Dynamic Viscoelastic Properties and Glass Transition Behavior of Corn Kernels

Abstract

Information on viscoelastic properties is useful in industrial processing of foods as it provides insight into the structure of the material as a function of processing conditions. Corn kernels were hand sanded to obtain smooth and closely uniform samples with minimum damage to the kernel structure. The sample was tested using a dynamic mechanical analyzer. A miniature three point bending clamp was used to test the samples over a temperature range of 5–110°C at frequencies of 1 and 10 Hz in the moisture content ranging from 11.9–%24.7% (db). The average value of storage modulus (G′) showed a decreasing trend with increase in the moisture content. The corresponding tan δ curve showed a peak representing the glass transition temperature. The kernels exhibited a transition in the range of 20–60°C, which is expected to be the glass transition range. The transition range is consistent with the results of similar studies on cornstarch. It embodies an average of several transition temperatures that arise due to the compositional diversity in the kernel. The individual corn kernels exhibit a wide sample-to-sample variation and structural heterogeneity. A very large number of test replications are required for consistent results. The accuracy and applicability of the dynamic testing in corn is limited due to the rigid sample specifications and testing conditions. Sample preparation and selection of appropriate loading conditions are also addressed.

Keywords:

• Viscoelasticity
• Glass transition
• Corn
• Storage modulus
• Loss modulus

INTRODUCTION

At various stages of grain handling, storage and processing, accurate knowledge of the factors affecting the material (property relationships) can lead to the optimization of the processes and help to minimize the losses and damage to the produce. The versatile applications of corn have motivated researchers to study the behavior and properties of corn over the past few decades. The complex structure and composition of kernel has seldom been considered while exploring the behavioral responses to the fluid transport processes (e.g., drying) and in predicting the stress cracking in corn. The properties of the material are influenced largely by the structure and the constitution of the material. The interactions among the heterogeneous constituents and the plasticizing effect of water present in the pore spaces and as integral part of the matrix affect the material behavior as a whole.

Corn kernel can be classified into three key structural components, namely - germ, endosperm (soft and hard) and pericarp. The kernel constitutes 5–6% pericarp, which is composed of dead cellulosic tubes. These tubes provide capillary interconnections between various cells thus, facilitating water absorption.[1] The corn endosperm constitutes about 82–84% of the kernel dry weight and is formed of 86–89% starch, depending on the variety.[2] The germ contributes about 11% of the kernel dry weight and is mainly composed of proteins and lipids.[3]

Dynamic testing of viscoelastic materials determines storage and loss moduli over short time scale; which provides information of structural changes over a wide range of operating conditions.[4] These tests facilitate the evaluation of the transition temperatures and lead to the determination of the dynamic viscoelastic properties as a function of time and temperature simultaneously.[5] Tests performed at low strain rates do not destroy the microstructure of the material and may provide insights into the micro structural property responses, without destroying the kernel structure.[6]

The Dynamic Mechanical Analyzer (DMA) can easily capture the abrupt change in the properties (storage and loss moduli or tan δ) indicating glass transition. For heterogeneous foods, the glass transition is displayed, not as a sharp but a gradual change in the slope of the graph. It embodies an average of several transition temperatures that arise due to the micro regions.[5] At lower temperatures, the molecular motions of polymer chains are immobilized in random conformations and the material is said to be glassy and brittle.[7] The transition from the glassy to the rubbery phase is also accompanied by increase in the free volume and entropy and a decrease in the viscosity.[6] The thermomechanical characteristics of a viscoelastic material can be determined by measuring the storage or elastic modulus and loss or viscous modulus as a function of temperature, time or frequency.[8] Dynamic mechanical thermal analyzer has been successfully used to measure the viscoelastic properties of soybeans,[6] the thermomechanical properties of oats,[9] and the glass transition in rice kernels.[7] Siebenmorgen et al.[7] developed a protocol for determining the glass transition behavior in brown rice using the DMTA. So far, the viscoelastic behavior in relation to glass transition behavior of corn has not been fully explored. A study on the glass transition in the corn kernels has suggested that corn kernels undergo glass transition similar to rice due to compositional similarity.[10] However, the influence of the glass transition on the viscoelastic properties and subsequent stress development during drying was not considered in any of the previous studies on corn.

The objective of this study was to explore the effect of temperature and moisture content on the dynamic viscoelastic properties and glass transition behavior of the whole corn kernel, with minimum modification of the grain structure.

MATERIALS AND METHODS

Test Materials

The cobs of corn variety Asgrow-715, harvested from the fields of Purdue Agricultural Research Station in October 2005 were hand shelled to cause minimum kernel damage and subsequently shipped to Texas Tech University. The samples were stored in double zip lock bags in the freezer at −4°C for about 3 months prior to testing with the Dynamic Mechanical Analyzer (DMA).

Sample Preparation

Crack free, uniform shaped kernels were handpicked and selected for sample preparation. The selected kernels were sanded using grade 5 commercial sand paper to a flat uniform smooth surface (Fig. 1). Sanding was performed with hand with gentle motion to avoid any stress or mechanical damage to the grain. This method involved scraping off some material from the outer layers of the whole kernels without causing large deformations in the sanded shaped cuboids. The sanded samples were placed in desiccators to obtain the three target levels of moisture content values of 12%, 17%, and 20%. Three saturated salt solutions, Potassium carbonate, Sodium chloride, and Potassium chloride were used to obtain different relative humidity levels of 49, 77, and 86% respectively. These relative humidity values were selected for obtaining target moisture contents.[11] The saturated salt solutions of 510 ml were prepared and kept in desiccators to attain constant temperature and relative humidity before the start of each conditioning experiment. A 15-ml bleach solution was added to the salt solutions to inhibit mold growth. A small tube containing 10-ml toluene was also placed in the center of the desiccators (260 mm in height and 230 mm in diameter) to prevent mold growth. The sanded samples were placed uniformly on the perforated plate. Desiccators were kept in ambient conditions at 25 ± 2°C. For the experiments involving glass transition behavior of anhydrous materials (e.g., soybean, corn, shrimp), it has been recommended that the temperature during conditioning be kept below T_g.[6] Higher temperatures may cause a permanent change in the viscoelastic properties due to the influence of moisture–dependent glass transition. Higher molecular weight biopolymers may undergo transition at higher temperatures (40–100°C) during conditioning.[12]

Figure 1 Picture of corn sample sanded in shape.

The samples were equilibrated for three weeks to reach desired levels of moisture content. The final moisture content levels were 11.9, 17.6, and 19.6% (db). A part of the samples were set aside to be tested as received after harvest and storage with moisture content of 24.7% (db). An average of about twelve kernels of similar shape, were randomly selected from each moisture level. The samples conditioned at each given moisture level were stored in double zip lock bags prior to DMA testing. Before the DMA runs, a part of the sample from the same bag was selected for the determination of moisture using the convection oven method at 103°C for 72 h.[11]

DMA experiments

The samples were tested in bending mode using the miniature three point bending clamp mounted on the DMA (Model Q800, TA Instruments, New Castle, DE) in the temperature range of 5–110°C, at a constant strain of 0.15% (within linear elastic limits) using the temperature ramp mode at the ramp rate of 3°C/min with frequencies of 1 and 10 Hz, respectively.

Each kernel was sprayed with Elmer's craft bond acid free multipurpose spray adhesive (Elmer's Products, Inc. Columbus, OH) to form a thin uniform coat, to control the moisture loss during the test runs. Elmer's craft glue spray, is water resistant glue, which forms an adhesive thin uniform film over the sample surface. No published data was available on the water barrier properties of this glue film. However, preliminary DMA experiments performed with corn kernels coated with this adhesive, showed a loss of less than 8% moisture (based on initial kernel weight) for the highest moisture content sample (24.7%) in comparison with the uncoated sample. This was less than 2% moisture loss when compared with an uncoated sample tested under similar conditions. The adhesive coat also helped to restrict the sample movement over the clamp during the test. During the preliminary runs without this coating, kernels showed a slight slip over the clamp at low normal force in the DMA attachment. The Elmer's glue (Elmer's Products, Inc., Columbus, OH) provided an adhesive coating and a moisture barrier to the sample. A comparative preliminary evaluation of the influence of glue and Vaseline thin film coating on the modulus was also conducted during this study. The results did not show any significant difference in the measured modulus values, but sample with Vaseline coating showed a slight slip at the end of the run.

The DMA was calibrated once a week using the standard procedure recommended by its manufacturer (TA Instruments, New Castle, DE). The DMA unit has a built in furnace for heating. The unit was also connected to a Gas Cooling Accessory (GCA) using liquid nitrogen to rapidly cool the furnace and to obtain the initial sub-ambient temperature (5°C) at the start of every DMA run. Two thermocouples are placed inside the furnace, which help to monitor the furnace temperature and the sample temperature. This DMA unit was fitted with a miniature three-point bending clamp to perform bending type tests on small samples (Fig. 2). About eight replications of bending type tests were performed for different moisture samples. To avoid biases introduced by the testing sequence, the samples were tested in the following random order: 24.7, 17.6, 11.9, and 19.6%. Each sample was mounted with the germ side facing up. The average values of width and thickness of each sample were entered into the DMA software. Initially, the samples were tested using the compression clamp, but it failed to exhibit the expected transition behavior, the hard samples required a large magnitude of force, which was close to the upper maximum limit of the force sensor of the instrument. This problem did not occur with the 3-point bending clamp, which had a wedge shaped sharper contact lines in comparison to the flat compression clamp surface. The sharper wedge surface did not cause any tearing of the corn samples during a given DMA run.

Figure 2 Miniature three-point bending clamp.

During a given DMA run, temperature sweeps were performed while deforming the samples simultaneously at two frequencies (1 Hz, 10 Hz). The DMA applied a sinusoidal deformation corresponding to 0.15% strain to the kernels and measured their force response. A simple stress-strain run was conducted on a sample with this clamp to select the test strain value of 0.15%. For this strain value, the behavior of the material was found to be within the linear viscoelastic range. The deformation and force values generated were converted into storage modulus values by the DMA software supplied by the manufacturer (TA Instruments, New Castle, DE). For all experiments, the temperature was increased from 5°C to 110°C at a rate of 3°C /min.

DISCUSSION

DMA tests with the three point bending application have been successfully used to explore the viscoelastic properties and glass transition behavior of different type of materials, including plant and animal tissues.[13–17] Thermomechanical properties of a paint layer on a composite material have been characterized using a three-point bending DMA clamp with a nearly precise quantification of the material T_g values.[18] There is no clamping of the sample mounted on DMA, which reduces the effect of the instrument on the measurement. Three-point bending is considered one of the most suitable geometries when an accurate modulus is required.[19]

A slight loss in moisture is expected due to increase in temperature during the DMA tests. The exposed area of samples was covered with Elmer's glue (Elmer's Products, Inc. Columbus, OH) to reduce the moisture loss. During the initial trials, the samples with highest moisture content (24.7% db) underwent about 2% decrease in the moisture content after the test when compared to an uncoated sample, which showed a moisture loss of about 8%. In lower moisture content samples, the moisture loss during a DMA run is expected to be less than 2%. Although, moisture loss during the DMA ramp is not desirable, it cannot be avoided. The results presented, are based on the initial sample moisture contents.

Temperature and Moisture Effects

The curves were plotted by taking the average of 4–6 DMA runs with different kernel samples, which were conditioned to attain the same moisture levels. Figure 3 is a typical representation, which shows the variation of storage modulus (G′) values with temperature. Corn samples with 24.7% average moisture content were deformed at two frequencies of 1 Hz and 10 Hz, respectively. In the temperature range of 10°C to 50°C, initially, a rapid decline in the G′ values was observed. The G′ values for samples with moisture content of 11.9, 17.6, and 19.6% decreased up to the temperature of 80, 70, and 60°C, respectively (Fig. 4).However, as heating continued, the rate of change in the G′ value decreased gradually. The samples with 11.9–19.6% moisture content, showed very little change in the G′ values beyond the temperature range of 60–80°C (Fig. 4). Corn kernels at 24.7% moisture content, exhibited the lowest G′ values. Samples with lower magnitudes of G′ were soft and easily deformable. The values of the storage moduli (G′) showed a decreasing trend with increase in temperature. On the application of heat, the molecular motion increases and the polymer becomes soft and flexible.[20]

Figure 3 Plot of Storage Modulus (Mpa) for Corn Samples at 24.7% m.c. tested at 1Hz and 10 Hz.

Figure 4 Plot of Storage Modulus (Mpa) for Corn Samples tested at 11.9–24.7% m.c.

The tan delta curve showed a peak at around 70°C (Fig. 5). This peak corresponds with the temperature at which the lowest G′ value for a given test sample was observed. Plasticization in synthetic polymers is accompanied by a drastic change in the G′ values, and the tan δ peak corresponds with the glass transition temperature. However, heterogeneous multicomponent polymeric systems tend to exhibit a broad range of transition temperatures and a relatively moderate drop in the G′ values.[12, 21–22]. Even though, tan δ peak may not be considered as the true indicator of the transition temperature,[21] but it can be useful in providing some insight into the transition behavior when considered along with the G′ values. Figure 5 shows an inflection in the G′ values and a corresponding peak in tan δ curve. Since, the samples with lower moisture content (11.9%) did not depict a rapid change in the G′ values with increase in temperature, a corresponding tan δ curve can provide useful insights on the transition range.

Figure 5 Plot of Storage Modulus and Tan Delta Vs Temperature (24.7%db).

The rapid decrease in the G′ values with temperature confirmed that the corn kernels were undergoing a transition from the glassy to the rubbery state in the temperature range of 30–60°C (Fig. 4). For all moisture contents, samples tested at 1 Hz showed lower G′ values than those tested at 10 Hz (Fig. 3). This was expected because the elastic component of the viscoelastic materials decreases with decrease in oscillation frequency. In the glass transition zone, the relaxation times for molecular motions are so large that the T_g values may also depend on the time scale of experiment.[4] As the frequency is increased, the T_g shifts to a higher temperature by a few degrees since the polymer chains require more energy to respond to shorter time scale stresses imposed at higher frequencies.[23]

DSC studies by[24] on the corn starch containing 25–50% moisture content (db), detected glass transition in the temperature range of 20–60°C. Studies from[25] reported that the average onset of glass transition in corn flakes occurred at 13°C at 9% moisture content (db). In Fig. 5, the glass transition is displayed, as a gradual change in slope of the plot of G′ and tan δ. It was observed that the average value of the onset of glass transition is around 22°C for the samples at 17.6% moisture content (db), and the transition stage continues to about 59°C (Table 1). These results are comparable to those of Shrogen[24] on corn starch. The effect of the moisture content on the onset and the end of the T_g could not be clearly identified. However, corn kernel, due to its complex composition might be undergoing numerous simultaneous transitions within different regional domains. Studies on corn structure have revealed that the soft endosperm occurs in the center surrounded by hard endosperm. The cellular structure of the floury endosperm constitutes of tiny air pockets surrounding the starch granules, which are embedded in the protein matrix. This matrix shrinks and collapses on moisture removal (i.e., at lower moisture). The protein matrix in the horny endosperm is tightly packed and remains intact. Corn contains about 4.3% lipids, which are mainly stored in the germ.[1, 26] Lipids along with moisture contribute towards the viscous component in the overall viscoelastic behavior of seeds,[27] the elastic component or the storage modulus values may be compensated in presence of fats and protein viscous contributions to the overall behavior.

Table 1 Moisture content and Tg value at onset and end

Moisture content (% db)	Average temperature at Tg onset	Average temperature at Tg end
11.9	25.7	58.4
17.6	21.9	59.3
19.6	21.2	63.8
24.7	23.0	54.5

The mechanical profile of the glass to rubber transition is said to be a strong function of the polysaccharide profile. If the polymer is network forming, then while considering the influence of polymer on rheology, it has been suggested that a “network T_g” value may have greater significance than the “calorimetric T_g” value.[28] This transition is observed as a range, which exhibits the combined effect of several transitions. The moisture content does not seem to have a clear effect on the onset of the transition, but the range of the transition temperatures increased with the decrease in the moisture. This indicates that moisture plays a significant role in defining the transition range of a material, resulting in numerous processing applications. Studies have indicated that the polymers exhibit a change in the heat capacity, diffusivity, expansion and contraction coefficients and viscoelastic properties during glass transition.[10, 24, 29–31] The onset of the glass transition started at about 20°C and continued until 80, 100, and 110°C corresponding to the moisture content of 24.6%, 19.6%, and 17.6%, respectively. The kernel softens with the increase in temperature. The presence of moisture facilitates the softening of the material thus ending the transition at a lower temperature as compared to the samples having lesser moisture.

Effect of Moisture on Average G′ Values

Figure 4 shows the effect of moisture on average G′ values. The G′ values were greater at 11.9% (db) moisture content than at other moisture levels. The data showed that in general the G′ decreased with an increase in moisture content. The G′ value for 11.9% moisture content kernels was about two to three times greater than the G′ value for higher moisture contents. The G′ values towards the beginning of the test were found to be 2 to 5 times larger than the G′ values at the end of the test (Fig 4). The loss in elastic modulus for samples with higher moisture content is higher; this is due to the plasticization effect of the moisture. This is consistent with literature studies on the viscoelastic properties of corn horny endosperm with a moisture content of 14.50–24.75% (db) at frequency of 1Hz conducted by.[32] Studies conducted on soybeans by[6] also yielded similar observations. The absolute value of modulus varied from 630 MPa to 188 MPa in the study. There was an appreciable drop in the modulus at higher moisture levels (16–17% db). Studies on viscoelastic properties of oats over the temperature range of -20–100 °C showed similar trends.[9]

When a sample is heated, differential moisture zones develop inside the grain. The surface moisture is lost quickly, the outer hard endosperm layers shrink and compact, forming a hard crust. The loosely packed cells in the floury endosperm have starch granules, which interact with the intercellular moisture present resulting in softening of the tissue, while the starch cells in the hard endosperm being very tightly packed may not be plasticized at lower moistures. This accounts for the higher moduli values for these samples. The hard endosperm, being stiff, mainly contributes to the storage modulus in the overall viscoelastic behavior. Experiments with a very large number of replications and precisely ground samples are needed to clearly substantiate the average G′ values. The loading conditions, type of clamp, the sample orientation, geometry and structural composition of the individual kernel are some of the factors that may have a significant effect on the average G′ values.

Implications on Processing

The glass transition temperature, T_g determination and its effect on the viscoelastic properties can become a key factor in optimizing the processing techniques. In preparation and storage of fortified products, glass transition is an important criterion in understanding nutrient stability and can provide additional insight into the factors affecting degradation during storage.[33] The kernel structure and geometry renders itself to the formation of transition zones that traverse across the matrix when subjected to thermal treatments. Studies on soybeans have shown that it exists in the glassy state below 10% moisture content.[34–36] Heating induces polymer relaxations within the microscopic domains at the cytoplasmic level and provides sufficient energy to cause some of the water to evaporate and escape from the regional domains within the kernel, which induces shrinkage. Glass transition mapping for rice kernels showed a clear relation between the transition and the stress zones inside the kernels.[37] At the cytoplasmic level, the polymers try to regain their original configuration; this may lead to a series of stress concentration zones to build up at the mesostructural level. The presence of insufficient moisture at the time of cooling may lead to an increased number of stress zones within the kernel.

During drying, the moisture transport exhibits Darcian (or Fickian) character in the glassy and rubbery states and non-Darcian (or non- Fickian) behavior near the glass transition zone.[38] Since, corn kernels exhibited glass-transition in a wide range of temperature ranging from 21.2–63.8°C and moisture contents, 11.9–24.7%db (Table 1); they are expected to exhibit non-Darcian moisture flow behavior during drying. During drying at a given temperature, the surface of the corn kernel may exist in a glassy state, with the center in a rubbery and the middle region in the transition state. A high moisture gradient and large magnitude of G′ in the glassy state may produce large viscoelastic stresses near the surface. Large contrasts in the viscoelastic properties within the kernel in the glassy and rubbery states due to glass transition may be responsible for quality changes during drying.[7, 37] Viscoelastic properties are affected by the polymer melting, change in the crystalline structure, composition and relaxation of long chain polymers or due to the plasticizing effect of water. The formation of a glassy crust near the surface may hinder further moisture removal from inside. The sudden loss of moisture during drying may cause the collapse of the biopolymer matrix producing compressive stresses inside the kernel. The moisture lost from the inner cells might be entrapped inside the kernel, bounded by the harder outer crust before it forces its way out. This may cause additional stress build up. The cumulative effect may lead to the formation of cracks in the kernel. NMR imaging studies on the drying of the kernel may provide additional insights on the moisture loss pattern and the buildup of the differential moisture zones inside the kernel.

Suggestions for Future DMA Studies with Corn

1. The structural heterogeneity of the corn kernels poses a serious limitation on the sample preparation method - no two kernels are exactly alike and the germ contributes a significant portion to the structure. However, it was noted that even in two similar shaped kernels, there is a variation of the germ size and position within each kernel. A way to minimize the variation could be hand shelling the kernels located at a specific location on the cob. This might reduce variation in sample properties to some extent.

2. To reduce sample–to–sample variation of the measured properties, a sample holding device and a grinding machine are required that would make the opposite faces of the samples parallel without damaging the kernel. Hand sanding method is precise but tedious and limited to smaller sample sizes. For large number of replications an automated device would be better suited.

3. Samples should be tested over a larger range of moisture starting from 9% (db) up to 25% (db). This wide range will help in clearly establishing the effect of moisture on the transition behavior.

4. The samples with higher moisture contents tend to become very soft at higher temperatures; therefore, the test strain should be selected carefully to ensure that the sample is being tested within linear elastic limit throughout the duration of the test.

CONCLUSIONS

Corn is a hetero polymer, composed of proteins, starch (amylose and amylopectin) and cellulose layer as pericarp, which may undergo transition at different temperatures. In the moisture content range of 11.9–24.7%, a two to three times decrease in G′ was observed due to increase in temperature from 10°C up to 70–80°C. Moisture played a significant role in heating induced material softness. The tan δ curves showed a peak in the temperature range of 40–60°C. The peak shifted by 2–3°C to a higher temperature upon increase in frequency from 1 Hz to 10 Hz, which is consistent with the material undergoing glass transition in this temperature range. The G′ value is directly influenced by the test frequency, showing a higher average G′ value corresponding to 10Hz test frequency irrespective of the moisture content of the sample. The specimen size requirements, appropriate loading conditions, loss of moisture during testing are some of the limitations that should be appropriately addressed while using the DMTA. Glass transition behavior of individual kernel components may provide useful insights on how drying rates and corn quality are related.

NOMENCLATURE

DMA	=	Dynamic mechanical analyzer
DSC	=	Differential scanning calorimeter
ESR	=	Electron–spin resonance
G′	=	Elastic (or storage) modulus (MPa)
G″	=	Loss modulus (MPa)
NMR	=	Nuclear magnetic resonance
tanδ	=	Ratio of loss modulus to storage modulus (= G″/G′) (unit less)
Tg	=	Glass transition temperature (°C)

ACKNOWLEDGMENTS

Thanks to USDA-CSREES for providing the financial support under the award number 2003-35503-13963. Thanks to Purdue University's grain laboratory for providing the corn samples.