- Email: [email protected]
New insights into cooked rice quality by measuring modulus, adhesion and cohesion at the level of an individual rice grain
Journal of Food Engineering 240 (2019) 21–28
Contents lists available at ScienceDirect
Journal of Food Engineering journal homepage: www.elsevier.com/locate/jfoodeng
New insights into cooked rice quality by measuring modulus, adhesion and cohesion at the level of an individual rice grain
T
L. Yua, T. Witta, M. Rincon Bonillab,1, M.S. Turnera,c, M. Fitzgeralda, J.R. Stokesb,∗ a
School of Agriculture and Food Sciences, The University of Queensland, Brisbane, 4072, Queensland, Australia School of Chemical Engineering, The University of Queensland, Brisbane, 4072, Queensland, Australia c Queensland Alliance for Agriculture and Food Innovation, The University of Queensland, Brisbane, 4072, Queensland, Australia b
A R T I C LE I N FO
A B S T R A C T
Keywords: Rice Cooked rice Mechanics Adhesion Cohesion Modulus Processing
Causal relationships between physical properties and structure/composition of cooked rice are difficult to quantify when mechanical measurements are performed on bulk samples using large deformations that alter the structure irreversibly. We demonstrate here methods involving small-deformation to characterise the elastic modulus (E), adhesion and cohesion at the individual grain level, and show distinct differences between freshly cooked rice and shelf-stable retorted rice. On average, retorted rice is harder and less adhesive and cohesive than freshly cooked rice, but their distributions in each of these mechanical properties overlap. E is independent of adhesion and weakly correlated with moisture content. In addition, a ring-shear tester is shown to distinguish the bulk cohesion and flowing ability between rice samples. Measuring the inherent physical properties of individual grains has the potential to enable a more sensitive evaluation of new processes and grain varieties, and development of quantitative structure-property-processing relationships for rational design of products to perform optimally at different stages, from manufacturing through to oral processing.
1. Introduction The desire for processed food to have the character of freshly prepared food is driving manufacturers to explore new processing methods and/or formulations. The distribution of ready-to-eat rice in shelfstable, chilled, and frozen forms is particularly desirable given its one of the most consumed staple foods in the world. In designing and evaluating innovative new processing methods to achieve this goal, such as high pressure processing (Yu et al., 2017), there is a need for reliable and sensitive measurements of the physical properties of cooked rice and its structure. Specifically, the mechanical and surface properties of cooked rice are anticipated to contribute to their flow behaviour during processing and consumption. Causal structure-property-processing relationships provide a means in which to rationally evaluate and optimise the specific effect of process and formulation variables. To predict the sensory perception of texture, and to assist in the process of designing a particular structure of food, instrumental procedures have been designed to measure texture-relevant physical properties. However, these are usually based on imitative techniques such as texture profile analysis (TPA), which have limited sensitivity and do not
measure inherent material properties of cooked rice. In this paper, we aim to demonstrate new approaches to evaluate the material properties of cooked rice grains based on measuring the elastic modulus, adhesion and cohesion at the level of individual rice grains. For cooked rice, uniaxial compression tests (including those embedded within TPA routines) as well as others such as puncture tests or extrusion tests, have been used to evaluate how the measured mechanical properties of cooked rice relate to the sensory perception of texture (Li et al., 2016; Lyon et al., 2000; Meullenet et al., 1998; Perez et al., 1993). Li et al. (2016) reported that 80% compression in TPA measured on 1 g of rice placed as a single layer can differentiate between sticky and non-sticky rice. An observed correlation between TPA and sensory measure of ‘hardness’ and ‘stickiness’ largely depends on the significant difference in amylose content in waxy (no amylose) and high amylose (ca. 30% of the starch content) rices. For rice varieties with an amylose content between 18 and 26%, they observed that the hardness measures no longer correlate while the stickiness correlation becomes weaker. Weak correlations between instrumental TPA-assessment of hardness, adhesiveness and cohesiveness with sensory hardness, adhesiveness and cohesiveness of mass were also reported by Lyon
∗
Corresponding author. E-mail addresses: [email protected] (T. Witt), [email protected] (M. Rincon Bonilla), [email protected] (M.S. Turner), m.fi[email protected] (M. Fitzgerald), [email protected] (J.R. Stokes). 1 Current address: Basque Centre for Applied Mathematics, Alameda de Mazarredo 14 (48009) Bilbao, Bizkaia, Spain. https://doi.org/10.1016/j.jfoodeng.2018.07.010 Received 17 July 2017; Received in revised form 30 April 2018; Accepted 6 July 2018 Available online 09 July 2018 0260-8774/ © 2018 Published by Elsevier Ltd.
Journal of Food Engineering 240 (2019) 21–28
L. Yu et al.
performed by confining a sample to a certain extent between two surfaces, and then the surfaces are separated at a set speed. Adhesion is reported as either the maximum negative force required to detach the rice grain from the device surface, or the negative area under the forcedistance or force-time curve (Fiszman and DamÁSio, 2000; Okabe, 1979). These differences in definition, as well as choice of extent of confinement and detachment speed, make comparison of literature values difficult. Thus a constant definition and a clear interpretation of how each physical property relates to each sensory property would be beneficial, although it is not well known if the different definitions actually lead to a different conclusion. The physical measurement of adhesion of individual grains to a surface is recommended because Mossman et al. (1983) showed that measuring 40 grains resulted in less variation in adhesion than measuring an individual grain. Such homogeneity may lead to inconclusive results when comparing differences between rices as a function of process variables or rice variety. To evaluate the sensory adhesion of rice to surfaces such as to a spoon, molars or the lips, it is relevant to measure the surface properties of cooked rice grains (Okadome et al., 1999). Large strain deformation tests that are used in many studies will alter the integrity of the rice and the measured adhesion will not necessarily be related to its surface properties. The purpose of this study is to overcome some of the limitations described above by developing a more sensitive measure of the mechanical and surface properties of cooked rice grains. Individual rice grains of the same rice variety cooked in a rice cooker or retorted are measured in a small strain uniaxial compression test to obtain objective and inherent physical and surface properties of individual grains. The elastic modulus is obtained within the elastic limit of the rice grains. The two common ways of analysing rice adhesion are compared to evaluate which is more suitable to differentiate between the two rice samples when they are in contact with a metal surface. We also seek to evaluate cohesion as a surface property between two individual rice gains, and compare this to ‘bulk’ values obtained using a ring-shear tester. Whilst the ring-shear tester is normally used to measure flowability and cohesion in powders (Schulze, 2008), including powdered food (Iqbal and Fitzpatrick, 2006), it was recently used to provide insights into these properties in hydrated semi-solid foods (Tobin et al., 2017).
et al. (2000) for different rice varieties. The authors highlight the need for more sensitive instrumental mechanical methods to capture the subtle, yet detectable differences in sensory texture between samples (Lyon et al., 2000). We note in particular that both the extrusion test and the TPA approach commonly measure the mechanical response of cooked rice grains while they are being destroyed during exposure to high compression. In this case, the rice is transformed dramatically, well-beyond its elastic limit, as it is compressed into a ‘single’ bulk soft material constructed of all components of the grain. This transformed product bears little resemblance to the original rice grains, and so it is perhaps unsurprising that measurements obtained on the rice in this state do not strongly relate to adhesiveness and cohesiveness of rice that we would hypothesise to be closely related to the surface properties of the grain. Despite the wide utilisation of uniaxial compression tests to evaluate texture-related mechanical properties of cooked rice and other food systems, the sample preparation, sample size and amount, plunger size, compression speed, and degree of compression vary highly between study groups. This renders the results difficult to compare, particularly when the inherent material properties (e.g. modulus) are not extracted from measured data (Champagne et al., 1999; Meullenet et al., 2001; Mossman et al., 1983; Okadome et al., 1999). Due to such variability in parameters combined with the tendency to perform measurements at high compressive strains, where rice is transformed to a new structural state, such mechanical measurements do not necessarily provide information about the inherent physical and surface properties of the rice. This limits the establishment of a unifying structure-property relationship that is relevant to a full spectrum of sensory texture attributes. The major limitations of current uniaxial compression methodologies are the poor definition of the contact area, the low likelihood of capturing the true variation of the sample, and the measurement of irrelevant sensory properties. These all need to be overcome in order to predict relevant sensory texture properties accurately. The arrangement of some number of rice grains in a single layer equidistant from each other is a commonly practiced experimental method (Boluda-Aguilar et al., 2013; Li et al., 2016; Meullenet et al., 2001; Patindol et al., 2010). However rice kernels are not flat, so the contact area with the plunger increases as the compression progresses (Tsuji, 1981). This means that it is important for all samples to have the same geometry when submitted to a compression test because stress is defined as the force divided by contact area (Bourne, 2002). A difference in the contact area will lead to a difference in the measured stress, even when the modulus is the same. Shaping each sample into the same geometry is applied to other soft foods such as cheddar cheese. In this case, good predictions of sensory springiness, cohesiveness, cohesiveness of mass, roughness of mass and tooth-pull can be obtained with TPA (Moiny et al., 2002). As well established in material science, differences in forces may be due to differences in contact area and not inherent differences in modulus and fracture properties. When a layer of rice grains with a distribution of radii are compressed, the plunger makes contact with the rice grain of the largest diameter first, so the reactive force is an ambiguous sum of rice grains at differing levels of compression. This is problematic as a distribution of hard and soft grains under compression can easily lead to measurements that correspond to hard rice grains even when most grains are actually soft. Due to the non-uniformity between grains, a better approach to investigate intrinsic properties relevant to sensory perception, as recommended by Szczesniak and Hall (1975), would be to measure the distribution of hardness by performing measurements on large numbers of individual grains. There are a number of different terms used to describe stickiness or adhesiveness in the literature; here we define adhesion as the stickiness between the rice grain and the surface, and cohesion as the stickiness between individual rice grains. Adhesion is a measure of force, work or energy required to detach rice grains from a surface but the actual definition of the measurement varies. Typically, a “pull-off” test is
2. Material and methods 2.1. Material The samples used (variety Langi) were grown in the Riverina, NSW, Australia and harvested in 2014. The grains were provided after milling. The retorted rice of the same variety and same batch was kindly provided by SunRice and contained rice, water, 2.5% vegetable oil and 0.5% distilled monoglyceride. The chemical composition of raw rice was provided by SunRice showing a total starch content of 76.9% with 17.8% amylose and a moisture content of 12.1%. 2.2. Sample preparation For freshly cooked rice, 300 g of milled rice was weighed into a rice cooker bowl and washed three times with 1.5 L of deionised water for 10 s. After each washing step, the rice was drained in a strainer for 10 s. Milled rice was cooked with a water-to-rice ratio of 1.75 in an electric rice cooker (Tefal intelligent rice cooker, Tefal S.A.S, Rumilly, France). The cooking time was 45 min. After cooking, the outer and bottom 1 cm of rice was discarded. The rest was mixed gently with a plastic paddle to distribute the moisture homogenously, then rice was kept warm for another 10 min in the cooker. Before instrumental analysis, retorted rice samples were warmed in the pouch according to the recommendation on the package in a microwave at 900 W for 90 s. All samples were cooled to approximately 40 °C and kept in a water bath 22
Journal of Food Engineering 240 (2019) 21–28
L. Yu et al.
set to 43 °C during measurements to retard retrogradation. 2.3. Moisture content To measure moisture content 10 ± 0.1 g of each rice sample was weighed, then dried at 103 °C for 24 h and weighed again (n = 4). To investigate the correlation between elastic modulus and moisture content, 20 individual grains were dried after the compression test at the same oven conditions. Immediately after drying, the samples were weighed again. The moisture content on a wet basis was calculated as the difference in water weight divided by the original sample weight. Fig. 1. An example of model fitting (red line) of the experimental data (blue dots) within the linear viscoelastic region of a force-indentation curve. The fitting curves have an R2 value ≥ 0.995.
2.4. Rice grain size Rice grains were assumed to be ellipsoids, and the length and diameter were measured with a calliper (SUC23-200, ProSciTech, Kirwan, Australia). Sixty random grains from 3 preparations of freshly cooked and 3 packages of retorted rice, respectively, were measured and the average is reported. The aspect ratio (α) of the rice grains was calculated as the average length (R) divided by the average diameter (R’). Rice grains with obvious defects were not measured.
where b is the minor axis of the elliptical contact, and is related to the estimated contact area by (5)
A = πα 4/3b2 Combining equation (1), (3) and (4), produced equation (6)
4 E R β δ 1.5 3 (1 − ν 2)
2.5. Mechanical and surface properties of rice
F=
To develop a method to measure the mechanics of single grains, freshly cooked rice was compared to warmed, retorted rice. To measure elastic modulus (E), individual rice grains were uniaxially compressed with a smooth parallel plate (40 mm diameter) on a stress-controlled rheometer (AR-G2, TA Instruments Ltd., UK). A set compression speed of 10 μm/s was applied until the normal force reached 1.0 ± 0.01 N, then the plates were separated at a speed of 20 μm/s. The normal force of 1.0 N was chosen following extensive preliminary tests that found that for 90% of the cooked rice grains measured, compression was within the linear viscoelastic region, and thus was before the grains plastically deformed or fractured. To measure E and adhesion, intact grains were randomly taken from at least 3 different preparations of freshly cooked and 3 different packages of retorted rice.
where beta in equation (7) represents the contribution from the aspect ratio 1/3
β=
2 21.5α 2 ⎛α + 1⎞ + 1)1.5 ⎝ 2α 4 ⎠
(α 2/3
⎜
⎟
(7)
and becomes unity for a circular contact area. Fig. 1 shows force against indentation (δ 1.5) data for a particular rice grain. E was extracted from the slope of the force F vs. the indentation δ 1.5 curve fitted with MATLAB (R2014b, The MathWorks Inc.) and by manually choosing the zero point; this is the point where the plate first came into contact with the grain, and the force began to increase. In order to eliminate error due to uncertainty in the initial gap, a section of the curve with a minimum of 10 points was selected for fitting in such a way that r2 was always above 0.995.
2.5.1. Elastic modulus E The elastic modulus E for individual rice grains was estimated using the Hertzian contact model for compression of a spherical body with a flat punch (Sneddon, 1965). However, since the grains are ellipsoidal, the correction introduced by Greenwood (1997) for bodies where the contact shape is an ellipse, was employed. Here, an equivalent sphere radius Re is calculated in order to preserve the relationship between force F and contact area A provided by the Hertzian model, i.e., equation (1)
3FR e 2/3 ⎞ A = π⎛ ⎝ 4E ∗ ⎠
(6)
2.5.2. Adhesive stress and adhesion The two methods for quantifying adhesion from a “pull-off” test are evaluated in Fig. 2. In this study, we define adhesive stress as the absolute of the maximum negative force for the detachment curve, divided by the contact area between the rice grain and the plate. The estimated area is calculated using equation (4); note this is an ‘engineering’ stress because it does not account for the increase in contact area that arises during compression. Adhesion is defined here as the adhesive energy, which is the area under the negative curve divided by the contact area between rice grain and plate. The starting point to
(1)
where
E ∗ = E/(1 − ν 2)
(2)
with E as the elastic modulus and ν the Poisson's ratio. Incompressibility is assumed (ν = 0.5) because of the high water content of the grain. Re is a function of the radii of curvature and can be expressed as a function of the aspect ratio α according to equation (2) 1/3
2α 4 ⎞ Re = R ⎛ 2 ⎝α + 1⎠ ⎜
⎟
(3)
whereby R is half the thickness of the rice grain (i.e., the minor radius of the ellipsoid). To estimate force from indentation δ, an additional correction was required (Greenwood, 1997) using equation (3)
b2 2/3 δ= (α + 1) 2R
Fig. 2. An example of calculating the area under the negative curve as adhesive energy (grey) and the maximum negative force (black line) indicated by the red arrow.
(4) 23
Journal of Food Engineering 240 (2019) 21–28
L. Yu et al.
2.7. Statistical analysis Statistical differences were calculated with GraphPad Prism Version 6 (GraphPad Software, Inc., La Jolla, USA) with t-test and KolmogorovSmirnov test (non-parametric, two-tailed, heteroscedastic). Due to a high percentage of zero values in adhesion, cohesion and cohesive stress especially for retorted rice, the statistical difference was calculated for the population greater than zero and the average is presented with standard deviation. Statistical power analysis, which is applied to determine the minimum sample size required to detect an effect of a given size with a given degree of confidence, was calculated in R studio (R studio, Boston, US). Pearson correlations were calculated in Microsoft Excel.
Fig. 3. (a). An image of the experimental setup used to measure cohesive stress between two perpendicular single rice grains compressed against each other; (b): Schematic to show the calculated area between two perpendicular single grains in grey.
integrate the area was automatically determined when the mean of six consecutive data points fell below 2.5 times the standard deviation of the mean of these six points, and the integration stopped when the curve reached the mean again. The noise range for zero compression force is 0.01 N, thus values < 0.01N were regarded as equal to zero and therefore not adhesive.
3. Results 3.1. Size distribution of grains The length of freshly cooked rice grains varies between 8.8 and 12 mm with an average of 10 ± 0.7 mm. The freshly cooked grains are significantly longer (p < 0.0001) than the retorted rice which vary from 7.2 to 9.7 mm, with an average of 8.3 ± 0.6 mm. However, the diameter is not significantly different between the two types of grain. The average aspect ratio α is 4.0 ± 0.4 for freshly cooked rice and 3.3 ± 0.3 for retorted rice.
2.5.3. Cohesive stress and cohesion Cohesive stress and cohesion between two individual grains are compared to determine which method characterizes cohesion better. A schematic setup is shown in Fig. 3a and b. The cohesive stress is defined as the maximum negative force between two rice grains that are separated from each other at a right angle after compression force reached 0.1 N, divided by the contact area between the two rice grains. To calculate the contact area between two grains, the gap at 0.1 N compression is assumed to be the diameter of both grains. For simplification the two grains are assumed to have the same diameter. The two grains are perpendicular, so the contact area is assumed to be the square of the diameter, which slightly overestimates the contact area. Cohesion is defined here as the cohesive energy, calculated in the same way as for adhesion, divided by the contact area between two rice grains. The force of 0.1 N was chosen because the surfaces of adjacent grains were already touching and a higher force fractured the grains. The upper plate was not fixed, thus slight rotation upon contact between the grains cannot be excluded. The noise range for zero compression force is 0.01 N, thus values < 0.01N were regarded as equal to zero and therefore not cohesive.
3.2. Mechanical properties 3.2.1. Elastic modulus E The distribution of the elastic modulus (E) of freshly cooked rice is narrower and has a lower average than retorted rice (p < 0.0001) (Fig. 5). For freshly cooked rice, E ranges from 0.3 to 1.5 MPa and the average is 0.89 ± 0.26 MPa, while for retorted rice, E ranges from 0.62 to 8.2 MPa with an average of 1.4 ± 1 MPa. Both grains show a skewed distribution of elastic moduli, with a greater number of soft grains than hard grains. 3.2.2. Elastic modulus and moisture content A reasonable hypothesis for the broad distribution of E is that it arises from an inhomogeneous moisture distribution within a batch, thus the moisture content of 20 individual grains was measured after compression to investigate whether E is dependent on the moisture content. Fig. 6 shows that the moisture content of freshly cooked rice ranges between 57% and 66% and is narrower than the moisture content of retorted rice grains, which varies between 39% and 62%.
2.6. Cohesion of bulk rice grains Cohesion of bulk rice was measured with a Ring Shear Tester (RSTXS, Dietmar Schulze, Wolfenbuettel, Germany) following the methods described in Schulze (2008). The XS-Lr shear cell was filled, weighed and then placed on the motor-driven base of the Ring Shear Tester. 47 ± 1.0 g of bulk rice is used that corresponds to a volume of 72.99 cm3. The sample were pre-sheared under a consolidation force of 1000 N until a steady-state was reached, and then sheared until incipient flow was attained under normal forces of 200 N, 320 N, 440 N, 560 N, 680 N and 800 N at a speed of 75 mm/min. All rice samples were measured in triplicate with each sample measurement involving duplicate measurements of each normal force. The cohesion was the intercept of the yield loci from a linear fit of the shear force required to induce incipient flow with the applied normal forces; the cohesion is defined as the shear stress required to cause flow in the material under zero normal force load (Schulze, 2008; Teunou et al., 1999). An example of this is shown in Fig. 4 with the cohesions of 214 Pa and 464 Pa measured for retorted and freshly cooked rice respectively. The slope of ca. 0.6 was similar for both samples and is generally related to the angle of internal friction; since it was not differentiating between samples, we do not consider it further.
3.2.3. Adhesive stress and adhesion Fig. 7a and b shows the distributions of the adhesive stress and adhesion of individual cooked and retorted rice grains, respectively. Adhesive stress between freshly cooked rice and the stainless steel plate ranges from 8.2 to 53 mPa with an average of 25 ± 9 mPa, and for retorted rice from 0.8 to 45 mPa with an average of 11 ± 8 mPa. Adhesion is below the detection limit with adhesion being zero for 7% of freshly cooked rice grains and 33% of retorted rice, therefore the average and standard deviation reported here represent the population with adhesion greater than zero. Adhesion between freshly cooked rice grain and the stainless steel surface ranges from 0 to 9.5 J/m2 with an average of 4.2 ± 2 J/m2 and for retorted rice from 0 to 2.0 J/m2 with an average of 0.7 ± 0.6 J/m2. The results show that freshly cooked rice is significantly more adhesive to a stainless steel plate than retorted rice (p < 0.0001), and that most of the retorted rice grains are essentially non-adhesive. 3.2.4. Cohesive stress and cohesion Cohesive stress is below detection limit for 10% of freshly cooked rice grains and for 65% of retorted rice with cohesive stress being zero; 24
Journal of Food Engineering 240 (2019) 21–28
L. Yu et al.
Shear Stress (Pa)
900
y = 0.617x + 464
700 500
y = 0.627x + 218
300 100 0
200
400 600 Normal Stress (Pa)
800
Fig. 4. An example of calculating the bulk cohesion as the intercept of the yield locus on the shear stress axis (y-axis) of freshly cooked (■) and warmed, retorted rice (●).
Fig. 7. (a). The distribution of adhesive stresses between plate and rice of freshly cooked rice (■) (n = 60) and warmed, retorted rice (■) (n = 60). This shows that freshly cooked rice grains are significantly more adhesive than retorted rice grains (p < 0.0001). (b) The distribution of adhesive energy between plate and rice surface of freshly cooked rice (■) (n = 60) and warmed, retorted rice (■) (n = 60). This shows that the freshly cooked rice grains have a broader distribution and are significantly more adhesive than retorted rice grains (p < 0.0001).
Fig. 5. The distribution of elastic moduli of individual grains of freshly cooked rice (■) (n = 60) and warmed, retorted rice (■) (n = 60). The mean elastic modulus of freshly cooked rice is significantly lower than retorted rice (p < 0.0001).
Fig. 6. A plot of the negative correlation between moisture content and elastic modulus of warmed, retorted rice (●) and freshly cooked rice (■) (n = 20), R2 = 0.6637.
Fig. 7. (continued)
and cohesion is below detection limit for 35% of retorted rice grains with cohesion being zero. Therefore, the average and standard deviation reported here represent the population with cohesive stress or cohesion greater than zero. The cohesion for the population of rice granule pairs is shown in Fig. 8. Pairs of freshly cooked and retorted rice grains cohesive stress (4040 ± 1260 Pa) and cohesion (0.84 ± 0.58 J/m2) are significantly higher (p < 0.05, p < 0.0001) between two freshly cooked rice grains than two retorted rice grains (cohesive stress 2770 ± 650 Pa, cohesion 0.080 ± 0.09 J/m2). A broader distribution is observed for the freshly cooked than retorted rice, which is accompanied by a larger distribution of contact areas between two freshly cooked grains; from 2.7 to 5 mm2 for freshly cooked rice pairs compared to 3.5–4.9 mm2 for retorted rice grain pairs.
Another approach to understand cohesion is to use a method which provides an average of the bulk interactions and flowing ability of the rice grains, one such method is to use a ring shear tester. Similar to the trend observed in the individual grain cohesion results, the bulk cohesion of freshly cooked rice at 470 ± 34 Pa is significantly greater than for retorted rice at 206 ± 17 Pa (p < 0.0001). This technique measures the interaction between grains as well as between grains and shear cell wall, and assesses how sticky the bulk is.
4. Discussion Measuring the inherent physical properties at the level of individual rice grains can be applied to characterise and compare the mechanical 25
Journal of Food Engineering 240 (2019) 21–28
L. Yu et al.
60
Percentage
50 40 30 20 10 0 0
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
1
1.1 1.2 1.3 1.4 >1.5
Cohesion (J/m2) Fig. 8. The distribution of cohesion between rice grains pairs of freshly cooked rice (■) (n = 60) and warmed, retorted rice (■) (n = 60). This shows that the freshly cooked rice grains have a broader distribution and are significantly more cohesive than retorted rice grains (p < 0.0001).
properties as a function of processing variables, as demonstrated here, but also between different rice varieties and their composition. Measuring material functions of individual grains, including E, adhesion and cohesion at low deformation (i.e. prior to plastic deformation), ensures that valid comparisons can be made between results obtained from different studies and research groups. The power of performing measurements on individual grains, or grain pairs is demonstrated by analysing the interrelationship between variables and leading to information on the structural or compositional drivers for particular physical properties. Fig. 9. A plot of adhesion and elastic modulus for individual grains of freshly cooked rice (■) (n = 60) and warmed, retorted rice (●) (n = 60). Neither rice grain population displayed a correlation.
4.1. Mechanics The wider distributions for E of retorted rice indicates a greater heterogeneity in the retorted rice grains, which is possibly due to the difference in processing. Skewed distribution of rice grains was reported by Perez et al. (1996), who recorded the maximum penetration force of close to 100 individual cooked rice grains in a puncture test to compare rice varieties with similar physicochemical properties, and the distributions produced display the same skew toward soft grains as seen in the current study. The wider range of moisture content of retorted rice matches with the wider range of E and is another indicator that moisture affects E. A weak, negative correlation between E and moisture content is evident in the joint population of both samples. A similar result using 15 different rice varieties was reported by Perez et al. (1996) with the maximum penetration force of cooked, individual rice grains with an average moisture content of 69% exhibiting a higher penetration force than rice with an average moisture content of 75% (Perez et al., 1996). Many studies have correlated amylose, amylopectin or protein content to cooked rice hardness (Sowbhagya et al., 1987; Ramesh et al., 1999; Jane et al., 1999; Okadome et al., 1999), however, only one rice variety was used in our study, thus any effect of compositional differences on hardness between rice varieties can be ruled out. The weak, negative correlation between moisture content and E in freshly cooked and retorted rice grains indicates that the E of individual grains may also be dependent on other properties such as, for example, moisture distribution within the grain, starch crystallinity, hollows that occur during cooking inside the grain, and the grain morphology. On average, the freshly cooked rice is softer, has a higher moisture content and is more adhesive than retorted rice. We therefore evaluate whether E is related to adhesion in Fig. 9, but find no correlation between them. That is, for individual grains of the same modulus, freshly cooked rice is much more adhesive than retorted rice. This is an important finding; the measured adhesion is likely to depend on a surface layer coating the bulk grain. The results suggest that it is this property that is largely affected by the difference in processing, and this is likely
to have a major effect on the behaviour of the rice in processing and during consumer use and consumption. We should point out that the adhesion and E measure here are essentially intrinsic material properties of the system whereby they are determined at low deformation. The current experiments are thus more likely to describe properties related to rice textural attributes which are prior to chewing the rice into a bolus, and thus the measurements do not represent the maximum force and adhesiveness as measured using TPA which may be negatively correlated as reported by Tsuji (1981). 4.2. Adhesion and cohesion Adhesion results are less broadly distributed than the adhesive stress results, and provide a clearer distinction between the two samples. The ability differentiate more clearly between adhesive and nonadhesive samples leads to our recommendation to use adhesion in comparison between samples rather than adhesive stress. To evaluate rice adhesion to a surface such as stainless steel, we recommend considering the contact area upon compression. This is further supported by Brenner and Nishinari (2014), who state that the dependence of adhesion on the contact area between sample and the plunger after compression needs to be accounted for, but is not normally performed in TPA. The small strain compression used in this study distinguishes between the adhesion property of the grain surface rather than following destruction of the sample that is considered in TPA. Surface adhesion is useful in a sensory study that evaluates how rice grains adhere to different surfaces such as a container, spoon (Windham et al., 1997), lips (Lyon et al., 2000) or molars, but is less likely to be relevant to chewed samples. For the first time, the cohesion between rice grains as an inherent surface property was measured. It should be noted that the cohesion between two perpendicular rice grains measures the minimum contact 26
Journal of Food Engineering 240 (2019) 21–28
L. Yu et al.
grains often does not provide an unambiguous result of the population. We also found that the difference in E was only partially dependent on moisture content. In addition, measuring individual grains has uniquely uncovered that a major difference between freshly cooked rice and retorted rice is that the latter in predominantly non-adhesive, yet there is no dependence of adhesion on E. This suggests that whilst strategies to obtain a more uniform moisture distribution may lower E of more rice grains in retorted samples, this will not necessarily increase the adhesion of the rice. For the first time, cohesion between two individual grains are measured and indicates that the bulk cohesion is the sum of cohesive forces between individual grains. To evaluate bulk cohesion, a new method was developed using a ring shear tester that measures bulk cohesion as an inherent physical property, and this may be a useful method to study the flowing ability of bulk rice. These newly developed techniques improve the ability to quantify the mechanical properties of cooked rice unambiguously; such measurements are not obtained using TPA that exposes samples to large non-linear deformations. Whilst a lack of adhesion and cohesion is desirable for filling and packaging of grains, strategies may be needed to build this into the rice samples at the time of home cooking/consumption to resemble freshly cooked rice as closely as possible. The findings highlight that the unique feature of measuring material properties of individual grains is the potential for isolation of causal relationships. Whilst not addressed here, the next stage is to determine the structural drivers for the observed measurements, and in particular to consider the nature of the surface layer that coats freshly cooked rice grains. The techniques presented provide a basis for comparing and quantifying how material and surface properties are affected by different types of processing, including high pressure processing, to develop new and improved formats of ready-to-eat rice. This may also provide the necessary tools to apply in breeding programs for enhancing selection for rice varieties with particular product propositions. Future work may also address the use of individual grain measurements for building causal correlations with textural attributes.
area between them, and thus underestimates the sum of cohesive stresses in bulk samples of grains. The cohesive stress and the cohesion are broadly distributed, however the distribution of cohesion of the retorted rice is narrower than the cohesive stress, therefore we recommended to measure cohesion rather than cohesive stress for a clearer differentiation between two samples. In bulk samples, rice grains adhere to each other at various angles and each grain has more than one contact point, leading to a network of cohesive stresses; we performed novel measurements using a ring shear tester to evaluate this property. Measuring bulk cohesion with a ring shear tester considers the surface and geometric properties of rice grains affecting cohesion and is the sum of cohesion between individual grains. This technique can be applied to determine and design a nonadhesive rice, which may be useful during manufacture since adhesion and cohesion may affect filling and packaging operations, although in many cases it may be desirable to have features akin to freshly cooked rice at the time of consumption. This points to the potential for future research to develop strategies to consider controlling the rice surfaces at various stages including cooking/reheating at home. In addition, the methods presented here will be invaluable for evaluating the influence on the material properties of rice using other processing methods, for example, emerging high pressure processing technologies. 4.3. Texture Our focus here has not been on generating measurements that relate to sensory properties per se, although this is a major focus in the literature for performing mechanical evaluations. In particular, we do not analyse measurements at a large deformation that may be more relevant to in-mouth sensations arising during oral processing, although this is also possible to perform on individual rice grains. However, we still anticipate that several textural properties will be dependent on the material functions obtained here. E in particular is likely to relate to sensory hardness/firmness, which is often defined based on compressing a sample lightly between teeth without rupturing, i.e. under relatively small deformation. The methods for adhesion and cohesion developed here might be helpful to compare to sensory analysis results such as ‘visual adhesiveness’ defined as the force required to separate individual grains adhering to each other, or ‘manual adhesiveness’ defined as the degree to which the grains stick together in a mass (Lyon et al., 2000), without consuming the sample. Performing quantitative comparisons to sensory attributions requires a large number of measurements on individual grains, which may be prohibitive for many researchers. Using statistical power analysis, the minimum required sample number to obtain a significant difference between two samples with a certain degree of confidence can be calculated. We found that a at least 33 grains were needed to be measured individually to obtain a statistical power of 0.80 for E and adhesion, but by measuring 60 grains, a statistical power of 0.97 was obtained that was sufficient with this set of sample. We therefore recommended to test for statistical power and ensure the minimum number necessary to obtain a valid and rigorous conclusion.
Acknowledgements This research was financially supported by the Australian Research Council (ICI30100011), Industrial Transformation Training Centre ‘Agents of change: Transforming the food industry for Australia, Asia and beyond’, the University of Queensland (UQ), Australia and SunRice. The authors thank SunRice for providing the samples. Further, the authors thank Dr. Tony Howes from the School of Chemical Engineering (UQ) for providing valuable assistance in developing the model used, Dr. Arthur Riedel from the School of Agriculture and Food Sciences (UQ) for valuable assistance in statistics, and Prof. Robert Gilbert from Queensland Alliance for Agriculture and Food Innovation (UQ) for valuable comments and discussion. References Boluda-Aguilar, M., Taboada-Rodríguez, A., López-Gómez, A., Marín-Iniesta, F., BarbosaCánovas, G.V., 2013. Quick cooking rice by high hydrostatic pressure processing. Lwt-Food Sci. Technol. 51 (1), 196–204. Bourne, M., 2002. Food Texture and Viscosity: Concept and Measurement. Academic Press. Brenner, T., Nishinari, K., 2014. A note on instrumental measures of adhesiveness and their correlation with sensory perception. J. Texture Stud. 45 (1), 74–79. Champagne, E.T., Bett, K.L., Vinyard, B.T., McClung, A.M., Barton, F.E., Moldenhauer, K., et al., 1999. Correlation between cooked rice texture and rapid visco analyser measurements. Cereal Chem. J. 76 (5), 764–771. https://doi.org/10.1094/CCHEM.1999. 76.5.764. Fiszman, S.M., DamÁSio, M.H., 2000. Instrumental measurement of adhesiveness in solid and semi-solid foods. A survey. J. Texture Stud. 31 (1), 69–91. https://doi.org/10. 1111/j.1745-4603.2000.tb00285.x. Greenwood, J.A., 1997. Analysis of elliptical Hertzian contacts. Tribol. Int. 30 (3), 235–237. https://doi.org/10.1016/S0301-679X(96)00051-5. Iqbal, T., Fitzpatrick, J.J., 2006. Effect of storage conditions on the wall friction characteristics of three food powders. J. Food Eng. 72 (3), 273–280. https://doi.org/10.
5. Conclusion In this study, we developed sensitive methods to determine the physical properties of cooked rice by measuring the mechanics and surface properties of individual grains of the same rice variety processed in two different ways. The elastic modulus (E), adhesion and cohesion are measured using deformation at small strains and within the linear viscoelastic region. This study found that on average the E of freshly cooked rice is significantly lower than for retorted rice, but they did have overlapping distributions. The overlapping distributions of each physical property of the two samples emphasizes the importance to measure single grains because measuring the average of several 27
Journal of Food Engineering 240 (2019) 21–28
L. Yu et al.
rice by single-grain puncture. J. Texture Stud. 27 (1), 1–13. https://doi.org/10.1111/ j.1745-4603.1996.tb00056.x. Perez, C.M., Juliano, B.O., Bourne, M.C., AnzaldÚA-Morales, A., 1993. Hardness of cooked milled rice by instrumental and sensory methods. J. Texture Stud. 24 (1), 81–94. https://doi.org/10.1111/j.1745-4603.1993.tb01279.x. Ramesh, M., Mitchell, J.R., Jumel, K., Harding, S.E., 1999. Amylose content of rice starch. Starch - Stärke 51 (8–9), 311–313. https://doi.org/10.1002/(SICI)1521-379X (199909)51:8/9<311::AID-STAR311>3.0.CO;2-E. Sneddon, I.N., 1965. The relation between load and penetration in the axisymmetric boussinesq problem for a punch of arbitrary profile. Int. J. Eng. Sci. 3 (1), 47–57. https://doi.org/10.1016/0020-7225(65)90019-4. Sowbhagya, C.M., Ramesh, B.S., Bhattacharya, K.R., 1987. The relationship between cooked-rice texture and the physicochemical characteristics of rice. J. Cereal. Sci. 5 (3), 287–297. https://doi.org/10.1016/S0733-5210(87)80029-2. Szczesniak, A.S., Hall, B.J., 1975. Application of the general foods texturometer to specific food products. J. Texture Stud. 6 (1), 117–138. https://doi.org/10.1111/j.17454603.1975.tb01121.x. Teunou, E., Fitzpatrick, J.J., Synnott, E.C., 1999. Characterisation of food powder flowability. J. Food Eng. 39 (1), 31–37. https://doi.org/10.1016/S0260-8774(98) 00140-X. Tobin, A.B., Heunemann, P., Wemmer, J., Stokes, J.R., Nicholson, T., Windhab, E.J., Fischer, P., 2017. Cohesiveness and flowability of particulated solid and semi-solid food systems. Food Funct. 8, 3647–3653. Tsuji, S., 1981. Texture measurement of cooked rice kernels using the multiple-point mensuration method 1. J. Texture Stud. 12 (2), 93–105. https://doi.org/10.1111/j. 1745-4603.1981.tb01224.x. Windham, W.R., Lyon, B.G., Champagne, E.T., Barton, F.E., Webb, B.D., McClung, A.M., McKenzie, K.S., 1997. Prediction of cooked rice texture quality using near-infrared reflectance analysis of whole-grain milled samples. Cereal Chem. J. 74 (5), 626–632. https://doi.org/10.1094/CCHEM.1997.74.5.626. Yu, L., Turner, M., Fitzgerald, M., Stokes, J., Witt, T., January 2017. Review of the effects of different processing technologies on cooked and convenience rice quality. Trends Food Sci. Technol. 59, 124–138.
1016/j.jfoodeng.2004.12.007. Jane, J., Chen, Y., Lee, L., McPherson, A., Wong, K., Radosavljevic, M., Kasemsuwan, T., 1999. Effects of amylopectin branch chain length and amylose content on the gelatinization and pasting properties of starch 1. Cereal Chem. 76 (5), 629–637. Li, H., Prakash, S., Nicholson, T.M., Fitzgerald, M.A., Gilbert, R.G., 2016. Instrumental measurement of cooked rice texture by dynamic rheological testing and its relation to the fine structure of rice starch. Carbohydr. Polym. 146, 253–263. https://doi.org/ 10.1016/j.carbpol.2016.03.045. Lyon, B.G., Champagne, E.T., Vinyard, B.T., Windham, W.R., 2000. Sensory and instrumental relationships of texture of cooked rice from selected cultivars and postharvest handling practices. Cereal Chem. 77 (1), 64–69. Meullenet, J.-F.C., Gross, J., Marks, B.P., Daniels, M., 1998. Sensory descriptive texture analyses of cooked rice and its correlation to instrumental parameters using an extrusion cell. Cereal Chem. 75 (5), 714–720. Meullenet, J.F., Griffin, V.K., Carson, K., Davis, G., Davis, S., Gross, J., Caicedo, A.L.V., 2001. Rice external preference mapping for asian consumers living in the United States. J. Sensory Stud. 16 (1), 73–94. https://doi.org/10.1111/j.1745-459X.2001. tb00291.x. Moiny, v., meullenet, j. f., xiong, r., 2002. Uniaxial compression of cheddar cheese at various loading rates and its correlation to sensory texture profiles. J. Texture Stud. 33 (3), 237–254. Mossman, A., Fellers, D., Suzuki, H., 1983. Rice stickiness. I. Determination of rice stickiness with an Instron tester. Cereal Chem. 60 (4), 286–292. Okabe, M., 1979. Texture measurement of cooked rice and its relationship to the eating quality. J. Texture Stud. 10 (2), 131–152. https://doi.org/10.1111/j.1745-4603. 1979.tb00241.x. Okadome, H., Toyoshima, H., Ohtsubo, K. i., 1999. Multiple measurements of physical properties of individual cooked rice grains with a single apparatus. Cereal Chem. 76 (6), 855–860. Patindol, J., Gu, X., Wang, Y.-J., 2010. Chemometric analysis of cooked rice texture in relation to starch fine structure and leaching characteristics. Starch - Stärke 62 (3–4), 188–197. https://doi.org/10.1002/star.200900181. Perez, C.M., Bourne, M.C., Juliano, B.O., 1996. Measuring hardness distribution of cooked
28
Contents lists available at ScienceDirect
Journal of Food Engineering journal homepage: www.elsevier.com/locate/jfoodeng
New insights into cooked rice quality by measuring modulus, adhesion and cohesion at the level of an individual rice grain
T
L. Yua, T. Witta, M. Rincon Bonillab,1, M.S. Turnera,c, M. Fitzgeralda, J.R. Stokesb,∗ a
School of Agriculture and Food Sciences, The University of Queensland, Brisbane, 4072, Queensland, Australia School of Chemical Engineering, The University of Queensland, Brisbane, 4072, Queensland, Australia c Queensland Alliance for Agriculture and Food Innovation, The University of Queensland, Brisbane, 4072, Queensland, Australia b
A R T I C LE I N FO
A B S T R A C T
Keywords: Rice Cooked rice Mechanics Adhesion Cohesion Modulus Processing
Causal relationships between physical properties and structure/composition of cooked rice are difficult to quantify when mechanical measurements are performed on bulk samples using large deformations that alter the structure irreversibly. We demonstrate here methods involving small-deformation to characterise the elastic modulus (E), adhesion and cohesion at the individual grain level, and show distinct differences between freshly cooked rice and shelf-stable retorted rice. On average, retorted rice is harder and less adhesive and cohesive than freshly cooked rice, but their distributions in each of these mechanical properties overlap. E is independent of adhesion and weakly correlated with moisture content. In addition, a ring-shear tester is shown to distinguish the bulk cohesion and flowing ability between rice samples. Measuring the inherent physical properties of individual grains has the potential to enable a more sensitive evaluation of new processes and grain varieties, and development of quantitative structure-property-processing relationships for rational design of products to perform optimally at different stages, from manufacturing through to oral processing.
1. Introduction The desire for processed food to have the character of freshly prepared food is driving manufacturers to explore new processing methods and/or formulations. The distribution of ready-to-eat rice in shelfstable, chilled, and frozen forms is particularly desirable given its one of the most consumed staple foods in the world. In designing and evaluating innovative new processing methods to achieve this goal, such as high pressure processing (Yu et al., 2017), there is a need for reliable and sensitive measurements of the physical properties of cooked rice and its structure. Specifically, the mechanical and surface properties of cooked rice are anticipated to contribute to their flow behaviour during processing and consumption. Causal structure-property-processing relationships provide a means in which to rationally evaluate and optimise the specific effect of process and formulation variables. To predict the sensory perception of texture, and to assist in the process of designing a particular structure of food, instrumental procedures have been designed to measure texture-relevant physical properties. However, these are usually based on imitative techniques such as texture profile analysis (TPA), which have limited sensitivity and do not
measure inherent material properties of cooked rice. In this paper, we aim to demonstrate new approaches to evaluate the material properties of cooked rice grains based on measuring the elastic modulus, adhesion and cohesion at the level of individual rice grains. For cooked rice, uniaxial compression tests (including those embedded within TPA routines) as well as others such as puncture tests or extrusion tests, have been used to evaluate how the measured mechanical properties of cooked rice relate to the sensory perception of texture (Li et al., 2016; Lyon et al., 2000; Meullenet et al., 1998; Perez et al., 1993). Li et al. (2016) reported that 80% compression in TPA measured on 1 g of rice placed as a single layer can differentiate between sticky and non-sticky rice. An observed correlation between TPA and sensory measure of ‘hardness’ and ‘stickiness’ largely depends on the significant difference in amylose content in waxy (no amylose) and high amylose (ca. 30% of the starch content) rices. For rice varieties with an amylose content between 18 and 26%, they observed that the hardness measures no longer correlate while the stickiness correlation becomes weaker. Weak correlations between instrumental TPA-assessment of hardness, adhesiveness and cohesiveness with sensory hardness, adhesiveness and cohesiveness of mass were also reported by Lyon
∗
Corresponding author. E-mail addresses: [email protected] (T. Witt), [email protected] (M. Rincon Bonilla), [email protected] (M.S. Turner), m.fi[email protected] (M. Fitzgerald), [email protected] (J.R. Stokes). 1 Current address: Basque Centre for Applied Mathematics, Alameda de Mazarredo 14 (48009) Bilbao, Bizkaia, Spain. https://doi.org/10.1016/j.jfoodeng.2018.07.010 Received 17 July 2017; Received in revised form 30 April 2018; Accepted 6 July 2018 Available online 09 July 2018 0260-8774/ © 2018 Published by Elsevier Ltd.
Journal of Food Engineering 240 (2019) 21–28
L. Yu et al.
performed by confining a sample to a certain extent between two surfaces, and then the surfaces are separated at a set speed. Adhesion is reported as either the maximum negative force required to detach the rice grain from the device surface, or the negative area under the forcedistance or force-time curve (Fiszman and DamÁSio, 2000; Okabe, 1979). These differences in definition, as well as choice of extent of confinement and detachment speed, make comparison of literature values difficult. Thus a constant definition and a clear interpretation of how each physical property relates to each sensory property would be beneficial, although it is not well known if the different definitions actually lead to a different conclusion. The physical measurement of adhesion of individual grains to a surface is recommended because Mossman et al. (1983) showed that measuring 40 grains resulted in less variation in adhesion than measuring an individual grain. Such homogeneity may lead to inconclusive results when comparing differences between rices as a function of process variables or rice variety. To evaluate the sensory adhesion of rice to surfaces such as to a spoon, molars or the lips, it is relevant to measure the surface properties of cooked rice grains (Okadome et al., 1999). Large strain deformation tests that are used in many studies will alter the integrity of the rice and the measured adhesion will not necessarily be related to its surface properties. The purpose of this study is to overcome some of the limitations described above by developing a more sensitive measure of the mechanical and surface properties of cooked rice grains. Individual rice grains of the same rice variety cooked in a rice cooker or retorted are measured in a small strain uniaxial compression test to obtain objective and inherent physical and surface properties of individual grains. The elastic modulus is obtained within the elastic limit of the rice grains. The two common ways of analysing rice adhesion are compared to evaluate which is more suitable to differentiate between the two rice samples when they are in contact with a metal surface. We also seek to evaluate cohesion as a surface property between two individual rice gains, and compare this to ‘bulk’ values obtained using a ring-shear tester. Whilst the ring-shear tester is normally used to measure flowability and cohesion in powders (Schulze, 2008), including powdered food (Iqbal and Fitzpatrick, 2006), it was recently used to provide insights into these properties in hydrated semi-solid foods (Tobin et al., 2017).
et al. (2000) for different rice varieties. The authors highlight the need for more sensitive instrumental mechanical methods to capture the subtle, yet detectable differences in sensory texture between samples (Lyon et al., 2000). We note in particular that both the extrusion test and the TPA approach commonly measure the mechanical response of cooked rice grains while they are being destroyed during exposure to high compression. In this case, the rice is transformed dramatically, well-beyond its elastic limit, as it is compressed into a ‘single’ bulk soft material constructed of all components of the grain. This transformed product bears little resemblance to the original rice grains, and so it is perhaps unsurprising that measurements obtained on the rice in this state do not strongly relate to adhesiveness and cohesiveness of rice that we would hypothesise to be closely related to the surface properties of the grain. Despite the wide utilisation of uniaxial compression tests to evaluate texture-related mechanical properties of cooked rice and other food systems, the sample preparation, sample size and amount, plunger size, compression speed, and degree of compression vary highly between study groups. This renders the results difficult to compare, particularly when the inherent material properties (e.g. modulus) are not extracted from measured data (Champagne et al., 1999; Meullenet et al., 2001; Mossman et al., 1983; Okadome et al., 1999). Due to such variability in parameters combined with the tendency to perform measurements at high compressive strains, where rice is transformed to a new structural state, such mechanical measurements do not necessarily provide information about the inherent physical and surface properties of the rice. This limits the establishment of a unifying structure-property relationship that is relevant to a full spectrum of sensory texture attributes. The major limitations of current uniaxial compression methodologies are the poor definition of the contact area, the low likelihood of capturing the true variation of the sample, and the measurement of irrelevant sensory properties. These all need to be overcome in order to predict relevant sensory texture properties accurately. The arrangement of some number of rice grains in a single layer equidistant from each other is a commonly practiced experimental method (Boluda-Aguilar et al., 2013; Li et al., 2016; Meullenet et al., 2001; Patindol et al., 2010). However rice kernels are not flat, so the contact area with the plunger increases as the compression progresses (Tsuji, 1981). This means that it is important for all samples to have the same geometry when submitted to a compression test because stress is defined as the force divided by contact area (Bourne, 2002). A difference in the contact area will lead to a difference in the measured stress, even when the modulus is the same. Shaping each sample into the same geometry is applied to other soft foods such as cheddar cheese. In this case, good predictions of sensory springiness, cohesiveness, cohesiveness of mass, roughness of mass and tooth-pull can be obtained with TPA (Moiny et al., 2002). As well established in material science, differences in forces may be due to differences in contact area and not inherent differences in modulus and fracture properties. When a layer of rice grains with a distribution of radii are compressed, the plunger makes contact with the rice grain of the largest diameter first, so the reactive force is an ambiguous sum of rice grains at differing levels of compression. This is problematic as a distribution of hard and soft grains under compression can easily lead to measurements that correspond to hard rice grains even when most grains are actually soft. Due to the non-uniformity between grains, a better approach to investigate intrinsic properties relevant to sensory perception, as recommended by Szczesniak and Hall (1975), would be to measure the distribution of hardness by performing measurements on large numbers of individual grains. There are a number of different terms used to describe stickiness or adhesiveness in the literature; here we define adhesion as the stickiness between the rice grain and the surface, and cohesion as the stickiness between individual rice grains. Adhesion is a measure of force, work or energy required to detach rice grains from a surface but the actual definition of the measurement varies. Typically, a “pull-off” test is
2. Material and methods 2.1. Material The samples used (variety Langi) were grown in the Riverina, NSW, Australia and harvested in 2014. The grains were provided after milling. The retorted rice of the same variety and same batch was kindly provided by SunRice and contained rice, water, 2.5% vegetable oil and 0.5% distilled monoglyceride. The chemical composition of raw rice was provided by SunRice showing a total starch content of 76.9% with 17.8% amylose and a moisture content of 12.1%. 2.2. Sample preparation For freshly cooked rice, 300 g of milled rice was weighed into a rice cooker bowl and washed three times with 1.5 L of deionised water for 10 s. After each washing step, the rice was drained in a strainer for 10 s. Milled rice was cooked with a water-to-rice ratio of 1.75 in an electric rice cooker (Tefal intelligent rice cooker, Tefal S.A.S, Rumilly, France). The cooking time was 45 min. After cooking, the outer and bottom 1 cm of rice was discarded. The rest was mixed gently with a plastic paddle to distribute the moisture homogenously, then rice was kept warm for another 10 min in the cooker. Before instrumental analysis, retorted rice samples were warmed in the pouch according to the recommendation on the package in a microwave at 900 W for 90 s. All samples were cooled to approximately 40 °C and kept in a water bath 22
Journal of Food Engineering 240 (2019) 21–28
L. Yu et al.
set to 43 °C during measurements to retard retrogradation. 2.3. Moisture content To measure moisture content 10 ± 0.1 g of each rice sample was weighed, then dried at 103 °C for 24 h and weighed again (n = 4). To investigate the correlation between elastic modulus and moisture content, 20 individual grains were dried after the compression test at the same oven conditions. Immediately after drying, the samples were weighed again. The moisture content on a wet basis was calculated as the difference in water weight divided by the original sample weight. Fig. 1. An example of model fitting (red line) of the experimental data (blue dots) within the linear viscoelastic region of a force-indentation curve. The fitting curves have an R2 value ≥ 0.995.
2.4. Rice grain size Rice grains were assumed to be ellipsoids, and the length and diameter were measured with a calliper (SUC23-200, ProSciTech, Kirwan, Australia). Sixty random grains from 3 preparations of freshly cooked and 3 packages of retorted rice, respectively, were measured and the average is reported. The aspect ratio (α) of the rice grains was calculated as the average length (R) divided by the average diameter (R’). Rice grains with obvious defects were not measured.
where b is the minor axis of the elliptical contact, and is related to the estimated contact area by (5)
A = πα 4/3b2 Combining equation (1), (3) and (4), produced equation (6)
4 E R β δ 1.5 3 (1 − ν 2)
2.5. Mechanical and surface properties of rice
F=
To develop a method to measure the mechanics of single grains, freshly cooked rice was compared to warmed, retorted rice. To measure elastic modulus (E), individual rice grains were uniaxially compressed with a smooth parallel plate (40 mm diameter) on a stress-controlled rheometer (AR-G2, TA Instruments Ltd., UK). A set compression speed of 10 μm/s was applied until the normal force reached 1.0 ± 0.01 N, then the plates were separated at a speed of 20 μm/s. The normal force of 1.0 N was chosen following extensive preliminary tests that found that for 90% of the cooked rice grains measured, compression was within the linear viscoelastic region, and thus was before the grains plastically deformed or fractured. To measure E and adhesion, intact grains were randomly taken from at least 3 different preparations of freshly cooked and 3 different packages of retorted rice.
where beta in equation (7) represents the contribution from the aspect ratio 1/3
β=
2 21.5α 2 ⎛α + 1⎞ + 1)1.5 ⎝ 2α 4 ⎠
(α 2/3
⎜
⎟
(7)
and becomes unity for a circular contact area. Fig. 1 shows force against indentation (δ 1.5) data for a particular rice grain. E was extracted from the slope of the force F vs. the indentation δ 1.5 curve fitted with MATLAB (R2014b, The MathWorks Inc.) and by manually choosing the zero point; this is the point where the plate first came into contact with the grain, and the force began to increase. In order to eliminate error due to uncertainty in the initial gap, a section of the curve with a minimum of 10 points was selected for fitting in such a way that r2 was always above 0.995.
2.5.1. Elastic modulus E The elastic modulus E for individual rice grains was estimated using the Hertzian contact model for compression of a spherical body with a flat punch (Sneddon, 1965). However, since the grains are ellipsoidal, the correction introduced by Greenwood (1997) for bodies where the contact shape is an ellipse, was employed. Here, an equivalent sphere radius Re is calculated in order to preserve the relationship between force F and contact area A provided by the Hertzian model, i.e., equation (1)
3FR e 2/3 ⎞ A = π⎛ ⎝ 4E ∗ ⎠
(6)
2.5.2. Adhesive stress and adhesion The two methods for quantifying adhesion from a “pull-off” test are evaluated in Fig. 2. In this study, we define adhesive stress as the absolute of the maximum negative force for the detachment curve, divided by the contact area between the rice grain and the plate. The estimated area is calculated using equation (4); note this is an ‘engineering’ stress because it does not account for the increase in contact area that arises during compression. Adhesion is defined here as the adhesive energy, which is the area under the negative curve divided by the contact area between rice grain and plate. The starting point to
(1)
where
E ∗ = E/(1 − ν 2)
(2)
with E as the elastic modulus and ν the Poisson's ratio. Incompressibility is assumed (ν = 0.5) because of the high water content of the grain. Re is a function of the radii of curvature and can be expressed as a function of the aspect ratio α according to equation (2) 1/3
2α 4 ⎞ Re = R ⎛ 2 ⎝α + 1⎠ ⎜
⎟
(3)
whereby R is half the thickness of the rice grain (i.e., the minor radius of the ellipsoid). To estimate force from indentation δ, an additional correction was required (Greenwood, 1997) using equation (3)
b2 2/3 δ= (α + 1) 2R
Fig. 2. An example of calculating the area under the negative curve as adhesive energy (grey) and the maximum negative force (black line) indicated by the red arrow.
(4) 23
Journal of Food Engineering 240 (2019) 21–28
L. Yu et al.
2.7. Statistical analysis Statistical differences were calculated with GraphPad Prism Version 6 (GraphPad Software, Inc., La Jolla, USA) with t-test and KolmogorovSmirnov test (non-parametric, two-tailed, heteroscedastic). Due to a high percentage of zero values in adhesion, cohesion and cohesive stress especially for retorted rice, the statistical difference was calculated for the population greater than zero and the average is presented with standard deviation. Statistical power analysis, which is applied to determine the minimum sample size required to detect an effect of a given size with a given degree of confidence, was calculated in R studio (R studio, Boston, US). Pearson correlations were calculated in Microsoft Excel.
Fig. 3. (a). An image of the experimental setup used to measure cohesive stress between two perpendicular single rice grains compressed against each other; (b): Schematic to show the calculated area between two perpendicular single grains in grey.
integrate the area was automatically determined when the mean of six consecutive data points fell below 2.5 times the standard deviation of the mean of these six points, and the integration stopped when the curve reached the mean again. The noise range for zero compression force is 0.01 N, thus values < 0.01N were regarded as equal to zero and therefore not adhesive.
3. Results 3.1. Size distribution of grains The length of freshly cooked rice grains varies between 8.8 and 12 mm with an average of 10 ± 0.7 mm. The freshly cooked grains are significantly longer (p < 0.0001) than the retorted rice which vary from 7.2 to 9.7 mm, with an average of 8.3 ± 0.6 mm. However, the diameter is not significantly different between the two types of grain. The average aspect ratio α is 4.0 ± 0.4 for freshly cooked rice and 3.3 ± 0.3 for retorted rice.
2.5.3. Cohesive stress and cohesion Cohesive stress and cohesion between two individual grains are compared to determine which method characterizes cohesion better. A schematic setup is shown in Fig. 3a and b. The cohesive stress is defined as the maximum negative force between two rice grains that are separated from each other at a right angle after compression force reached 0.1 N, divided by the contact area between the two rice grains. To calculate the contact area between two grains, the gap at 0.1 N compression is assumed to be the diameter of both grains. For simplification the two grains are assumed to have the same diameter. The two grains are perpendicular, so the contact area is assumed to be the square of the diameter, which slightly overestimates the contact area. Cohesion is defined here as the cohesive energy, calculated in the same way as for adhesion, divided by the contact area between two rice grains. The force of 0.1 N was chosen because the surfaces of adjacent grains were already touching and a higher force fractured the grains. The upper plate was not fixed, thus slight rotation upon contact between the grains cannot be excluded. The noise range for zero compression force is 0.01 N, thus values < 0.01N were regarded as equal to zero and therefore not cohesive.
3.2. Mechanical properties 3.2.1. Elastic modulus E The distribution of the elastic modulus (E) of freshly cooked rice is narrower and has a lower average than retorted rice (p < 0.0001) (Fig. 5). For freshly cooked rice, E ranges from 0.3 to 1.5 MPa and the average is 0.89 ± 0.26 MPa, while for retorted rice, E ranges from 0.62 to 8.2 MPa with an average of 1.4 ± 1 MPa. Both grains show a skewed distribution of elastic moduli, with a greater number of soft grains than hard grains. 3.2.2. Elastic modulus and moisture content A reasonable hypothesis for the broad distribution of E is that it arises from an inhomogeneous moisture distribution within a batch, thus the moisture content of 20 individual grains was measured after compression to investigate whether E is dependent on the moisture content. Fig. 6 shows that the moisture content of freshly cooked rice ranges between 57% and 66% and is narrower than the moisture content of retorted rice grains, which varies between 39% and 62%.
2.6. Cohesion of bulk rice grains Cohesion of bulk rice was measured with a Ring Shear Tester (RSTXS, Dietmar Schulze, Wolfenbuettel, Germany) following the methods described in Schulze (2008). The XS-Lr shear cell was filled, weighed and then placed on the motor-driven base of the Ring Shear Tester. 47 ± 1.0 g of bulk rice is used that corresponds to a volume of 72.99 cm3. The sample were pre-sheared under a consolidation force of 1000 N until a steady-state was reached, and then sheared until incipient flow was attained under normal forces of 200 N, 320 N, 440 N, 560 N, 680 N and 800 N at a speed of 75 mm/min. All rice samples were measured in triplicate with each sample measurement involving duplicate measurements of each normal force. The cohesion was the intercept of the yield loci from a linear fit of the shear force required to induce incipient flow with the applied normal forces; the cohesion is defined as the shear stress required to cause flow in the material under zero normal force load (Schulze, 2008; Teunou et al., 1999). An example of this is shown in Fig. 4 with the cohesions of 214 Pa and 464 Pa measured for retorted and freshly cooked rice respectively. The slope of ca. 0.6 was similar for both samples and is generally related to the angle of internal friction; since it was not differentiating between samples, we do not consider it further.
3.2.3. Adhesive stress and adhesion Fig. 7a and b shows the distributions of the adhesive stress and adhesion of individual cooked and retorted rice grains, respectively. Adhesive stress between freshly cooked rice and the stainless steel plate ranges from 8.2 to 53 mPa with an average of 25 ± 9 mPa, and for retorted rice from 0.8 to 45 mPa with an average of 11 ± 8 mPa. Adhesion is below the detection limit with adhesion being zero for 7% of freshly cooked rice grains and 33% of retorted rice, therefore the average and standard deviation reported here represent the population with adhesion greater than zero. Adhesion between freshly cooked rice grain and the stainless steel surface ranges from 0 to 9.5 J/m2 with an average of 4.2 ± 2 J/m2 and for retorted rice from 0 to 2.0 J/m2 with an average of 0.7 ± 0.6 J/m2. The results show that freshly cooked rice is significantly more adhesive to a stainless steel plate than retorted rice (p < 0.0001), and that most of the retorted rice grains are essentially non-adhesive. 3.2.4. Cohesive stress and cohesion Cohesive stress is below detection limit for 10% of freshly cooked rice grains and for 65% of retorted rice with cohesive stress being zero; 24
Journal of Food Engineering 240 (2019) 21–28
L. Yu et al.
Shear Stress (Pa)
900
y = 0.617x + 464
700 500
y = 0.627x + 218
300 100 0
200
400 600 Normal Stress (Pa)
800
Fig. 4. An example of calculating the bulk cohesion as the intercept of the yield locus on the shear stress axis (y-axis) of freshly cooked (■) and warmed, retorted rice (●).
Fig. 7. (a). The distribution of adhesive stresses between plate and rice of freshly cooked rice (■) (n = 60) and warmed, retorted rice (■) (n = 60). This shows that freshly cooked rice grains are significantly more adhesive than retorted rice grains (p < 0.0001). (b) The distribution of adhesive energy between plate and rice surface of freshly cooked rice (■) (n = 60) and warmed, retorted rice (■) (n = 60). This shows that the freshly cooked rice grains have a broader distribution and are significantly more adhesive than retorted rice grains (p < 0.0001).
Fig. 5. The distribution of elastic moduli of individual grains of freshly cooked rice (■) (n = 60) and warmed, retorted rice (■) (n = 60). The mean elastic modulus of freshly cooked rice is significantly lower than retorted rice (p < 0.0001).
Fig. 6. A plot of the negative correlation between moisture content and elastic modulus of warmed, retorted rice (●) and freshly cooked rice (■) (n = 20), R2 = 0.6637.
Fig. 7. (continued)
and cohesion is below detection limit for 35% of retorted rice grains with cohesion being zero. Therefore, the average and standard deviation reported here represent the population with cohesive stress or cohesion greater than zero. The cohesion for the population of rice granule pairs is shown in Fig. 8. Pairs of freshly cooked and retorted rice grains cohesive stress (4040 ± 1260 Pa) and cohesion (0.84 ± 0.58 J/m2) are significantly higher (p < 0.05, p < 0.0001) between two freshly cooked rice grains than two retorted rice grains (cohesive stress 2770 ± 650 Pa, cohesion 0.080 ± 0.09 J/m2). A broader distribution is observed for the freshly cooked than retorted rice, which is accompanied by a larger distribution of contact areas between two freshly cooked grains; from 2.7 to 5 mm2 for freshly cooked rice pairs compared to 3.5–4.9 mm2 for retorted rice grain pairs.
Another approach to understand cohesion is to use a method which provides an average of the bulk interactions and flowing ability of the rice grains, one such method is to use a ring shear tester. Similar to the trend observed in the individual grain cohesion results, the bulk cohesion of freshly cooked rice at 470 ± 34 Pa is significantly greater than for retorted rice at 206 ± 17 Pa (p < 0.0001). This technique measures the interaction between grains as well as between grains and shear cell wall, and assesses how sticky the bulk is.
4. Discussion Measuring the inherent physical properties at the level of individual rice grains can be applied to characterise and compare the mechanical 25
Journal of Food Engineering 240 (2019) 21–28
L. Yu et al.
60
Percentage
50 40 30 20 10 0 0
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
1
1.1 1.2 1.3 1.4 >1.5
Cohesion (J/m2) Fig. 8. The distribution of cohesion between rice grains pairs of freshly cooked rice (■) (n = 60) and warmed, retorted rice (■) (n = 60). This shows that the freshly cooked rice grains have a broader distribution and are significantly more cohesive than retorted rice grains (p < 0.0001).
properties as a function of processing variables, as demonstrated here, but also between different rice varieties and their composition. Measuring material functions of individual grains, including E, adhesion and cohesion at low deformation (i.e. prior to plastic deformation), ensures that valid comparisons can be made between results obtained from different studies and research groups. The power of performing measurements on individual grains, or grain pairs is demonstrated by analysing the interrelationship between variables and leading to information on the structural or compositional drivers for particular physical properties. Fig. 9. A plot of adhesion and elastic modulus for individual grains of freshly cooked rice (■) (n = 60) and warmed, retorted rice (●) (n = 60). Neither rice grain population displayed a correlation.
4.1. Mechanics The wider distributions for E of retorted rice indicates a greater heterogeneity in the retorted rice grains, which is possibly due to the difference in processing. Skewed distribution of rice grains was reported by Perez et al. (1996), who recorded the maximum penetration force of close to 100 individual cooked rice grains in a puncture test to compare rice varieties with similar physicochemical properties, and the distributions produced display the same skew toward soft grains as seen in the current study. The wider range of moisture content of retorted rice matches with the wider range of E and is another indicator that moisture affects E. A weak, negative correlation between E and moisture content is evident in the joint population of both samples. A similar result using 15 different rice varieties was reported by Perez et al. (1996) with the maximum penetration force of cooked, individual rice grains with an average moisture content of 69% exhibiting a higher penetration force than rice with an average moisture content of 75% (Perez et al., 1996). Many studies have correlated amylose, amylopectin or protein content to cooked rice hardness (Sowbhagya et al., 1987; Ramesh et al., 1999; Jane et al., 1999; Okadome et al., 1999), however, only one rice variety was used in our study, thus any effect of compositional differences on hardness between rice varieties can be ruled out. The weak, negative correlation between moisture content and E in freshly cooked and retorted rice grains indicates that the E of individual grains may also be dependent on other properties such as, for example, moisture distribution within the grain, starch crystallinity, hollows that occur during cooking inside the grain, and the grain morphology. On average, the freshly cooked rice is softer, has a higher moisture content and is more adhesive than retorted rice. We therefore evaluate whether E is related to adhesion in Fig. 9, but find no correlation between them. That is, for individual grains of the same modulus, freshly cooked rice is much more adhesive than retorted rice. This is an important finding; the measured adhesion is likely to depend on a surface layer coating the bulk grain. The results suggest that it is this property that is largely affected by the difference in processing, and this is likely
to have a major effect on the behaviour of the rice in processing and during consumer use and consumption. We should point out that the adhesion and E measure here are essentially intrinsic material properties of the system whereby they are determined at low deformation. The current experiments are thus more likely to describe properties related to rice textural attributes which are prior to chewing the rice into a bolus, and thus the measurements do not represent the maximum force and adhesiveness as measured using TPA which may be negatively correlated as reported by Tsuji (1981). 4.2. Adhesion and cohesion Adhesion results are less broadly distributed than the adhesive stress results, and provide a clearer distinction between the two samples. The ability differentiate more clearly between adhesive and nonadhesive samples leads to our recommendation to use adhesion in comparison between samples rather than adhesive stress. To evaluate rice adhesion to a surface such as stainless steel, we recommend considering the contact area upon compression. This is further supported by Brenner and Nishinari (2014), who state that the dependence of adhesion on the contact area between sample and the plunger after compression needs to be accounted for, but is not normally performed in TPA. The small strain compression used in this study distinguishes between the adhesion property of the grain surface rather than following destruction of the sample that is considered in TPA. Surface adhesion is useful in a sensory study that evaluates how rice grains adhere to different surfaces such as a container, spoon (Windham et al., 1997), lips (Lyon et al., 2000) or molars, but is less likely to be relevant to chewed samples. For the first time, the cohesion between rice grains as an inherent surface property was measured. It should be noted that the cohesion between two perpendicular rice grains measures the minimum contact 26
Journal of Food Engineering 240 (2019) 21–28
L. Yu et al.
grains often does not provide an unambiguous result of the population. We also found that the difference in E was only partially dependent on moisture content. In addition, measuring individual grains has uniquely uncovered that a major difference between freshly cooked rice and retorted rice is that the latter in predominantly non-adhesive, yet there is no dependence of adhesion on E. This suggests that whilst strategies to obtain a more uniform moisture distribution may lower E of more rice grains in retorted samples, this will not necessarily increase the adhesion of the rice. For the first time, cohesion between two individual grains are measured and indicates that the bulk cohesion is the sum of cohesive forces between individual grains. To evaluate bulk cohesion, a new method was developed using a ring shear tester that measures bulk cohesion as an inherent physical property, and this may be a useful method to study the flowing ability of bulk rice. These newly developed techniques improve the ability to quantify the mechanical properties of cooked rice unambiguously; such measurements are not obtained using TPA that exposes samples to large non-linear deformations. Whilst a lack of adhesion and cohesion is desirable for filling and packaging of grains, strategies may be needed to build this into the rice samples at the time of home cooking/consumption to resemble freshly cooked rice as closely as possible. The findings highlight that the unique feature of measuring material properties of individual grains is the potential for isolation of causal relationships. Whilst not addressed here, the next stage is to determine the structural drivers for the observed measurements, and in particular to consider the nature of the surface layer that coats freshly cooked rice grains. The techniques presented provide a basis for comparing and quantifying how material and surface properties are affected by different types of processing, including high pressure processing, to develop new and improved formats of ready-to-eat rice. This may also provide the necessary tools to apply in breeding programs for enhancing selection for rice varieties with particular product propositions. Future work may also address the use of individual grain measurements for building causal correlations with textural attributes.
area between them, and thus underestimates the sum of cohesive stresses in bulk samples of grains. The cohesive stress and the cohesion are broadly distributed, however the distribution of cohesion of the retorted rice is narrower than the cohesive stress, therefore we recommended to measure cohesion rather than cohesive stress for a clearer differentiation between two samples. In bulk samples, rice grains adhere to each other at various angles and each grain has more than one contact point, leading to a network of cohesive stresses; we performed novel measurements using a ring shear tester to evaluate this property. Measuring bulk cohesion with a ring shear tester considers the surface and geometric properties of rice grains affecting cohesion and is the sum of cohesion between individual grains. This technique can be applied to determine and design a nonadhesive rice, which may be useful during manufacture since adhesion and cohesion may affect filling and packaging operations, although in many cases it may be desirable to have features akin to freshly cooked rice at the time of consumption. This points to the potential for future research to develop strategies to consider controlling the rice surfaces at various stages including cooking/reheating at home. In addition, the methods presented here will be invaluable for evaluating the influence on the material properties of rice using other processing methods, for example, emerging high pressure processing technologies. 4.3. Texture Our focus here has not been on generating measurements that relate to sensory properties per se, although this is a major focus in the literature for performing mechanical evaluations. In particular, we do not analyse measurements at a large deformation that may be more relevant to in-mouth sensations arising during oral processing, although this is also possible to perform on individual rice grains. However, we still anticipate that several textural properties will be dependent on the material functions obtained here. E in particular is likely to relate to sensory hardness/firmness, which is often defined based on compressing a sample lightly between teeth without rupturing, i.e. under relatively small deformation. The methods for adhesion and cohesion developed here might be helpful to compare to sensory analysis results such as ‘visual adhesiveness’ defined as the force required to separate individual grains adhering to each other, or ‘manual adhesiveness’ defined as the degree to which the grains stick together in a mass (Lyon et al., 2000), without consuming the sample. Performing quantitative comparisons to sensory attributions requires a large number of measurements on individual grains, which may be prohibitive for many researchers. Using statistical power analysis, the minimum required sample number to obtain a significant difference between two samples with a certain degree of confidence can be calculated. We found that a at least 33 grains were needed to be measured individually to obtain a statistical power of 0.80 for E and adhesion, but by measuring 60 grains, a statistical power of 0.97 was obtained that was sufficient with this set of sample. We therefore recommended to test for statistical power and ensure the minimum number necessary to obtain a valid and rigorous conclusion.
Acknowledgements This research was financially supported by the Australian Research Council (ICI30100011), Industrial Transformation Training Centre ‘Agents of change: Transforming the food industry for Australia, Asia and beyond’, the University of Queensland (UQ), Australia and SunRice. The authors thank SunRice for providing the samples. Further, the authors thank Dr. Tony Howes from the School of Chemical Engineering (UQ) for providing valuable assistance in developing the model used, Dr. Arthur Riedel from the School of Agriculture and Food Sciences (UQ) for valuable assistance in statistics, and Prof. Robert Gilbert from Queensland Alliance for Agriculture and Food Innovation (UQ) for valuable comments and discussion. References Boluda-Aguilar, M., Taboada-Rodríguez, A., López-Gómez, A., Marín-Iniesta, F., BarbosaCánovas, G.V., 2013. Quick cooking rice by high hydrostatic pressure processing. Lwt-Food Sci. Technol. 51 (1), 196–204. Bourne, M., 2002. Food Texture and Viscosity: Concept and Measurement. Academic Press. Brenner, T., Nishinari, K., 2014. A note on instrumental measures of adhesiveness and their correlation with sensory perception. J. Texture Stud. 45 (1), 74–79. Champagne, E.T., Bett, K.L., Vinyard, B.T., McClung, A.M., Barton, F.E., Moldenhauer, K., et al., 1999. Correlation between cooked rice texture and rapid visco analyser measurements. Cereal Chem. J. 76 (5), 764–771. https://doi.org/10.1094/CCHEM.1999. 76.5.764. Fiszman, S.M., DamÁSio, M.H., 2000. Instrumental measurement of adhesiveness in solid and semi-solid foods. A survey. J. Texture Stud. 31 (1), 69–91. https://doi.org/10. 1111/j.1745-4603.2000.tb00285.x. Greenwood, J.A., 1997. Analysis of elliptical Hertzian contacts. Tribol. Int. 30 (3), 235–237. https://doi.org/10.1016/S0301-679X(96)00051-5. Iqbal, T., Fitzpatrick, J.J., 2006. Effect of storage conditions on the wall friction characteristics of three food powders. J. Food Eng. 72 (3), 273–280. https://doi.org/10.
5. Conclusion In this study, we developed sensitive methods to determine the physical properties of cooked rice by measuring the mechanics and surface properties of individual grains of the same rice variety processed in two different ways. The elastic modulus (E), adhesion and cohesion are measured using deformation at small strains and within the linear viscoelastic region. This study found that on average the E of freshly cooked rice is significantly lower than for retorted rice, but they did have overlapping distributions. The overlapping distributions of each physical property of the two samples emphasizes the importance to measure single grains because measuring the average of several 27
Journal of Food Engineering 240 (2019) 21–28
L. Yu et al.
rice by single-grain puncture. J. Texture Stud. 27 (1), 1–13. https://doi.org/10.1111/ j.1745-4603.1996.tb00056.x. Perez, C.M., Juliano, B.O., Bourne, M.C., AnzaldÚA-Morales, A., 1993. Hardness of cooked milled rice by instrumental and sensory methods. J. Texture Stud. 24 (1), 81–94. https://doi.org/10.1111/j.1745-4603.1993.tb01279.x. Ramesh, M., Mitchell, J.R., Jumel, K., Harding, S.E., 1999. Amylose content of rice starch. Starch - Stärke 51 (8–9), 311–313. https://doi.org/10.1002/(SICI)1521-379X (199909)51:8/9<311::AID-STAR311>3.0.CO;2-E. Sneddon, I.N., 1965. The relation between load and penetration in the axisymmetric boussinesq problem for a punch of arbitrary profile. Int. J. Eng. Sci. 3 (1), 47–57. https://doi.org/10.1016/0020-7225(65)90019-4. Sowbhagya, C.M., Ramesh, B.S., Bhattacharya, K.R., 1987. The relationship between cooked-rice texture and the physicochemical characteristics of rice. J. Cereal. Sci. 5 (3), 287–297. https://doi.org/10.1016/S0733-5210(87)80029-2. Szczesniak, A.S., Hall, B.J., 1975. Application of the general foods texturometer to specific food products. J. Texture Stud. 6 (1), 117–138. https://doi.org/10.1111/j.17454603.1975.tb01121.x. Teunou, E., Fitzpatrick, J.J., Synnott, E.C., 1999. Characterisation of food powder flowability. J. Food Eng. 39 (1), 31–37. https://doi.org/10.1016/S0260-8774(98) 00140-X. Tobin, A.B., Heunemann, P., Wemmer, J., Stokes, J.R., Nicholson, T., Windhab, E.J., Fischer, P., 2017. Cohesiveness and flowability of particulated solid and semi-solid food systems. Food Funct. 8, 3647–3653. Tsuji, S., 1981. Texture measurement of cooked rice kernels using the multiple-point mensuration method 1. J. Texture Stud. 12 (2), 93–105. https://doi.org/10.1111/j. 1745-4603.1981.tb01224.x. Windham, W.R., Lyon, B.G., Champagne, E.T., Barton, F.E., Webb, B.D., McClung, A.M., McKenzie, K.S., 1997. Prediction of cooked rice texture quality using near-infrared reflectance analysis of whole-grain milled samples. Cereal Chem. J. 74 (5), 626–632. https://doi.org/10.1094/CCHEM.1997.74.5.626. Yu, L., Turner, M., Fitzgerald, M., Stokes, J., Witt, T., January 2017. Review of the effects of different processing technologies on cooked and convenience rice quality. Trends Food Sci. Technol. 59, 124–138.
1016/j.jfoodeng.2004.12.007. Jane, J., Chen, Y., Lee, L., McPherson, A., Wong, K., Radosavljevic, M., Kasemsuwan, T., 1999. Effects of amylopectin branch chain length and amylose content on the gelatinization and pasting properties of starch 1. Cereal Chem. 76 (5), 629–637. Li, H., Prakash, S., Nicholson, T.M., Fitzgerald, M.A., Gilbert, R.G., 2016. Instrumental measurement of cooked rice texture by dynamic rheological testing and its relation to the fine structure of rice starch. Carbohydr. Polym. 146, 253–263. https://doi.org/ 10.1016/j.carbpol.2016.03.045. Lyon, B.G., Champagne, E.T., Vinyard, B.T., Windham, W.R., 2000. Sensory and instrumental relationships of texture of cooked rice from selected cultivars and postharvest handling practices. Cereal Chem. 77 (1), 64–69. Meullenet, J.-F.C., Gross, J., Marks, B.P., Daniels, M., 1998. Sensory descriptive texture analyses of cooked rice and its correlation to instrumental parameters using an extrusion cell. Cereal Chem. 75 (5), 714–720. Meullenet, J.F., Griffin, V.K., Carson, K., Davis, G., Davis, S., Gross, J., Caicedo, A.L.V., 2001. Rice external preference mapping for asian consumers living in the United States. J. Sensory Stud. 16 (1), 73–94. https://doi.org/10.1111/j.1745-459X.2001. tb00291.x. Moiny, v., meullenet, j. f., xiong, r., 2002. Uniaxial compression of cheddar cheese at various loading rates and its correlation to sensory texture profiles. J. Texture Stud. 33 (3), 237–254. Mossman, A., Fellers, D., Suzuki, H., 1983. Rice stickiness. I. Determination of rice stickiness with an Instron tester. Cereal Chem. 60 (4), 286–292. Okabe, M., 1979. Texture measurement of cooked rice and its relationship to the eating quality. J. Texture Stud. 10 (2), 131–152. https://doi.org/10.1111/j.1745-4603. 1979.tb00241.x. Okadome, H., Toyoshima, H., Ohtsubo, K. i., 1999. Multiple measurements of physical properties of individual cooked rice grains with a single apparatus. Cereal Chem. 76 (6), 855–860. Patindol, J., Gu, X., Wang, Y.-J., 2010. Chemometric analysis of cooked rice texture in relation to starch fine structure and leaching characteristics. Starch - Stärke 62 (3–4), 188–197. https://doi.org/10.1002/star.200900181. Perez, C.M., Bourne, M.C., Juliano, B.O., 1996. Measuring hardness distribution of cooked
28