and power output of consumer
microwave ovens varies widely. Allred-Coyle et al. studied the effects of bag capacity on
expansion volume of a single popcorn hybrid that was conditioned to three different
moisture contents (11%, 12%, and 13%).(80) The conditioned popcorn kernels were
added to microwave paper bags containing an inlaid metallic susceptor with three internal
cavity sizes (1400 cm3, 2400 cm3, and 3500 cm3) and were popped in microwave ovens
with three power outputs (500 W, 800 W, and 1000 W). Results showed that moisture
content, microwave oven wattage, and bag capacity all had a significant (p<0.05) effects
30
on expansion volume, with a significant (p<0.05) interaction effect between all three
design factors. In general, expansion volume increased with relative increases in bag
capacity, which may be due to the larger available space permitting popcorn to more
freely expand inside the bag, whereas a more constricted bag restricts pop expansion.(80)
This finding supports the assertion that bag design is also a critical factor affecting
popcorn quality. Overall, the maximum expansion volume was achieved with the largest
capacity bags and the highest microwave wattage.(41)
1.7. Conclusions
The mechanism describing thermal, physical, and chemical transformations from
kernel to edible popcorn flake can be understood by considering a multistep process
where superheated steam is forced into the hilum of starch granules in the translucent
portion of kernel endosperm, facilitating a phase transition of starch from a glassy to
rubbery state. When the pressure and temperature inside the kernel reaches
approximately 177-185°C and 760-930 kPA, the pericarp ruptures. Subsequently,
superheated steam vaporizes and forms rapidly expanding bubbles at nuclei within the
hilum of starch in the translucent endosperm and in the open spaces around the starch
granules of the opaque endosperm. Expansion caused by the vaporizing moisture inside
individual starch granules of the translucent endosperm creates a porous cell of foam.
Evaporation of the superheated water cools the viscoelastic foam matrix to provide the
familiar solid flake.
The most important attributes of high-quality popcorn are high expansion volume,
low occurrence of unpopped kernels, good hull dispersion, desirable appearance, crisp
31
and tender texture, and delicate flavor. Selection of hybrid, growing environment, and
kernel conditioning are critical factors affecting end-use quality. Physical characteristics
of kernels such as kernel size, shape, density, and pericarp damage and thickness have
been shown to affect popping performance, as have innate compositional factors
including kernel moisture, protein, and fatty acid content. In addition, the end-use
quality of popcorn is influenced by extrinsic variables including the popping method and
ingredient additives.
While the popping mechanism and many quality traits for popcorn have been well
established, breeders and producers and popcorn manufacturers continually seek new or
enhanced quality attributes to increase consumer satisfaction or create market differentiation.
Focusing on new or different quality attributes can increase consumer relevance and enhance
market success. For example, food categories such as pasta and frozen potatoes have
successfully innovated by selling an assortment of different shapes. (81) Moreover, identifying
new quality attributes that might increase popcorn‟s relevance and consumption is a viable
means of addressing the gap in whole grain consumption in the American diet.(58) The
significance of investigating and characterizing new quality attributes for popcorn is that it
might inspire new innovation to help lead to renewed growth of popcorn consumption.
1.8. References
1. Hoseney, R.C. Principles of Cereal Science and Technology, edition 2; American
Association of Cereal Chemists: St. Paul, MN, 1994; 345-347.
2. Matz, S.A. Snack Food Technology, edition 2; AVI Pub. Co.: Westport, CT, 1984; 13, 133-144.
32
3. Rooney, L.W; Serna-Saldivar, S.O. Uses of whole corn and dry-milled fractions. In
Corn Chemistry and Technology; White, P.J.; Johnson, L.A., Eds.; American
Association of Cereal Chemists: St. Paul, MN, 1987; 399-429.
4. Ziegler, K. E. Popcorn. In Specialty Corn; Hallauer, A., Ed.; CRC Press: Boca Raton,
FL, 2001; 199-234.
5. Home Theater Express: The history of popcorn machines, Portland, OR,
http://www.ht-express.com/howto-history-popcornmachine.htm (accessed November
23, 2009).
6. The Popcorn Board, http://www.popcorn.org (accessed November 26, 2011).
7. Foer, J. Arming nature's grenade. Discover Magazine 2005, 26, 24-25.
8. Ziegler, K.E. Popcorn. In Corn: Chemistry and Technology; White, P.J.; Johnson,
L.A., Eds.; American Association of Cereal Chemists: St. Paul, MN, 2003.
9. Watson, S.A. Description, development, structure and composition of the corn kernel.
In Corn Chemistry and Technology; White, P.J.; Johnson, L.A., Eds.; American
Association of Cereal Chemists: St. Paul, MN, 2003.
10. Hoseney, R.C.; Zeleznak, K.; Abdelrahman, A. Mechanism of popcorn popping.
Journal of Cereal Science 1983, 1, 43-52.
11. Wu, P.J.; Schwartzberg, H.G. Popping behavior and zein coating of popcorn. Cereal
Chemistry 1992, 69, 567-573.
12. Doebley, J.F.; Iltis, H.H. Taxonomy of Zea (Gramineae). I. A subgeneric
classification with key to taxa. American Journal of Botany 1980, 67, 982-993.
13. Richardson, D.L. Purdue hybrid performance trials encouraging. Popcorn
Concessions Merchandiser 1957, 12, 10-17.
33
14. Tracy, W.F.; Galinat, W.C. Thickness and cell layer number of the pericarp of sweet
corn and some of its relatives. Horticultural Science 1987, 22, 645-647.
15. Park, D.; Allen, K.G.D.; Stermitz, F.R.; Maga, J.A. Chemical composition and
physical characteristics of unpopped popcorn hybrids. Journal of Food Composition
and Analysis 2000, 13, 921-934.
16. Borras, F.; Seetharaman, K.; Yao, N.; Robutti, J.L.; Percibaldi, N.M.; Eyherabide,
G.H. Relationship between popcorn composition and expansion volume and
discrimination of corn types by using zein properties. Cereal Chemistry 2006, 83, 8692.
17. Sweley, J.C.; Rose, D.J.; Jackson, D.S. Hybrid and environment effects on popcorn
kernel physiochemical properties and their relationship to microwave popping
performance. Journal of Cereal Science 2012, doi:10.1016/j.jcs.2011.11.006.
18. Heller, S.N.; Rivers, J.M.; Hackler, L.R. Dietary fiber: the effect of particle size and
pH on its measurement. Journal of Food Science 1977, 42, 436-439.
19. Ziegler, K.E.; Ashman, R.B.; White, G.M. Wysong, D.S. Popcorn production and
marketing. In National Corn Handbook; Purdue University Cooperative Extension
Service, 1984.
20. Ceylan, M.; Karababa, E. Comparison of sensory properties of popcorn from various
types and sizes of kernel. Journal of the Science of Food and Agriculture 2001, 82,
127-133.
21. D‟Croz-Mason, N.; Waldren, R. Popcorn production. In NebGuide, University of
Nebraska Cooperative Extension, 2007; G78-426.
34
22. Eldredge, J.C.; Thomas, W.I. Popcorn… Its production, processing and utilization. In
Iowa Agricultural Experiment Station Bulletin; Iowa State University: Ames, IA,
1959; 127.
23. Pordesimo, L.O.; Anantheswaran, R.C.; Mattern, P.J. Quantification of horny and
floury endosperm in popcorn and their effects on popping performance in a
microwave oven. Journal of Cereal Science 1991, 14, 189-198.
24. DaSliva, W.J.; Vidal, B.C.; Martins, M.E.Q. What makes popcorn pop. Nature 1993,
362, 417.
25. Schwartzberg, H.G.; Wu, J.P.C.; Nussinovitch, A.; Mugerwa, J. Modelling
deformation and flow during vapor-induced puffing. Journal of Food Engineering
1995, 25, 329-372.
26. Kokini, J.L.; Chang, C.N.; Lai, L.S. The role of rheological properties on extrudate
expansion. In Food Extrusion Science and Technology; Kokini, J.L.; Karwe, M.V.,
Eds.; Marcel Dekker, Inc.: New York, NY, 1992.
27. Tandjung, A.S.; Janaswamy, S.; Chandrasekaran, R.; Abaoubacar, A.; Hamaker, B.R.
Role of the pericarp cellulose matrix as a moisture barrier in microwaveable popcorn.
Biomacromolecules 2005, 6, 1654-1660.
28. Hoover, R. Composition, molecular structure, and physiochemical properties of tuber
and root starches: A review. Carbohydrate Polymers 2001, 5, 253-267.
29. Wang, S.S.; Chiang, W.C.; Yeh, A.I.; Zhao, B.L.; Cho, M.H. Experimental analysis
and computer simulation of starch-water interaction during phase transition. Journal
of Food Science 1991, 56, 121-124.
35
30. Shimoni, E.; Dirks, E.M.; Labuza, T.P. The relation between final popped volume of
popcorn and thermal-physical parameters. Lebensmittel-Wissenschaft & Technologie
2002, 35, 93-98.
31. Reeve, R.M; Walker Jr., H.G. The microscopic structure of popped cereals. Cereal
Chemistry 1969, 46, 227-241.
32. Hoseney, R.C.; Mason, W.R.; Lai, C.S.; Guetzlaff, J. Factors affecting the viscosity
and structure of extrusion-cooked wheat starch. In Food Extrusion Science and
Technology; Kokini, J.L.; Karwe, M.V., Eds.; Marcel Dekker, Inc.: New York, NY,
1992.
33. Madeka, H.; Kokini, J.L. Effect of addition of zein and gliadin on the rheological
properties of amylopectin starch with low-to-intermediate moisture. Cereal Chemistry
1992, 69, 489-494.
34. Byrd, J.E.; Perona, M.J. Kinetics of popping of popcorn. Cereal Chemistry 1995, 82,
53-59.
35. Moraru, C.I.; Kokini, J.L. Nucleation and expansion during extrusion and microwave
heating of cereal foods. Comprehensive Reviews in Food Science and Food Safety
2003, 2, 147-165.
36. Warburton, S.C.; Donald, A.M.; Smith, A.C. The deformation of brittle starch foams.
Journal of Materials Science 1990, 25, 4001-4007.
37. Zobel, H.F.; Young, S.N.; Rocca, L.A. Starch gelatinization: An x-ray diffraction
study. Cereal Chemistry 1988, 65, 443-446.
38. Biliaderis, C.G.; Galloway, G. Crystallization behavior of amylose-V complexes:
Structure-property relationships. Carbohydrate Research 1989, 189, 31-48.
36
39. Song, A.; Eckhoff, S.R.; Paulsen, M.; Litchfield, J.B. Effects of kernel size and
genotype on popcorn popping volumes and number of unpopped kernels. Cereal
Chemistry 1991, 68, 464-467.
40. Lyerly, P.J. Some genetic and morphological characters affecting the popping
expansion of popcorn. Journal of the American Society of Agronomy 1942, 34, 986995.
41. Levy, B. A new perspective on popcorn. Snack World 1988, 45, 24.
42. Dofing, S.M.; D‟Croz-Mason, N.; Buck, J.S. Inheritance of expansion volume and
yield in two popcorn X dent corn crosses. Crop Science 1991, 31, 715-418.
43. Dofing, S.M.; Thomas-Compton, M.A.; Buck, J.S. Genotype x popping method
interaction for expansion volume in popcorn. Crop Science 1990, 30, 62-65.
44. Mohamed, A.A.; Ashman, R.B.; Kirleis, A.W. Pericarp thickness and other kernel
physical characteristics relate to microwave popping quality of popcorn. Journal of
Food Science 1993, 58, 342-346.
45. Metzger, D.D.; Hsu, K.H.; Ziegler, K.E.; Bern, C.J. Effect of moisture content on
popcorn popping volume for oil and hot-air popping. Cereal Chemistry 1989, 66,
247-248.
46. Pordesimo, L.O.; Anantheswaran, R.C.; Fleishmann, A.M.; Lin, Y.E.; Hanna, M.A.
Physical properties as indicators of popping characteristics of microwave popcorn.
Journal of Food Science 1990, 55, 1352-1355.
47. Tian, T.; Buriak, P.; Eckhoff, S.R. Effect of hybrid and physical properties of
individual popcorn kernels on expansion volume. Cereal Chemistry 2001, 78, 578582.
37
48. Boac, J.M.; Casada, M.E.; Maghirang, R.G.; Harrier III, J.E. Material and interaction
properties of selected grains and oilseeds for modeling discrete particles. Transactions
of the ASABE 2010, 53, 1201-1216.
49. Wang, P.; Song, C.; Jin, Y.; Makse, H.A. Jamming II: Edwards‟ statistical mechanics
of random packings of hard spheres. Physica A 2011, 390, 427-455.
50. Gan. M.; Gopinathan, N.; Jia, X.; Williams, R.A. Predicting packing characteristics of
particles of arbitrary shapes. KONA 2004, 22, 82-91.
51. Skloot, R. Why popcorn doesn‟t pop. New York Times. 2005,
www.nytimes.com/2005/12/11/magazine/11ideas_section416.html?scp=1&sq=Why+popcorn+doesn%27t+pop&st=nyt (accessed November 26,
2011).
52. Quinn Sr., P.V.; Hong, D.C.; Both, J.A. Increasing the size of a piece of popcorn.
Physica A 2005, 353, 637-648.
53. AACC International. Approved Methods of Analysis, edition 11. Method 66-20.01.
Determination of Granularity of Semolina and Farina: Sieving method. Approved
November 3, 1999. AACC International, St. Paul, MN, U.S.A. doi:
10.1094/AACCIntMethod-66-20.01.
54. Sweley, J.C.; Rose, D.J.; Jackson, D.S. Composition and sensory evaluation of
popcorn flake polymorphisms for a select butterfly-type hybrid. Cereal Chemistry
2011, 88, 321-327.
55. Waldrat, J.P.; Lindsay, R.C.; Libbey, L.M. Popcorn flavor: Identification of volatile
compounds. Journal of Agricultural and Food Chemistry 1970, 18, 926.
38
56. Buttery, R.G.; Ling, L.C.; Stern, D.J. Studies on popcorn aroma and flavor volatiles.
Journal of Agricultural and Food Chemistry 1997, 45, 837-843.
57. Schieberle, P. Primary odorants in popcorn. Journal of Agricultural and Food
Chemistry 1991, 39, 1141-1144.
58. US Department of Agriculture and US Department of Health and Human Services.
Dietary Guidelines for Americans, edition 7; US Government Printing Office:
Washington D.C., 2010.
59. Jones, J. M., Reicks, M., and Adams, J. The importance of promoting whole grain
foods message. Journal of American College of Nutrition 2002, 21, 293-297.
60. Shandera, D.L.; Jackson, D.S.; Johnson, B.E. Quality factors impacting processing of
maize dent hyrids. Maydica 1997, 42, 281-289.
61. Allred-Coyle, T.A.; Toma, R.B.; Reiboldt, W.; Thaku, M. Effects of moisture
content, hybrid variety, kernel size, and microwave wattage on the expansion volume
of microwave popcorn. International Journal of Food Sciences & Nutrition 2000, 51,
389-394.
62. Kandala, C.V.K.; Nelson, S.O.; Lawrence, K.C. Nondestructive moisture
determination in small samples of popcorn by RF impedance measurement.
Transactions of the ASAE 1994, 37, 191-194.
63. Soylu, S.; Tekkanat, A. Interactions among expansion volume and kernel properties
in various popcorn genotypes. Journal of Food Engineering 2007, 80, 336-341.
64. Ertas, N.; Soylu, S.; Bilgicli, N. Effects of kernel properties and popping methods on
popcorn quality of different corn cultivars. Journal of Food Process Engineering
2009, 32, 478-496.
39
65. Lu, H.-J.; Bernardo, R.; Ohm, H.W. Mapping QTL for popping expansion in popcorn
with simple sequence repeat markers. Theoretical and Applied Genetics 2003, 106,
423-427.
66. Eldredge, J.C., and Lyerly, P.J. Popcorn in Iowa. Iowa Agric. Exp. Stn. Bull 1943,
54, 753.
67. Song, A.; Eckhoff, S.R. Optimum popping moisture content for popcorn kernels of
different sizes. Cereal Chemistry 1994, 71, 458-460.
68. Gökmen, S. Effects of moisture content and popping method on popping
characteristics of popcorn. Journal of Food Engineering 2004, 65, 357-362.
69. Song, A.; Eckhoff, S.R. Individual kernel moisture content of preshelled and shelled
popcorn and equilibrium isotherms of popcorn kernels of different sizes. Cereal
Chemistry 1994, 71, 461-463.
70. Willier, J.G.; Brunson, A.M. Factors affecting the popping quality of popcorn.
Journal of Agricultural Research 1927, 35, 615-624.
71. Haugh, C.G.; Lien, R.M.; Hanes, R.E.; Ashman, R.B. Physical properties of popcorn.
Transactions of the ASAE 1976, 19, 168-171, 176.
72. Singh, V.; Barreiro, L.; McKinstry, J.; Buriak, P.; Eckhoff, S.R. Effect of kernel size,
location, and type of damage on popping characteristics of popcorn. Cereal Chemistry
1997, 74, 672-675.
73. Lin, Y.E.; Anatheswaran, R.C. Studies on popping of popcorn in a microwave oven.
Journal of Food Science 1988, 53, 1746-1749.
74. Zhang, W.; Hoseney, R.C. Factors affecting expansion of corn meals with poor and
good expansion properties. Cereal Chemistry 1998, 75, 639-643.
40
75. Dombrink-Kurtzman, M.A.; Bietz, J.A. Zein composition in hard and soft endosperm
of maize. Cereal Chemistry 1993, 73, 775-778.
76. U.S. Department of Agriculture, Agriculture Research Service. USDA National
Nutrient Database for Standard Reference, release 23; US Government Printing
Office: Washington D.C., 2010.
77. Singh, J.; Singh, N. Effects of different ingredients and microwave power on popping
characteristics of popcorn. Journal of Food Engineering 1999, 42, 161-165.
78. Ceylan, M.; Karababa, E. The effects of ingredients on popcorn popping
characteristics. International Journal of Food Engineering 2004, 39, 361-370.
79. Hong, D.C.; Both, J.A. Controlling the size of popcorn. Physica A 2001, 289, 557560.
80. Allred-Coyle, T.A.; Toma, R.B.; Reiboldt, W.; Thaku, M. Effects of bag capacity,
storage time and temperature, and salt on the expansion volume of microwave
popcorn. Journal of the Science of Food and Agriculture 2001, 81, 121-125.
81. Berkowitz, M. Product shape as a design innovation strategy. Journal of Product
Innovation Management 1987, 4, 274-283.
Figure 1.1. Sales of unpopped popcorn 1970 – 2009.(6)
41
42
Figure 1.2. Various popcorn flake polymorphisms identified by qualitative observation.
Clockwise from upper-left: mushroom shape, unilaterally-expanded, bilaterallyexpanded, and multilaterally-expanded. The mushroom shape was identified from
popped mushroom-type hybrid, while the other three shapes were identified from a
butterfly-type hybrid. Scale: each square = 10 cm x 10 cm.
Figure 1.3. Scanning electron microscopic images of various popcorn flake polymorphisms viewed at 200x (top row) and 500x
(bottom row) resolution. All popcorn samples were mounted on metal stubs immediately after popping and coated with goldpalladium and observed using a Hitachi TM-3000 SEM. Images were captured by automatic image capturing software.
43
Table 1.1. Quality Traits for Popped Popcorn
Trait
Expansion volume
Description
Bulk volume of popped flakes produced from a known
weight of unpopped kernels. Larger expansion volume
is desirable.
Commonly reported values
35 – 55 cm3/g
Example references
Dofing et al.(43);
Metzger et al.(45)
Flake size
Average size of popped popcorn flakes, determined by
dividing bulk volume by total number popped flakes.
Larger flake size is desirable.
4 – 8 cm3
Pordesimo et al.(46);
Mohamed et al.(44)
Unpopped kernels
Kernels that do not pop (or pop very little) and pass
through 7.14 mm square-hold screen; reported as % total
kernels or kernel weight before popping. Less unpopped
kernels are desirable.
10 – 12%
Song et al.(39)
Hull dispersion
Size distribution of shattered pericarp pieces after
popping. Evaluated using qualitative visual assessment.
Numerous small pieces are desirable.
(Qualitative) Acceptable or
unacceptable
Ziegler(4)
Flake shape
Shape made after popping. Evaluated using qualitative
visual assessment. Mushroom shape desirable for
coating applications; butterfly-type shapes preferred for
non-coating applications
(Qualitative) Mushroom- or
Butterfly-Type; butterfly-type can be
further classified as unilateral,
bilateral, or multilaterally expanded
Ziegler(8); Sweley et
al.(54)
Sensory properties
(taste, color, texture)
High consumer acceptability for popcorn is correlated
with good color, texture and flavor. Desirable popcorn
texture is associated with tenderness and freedom from
hulls. Flavor is associated with aroma, particularly
pyrazines compounds, which can be quantified using
GC-MS.
Color: yellow or white
Texture: crispness, tenderness
Flavor: popcorn flavor, butter flavor
Ceylan and
Karababa(20)
44
Table 1.2. Inherent Kernel Factors Affecting Popping Performance of Popcorn
Factor
Hybrid
Description
Hybrid selection can affect all popping performance
measures. Expansion volume is a quantitative trait that can
be improved with plant breeding and hybrid selection.
Select References
Song et al.(39); Ziegler(4); Lu et al.(65);
Ertas et al.(64)
Moisture content
Maximum expansion volume occurs at 13.0 – 14.5%
moisture.
Eldredge and Lyerly(66); Metzger et
at.(45); Song and Eckhoff(67)
Kernel size
Small to medium-sized kernels show greater expansion
volume than larger sized kernels.
Willier and Brunson(70); Lyerly(40); Lin
and Anatheswaran(73); Pordesimo et
al.(46); Song et al.(39)
Specific gravity
Increased test weight is correlated to increased expansion
volume.
Haugh et al.(1); Pordesimo et al.(46)
Kernel sphericity
Increased sphericity is positively correlated to expansion
volume.
Pordesimo et al.(46); Mohamed et al.(44);
Ertas et al.(64)
Pericarp thickness
Thicker pericarp is positively correlated to expansion
volume and unpopped kernels.
Mohamed et al.(44); DaSliva et al.(24)
Kernel damage
Kernel damage to either the pericarp or seed coat decreases
expansion volume and increases popping time.
Lin and Anatheswaran(73); Singh et
al.(72)
Translucent and
Opaque Endosperm
Increased ratio of translucent to opaque endosperm is
associated with increased expansion volume.
Hoseney et al.(10); Pordesimo et al.(27)
Kernel lipids
Increased levels of linoleic and oleic acid are positively
correlated to expansion volume.
Borras et al.(16)
Kernel proteins
Increased levels of α-zein and glutelin proteins have a
positive effect on expansion volume.
Borras et al.(16)
45
Table 1.3. Extrinsic Variables Affecting Popping Performance of Popcorn
Factor
Heating method
Description
Popping method (air, oil, or microwave) has a significant effect on
expansion volume, unpopped kernels, and flake size. No heating
method shown to consistently perform the best.
Select References
Dofing et al.(43); Mohamed et al.(44);
Ertas et al.(64)
Ingredient additions
Expansion volume and unpopped kernels are positively correlated
to butter addition and negatively correlated to oil addition. Salt
(NaCl) addition of <2% acts synergistically with added oil to
decrease unpopped kernels.
Lin and Anatheswaran(73); Singh
and Singh(77); Ceylan and
Karababa(78)
Atmospheric pressure
Decreased atmospheric pressure increases expansion volume and
flake size and decreases unpopped kernels.
Hong and Both(79); Quinn Sr. et
al.(52)
Microwave Power
Increassed microwave power is positively correlated to expansion
volume and larger flake size.
Allred-Coyle et al.(61); Allred-Coyle
et al.(80)
Packaging
Expansion volume increases with larger microwave bag capacity
(i.e., less restriction on popcorn kernels to expand).
Levy(41); Allred-Coyle et al.(80)
46
47
CHAPTER 2. COMPOSITION AND SENSORY EVALUATION OF POPCORN
FLAKE POLYMORPHISMS FOR A SELECT BUTTERFLY-TYPE HYBRID
2.1. Abstract
The objective of this study was to identify and characterize different popped
popcorn flake shapes, or polymorphisms, arising from a yellow butterfly popcorn hybrid
(YP-213), and then to determine the impact of popcorn flake shape on composition and
sensory characteristics. Kernels were popped using a microwave oven and visually
sorted into three different polymorphisms depending on whether the appendages were
expanded unilaterally, bilaterally, or multilaterally. When popped, 9.0 ± 3.1%, 71.2 ±
5.9%, and 12.3 ± 3.8% of kernels were expanded unilaterally, bilaterally, and
multilaterally, respectively, while 7.6 ± 1.4% of kernels remained unpopped. Expansion
volumes for the unilaterally, bilaterally, and multilaterally expanded polymorphisms were
28.6 ± 3.84, 43.0 ± 0.84, and 53.5 ± 2.5 cm3/g, respectively. Unilateral popcorn flakes
retained the most fat, saturated fat, and sodium, while multilaterally expanded flakes had
the highest levels of protein, total carbohydrate, and popcorn-like aromatic pyrazines.
Sensory evaluation revealed significant differences among polymorphisms for flavor and
texture attributes, with the unilaterally expanded polymorphism receiving the highest
overall product liking. These data show that different popcorn flake polymorphisms
produced from a single hybrid of popcorn affect sensory and compositional profiles.
More research is necessary to elucidate the factors that affect popcorn flake
polymorphisms and support development of new varieties or techniques to produce the
most desirable microwave popcorn.
48
2.2. Introduction
Popcorn (Zea mays everta) is a familiar snack food with sizable commercial
business in the United States (US). For instance, annual domestic consumption of
popcorn is nearly 4.5 × 108 kg of popped popcorn each year, or about 1.36 kg annually
per person, making popcorn a favorite snack food consumed in the US by volume
consumed (The Popcorn Board 2010).
The most studied and reported quality factor for popcorn is expansion volume.
From a commercial processor‟s outlook, increased pop expansion produces greater
profits since popcorn is purchased by weight yet sold by bulk volume (Lyerly 1942, Song
et al 1991). Additionally, high pop expansion volume is correlated with desirable
consumer attributes including tender and fluffy texture (Lyerly 1942, Levy 1988) and
improved tenderness and palatability (Dofing et al 1991, Ceylan and Karababa 2001). As
a result, “lighter and fluffier” popcorn (with higher pop expansion) has the dual benefit of
both increasing profits for the manufacturer and increasing enjoyment for the consumer.
While displacement has been used as a method for measuring expansion volume
in select cases (Tian et al 2001), the most widespread method for reporting popcorn
expansion volume is by measuring bulk density. Other quality factors that are often
considered in popcorn include the number or percentage of unpopped kernels, hull
dispersion, and the size, shape, color, and texture of the popcorn flake (Ziegler 2001,
Quinn Sr. 2005).
Popcorn hybrids are commonly classified as yellow mushroom, white butterfly, or
yellow butterfly depending on the type of popcorn flake produced upon popping (Ziegler
et al 1984, D‟Croz-Mason and Waldren 1990), and specific hybrids are available to
49
produce popcorn flakes of each morphological type (Ziegler 2001). The butterfly-shaped
flakes are characterized as having appendages spreading in various directions and are
typically preferred for home consumption because they have more tender texture and less
prominent hull pieces than mushroom-shaped popcorn (Ziegler 2001). Mushroomshaped corn is conventionally used for commercial pre-popped markets because the
popcorn flakes are less susceptible to breakage than butterfly-shaped popcorn during
mixing, coating, and packaging (Eldredge and Thomas 1959).
Previous researchers have studied the importance of various sensory attributes of
popcorn flakes using consumer response evaluations. Matz (1984) and Watson (1988)
indicated that consumer acceptability for popcorn flake texture is positively correlated
with high pop expansion, freedom from hulls, and formation of butterfly-shaped flakes.
Ceyland and Karababa (2001) evaluated the sensory attributes of mushroom and butterfly
hybrids including popcorn flake size, uniformity, color, texture, and taste, and found that
flake shape was not an independent factor for consideration in the statistical analysis.
Park and Maga (2001) also compared sensory attributes from different popcorn hybrids,
and they determined that flake polymorphism (limited to mushroom versus butterfly
shape) did not significantly affect consumer preference.
Nevertheless, during numerous informal observations, the authors have noted that
many consumers appear to hunt for distinct shapes of popped popcorn within a bag or
bowl of microwave popcorn. Moreover, the visual perceptions of a food product clearly
influence the sensorial experience of eating (Hutchins 1977).
Ziegler (2001) illustrated that butterfly-shaped popcorn flakes can assume several
distinct polymorphisms; however, no previous studies have sought to characterize the
50
importance or classify differences in flake polymorphism. Accordingly, the purpose of
this study was to describe differences in compositional and sensory attributes of different
popcorn flake polymorphs obtained from a butterfly-type hybrid, and then to establish
relationships between the compositional and sensory measures.
2.3. Materials and Methods
2.3.1. Materials
Samples of a commercial US yellow butterfly popcorn hybrid (YP-213) were
obtained from ConAgra Foods (Omaha, NE). The hybrid used in this study was a
composite sample from seed production grown in Nebraska, Iowa, and Ohio in 2009. For
compositional analysis, unpopped popcorn kernels were milled using a two-step process
of chopping for 20 sec in a Blixer 3 blender (Robot Coupe, USA) and then grinding at the
highest setting using a Kitchen Mill Model 91 (Blentech, USA). The compositional
characteristics of the kernels (Table 2.1) were evaluated using previously published
methods: moisture (Approved Method 950.46, AOAC International 2010); lipids
(Approved Method 996.06, AOAC International 2010); protein (Approved Method
992.15, AOAC International 2010), with a % nitrogen to % protein conversion factor of
5.65 (Park et al 2000); amylose:amylopectin ratio (Zho et al 2008); and minerals (Na, K,
Ca, Fe, P, Approved Method 984.27, AOAC International 2010). Total carbohydrate was
measured by difference [100% - (moisture% + protein% + lipid% + minerals%)]. Total
sugars were determined by extraction using a 50/50 water and ethanol mixture while
shaking in a water bath at 60°C, then quantified by HPLC using an amine phase column
(Prevail Carbohydrate ES 5μm, Grace Discovery Sciences #35101), mobile phase of
51
75/25: Acetonitrile/Water at 1.0 ml/min and an Evaporative Light Scattering Detector
(Alltech #3300). Total starch was determined by enzymatic conversion of starch to
glucose using thermostable α-Amylase (3000U/ml, Megazyme #E-BLAAM) and
Amyloglucosidase (3260 U/ml, Megazyme #E-AMGDF), then quantified using the
HPLC procedure noted above for total sugars.
Kernels were tempered to 14% moisture (weight by volume) content in a
controlled cabinet at 21.5˚C and 73% relative humidity (rh) for approximately 30 days,
and then weighed into pre-folded, commercially standard, bi-layer microwavable paper
bags (14.92 x 29.53 cm2) with an internal capacity of 2800 cm3. The bags contained an
inlaid aluminum-coated polyester susceptor (13.65 x 16.51 cm2) positioned on the bottom
center of the bag gusset. A slurry of oil, salt, butter flavor, and color was added to the
popcorn kernels in each bag, as summarized in the formula shown in Table 2.2. Bags
were sealed using heat induction and stored at 21.5 ˚C and 73% rh until used for
experimentation (maximum of 4 months).
2.3.2. Popping and Sorting
All popping tests were completed using a series of seven different GE 1150 Watts
microwave ovens (Model JE146WF03). The actual power output for ovens varied
between 1110 and 1170 Watts as calculated by the method outlined by Allred-Coyle et al
(2000). To standardize popping performance between ovens, 240 ml of water was heated
in a ceramic coffee mug for 1.5 min before the first popping of the day. Microwave use
was rotated between the seven ovens for popping tests and doors were left open between
uses to ensure microwave ovens cooled down between use and constant oven temperature
was maintained for all popping tests performed.
52
Unfolded bags were placed in the center of a microwave oven with susceptor-side
down and popped on high power until the interval between pops slowed to 2-3 sec. The
contents were poured into a steel sieve with round-hole, 7.94 mm openings (Seedburo,
USA Model 0070) to remove unpopped kernels, which were counted and weighed. The
total popped volume of the sample was measured in a 4 L graduated cylinder and
rounded to the nearest 25 ml graduation. Next, the popcorn flakes were visually
inspected and segregated into three shapes based on the whether the flake appendages
were expanded unilaterally, bilaterally, or multilaterally (Figure 2.1).
For each shape, the number of popcorn flakes was counted, weighed, and
volumetrically measured five times using a 4 L graduated cylinder and rounded to the
nearest 25 ml graduation. The average bulk density of each polymorphous shape was
calculated by dividing the pop volume for that shape by the weight of popcorn flakes of
that particular shape; the percentage of each polymorphous shape was calculated by
dividing the number of flakes obtained of a particular shape by the total number of
popped and unpopped kernels; the percent unpopped kernels were calculated by dividing
the number of unpopped kernels by the total number of popped and unpopped kernels.
2.3.3. Compositional and Flavor Analysis of Popped Kernels
Popped kernels of each flake polymorphism were homogenized in a Blixer 3
blender (Robot Coupe, USA) for approximately 20 sec, and then analyzed for
compositional attributes as described for unpopped kernels. Additionally, amino acids
were determined by acid hydrolysis in 6 N HCl followed by derivatization using
orthophthal-aldehyde-9-fluorenylmethyloroformate, and then quantified using HPLC
with fluorescence detection following the method described by Henderson et al (2000).
53
Total dietary fiber was determined using the enzymatic-gravimetric method (Approved
Method 985.29, AOAC International 2010).
Volatile compounds for each popcorn flake polymorphism were characterized
using solid phase microextracton (SPME) coupled to gas chromatography-mass
spectrometry (GC-MS). SPME is a semi-quantitative technique widely applied for flavor
analysis, as it simplifies extraction without using solvents by using an adsorbent silica
fiber to partition analytes which can be desorbed using GC (Steffen and Pawliszyn 1996).
A 0.5 g sample of homogenized popcorn was sealed in a 20 ml magnetic crimp top
headspace vial with PTFE/silicone septa (Gerstel, USA). A 23 gauge, 85 µm
carboxen/polydimethylsiloxane fiber was used for headspace sampling (Supelco, USA).
The fiber was conditioned at 300 °C for 1 hr prior to use according to manufacturers
recommendations. The samples were analyzed with an Agilent 6890 GC coupled to an
Agilent 5973 MSD, equipped with a Gerstel MPS2 autosampler and a Phenomenex, ZB-5
capillary column (60 m, 0.25 mm i.d., 0.25 µm film thickness). The samples were
equilibrated to 70 °C for 10 min before exposing the fiber to the headspace for 25 min at
70 °C. Equilibrium temperature of 70 °C was used in order to increase partitioning of less
volatile compounds into the headspace while minimizing artifact formation. Then the
fiber was immediately extracted in the GC injector in splitless mode at 280 °C for 5 min,
purge was switched on after 3 min to 30 ml/min. Helium was used as the carrier gas at
constant flow (1 ml/min) with an average linear velocity of 26 cm/s. The oven
temperature was kept at 40 °C for 3 min, then increased 4 °C/min to 260 °C, and held for
2 min (run time 60 min). The mass selective detector (MSD) interface was held at 280
°C. Using Chemstation MSD software (Agilent, USA), electron ionization mass spectra
54
were acquired at 70 eV in the m/z range 30–350 mass units. Flavor compounds were
identified by comparing mass spectra to the spectral library found in the NIST Standard
Reference, Database Number 69 (Stein 2008).
2.3.4. Sensory Protocol and Evaluation
Using a central location test held over the course of a single day in Omaha, NE,
USA, 150 consumers evaluated 3 samples of the 3 popcorn flake shapes using a
sequential monadic test design, in which samples were assigned random three-digit codes
and evaluated blindly by panelists with serving order balanced and rotated to account for
order and position effect (Peryam 1958). Consumers were asked to evaluate overall
liking, appearance, shape, flavor, and texture liking for each of the three flake
polymorphs using a 9-point hedonic scale, ranging from 1 (dislike extremely) to 9 (like
extremely) with midpoint of 5 (neither like or nor dislike). In addition, intensity
measures for overall flavor, butter flavor, saltiness, texture, crispness, crunchiness, and
dissolvability were evaluated on a 10-point scale. Intensity measures degree of an
attribute and not liking, and so intensity ranges were described uniquely for each attribute
tested. All panelists were recruited from the Automated Recruitment Calling System
(ARCS) national
database as head of household shoppers over age 18 that had purchased
and eaten microwave popcorn at least three times in the past month and were acceptors of
butter-flavored popcorn. Consumers were given a modest monetary compensation for
their responses.
Immediately after popping in one of the seven aforementioned GE 1150 Watt
microwave ovens, popcorn was poured onto cooking trays that had been stored at 70 ˚C
in an industrial baking unit. Trained servers were assigned to sort for specific shapes to
55
minimize sorting variability. Twelve pieces of each shape were sorted and served in 6
ounce polystyrene containers to consumers. Consumers were instructed to drink water
and consume saltine crackers (unsalted tops) between each sample. Eight bags of fresh
product were popped for every serving position, with additional bags popped on an asneeded basis. In this way, panelists received all samples at a similar temperature within
ten minutes of popping.
2.3.5. Data Analysis
All statistical analyses were performed using SAS software (version 9.1, SAS
Institute, Cary, NC). Pop performance and compositional data were analyzed using
analysis of variance (ANOVA). Fisher‟s LSD test for mean separation was used to
determine significance (p ≤ 0.05) for all pop performance, compositional, and sensory
evaluation data.
Volatile flavor data from SPME procedures were standardized to a baseline of 1.0
for each compound extracted from the bilaterally expanded flakes, which permitted
relative comparisons to be made between peak areas for a given compound. In addition,
volatile flavor analysis from GC-MS was imported and analyzed by principal component
analysis (PCA), which simplified the measured flavor volatiles into a smaller number of
principal components (PC) that account for most of the variance observed. PCA was
performed in SAS using FACTOR procedures with a varimax rotation.
A mass balance was used to compare and identify any divergence between the
weighted average composition of popcorn flake polymorphs to the expected composition
from raw materials (unpopped kernels and slurry). The weighted average composition
56
for popcorn flakes was calculated by multiplying measured analytical values for each
flake polymorphism by the percentage of the shape.
2.4. Results and Discussion
2.4.1. Popping Performance by Flake Morphology
The total average expansion volume prior to sorting was 43.5 cm3/g. Most
kernels were bilaterally expanded (Table 2.3). Unilaterally and multilaterally expanded
kernels were significantly less prevalent, while 7.6 ± 1.4% of kernels remained
unpopped. The multilaterally expanded shape had the largest expansion volume, while
the unilaterally expanded polymorphism had the smallest expansion volume.
During sorting, the unilaterally-expanded shape appeared predominantly towards
the bottom of the bag and was more yellow, shiny, and smooth in appearance than the
other shapes, while the multilaterally-expanded flakes were found more towards the top
of the bag and had more pericarp visually exposed, giving a dry and rough appearance to
the popcorn flakes. Granular mixes of different sized objects are known to separate by
size in response to shaking or stirring with larger particles rising to the top (Rosato et al
1987, Jullien and Meakin 1990). One possible explanation for the distribution of flake
shapes within the microwave popcorn bag is that the explosive force of kernels popping
creates sufficient jostling within the bag to facilitate sedimentation, i.e., the larger
multilateral expanded kernels rise towards the top surface, while the smaller unilaterally
expanded kernels shift into voids created underneath larger flakes and tend towards the
bottom of the bag.
57
Another hypothesis explaining popcorn flake distribution is that kernels may pop
differently depending on location within the bag. Those that expand unilaterally are
those that are inhibited from full, multidirectional expansion because of the weight of the
kernels above them. Bilaterally expanded kernels are those in the middle of the bag,
where some inhibition would be experienced from the kernels on the top, but less than
those on the bottom. Finally, multilaterally expanded kernels are those on the top of the
bag and uninhibited from expansion.
2.4.2. Compositional and Flavor Analysis of Popped Kernels
The unilaterally-expanded shape had significantly higher levels of total fat,
saturated fat, total ash, potassium, and sodium than the bilaterally- and multilaterallyexpanded popcorn flakes (Table 2.4). In contrast, the multilaterally-expanded shape had
the highest levels of protein and total carbohydrates and the lowest levels of fat, saturated
fat, and sodium. While further investigation is required to determine the mechanisms and
factors for forming various popcorn flake polymorphisms, it is noteworthy that
unilaterally-expanded flake shapes had the smallest expansion volume and highest levels
of sodium and fat, while the multilaterally-expanded polymorphism had the largest
expansion volume and lowest level of sodium and fat. These results are consistent with
the observations described above during sorting and suggest there may be an unbalanced
distribution of oil slurry on kernels during the microwave popping. The unilaterally
expanded kernels were located on the bottom of the bag where they were more exposed
to the added fat and sodium, while the multilaterally expanded kernels tended
predominantly to the top of the bag where there would be less exposure to added
ingredients.
58
The weighted average composition of popcorn flakes using the composition for
each polymorph was 30.93% fat, 7.53% protein, 59.05% carbohydrate, and 2.60% ash
(Table 2.5). The mass balance calculations using raw materials revealed that achieving
equivalency to the weighted average flake composition necessitated about 7.5% slurry
not remaining with the popped kernels. While no published research has been found
which quantifies the amount of slurry lost, observations in our laboratory suggest that 510% of slurry adheres to the inside walls of the packaging material during popping.
Reducing this yield loss would seemingly offer a significant cost savings opportunity for
the producer.
SPME-GC-MS procedures identified 48 volatile compounds in the extracts of
popped popcorn flakes, including several volatiles classified as aldehydes, pyrazines, and
alcohols, and a singular lactone and glyceride compound (Table 2.6). Multilaterallyexpanded popcorn flakes had significantly greater levels of 2,5-dimethyl-pyrazine, 2ethyl-5-methyl-pyrazine, 3-ethyl-2,5-dimethyl-pyrazine, and 2-methyl-pyrazine than the
unilaterally-and bilaterally-expanded polymorphisms. On the other hand, multilateral
flake shapes had significantly less of the lactone 2H-Pyran-2-one, tetrahydro-6-pentyl
than did the other two shapes.
PCA of these volatiles revealed grouping among similar popcorn flake shapes and
differentiation between different flake polymorphs (Figure 2.2). Principle component
(PC) factors 1 and 2 accounted for a combined 73.9% of all variation in data. PC factors
3 and 4 accounted for an additional 9.9% and 5.0% of variation, respectively (not shown).
While it is known that most foods are composed of many volatiles, only a limited number
of particular volatile compounds are important to determine flavor (Grosch and
59
Schieberle 1997). The main loadings for PC factor 1 included all pyrazine compounds,
including 2,5-dimethyl-pyrazine, 2-ethyl-5-methyl-pyrazine, 3-ethyl-2,5-dimethylpyrazine, and 2-methyl-pyrazine. Pyrazines have been associated with imparting
characteristic popcorn-like aroma (Schieberle 1991, Buttery et al 1997). Loading for PC
factor 3 included the lactone 2H-Pyran-2-one, tetrahydro-6-pentyl. Lactones have been
shown to characterize sweet cream butter flavor (Mallia et al 2008).
2.4.3. Sensory Evaluation
Overall consumer liking for the different popcorn flake shapes ranged from 6.3 to
7.4 on a 9-point Hedonic scale (Table 2.7). The unilaterally-expanded polymorphism
received significantly higher hedonic scores than the bilateral and multilateral shapes for
overall liking and flavor liking. The unilateral and bilateral polymorphisms scored
significantly higher than the multilateral shape for texture liking. Because test
methodology presented samples to consumers without calling attention to shape, it is
uncertain whether consumers detected flake shape differences. However, no statistical
significance was observed for either appearance or shape liking.
Significant differences were also observed between attribute intensities for the
various popcorn flake polymorphisms (Table 2.8). The unilaterally-expanded flake
polymorph had the highest mean score intensities for overall flavor, butter flavor,
saltiness, texture density, crispness, and crunchiness compared to the bilaterallyexpanded and multilaterally-expanded flake shapes. In contrast, the multilaterallyexpanded flake polymorph had the lowest mean score intensities for overall flavor, butter
flavor, saltiness, and crispness. Intensity scores for crunchiness and texture density for
the bilaterally-expanded and multilaterally-expanded popcorn flakes were not
60
significantly different from one another. Bilaterally-expanded popcorn flakes were
shown to have significantly faster dissolvability than unilaterally-expanded flakes.
2.4.4. Relationship between Composition and Sensory Measures
Unilaterally-expanded popcorn flakes were shown to have the highest overall
product liking in consumer testing despite having the smallest expansion volume,
although the unilaterally-expanded polymorph had the highest levels of total fat, saturated
fat, and sodium, which correlates with having the highest intensity mean scores for
overall flavor, butter flavor, and saltiness. In contrast, multilaterally-expanded flake
polymorphs had the lowest overall product liking, highest expansion volume, lowest
levels of total fat and saturated fat, , and highest levels of protein, total carbohydrates,
and popcorn-like aromatic pyrazines. Consumer panel mean score intensities for butter
flavor correlate with the volatile flavor results showing that multilaterally-expanded
flakes had the lowest levels of the lactone compound 2H-Pyran-2-one, tetrahydro-6pentyl.
Bilaterally-expanded popcorn flakes showed overall product liking scores and
fat and sodium levels between the unilateral and multilaterally-expanded polymorphs.
The results from this study suggest that quality attributes beyond just expansion
volume and texture density of flakes are important considerations for consumer liking.
To the contrary, the sensory evaluation showed that multilaterally expanded flakes had
the least preferred texture despite having the highest expansion volume, although it may
be that consumers simply prefer more oil and salt on popcorn since the sensory attribute
intensity results and compositional data showed that unilaterally-expanded flake
polymorphs had the highest levels of total fat and sodium.
61
2.5. Conclusions
Discernable popcorn flake polymorphisms can be identified and characterized
even within a single hybrid of butterfly popcorn. Significant differences in expansion
values, sensory hedonics, and compositional and flavor attributes exist between different
popcorn flake shapes sorted after popping in a microwave oven. These results may assist
commercial manufacturers in developing better methods for characterizing popcorn flake
polymorphisms. Furthermore, given the potential impact to consumer acceptability, it
may prove useful to research and gain a deeper understanding of factors that affect
popcorn flake polymorphism.
2.6. Literature Cited
Allred-Coyle, T. A., Toma, R. B., Reiboldt, W., and Thaku, M. 2000. Effects of
moisture content, hybrid variety, kernel size, and microwave wattage on the
expansion volume of microwave popcorn. International J. of Food Sci. and
Nutrition. 51:389-394.
AOAC. 2010. Official Methods of Analysis. 18th Ed. Association of Official
Analytical Chemists. Washington, DC.
Borras, F. Seetharaman, K., Yao, N., Robutti, J. L., Percibaldi, N. M., and Eyherabide, G.
H. 2006. Relationship between popcorn composition and expansion volume and
discrimination of corn types by using zein properties. Cereal Chem. 83(1):86-92.
Buttery, R. G., Ling, L. C., and Stern, D. J. 1997. Studies on popcorn aroma and flavor
volatiles. J. Agr. Food Chem. 45:837-843.
62
Ceylan, M., and Karababa, E. 2004. The effects of ingredients on popcorn popping
characteristics. International J. Food Eng. 39:361-370.
D‟Croz-Mason, N. and Waldren, R. 1990. Popcorn Production. NebGuide G78-426.
University of Nebraska, Cooperative Extension, Lincoln, NE.
Dofing, S. M., D‟Croz-Mason, N., and Buck, J. S. 1990. Inheritance of expansion
volume and yield in two popcorn X dent corn crosses. Crop Sci. 31:715.
Eldredge, J. C. and Thomas, W. I. 1959. Popcorn – its production, processing, and
utilization. Iowa Agric. Exp. Stn. Bull P127. 1-16.
Grosch, W. and Schieberle, P. 1997. Flavor of cereal products – a review. Cereal Chem.
74(2)91-97.
Henderson, J. W., Ricker, R. D., Bidlingmeyer, B. A., and Woodward, C. 2000. Amino
acid analysis using Zorbax Eclipse-AAA columns and the Agilent 1100 HPLC.
Application note from Agilent Technologies.
Hutchins, J. B. 1977. The importance of visual appearance of foods to the processor and
the consumer. J. Food Quality. 1(3):267-278.
Jullien, R. and Meakin, P. 1990. A mechanism for particle size segregation in three
dimensions. Nature. 344:425-427.
Levy, B. 1988. A new perspective on popcorn. Snack World. 45(1):24.
Lin, Y. E. and Anatheswaran, R. C. 1988. Studies on popping of popcorn in a
microwave oven. J. Food Sci. 53(6):1746-1749.
Lyerly, P. J. 1942. Some genetic and morphological characters affecting the popping
expansion of popcorn. J. Am. Soc. Agron. 34:986.
63
Mallia, S., Escher, F., and Schlichtherle-Cerny, H. 2008. Aroma-active compounds of
butter: a review. European Food Research & Tech. 226(3):315-325.
Matz, S. A. 1984. Snacks based on popcorn. Pages 133-144 in: Snack Food Technology,
2nd Ed. AVI Pub. Co.: Westport, CT.
Mohamed, A. A., Ashman, R. B., and Kirleis, A. W. 1993. Pericarp thickness and other
kernel physical characteristics relate to microwave popping quality of popcorn. J.
Food Sci. 58(2):342-346.
Pace, W. E., Westphal, W. B., and Goldblith, S. A. 1968. Dielectric properties of
commercial cooking oils. J. Food Sci. 33:30.
Park, D., and Maga, J. A. 2001. Color, texture, and sensory evaluation of selected
hybrids of popped popcorn. J. Food Quality. 24(6):563-574.
Park, D., Allen, K. G. D., Stermitz, F. R., and Maga, J. A. 2000. Chemical composition
and physical characteristics of unpopped popcorn hybrids. J. Food Composition
and Analysis. 13:921-934.
Peryam, D. R. 1958. Sensory difference tests. Food Tech. 12(5):231-236.
Peryam, D. R. and Pilgrim F. J. 1957. Hedonic scale method of measuring food
preferences. Food Tech. 11(9):9-14.
Rosato, A., Strandburg, K. J., Prinz, F., and Swendsen, R. H. 1987. Why the Brazil nuts
are on top: size segregation of particulate matter by shaking. Physical Review
Letters. 58:1038-1040.
Quinn Sr., P. V. Hong, D. C., and Both, J. A. 2005. Increasing the size of a piece of
popcorn. Physica A. 353:637-648.
Schieberle, P. 1991. Primary odorants in popcorn. J. Agr. Food Chem. 39:1141-1144.
64
Schiffmann, R. F. 1986. Food product development for microwave processing. Food
Technol. 40(6):94.
Song, A., Eckhoff, S. R., Paulsen, M., and Litchfield, J. B. 1991. Effects of kernel size
and genotype on popcorn popping volumes and number of unpopped kernels.
Cereal Chem. 68(5):464-467.
Steffen, A., and Pawliszyn, J. 1996. Analysis of flavor volatiles using headspace solidphase microextraction. J. Agric. Food Chem. 44:2187-2193.
Stein, S. E. 2008. Mass Spectra in: NIST Chemistry, NIST Standard Reference
Database Number 69. P. J. Linstrom and W. G. Mallard, eds. National Institute
of Standards and Technology: Gaithersburg, MD.
The Popcorn Board [On-line]. 2010. Available: http://www.popcorn.org.
Tian, T., Buriak, P., and Eckhoff, S. R. 2001. Effect of hybrid and physical properties of
individual popcorn kernels on expansion volume. Cereal Chem. 78(5):578-582.
Watson, S. A. 1988. Corn marketing, processing, and utilization. Page 885 in: Corn and
Corn Improvement. G. Sprague and J. Dudley, eds. American Society of
Agronomy: Madison, WI.
Whistler, R. L., and Daniel, J. R. 1985. Carbohydrates. Pages 69 in: Food Chemistry. O.
Fennema,ed. Marcel Dekker: New York.
Zhu, T., Jackson, D.S., Wehling, R.L., and Geera, B. 2008. Comparison of amylose
determination methods and the development of a dual wavelength iodine binding
technique. Cereal Chem. 85(1)51-58.
65
Ziegler, K.E., Ashman, R.B., White, G.M., and Wyson, D.S. 1984. Popcorn Production
and Marketing. Cooperative Extension Service, Purdue University, West
Lafayette, Indiana. National Corn Handbook.
Ziegler, K. E. Chapter 7: Popcorn. 2001. Pages 199-234 in: Specialty Corn. A. R.
Hallauer, ed. CRC Press LLC: Boca Raton, FL.
Ziegler, K. E. 2003. Popcorn Chapter in: Corn: Chemistry and Technology. P. J. White
and L. A. Johnson, eds. American Association of Cereal Chemists: St. Paul, MN.
66
Figure 2.1. Popcorn flakes inspected and segregated into three shapes: expanded
unilaterally (A), bilaterally (B), or multilaterally (C).
67
Figure 2.2. PCA of flavor volatiles measured by GC-MS grouped among similar popcorn
shapes and differentiation between different flake polymorphs. Vector loadings for
pyrazine compounds also illustrated. (circles = bilateral; squares = unilateral; triangles =
multilateral)
68
Table 2.1. Composition of Butterfly Popcorn Hybrid
Analyte
Mean Valuea
per 100g sample
13.91 ± 0.06 g
2.95 ± 0.15 g
0.53 ± 0.02 g
Moisture
Total Fat
Saturated Fat
Fatty Acids
Palmitic (%) C16:0
0.49 ± 0.02 g
Stearic (%) C18:0
0.05 ± 0.00 g
Oleic (%) C18:1
0.61 ± 0.04 g
Linoleic (%) C18:2
1.71 ± 0.09 g
Linolenic (%) C18:3
0.04 ± 0.00 g
All Other Fatty Acids (%)
0.04 ± 0.00 g
Total Carbohydrate
73.20 ± 0.15 g
Total Starch
70.54 ± 2.84 g
b
Amylose
11.92 ± 0.35 g
c
Amylopectin
58.62 ± 0.35 g
Sugarsd
0.44 ± 0.09 g
Total Protein
9.05 ± 0.18 g
Total Ash/Minerals
0.90 ± 0.07 g
Sodium
Trace
Potassium
232 ± 20 mg
Calcium
4.07 ± 1.66 mg
Iron
2.07 ± 0.22 mg
Phosphorus
237mg ± 21 mg
a
Analyses run in triplicate.
b
Amylose determined by method described by Zhu et al (2008).
c
Amylopectin determined by total starch – amylose.
d
Only sucrose was detected.
69
Table 2.2. Microwave Popcorn Formulation
Amount Added
Total Weight
Per Bag (g)
(%)
Popcorn kernels
62.0
67.00%
a
Oil
28.0
30.26%
Salt (NaCl)
2.10
2.27%
Flavorb
0.40
0.43%
c
Color
0.04
0.04%
Total
92.54
100.00%
a
Oil used was blend of palm oil and palm oil stearine from ADM (#84-650-0).
b
Flavor used was butter-type flavor from Gilroy Foods (Gen-4981).
c
Color used was oil suspension of annatto, turmeric, and paprika oleoresin (ATP-1279,
Chr.Hansen, Denmark).
Ingredient
70
Table 2.3. Distribution and Pop Performance Measures of Three Popcorn Flake
Polymorphisms and Unpopped Kernels Produced After Microwave Popping
Means Valuesa
Unilaterally
Bilaterally
Multilaterally Unpopped
Expanded
Expanded
Expanded
Kernels
Count
32.8b
263.2a
45.0b
27.8
Shape Type or Unpopped (%)
9.0b
71.2a
12.3b
7.6
Mass (g)
8.1b
54.9a
9.0b
4.0
b
Pop Volume (mL)
232c
2360a
482b
c
3
Expansion Volume (cm /g)
28.6c
43.0b
53.5a
a
Values in same row followed by different letter indicate significant difference at P≤0.05 using
Fisher‟s LSD; n=5. (-) Means do not include unpopped data.
71
Table 2.4. Chemical Composition of Three Popcorn Flake Polymorphisms After
Microwave Popping
Mean Valuesa
Unilaterally Bilaterally Multilaterally
Analyte
Expanded
Expanded
Expanded
Total Fat (%)
34.41a
30.24b
27.23c
Saturated Fat (%)
16.81a
14.73b
13.05c
Trans Fat (%)
0.09a
0.08ab
0.07b
Protein (%)
6.75c
7.31b
8.12a
Total Ash (%)
2.76a
2.52b
2.54b
Sodium (mg / 100g)
767a
662b
621c
Potassium (mg / 100g)
216a
147b
202ab
Calcium (mg / 100g)
16.9a
17.2a
15.3a
Iron (mg / 100 g)
2.1a
1.5a
2.0a
Phosphorus (mg / 100g)
260a
173a
263a
Total Carbohydrates (%)
54.39c
57.71b
60.38a
Digestible Fiber (%)
8.95a
8.94a
10.75a
Moisture (%)
1.69a
2.22a
1.73a
a
Values in same row with different letter indicate significant difference at P≤0.05 using
Fisher‟s LSD; n=5.
Table 2.5. Mass Balance Comparison of Raw Materials to Average Popcorn Flake Composition
Weighted Average Composition
Theoretical Composition
Theoretical
Using Flake Polymorph
Using Raw Materials (no
Composition with 7.5%
Analyticalsa
yield loss)b
Slurry Loss Applied
Total Fat (g/100g)
30.93
33.05
30.83
Protein (g/100g)
7.53
7.04
7.30
Total Ash (g/100g)
2.60
2.94
2.80
Total Carbohydrates (g/100g)
59.05
56.97
59.07
Moisture (g/100g) c
0.00
0.00
0.00
Total Weight (g)
100.00
100.00
100.00
a
Calculated by weighting actual analytical for each popcorn flake polymorphism with weighted 77.0% bilateral, 13.3%
multilateral and 9.7% unilateral.
b
Calculated by weighting analytical values for raw materials.
c
All mass balance calculations done on dry basis.
72
Table 2.6. Volatiles Obtained from Three Popcorn Flake Polymorphisms After Microwave Popping (see next page for footnotes)
Analyte
Unilaterally Expanded
Multilaterally Expanded
1.19b
0.98a
1.07a
0.94ab
1.04a
1.23b
0.69c
0.68c
1.01a
1.07a
0.96a
1.05ab
1.15b
1.23a
1.10a
0.72c
1.08a
0.71b
1.02a
1.07a
0.89b
1.01a
0.88a
0.85b
1.61b
0.98b
0.75c
1.00c
1.00a
1.00a
1.00a
1.00a
1.00b
1.00b
1.00b
1.00a
1.00ab
1.00a
1.00b
1.00c
1.00a
1.00a
1.00b
1.00a
1.00b
1.00a
1.00a
1.00a
1.00a
1.00a
1.00a
1.00c
1.00b
1.00b
1.59a
0.75b
0.65b
0.91b
0.65a
2.50a
1.50a
1.21a
0.86b
0.92b
0.73b
1.09a
1.37a
1.19a
1.10a
1.53a
0.54b
1.66a
0.45a
0.85b
0.94ab
0.65b
0.97a
0.85b
1.90a
1.37a
1.53a
73
1-Decene
1-Dodecene
2,5-Cyclohexadiene-1,4dione,2-(1,1-dimethyetyl)
2,6-Dimethyldecane
2-Dodecene, (Z)2-Methoxy-4-vinylphenol
2-Nonene, 3-methyl-, (E)2-Trifluoromethylbenzone
2H-Pyran-2-one, tetrahydro-6-pentyl
3,5-Dimethyldodecane
3,5-Dimethyldodecane1
Acetaldehyde
Acetonitrile
Butane, 2,2,3,3-tetramethyl
Butanoic acid, 2-butoxy-1-methyl-2-oxoethylester
Cyclododecane
Cyclopropane, nonylCyclopropane, octylCyclotetrasiloxane, octamethyl
Decane
Decane, 4-ethylDodecane
Dodecane, 4-methylDodecane1
Ethanol
Furan, tetrahydroFurfural
Mean Valuesa
Bilaterally Expanded
Table 2.6 (continued). Volatiles Obtained from Three Popcorn Flake Polymorphisms After Microwave Popping
Analyte
a
Hexadecane
Hexanal
Hexane
Methyl Isobutyl Ketone
Nonanal
Nonane, 3-methylNonane, 3-methyl-5-propylOxirane, (ethoxymethyl)Propanoic acid, 2-methylPyrazine, 2,5-dimethylPyrazine, 2-ethyl-5-methylPyrazine, 3-ethyl-2,5-dimethyl
Pyrazine, methylTetradecane
Toluene
Triacetin
Tridecane, 3-methylTridecane, 3-methyl-1
Tridecane, 3-methyleneUndecane, 3-methylUndecane, 3-methylene-
Unilaterally Expanded
Mean Valuesa
Bilaterally Expanded
Multilaterally Expanded
1.26a
0.95b
1.18a
1.27a
1.05b
0.92a
0.88b
0.95a
0.87a
1.03b
0.90b
1.05b
0.76b
1.11a
1.32a
1.06a
0.93a
1.18a
0.86b
0.89b
0.79b
1.00b
1.00
1.00a
1.00b
1.00b
1.00a
1.00a
1.00a
1.00a
1.00b
1.00b
1.00b
1.00b
1.00a
1.00a
1.00ab
1.00a
1.00b
1.00a
1.00a
1.00a
0.48c
1.59a
1.11a
1.36a
1.60a
0.93a
0.98a
1.29a
0.90a
1.86a
1.70a
1.65a
1.38a
0.49b
1.45b
0.96b
0.82b
0.68c
0.87b
0.87b
1.10b
Reported peak area standardized for each analyte relative to 1.00 for the bilaterally expanded values. Values in same row followed by different
letter indicate significant difference at P≤0.05 using Fisher‟s LSD; n=4.
74
Table 2.7. Consumer Sensory Mean Scores for Three Popcorn Flake Polymorphisms
Mean Valuesa
Measured Likingb
Unilateral
Bilateral
Multilateral
Overall Product Liking
7.4a
7.0b
6.3b
Appearance Liking
7.6a
7.6a
7.4a
Shape Liking
7.5a
7.5a
7.4a
Flavor Liking
7.3a
6.8b
6.3b
Texture Liking
7.3a
7.0a
6.5b
a
Values in the same row followed by different letters indicate significant difference
at P≤0.05 using Fisher‟s LSD.
b
Liking scale of 1-9 for “dislike extremely” to “like extremely”; n=150 total.
75
Table 2.8. Attribute Intensity Means Scores for Three Popcorn Flake Polymorphisms
Mean Valuesa
Intensity Measurementb
Overall Flavor
(1=Extremely Weak to 10=Extremely Strong)
Butter Flavor
(1=Extremely Weak to 10=Extremely Strong)
Saltiness
(1=Not at all Salty to 10= Extremely Salty)
Texture
(1=Extremely Light/Fluffy to 10= Extremely Dense/Compact)
Crispness
(1=Not at all Crisp to 10= Extremely Crisp)
Crunchiness
(1=Not at all Crunchy to 10=Extremely Crunchy)
Dissolvability
(1=Extremely Slow to 10=Extremely Fast)
a
b
Unilaterally
Expanded
Bilaterally
Expanded
Multilaterally
Expanded
7.1a
5.7b
4.9c
7.0a
5.7b
4.8c
5.7a
4.8b
4.1c
5.7a
4.7b
4.7b
7.2a
6.3b
5.7c
7.0a
6.0b
5.5b
5.9b
6.3a
6.0ab
Values in same row followed by different letters indicate significant difference at P≤0.05 using Fisher‟s LSD.
Intensity measured on a 10-pt scale of 1-10. Intensity measures degree of an attribute (not liking); n=150 total.
76
77
CHAPTER 3. HYBRID AND ENVIRONMENT EFFECTS ON POPCORN KERNEL
PHYSIOCHEMICAL PROPERTIES AND THEIR RELATIONSHIP TO
MICROWAVE POPPING PERFORMANCE
3.1. Abstract
The objective of this study was to characterize the effect of hybrid and
environment on physical and chemical characteristics of popcorn kernels that have shown
importance in predicting end-use quality. Three popcorn hybrids grown in three different
environments were tested for physiochemical attributes and popping performance.
Hybrid had a significant effect on kernel sphericity, time-to-grind, dietary fiber, sugars,
and starch. Environment effect alone affected total mineral content. Hybrid and
environment main effects influenced test weight, tangential abrasive dehulling device
index, thousand kernel weight, total carbohydrates, and kernel protein content. Oil
adherence to the bag averaged 15.8% and was proportional to oil amount added prior to
microwave popping. Unpopped kernels averaged 11.4 ± 5.3%. Most unpopped kernels
were observed to successfully pop when heated a second time in microwave tests.
Expansion volume was 44.7 ± 3.7 and 47.3 ± 6.4 cm3/g, depending on the method of
determination. Expansion volume was correlated (p<0.05) with several kernel
physiochemical parameters that were influenced by hybrid effect. Sphericity, thousandweight, and total fat are physiochemical characteristics that appear to be good predictors
(p<0.05) of expansion volume.
78
3.2. Introduction
The most studied and reported quality factors for popcorn (Zea mays Everta) have
been expansion volume and the number of unpopped kernels (Ziegler, 2001). High pop
expansion volume is correlated with desirable consumer attributes including improved
texture and consumer acceptability (Dofing et al., 1990; Ceylan and Karababa, 2001),
while unpopped kernels are an undesirable nuisance to consumers and unrealized profit
for producers.
Popcorn expansion volume has been shown to be a quantitative trait with high
heritability and influenced by three to five major genes (Ziegler, 2001; Lu et al., 2003).
Breeding methods used to improve dent corn are commonly used for popcorn, with the
primary distinction being that the trait of emphasis in popcorn has historically been
popping performance whereas dent corn breeders have usually focused on grain yield
(Ziegler, 2001). While expansion volume tends to be negatively correlated with grain
yield, breeding methods which use additive genetic variation for expansion volume and
dominance variation for grain yield will likely result in simultaneous improvement in
both producer yield and end-use quality traits (Dofing et al., 1990).
Correlating intrinsic kernel characteristics with final performance quality
measures can provide simple and reliable prediction methods to breeders and producers
(Dorsey-Redding et al., 1991). As such, several studies have investigated and elucidated
popcorn kernel characteristics that influence expansion volume. Some of the phenotypic
characteristics shown to affect pop volume include physical traits such as kernel size,
shape, density, hardness, and pericarp thickness, as well as composition attributes such as
levels of zein protein and the types of fatty acid in the kernel (Borras et al., 2006).
79
Evaluating genetic and environmental mechanisms which influence phenotypic
characteristics and predict performance is an essential part of plant breeding (Stuber et
al., 1992). The physical and biochemical characteristics of popcorn kernels are known to
vary depending on the genetics, growing environment, and agronomic practices used
(Ziegler, 2001). Yet while end-quality popping performance has been the primary focus
of growers and popcorn breeding programs (Dofing et al., 1991), the effect of hybrid and
environment on physical and biochemical properties of popcorn kernels and the
relationship among these properties has not been comprehensively studied. The
objectives of this investigation were to determine the effects of hybrid and growing
environment on the physical properties and biochemical composition of popcorn kernels
and establish the relationships among intrinsic kernel quality characteristics, as well as to
relate kernel properties to end-use microwave popping performance.
3.3. Experimental
3.3.1. Popcorn Samples
Three commercial, butterfly-type popcorn hybrids, YPK-213, YPK-313, and
YPK-321, were supplied by ConAgra Foods, Inc. (Omaha, NE USA). Hybrids were
planted and grown in strip plots in Nebraska, Indiana, and Ohio, three states in the
Midwestern USA, during 2009. These three states annually produce over 250,000 kg of
popcorn, which represents 65% of the commercial popcorn grown in the USA (National
Agricultural Statistics Service, 2007). The Nebraska strip plot was irrigated land, while
Indiana and Ohio strip plots were non-irrigated dryland. Fertilization, weed, and pest
control were applied according to standard agronomic practices at each location. All
80
hybrids were harvested and thrashed by combine at commercial maturity when their field
moisture contents were less than 18%. Samples (a total of three from each plot) were
cleaned, bulked packed, and tempered to 14% moisture by storing at 21.5 ˚C and 73% rh
until used for experimentation, which occurred from May 2010 – January 2011.
3.3.2. Composition and Physical Property Testing
Compositional and physical properties analyses were performed on conditioned
popcorn samples after equilibrating for a minimum of 120 d at storage conditions and
verifying desired moisture content of 14.0 ± 0.5% had been achieved as measured using a
Dickey-john GAC III moisture analyzer. More accurate kernel moisture contents were
determined using the method described below. Kernel moisture content of 14.0% was
selected for the target moisture content as it has been reported to be the optimum level for
microwave popping performance (Ziegler, 2001; Gökmen, 2004). Thousand kernel
weight (g) was measured by randomly selecting and weighing 1,000 intact kernels. Test
weight (kg/hl) was measured using the United States Department of Agriculture (USDA)
Federal Grain Inspection Service Method (1990). True kernel density (g/cm3) was
determined using an air-comparison pycnometer (Pomeranz et al., 1984; Quantachrom
pycnometer Boynton Beach, FL) equipped with ultra-pure nitrogen as displacement gas.
Tangential abrasive dehulling device (TADD) index was determined by calculating the
percentage of kernel weight remaining after abrading 40g of sample for 10 min using a
TADD instrument (Venables Machine Works LTD, Saskatoon, Saskatchewan, Canada)
with no. 36 grit disk under constant vacuum aspiration (Reichert et al., 1986). Stenvert
kernel hardness was determined by measuring time required to grind 20 g of intact
popcorn kernels and the ground sample height (mm) using a Kinematica AG Polymix
81
Grinder (Model PX-MFC 90 D, Lucerne, Switzerland) with a 2 mm screen and operated
at 3,600 rpm measured using the integrated tachometer (Pomeranz et al., 1985). Average
kernel sphericity was calculated as the ratio of geometric mean diameter (cubic root of
the multiplied length, width, and thickness) and kernel length for 30 randomly-selected
kernels per sample, as measured with a digital micrometer (Mitutoya America Corp.,
Aurora, IL). Five observations were made for each experimental unit for thousand
kernel weight, pycnometer density, and Stenvert hardness tests. Six observations were
taken per sample for TADD index, seven sample reps were observed for test weight, and
thirty observations per sample were made for sphericity.
Compositional characteristics were measured for the popcorn kernels after using a
two-step milling process of chopping for 20 sec in a Blixer 3 blender (Robot Coupe,
USA) and then grinding with a Kitchen Mill Model 91 (Blentech, USA). Analyses of
total fat, saturated fat, fatty acids, fiber, sugar, moisture, and minerals were determined
using standard methods (AOAC Approved Methods 996.06, 950.46, 992.15, 984.27),
while protein, total carbohydrates, sugars, and starch were determined using the
preparation and method modifications for popcorn as described previously (Sweley et al.,
2011). Amylose and amylopectin were measured using the modified dual-wavelength
iodine binding procedure described by Zhu et al. (2008). For all compositional analysis,
three replications were made per sample, and all reported data were adjusted to 14.0%
moisture basis (Dorsey-Redding et al., 1990).
3.3.3. Pop Performance Testing
A randomized, D-optimal design requiring a total of 65 runs was developed using
Design Expert statistical software (version 8.0, Stat-Ease, Inc., Minneapolis, MN) in
82
order to assess popping performance across a range of prevalent microwave wattages and
different amounts of oil added to the kernels. The 9 hybrid*environment popcorn
samples previously described were used as the kernels for experiments. Popcorn kernels
and oil were weighed into pre-folded, commercially standard, bi-layer microwavable
paper bags (14.92 cm x 29.53 cm) containing an aluminum-coated polyester susceptor
(13.65 cm x 16.51 cm) embedded at the bottom center of the bag gusset. The total weight
of kernels and oil was fixed at 77 g, which was established after conducting pretesting to
verify that popping would not be constricted by the bag. The corn-oil ratios used in this
study (2:1 to 16:1) were selected to encompass the prevalent range of oil levels (6% to
30%) found in microwave popcorn products being sold in the US market as found in the
USDA National Nutrient Database for Standard Reference (2010) for retail microwave
popcorn products. The oil used for all testing was a commercially-available blend of
palm oil and palm oil stearine (Archer Daniels Midland Company product 845600,
Decatur, IL).
Popping tests were completed using the following three microwave ovens with
distinct power outputs, internal cavity dimensions, and manufacturers: (a) Panasonic
model NN-H765BF (1250 W, 0.45 m3 internal cavity), (b) General Electric model
JE1160WC (1100 W, 0.31 m3 internal cavity), and (c) Samsung MW830WA (800-Watt,
0.17 m3 internal cavity). The actual power output for the three microwaves was 1240,
1050, and 750 W, respectively, which was determined by taking 70% of the measured
temperature change of 1000 g of deionized water (20˚C) after heating for 62 s on high
power (Schiffmann, 1987).
83
For each experimental run, microwave popping tests were performed immediately
after bag preparation by placing unfolded bags in the center of microwave with
susceptor-side down and popping on high power until the interval between pops slowed
to 2 s. After popping, bag contents were poured into a steel sieve with round-hole, 7.94
mm openings (Seedburo, USA Model 0070) to remove unpopped kernels. The empty
paper bag and any residual oil was weighed, and used to calculate the percentage (by
weight) of original oil lost to the bag during popping. Popped popcorn flakes were
weighed and then poured into a 4 L graduated cylinder, which was inverted once and the
resultant pop volume rounded to the nearest 25 mL graduation.
The conventional manner of reporting microwave popcorn expansion volume has
been the volume of popped corn per original weight of unpopped popcorn (Dofing et al,
1990; Mohamed et al., 1993). An alternative method for determining expansion volume
has been reported by Pordesimo et al. (1990) as the ratio of pop volume per weight of
popped kernels. Both methods of determining expansion volume were calculated and
reported in this study.
In order to further examine the potential pop viability of kernels which did not
pop, unpopped kernels from each hybrid*environment sample were collected and
combined. Using the same paper bags described previously, 28 g unpopped kernels were
weighed without any additional oil and popped using the 1250 W Panasonic microwave.
The resultant number of unpopped kernels were counted, weighed, and converted to a
percentage of the preceding 28 g starting weight. The limited amount of unpopped
kernels collected during initial experiments constrained the starting kernel weight and
permitted only a single sample replicate per treatment combination for re-pop testing,
84
although two-way ANOVA analysis could still be used for mean comparisons between
hybrids and environment effects (n=3).
3.3.4. Statistical Analysis
All statistical analyses were performed using SAS software (version 9.2, SAS
Institute, Cary, NC). Kernel physical and compositional properties and pop performance
comparisons used two-way analysis of variance (ANOVA) with fixed effects for hybrid
and environment to make comparisons using Fisher‟s least significant differences (LSD)
test at α = 0.05. Only main effects of hybrid and environment were included ANOVA
calculations since the experimental design did not include field plot replication. The
relationship of kernel hybrid and environment to kernel physiochemical parameters, as
well as the relationship between physiochemical parameters and popping performance
measures, were analyzed using the SAS correlation (CORR) procedure.
3.4. Results and Discussion
3.4.1. Kernel Physicochemical Parameters
Significant differences in physical and compositional properties between the
popcorn kernels were observed (Table 3.1). Test weight varied significantly among the
samples (p<0.05), which might be due to test weight being a measure of bulk density and
therefore dependent on many factors beyond the kernel density including kernel size and
shape, packing, and void volume, among others (Lee et al., 2007). While thousand kernel
weight eliminates the variability in measurement caused by packing and void volume, the
results of this study showed thousand kernel weight measures to be more variable than
85
test weight. In contrast, pycnometer density was shown to have the highest
reproducibility among all physical properties tested, which is consistent with results
reported for conventional maize (Lee et al., 2007). Because pycnometer density provides
an indirect indication of the percent of hard and soft endosperm found in the kernel (Lee
et al., 2007), it may be especially relevant for popcorn since it has been proposed that
popcorn kernels with larger amounts of hard endosperm will have greater expansion ratio
(Hoseney et al., 1983).
The values for thousand kernel weight, test weight, pycnometer density, and
kernel sphericity were consistent with the range of values reported previously for popcorn
kernels (Pordesimo et al., 1990; Mohamed et al., 1993; Ceylan and Karababa, 2001).
Ertas et al. (2009) reported lower values for thousand kernel weight and test weight than
the results observed in this study, probably due to testing by Ertas et al. being conducted
on kernels conditioned to below 8% moisture content versus the 14% moisture content
used in this study and recommended for optimum pop performance. Values for kernel
sphericity were also consistent with the range of values previously reported previously
for popcorn kernels (Pordesimo et al., 1990; Mohamed et al., 1993; Ceylan and
Karababa, 2001; Ertas et al. 2009).
TADD index has not been commonly reported for popcorn, but it showed
relatively low variability in test measurements and provided delineation between different
hybrids and growing environments. In addition, the post-abrasion samples revealed a
remarkably vibrant yellow appearance in the kernel endosperm that was not observed
with grinding or fracturing kernels. The range of TADD index values observed for
experimental units in this study was 38.2 – 49.1% with an average of 43.5 ± 3.4%, which
86
is greater than the range of mean TADD indices (27.8 – 31.9%) reported for four clusters
of diverse conventional maize hybrids (Lee et al., 2007). This is to be expected since the
TADD index measures hardness by abrading the outer kernel layers, and the cell wall
matrix found in popcorn pericarp has greater structural organization than conventional
corn, which is a distinguishing featuring that enables pop mechanics (Hoseney et al.,
1983; Tandjung et al., 2005). Thus, it may be that TADD index provides a rapid, indirect
measure of the relative pericarp organization for popcorn.
The compositional analysis also revealed significant differences among the
popcorn samples (Table 3.2). The ranges of protein content and starch observed for
individual popcorn kernel samples in this study were similar to those reported previously
for popcorn by Park et al. (2000). Amylose and amylopectin concentrations were lower
than previously reported for popcorn kernels (Park et al., 2000; Borras et al., 2006),
which may simply be attributed to sample variation between the studies, or it may be due
to differences in analytical methodologies.
Across all hybrid and environment treatments, the fatty acid composition of
popcorn kernel lipids consisted primarily of linoleic, oleic, and palmitic acids, while
stearic, linolenic, and all other fatty acids were less than 2% (Table 3.2). The
preponderance of fatty acids are similar to values reported previously (Park et al., 2000;
Borras et al. 2006).
Several notable relationships are observed from analysis of the correlation
coefficients among the popcorn kernel physical and chemical properties (Table 3.3).
Protein had a significant inverse correlation with both sugars and total carbohydrates.
Total carbohydrates were positively correlated with TADD index but negatively
87
correlated with time-to-grind, while protein was negatively correlated with TADD index
and positively correlated with time-to-grind. The positive correlation between protein
and time-to-grind is consistent with previous studies in maize (Dorsey-Redding et al.,
1991; Shandera et al., 1997). Starch was positively correlated with amylopectin and
negatively correlated with dietary fiber, but no relationships between starch and physical
measures were observed. While total fat was predictably correlated with several fatty
acids, total fat was not related to any other physical or chemical measure, whereas
relationships between oil and starch and oil and kernel density have been reported for
conventional maize (Dorsey-Redding et al., 1991). Kernel sphericity had a significant
inverse correlation with sugars and oleic acid, while test weight and thousand-weight
were not correlated with compositional measures.
3.4.2. Hybrid and Environment Influence on Kernel Properties
Hybrid and/or growing location had a significant effect on most physiochemical
parameters for popcorn kernels measured in this study, except percent total fat,
pycnometer density, and Stenvert column height. Both hybrid and environment main
effects had a significant (p<0.01) effect on physical kernels tests of thousand kernel
weight, test weight, TADD index, as well as the levels of total carbohydrates and protein.
Hybrid main effect alone was shown to influence sphericity, time-to-grind, dietary fiber,
sugars, and total starch. These results are generally consistent with previous studies on
popcorn showing that hybrid has a significant effect on some select kernel physical
and/or compositional characteristics (Park et al., 2000; Ceylan and Karababa, 2001;
Soylu and Tekkanat, 2007), although Park et al. (2000) also found hybrid to have an
88
effect on total fat, and no significant effects were observed in kernel fat in this study,
perhaps due to the specific hybrids used.
Environment was shown to be the only effect showing significance for total
minerals. Further analysis in this study specifically uncovered that iron, phosphorus, and
potassium contents of popcorn kernels were significantly affected by environment. These
results agree with previous findings in other cereal grains that levels of potassium,
phosphorus, and iron vary as a result of environmental factors such as the use of
irrigation versus rainfall, fertilizer content, and the nature of the soil itself (Greaves and
Hirst, 1929).
The advancement in quantitative genetics and mapping of quantitative trait loci
(QTL) for popcorn expansion volume might suggest that estimating the magnitude of
variance for kernel physicochemical attributes is immaterial, but it is unlikely that a
single genotype exists that will perform best in all environments, even when those
environments are relatively similar (Allard and Bradshaw, 1964). To wit, investigation
of genetic mechanisms for popcorn by Li et al. (2006) revealed that only 22.7% of QTLs
identified as influencing expansion volume were expressed in different environments,
suggesting a complex relationship between genetics, environment, and agronomic inputs
for popcorn. Moreover, it is possible that new end-use quality factors beyond expansion
volume will emerge as desirable traits for popcorn that will place new demands on
breeders (Ziegler, 2001). Plant breeding efforts to fulfill new demands might be achieved
through better understanding the intermediate factors and pathways for achieving end
performance and making more precise selections (Allard and Bradshaw, 1964).
89
3.4.3. Microwave Popping Performance
Across all hybrids and environments, unpopped kernels averaged 11.4% and
showed a high coefficient of variation (46.5%), which is consistent with the 10-12%
unpopped kernels that consumers typically find in home microwave popping (Quinn Sr.
et al., 2005) and the high variability reported in previous studies for unpopped kernels
after microwave popping (Dofing et al., 1990; Mohamed et al., 1993; Gökmen 2004;
Soylu and Tekkanat 2006; Ertas et al., 2009).
It has been previously proposed that popping temperature is the most critical
factor determining the viability of popcorn kernels to pop (Hoseney et al., 1983). Most
unpopped kernels were observed to successfully pop when collected and heated a second
time in microwave tests (Table 3.4), performing with a comparable average percentage of
unpopped kernels (11.5 ± 2.6%) as the original testing (11.4 ± 5.3%). This suggests that
most unpopped kernels are not inherently unviable, but rather that not all kernels achieve
the minimum thermodynamic requirements for popping during microwave heating. In
fact, the power absorbance and threshold internal pressure and temperature required for
popping is known to vary for individual kernels (Byrd and Perona, 1995). Thus, it may
be that unpopped kernels are those kernels in the bag that are shielded from microwaves
by other kernels or are positioned along the sides of the bag and do not absorb sufficient
reflective energy from the susceptor to achieve critical temperatures required for popping.
This may explain why none of the kernel physiochemical parameters correlated with
unpopped kernels (Table 3.5).
Any consumer that has ever reached inside a microwave popcorn bag can attest to
the dissatisfaction of messy oil transfer when the hand brushes against the inside of the
90
bag. An average of 15.8 ± 2.3% by weight of oil remains with the bag after popping
(Table 3.4). As illustrated in Figure 3.1, the amount of oil adhering to the bag has a
linear relationship with the original oil amount within the constraints used in this study.
Oil loss to the bag represents a considerable opportunity cost for commercial producers.
It is probable that using alternative packaging materials or coatings on the bag and/or
using different types of oil would affect the oil lost to the bag during microwave popping
and would be worth investigating.
Popping performance results also showed significant differences between samples
for expansion volume when using both of the methods of determining expansion volume
(Table 3.4). Expansion volumes averaged 44.7 ± 3.7 cm3/g across all runs when the
using the prevalent method of calculating expansion volume as a function of only the
original kernel weight.
Despite the prevalence of this method of determination, it does
not account for the variation in popped flake density as a result of other ingredient
additions, nor does it account for loss factors during popping. As discussed above, the
percentage of unpopped kernels that do not pop can be significant and highly variable. In
addition, the density of popped popcorn flakes will vary depending on the amount of oil
coverage. This is an important consideration since most prepackaged microwave
popcorn sold in the US is coated with a variety of ingredients such as oil, butter, and salt
to improve the sensory quality of finished product, and variations in particle density have
been suggested to explain the majority of variation in bulk density measurements of other
cereal grains (Doehlert and McMullen, 2008). Thus, while determining expansion
volume based on original kernel weight may be reasonable for breeders or producers that
sell kernels to commercial venues like movie theatres and athletic venues where popped
91
popcorn is ultimately sold by volume (Hoseney et al., 1983), it may be less germane to
producers of packaged goods for retail markets where popcorn must be labeled and sold
by weight.
The method proposed by Pordesimo et al. (1990) divides pop volume by the
weight of popped flakes only, thereby eliminating the influence of unpopped kernels and
the weight of oil lost to the bag during popping in calculations. Using this method of
determination resulted in expansion values of 47.3 ± 6.4 cm3/g across all runs in this
study. In effect, this method of reporting expansion volume effectively expresses the
inverse of bulk density. The primary advantages of this method are that it is not biased
by loss factors due to oil retained by the bag or unpopped kernels, and it accounts for the
contribution of other ingredients to the bulk mass. Ultimately, it is important to
recognize that expansion volume is a complex trait that is also a function packing
efficiency and the measurement of pop volume itself, which is influenced by factors such
as the size and shape of individual flakes and void spaces between flakes.
Thousand kernel weight, sphericity, total fat, oleic acid, linoleic acid exhibited
significant correlations with both expressions of expansion volume (Table 3.5). The
negative correlation between thousand kernel weight and expansion volume agrees with
previous findings (Ceylan and Karababa, 2001). In addition, the negative correlation
between levels of oleic acid and pop expansion volume agrees with previous results
reported by Borras et al. (2006) for seven Argentinean popcorn hybrids. However, the
findings in this study disagree and are opposite to the findings by Borras et al. of a
significant positive correlation between expansion volume and levels of linoleic acid.
This divergence may be a result of the different hybrids or popping methods
92
(conventional oil popping vs. microwave popping) used in the two studies, or because
this study determined correlations to fatty acids as a percentage of the total kernel
composition instead of fatty acids as a percentage of total lipids.
Kernel sphericity was positively correlated with expansion volume, which agrees
with the findings from several previous studies (Pordesimo et al., 1990; Mohamed et al.,
1993; Ertas et al., 2009). The importance of structural geometry of popcorn kernels can
be explained by considering that dynamic fragmentation of granular materials can be
predicted from the stored elastic and kinetic energy (Griffith, 1921). Probability
modeling from fracture mechanics predicts the failure mode for pressurized vessels to
occur at the weakest point in the structure (Bargmann, 1986) and reveals that critical
pressures for fragmentation to be a function of surface area to volume ratio of the
particles (Pugno and Carpinteri, 2008). Moreover, previous work on popcorn kernels has
shown that increasing the difference between internal kernel pressure and the external
atmospheric pressure increases expansion volume (Quinn Sr. et al., 2005). Thus it may
be that more spherical popcorn kernels have greater structural stability due to a smaller
surface area to volume ratio and are thus able to build greater energy potentiation before
popping occurs.
Expansion volume is the only quality trait of popcorn that can be readily
measured, and evaluation of early generation popcorn hybrids is often limited by small
sample sizes which makes measuring pop volume precarious (Ziegler, 2001). While this
study was limited to three hybrids and growing locations in the USA, understanding
kernel physiochemical parameters that correlate with expansion volume may help lead to
the development and use of simple and non-destructive methods that can be leveraged
93
universally by popcorn breeders and producers for predicting expansion volume. For
example, total fat might be nondestructively estimated using near infrared spectroscopy,
and an automatic seed counter might be used to rapidly segregate kernels for thousand
weight measurement. In addition, the positive correlation between kernel sphericity and
expansion volume established previously and corroborated in this study suggests that
developing a simple and rapid method of determining kernel sphericity would be
beneficial.
3.5. Conclusions
Significant differences were observed in the physical and biochemical
characteristics of popcorn kernels used in this study. Significant positive correlations
were observed between Stenvert time-to-grind and protein, and a positive correlation
between time-to-grind and total carbohydrates and test weight. Protein was negatively
correlated with both sugars and total carbohydrates. Starch and fiber were also
negatively correlated.
Hybrid and environment effects on popcorn kernel physiochemical parameters are
varied. Among some key characteristics, the main effect of hybrid alone affected
sphericity, time-to-grind, dietary fiber, sugars, and starch. Environment main effect was
observed to influence total minerals. The main effects of both hybrid and environment
affected test weight, TADD Index, thousand-weight, protein, and total carbohydrates.
Determining microwave popcorn expansion volume as a function of pop volume
per popped flake weight may be a more appropriate method than on the basis of original
kernel weight, since it accounts for unpopped kernel loss factors and the contribution of
94
ingredient additions to the kernel which contributes to the bulk density. For both
methods of determination, expansion volume is positively correlated with kernel
sphericity and negatively correlated with thousand-kernel weight, total fat, and levels of
oleic and linoleic acid. The amount of oil retained by paper bags during microwave
popping is noteworthy and increases as a function of starting oil amount.
3.6. References
AOAC, 2010. Official Methods of Analysis. 18th ed. Association of Official Analytical
Chemists. Washington, DC.
Allard, R.W., Bradshaw, A.D., 1964. Implications of genotype-environmental
interactions in applied plant breeding. Crop Science 4, 503-508.
Bargmann, H.W., 1986. Prediction of pressure vessel failure: a critical review of the
probabilistic approach. Theoretical and Applied Fracture Mechanics 5, 1-16.
Borras, F., Seetharaman, K., Yao, N., Robutti, J.L., Percibaldi, N.M., Eyherabide, G. H.,
2006. Relationship between popcorn composition and expansion volume and
discrimination of corn types by using zein properties. Cereal Chemistry 83, 86-92.
Byrd, J.E., Perona, M. J., 1995. Kinetics of popping of popcorn. Cereal Chemistry 82,
53-59.
Ceylan, M., Karababa, E., 2001. Comparison of sensory properties of popcorn from
various types and sizes of kernel. Journal of the Science of Food and Agriculture
82, 127-133.
95
Doehlert, D.C., McMullen, M.S., 2008. Oat grain density measurement by sand
displacement and analysis of physical components of test weight. Cereal
Chemistry 85, 654-659.
Dofing, S.M., D‟Croz-Mason, N., Buck, J.S., 1991. Inheritance of expansion volume and
yield in two popcorn X dent corn crosses. Crop Science 31, 715-418.
Dofing, S.M., Thomas-Compton, M.A., Buck, J.S., 1990. Genotype x popping method
interaction for expansion volume in popcorn. Crop Science 30, 62-65.
Dorsey-Redding, C., Hurburgh, C.R., Johnson, L.A., Fox, S.R., 1990. Adjustment of
maize quality data for moisture content. Cereal Chemistry 67, 292-295.
Dorsey-Redding, C., Hurburgh, C.R., Johnson, L.A., Fox, S.R., 1991. Relationship
among maize quality factors. Cereal Chemistry 68, 602-605.
Ertas, N., Soylu, S., Bilgicli, N., 2009. Effects of kernel properties and popping methods
on popcorn quality of different corn cultivars. Journal of Food Process
Engineering 32, 478-496.
Greaves, J.E., and Hirst, C.T., 1929. The mineral content of grain. Journal of Nutrition
1, 293-298.
Griffith, A.A., 1921. The phenomena of rupture and flow in solids. Philosophical
Transaction of the Royal Society A 221, 163-198.
Gökmen, S., 2004. Effects of moisture content and popping method on popping
characteristics of popcorn. Journal of Food Engineering 65, 357-362.
Hoseney, R.C., Zeleznak, K., Abdelrahman, A., 1983. Mechanism of popcorn popping.
Journal of Cereal Science 1, 43-52.
96
Kirby, E.A., and Hughes, A.D., 1970. Some aspects of ammonium and nitrate nutrition
in plant metabolism. In: Kirby, A. (Ed.), Nitrogen Nutrition of the Plant.
University of Leeds, leeds, pp. 69-77.
Lee, K.M., Herrman, T.J., Rooney, L., Jackson, D.S., Lingenfelser, J., Rausch, K.D.,
McKinney, J., Iiams, C., Byrum, L., Hurburgh, C.R., Johnson, L.A., Fox, S.R.,
2007. Corroborative study on maize quality, dry-milling and wet-milling
properties of selected maize hybrids. Journal of Agricultural and Food Chemistry
55, 10751-10763.
Li, Y., Dong, Y., Niu, S., 2006. QTL analysis of popping fold and the consistency of
QTLs under two environments in popcorn. Acta Genetica Sinica 33, 724-732.
Lu, H.J., Bernardo, R., and Ohm, H.W., 2003. Mapping QTL for popping expansion in
popcorn with simple sequence repeat markers. Theoretical and Applied Genetics
106, 423-427.
Mohamed, A.A., Ashman, R.B., Kirleis, A.W., 1993. Pericarp thickness and other kernel
physical characteristics relate to microwave popping quality of popcorn. Journal
of Food Science 58, 342-346.
National Agricultural Statistics Service, 2007. Census of agriculture. USDA, Washington
D.C. Retrieved April 7, 2010 from:
http://www.agcensus.usda.gov/Publications/2007/Full_Report/Volume_1,_Chapte
r_2_US_State_Level/st99_2_026_026.pdf.
Park, D., Allen, K.G.D., Stermitz, F.R., Maga, J.A., 2000. Chemical composition and
physical characteristics of unpopped popcorn hybrids. Journal of Food
Composition and Analysis 13, 921-934.
97
Pomeranz, Y., Czuchajowska, Z., Martin, C.R., Lai, F.S., 1985. Determination of corn
hardness by the Stenvert hardness tester. Cereal Chemistry 62, 108-112.
Pomeranz, Y., Martin, C.R., Traylor, D.D., Lai, F.S., 1984. Corn hardness determination.
Cereal Chemistry 61, 147-150.
Pordesimo, L.O., Anantheswaran, R.C., Fleishmann, A.M., Lin, Y.E., Hanna, M.A.,
1990. Physical properties as indicators of popping characteristics of microwave
popcorn. Journal of Food Science 55, 1352-1355.
Pungno, N.M., Carpinteri, A., 2008. On linear elastic fragmentation mechanics under
hydrostatic compression. International Journal of Fracture 149, 113-117.
Quinn Sr., P.V., Hong, D.C., Both, J.A., 2005. Increasing the size of a piece of popcorn.
Physica A 353, 637-648.
Reichert, R.D., Tyler, R.T., York, A.E., Schwab, D.J., Tatarynovich, J.E., Mwasaru,
M.A., 1986. Description of a production model of the tangential abrasive
dehulling device and its application to breeder‟s samples. Cereal Chemistry 63,
201-207.
Schiffmann, R.F., 1987. Performance testing of products in microwave ovens.
Microwave World 8, 7-9, 14.
Shandera, D.L., Jackson, D.S., Johnson, B.E., 1997. Quality factors impacting processing
of maize dent hybrids. Maydica 42, 281-289.
Soylu, S., Tekkanat, A., 2007. Interactions amongst kernel properties and expansion
volume in various popcorn genotypes. Journal of Food Engineering 80, 336-341.
98
Sweley, J.C., Rose, D.J., Jackson, D.S., 2011. Composition and sensory evaluation of
popcorn flake polymorphisms for a select butterfly-type hybrid. Cereal Chemistry
88, 321-327.
Stuber, C.W., Lincoln, S.E., Wolff, D.W., Helentjaris, T., Lander, E.S., 1992.
Identification of genetic factors contributing to heterosis in a hybrid from two
elite maize inbred lines using molecular markers. Genetics 132, 823-839.
Tandjung, A.S., Janaswamy, S., Chandrasekaran, R., Abaoubacar, A., Hamaker, B.R.,
2005. Role of the pericarp cellulose matrix as a moisture barrier in microwaveable
popcorn. Biomacromolecules 6, 1654-1660.
U.S. Department of Agriculture, Agriculture Research Service, 2010. USDA National
Nutrient Database for Standard Reference, Release 23. Retrieved May 15, 2011
from: http://www.ars.usda.gov/ba/bhnrc/ndl.
USDA, 1990. General Information. In: Grain Grading Procedures. USDA Federal Grain
Inspection Service, Washington, DC, pp. 15-17.
Zhu, T., Jackson, D.S., Wehling, R.L., Geera, B., 2008. Comparison of amylose
determination methods and the development of a dual wavelength iodine binding
technique. Cereal Chemistry 85, 51-58.
Ziegler, K.E., 2001. Popcorn. In: Hallauer, A. (Ed.), Specialty Corn. CRC Press, Boca
Raton, FL, pp. 199-234.
Figure 3.1. Oil amount lost to bag during microwave popping
99
Table 3.1. Physical Properties of Three Popcorn Hybrids Grown in Three Different Environments a
Hybrid
Environment
YPK213
YPK313
YPK321
Indiana
Nebraska
Thousand Weight (g) b
159.5 ± 9.3 c
172.6 ± 7.8 a
166.0 ± 8.4 b
175.6 ± 5.7 a
156.3 ± 6.9 c
166.3 ± 5.1 b
Test Weight (kg/hl) c
87.0 ± 1.9 b
89.2 ± 0.8 a
86.7 ± 2.6 b
85.8 ± 2.6 b
88.9 ± 0.8 a
88.3 ± 1.2 b
1.38 ± 0.01 a
1.38 ± 0.01 a
1.38 ± 0.01 a
1.38 ± 0.01 a
1.38 ± 0.01 a
1.38 ± 0.01 a
14.6 ± 0.6 a
13.8 ± 0.7 b
14.3 ± 0.7 ab
14.4 ± 0.7 a
13.9 ± 0.8 b
14.3 ± 0.6 ab
Pycnometer Density
(g/cm3) b
Stenvert Time to Grind (sec) b
Stenvert Height (cm)
b
Ohio
7.0 ± 0.2 a
7.0 ± 0.1 a
7.0 ± 0.1 a
7.0 ± 0.1 a
7.0 ± 0.1 a
7.0 ± 0.1 a
TADD index d
40.9 ± 2.5 c
45.8 ± 2.7 a
43.8 ± 2.8 b
40.5 ± 2.1 c
46.5 ± 2.3 a
43.5 ± 2.3 b
Sphericity (%)e
74.6 ± 0.0 a
71.9 ± 0.0 b
72.4 ± 0.1 b
72.9 ± 0.1 a
73.2 ± 0.1 a
72.8 ± 0.1 a
a
Mean values ± SD in same row followed by the same letter within each factor (hybrid or environment) are not significantly
different (P<0.05). n=3.
100
Table 3.2. Biochemical Composition of Three Popcorn Hybrids Grown in Three Different Environments a
Hybrid
YPK213
YPK313
Environment
YPK321
Indiana
Nebraska
Ohio
Total Fat (%)
2.8 ± 0.2 a
2.9 ± 0.2 a
3.0 ± 0.3 a
3.0 ± 0.2 a
2.9 ± 0.2 a
3.0 ± 0.2 a
Total Carbohydrate (%)
72.3 ± 0.4 c
73.9 ± 0.8 a
73.1 ± 0.7 b
72.7 ± 0.6 b
73.7 ± 1.3 a
72.9 ± 0.6 b
Starch (%)
61.5 ± 2.7 b
63.4 ± 3.8 ab
65.3 ± 1.8 a
63.5 ± 3.0 a
63.6 ± 3.9 a
63.1 ± 2.9 a
Amylose (%)
9.1 ± 0.5 b
9.6 ± 0.8 ab
9.8 ± 1.0 a
9.5 ± 0.5 ab
10.1 ± 0.8 a
9.0 ± 0.8 b
Amylopectin (%)
52.3 ± 2.7 b
53.8 ± 3.5 ab
55.5 ± 2.2 a
55.0 ± 2.7 a
53.5 ± 3.6 a
54.1 ± 3.1 a
Dietary Fiber (%)
11.1 ± 0.6 a
11.4 ± 0.3 a
10.1 ± 0.3 b
11.0 ± 0.6 a
11.0 ± 0.8 a
10.7 ± 0.8 a
Sugars (%)
0.46 ± 0.02 c
0.65 ± 0.08 a
0.51 ± 0.04 b
0.56 ± 0.11 a
0.56 ± 0.11 a
0.50 ± 0.07 b
Protein (%)
9.9 ± 0.2 a
8.1 ± 0.8 c
8.8 ± 0.6 b
9.2 ± 0.6 a
8.4 ± 1.3 b
9.2 ± 0.5 a
Total Minerals (%)
1.0 ± 0.1 a
1.0 ± 0.1 a
1.1 ± 0.1 a
1.1 ± 0.0 a
1.1 ± 0.1 a
0.9 ± 0.1 b
Palmitic Acid (%)
16.4 ± 0.4 a
15.0 ± 0.2 c
15.8 ± 0.4 b
15.8 ± 0.8 a
15.9 ± 0.7 a
15.7 ± 0.5 a
Stearic Acid (%)
1.9 ± 0.1 b
1.7 ± 0.1 c
2.1 ± 0.1 a
1.9 ± 0.2 a
1.8 ± 0.1 b
2.0 ± 0.2 a
Oleic Acid (%)
20.8 ± 0.8 b
22.0 ± 0.7 a
21.7 ± 0.7 a
21.3 ± 0.7 b
21.1 ± 0.7 b
22.2 ± 0.8 a
Linoleic Acid (%)
58.3 ± 0.3 a
58.5 ± 0.8 a
57.7 ± 0.5 b
58.4 ± 0.7 a
58.5 ± 0.5 a
57.6 ± 0.4 b
Linolenic Acid (%)
1.4 ± 0.1 a
1.3 ± 0.0 b
1.3 ± 0.1 ab
1.3 ± 0.1 a
1.4 ± 0.1 a
1.3 ± 0.1 a
All Other (%)
1.2 ± 0.4 a
1.5 ± 0.3 a
1.4 ± 0.4 a
1.3 ± 0.4 a
1.4 ± 0.3 a
1.3 ± 0.3 a
Fatty Acid Compositionb
a
Mean values ± SD in same row followed by the same letter within each factor (hybrid or environment) are not
significantly different (P<0.05). All values reported on 14.0% moisture content (wb). n=3.
b
Fatty acids reported as percentage of total fat in kernel. n=3.
101
Table 3.3. Correlation coefficientsa between popcorn kernel physical and biochemical properties b
Thousand
Kernel
Weight
Sphericity
-0.040
0.467
-0.514
0.326
0.822**
0.099
-0.532
-0.371
0.436
0.362
0.238
-0.649
0.309
-0.554
-0.023
0.468
-0.102
-0.205
-0.037
-0.319
-0.213
0.476
0.232
0.290
-0.629
0.114
0.299
0.373
-0.199
-0.332
0.057
0.068
0.131
0.347
0.382
0.116
-0.792*
0.291
0.574
0.513
-0.704*
-0.302
-0.432
0.088
0.891*
-0.385
-0.802*
-0.160
0.586
Total Minerals
-0.231
0.051
-0.208
0.190
0.0741
0.066
-0.055
Linolenic
-0.621
-0.205
0.567
0.086
-0.423
-0.280
0.377
0.440
-0.747*
-0.550
0.73 *
-0.289
-0.367
-0.233
-0.793*
0.218
-0.012
-0.35
0.034
-0.901**
0.006
0.589
-0.112
0.476
-0.086
-0.291
-0.516
Total Fat
Total
Carbohdydrate
Starch
Amylose
Amylopectin
Dietary Fiber
Sugars
Protein
Test Weight
Pycnometer
Density
Total
Carbohdrates
Starch
Amylose
Amylopectin
Dietary
Fiber
Sugars
Protein
Total
Minerals
Test
Weight
Pycnometer
Density
Stenvert
Time
Stenvert
Height
-0.041
0.481
-0.034
0.528
-0.450
0.167
-0.086
0.036
-0.454
-0.023
0.088
0.405
0.512
0.544
0.368
0.009
0.767*
-0.988**
0.211
0.523
-0.092
-0.906**
0.373
0.950**
-0.717*
0.337
-0.583
0.283
0.015
-0.200
0.065
-0.048
0.344
-0.589
0.69*
0.159
-0.755*
0.246
-0.430
0.071
0.368
0.036
-0.800**
Stenvert Time
Stenvert
Height
TADD index
Thousand
Kernel Weight
a
b
TADD
index
-0.529
Correlation coefficients greater than |r| = 0.45 are bold-printed. * Significance at p < 0.05; **significance at P < 0.01.
Analysis used compositional data on the basis of 14.0% moisture content (wb) in kernels. n=9.
102
Table 3.4. Pop Performance Measures for Three Popcorn Hybrid Grown in Three Different Environments a
Expansion Volume
Factor
Level
Pop Volume (cm3) per
popped flake weight (g)
Hybrid
YPK-213
49.4 ± 6.5 a
46.5 ± 4.6 a
12.5 ± 6.2 a
16.5 ± 3.2 a
9.6 ± 2.3 b
YPK-313
46.7 ± 5.8 a
44.1 ± 3.5 b
11.6 ± 4.7 a
15.9 ± 2.2 a
14.0 ± 3.8 a
YPK-321
45.9 ± 7.2 a
43.8 ± 3.4 b
10.1 ± 5.3 a
15.0 ± 2.1 a
10.7 ± 2.0 ab
Indiana
51.8 ± 7.8 a
47.6 ± 5.3 a
13.2 ± 4.7 a
17.5 ± 2.6 a
9.3 ± 1.5 b
Nebraska
45.6 ± 7.1 b
43.5 ± 3.9 b
10.9 ± 6.2 a
15.1 ± 2.1 b
11.1 ± 3.6 ab
Environment
Pop Volume (cm3) per
original kernel weight (g)
Unpopped
Kernels (%)
Oil Lost to Bag (%)
%Unpopped kernels
after repopping b
Ohio
44.5 ± 6.5 b
43.2 ± 3.4 b
10.1 ± 5.6 a
14.7 ± 3.5 b
14.0 ± 2.6 a
Average across all runs
47.3 ± 6.4
44.7 ± 3.7
11.4 ± 5.3
15.8 ± 2.3
11.5 ± 2.6
a
Mean values ± SD followed by the same letter within the same column for each design factor (hybrid or environment) are not significantly different (P < 0.05).
n=3.
b
Unpopped kernels from initial testing were collected, then 28 g samples were re-popped using a 1250 W microwave oven. n=3.
103
Table 3.5. Correlationsa Between Popcorn Kernel Physiochemical Parameters and Unpopped Kernels and Expansion Volume b
Unpopped
Kernels
Physical Parameters
Thousand Weight
Test Weight
Pyncometer Density
Stenvert Time to Grind
Stenvert Height
TADD index
Sphericity
Composition factor
Total Fat
Total Carbohydrate
Starch
Amylose
Amylopectin
Dietary Fiber
Sugars
Protein
Total Minerals
Palmitic Acid
Stearic Acid
Oleic Acid
Linoleic Acid
Linolenic Acid
Expansion Volume
Pop Volume (cm3) per
Pop Volume (cm3) per
popped flake weight (g) original kernel weight (g)
-0.565
0.226
-0.348
0.297
-0.242
-0.138
0.469
-0.729 *
0.107
-0.187
0.310
-0.381
-0.224
0.788 *
-0.708 *
0.077
-0.187
0.217
-0.314
-0.167
0.835 *
-0.573
-0.273
-0.168
-0.519
-0.006
-0.209
-0.584
0.381
-0.491
-0.115
-0.156
-0.479
-0.645
0.083
-0.797 *
-0.348
-0.427
-0.131
-0.415
0.080
-0.619
0.440
-0.040
-0.053
-0.381
-0.890 **
-0.801 **
0.182
-0.824 **
-0.253
-0.525
-0.007
-0.562
0.269
-0.477
0.337
0.118
-0.105
-0.521
-0.966 **
-0.791 *
0.087
a
Correlation coefficients greater than |r|=0.45 are bold-printed. * Significance at p<0.05; **significance at P<0.01.
Analysis used compositional data on the basis of 14.0% moisture content (wb) in kernels.n=9.
b
104
105
CHAPTER 4. EFFECTS OF HYBRID, ENVIRONMENT, OIL ADDITION, AND
MICROWAVE WATTAGE ON POPPED POPCORN MORPOLOGY
4.1. Abstract
The objective of this study was to understand what factors influence the formation
of different shapes of popped popcorn through the development of statistical models.
Microwave popcorn popping was conducted across a range of microwave wattages (7501240 W) and oil additions (0-30%) using a set of three popcorn hybrids grown in three
environments. After popping, expansion volume was measured and the relative
proportion of different popped shapes was enumerated by visual characterization of
popped flakes, namely: unilateral, bilateral, or multilaterally-expanded. The percentage
of flake morphologies varied from 1-24% unilateral, 20-55% bilateral, and 31-68%
multilateral across all runs. The relative percentage of each shape was influenced by
hybrid, growing location, corn:oil ratio, and microwave wattage. The proportion of
unilateral flakes was positively correlated to oleic acid in the kernel and negatively
correlated to kernel sphericity, while bilateral flakes were positively correlated to dietary
fiber in the kernel. Expansion volume was positively correlated to occurrence of
bilaterally-expanded flakes and negatively correlated to unilateral shape. These data may
support the development of new hybrids or varieties of popcorn that produce the most
desirable amounts of popped shapes in order to optimize consumer liking or create
differentiated products and market new usages for popcorn.
106
4.2. Introduction
Visual appearance plays an important role in the sensory perceptions and
consumer acceptability of food (Imram, 1999). Product shape is an element of food
appearance that is often varied to reach broader market segments and infer different
functional and emotive benefits to consumers (Berkowitz, 1987). For example, pasta is
sold in an assortment of different shapes such as penne, rotini, farfalle, and so on, and
individual consumer preferences for pasta shape have been shown to significantly affect
consumption and satiety (Rolls et al., 1982). Many other food categories have augmented
product choices by offering different shapes, including pretzels, tortilla chips, and frozen
potatoes, among others.
The shape of popped popcorn (i.e., flakes) has been long held as an important
consideration. Popcorn hybrids are commonly classified by the shape made upon
popping as either mushroom or butterfly-type (Ziegler et al., 1984; D‟Croz-Mason and
Waldred, 1990). Mushroom hybrids produce more durable flakes and thus are typical in
mixing and coating applications (Eldredge and Thomas, 1959), whereas butterfly-type
hybrids are tender and delicate and commonly found at movie theatres and in retail
microwave popcorn (Ziegler, 2001). Even within a single hybrid of butterfly-type
hybrid, it has been observed that distinct polymorphisms can be identified and
characterized as having differences in compositional properties and sensory attributes
depending on whether the appendages are expanded unilaterally, bilaterally, or
multilaterally (Sweley et al., 2011).
This raises the question of what factors influence the formation of different shapes
of popcorn flakes within butterfly-type hybrids. While some factors affecting the
107
morphology of extruded and microwaved starch pellets have been reported (Lee et al,
2000), no study has investigated factors influencing shape formation for microwave
popcorn. Understanding these relationships may lead to a means of evaluating potential
polymorphic quantities that form in microwave popcorns, and it may lead to improved
selection of hybrids that produce popcorn with a greater proportion of more desirable
flake polymorphs. Accordingly, the present study was undertaken to understand the
influence of key intrinsic factors and extrinsic variables on the formation of flake
polymorphisms in microwave popcorn.
4.3. Experimental
4.3.1. Popcorn Samples
Three butterfly-type popcorn hybrids (YPK213, YPK313, YPK321) grown in
strip plot trials in Nebraska, Indiana, and Ohio in 2009 were obtained from ConAgra
Foods, Inc. (Omaha, NE USA). The physical and chemical characteristics of popcorn
samples were measured and evaluated after conditioning kernels to 14% moisture (wet
basis) by storage at 21.5 ˚C and 73% rh. Physical testing included test weight, thousandweight, true kernel density measured by pycnometer, Stenvert height and time-to-grind,
tangential abrasive dehulling device (TADD) index, and sphericity. Compositional
characterization included total fat, total carbohydrates, starch, amylose, amylopectin,
dietary fiber, sugars, protein, total minerals, and fatty acid composition. The results of
the physiochemical testing for this set of popcorn hybrids and growing environment have
been reported (Sweley et al., 2012).
108
4.3.2. Test Design and Microwave Popping
In order to assess the effect of both intrinsic kernel properties as well as extrinsic
factors on popcorn flake shape formation, an experimental design was developed to
consider four design factors: hybrid, environment, microwave wattage, and oil addition.
A D-optimal design was developed to predict responses using a cubic model with 65
runs, including 6 for lack of fit and 5 for pure error. Hybrid, environment, and
microwave wattage factors each had three levels in the experimental design, while oil
addition was modeled as a continuous variable and included twelve levels (Table 4.1).
A detailed description of the popping methods and experimental design has been
described previously (Sweley et al., 2012). In brief, hybrid and environment factors were
represented by the popcorn samples formerly described. Popcorn kernels and oil were
weighed into pre-folded microwave popcorn bags. The type of oil used was a blend of
palm and palm stearine (Archer Daniels Midland Company product 845600, Decatur, IL).
Design limits on oil addition were established to encompass the prevalent range of oil
levels (6% to 30%) found in US microwave popcorn products (USDA, 2010). To
maintain consistency and ensure bag constriction did not affect popping performance, the
combined weight of popcorn and oil was 77 g for all experiments. Pop testing was
completed using three microwave ovens: Panasonic NN-H765BF, General Electric
JE1160WC, and Samsung MW830WA. The actual power output for the microwaves was
1240 W, 1050 W, and 750 W, respectively, for the three microwaves using the method of
Schiffmann (1987). Statistical design and modeling used the actual microwave power
output instead of the manufacturer listed output.
109
For each experimental run, microwave popping was performed until the time
between audible pops was 2 s, and then bag contents were poured into a steel sieve with
round-holes (7.94 mm openings; Seedburo, USA Model 0070) to remove unpopped
kernels. The popped popcorn was weighed and measured volumetrically to the nearest
25 mL using a 4 L graduated cylinder. Next, individual popcorn flakes were visually
inspected and segregated by shape depending on whether the flake appendages were
expanded unilaterally, bilaterally, or multilaterally (Fig. 4.1), and the number of each
shape type was counted and weighed. The counts of each popcorn flake shape were
converted to a percentage of total popped kernels. This was necessary since the corn-tooil design factor varied the amount of corn in the bags to maintain a constant total
ingredient weight for all runs.
4.3.3. Statistical Analysis
All statistical analyses were performed using SAS software (version 9.2, SAS
Institute, Cary, NC). The relationship between popped shape and the kernel
physiochemical attributes was analyzed using the SAS correlation (CORR) procedure.
CORR procedure was also used to establish the relationship between the different flake
polymorphisms, as well as the relationship between popped shape and expansion volume.
Analysis of variance (ANOVA) was performed on the individual dependent
observations of percent unilateral flakes, percent bilateral flakes, and percent multilateral
flakes using SAS general linear models (GLM) procedure. This analysis was used to
predict the relationship between each of the flake polymorphs and the independent
variables. Hybrid, environment, and microwave wattage were regarded as categorical
variables, while corn-oil ratio was treated as a continuous variable. Model selection for
110
each popcorn shape was achieved by backwards elimination from all main effects,
quadratic and two-way cross product interactions. Only effects that showed significance
at p < 0.05 were included in the fitted hierarchical model for each dependent variable.
Thus, the ANOVA for flake shape is distinctive depending on the factors which showed
statistical significance for each shape‟s polymorphic formation.
Though these ANOVAs described the variation in proportion of each flake, this
approach was univariate in a multivariate scenario and so changes in the relative amount
of any one particular flake polymorph inevitably affected the relative amount of other
shapes. Therefore, in addition to fitting ANOVAs for each polymorph individually, a
two-stage least squares methodology was used to simultaneously fit regression models to
multiple shapes of popcorn flakes using the SAS procedure for estimating parameters in
interdependent systems of linear regression equations (SYSLIN). In order to accomplish
this regression, the categorical design factors of hybrid and environment were
reparameterized as shown in Eq. (1-4).
xh1 = 1 if the observation is from Hybrid VY213, else 0
(1)
xh2 = 1 if the observation is from Hybrid VY313, else 0
(2)
xe1 = 1 if the observation is from Indiana, else 0
(3)
xe2 = 1 if the observation is from Ohio, else 0
(4)
As before, model selection was achieved by backwards elimination from all main
effects, quadratic and two-way cross product interactions. Because effects were fitted to
the model for different flake polymorphs concurrently, effects which showed significance
at p < 0.05 for any of the dependent variables were included in the final model.
111
4.4. Results and Discussion
4.4.1. Relationship of Kernel Properties to Popped Morphology
Correlation analysis between kernel physiochemical properties and popped flake
morphology (Table 4.2) revealed a significant positive correlation between dietary fiber
and bilateral shape (p<0.05, r=0.682) and positive correlation between oleic acid in the
kernel and unilateral shape (p<0.01, r=0.867). Sphericity had a significant inverse
correlation to unilateral shape (p<0.01, r=-0.822), but did not have a significant influence
on occurrence of other flake shapes. Previous research has indicated that the morphology
of microwave-expanded starch extrudates is highly dependent on both the pellet
appearance before puffing and the rate of pressure and water vapor release following
outer cell rupture (Lee et al., 2000). Similarly, the model describing popping mechanics
for popcorn has been described as the kernel acting as a pressure vessel in which the
pericarp is so rigid that during heating, moisture inside the kernel converts to superheated
steam that builds temperature and pressure until it can overcome the combined force of
the pericarp and atmosphere (Hoseney et al., 1983). In addition, engineering fracture
mechanics suggest that pressure vessels fail at the weakest point in the structure
(Bargmann, 1986) and critical pressures for particle fragmentation depend on surface area
to volume ratio (Pugno and Carpinteri, 2008). Thus less spherical kernels may have a
predominant failure point in the pericarp through which molten starch of popcorn
endosperm tends to expand and create flake morphology with appendages in only one
direction, whereas more spherical kernels tend to have an increased number of
fragmentation points, allowing the endosperm to expand and form appendages in multiple
directions.
112
4.4.2. Predictive Models for Popped Morphology
4.4.2.1. Unilateral Shape
The regression model for unilaterally-expanded popcorn flakes was:
Y1 = b0 + b1xe + b2xh + b3xw + b4xw2
(5)
where Y1 is percentage of unilaterally-expanded flake polymorphs, xe is environment
effect, xh is hybrid effect, xw is microwave wattage effect, and bi are the coefficients for
all main effects and interaction terms (Table 4.3). The analysis of variance (Table 4.4)
showed the highest order effects that were significant were hybrid*environment*
microwave wattage (p<0.01) and corn:oil*environment*microwave (p<0.01). Because
model selection was fitted using backwards selection, all lower order effects that
contributed to the three-way interactions were included in the final hierarchical model.
The lack of fit test showed the model was adequate for predicting percentage unilaterallyexpanded flakes within the range of variables studied (p<0.01), and the coefficient of
determination (R2 = 0.86) showed a good fit of the model to the data.
While the results of this modeling reveal a somewhat complex relationship
between design factors, they suggest that the relative percentage of unilateral polymorphs
can be altered by selecting hybrids and environments for the desired microwave popcorn.
For example, the ratio of unilateral flakes can be maximized by opting for hybrids
VYP313 or VYP321 instead of VYP213, selecting popcorn grown in the Ohio strip plots,
and popping in a lower wattage microwave. Increasing the proportion of unilateral
popcorn flakes may be of considerable interest, as previous sensory testing revealed the
unilateral shape to have improved texture, flavor, and overall liking relative to other
popcorn flake shapes (Sweley et al, 2011).
113
4.4.4.2. Bilateral and Multilateral Shapes
Regression analysis performed individually on the responses for bilateral and
multilaterally-expanded popcorn flakes resulted in selection and fitting to models with
the same generalized equation (Eq. 6).
Y2,3 = b0 + b1xco + b2xh + bexco2
(6)
where Y2 and Y3 are the percentage of bilateral or multilateral polymorphs, respectively, xh
are hybrid effects, xco is the effect of corn: oil ratio, and , and bi are the coefficients.
While the general equation defining the model bilateral and multilateral shapes is the
same, the coefficients were distinct for each shape (Table 4.5). The sum of squares for
the main effect „hybrid‟ and the quadratic effect „corn:oil ratio‟ were significant (p<0.05)
and impacted the formation of both bilateral and multilateral flakes (Table 4.6).
However, while parameter estimates were significant and the lack of fit test indicated
both models were sufficiently accurate to predict performance (p< 0.01), the coefficients
of determination for bilateral and multilateral polymorphs (R2 = 0.36, 0.33, respectively)
showed only moderate fit of each models to the actual data. This suggests that the
variation in the percentage of bilateral and multilaterally-expanded flakes may be
attributed to factors and/or ranges of factors beyond those in this study. For example,
variation in kernel position within the microwave package might affect heat transfer for
individual kernels and give rise to different shape formation. Furthermore, pop
mechanics that result in bilateral or multilateral polymorphs may be related and thus
make it difficult to model individual responses. In fact, while the models for bilateral and
multilateral have the same general equations, the parameter estimates are very nearly
opposite, and correlation analysis suggests a significant negative correlation between
114
percent bilateral and multilateral flakes across all factors considered in this study (r=0.813, p<0.01). As a result of this relationship, the sum of percent bilateral and
multilateral polymorphs was also modeled. Regression analysis resulted in a model with
equivalent fit and ANOVA values obtained for the unilateral model (Eq. 5), which is
expected since the sum of the bilateral and multilateral flakes is intrinsically defined as a
mixture where [%bilateral] + [%multilateral] = 1 – [%unilateral].
4.4.4.3. Universal model for popcorn shape
Two-stage least-squares regression led to the selection of a universal model for
popcorn flake polymorphism defined by Eq. (7).
Yp = b0 + b1xco + b2xw + b3xh1 + b4xh2 + b5xe1 + b6xe2 + b7xco2
(7)
where Yp is percentage of each polymorphic shape, bi are the coefficients, xco is the corn:
oil ratio effect, xw is microwave wattage effect, xh is hybrid effect, and xe is environment
effect. The coefficients for each shape in the universal model are summarized in Table
4.7. Note that the coefficients showing significance (p<0.05) for either unilateral or
bilateral shape were included in the final model. Regression analysis for the predicted
percentage of unilateral and bilateral polymorphs showed a reasonably good fit to the
data (R2 = 0.54, 0.41, respectively) considering that multiple responses were fitted
simultaneously. Percent multilateral polymorph was not regressed since this response is
intrinsically defined by the other two polymorphs. This model shows that the proportion
of flakes can be optimized for the desired proportion of shapes by selections in popcorn
hybrid, environment, microwave wattage, or the amount of oil added (Fig. 4.2). Notably,
the same input design factors for maximizing the proportion of unilateral flakes (Eq. 5)
115
are included in this universal model (i.e., hybrids VYP313 or VYP321, Ohio strip plot,
low microwave wattage).
While no previous study has investigated factors affecting morphology of popped
flakes, several studies have investigated factors influencing other end quality measures
for popcorn, particularly expansion volume and unpopped kernels. These studies reveal
that microwave popping of popcorn is a complicated system that is dependent on a
variety of both intrinsic factors including hybrid and extrinsic variables including
microwave wattage and oil added (Dofing et al., 1990; Song et al., 1991; Singh and
Singh, 1999; Ceylan and Karababa, 2004). This suggests that formation of popcorn flake
polymorphisms is influenced by many of the same critical factors which have been
established for affecting popcorn expansion volume and unpopped kernels.
4.4.3. Relationship of shape to expansion volume
Since expansion volume is the most important quality measure for popcorn
(Borras et al., 2006), it is relevant to understand the role that varying proportions of flake
polymorphisms have on expansion volume. As shown in Fig. 4.3, unilaterally-expanded
polymorphs were negatively correlated to expansion volume (r=-0.565, p<0.01), while
bilateral polymorphisms were positively correlated (r=0.323, p<0.01) and multilaterallyexpanded popcorn flakes did not have a significant effect (r=0.013, p=0.92).
Previous studies have demonstrated that the combination and interactions of
individual particle shapes in mixtures has an effect on the packing characteristics and
bulk density of grains (Boac et al., 2010; Doehlert et al., 2008). While the statistical
contribution of unilateral and bilateral shape in this study was moderate, the analysis was
constrained by the relative quantities of each shape produced naturally during popping;
116
the preponderance of unilateral flakes varied from 1.13 to 23.5%, percent bilateral ranged
from 19.6 to 54.7%, and multilateral ranged from 30.5 to 68.0% across all runs in this
study. Mixture studies that expand the design space outside these limits may introduce
relationships not discovered in this study.
Research efforts by plant breeders and growers have selected popcorn hybrids
that pop nearly twice as large as those from just 50 years ago (Foer, 2005). While the
proportions of different shapes that would have been found in older, low expansion
volume hybrids is speculative, it may be that a corollary effect of breeding efforts
focused on increasing expansion volume has been to decrease the relative fraction of
unilateral shapes. This might explain why unilateral shape tended to be the least
occurring shape polymorph. This presents a paradoxical quandary, since increasing
expansion volume is a desirable quality for popcorn, yet the unilaterally-shaped flakes
have been shown to have preferred texture and flavor liking (Sweley et al., 2011). Thus,
further research to understand how the proportion of popped flake shapes could be
optimized simultaneously for the highest consumer liking and best expansion volume
would be warranted.
4.5. Conclusions
Popcorn breeders and producers and microwave popcorn manufacturers
continually seek to improve quality attributes that will increase consumer satisfaction or
create market differentiation. The morphology of popped flakes may be a quality
attribute for popcorn that could provide a market advantage. Research efforts to
understand factors affecting flake shape formation can provide insights that can be used
117
by popcorn growers, manufacturers, and snack industry professionals to further improve
the acceptability of current products or enable new product development and usage
occasions.
The results of this study reveal that the shape formation of popped popcorn is
complex and dependent on both intrinsic and external factors, as well as interaction
effects. Varying the relative proportion of different flake polymorphs in microwave
popping is notionally possible by selecting the optimum hybrids, growing location and
environment, corn-oil ratio, and microwave wattage. Among physiochemical
characteristics, kernel sphericity was negatively correlated to unilateral shape, while oleic
acid was positively correlated to unilateral shape and dietary fiber was correlated to
bilateral shape. Unilaterally-expanded popcorn flakes occurred least frequently and were
negatively correlated to expansion volume.
4.5. References
Bargmann, H.W., 1986. Prediction of pressure vessel failure: a critical review of the
probabilistic approach. Theoretical and Applied Fracture Mechanics 5, 1-16.
Berkowitz, M., 1987. Product shape as a design innovation strategy. Journal of Product
Innovation Management 4, 274-283.
Boac, J.M., Casada, M.E., Maghirang, R.G., Harrier III, J.E., 2010. Material and
interaction properties of selected grains and oilseeds for modeling discrete
particles. Transactions of the ASABE 53, 1201-1216.
Borras, F., Seetharaman, K., Yao, N., Robutti, J. L., Percibaldi, N. M., Eyherabide, G. H.,
2006. Relationship between popcorn composition and expansion volume and
118
discrimination of corn types by using zein properties. Cereal Chemistry 83, 8692.
Ceylan, M., Karababa, E., 2004. The effects of ingredients on popcorn popping
characteristics. International Journal of Food Engineering 39, 361-370.
D‟Croz-Mason, N., Waldren, R., 1990. Popcorn Production. NebGuide G78-426.
University of Nebraska, Cooperative Extension, Lincoln, NE.
Doehlert, D.C., McMullen, M.S., 2008. Oat grain density measurement by sand
displacement and analysis of physical components of test weight. Cereal
Chemistry 85, 654-659.
Dofing, S.M., Thomas-Compton, M.A., Buck, J.S., 1990. Genotype x popping method
interaction for expansion volume in popcorn. Crop Science 30, 62-65.
Eldredge, J.C., Thomas, W.I., 1959. Popcorn – its production, processing, and utilization.
Iowa Agric. Exp. Stn. Bull P127, 1-16.
Foer, J., 2005. Arming nature's grenade. Discover Magazine 26, 24-25.
Hoseney, R.C., Zeleznak, K., Abdelrahman, A., 1983. Mechanism of popcorn popping.
Journal of Cereal Science 1, 43-52.
Imran, N., 1999. The role of visual cues in consumer perception and acceptance of a food
product. Nutrition & Food Science 99, 224-230.
Lee, E.Y., Lim, K.I., Lim., J.-k., Lim, S.-T., 2000. Effects of gelatinization and moisture
content of extruded starch pellets on morphology and physical properties of
microwave-expanded products. Cereal Chemistry 77, 769-773.
Pungno, N.M., Carpinteri, A., 2008. On linear elastic fragmentation mechanics under
hydrostatic compression. International Journal of Fracture 149, 113-117.
119
Rolls, B.J., Rowe, E.A., Rolls, E.T., 1982. How sensory properties of foods affect human
feeding behavior. Physiology & Behavior 29, 409-417.
Schiffmann, R.F., 1987. Performance testing of products in microwave ovens.
Microwave World 8, 7-9, 14.
Singh, J., Singh, N., 1999. Effects of different ingredients and microwave power on
popping characteristics of popcorn. Journal of Food Engineering 42, 161-165.
Song, A., Eckhoff, S.R., Paulsen, M., Litchfield, J.B., 1991. Effects of kernel size and
genotype on popcorn popping volumes and number of unpopped kernels. Cereal
Chemistry 68, 464-467.
Sweley, J.C., Rose, D.J., Jackson, D.S., 2011. Composition and sensory evaluation of
popcorn flake polymorphisms for a select butterfly-type hybrid. Cereal
Chemistry 88, 321-327.
Sweley, J.C., Rose, D.J., Jackson, D.S., 2012. Hybrid and environment effects on
popcorn kernel physiochemical properties and their relationship to microwave
popping performance. Journal of Cereal Science 55, 188-194.
U.S. Department of Agriculture, Agriculture Research Service, 2010. USDA National
Nutrient Database for Standard Reference, Release 23.
Ziegler, K.E., Ashman, R.B., White, G.M., Wyson, D.S., 1984. Popcorn Production and
Marketing. Cooperative Extension Service, Purdue University, West Lafayette,
Indiana. National Corn Handbook.
Ziegler, K.E., 2001. Popcorn. In: Hallauer, A. (Ed.), Specialty Corn. CRC Press, Boca
Raton, FL, pp. 199-234.
Figure 4.1. Categorization of popped popcorn shapes from butterfly-type hybrids. From left to right: multilaterally-expanded;
bilaterally-expanded; unilaterally-expanded. (scale: each square measures 7 cm x 7 cm)
120
121
Figure 4.2. Predictive distribution of popcorn shapes %unilateral, %bilateral, and %multilateral polymorphs
in popcorn as a function of hybrid, environment, microwave wattage, and corn:oil ratio.
Figure 4.3. Correlation between expansion volume and %uniltateral (r=-0.565, p<.001), %bilateral (r=0.323, p=0.008), and
%multilateral flakes (r=0.013, p=0.92). Dotted lines represent 95% confidence interval around the mean.
122
Table 4.1. Table of Experimental Design a
Hybrid
VYP213
Environment
IN
NE
OH
VYP313
IN
NE
OH
VYP321
IN
NE
OH
a
Microwave
Wattage
900
1100
1300
900
1100
1300
900
1100
1300
900
1100
1300
900
1100
1300
900
1100
1300
900
1100
1300
900
1100
1300
900
1100
1300
6.0
2
11.5
12.4
12.6
12.6
12.7
Corn-to-Oil (%)
12.8 18.0 22.9
23.0
23.2
23.3
24.0
24.4
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
1
1
1
1
1
1
1
1
2
1
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
30.0
2
1
1
1
1
Numbers (1 or 2) in table indicate runs in experimental design with that set of conditions. Total number of runs = 65.
123
Table 4.2. Correlation Coefficients between Popcorn Kernel Physiochemical Properties and Popped Flake Shape a
%Unilateral
%Bilateral
Physical Parameters
Thousand Weight
0.356
0.316
Test Weight
0.329
-0.121
Pyncometer Density
0.179
-0.295
Stenvert Time to Grind
0.027
-0.497
Stenvert Height
0.153
0.292
TADD Index
-0.271
0.538
Sphericity
-0.822b
0.366
Composition factor
Total Fat
-0.341
0.572
Total Carbohydrate
0.039
0.519
Starch
0.560
-0.589
Dietary Fiber
-0.348
0.682 c
Protein
0.004
-0.542
Total Minerals
-0.362
-0.004
Palmitic Acid
-0.162
-0.379
Stearic Acid
0.348
-0.592
Oleic Acid
0.867 b
-0.358
Linoleic Acid
-0.216
0.496
Linolenic Acid
-0.184
-0.486
a
Correlation coefficients greater than |r| = 0.45 are bold-printed.
b
significance at P < 0.01.
c
significance at p < 0.05.
%Multilateral
-0.375
-0.284
0.323
0.245
-0.294
-0.075
0.084
0.191
-0.327
0.233
-0.349
0.246
0.531
0.626
0.444
-0.085
0.143
0.656
124
125
Table 4.3. Estimated Coefficients of Fitted Equation for %Unilateral Shape
Parameter
Intercept
Corn-Oil x Environment (IN)
Corn-Oil x Environment (NE)
Corn-Oil x Environment (OH)
Corn-Oil x MW Wattage (1050)
Corn-Oil x MW Wattage (1240)
Hybrid (VYP213) x Environment (IN)
Hybrid (VYP213) x Environment (NE)
Hybrid (VYP213) x Environment (OH)
Hybrid (VYP313) x Environment (IN)
Hybrid (VYP313) x Environment (NE)
Hybrid (VYP313) x Environment (OH)
Hybrid (VYP313) x Environment (IN)
Hybrid (VYP313) x Environment (NE)
Hybrid (VYP213) x MW Wattage (1050)
Hybrid (VYP213) x MW Wattage (1240)
Hybrid (VYP313) x MW Wattage (1050)
Hybrid (VYP313) x MW Wattage (1240)
Hybrid (VYP321) x MW Wattage (1050)
Hybrid (VYP321) x MW Wattage (1240)
Environment (IN) x MW Wattage (1050)
Environment (IN) x MW Wattage (1240)
Environment (NE) x MW Wattage (1050)
Environment (NE) x MW Wattage (1240)
Corn-Oil x Environment (IN) x MW Wattage (1050)
Corn-Oil x Environment (IN) x MW Wattage (1240)
Corn-Oil x Environment (NE) x MW Wattage (1050)
Corn-Oil x Environment (NE) x MW Wattage (1240)
Hybrid (VYP213) x Environment (IN) x WW Wattage (1050)
Hybrid (VYP213) x Environment (IN) x WW Wattage (1240)
Hybrid (VYP213) x Environment (NE) x WW Wattage (1050)
Hybrid (VYP213) x Environment (NE) x WW Wattage (1240)
Hybrid (VYP313) x Environment (IN) x WW Wattage (1050)
Hybrid (VYP313) x Environment (IN) x WW Wattage (1240)
Hybrid (VYP313) x Environment (NE) x WW Wattage (1050)
Hybrid (VYP313) x Environment (NE) x WW Wattage (1240)
a
significant at p < 0.01;
b
significant at p < 0.05
Estimate
0.246
0.002
-0.002
-0.004
0.005
0.009
-0.172
-0.141
-0.070
-0.162
-0.048
-0.027
-0.121
-0.069
-0.067
-0.077
-0.081
-0.106
-0.119
-0.125
0.162
0.118
0.032
0.074
-0.012
-0.010
0.003
-0.008
-0.107
-0.079
-0.015
-0.019
0.008
0.067
-0.079
-0.034
Standard
Error
0.021
0.002
0.002
0.002
0.003
0.003
0.029
0.029
0.024
0.032
0.029
0.025
0.029
0.033
0.033
0.033
0.033
0.028
0.029
0.032
0.041
0.045
0.046
0.046
0.004
0.004
0.004
0.004
0.049
0.051
0.054
0.051
0.051
0.051
0.051
0.048
t Value
11.74 a
1.45
-1.01
-2.09 b
1.96
3.65 a
-6.01 a
-4.81 a
-2.89 a
-5.08 a
-1.67
-1.09
-4.18 a
-2.1 b
-2.02
-2.32 b
-2.44 b
-3.73 a
-4.06 a
-3.91 a
4a
2.59 b
0.7
1.62
-3.36 a
-2.88 a
0.67
-2.18 b
-2.2 b
-1.55
-0.27
-0.38
0.16
1.3
-1.55
-0.72
Table 4.4. Analysis of Variance for Unilaterally-Expanded Popcorn Flakes
Source
Model
Corn-to Oil
Hybrid
Environment
Wattage
Corn-to Oil x Environment
Corn-to Oil x Wattage
Hybrid x Environment
Hybrid x Wattage
Environment x Wattage
Corn-to Oil x Environment x Wattage
Hybrid x Environment x Wattage
Error
Corrected Total
DF
Type III Sum
of Squares
35
0.127
1
0.001
2
0.037
2
0.007
2
0.011
2
0.001
2
0.004
4
0.006
4
0.002
4
0.018
4
0.018
8
0.014
29
0.020
64
0.147
a
significant at p < 0.01
b
significant at p < 0.05
Mean Square
F Value
0.004
0.001
0.018
0.003
0.005
<0.001
0.002
0.001
<0.001
0.005
0.004
0.002
0.001
5.24 a
1
26.59 a
4.89 b
7.71 a
0.58
2.58
1.98
0.53
6.49 a
6.33 a
2.56 b
126
Table 4.5. Estimated Coefficients of Fitted Equation for Bilateral and Multilateral Flakes (each response modeled individually)
Parameter
Intercept
Hybrid VYP213
Hybrid VYP313
Corn/Oil Ratio
(Corn/Oil Ratio)2
Bilateral Polymorph
Standard
Estimate
t Value
Error
0.425
0.093
0.078
-0.028
0.001
0.032
0.020
0.019
0.010
0.001
a
b
13.4 a
4.8 a
4.0 a
-2.7 a
2.4 b
Multilateral Polymorph
Standard
Estimate
t Value
Error
0.431
-0.036
-0.080
0.027
-0.001
0.030
0.018
0.018
0.010
0.001
14.5 a
-2.0 b
-4.4 a
2.8 a
-2.5 b
significant at p < 0.01
significant at p < 0.05
127
Table 4.6. Analysis of Variance for Bilateral and Multilateral Flake Shapes (note: each response modeled individually)
Bilateral Polymorph
Sources of variation
DF
Sum of
squares
Mean
Squares
Multilateral Polymorph
Sum of
squares
F Value
Mean
Squares
F Value
Model
4
0.158
0.039
9.96 a
0.112
0.028
7.7 a
Hybrid
2
0.1088
0.0544
13.21 a
0.070
0.035
9.7 a
a
0.028
0.028
7.8 a
0.023
0.023
6.5 b
0.216
0.004
Corn/Oil Ratio
1
0.0306
0.0306
7.43
(Corn/Oil Ratio)2
Error
1
0.0245
0.0245
5.95 b
60
0.247
0.004
Corrected Total
64
0.405
a
0.328
significant at p < 0.01
b
significant at p < 0.05
128
Table 4.7. Estimated Coefficients in Universal Model for %Unilateral, %Bilateral and %Multilateral Flake Shapes
Unilateral
Variable
Description
Estimate
Standard
Error
Multilateral c
Bilateral
t Value
Estimate
Standard
Error
t Value
Model
0.080
0.011
9.77 a
0.12290
0.02386
5.71 a
a
b0
Intercept
0.1984
0.028
7.13
0.3750
0.052
7.14 a
b1
Corn-oil ratio
0.0004
0.005
0.08
-0.0274
0.010
-2.67 a
a
b2
MW Wattage
-0.0001
<0.001
-2.83
0.0000
<0.000
1.11
b3
xh1
-0.0576
0.010
-5.49 a
0.0929
0.020
4.69 a
b4
xh2
0.0004
0.010
0.04
0.0778
0.020
3.99 a
b5
xe1
-0.0104
0.011
-0.99
0.0170
0.020
0.86
b6
xe2
0.0347
0.010
3.32 a
-0.0011
0.020
-0.06
b7
xco2
0.0000
<0.001
0.01
0.0013
0.001
2.38 b
Error
0.0669
0.001
0.2381
0.004
Corrected Total
0.1471
0.4051
a
significant at p < 0.01
b
significant at p < 0.05
c
coefficient estimates for multilateral intrinsically defined by %unilteral and %bilateral, so regression statistics were not obtained
Estimate
0.4268
0.0270
0.0000
-0.0353
-0.0782
-0.0066
-0.0336
-0.0013
129
130
CHAPTER 5. PACKING CHARACTERISTICS OF POPPED POPCORN WITH
DIFFERENT FLAKE SHAPES AND SIZE
5.1. Abstract
The objective of this research was to understand how size and shape of popped
popcorn affects bulk packing properties and expansion volume measurements, the most
important quality attribute for popcorn. Discrete sizes and shapes of popped popcorn
were physically separated and expansion volume was measured on each fraction. Void
volume and packing density of popcorn were derived using a new method involving sand
displacement. Results showed that both size and shape had a significant (p<0.05) effect
on measured responses. Small unilateral flakes had the highest packing density at 43.0%
and lowest expansion volume at 24.5 cm3/g, while large bilateral and multilateral flakes
had packing density <40% and expansion volumes >65 cm3/g. Empirical experiments
were also conducted using a movie theatre-style popcorn packaging tub. Results showed
the number of pieces of popcorn required to fill the package varied from 250-947
depending on the size and shape of popcorn used, with large bilateral or multilateral
flakes requiring the fewest popped kernels.
5.2. Introduction
Expansion volume is the most studied and reported quality attribute for popcorn.
Tender and fluffy popcorn associated with high expansion volume has been shown more
desirable by consumers (Lyerly 1942; Levy 1988), and increased expansion volume
produces greater profits for commercial venues like movie theatres where popcorn is
131
purchased by weight and sold by bulk volume (Song et al 1991; Hoseney et al 1983).
The most widespread method for determining expansion volume has been by measuring
bulk volume of popped popcorn (i.e., popcorn flakes) in a graduated cylinder and then
dividing by the original kernel weight (Dofing et al 1990; Song et al 1991; Mohamed et
al 1993).
The physical basis for popcorn expansion volume has not been comprehensively
reported, although several previous studies have pointed out that expansion volume
measurements are a function of many factors including flake size (Dofing et al 1990),
flake density (Sweley et al 2012), and the void spaces between kernels (Tian et al 2001).
The complexity of factors contributing to popcorn expansion volume measurements
parallels research of other cereal grains and legumes, which has shown bulk volume and
density measurements to be affected by a number of factors including 1) the combination
and distribution of individual sizes and shapes in a mixture, 2) the interaction effects
between grain particles and adjacent particles and 3) the container boundaries (Boac et al
2010).
Random packing of granular materials is a pervasive research topic in engineering
and physical sciences with application in both food and non-food systems. Computerassisted modeling using probabilistic algorithms or dynamic simulations have commonly
been used for understanding random packing of geometrically-symmetrical particles
(Gan et al 2004). Such studies predict packing density for spheres and ellipsoids ranging
from ~55% for random loose packing up to 64% for random close packing (Wang et al
2011). Compared to the extensive research on spherical or nearly spherical particles,
very few simulations have been done on packing of irregular shaped objects such as
132
popped popcorn, perhaps due to the complexity and cost of such modeling (Gan et al
2004). In fact, popcorn is an attractive candidate for modeling irregular-shaped particle
packing, because popcorn flakes have been shown to assume discrete, irregular shapes
when popped (Sweley et al 2011) and are relatively large particles (in comparison to, say,
sand particles) which makes it possible to utilize popcorn flakes for simple, controlled
experiments.
Thus, the purpose of this study was to investigate how material properties such as
size, shape, and distribution of popcorn flakes affect packing properties and expansion
volume measurements of popped popcorn through the development of statistical-based
models. Subsequently, the empirically-derived results from these studies were applied to
understand the packing properties of popcorn in a typical movie theatre packaging tub.
The results of this study may be particularly useful to purveyors of popcorn kernels at
commercial venues, since understanding how to increase expansion volume might lead to
new opportunities for increasing both consumer satisfaction and economic profitability of
popcorn sales. In addition, the models and observations developed for popped popcorn
might provide insights to the physical basis of packing irregular-shaped particles that can
be applied to other fields of study.
5.3. Materials and Methods
5.3.1. Popcorn Samples
Samples of a commercial, butterfly-type popcorn hybrid (YP-213) were obtained
from ConAgra Foods (Omaha, NE). The hybrid used in this study was a composite
sample from seed production grown in Nebraska, Iowa, and Ohio in 2010. Popcorn
133
kernels were tempered to 14.0% moisture content by storing at 21.5 °C and 73% relative
humidity until used for experimentation. Sixty g of kernels were then added to standard
microwave popcorn bags without addition of other ingredients and popped using a GE
1200 W microwave oven (Model PEB2060) until the interval between pops slowed to 2-3
s. After popping, the contents were poured into a steel sieve with a 7.94 mm round-hole
opening (Seedburo, USA Model 007) to remove unpopped kernels.
5.3.2. Flake Size and Shape Determination
Popcorn flakes were visually inspected and segregated into three shapes
previously identified for popped popcorn, depending on whether the flake appendages
were expanded unilaterally, bilaterally, or multilaterally (Sweley et al 2011). A total of
56 bags of popcorn were popped and sorted, resulting in several thousand pieces of each
shape type. Next, four hundred flakes of each shape type were randomly selected and
measured along three mutually perpendicular axes using digital calipers to the nearest
0.01 mm (Mitutoya America Corp., Aurora, IL). Flake size was taken to be the
geometric mean diameter of the three orthogonally-oriented dimensions, which was
calculated by the cubic root of the multiplied length, width, and thickness. Using the
distribution of average diameter sizes, popcorn flakes of each shape type were further
sorted and designated as being small, medium, or large, with size delineations established
at ± 0.5 standard deviations from the mean geometric diameter for each shape variety
(Figure 5.1, Figure 5.2).
134
5.3.3. Test Design and Packing Experiments
A completely randomized, full-factorial design requiring a total of 25 runs was
developed using Design Expert statistical software (version 8.0, Stat-Ease, Inc.,
Minneapolis, MN) to assess the effect of popcorn flake size and shape on expansion
volume and packing characteristics. Flake shape (unilateral, bilateral, and multilateral)
was designated as a categorical variable, while flake size was taken as a continuous
variable using the estimated average geometric diameter for flakes of each shape. This
was done to account for the differences in size distribution among flake shapes. For
simplicity and to accentuate differences, each run was comprised of only one flake size or
shape and only small and large flakes for each shape type were included in the experimental
design.
The total number of popcorn flakes was held constant at 300 pieces for all runs.
For each run, the required number of popcorn flakes of each size and shape were counted
and mixed together for 60 s using a rotating vertical tumbler drum (Krispy Kist Machine
Co., Chicago, IL) at 30 rpm to ensure random mixing of particles. The mixed popcorn
was weighed and then poured into a 4 L graduated cylinder (Nalgene, Model 3663-4000,
Rochester, NY, USA) which was inverted once and the resultant pop volume rounded to
the nearest 25 mL graduation. Expansion volume was determined by dividing total
popped volume by the weight of popped flakes (Eq. 1; Pordesimo et al 1990).
Expansion Volume =
(1)
The graduated cylinder still containing popcorn used for measuring pop volume
was fitted with a funnel having a stem opening with internal diameter of 11.1 mm (Blitz
USA, Model 05064, Miami, OK, USA). Next, the graduated cylinder was affixed to an
135
orbital shaker (VWR International, Model DS-500, Randor, PA, USA) set to 100 rpm
with a 1” planar rotation. Fine white silica sand purchased at a local hardware store was
selected as a placement medium. Testing revealed the sand had mean particle size of
0.329 ± 0.001 mm as determined by ASTM International sieving method (Pope and Ward
1998) and particle size calculation defined by Ensor et al (1970). Sand was added by
manually through the funnel at ~450 g/min until the sand level reached the recorded
graduation level of popcorn (Fig. 5.3). The mass of sand used for displacement was
divided by the measured density of the sand (1.69 g/cm3) to determine the total volume of
sand added.
Percentage void volume was calculated by dividing the volume of sand added to
the popcorn by the total pop volume (Eq. 2). Packing density was defined as the
proportion of physical space occupied by particles in bulk packed systems (Gan et al
2004) and expressed as a percentage (Eq. 3).
Void Volume (%) =
x 100%
Packing Density (%) = 100% – Void Volume
(2)
(3)
Archimedes‟ principle is commonly applied to determine the volume and density
of objects in a displacement medium. As such, the data collected during this research
also made it possible to determine the average density for each flake size/shape type,
which was taken to be the weight of popped flakes divided by the actual space occupied
by the popped flakes (Eq. 4).
Average Flake Density =
(4)
136
5.3.4. Movie Theatre-Style Packaging Application
To demonstrate the pragmatic implications of varying popcorn flake size and
shape, a standard “medium sized” movie theatre popcorn packaging tub was obtained
from the Solo Cup Company (Spec No. VP130-00061, Lake Forest, IL). The packaging
tub was measured and shown to have interior cavity height of 15.4 cm, bottom rim
diameter 14.0 cm, and upper rim diameter of 17.0 cm. These dimensions were used to
establish the interior cavity volume of 2,915 cm3.
Next, popcorn flakes comprised of all homogeneous sizes and shapes were
manually added to the packaging tub at an approximate rate of 1 flake per second.
Individual flakes were dropped with random initial orientation across the entire area of
the packaging tub from a height approximately 10 cm above the rim. This method of
flake addition was used to ensure consistency between runs and to ensure the packaging
tub was more evenly filled to the rim, since it is known that adding particles from a fixed
point above the container would result in undesirable heaping (Gan et al 2004).
Flake addition continued until the packaging container was filled up to – but did
not exceed – the interior tub height, as determined using qualitative observation at eyelevel of the rim. The resultant number and weight of added popcorn flakes were
recorded, and expansion volume and packing density were calculated using Eqs. 5 and 6,
respectively. Average flake density for each flake size/shape from the sand displacement
test was used for calculations of packing density in the movie theatre tub. Flake density
for medium-sized flakes was taken to be the average of small and large-sized flake
densities.
Expansion Volume =
(5)
137
Packing Density =
(6)
Three replicates of all nine popcorn flake size and shape combinations were tested using
the movie theatre packaging tub.
5.3.5. Data Analysis
Statistical analyses for sizing flake shapes and empirical modeling were
performed using SAS software (version 9.2, SAS Institute, Cary, NC). Fisher‟s LSD test
for mean separation was used to determine significance (p ≤ 0.05) for all flake size
measurements. The distribution of size determinations for each shape type were plotted
in a histogram and tested for normality using the Shapiro-Wilk test (Shapiro and Wilk,
1965).
The relationships between flake size, popcorn weight, flake density, pop volume,
void volume, packing density, and expansion volume were analyzed using the SAS
correlation (CORR) procedure. Flake shape was not included in correlation analysis
because it was a categorical variable. For both sand displacement and movie theatre tub
packing experiments, analysis of variance (ANOVA) was performed independently on
the response variables using SAS general linear mixed models (GLIMMIX) procedure.
Mean values were compared using Fisher‟s LSD with significance determined at p ≤
0.05. ANOVA was also used to develop models predicting the relationship between
flake size and shape and the primary response variables (expansion volume and
expansion volume). Significance for model selection was defined as p<0.05.
138
5.4. Results and Discussion
5.4.1. Flake Size and Shape
Significant differences in the dimensions of popcorn flake shapes were observed
(Table 5.1). Multilateral popcorn flakes had the largest mean geometric diameter, and
unilateral flakes had the smallest geometric mean diameter of the three shapes. The
Shapiro-Wilk test upheld that geometric mean diameters were normally distributed for all
three shapes. The average ratio of length-to-diameter dimensions was close to 1.1 for
both unilateral and multilateral shapes, whereas the length of bilateral-shaped flakes was
1.7x greater than the diameter, and the length of bilateral flakes was the largest dimension
measured in all three shapes. The higher length-to-diameter dimension of bilateral flakes
is an important consideration, since multiple studies on random packing of cylindrical
particles have shown that packing density decreases as the aspect ratio (length-todiameter ratio) increases (Gan et al 2004).
5.4.2. Displacement Testing
Packing density of bulk particles is commonly determined empirically by filling
interstitial voids with displacement media. While water, mercury, and compressed gas
have commonly been used as displacement media, popcorn is porous which adversely
affects accurate determination of void volume by fluid displacement (Chang 1988).
Because displacement with a flowable, solid medium that would not penetrate the
popcorn was desirable, pre-testing was conducted using various displacement media
including rapeseed, amaranth seed, and sand. While rapeseed displacement is an
established method for measuring volume of baked goods (AACC International, 2010),
139
significant bridging was observed along the graduated cylinder wall during displacement
testing with rapeseed, so it was eliminated from consideration. No bridging was
observed with amaranth seed or sand displacement, but sand was ultimately selected for
this study since it has been previously established as an effective displacement medium
for measuring packing efficiency in other porous grains (Doehlert and McMullen 2008).
It was speculated that gyration of the orbital shaker and addition of sand might
increase popcorn compaction during displacement testing. However, most popcorn
maintained its bulk volume during displacement testing, as indicated by the observed
graduation level changing no more than 25 cm3 (with many runs showing no change)
after sand addition, suggesting that minimum packing occurred during displacement
testing. One possible explanation for this may be that sand was added sufficiently slowly
to form a close fitting envelope around individual popcorn pieces and effectively held the
popcorn in place during displacement testing. In fact, the effectiveness of mercury
displacement to measure bulk volume and density of packed particles is ascribed to
mercury‟s ability to conform to the surface of irregular objects and form a tight envelope
that holds bulk particles in place (Webb 2001).
Another possible explanation for popcorn having mostly maintained its bulk
volume during sand addition may be found in the method used for measuring expansion
volume, which introduced popcorn flakes from the top of the 60 cm tall graduated
cylinder and inverted the cylinder once before taking measurements. Dropping particles
from a greater height and vertical jamming or tapping of particles are known to increase
packing density (Zhang et al 2001; Wouterse and Philipse 2006). So it may be that sand
addition during displacement had little effect on packing density because the popcorn
140
pieces in graduated cylinder measurements are more densely jammed than would be
observed in random loose particle packing.
5.4.3. Packing Performance
Significant differences were noted between packing characteristics of different
popcorn flake sizes and shapes (Table 5.2). Large bilateral and large multilateral flakes
produced the highest pop volumes, expansion volumes, and void volumes of all flake
types. These observed values for expansion volume are significantly greater than the 3655 cm3/g expansion volumes previously reported for microwave popcorn and experienced
by consumers at home (Mohamed et al 1993; Metzger et al 1989), which suggests that
optimizing flake size and shape might lead to increases in expansion volume.
Average packing density ranged from 35.5% for large bilateral flakes to 43.0%
for small unilateral flakes; across all runs and flake types, packing density averaged 39.5
± 3.0%. Packing density of popcorn has not been previously reported, although these
pack densities for popcorn are less than the 53-55% packing density reported for oats
(Dohelert and McMullen 2008) and significantly lower than the 68% random packing
density predicted from modeling oblate spheroids such as M&M‟s® candies (Chaikin et
al 2006). It is known that interacting particles create an excluded volume effect
(Williams and Philipse 2003), and it is likely that the extended appendage arms found in
popcorn gave rise to larger interstitial void spaces. In addition, while graduated cylinders
are the established equipment for measuring popcorn expansion volume, the narrow
diameter of the graduated cylinder (9.9 cm) may have resulted in more void spaces than
would be present in typical popcorn containers such as those found in movie theatres due
to greater interactions with the walls of the cylinder (see below).
141
Small unilateral flakes had the lowest pop volume and expansion volume and the
highest packing density of all flake types. Small and large unilateral popcorn flakes had
significantly greater average flake density than bilateral and multilateral flakes, which are
consistent with sensory results showing unilaterally-expanded popcorn flakes had the
highest intensity scores for texture density, crispness, and crunchiness attributes
compared to bilateral and multilateral flakes (Sweley et al 2011).
Across all runs, average flake density was 0.054 ± 0.021 g/cm3. No previous
studies have reported the density of popped popcorn based on weight per actual physical
space occupied by the flake. Determining average popcorn flake density based on
average individual kernel expansion rather than using bulk measurements might be used
by breeders to assess popping performance to make hybrid selections.
5.4.4. Relationship of Flake Properties to Packing Characteristics
Correlation analysis revealed significant (p<0.01) relationships between many
popcorn flake properties and packing characteristics (Table 5.3). Larger popcorn flake
size was positively correlated to flake weight and negatively correlated to average flake
density. Correlation analysis also showed that expansion volume was positively
correlated to void volume and negatively correlated to packing density. Larger flake size
was positively correlated to pop volume and expansion volume, but negatively correlated
to packing density. This analysis supports the premise that flake size is an important
contributor to packing density and expansion volume, as the results suggest that flake size
accounts for 62% of the variation in packing density and 94% of the variation in
expansion volume measurements given a homogeneous mix of popcorn shapes.
Similarly, 94% of variation in expansion volume could be attributed to popcorn flake
142
density. This result is consistent with previous results in oats showing that grain density
variation comprised nearly 80% of bulk density measurement (Doehlert and McMullen
2008).
5.4.5. Flake Size and Shape Effects on Expansion Volume and Packing Density
Analysis of variance on the factorial design for homogeneous flake types (Table
5.4) showed good fit of the models to data for both expansion volume and packing
density (R2 = 0.99, 0.77, respectively). The parameter estimates for the models are
shown in Table 5.5. For packing density, the interaction of flake size and shape was not
significant (p>0.05), and so the final model includes only the main effects of size and
shape. While both size and shape were statistically significant (p<0.01), the mean sum of
squares indicated that flake size contributed ~4.5x more to the estimated packing density
than flake shape. For expansion volume, the interaction of flake size and shape was the
highest order effect showing significance (p<0.01), so simple effects must be used for
understanding the relationship between design factors and expansion volume.
The results of statistical modeling indicate that expansion volume increased and
packing density decreased for each shape type as flake size became larger. Unilaterallyexpanded popcorn flakes had the lowest predicted expansion volume and highest packing
density of all three shape types within the design limits for unilateral sizes. It may be that
the low aspect ratio and limited appendages emanating from unilateral flakes facilitated
denser particle packing, leading to the lower expansion volume. Between the other two
popcorn flake types, bilaterally-expanded shapes had greater expansion volume and
lower predicted packing density than multilateral shapes across all flake sizes. These
predictions provide empirical support to simulated packing research showing that
143
increased aspect ratio (length-to-diameter ratio) decreases packing density. Moreover,
the findings suggest that popcorn expansion volume might be increased by optimizing the
proportion of bilateral-shaped flakes and maximizing average flake size.
5.4.6. Packing Performance in Movie Theatre-Style Packaging
The number of popped popcorn pieces needed to fill the movie-theatre packaging
tub varied significantly depending on the shape and size of popcorn flakes added (Table
5.6). Large bilateral and large multilateral shapes required the least number of popped
flakes to fill the package (269 and 266 flakes on average, respectively). In contrast, more
than 600 medium-sized and over 900 small-sized unilateral flakes were required to fill
the package. Beyond simply requiring more flakes to fill the packaging tub, it was
observed that packing density was comparatively higher for unilateral flakes than the
other two shapes. In fact, all sizes of unilateral shapes had significantly (p<0.05) lower
estimated expansion volume than bilateral and multilateral flakes. Large bilateral and
multilateral flakes had the largest expansion volume and smallest packing density of all
flake types, which is consistent with results found in the graduated cylinder testing.
In general, the estimated expansion volume and packing densities in the movie
theatre tub were consistent with values illustrated in Table 5.2 for sand displacement
testing in the graduated cylinder. It was hypothesized that packing density might be
higher in the movie theatre tub, since it is known that a slower rate of particle
introduction to a packaging container increases pack density (Zhang et al 2001) and
particle contact with container walls reduces the efficiency of particle-particle
interactions (Wouterse and Philipse 2006). One possible explanation for this unexpected
similarity in pack density is that the graduated cylinder was inverted before taking
144
measurements, whereas the movie theatre tub remained stationary during filling, which
may have resulted in a looser pack density than could have been achieved with shaking.
Ultimately, these results suggest that measurements made in graduated cylinder testing
and conventionally used by popcorn breeders and kernel wholesalers for reporting
expansion volume provide good indication of the bulk packing properties of popcorn in
movie theatre style packaging tubs.
While the design of this experiment used only flakes with similar size and
morphology, popcorn found in uncontrolled settings (such as purchased at a movie
theatre or made at home in microwave) would be expected to contain a mixture of
popcorn flakes with different sizes and shapes (Sweley 2012), and linear packing models
predicts loose pack density tends to (but does not always) increase in multi-sized
mixtures with three nonspherical geometric particles due the increased chance of particles
fitted together more tightly (Stovall et al 1986; Yu and Zou 1996). Thus, further
investigation to characterize the relationship between heterogeneous mixtures of different
shaped popcorn flakes in bulk packing would be meaningful.
5.5. Conclusions
This study showed that size and shape of popcorn flakes has a significant effect
on expansion volume measuring and bulk packing characteristics. Bilateral and
multilaterally- shaped popcorn flakes produce greater expansion volumes than unilateral
shapes. Models predict that for all shape types, increasing flake size increases the
expansion volume and decreases packing efficiency. Analysis shows that larger-sized
145
popcorn flakes have lower average flake density, and flake density varies 0.054 ± 0.021
g/cm3 depending on the size and shape.
The number of popcorn flakes required to fill a standard movie-theatre style tub
varies significantly depending on the size and shape of flakes added. Small unilateral
flakes pack together the most tightly and require over 900 popcorn pieces to fill the
packing tub, while large bilateral and large multilateral flakes pack loosest and require
less than 300 pieces. Packing density in the movie theatre packing tub averaged 38.2 ±
2.6% across all flake size-shape types, which is significantly less than the 53-55%
packing density reported for oats (Dohlert and McMullen 2008) and the 55-64% density
reported for random packing of hard spheres (Wang et al 2011).
Greater popcorn expansion volume is both desirable to consumers and
economically beneficial to commercial sellers of popped popcorn. Therefore, additional
research to help further optimize packing characteristics of multi-sized mixtures of
different shapes of popped popcorn in various packaging types would be valuable.
5.6. Literature Cited
AACC International. 2010. Approved Methods of Analysis, 11th Ed. Method 10-05.01.
AACC International, St. Paul, MN, U.S.A.
Boac, J. M., Casada, M. E., Maghirang, R. G., and Harrier III, J. E. 2010. Material and
interaction properties of selected grains and oilseeds for modeling discrete particles.
Trans. ASABE 53:1201-1216.
146
Chaikin, P. M., Donev, A., Man, W., Stillinger, F. H., and Torquato, S. 2006. Some
observations on the random packing of hard ellipsoids. Ind. Eng. Chem. Res.
45:6960-6965.
Chang, C. S. 1988. Measuring density and porosity of grain kernels using a gas
pycnometer. Cereal Chem. 65:13-15.
Doehlert, D. C., and McMullen, M. S. 2008. Oat grain density measurement by sand
displacement and analysis of physical components of test weight. Cereal Chem.
85:654-659.
Dofing, S. M., Thomas-Compton, M. A., and Buck, J. S. 1990. Genotype x popping
method interaction for expansion volume in popcorn. Crop Sci. 30:62-65.
Ensor, W. L., Olson, H. H., and Colenbrander, V. F. 1970. A report: committee on
classification of particle size in feedstuffs. J. Dairy Sci. 53:689-690.
Gan. M., Gopinathan, N., Jia, X., and Williams, R. A. 2004. Predicting packing
characteristics of particles of arbitrary shapes. KONA 22:82-91.
Hoseney, R. C., Zeleznak, K., and Abdelrahman, A. 1983. Mechanism of popcorn
popping. J. Cereal Sci.1:43-52.
Levy, B. 1988. A new perspective on popcorn. Snack World 45:24.
Lyerly, P. J. 1942. Some genetic and morphological characters affecting the popping
expansion of popcorn. J. Am. Soc. Agron. 34:986-999.
Metzger, D. D., Hsu, K. H., Ziegler, K. E., and Bern, C. J. 1989. Effect of moisture
content on popcorn popping volume for oil and hot-air popping. Cereal Chem.
66:247-248.
147
Mohamed, A. A., Ashman, R. B., and Kirleis, A. W. 1993. Pericarp thickness and other
kernel physical characteristics relate to microwave popping quality of popcorn. J.
Food Sci. 58:342-346.
Pope, L.R., and Ward, C.W. 1998. Manual on Testing Sieving Methods: Guidelines for
Establishing Sieve Analysis Procedures: 4th Edition, ASTM International.
Pordesimo, L. O., Anantheswaran, R. C., Fleishmann, A. M., Lin, Y. E., and Hanna, M.
A. 1990. Physical properties as indicators of popping characteristics of microwave
popcorn. J. Food Sci. 55:1352-1355.
Shapiro, S. S., and Wilk, M. B. 1965. An analysis of variance test for normality
(complete samples). Biometrika 52:591-611.
Song, A., Eckhoff, S. R., Paulsen, M., and Litchfield, J. B. 1991. Effects of kernel size
and genotype on popcorn popping volumes and number of unpopped kernels. Cereal
Chem. 68:464-467.
Stovall, T., De Larrard, F., and Buil, M. 1986. Linear packing density of grain mixtures.
Powder Tech. 48:1-12.
Sweley, J. C., Rose, D. J., and Jackson, D. S. 2011. Composition and sensory evaluation
of popcorn flake polymorphisms for a select butterfly-type hybrid. Cereal Chem.
88:321-327.
Sweley, J. C., Rose, D. J., and Jackson, D. S. 2012. Hybrid and environment effects on
popcorn kernel physiochemical properties and their relationship to microwave
popping performance. J. Cereal Sci. doi:10.1016/j.jcs.2011.11.006.
Tian, Y., Buriak, P., and Eckhoff, S. R. 2001. Effect of hybrid and physical properties of
individual popcorn kernels on expansion volume. Cereal Chem. 78:578-582.
148
Wang, P., Song, C., Jin, Y., and Makse, H. A. 2011. Jamming II: Edwards‟ statistical
mechanics of random packings of hard spheres. Physica A 390:427-455.
Webb, P. A. 2001. Volume and density determinations for particle technologists.
http://www.particletesting.com/docs/density_determinations.pdf. Accessed January
21, 2012.
Williams, S. R., and Philipse, A. P. 2003. Random packings of spheres and
spherocylinders simulated by mechanical contraction. Phys. Rev. E 67:051301(1-9).
Wouterse, A., and Philipse, A.P. 2006. Geometric cluster ensemble analysis of random
sphere packings. J. Chem. Phys. 125:194709.
Yu, A. B., and Zoh, R. P. 1996. Modifying the linear packing model for predicting the
porosity of nonspherical particle mixtures. Ind. Eng. Chem. Res. 35:3730-3741.
Zhang, Z. P., Liu, L. F., Yuan, Y. D., and Yu, A. B. 2001. A simulation study of the
effect of dynamic variables on the packing of spheres. Powder Tech. 116:23-32.
Figure 5.1. Average flake size for 400 popped popcorn flakes. Size limits (small, medium, and large) established by geometric mean
diameter ± 0.5SD for each shape.
149
Figure 5.2. Popcorn flake shapes and sizes delineated for packing experiments. Medium-sized flakes were excluded from factorial
design but used in movie theatre package testing.
150
151
Figure 5.3. Visual depiction of apparatus used for measuring popcorn expansion volume
and packing characteristics (void volume, packing density) using sand displacement.
Table 5.1. Average Dimensions for Three Shapes of Popped Popcorn Measured by Digital Calipers a
Length (mm)
Width (mm)
Height (mm)
Length:Diameter
Ratio b
Geometric Mean
Diameter (mm) c
Unilateral
20.4 ± 4.8c
16.3 ± 3.5c
20.5 ± 4.1b
1.1 ± 0.3b
18.8 ± 3.0c
Bilateral
32.4 ± 4.0a
20.7 ± 3.0b
17.9 ± 3.3c
1.7 ± 0.4a
22.8 ± 2.3b
Shape
Multilateral
25.2 ± 3.8b
22.2 ± 3.5a
25.4 ± 4.1a
1.1 ± 0.2b
24.0 ± 2.4a
Mean values ± SD in same column followed by the same letter are not significantly different (P<0.05). n=400.
b
Calculated as length / [(width x height) / 2]
c
Calculated as cubic root of length x width x height.
a
152
Table 5.2. Pop Volume and Packing Characteristics for Different Sizes and Shapes of Popped Popcorn
Popcorn Flake
Size & Shape a
n
Pop
Volume
(cm3)
Popcorn
Weight
(g)
Expansion
Volume
(cm3/g)
Void
Volume
(%) b
Packing
Density
(%) c
Average
Popcorn Flake
Density (g/cm3)
Small Unilateral
5
910 d
37.1 d
24.5 d
57.0% c
43.0% a
0.0950 a
Large Unilateral
4
2100 bc
46.1 b
45.5 c
59.4% b
40.6% b
0.0541 b
Small Bilateral
4
2013 c
35.9 e
56.0 b
60.7% b
39.4% b
0.0453 c
Large Bilateral
4
3125 a
46.8 ab
66.8 a
64.5% a
35.5% c
0.0422 c
Small Multilateral
4
2225 b
39.8 c
55.9 b
59.1% b
40.9% b
0.0437 c
Large Multilateral
4
3175 a
47.8 a
66.4 a
63.6% a
36.4% c
0.0414 c
2204
42.0
52.5
60.5%
39.5%
0.0536
Average
St Dev
814
5.0
15.8
3.0%
3.0%
0.0208
For each run, 300 popcorn flakes of homogeneous size and shape were tested. Run order was randomized.
b
Void volume determined by sand displacement.
c
The actual physical space occupied by popcorn in pop volume measurements, calculated as 100% - Void Volume.
d
Mean values in the same column followed by the same letter are not significantly different (P<0.05).
a
153
Table 5.3. Correlation Coefficients between Some Popcorn Flake Properties and Packing Characteristics a
Flake Size
Flake
Weight
Average Flake
Density
Pop
Volume
Void
Volume
Flake Weight
0.61
Average Flake Density
-0.92
-0.53
Pop Volume
0.96
0.78
-0.90
Void Volume
0.79
0.63
-0.72
0.86
Expansion Volume
0.97
0.57
-0.97
0.96
0.82
Packing Density
-0.79
-0.63
0.72
-0.86
-1.00
a
Expansion
Volume
-0.82
All correlations showed significance at p < 0.01
154
Table 5.4. Analysis of Variance for Expansion Volume and Packing Density a
Expansion Volume
Sources of
variation
a
b
DF
Sum of
squares
Packing Density
Mean
Squares
F Value
DF
Sum of
squares
Mean
Squares
F Value
Model
5
5771
1154
478 b
3
0.0168
0.0056
23.0 b
Size
1
1090
1090
423 b
1
0.0073
0.0073
29.9 b
Shape
2
54
27
10 b
2
0.0032
0.0016
6.5 b
Size*Shape
2
55
27
11 b
Error
19
49
3
21
0.0051
0.0002
Corrected Total
24
5820
24
0.0219
Each response modeled individually
Significant at p < 0.01
155
Table 5.5. Parameter Estimates for Models of Expansion Volume and Packing Density
Expansion Volume
Parameter
Packing Density
Estimate
Standard
Error
t Value
Estimate
Standard
Error
t Value
-93.8
6.6
-14.3 b
99.665
0.045
2212.7 b
6.9
0.4
19.5 b
-0.013
0.002
-5.5 b
47.5
13.2
3.6 b
-0.009
0.012
0.7
47.9
13.4
3.6
b
0.038
0.015
2.6 c
0.0
.
.
0.0
.
.
b
Intercept
Size
Shape Bilateral
Shape Multilateral
Shape Unilateral
Size * Shape Bilateral
-2.1
0.6
-3.5
Size * Shape Multilateral
-2.4
0.6
-4.0 b
0.0
.
.
Size * Shape Unilateral
Each response modeled individually
b
Significant at p < 0.01
c
Significant at p < 0.05
a
a
156
Table 5.6. Popcorn Flake Packing in a Standard Movie Theatre-Style Packaging Tub a
Flakes Needed to
Fill Tub
Flake Weight (g)
Expansion Volume
(cm3/g) b
Packing Density
(%) c
Small
934 ± 11 a
115.8 ± 2.2 a
25.2 ± 0.5 e
41.8 ± 0.8 a
Medium
636 ± 27 b
88.0 ± 3.6 b
33.2 ± 1.4 d
40.5 ± 1.6 ab
Large
402 ± 18 c
61.9 ± 1.8 c
47.1 ± 1.4 c
39.3 ± 1.2 bc
Small
416 ± 12 c
49.4 ± 1.3 d
59.0 ± 1.6 b
37.4 ± 1.0 cd
Medium
353 ± 7 de
48.5 ± 1.2 d
60.1 ± 1.5 b
38.1 ± 1.0 bc
Large
269 ± 7 g
42.1 ± 1.1 e
69.3 ± 1.8 a
34.2 ± 0.9 e
Small
383 ± 16 cd
49.7 ± 1.8 d
58.7 ± 2.1 b
39.0 ± 1.4 bc
Medium
343 ± 15 ef
50.3 ± 2.8 d
60.7 ± 1.7 b
38.8 ± 1.1 bc
Shape
Size
Unilateral
Bilateral
Multilateral
Large
266 ± 12 fg
42.4 ± 1.8 e
68.9 ± 3.0 a
35.1 ± 1.5 de
Mean values ± SD in same column followed by the same letter are not significantly different (P<0.05). n=3.
b
Calculated by dividing packaging tub volume by flake weight
c
Estimated by dividing flake weight by averaged flake density, then dividing by packaging tub volume.
a
157
158
GENERAL CONCLUSIONS
Conclusions
The present dissertation has described the testing of three hypotheses directed at
understanding and characterizing flake polymorphisms as an end-use quality attribute for
butterfly-type popcorn hybrids. The first hypothesis was that popcorn flake
polymorphisms could be identified, characterized, and differentiated. The second
hypothesis was the formation of popcorn flake polymorphisms could be influenced by
selection of both intrinsic and external factors. The final hypothesis was that flake shape
affects both consumer acceptability and packing characteristics of popcorn.
For the first hypothesis, it was demonstrated that distinct popcorn flake
polymorphisms can be identified by visual inspection based on whether the flake
appendages are expanded unilaterally, bilaterally, or multilaterally. Bilateral flakes could
be differentiated as those flakes having a higher aspect ratio (length-to-diameter ratio)
than unilateral and multilateral flakes. Significant differences in compositional and
flavor attributes exist and can be characterized between different popcorn flake shapes.
When popcorn flakes were isolated after microwave popping kernels with 30% oil added
to kernels, unilateral popcorn flakes retained the most fat, saturated fat, and sodium,
while multilaterally expanded flakes had the highest levels of protein, total carbohydrate,
and popcorn-like aromatic pyrazines.
For the second hypothesis, it was shown that formation of different flake
polymorphisms is dependent on both inherent kernel characteristics and external variables,
including hybrid, growing location and environment, amount of oil added to kernels, and
159
microwave oven wattage. There is an interaction effect between intrinsic and extrinsic
factors, which makes optimizing the relative proportion of different flake shapes complex
and particular for the desired outcome. For example, the relative proportion of unilateral
flakes can be increased by popping in reduced wattage microwave ovens.
For the third hypothesis, it was demonstrated that the shape of popcorn flakes
affects both consumer acceptability and packing characteristics of popcorn. While
sensory hedonics indicate no significant preference for the appearance of different flake
polymorphisms, the unilaterally-expanded shape received the highest consumer
acceptability scores for overall liking and flavor liking, perhaps owing to the higher
levels of fat and sodium. The unilateral and bilateral polymorphisms scored significantly
higher than the multilateral shape for texture liking in consumer testing. Unilateral flakes
were shown to have lower expansion volume and greater packing density, on average,
than bilateral and multilateral flakes. Empirical modeling revealed both flake size and
shape have a significant effect on bulk packing characteristics of popcorn. Small
unilateral flakes pack together the most tightly and require three times as many popped
flakes to fill a standard movie theatre packaging tub as large bilateral and multilateral
flakes.
Beyond understanding the shape of popcorn flakes, this research produced several
other notable discoveries. Expansion volume was positively correlated to kernel
sphericity and negatively correlated to thousand-kernel weight and total fat in the kernel.
Most unpopped kernels successfully pop when heated by microwave a second time.
Finally, the amount of oil retained by paper bags during microwave popping was shown
to be significant and increase linearly as a function of starting oil amount.
160
Significance of Findings and Suggestions for Future Research
Peter Drucker defined innovation as change that creates a new dimension of
performance. The significance of investigating and characterizing new quality attributes
for microwave popcorn is that it might inspire new innovation to help reverse the erosion
of popcorn consumption that has occurred during the past eighteen years. Increasing
relevance and consumption of popcorn as a result of better understanding popcorn shape
can be conceptualized in several ways. Firstly, popcorn breeders and producers and
microwave popcorn manufacturers continually seek to improve quality attributes which
will increase consumer satisfaction or create market differentiation. Popped popcorn
flake shapes may be one factor of considerable interest. Enhancing the understanding of
popcorn flake polymorphism as a quality attribute can support the development of new
hybrids or varieties and techniques to produce the most desirable microwave popcorn.
Secondly, providing a fundamental understanding of flake shape might also reveal
innovative opportunities to differentiate existing products or market new usage for
popcorn.
While these studies show promise for flake shape as an innovative quality
attribute for differentiating and optimizing popcorn quality, additional research to
establish a rapid and cost-effective method for quantifying flake shape would increase the
utility and likelihood of adoption by popcorn breeders and commercial popcorn sellers.
While some factors affecting shape formation were studied, investigation of additional
influences on popped shape would be valuable. For example, additional research focused
on the effect of differing kernel sizes, varying moisture contents from 14.0%, and
161
different methods of popping (including air popping and conventional oil popping) on the
prevalence of different popcorn polymorphisms would be meaningful. In addition,
further research on different microwave packaging structure and materials might provide
insight into whether shape formation is attributable to kernel placement in the bag or is an
inherent quality of the popcorn. Moreover, investigation focused on packaging materials
might reveal opportunities for minimizing oil stuck to the bag. Finally, additional
research on sensory hedonics and physiochemical and volatile flavor attributes of flake
shapes produced when popping popcorn without oil addition might elucidate whether
consumer preference can be ascribed to innate characteristics of the flake polymorphisms
(such as texture liking of popcorn-like flavor) or if consumer simply prefer more oil and
salt.
Beyond the shape of popped flakes, more research could be devoted to elucidating
other quality attributes for popcorn besides expansion volume and unpopped kernels. Hull
dispersion has been noted as an important quality factor for the eatability of popcorn, yet it
has not been comprehensively studied. Research to establish a standard method for
measuring and reporting hull dispersion might offer new tools for popcorn breeders to
improve or differentiate hybrids. Given the need for increasing healthful foods in the
American diet, additional research to evaluate nutritional aspects of popcorn would also be
meaningful. For example, research on differences in starch and dietary fiber structure or
nutraceutical components like antioxidants in popcorn would be relevant and valuable to
promoting the health benefits of popcorn beyond just whole grain.
162
APPENDICES
Appendix A. Sensory Testing Protocol and Ballot
163
Appendix A. Sensory Test Methodology and Ballot (continued)
Following instructions (screen 0), questions 1 – 13 are asked successively for each of the 3 popcorn samples presented.
Screen /
Question
0
Type
Attribute
Instruction Today you will be evaluating Butter Microwave Popcorn…
Scale
n/a
Scale labels
Scale Graduations
n/a
n/a
1
Hedonic
OVERALL, how much did you like or dislike the popcorn?
1–9
Dislike Extremely---Like Extremely
All scalepoints labeled
2
Hedonic
How much did you like or dislike APPEARANCE of the popcorn?
1–9
Dislike Extremely---Like Extremely
All scalepoints labeled
3
Hedonic
How much did you like or dislike the SHAPE of the popcorn?
1–9
Dislike Extremely---Like Extremely
All scalepoints labeled
4
Hedonic
How much did you like or dislike the FLAVOR of the popcorn?
1–9
Dislike Extremely---Like Extremely
All scalepoints labeled
5
Hedonic
How much did you like or dislike the TEXTURE of the popcorn?
1–9
Dislike Extremely---Like Extremely
All scalepoints labeled
6
Intensity
Describe the Overall Flavor of the popcorn
1 – 10 Extremely Weak -- Extremely Strong
Only endpoints labeled
7
Intensity
Describe the Butter Flavor of the popcorn
1 – 10 Extremely Weak – Extremely Strong
Only endpoints labeled
8
Intensity
Describe the Saltiness of the popcorn
1 – 10 Not at all Salty -- Extremely Salty
Only endpoints labeled
9
Intensity
Describe the Texture of the popcorn
1 – 10 Extremely Dense/Compact—Extremely
Light/Fluffy
Only endpoints labeled
10
Intensity
Describe the Crispness of the popcorn
1 – 10 Not at all Crisp—Extremely Crisp
Only endpoints labeled
11
Intensity
Describe the Crunchiness of the popcorn
1 – 10 Not at all Crunchy—Extremely Crunchy
Only endpoints labeled
12
Intensity
Describe the rate at which the popcorn dissolves in your mouth:
1 – 10 Extremely Slow –Extremely Fast
Only endpoints labeled
13
Verbatim
What about this product did you like?
14
Demographic Age,
education
Whatgender,
about this
product did you dislike?
n/a
n/a
n/a
n/a
n/a
n/a
164
165
Appendix B. Select SAS Code
B.1. Mean and Mean Difference Testing
Analysis of variance was used to determine and report mean, standard deviation,
and mean comparisons between groups of data. SAS general linear models (GLM)
procedure was used for data analysis in Chapters 2-4, while SAS general linear mixed
models (GLIMMIX) was used in Chapter 5. While both procedures for data analysis are
appropriate, GLIMMIX is a relatively newer procedure and the current preference of
statisticians. Fisher‟s LSD test for mean separation was used to determine significance (p
≤ 0.05) using lsmeans. The following is a sample of SAS code used for obtaining means
and determining significant differences for packing characteristics of different popcorn
flake shapes and sizes (Table 5.2). Similar code was used for Tables 2.3, 2.4, 2.6, 2.7,
2.8, 3.1, 3.2, 3.4, 5.1, and 5.6.
DATA example_B_1;
INPUT size shape pop_volume pop_weight exp_volume
void_volume pack_density flake_density;
DATALINES;
[data];
PROC GLIMMIX;
CLASS size shape;
MODEL size shape pop_volume pop_weight exp_volume
void_volume pack_density flake_density=size shape;
LSMEANS size shape/ PDIFF lines;
ODS OUTPUT LSMLINES=LSMLINES;
RUN;
B.2. Correlation Procedure
The relationship between nonparametric measures (design factors, attributes,
and/or responses) was determined using SAS correlation (CORR) procure. Pearson
product-moment correlations measure only linear association between two variables; it is
not intended to show causality and may incorrectly estimate the strength of nonlinear
166
relationships. The following is a sample of SAS code used for determining the
relationship between kernel properties and popped flake morphology (section 4.4.1.
results & discussion). Similar code was used for Tables 3.3, 3.5, 5.3, and 5.6.
DATA example_B_2;
INPUT size shape pop_volume pop_weight exp_volume
void_volume pack_density flake_density;
DATALINES;
PROC MEANS;
NOPRINT
CHARTYPE
NWAY
MEAN NONOBS;
VAR Uni Bi Multi Total_Fat Total_Carb Starch Amylose
Amylopectin Fiber Sugars Protein Total_Ash
Test_Weight True_Density Stenv_Time Stenv_Height
TADD_Index Thousand_Wt Sphericity;
ORDER=UNFORMATTED ASCENDING;
OUTPUT OUT=WORK.MEANSummaryStats
MEAN()=/AUTONAME AUTOLABEL WAYS INHERIT;
RUN;
PROC CORR;
PEARSON
OUTP=WORK.CORRS
VARDEF=DF
VAR Uni_Mean Bi_Mean Multi_Mean
WITH Total_Fat_Mean Total_Carb_Mean Starch_Mean
Amylose_Mean Amylopectin_Mean Fiber_Mean
Sugars_Mean Protein_Mean Total_Ash_Mean
Test_Weight_Mean True_Density_Mean
Stenv_Time_Mean Stenv_Height_Mean TADD_Index_Mean
Thousand_Wt_Mean Sphericity_Mean;
RUN;
B.3. PCA of Flavor Volatiles
In Chapter 2, volatile flavor analysis from GC-MS was analyzed by principal
component analysis. PCA was used to simplify the measured flavor volatiles into a
smaller number of principal components (PC) that account for most of the variance
observed. PC are linear combinations of optimally-weighted observed variables. The
first PC accounts the highest percentage of sample variance, the second component
accounts for the maximum amount of variance not described by the first component, and
167
so on. PCA was performed in SAS using FACTOR procedures as illustrated below. The
code “nfactors=4” was used because several iterations in SAS revealed that the four PC
accounted for 91.2% of the variation in the dataset. The PC were rotated using an
Orthovarimax rotation, so that the PC could be plotted orthogonally.
PROC FACTOR DATA=example_B_3;
METHOD=PRIN
VARDEF=DF
SINGULAR=1E-08
NFACTORS=4
PRIORS=ONE
ROTATE=VARIMAX
NORM=KAISER
OUT=output_data;
VAR Acetaldehyde Oxirane Ethanol Acetonitrile
Hexane Furan Methyl_Isobutyl_Ketone Toluene Hexanal
methylPyrazine Furfural dimethylPyrazin
methylNonane methyl2Nonene Decene Decane Nonanal
ethylDecane Dimethyldecane methylDodecane
methylUndecane methyleneUndecane Dodecene Dodecane
Z2Dodecene Cyclododecane octylCyclopropane
Methoxyvinylphenol Domethyldodecane Triacetin
methylTridecane methyleneTridecane
nonylCyclopropane Tetradecane Hexadecane;
RUN;
B.4. Regression Modeling for Shape Formation
The methods used for analysis of variance model selection and fitting for the
formation of each popcorn shape is comprehensively described in Section 4.3.3.
ANOVA was performed using SAS general linear models (GLM) procedure. Model
selection for each popcorn shape was achieved by backwards elimination from all main
effects, quadratic and two-way cross product interactions. Only effects that showed
significance at p < 0.05 were included in the fitted hierarchical model for each dependent
variable. The following is a sample of SAS code used for obtaining ANOVA and
parameter estimates after model reduction for % unilateral shape (Tables 4.1 and 4.2).
Note that hybrid, geography, and microwave wattage were defined as categorical
168
variables for this regression, while corn-oil ratio was quantitative. Similar code and
model fitting was used for the univariate regression on bilateral and multilateral shapes.
PROC GLM DATA=example_B_4;
CLASS Hybrid Geo Wattage;
MODEL Unilateral=cornoilratio Hybrid Geo Wattage
cornoilratio*Geo cornoilratio*Wattage
Hybrid*Geo Hybrid*Wattage Geo*Wattage
cornoilratio*Geo*Wattage Hybrid*Geo*Wattage
/SS3
SOLUTION;
OUTPUT OUT=WORK.Pred
PREDICTED=predicted_Unilateral
RUN;
As also described in Section 4.3.3., regression analysis to model the formation of
multiples shapes of popcorn simultaneously used two-stage least squares methodology
using the SAS procedure for estimating parameters in interdependent systems of linear
regression equations (SYSLIN). The efficiency of the parameter estimates are improved
by accounting for the correlated errors in this system of equations. Model selection was
achieved by backwards elimination from all main effects, quadratic and two-way cross
product interactions, and only effects showing significance at p < 0.05 for any of the
dependent variables were included in the final model. The SAS used for this analysis and
shown in Table 4-5 is illustrated below:
DATA example_B_4;
SET SASUSER.IMPW_00C1 WORK.FOR_PREDICTION;
cfsi=Bulk_Density_mL/(Unilateral+Bilateral+Multil
ateral);
IF Hybrid='VYP213' THEN hybrid_dummy=1;
IF Hybrid='VYP313' THEN hybrid_dummy=2;
IF Hybrid='VYP321' THEN hybrid_dummy=3;
IF Geo='IN' THEN geo_dummy=1;
IF Geo='OH' THEN geo_dummy=2;
IF Geo='NE' THEN geo_dummy=3;
wattage_dummy=Wattage*1; a_sq=COR**2;
RUN;
PROC SYSLIN DATA= example_B_4
OUTEST=e 2sls
OUT=predse;
ENDOGENOUS Unilateral Bilateral Multilateral;
169
INSTRUMENTS cornoilratio;
model1: MODEL Unilateral=cornoilratio wattage_dummy
b1 b2 c1 c2 a_sq ;
OUTPUT p=puni;
model2: MODEL Bilateral=cornoilratio wattage_dummy b1
b2 c1 c2 a_sq ;
OUTPUT p=pbil;
RUN;
B.5. Modeling Flake Size and Shape Effects on Expansion Volume and Packing Density
The methods used for analysis of variance model selection and fitting for the
effect of popcorn flake size and shape on expansion volume and packing density are
described in Section 5.3.5. ANOVA was performed using SAS general linear mixed
models (GLIMMIX) procedure. Main effects and interaction (size x shape) were
modeled. Flake shape was defined as a class variable since it was categorical; flake size
was taken as a quantitative factor based on the data transformations shown below.
Significance was established at p<0.05. The following is a sample of SAS code used for
obtaining ANOVA and parameter estimates for expansion volume (Tables 5-4 and 5-5).
ANOVA was also independently performed on the packing density response variable.
DATA example_B_5;
SET WORK.POPCORN_SHAPE_ANALYSIS_WITH_SIZ2;
IF size='Big' and shape='Mult' THEN size=25.22;
IF size='Big' and shape='Bi' THEN size=23.89;
IF size='Big' and shape='uni' THEN size=20.33;
IF size='Small' and shape='Mult' THEN size=22.87;
IF size='Small' and shape='Bi' THEN size=21.63;
IF size='Small' and shape='uni' THEN size=17.26;
Expansion_Volume=Pop_Volume/Popcorn_Weight;
Packing_Density=1-Void_Volume;
Average_Popcorn_Flake_Density=Popcorn_Weight/(
Pop_Volume*Packing_Density*300);
RUN;
PROC GLIMMIX DATA=example_B_5;
CLASS Shape;
MODEL Pop_Volume=size Shape size*Shape/SS3
SOLUTION;
MEANS Shape / tukey lines;
RUN;
Документ
Категория
Без категории
Просмотров
0
Размер файла
5 332 Кб
Теги
sdewsdweddes
1/--страниц
Пожаловаться на содержимое документа