- Email: [email protected]
Dynamic rheological properties of plant cell-wall particle dispersions
Colloids and Surfaces B: Biointerfaces 81 (2010) 461–467
Contents lists available at ScienceDirect
Colloids and Surfaces B: Biointerfaces journal homepage: www.elsevier.com/locate/colsurfb
Dynamic rheological properties of plant cell-wall particle dispersions Li Day, Mi Xu, Sofia K. Øiseth, Leif Lundin, Yacine Hemar ∗ CSIRO Food & Nutrition Sciences, and Food Futures National Research Flagship, 671, Sneydes Road, Werribee, Victoria 3030, Australia
a r t i c l e
i n f o
Article history: Received 4 January 2010 Received in revised form 16 July 2010 Accepted 20 July 2010 Available online 30 July 2010 Keywords: Plant cell-wall Rheology Confocal microscopy Complex modulus Volume fraction
a b s t r a c t The rheological behaviour of plant cell-wall particle dispersions was investigated using dynamic oscillatory measurements. Two starting plant materials, broccoli stem and carrot were used and two types of particles were obtained by mechanically shearing blanched (80 ◦ C, 10 min) or cooked (100 ◦ C, 15 min) plant tissues. Blanching resulted in cell-wall particles made up of a collection of clusters of cells with an average particles size of ∼200 m, while cooking generated nearly all single-cell particles with an average particle size of ∼80 m. The rheological measurements showed that in the range of weight concentrations considered (∼0.5% to ∼8%) the dispersions behaved as elastic materials with the elastic modulus G higher than G within the frequency range (0.01–10 Hz). This study shows that the behaviour of the complex modulus G* as a function of the effective volume fraction can be modelled using different theoretical equations. To do so, it is assumed that below a critical volume fraction c a network of plant cell-wall particles was formed and G* as a function of obeys a power-law relationship. However above c , where the particles were highly packed, G* could be modelled using theoretical equations developed for concentrated emulsions and elastic particle dispersions. Crown Copyright © 2010 Published by Elsevier B.V. All rights reserved.
1. Introduction The plant cell-wall is the major structural component of fruit and vegetables, and along with the turgor pressure, play a key role in contributing to textural properties, which in turn imparts sensory attributes and the eating quality to plant foods [1]. Plant cell-walls are structures that surround and separate cells and provide support for cell expansion and plant growth [2,3]. They are complex and highly sophisticated composite materials made of cellulose, hemicellulose and pectin, that form a scaffold matrix with intertwined structure [4]. The ordered array of cellulose microfibrils providing the ‘steel rod’ like matrix structure of the wall and the hemicellulose forms the link between the cellulose microfibrils providing a strengthening effect with some degree of extensibility to the wall. This matrix is embedded in a pectin network. These pectins are connected by covalent and ionic cross-links to form structural networks that are independent but synergistic with the cellulose–hemicellulose networks [4–6]. Due to their ability to bind large amounts of water, plant cellwalls can be used as swelling and thickening agents to provide desirable food texture [7]. They are also a major contributor to the dietary fibre part of our food intake. Thus, they have both physical and physiological roles in human food sources [8]. By refining existing conventional processes and developing new novel food
∗ Corresponding author. Tel.: +61 3 97313435; fax: +61 3 39731 3200. E-mail addresses: [email protected], [email protected] (Y. Hemar).
manufacturing technologies plant cell-wall materials have potential for manufacturing food products with a healthier composition and without added stabilisers or thickeners. In order to obtain critical information which can be used for product and process performance such as processing efficiency, product physical stability and sensory properties, a fundamental understanding and control of plant cell-wall microstructure and its rheological properties is required. Plant cell-wall particle dispersions can be seen as colloidal dispersions of irregular deformable particles. The rheological properties of the dispersions depend on the particle shape, size distribution, particle concentration and deformability, particle–particle interactions, and hydrodynamic forces arising from the relative motion of particles to the surrounding fluid [9–11]. For example, in a concentrated form such as tomato paste, the viscosity of the system is largely controlled by the deformability of closepacked particles [12,13]. Kunzek and co-workers [14–16] have extensively investigated the rheological behaviour of cell-wall dispersions and reported that concentrated dispersions of cell-wall materials showed dominant elastic properties [15]. They have also reported that the rheological properties of rehydrated apple plant cell-wall dispersions are dependent on the concentration, stiffness and elastic properties of the solid phase, which is primarily determined by the interactions between the solid particles [14]. As in the present study, they investigated the rheological behaviour of rehydrated dispersions of carrot cell-wall materials and also reported that these dispersions showed an elastic behaviour at the concentration of 5% total solids [17]. However, to the best of our
0927-7765/$ – see front matter. Crown Copyright © 2010 Published by Elsevier B.V. All rights reserved. doi:10.1016/j.colsurfb.2010.07.041
462
L. Day et al. / Colloids and Surfaces B: Biointerfaces 81 (2010) 461–467
knowledge, there are no systematic studies on the rheology of plant cell-wall particle dispersions as a function of concentration. Despite renewed interest in recent years in the rheology of plant cell-wall dispersions, a theoretical framework for the interpretation of experimental results is lacking. Existing theoretical models deal with either the rheology of suspensions of hard spherical particles or with dispersions of spherical soft droplets, as found in emulsions where the elasticity of the droplet is due to the oil–water surface tension [18]. In the case of elastic particle dispersions, a theoretical model was first proposed by Evans and Lips [19] and was found to fit qualitatively well for the dispersions of sephadex gel microbeads. However, this model failed to predict accurately the rheological behaviour of concentrated dispersions of model microgel particles of agar with well characterised elasticity [20]. The present work reports on the study of two types of cell-wall particle dispersions obtained using a combination of heating and shear processing. Two starting raw materials, carrot and broccoli (stem only) were used. Particle size and confocal laser scanning microscopy were used to investigate the size and morphology of the resulting cell-wall particles. The aim of the work was to evaluate the effect of particle size, shape and elasticity on the viscoelastic properties of plant cell dispersions of varying solid fractions. The results of the rheological measurements were analysed using existing models, including a model adopted from the viscoelasticity of highly-concentrated emulsions.
The dynamic rheological properties of the suspensions were measured using an Anton Paar Physica controlled-stress rheometer (MCR 300, Physica Meßtechnik GmbH, Stuttgart, Germany) with a double gap geometry DG26.7/T200/AL. The double gap geometry is made of a bob having an internal radius R2 = 12.330 mm and an external radius R3 = 13.330 mm and a cup with an internal radius R1 = 11.913 mm and an external radius R4 = 13.796 mm. This gives an inner gap of 0.417 mm and an outer gap of 0.466 mm. All measurements were performed at least in triplicate at a constant temperature of 25 ◦ C. Frequency-sweep measurements (0.01–10 Hz) were carried out at a constant strain of 0.1%. Strainsweep (0.01–100%) was carried out at a constant frequency of 1 Hz.
2. Materials and methods
2.6. Total solid content
2.1. Plant materials
The total solid content was determined using the AOAC official method 964.22. About 15 g (1 tablespoon) of washed sand was added to a flat bottom metal dish and dried at 102 ◦ C for a minimum of 30 min. The dish was then allowed to cool down in a desiccator before the dispersion sample (about 3 g) was added to each dish, and mixed using a glass rod to ensure an even distribution of the sample over the base of the dish. A small amount of deionised water was added if the sample was difficult to mix. The dish was then placed in a boiling water bath to remove most of the moisture, followed by drying in a 70 ◦ C vacuum oven for 2–3 h with a dry air flow at a maximum pressure of 700 mbar. After drying, the dish was transferred to a desiccator and weighed as soon as the sample reached room temperature to minimize moisture uptake. Two replicates of each sample were analysed.
Fresh broccoli (Brassica oleracea cv. italica) grown during the 2008–2009 summer were kindly supplied by a local farm (Fresh Select, Werribee South, Victoria, Australia). Broccoli crops were hand picked by the farmer on the day of harvest. They were collected from the wholesale plant, transferred to the cool store (4 ◦ C) on site and were used within 4 weeks of storage. Fresh carrots roots (Daucus carota, cv. Kuroda) grown in the southern part of Australia during the 2008–2009 summer were purchased in one batch from a local supermarket. The carrots were then stored at 4 ◦ C and used within 4 weeks. 2.2. Preparation of plant dispersions The broccoli flower heads were removed and only the stems were used for the experiment. The carrot peel was also removed. Both broccoli stems and carrots were cut into ∼2 cm pieces. Water (200 g) was pre-heated to 80 ◦ C or 100 ◦ C. Broccoli or carrot pieces (200 g) were then added to the water and further heated at 80 ◦ C for 10 min or 100 ◦ C for 30 min, and cooled immediately in an ice bath. After cooling, additional water was added to make up for the evaporated volume. Heat-treated broccoli or carrot and the water was then transferred to a kitchen blender (Philips Blender HR 2835 400 W) and blended for a total of 8 min in 1 min intervals. Sodium azide (0.01 wt%) was added to the dispersions to prevent microbial growth. 2.3. Particle size distribution The particle size distribution of the dispersions was measured by laser light scattering using a Malvern Mastersizer 2000 instrument (Malvern Instruments Ltd, Worcestershire, UK). A refractive index of 1.33 for water was used and a refractive index of 1.560 with absorption of 0.1 was used as the optical properties of the particles. The particle calculation was set for irregular particles and the dispersions were measured in duplicate.
2.4. Confocal laser scanning microscopy Confocal laser scanning microscopy (CLSM) was used to characterise the microstructure of the particles in the dispersions. Samples were stained with the fluorescent dye congo red (0.005%), then observed at room temperature under a HC PL APO 20× or a HCX PL APO 63× objective using a Leica TCS SP5 confocal laser scanning microscope (Leica Microsystems, Wetzlar, Germany). The fluorescent dye was excited by an argon 488 nm laser and the emitted light was collected at 544–663 nm. 2.5. Rheology measurements
3. Results 3.1. Microstructural properties The morphologies of carrot plant tissue (Fig. 1A) and broccoli plant tissue (Fig. 1B) from the starting material are made of a cellular architecture (Fig. 1A). Mechanical shear through blending of the blanched plant tissues resulted in large cell cluster particles each containing several individual cells (Fig. 2A and B). The blending of the cooked tissues resulted in many more smaller particles which are in majority made up of single-cells (Fig. 3A and B). For the sake of simplicity, the cell-wall particles obtained after blanching will be referred to as cluster-cell particles and those obtained after cooking will be termed single-cell particles. The shapes of both the single-cell and the cluster-cell particles are not spherical. In fact, for example, the image analysis performed on the confocal micrographs of the carrot dispersions, using analySIS 3.2 software, showed that the mean sphericity of the cluster-cell particles was 0.36 and that of the single-cell particles was 0.27, indicating clearly that these particles are not spherical (spherical particles present a value of a sphericity equal to 1). Particle size measurements performed on all the dispersions showed that, both cluster-cell and single-cell dispersions exhibited
L. Day et al. / Colloids and Surfaces B: Biointerfaces 81 (2010) 461–467
463
Fig. 1. Confocal micrographs showing the effect of heat and shear processes on carrot (A) and broccoli (B) plant cell-wall particle morphologies. Pictures are for: raw plant tissue (1), cluster-cell particles (2) and single-cell particles (3). Scale bar represent 250 m.
a monomodal size distribution (Fig. 2). For the broccoli cluster-cell dispersions the particle size diameter ranged from 5 to 1000 m, with a main peak at 200 m. The particle size distribution of singlecell dispersions also showed a monomodal dispersion ranging from 2 to 800 m and a main peak at 80 m. However, the particle size distribution of the broccoli single-cell dispersion also showed a shoulder in the range 200–800 m, with the majority of the singlecell particles size ranging from 2 to 200 m. Similar particle sizes were obtained for dispersions produced from carrot cell-wall materials. For the carrot cluster-cell dispersion the diameter size of particles ranged from 5 to 600 m, with a main peak at 200 m.
Fig. 3. Elastic G (solid symbols) and loss G (open symbols) at 1 Hz as a function of the applied strain for carrot cluster-cells (A) and single-cells (B) dispersions. Concentrations are: 1 wt% (), 1.5 wt% (䊉), 2wt% () and 2.5 wt% ().
Fig. 2. Particle size distributions of broccoli (, ) and carrot (䊉, ) cell-wall materials. Open symbols are for single-cells and solid symbols are for cluster-cells.
464
L. Day et al. / Colloids and Surfaces B: Biointerfaces 81 (2010) 461–467
The particle size distribution of the carrot single-cell dispersion was also monomodal, with a diameter size ranging from 4 to 400 m and a main peak at 80 m. The fact that the particle size distributions of the single-cell and the cluster-cell dispersions obtained from broccoli stem and carrot raw material are very similar is an indication that for these plant materials the particle size is influenced by the process heat and blending condition rather than the type of starting raw material, indicating material independence. 3.2. Viscoelastic properties The strain-sweep results performed on the cluster-cell and single-cell dispersions are reported in Fig. 3. At low applied strain, the elastic modulus G and the loss modulus G were constant with G higher than G indicating that these dispersions have a solid-like response. This low-strain region (≤2%), where the values of both G and G do not vary is the linear viscoelastic region of the plant dispersions. Upon increasing the applied strain, G started to decrease while G started to increase. Further increase in the strain leads to a decrease in both G and G , to eventually reach a cross-over point after which G became higher than G . This indicates that when a high strain is applied the dispersion flows. The increase followed by the decrease in G observed at intermediate applied strain (a maxima in G could be clearly seen at a strain ≈ 2% in Fig. 3A) was previously reported for different systems such as polymers, colloidal glasses emulsions and gels [21]. Although the exact origin of this peculiar behaviour still to be fully understood, it has been suggested, in the case of concentrated colloidal dispersions, that this could be due to the break-down of aggregates into small clusters which are more dissipative [22]. Qualitatively the same strain-sweep behaviour was observed for both carrot dispersions made of cluster-cells (Fig. 3A) and single-cells (Fig. 3B), although some differences can be seen. For instance, the cross-over strain c at which G = G was higher for the single-cell dispersions (Fig. 3A) than the cluster-cell dispersions (Fig. 3B), indicating that singlecell dispersions i.e. smaller particle size leads to a larger number of strain bearing entanglements in comparison with the cluster-cell dispersion. The frequency-sweep measurements performed on carrot particle dispersions are reported in Fig. 4. These measurements were obtained at a constant strain of 0.1% which is within the linear viscoelastic region. At the concentrations between 1% and 2.5%, for both the cluster-cell and single-cell dispersions, the typical behaviour of a weak gel is exhibited, with G higher than G at all the frequencies, with G lower than G by less than 10-folds [23]. Qualitatively, strain-sweep and frequency-sweep measurements performed on the broccoli cell-wall dispersions yielded similar rheological behaviour to that obtained on carrot cell-wall dispersions (results not shown). In order to understand the effect of the concentration on the rheological behaviour of the cell-wall dispersion, the complex modulus G* (=(G2 + G2 )1/2 ) as a function of the weight concentration c are reported in Fig. 5 for single-cell and cluster-cell dispersions prepared from both carrot (Fig. 5A) and broccoli plant cell-wall materials (Fig. 5B). At low concentrations and for all the plant cell material dispersions, the complex modulus G* increased steeply with c. At high c, only a small increase is observed. The transition from one behaviour of G* as a function of c to another occurs at a critical concentration c*, which depends both on the type and size of plant cell-wall particles (see Fig. 5). For instance, in the case of the cluster-cell dispersion made from carrot cell-wall materials, the transition occurred at ∼2.5%, while for dispersions made of single-cell particles the transition occurred at ∼4.75%. Similarly, in the case of dispersions made of broccoli cell-wall materials, the transition occurs at ∼1.5% for cluster-cell dispersions, and at ∼4% for single-cell dispersions. In addition to the differ-
Fig. 4. Elastic G (solid symbols) and loss G (open symbols) as function of the frequency for carrot cluster-cells (A) and single-cells (B) dispersions. Concentrations are: 1 wt% (), 1.5 wt% (䊉), 2 wt% (), and 2.5 wt% ().
ence in the transition concentration c*, the rate of increase in G* as a function of c is also dependent on the size of the cell-wall particles in the dispersions, and it is higher for dispersions made with cluster-cell particles compared to the dispersions made with single-cell particles. However, despite these differences the plateau value of G* attained at high c is the same for the dispersions made from the same raw cell-wall materials, independent of the particle sizes. 4. Discussion As mentioned in Section 1, despite the industrial relevance of plant cell-wall dispersions, to date their complex rheological behaviour has not been investigated using theoretical models. Although normally the rheological behaviour of dispersions is modelled as a function of the volume fraction occupied by the particles, in the present work we attempted to adapt existing models to understand the complex modulus of the plant cell-wall dispersions as a function of solid concentration. This is because it was not possible to determine accurately the volume fraction from the concentration c. due to the soft nature of the cell-wall particles and their ability to hold water within the plant cell-wall matrix. Conventional methods such as centrifugation cause the water to be transported out of the cell particles leading to changes in the shape and size of the cell-wall particles, thus an inaccurate representative of the volume fraction of particles as to they are suspended in the dispersion. The difficulty to obtain an accurate determination of the volume fraction of soft particles has been previously discussed by Adams et al. [20]. With the aim to compare the behaviour of carrot and broccoli particle dispersions in relation to difference in their size, shape and elastic properties, the solids concentration c** at the plateau value of G* was assumed equivalent to a volume fraction equal to 1. It should be noted that depending on the dispersion,
L. Day et al. / Colloids and Surfaces B: Biointerfaces 81 (2010) 461–467
465
Fig. 6. Complex modulus G* replotted as a function of the effective volume fraction for (A) carrot and (B) and broccoli cell-wall material dispersions. Symbols are: single-cells () and cluster-cells (). Solid lines are obtained using Eqs. (1) and (2). Fig. 5. Complex modulus G* as a function of concentration c for (A) carrot and (B) broccoli CWM dispersions. Symbols are: single-cells () and cluster-cells (䊉). Error bars correspond to standard deviations. c* marks transition concentrations and c** indicates the critical concentration chosen to be equivalent to a volume fraction equal to 1.
a plateau value for G* are in practice often attained at volume fractions lower than 1. Thus, for each dispersion, the relative volume fraction was calculated by dividing c by c**. In the case of the carrot cell-wall particle dispersions, the value of c** used was 4.5% and 7.30% for cluster-cell particles and for the single-cell particles respectively. In the case of the broccoli cell-wall particle dispersions, the value of c** used was 2.90% and 5.20% for cluster-cell particles and for the single-cell particles respectively. The result of this exercise could be seen in Fig. 6 where G* is plotted as a function of the relative volume fraction . From Fig. 6, it can be clearly seen that there are two distinct regimes delimited by a critical volume fraction c . A first regime I, for ≤ c where the G* increase steeply with , and a second regime II, for ≥ c , where the behaviour of G* as a function of increases much less steeply to reach a plateau at high values of . It is likely that regime I is due to the formation of a network made of interconnected cell-wall material particles, which can occur even at relatively low concentrations. As shown in Fig. 7A, cell-wall particles of broccoli at a concentration of 2.5 wt%, are in contact with each other, although they do not occupy the whole space, since some voids can be clearly seen. This could be due to weak interaction, resulting in the aggregation of the particles. Regime II, corresponds to concentrations at which the particles are closely packed as shown in Fig. 7B, in which case the deformable particles
fill the whole space and the viscoelastic response is governed by the elasticity of the particles. To model the behaviour of G* as a function of , at ≤ c , we propose to use a power-law equation: G∗ ∝ ˛
(1)
This equation is generally used to model the elastic behaviour of particlate networks [24]. Note that Eq. (1) is a simplified form of more sophisticated theoretical models involving the fractal dimension of the particulate network and the critical volume fraction at which the network percolates [25,26]. The constant ˛, in Eq. (1), depends on the interactions between the particles and on their shape. Microcal Origin software was used to fit the data and the results are reported in region I of Fig. 6. In the case of carrot cell-wall dispersions, a value of ˛ of 3.0 ± 0.1 and 6.2 ± 0.2 was obtained for the dispersion made of single-cells and cluster-cells respectively; and in the case of broccoli cell-wall dispersions, a value of ˛ of 3.0 ± 0.3 and 6.4 ± 0.5 was obtained for the dispersions made of single-cells and cluster-cells respectively (Table 1). The high value obtained in the case of cluster-cell dispersions, for both carrot and broccoli cell-walls, is due to the non-spherical shape of these particles and the presence of dangling fragments at their surfaces. In addition, the values of ˛ found for both the carrot and broccoli dispersions are similar. When the particle effective volume fractions reached ≥ c , the dispersions present morphologies showing highly packed elastic cell-wall particles (Fig. 7B). The Mason et al. [27] equation which has been used extensively in modelling the elasticity of concentrated emulsions, was therefore applied to the cell-wall dispersions
466
L. Day et al. / Colloids and Surfaces B: Biointerfaces 81 (2010) 461–467
Fig. 7. Confocal micrographs of broccoli single-cell dispersions at (A) 1.9 wt% and (B) 3.8 wt% total solids. Scale bar represent 250 m. Table 1 (2) Critical concentration c** used to obtain the effective volume fraction and different parameters obtained by fitting the experimental data to Eqs. (1)–(3). Note that c and (3) c correspond to the critical volume fraction c obtained using Eqs. (2) and (3) respectively. Carrot
c** [%] ˛ [Pa] c (2) A [Pa] c (3)
Broccoli
Single-cell
Cluster-cell
Single-cell
Cluster-cell
7.3 3.0 ± 0.1
4.5 6.3 ± 0.2
5.2 3.0 ± 0.3
2.9 6.4 ± 0.5
1046 ± 132 0.39 ± 0.03 1833 ± 146 0.56 ± 0.02
1208 ± 127 0.33 ± 0.02 1706 ± 149 0.49 ± 0.02
to relate the behaviour of G* to through: G∗ = 2 ( − c )
(2)
In the case of the concentrated emulsion = 1.77/R, where is the droplet surface tension and R its radius; and c is the volume fraction at which the droplets start to touch and the emulsion starts to exhibit an elastic modulus. In the present work we assume that is a constant depending on the elasticity of the cell-wall particle. For the same range of volume fraction ( ≥ c ) the empirical equation proposed by Adams et al. [20] was also used:
G = A
1−
1/3 c
(3)
where A is an adjustable parameter and c has the same definition as above. The results of these two models are reported in region II of Fig. 6, where they are plotted as dashed line for Eq. (2) and as a solid line for Eq. (3). Using Eq. (2) the best fitting parameters were c = 0.39 ± 0.03 and = 1406 ± 132 Pa for carrot cell-wall materials, and c = 0.33 ± 0.02 and = 1208 ± 127 Pa for broccoli cell-wall materials. Using Eq. (3), c = 0.56 ± 0.02 and A = 1833 ± 146 Pa for carrot cell-wall materials, and c = 0.49 ± 0.02 and A = 1706 ± 149 Pa for broccoli cell-wall materials (Table 1). This shows that for each model and for both dispersions made from carrot or broccoli cell-wall materials, the values of c are close, which is not surprising since both particle size measurements and microscopy showed that the particles from these two different raw materials had similar particle morphologies and size distributions. The values of and A were also comparable for the dispersions made with broccoli and those made from carrot raw materials. This would indicate that the elasticity of the cell particle materials obtained from carrot or broccoli is not markedly different. Note however that the ratio /A (≈0.6 for carrot dispersions and ≈0.7 for
broccoli dispersions) is close for the two raw materials, confirming that although these two parameters and A do not have the same value, they are likely to represent the same physical properties, i.e. elasticity, of the individual cell unit. It needs to be emphasised that the values reported in Table 1 are under the assumption that the concentration c** is equivalent to a relative volume fraction of 1. Qualitatively, both the model of highly-concentrated emulsion (Eq. (2)) and the empirical Eq. (3) gave a reasonable description of G* as a function of , although some differences could be seen. For instance, Eq. (2) allows modelling of a wider range, and at high the values obtained by Eq. (3) increases less than those obtained by Eq. (2). These discrepancies between the results obtained using Eqs. (2) and (3) and the experimental results could be due on one hand to the highly deformable nature of the cell-wall material particles, where the cell-wall particles could change the shapes and perceived sizes (e.g. less spherical, or smaller sizes due to the water in the cell being forced out through packing). In this case, both the packing of the particles and their elasticity can be markedly affected, which in turn can affect the behaviour of G* as a function of [28]. On the other hand, the equations used in this study were developed for dispersions made of model monodisperse single spherical elastic particles. It would be worth repeating the study on dispersions made of pure single-cells, which could be obtained for example by enzymatic processing.
5. Conclusion In the present study the rheological behaviour of cell-wall particle dispersions was investigated using dynamic oscillatory measurements. Two starting plant materials, broccoli stem and carrot root were used and two types of particles were obtained by mechanically shearing blanched or cooked plant materials. The
L. Day et al. / Colloids and Surfaces B: Biointerfaces 81 (2010) 461–467
blanched material resulted in cell-wall particles mainly made up of a collection of individual cluster-cells with a larger average particle size, and the cooked material which generated nearly all single-cell particles. The rheological measurements showed that in the range of weight concentrations considered (∼0.5% to ∼8%) the dispersion behaved as an elastic material similarly to weak gels. The complex elastic modulus G* was found to increase with the weight concentration c, where below a critical concentration c* a steep increase in G* was observed and above c* the increase in G* tended to plateau at very high concentrations. To model this rheological behaviour, G* is first plotted as a function of a relative volume fraction which is determined by assuming that G* reaches a maximum value when = 1. For values below a critical volume fraction c , the cell-wall particles are assumed to form a network and G* as a function of is modelled by a power-law equation. For values above c, an empirical equation developed for deformable elastic particles and a second equation developed for concentrated emulsions were applied to investigate the behaviour of the plant cell-wall particle dispersions. The results demonstrate that these equations can qualitatively model the behaviour of G* as a function of with some limitations due to the complexity of plant cell-wall particle deformation and subsequent changes in particle shape, size and effective packing when they reached high concentrations. It would be of interest to compare the rheological behaviour of dispersions made of pure single-cells with maintained turgor pressure and known elasticity to that of dispersions made of emulsion droplets and/or monodisperse elastic particles. In addition, the challenge to adequately estimate the phase volumes of soft deformable particles as a function of concentrations still remains and should be investigated in the future. Acknowledgements We thank Ms. Lilliane Talbot (Polytech’Montpellier, France) for her assistance with sample preparation and rheological measurements of broccoli dispersions as part of her internship with CSIRO Food and Nutritional Sciences, and Prof. Mike Gidley (University of Queensland, Australia) and Ingrid Appelqvist (CSIRO Food Futures Flagship) for fruitful discussions.
467
References [1] K.W. Waldron, M.L. Parker, A.C. Smith, Comprehensive Reviews In Food Science And Food Safety 2 (2003) 128–146. [2] N.C. Carpita, D.M. Gibeaut, The Plant Journal 3 (1993) 1–30. [3] D.J. Cosgrove, Nature Reviews Molecular Cell Biology 6 (2005) 850–861. [4] D.J. Cosgrove, Annual Review of Cell and Developmental Biology 13 (1997) 171–201. [5] C. Somerville, S. Bauer, G. Brininstool, M. Facette, T. Hamann, J. Milne, E. Osborne, A. Paredez, S. Persson, T. Raab, S. Vorwerk, H. Youngs, Science 306 (2004) 2206–2211. [6] N.C. Carpita, D.M. Gibeaut, Plant Journal 3 (1993) 1–30. [7] P.J. Harris, B.G. Smith, International Journal of Food Science and Technology 41 (2006) 129–143. [8] G.J. McDougall, I.M. Morrison, D. Stewart, J.R. Hillman, Journal of the Science of Food and Agriculture 70 (1996) 133–150. [9] W. Pabst, E. Gregorova, C. Berthold, Journal of the European Ceramic Society 26 (2006) 149–160. [10] D.B. Genovese, J.E. Lozano, M.A. Rao, Journal of Food Science 72 (2007) R11–R20. [11] J.R. Stokes, W.J. Frith, Soft Matter 4 (2008) 1133–1140. [12] C. Valencia, M.C. Sanchez, A. Ciruelos, A. Latorre, J.M. Madiedo, C. Gallegos, Food Research International 36 (2003) 911–919. [13] B. Yoo, M.A. Rao, Journal of Texture Studies 25 (1994) 421–436. [14] H. Kunzek, H. Opel, B. Senge, Zeitschrift für Lebensmittel - Untersuchung und Forschung A-Food Research Technology 205 (1997) 193–203. [15] S. Vetter, H. Kunzek, European Food Research and Technology 216 (2003) 129–137. [16] S. Schalow, H. Kunzek, European Food Research and Technology 219 (2004) 329–340. [17] C. Pickardt, G. Dongowski, H. Kunzek, European Food Research and Technology 219 (2004) 229–239. [18] M.D. Lacasse, G.S. Grest, D. Levine, T.G. Mason, D.A. Weitz, Physical Review Letters 76 (1996) 3448–3451. [19] I.D. Evans, A. Lips, Journal of the Chemical Society, Faraday Transactions 86 (1990) 3413–3417. [20] S. Adams, W.J. Frith, J.R. Stokes, Journal of Rheology 48 (2004) 1195–1213. [21] K. Miyazaki, H.M. Wyss, D.A. Weitz, D.R. Reichman, Europhysics Letters 75 (2006) 915–921. [22] F. Yziquel, P.J. Carreau, M. Moan1, P.A. Tanguy, Journal of Non-Newtonian Fluid Mechanics 86 (1999) 133–155. [23] R. Lapasin, S. Pricl, Rheology of Industrial Polysaccharides: Theory and Applications, Aspen Publishers Inc., 1995. [24] J.A. Yanez, E. Laarz, L. Bergstrom, Journal of Colloid and Interface Science 209 (1999) 162–172. [25] W.H. Shih, W.Y. Shih, S.I. Kim, J. Liu, I.A. Aksay, Physical Review A 42 (1990) 4772–4779. [26] A.G. Marangoni, Physical Review B 62 (2000) 13951–13955. [27] T.G. Mason, J. Bibette, D.A. Weitz, Physical Review Letters 75 (1995) 2051–2054. [28] W.J. Frith, A. Lips, Advances in Colloid and Interface Science 61 (1995) 161–189.
Contents lists available at ScienceDirect
Colloids and Surfaces B: Biointerfaces journal homepage: www.elsevier.com/locate/colsurfb
Dynamic rheological properties of plant cell-wall particle dispersions Li Day, Mi Xu, Sofia K. Øiseth, Leif Lundin, Yacine Hemar ∗ CSIRO Food & Nutrition Sciences, and Food Futures National Research Flagship, 671, Sneydes Road, Werribee, Victoria 3030, Australia
a r t i c l e
i n f o
Article history: Received 4 January 2010 Received in revised form 16 July 2010 Accepted 20 July 2010 Available online 30 July 2010 Keywords: Plant cell-wall Rheology Confocal microscopy Complex modulus Volume fraction
a b s t r a c t The rheological behaviour of plant cell-wall particle dispersions was investigated using dynamic oscillatory measurements. Two starting plant materials, broccoli stem and carrot were used and two types of particles were obtained by mechanically shearing blanched (80 ◦ C, 10 min) or cooked (100 ◦ C, 15 min) plant tissues. Blanching resulted in cell-wall particles made up of a collection of clusters of cells with an average particles size of ∼200 m, while cooking generated nearly all single-cell particles with an average particle size of ∼80 m. The rheological measurements showed that in the range of weight concentrations considered (∼0.5% to ∼8%) the dispersions behaved as elastic materials with the elastic modulus G higher than G within the frequency range (0.01–10 Hz). This study shows that the behaviour of the complex modulus G* as a function of the effective volume fraction can be modelled using different theoretical equations. To do so, it is assumed that below a critical volume fraction c a network of plant cell-wall particles was formed and G* as a function of obeys a power-law relationship. However above c , where the particles were highly packed, G* could be modelled using theoretical equations developed for concentrated emulsions and elastic particle dispersions. Crown Copyright © 2010 Published by Elsevier B.V. All rights reserved.
1. Introduction The plant cell-wall is the major structural component of fruit and vegetables, and along with the turgor pressure, play a key role in contributing to textural properties, which in turn imparts sensory attributes and the eating quality to plant foods [1]. Plant cell-walls are structures that surround and separate cells and provide support for cell expansion and plant growth [2,3]. They are complex and highly sophisticated composite materials made of cellulose, hemicellulose and pectin, that form a scaffold matrix with intertwined structure [4]. The ordered array of cellulose microfibrils providing the ‘steel rod’ like matrix structure of the wall and the hemicellulose forms the link between the cellulose microfibrils providing a strengthening effect with some degree of extensibility to the wall. This matrix is embedded in a pectin network. These pectins are connected by covalent and ionic cross-links to form structural networks that are independent but synergistic with the cellulose–hemicellulose networks [4–6]. Due to their ability to bind large amounts of water, plant cellwalls can be used as swelling and thickening agents to provide desirable food texture [7]. They are also a major contributor to the dietary fibre part of our food intake. Thus, they have both physical and physiological roles in human food sources [8]. By refining existing conventional processes and developing new novel food
∗ Corresponding author. Tel.: +61 3 97313435; fax: +61 3 39731 3200. E-mail addresses: [email protected], [email protected] (Y. Hemar).
manufacturing technologies plant cell-wall materials have potential for manufacturing food products with a healthier composition and without added stabilisers or thickeners. In order to obtain critical information which can be used for product and process performance such as processing efficiency, product physical stability and sensory properties, a fundamental understanding and control of plant cell-wall microstructure and its rheological properties is required. Plant cell-wall particle dispersions can be seen as colloidal dispersions of irregular deformable particles. The rheological properties of the dispersions depend on the particle shape, size distribution, particle concentration and deformability, particle–particle interactions, and hydrodynamic forces arising from the relative motion of particles to the surrounding fluid [9–11]. For example, in a concentrated form such as tomato paste, the viscosity of the system is largely controlled by the deformability of closepacked particles [12,13]. Kunzek and co-workers [14–16] have extensively investigated the rheological behaviour of cell-wall dispersions and reported that concentrated dispersions of cell-wall materials showed dominant elastic properties [15]. They have also reported that the rheological properties of rehydrated apple plant cell-wall dispersions are dependent on the concentration, stiffness and elastic properties of the solid phase, which is primarily determined by the interactions between the solid particles [14]. As in the present study, they investigated the rheological behaviour of rehydrated dispersions of carrot cell-wall materials and also reported that these dispersions showed an elastic behaviour at the concentration of 5% total solids [17]. However, to the best of our
0927-7765/$ – see front matter. Crown Copyright © 2010 Published by Elsevier B.V. All rights reserved. doi:10.1016/j.colsurfb.2010.07.041
462
L. Day et al. / Colloids and Surfaces B: Biointerfaces 81 (2010) 461–467
knowledge, there are no systematic studies on the rheology of plant cell-wall particle dispersions as a function of concentration. Despite renewed interest in recent years in the rheology of plant cell-wall dispersions, a theoretical framework for the interpretation of experimental results is lacking. Existing theoretical models deal with either the rheology of suspensions of hard spherical particles or with dispersions of spherical soft droplets, as found in emulsions where the elasticity of the droplet is due to the oil–water surface tension [18]. In the case of elastic particle dispersions, a theoretical model was first proposed by Evans and Lips [19] and was found to fit qualitatively well for the dispersions of sephadex gel microbeads. However, this model failed to predict accurately the rheological behaviour of concentrated dispersions of model microgel particles of agar with well characterised elasticity [20]. The present work reports on the study of two types of cell-wall particle dispersions obtained using a combination of heating and shear processing. Two starting raw materials, carrot and broccoli (stem only) were used. Particle size and confocal laser scanning microscopy were used to investigate the size and morphology of the resulting cell-wall particles. The aim of the work was to evaluate the effect of particle size, shape and elasticity on the viscoelastic properties of plant cell dispersions of varying solid fractions. The results of the rheological measurements were analysed using existing models, including a model adopted from the viscoelasticity of highly-concentrated emulsions.
The dynamic rheological properties of the suspensions were measured using an Anton Paar Physica controlled-stress rheometer (MCR 300, Physica Meßtechnik GmbH, Stuttgart, Germany) with a double gap geometry DG26.7/T200/AL. The double gap geometry is made of a bob having an internal radius R2 = 12.330 mm and an external radius R3 = 13.330 mm and a cup with an internal radius R1 = 11.913 mm and an external radius R4 = 13.796 mm. This gives an inner gap of 0.417 mm and an outer gap of 0.466 mm. All measurements were performed at least in triplicate at a constant temperature of 25 ◦ C. Frequency-sweep measurements (0.01–10 Hz) were carried out at a constant strain of 0.1%. Strainsweep (0.01–100%) was carried out at a constant frequency of 1 Hz.
2. Materials and methods
2.6. Total solid content
2.1. Plant materials
The total solid content was determined using the AOAC official method 964.22. About 15 g (1 tablespoon) of washed sand was added to a flat bottom metal dish and dried at 102 ◦ C for a minimum of 30 min. The dish was then allowed to cool down in a desiccator before the dispersion sample (about 3 g) was added to each dish, and mixed using a glass rod to ensure an even distribution of the sample over the base of the dish. A small amount of deionised water was added if the sample was difficult to mix. The dish was then placed in a boiling water bath to remove most of the moisture, followed by drying in a 70 ◦ C vacuum oven for 2–3 h with a dry air flow at a maximum pressure of 700 mbar. After drying, the dish was transferred to a desiccator and weighed as soon as the sample reached room temperature to minimize moisture uptake. Two replicates of each sample were analysed.
Fresh broccoli (Brassica oleracea cv. italica) grown during the 2008–2009 summer were kindly supplied by a local farm (Fresh Select, Werribee South, Victoria, Australia). Broccoli crops were hand picked by the farmer on the day of harvest. They were collected from the wholesale plant, transferred to the cool store (4 ◦ C) on site and were used within 4 weeks of storage. Fresh carrots roots (Daucus carota, cv. Kuroda) grown in the southern part of Australia during the 2008–2009 summer were purchased in one batch from a local supermarket. The carrots were then stored at 4 ◦ C and used within 4 weeks. 2.2. Preparation of plant dispersions The broccoli flower heads were removed and only the stems were used for the experiment. The carrot peel was also removed. Both broccoli stems and carrots were cut into ∼2 cm pieces. Water (200 g) was pre-heated to 80 ◦ C or 100 ◦ C. Broccoli or carrot pieces (200 g) were then added to the water and further heated at 80 ◦ C for 10 min or 100 ◦ C for 30 min, and cooled immediately in an ice bath. After cooling, additional water was added to make up for the evaporated volume. Heat-treated broccoli or carrot and the water was then transferred to a kitchen blender (Philips Blender HR 2835 400 W) and blended for a total of 8 min in 1 min intervals. Sodium azide (0.01 wt%) was added to the dispersions to prevent microbial growth. 2.3. Particle size distribution The particle size distribution of the dispersions was measured by laser light scattering using a Malvern Mastersizer 2000 instrument (Malvern Instruments Ltd, Worcestershire, UK). A refractive index of 1.33 for water was used and a refractive index of 1.560 with absorption of 0.1 was used as the optical properties of the particles. The particle calculation was set for irregular particles and the dispersions were measured in duplicate.
2.4. Confocal laser scanning microscopy Confocal laser scanning microscopy (CLSM) was used to characterise the microstructure of the particles in the dispersions. Samples were stained with the fluorescent dye congo red (0.005%), then observed at room temperature under a HC PL APO 20× or a HCX PL APO 63× objective using a Leica TCS SP5 confocal laser scanning microscope (Leica Microsystems, Wetzlar, Germany). The fluorescent dye was excited by an argon 488 nm laser and the emitted light was collected at 544–663 nm. 2.5. Rheology measurements
3. Results 3.1. Microstructural properties The morphologies of carrot plant tissue (Fig. 1A) and broccoli plant tissue (Fig. 1B) from the starting material are made of a cellular architecture (Fig. 1A). Mechanical shear through blending of the blanched plant tissues resulted in large cell cluster particles each containing several individual cells (Fig. 2A and B). The blending of the cooked tissues resulted in many more smaller particles which are in majority made up of single-cells (Fig. 3A and B). For the sake of simplicity, the cell-wall particles obtained after blanching will be referred to as cluster-cell particles and those obtained after cooking will be termed single-cell particles. The shapes of both the single-cell and the cluster-cell particles are not spherical. In fact, for example, the image analysis performed on the confocal micrographs of the carrot dispersions, using analySIS 3.2 software, showed that the mean sphericity of the cluster-cell particles was 0.36 and that of the single-cell particles was 0.27, indicating clearly that these particles are not spherical (spherical particles present a value of a sphericity equal to 1). Particle size measurements performed on all the dispersions showed that, both cluster-cell and single-cell dispersions exhibited
L. Day et al. / Colloids and Surfaces B: Biointerfaces 81 (2010) 461–467
463
Fig. 1. Confocal micrographs showing the effect of heat and shear processes on carrot (A) and broccoli (B) plant cell-wall particle morphologies. Pictures are for: raw plant tissue (1), cluster-cell particles (2) and single-cell particles (3). Scale bar represent 250 m.
a monomodal size distribution (Fig. 2). For the broccoli cluster-cell dispersions the particle size diameter ranged from 5 to 1000 m, with a main peak at 200 m. The particle size distribution of singlecell dispersions also showed a monomodal dispersion ranging from 2 to 800 m and a main peak at 80 m. However, the particle size distribution of the broccoli single-cell dispersion also showed a shoulder in the range 200–800 m, with the majority of the singlecell particles size ranging from 2 to 200 m. Similar particle sizes were obtained for dispersions produced from carrot cell-wall materials. For the carrot cluster-cell dispersion the diameter size of particles ranged from 5 to 600 m, with a main peak at 200 m.
Fig. 3. Elastic G (solid symbols) and loss G (open symbols) at 1 Hz as a function of the applied strain for carrot cluster-cells (A) and single-cells (B) dispersions. Concentrations are: 1 wt% (), 1.5 wt% (䊉), 2wt% () and 2.5 wt% ().
Fig. 2. Particle size distributions of broccoli (, ) and carrot (䊉, ) cell-wall materials. Open symbols are for single-cells and solid symbols are for cluster-cells.
464
L. Day et al. / Colloids and Surfaces B: Biointerfaces 81 (2010) 461–467
The particle size distribution of the carrot single-cell dispersion was also monomodal, with a diameter size ranging from 4 to 400 m and a main peak at 80 m. The fact that the particle size distributions of the single-cell and the cluster-cell dispersions obtained from broccoli stem and carrot raw material are very similar is an indication that for these plant materials the particle size is influenced by the process heat and blending condition rather than the type of starting raw material, indicating material independence. 3.2. Viscoelastic properties The strain-sweep results performed on the cluster-cell and single-cell dispersions are reported in Fig. 3. At low applied strain, the elastic modulus G and the loss modulus G were constant with G higher than G indicating that these dispersions have a solid-like response. This low-strain region (≤2%), where the values of both G and G do not vary is the linear viscoelastic region of the plant dispersions. Upon increasing the applied strain, G started to decrease while G started to increase. Further increase in the strain leads to a decrease in both G and G , to eventually reach a cross-over point after which G became higher than G . This indicates that when a high strain is applied the dispersion flows. The increase followed by the decrease in G observed at intermediate applied strain (a maxima in G could be clearly seen at a strain ≈ 2% in Fig. 3A) was previously reported for different systems such as polymers, colloidal glasses emulsions and gels [21]. Although the exact origin of this peculiar behaviour still to be fully understood, it has been suggested, in the case of concentrated colloidal dispersions, that this could be due to the break-down of aggregates into small clusters which are more dissipative [22]. Qualitatively the same strain-sweep behaviour was observed for both carrot dispersions made of cluster-cells (Fig. 3A) and single-cells (Fig. 3B), although some differences can be seen. For instance, the cross-over strain c at which G = G was higher for the single-cell dispersions (Fig. 3A) than the cluster-cell dispersions (Fig. 3B), indicating that singlecell dispersions i.e. smaller particle size leads to a larger number of strain bearing entanglements in comparison with the cluster-cell dispersion. The frequency-sweep measurements performed on carrot particle dispersions are reported in Fig. 4. These measurements were obtained at a constant strain of 0.1% which is within the linear viscoelastic region. At the concentrations between 1% and 2.5%, for both the cluster-cell and single-cell dispersions, the typical behaviour of a weak gel is exhibited, with G higher than G at all the frequencies, with G lower than G by less than 10-folds [23]. Qualitatively, strain-sweep and frequency-sweep measurements performed on the broccoli cell-wall dispersions yielded similar rheological behaviour to that obtained on carrot cell-wall dispersions (results not shown). In order to understand the effect of the concentration on the rheological behaviour of the cell-wall dispersion, the complex modulus G* (=(G2 + G2 )1/2 ) as a function of the weight concentration c are reported in Fig. 5 for single-cell and cluster-cell dispersions prepared from both carrot (Fig. 5A) and broccoli plant cell-wall materials (Fig. 5B). At low concentrations and for all the plant cell material dispersions, the complex modulus G* increased steeply with c. At high c, only a small increase is observed. The transition from one behaviour of G* as a function of c to another occurs at a critical concentration c*, which depends both on the type and size of plant cell-wall particles (see Fig. 5). For instance, in the case of the cluster-cell dispersion made from carrot cell-wall materials, the transition occurred at ∼2.5%, while for dispersions made of single-cell particles the transition occurred at ∼4.75%. Similarly, in the case of dispersions made of broccoli cell-wall materials, the transition occurs at ∼1.5% for cluster-cell dispersions, and at ∼4% for single-cell dispersions. In addition to the differ-
Fig. 4. Elastic G (solid symbols) and loss G (open symbols) as function of the frequency for carrot cluster-cells (A) and single-cells (B) dispersions. Concentrations are: 1 wt% (), 1.5 wt% (䊉), 2 wt% (), and 2.5 wt% ().
ence in the transition concentration c*, the rate of increase in G* as a function of c is also dependent on the size of the cell-wall particles in the dispersions, and it is higher for dispersions made with cluster-cell particles compared to the dispersions made with single-cell particles. However, despite these differences the plateau value of G* attained at high c is the same for the dispersions made from the same raw cell-wall materials, independent of the particle sizes. 4. Discussion As mentioned in Section 1, despite the industrial relevance of plant cell-wall dispersions, to date their complex rheological behaviour has not been investigated using theoretical models. Although normally the rheological behaviour of dispersions is modelled as a function of the volume fraction occupied by the particles, in the present work we attempted to adapt existing models to understand the complex modulus of the plant cell-wall dispersions as a function of solid concentration. This is because it was not possible to determine accurately the volume fraction from the concentration c. due to the soft nature of the cell-wall particles and their ability to hold water within the plant cell-wall matrix. Conventional methods such as centrifugation cause the water to be transported out of the cell particles leading to changes in the shape and size of the cell-wall particles, thus an inaccurate representative of the volume fraction of particles as to they are suspended in the dispersion. The difficulty to obtain an accurate determination of the volume fraction of soft particles has been previously discussed by Adams et al. [20]. With the aim to compare the behaviour of carrot and broccoli particle dispersions in relation to difference in their size, shape and elastic properties, the solids concentration c** at the plateau value of G* was assumed equivalent to a volume fraction equal to 1. It should be noted that depending on the dispersion,
L. Day et al. / Colloids and Surfaces B: Biointerfaces 81 (2010) 461–467
465
Fig. 6. Complex modulus G* replotted as a function of the effective volume fraction for (A) carrot and (B) and broccoli cell-wall material dispersions. Symbols are: single-cells () and cluster-cells (). Solid lines are obtained using Eqs. (1) and (2). Fig. 5. Complex modulus G* as a function of concentration c for (A) carrot and (B) broccoli CWM dispersions. Symbols are: single-cells () and cluster-cells (䊉). Error bars correspond to standard deviations. c* marks transition concentrations and c** indicates the critical concentration chosen to be equivalent to a volume fraction equal to 1.
a plateau value for G* are in practice often attained at volume fractions lower than 1. Thus, for each dispersion, the relative volume fraction was calculated by dividing c by c**. In the case of the carrot cell-wall particle dispersions, the value of c** used was 4.5% and 7.30% for cluster-cell particles and for the single-cell particles respectively. In the case of the broccoli cell-wall particle dispersions, the value of c** used was 2.90% and 5.20% for cluster-cell particles and for the single-cell particles respectively. The result of this exercise could be seen in Fig. 6 where G* is plotted as a function of the relative volume fraction . From Fig. 6, it can be clearly seen that there are two distinct regimes delimited by a critical volume fraction c . A first regime I, for ≤ c where the G* increase steeply with , and a second regime II, for ≥ c , where the behaviour of G* as a function of increases much less steeply to reach a plateau at high values of . It is likely that regime I is due to the formation of a network made of interconnected cell-wall material particles, which can occur even at relatively low concentrations. As shown in Fig. 7A, cell-wall particles of broccoli at a concentration of 2.5 wt%, are in contact with each other, although they do not occupy the whole space, since some voids can be clearly seen. This could be due to weak interaction, resulting in the aggregation of the particles. Regime II, corresponds to concentrations at which the particles are closely packed as shown in Fig. 7B, in which case the deformable particles
fill the whole space and the viscoelastic response is governed by the elasticity of the particles. To model the behaviour of G* as a function of , at ≤ c , we propose to use a power-law equation: G∗ ∝ ˛
(1)
This equation is generally used to model the elastic behaviour of particlate networks [24]. Note that Eq. (1) is a simplified form of more sophisticated theoretical models involving the fractal dimension of the particulate network and the critical volume fraction at which the network percolates [25,26]. The constant ˛, in Eq. (1), depends on the interactions between the particles and on their shape. Microcal Origin software was used to fit the data and the results are reported in region I of Fig. 6. In the case of carrot cell-wall dispersions, a value of ˛ of 3.0 ± 0.1 and 6.2 ± 0.2 was obtained for the dispersion made of single-cells and cluster-cells respectively; and in the case of broccoli cell-wall dispersions, a value of ˛ of 3.0 ± 0.3 and 6.4 ± 0.5 was obtained for the dispersions made of single-cells and cluster-cells respectively (Table 1). The high value obtained in the case of cluster-cell dispersions, for both carrot and broccoli cell-walls, is due to the non-spherical shape of these particles and the presence of dangling fragments at their surfaces. In addition, the values of ˛ found for both the carrot and broccoli dispersions are similar. When the particle effective volume fractions reached ≥ c , the dispersions present morphologies showing highly packed elastic cell-wall particles (Fig. 7B). The Mason et al. [27] equation which has been used extensively in modelling the elasticity of concentrated emulsions, was therefore applied to the cell-wall dispersions
466
L. Day et al. / Colloids and Surfaces B: Biointerfaces 81 (2010) 461–467
Fig. 7. Confocal micrographs of broccoli single-cell dispersions at (A) 1.9 wt% and (B) 3.8 wt% total solids. Scale bar represent 250 m. Table 1 (2) Critical concentration c** used to obtain the effective volume fraction and different parameters obtained by fitting the experimental data to Eqs. (1)–(3). Note that c and (3) c correspond to the critical volume fraction c obtained using Eqs. (2) and (3) respectively. Carrot
c** [%] ˛ [Pa] c (2) A [Pa] c (3)
Broccoli
Single-cell
Cluster-cell
Single-cell
Cluster-cell
7.3 3.0 ± 0.1
4.5 6.3 ± 0.2
5.2 3.0 ± 0.3
2.9 6.4 ± 0.5
1046 ± 132 0.39 ± 0.03 1833 ± 146 0.56 ± 0.02
1208 ± 127 0.33 ± 0.02 1706 ± 149 0.49 ± 0.02
to relate the behaviour of G* to through: G∗ = 2 ( − c )
(2)
In the case of the concentrated emulsion = 1.77/R, where is the droplet surface tension and R its radius; and c is the volume fraction at which the droplets start to touch and the emulsion starts to exhibit an elastic modulus. In the present work we assume that is a constant depending on the elasticity of the cell-wall particle. For the same range of volume fraction ( ≥ c ) the empirical equation proposed by Adams et al. [20] was also used:
G = A
1−
1/3 c
(3)
where A is an adjustable parameter and c has the same definition as above. The results of these two models are reported in region II of Fig. 6, where they are plotted as dashed line for Eq. (2) and as a solid line for Eq. (3). Using Eq. (2) the best fitting parameters were c = 0.39 ± 0.03 and = 1406 ± 132 Pa for carrot cell-wall materials, and c = 0.33 ± 0.02 and = 1208 ± 127 Pa for broccoli cell-wall materials. Using Eq. (3), c = 0.56 ± 0.02 and A = 1833 ± 146 Pa for carrot cell-wall materials, and c = 0.49 ± 0.02 and A = 1706 ± 149 Pa for broccoli cell-wall materials (Table 1). This shows that for each model and for both dispersions made from carrot or broccoli cell-wall materials, the values of c are close, which is not surprising since both particle size measurements and microscopy showed that the particles from these two different raw materials had similar particle morphologies and size distributions. The values of and A were also comparable for the dispersions made with broccoli and those made from carrot raw materials. This would indicate that the elasticity of the cell particle materials obtained from carrot or broccoli is not markedly different. Note however that the ratio /A (≈0.6 for carrot dispersions and ≈0.7 for
broccoli dispersions) is close for the two raw materials, confirming that although these two parameters and A do not have the same value, they are likely to represent the same physical properties, i.e. elasticity, of the individual cell unit. It needs to be emphasised that the values reported in Table 1 are under the assumption that the concentration c** is equivalent to a relative volume fraction of 1. Qualitatively, both the model of highly-concentrated emulsion (Eq. (2)) and the empirical Eq. (3) gave a reasonable description of G* as a function of , although some differences could be seen. For instance, Eq. (2) allows modelling of a wider range, and at high the values obtained by Eq. (3) increases less than those obtained by Eq. (2). These discrepancies between the results obtained using Eqs. (2) and (3) and the experimental results could be due on one hand to the highly deformable nature of the cell-wall material particles, where the cell-wall particles could change the shapes and perceived sizes (e.g. less spherical, or smaller sizes due to the water in the cell being forced out through packing). In this case, both the packing of the particles and their elasticity can be markedly affected, which in turn can affect the behaviour of G* as a function of [28]. On the other hand, the equations used in this study were developed for dispersions made of model monodisperse single spherical elastic particles. It would be worth repeating the study on dispersions made of pure single-cells, which could be obtained for example by enzymatic processing.
5. Conclusion In the present study the rheological behaviour of cell-wall particle dispersions was investigated using dynamic oscillatory measurements. Two starting plant materials, broccoli stem and carrot root were used and two types of particles were obtained by mechanically shearing blanched or cooked plant materials. The
L. Day et al. / Colloids and Surfaces B: Biointerfaces 81 (2010) 461–467
blanched material resulted in cell-wall particles mainly made up of a collection of individual cluster-cells with a larger average particle size, and the cooked material which generated nearly all single-cell particles. The rheological measurements showed that in the range of weight concentrations considered (∼0.5% to ∼8%) the dispersion behaved as an elastic material similarly to weak gels. The complex elastic modulus G* was found to increase with the weight concentration c, where below a critical concentration c* a steep increase in G* was observed and above c* the increase in G* tended to plateau at very high concentrations. To model this rheological behaviour, G* is first plotted as a function of a relative volume fraction which is determined by assuming that G* reaches a maximum value when = 1. For values below a critical volume fraction c , the cell-wall particles are assumed to form a network and G* as a function of is modelled by a power-law equation. For values above c, an empirical equation developed for deformable elastic particles and a second equation developed for concentrated emulsions were applied to investigate the behaviour of the plant cell-wall particle dispersions. The results demonstrate that these equations can qualitatively model the behaviour of G* as a function of with some limitations due to the complexity of plant cell-wall particle deformation and subsequent changes in particle shape, size and effective packing when they reached high concentrations. It would be of interest to compare the rheological behaviour of dispersions made of pure single-cells with maintained turgor pressure and known elasticity to that of dispersions made of emulsion droplets and/or monodisperse elastic particles. In addition, the challenge to adequately estimate the phase volumes of soft deformable particles as a function of concentrations still remains and should be investigated in the future. Acknowledgements We thank Ms. Lilliane Talbot (Polytech’Montpellier, France) for her assistance with sample preparation and rheological measurements of broccoli dispersions as part of her internship with CSIRO Food and Nutritional Sciences, and Prof. Mike Gidley (University of Queensland, Australia) and Ingrid Appelqvist (CSIRO Food Futures Flagship) for fruitful discussions.
467
References [1] K.W. Waldron, M.L. Parker, A.C. Smith, Comprehensive Reviews In Food Science And Food Safety 2 (2003) 128–146. [2] N.C. Carpita, D.M. Gibeaut, The Plant Journal 3 (1993) 1–30. [3] D.J. Cosgrove, Nature Reviews Molecular Cell Biology 6 (2005) 850–861. [4] D.J. Cosgrove, Annual Review of Cell and Developmental Biology 13 (1997) 171–201. [5] C. Somerville, S. Bauer, G. Brininstool, M. Facette, T. Hamann, J. Milne, E. Osborne, A. Paredez, S. Persson, T. Raab, S. Vorwerk, H. Youngs, Science 306 (2004) 2206–2211. [6] N.C. Carpita, D.M. Gibeaut, Plant Journal 3 (1993) 1–30. [7] P.J. Harris, B.G. Smith, International Journal of Food Science and Technology 41 (2006) 129–143. [8] G.J. McDougall, I.M. Morrison, D. Stewart, J.R. Hillman, Journal of the Science of Food and Agriculture 70 (1996) 133–150. [9] W. Pabst, E. Gregorova, C. Berthold, Journal of the European Ceramic Society 26 (2006) 149–160. [10] D.B. Genovese, J.E. Lozano, M.A. Rao, Journal of Food Science 72 (2007) R11–R20. [11] J.R. Stokes, W.J. Frith, Soft Matter 4 (2008) 1133–1140. [12] C. Valencia, M.C. Sanchez, A. Ciruelos, A. Latorre, J.M. Madiedo, C. Gallegos, Food Research International 36 (2003) 911–919. [13] B. Yoo, M.A. Rao, Journal of Texture Studies 25 (1994) 421–436. [14] H. Kunzek, H. Opel, B. Senge, Zeitschrift für Lebensmittel - Untersuchung und Forschung A-Food Research Technology 205 (1997) 193–203. [15] S. Vetter, H. Kunzek, European Food Research and Technology 216 (2003) 129–137. [16] S. Schalow, H. Kunzek, European Food Research and Technology 219 (2004) 329–340. [17] C. Pickardt, G. Dongowski, H. Kunzek, European Food Research and Technology 219 (2004) 229–239. [18] M.D. Lacasse, G.S. Grest, D. Levine, T.G. Mason, D.A. Weitz, Physical Review Letters 76 (1996) 3448–3451. [19] I.D. Evans, A. Lips, Journal of the Chemical Society, Faraday Transactions 86 (1990) 3413–3417. [20] S. Adams, W.J. Frith, J.R. Stokes, Journal of Rheology 48 (2004) 1195–1213. [21] K. Miyazaki, H.M. Wyss, D.A. Weitz, D.R. Reichman, Europhysics Letters 75 (2006) 915–921. [22] F. Yziquel, P.J. Carreau, M. Moan1, P.A. Tanguy, Journal of Non-Newtonian Fluid Mechanics 86 (1999) 133–155. [23] R. Lapasin, S. Pricl, Rheology of Industrial Polysaccharides: Theory and Applications, Aspen Publishers Inc., 1995. [24] J.A. Yanez, E. Laarz, L. Bergstrom, Journal of Colloid and Interface Science 209 (1999) 162–172. [25] W.H. Shih, W.Y. Shih, S.I. Kim, J. Liu, I.A. Aksay, Physical Review A 42 (1990) 4772–4779. [26] A.G. Marangoni, Physical Review B 62 (2000) 13951–13955. [27] T.G. Mason, J. Bibette, D.A. Weitz, Physical Review Letters 75 (1995) 2051–2054. [28] W.J. Frith, A. Lips, Advances in Colloid and Interface Science 61 (1995) 161–189.