Structural And Mechanical Characterization Of Nanoclay-reinforced Agarose Nanocomposites..ingles

INSTITUTE OF PHYSICS PUBLISHING

NANOTECHNOLOGY

Nanotechnology 16 (2005) 2020–2029

doi:10.1088/0957-4484/16/10/006

Structural and mechanical characterization of nanoclay-reinforced agarose nanocomposites Xiaodong Li1,3,4 , Hongsheng Gao1 , Wally A Scrivens2,3 , Dongling Fei2 , Vivek Thakur2 , Michael A Sutton1, Anthony P Reynolds1 and Michael L Myrick2 1

Department of Mechanical Engineering, University of South Carolina, 300 Main Street, Columbia, SC 29208, USA 2 Department of Chemistry and Biochemistry, University of South Carolina, 631 Sumter Street, Columbia, SC 29208, USA E-mail: [email protected] and [email protected]

Received 21 April 2005, in final form 1 June 2005 Published 3 August 2005 Online at stacks.iop.org/Nano/16/2020 Abstract Nanoclay-reinforced agarose nanocomposite films with varying weight concentration ranging from 0 to 80% of nanoclay were prepared, and structurally and mechanically characterized. Structural characterization was carried out by transmission electron microscopy (TEM), scanning electron microscopy (SEM) and atomic force microscopy (AFM). It was found that pre-exfoliated clay platelets were re-aggregated into particles (stacked platelets) during the composite preparation process. Each particle consists of approximately 11 clay platelets stacked together. The nanoclay particles were homogeneously dispersed within an agarose matrix. The clay particles were oriented with a slight preference of the stacked platelets being parallel to the composite film’s surface within the low loading composite films. Mechanical properties of the nanocomposite films were measured by tensile, three-point bending and nanoindentation tests. Mechanical testing results show that nanoclays provide a significant enhancement to the tensile modulus and strength. For the 60% clay nanocomposite, its elastic modulus increases up to 21.4 GPa, which is five times higher than that of the agarose matrix. Based upon the structural characterization, a theoretical model has been developed to simulate the mechanical behaviour of the nanoclay-reinforced polymer composites.

1. Introduction Clay reinforced polymer composites have spurred great interest since nylon 6–clay nanocomposites were developed and found applications in the automotive industry by the researchers at Toyota in the early 1990s [1–3]. With only 4.2 wt% clay dispersed in nylon 6 matrix, the elastic modulus increased by as much as 100% and the heat distortion temperature increased by 80 ◦ C compared to those of the unfilled nylon 6 matrix [1, 2]. To date, 3 Authors to whom any correspondence should be addressed. 4 www.me.sc.edu/research/nano

0957-4484/05/102020+10$30.00 © 2005 IOP Publishing Ltd

various clay reinforced nanocomposites or hybrids have been synthesized such as epoxy–clay nanocomposites [4, 5], polyimide–clay nanocomposites [6] and polyurethane–clay nanocomposites [7, 8], to name a few. Previous studies have found that nanoclay reinforced polymeric composites offer many improved properties over pristine polymers in tensile modulus and strength [2, 9–11], thermal properties and heat distortion temperature [2, 12], resistance to flammability [13], and reduced permeability to liquids or gases [14]. A common observation from these studies is that the magnitude of improvement is, to a large extent, dependent upon the state of dispersion of nanoclays in the matrix. Only when clays

Printed in the UK

2020

Structural and mechanical characterization of nanoclay-reinforced agarose nanocomposites

are homogeneously dispersed and randomly orientated within the matrix, i.e. nanoclays are completely exfoliated, could the enhancement be more significant. However, the clay dispersion effect on the resulting mechanical properties is not well understood, especially for nanocomposites with a higher clay concentration. The term clay refers to a class of materials made up of layered silicates, among which montmorillonite (MMT) is the most commonly used clay in the synthesis of nanocomposites. The MMT is composed of layered, nanometre thick platelets, and has the same crystalline structure as mica [15]. Although clay has many attractive properties such as high elastic modulus and wide availability at low cost, its application is limited largely because the clay dispersion cannot be easily achieved [16]. The full exfoliation of nanoclays in nylon 6 matrix was first achieved by Usuki et al [17] in 1993 and is believed to be a fundamental breakthrough for utilizing clay in practical applications. Although nanoclays have proven to be a prominent reinforcement, to date, only a few polymer–clay nanocomposites with well enhanced strength or stiffness have been reported. Agarose is a naturally occurring polysaccharide and has been mainly used as a gel medium for the separation of DNA and protein molecules via electrophoresis [18]. With an ever-increasing demand for advanced materials, enhancing the mechanical properties of natural polymers like agarose to make new, high-performance composites is important and compelling. The synthesis methods of polymer–clay nanocomposites are generally categorized into three groups according to the starting materials and the processing techniques: intercalation of polymer or pre-polymer from solution, in situ intercalative polymerization and melt intercalation [19]. In the present study, a group of montmorillonitereinforced agarose nanocomposite films with clay concentration up to 80 wt% were prepared and characterized, with the focus on their structural and mechanical properties. The objectives of this study were to investigate the effects of clay concentration and dispersion on the resulting mechanical properties, and to develop a theoretical model for predicting the mechanical properties of nanoclay reinforced polymer composites.

2. Experimental details 2.1. Sample preparation A 1 wt% solution of sodium montmorillonite (Cloisite Na + , Southern Clay Products) was dispersed in distilled water. The dispersion was sonicated (Branson 5510) for an hour and then centrifuged (Sorvall® Legand T) at 3500 rpm for 75 min. The transparent supernatant solution was decanted into a silated container. The weight concentration of this exfoliated montmorillonite (MMT) nanoclay suspension was 0.7%. Nanocomposites were prepared by a gelation method, i.e. agarose (CertifiedTM PCR) was added to the MMT suspension at a concentration of
1 wt%. The mixtures were first heated to 100 ◦ C to dissolve the agarose and then poured into Petri dishes. Agarose gels were formed when the mixtures were cooled to room temperature. These gels were dried at room temperature in the air, resulting in nanocomposite films after the gels dehydrated and collapsed. Water content

of the films was calculated based on the weight lost before and after placing in the vacuum oven at 100 ◦ C over 24 h. The water contents of pure agarose and 40% MMT/agarose films are 16.58% and 12.77% respectively. Different loadings (i.e. concentration) of MMT/agarose films were prepared by varying the ratio of MMT to the amount of dry agarose added. The thickness of the resulting MMT/agarose films is in the range from 60 to 80 µm. For structural characterization of the nanoclays by atomic force microscopy (AFM), the resulting exfoliated MMT suspension was further diluted with distilled water and then a drop of solution was put onto the surface of newly cleaved mica. 2.2. Structural and mechanical characterization The morphology of the nanoclays used in this study was characterized using a Dimension 3100 AFM system (Veeco Metrology Group) operated in the tapping mode. Crosssections of the nanocomposite films were studied with a Philips EM 430 ST transmission electron microscope (TEM). The morphology, characteristic length and thickness of the nanoclays before dispersion were investigated and compared with those of the clays within the nanocomposites. Mechanical properties of the composites were measured using an UMT-2 mechanical tester (CETR Inc.). Both tensile and three-point bending tests were preformed. The composite films were first cut into strips and then attached onto a frame, which was used for protecting the specimen from tension or twisting during handling. The frame was cut away once the specimen was mounted on the tester. The instrument is capable of applying up to 5000 mN force with a force resolution of 1.3 mN. The strain rate was 0.5 mm min−1 for all tensile tests. Fracture surfaces of the tensile specimens were examined using a Philips XL30 field emission environmental scanning electron microscope (ESEM) operated at 30 kV. Three-point bending tests were preformed using the same tester. A high-resolution load transducer with a force resolution of 0.1 mN was used. The bending elastic modulus was calculated by L3 d P (1) E= 48 · I dδ where L is the suspended length and I is the second moment of inertia. d P/dδ is the slope of the initial linear portion of the bending load–deflection curve. Nanoindentation tests were performed using a Triboscope nanomechanical testing system (Hysitron Inc.) in conjunction with the Veeco AFM system. The Hysitron nanoindenter monitors and records the load and displacement of the indenter, a diamond Berkovich three-sided pyramid, with a force resolution of about 1 nN and displacement resolution of about 0.1 nm. Hardness and elastic modulus were calculated from the recorded load–displacement data. A typical indentation experiment consists of four steps: approaching the surface; loading to peak load; holding the indenter at peak load for 5 s; finally unloading completely. The holding segment was included to avoid the influence of creep on the unloading characteristics since the unloading curve was used to obtain the elastic modulus of a material. For more details on nanoindentation experimental techniques, please see [20, 21]. 2021

X Li et al

Nanoindentation hardness is defined as the indentation load divided by the projected contact area of the indentation. It is the mean pressure that a material will support under load. From the load–displacement curve, hardness can be obtained at the peak indentation load as H=

Pmax A

(2)

where A is the projected contact area. For an indenter with a known geometry such as the Berkovich tip used in this study, the projected contact area is a function of contact depth, which is measured by the nanoindenter in situ during indentation [20, 21]. Therefore, the projected contact area A can be measured and calculated directly from the indentation displacement. The indentation elastic modulus was calculated using the Oliver–Pharr data analysis procedure [22] beginning by fitting the unloading curve to a power-law relation. The unloading stiffness can be derived from the slope of the initial portion of the unloading curve, S = d p/dh. Based on relationships developed by Sneddon [23] for the indentation of an elastic half space by any punch that can be described as a solid of revolution of a smooth function, a geometry independent relation involving contact stiffness, contact area, and elastic modulus can be derived as follows:
S π (3) Er = 2β A where β is a constant which depends on the geometry of the indenter (β = 1.034 for a Berkovich indenter) [20, 21], and E r is the reduced elastic modulus which accounts for the fact that elastic deformation occurs in both the sample and the indenter. E r is given by 1 − ν 2 1 − νi2 1 = + Er E Ei

(4)

where E and ν are the elastic modulus and Poisson’s ratio for the sample, respectively, and E i and νi are the same quantities of the indenter. For the diamond indenter, E i = 1140 GPa and νi = 0.07 [20, 21].

3. Results and discussion Figure 1 shows the representative TEM and AFM images as well as the morphology characteristics of the nanoclays used in this study. Direct observation of individual clay platelets or platelet stacks confirms that MMT has the layered structural characteristic with an irregular shape, which is in good agreement with previous studies [24]. The measured platelet thickness is about 1 nm, which is also consistent with the previously published value [25, 26]. The maximum length measured is 853 nm and the maximum width 430 nm. Since there are no standards regarding representing the length and width of irregular shape, in this study, the characteristic length of a platelet is represented by the average of its length and width that were measured in two perpendicular directions and correspond well to its shape. The histograms of platelet length and thickness are shown in figures 1(d) and (f), respectively. Figure 1(e) shows the histogram of the number of platelets per 2022

particle. It is clear that the majority of clay particles are singlelayer platelets before dispersion, as shown in figure 1(e). The average length of clay particles is 174.7 nm and the aspect ratio is 108.3. Previous studies have shown that the interlayer spacing is 0.96 nm for natural (unpolymerized) MMT with a platelet thickness of 0.94 nm [25, 26]. According to XRD analysis, the interlayer spacing of intercalated MMT within the agarose matrix is averaged to be 1.52 nm. In addition, the tensile modulus of individual MMT platelets is 178 GPa and its density is 2.83 g cm−3 [25]. Figure 2 shows the low magnification TEM images of the nanocomposites with 30, 60, and 80% clay loadings. In figure 2(b), it is shown more clearly that clay particles were, to a large extent, dispersed uniformly and oriented randomly within the matrix. Nevertheless, the clay particles within lower clay loading nanocomposites show a slight tendency of the stacked platelets to be parallel to the nanocomposite film surfaces. This tendency is diminished with increasing clay concentration. The higher the clay concentration, the smaller the deviation from uniform orientation found. When the clay weight concentration is higher than 60%, thicker platelet stacks can be found and the clay dispersion is less uniform with further increasing the clay content. The high magnification TEM images of the nanocomposites with 0, 30, 60, and 80% clay loadings are shown in figure 3. These images demonstrate clearly that the nanoclays within the agarose matrix were mainly in the form of platelet stacks composed of approximately 11 platelets in a stack and thus existed as intercalated particles. According to results measured by AFM, however, about 75% of clay particles were exfoliated single-layer platelets. It is evident that the pre-exfoliated MMT platelets were re-aggregated during the dispersion and polymerization processes. The high clay concentration in the matrix as well as the large aspect ratio and high surface-tovolume ratio of the clay platelets make it difficult for the agarose to wrap the platelets and keep them apart from one another. The characteristic length and thickness of the nanoclays within the nanocomposites were measured by TEM. Figure 4 shows the histogram of particle length and number of platelets per particle for the clay particles within the matrix. Comparison of figure 4 with figures 1(d) and (e) shows that both particle length and thickness are increased during the composite preparation process. As a result, the aspect ratio is decreased from 108.3 down to 16. The representative tensile stress–strain curves are shown in figure 5. It is noted that both tensile elastic modulus and strength increase with an increase in clay concentration and reach the peak values at 60% clay loading. With a further increase in clay concentration the modulus and strength decrease. The ductility decreases with increasing clay concentration, as shown in figure 6, indicating an increase in brittleness in the high clay loading composites. For a two-phase composite, there is an optimum volume fraction or volume ratio that provides good bonding between the two phases for load transfer. When clay loading is higher than a critical volume/weight percentage, the direct consequence is that the clay particles within nanocomposites cannot be completely exfoliated into platelets or even fully

Structural and mechanical characterization of nanoclay-reinforced agarose nanocomposites

Figure 1. Representative (a) TEM and (b) AFM images of the nanoclays before dispersion, (c) cross-section of a clay platelet, (d) histogram of particle length before dispersion, (e) histogram of platelet number per particle, and (f) histogram of particle thickness before dispersion. (This figure is in colour only in the electronic version)

Figure 2. Low magnification TEM images of nanocomposites with (a) 30% clay, (b) 60% clay, and (c) 80% clay concentration.

intercalated. TEM observation of the high clay loading composites has proven that when the clay concentration is higher than 60% there is less available agarose matrix material for intercalating into the nanoclays. Thus good bonding between the clay and agarose matrix can hardly be formed for these nanocomposites. Therefore, the agarose–

clay nanocomposites with clay loading higher than 60% have no practical meaning unless better dispersion techniques are developed. Figure 7 shows the representative load–deflection curves of three-point bending tests. The load was normalized with respect to the cross-section area of the specimen. By 2023

X Li et al

Figure 3. High magnification TEM images of (a) agarose matrix and nanocomposites with (b) 30% clay loading, (c) 60% clay loading, and (d) 80% clay loading.

200

(a)

Total clay number =378 Average length =236.61 nm

150

0.1

Stress (MPa)

Frequency

0.15

0.05

100

0% clay 30% clay 60% clay 80% clay

50 0

Frequency

0.5

40

100

(b)

160

220 280 340 Particle length (nm)

400

460

0

Total number =378 Average number per particle n=11.13

0

0.05

0.1

0.15

0.2

Strain (mm/mm)

Figure 5. Representative stress–strain curves of nanocomposites with 0, 30%, 60%, and 80% clay in the matrix.

0.25

0

7

9

11 13 15 17 No. of platelets per particle

19

Figure 4. Histogram of (a) length of clay particles and (b) number of platelets per particle.

normalizing, the slope of the load–deflection curve is directly correlated to the magnitude of the tensile modulus. It can be seen that the elastic modulus of the nanocomposite 2024

increases with an increase in clay concentration and reaches the maximum value at 60% clay concentration. Then the elastic modulus decreases with a further increase in clay concentration. The elastic moduli as functions of clay concentration measured by tensile and bending tests are plotted in figure 8(a). It is clear that the elastic moduli measured by tensile and bending tests are in good agreement for the nanocomposites with clay concentrations less than 60%. The elastic modulus reaches the maximum value of 21.4 ± 2.5 GPa at 60% clay loading, which is increased sixfold compared with that of the agarose matrix. It has been reported that human bone

Structural and mechanical characterization of nanoclay-reinforced agarose nanocomposites

40 Elastic modulus (GPa)

0.25

Ductility

0.2 0.15 0.1 0.05

30 20 10 0

0

0

20

40

60

80

100

Tensile tests Bending tests

0

Clay concentration (wt. %)

Normalized load (mN/mm2 )

100 80

0 % clay 30 % clay 60 % clay 80 % clay

100

20 40 60 80 Clay concentration (wt. %)

100

250 Tensile stress (MPa)

Figure 6. Ductility (percentage elongation at failure) as a function of clay concentration.

20 40 60 80 Clay concentration (wt. %)

200 150 100 50

60

0 40

Figure 8. (a) Elastic modulus as a function of clay concentration obtained by tensile and three-point bending tests, and (b) tensile strength as a function of clay concentration.

20 0 0

0

0.01

0.02 0.03 Displacement (mm)

0.04

Figure 7. Representative normalized load–deflection curves of bending tests.

has an elastic modulus of 11.4–21.2 GPa [27–29]. The nanoclay reinforced agarose nanocomposites have comparable mechanical properties with human bone. This may extend the application of agarose-based material to orthopaedic and other bioengineering areas. The tensile strength, for the composites with clay loading ranging from 40 to 60%, is increased by as much as 50% compared with that of agarose matrix, as shown in figure 8(b). The maximum tensile strength was found to be 183 MPa for the nanocomposite with about 50–60% clay loading. Noting, the wide range of clay loadings is very useful for tailoring the mechanical properties according to the design requirements. When clay concentration is higher than 60%, however, both elastic modulus and tensile strength decrease drastically with a further increase in clay concentration, indicating unfeasible volume fraction and poor interfaces between the agarose matrix and clay particles. Tang et al [30] developed the nanostructured polymer clay films, named artificial nacre, using a layer-by-layer assembly process in which 5 min adsorption of organic layers (polyelectrolytes) and 10 min dispersion of inorganic layers (clay MMT) were processed sequentially with a rinsing process in deionized water between every two sequences. The 2 min rinsing processes were used to remove the irregular adsorbed platelets loosely attached to the clay layer in order to obtain a better interface between the organic layer and inorganic clay platelets. During this process, the clay solution was refreshed

every 20th deposition cycle and other solutions also refreshed appropriately. A thickness of 4.9 µm was obtained for the typical 200-multilayer films. The measured elastic moduli of the artificial nacres were 10 ± 2 and 13 ± 2 GPa for 100and 200-layered films, respectively. The maximum tensile strength of these films was 109 MPa and was reported to have approached that of real nacre. By comparison of the elastic modulus of the artificial nacre and the nanocomposite films in this study, it is quite clear that the mechanical properties of the agarose–clay composite films are superior to those of the artificial nacre. From a synthesis point of view, the easy and simple route for the preparation of agarose–clay nanocomposites used in the present study has profound merits, which should find more industrial applications. Figure 9 shows the representative SEM images of fracture surfaces of the nanocomposites with 0, 40, 60, and 80% clay loadings. It can be seen that the fracture surface became increasingly rougher with an increase in clay concentration. For the composites with clay concentrations less than 60%, no debris particles were found on the fracture surfaces, indicating a strong bonding between the clay particles and the matrix. Clay-like debris particles were found on the fracture surfaces of the 80% clay composite. This further confirms that the agarose matrix is less available to intercalate into nanoclays for the 80% clay nanocomposite, thereby resulting in poor bonding between the nanoclay particles and the agarose matrix. Figure 10 shows the hardness and elastic modulus of the nanocomposites measured by nanoindentation tests. Both the nanoindentation hardness and elastic modulus decrease in a random manner as clay concentration is increased. It should be noted that the tensile and three-point bending tests measured 2025

X Li et al

Figure 9. Representative SEM images of the fracture surfaces of (a) agarose matrix and nanocomposites with (b) 40% clay, (c) 60% clay, and (d) 80% clay.

10

Hardness Elastic modulus

1.5

8 6

1 4 0.5 0

2 0

20

40 60 Percentage of clay

80

Elastic modulus (GPa)

Hardness (GPa)

2

0 100

Figure 10. Nanoindentation hardness and elastic modulus as a function of clay concentration.

the modulus in the direction parallel to the nanocomposite films, whereas the nanoindentation tests measured the elastic modulus in the direction perpendicular to the composite films. The nanoindenter probed a small volume of material and the indenter tip might or might not encounter the nanoclay particles; even if the indenter tip encountered the nanoclay particles, the nanoclay particles beneath the indenter tip might be bent down together with the indenter [31]. On the other hand, the nanoindentation elastic/plastic deformation zone is much smaller compared with that of the global tensile and bending tests, and there were no or only a few nanoclay particles in the nanoindentation deformation zone. Long-range load transfer was limited.

4. Modelling of elastic modulus The overall strategy in developing a model is to establish the relation of the physical properties between a composite 2026

and its individual components. For conventional composite systems, the theoretical frameworks have been well developed for predicting various properties of a composite based on the properties of its individual components. These theories are useful for evaluating the contribution of individual components and for optimizing the overall performance of a composite based on the matrix and filler moduli, volume fraction, filler aspect ratio and orientation etc. In all these theories, the underlying assumptions are (1) all components in a composite will act independently and (2) the interaction or bonding between the matrix and fillers is strong enough for load transfer. The two well known composite theories are Halpin– Tsai theory [32–34] and Mori–Tanaka theory [35]. Both theories have received considerable attention in the composite community [25, 26, 36–40]. However, none of them can be applied directly to the nanoclay reinforced nanocomposite materials, in which the roles of aspect ratio, dispersion, and orientation of the fillers are more critical compared with conventional composites. Halpin theory [32–34] has been widely used for predicting the elastic modulus of unidirectional fibre reinforced composites as functions of the fibre volume fraction and aspect ratio. The longitudinal and transverse elastic moduli, i.e. E 11 and E 22 , as shown in figure 11, are expressed in a general form: 1 + ξ ηVf E = Em 1 − ηVf

(5)

where E and E m are the elastic moduli of the composite and the matrix, respectively. Vf is the volume fraction of the filler. η is given by Ef −1 E (6) η = Emf +ξ Em

Structural and mechanical characterization of nanoclay-reinforced agarose nanocomposites Fibre

Clay platelet

1

2

2 3 Loading direction 1 2 3

to construct the model. AFM observation of the nanoclays shows that the rectangular shape with the length to width ratio of about 1.5 could best represent clay platelets. In general, for nanocomposite systems with large quantity of clay fillers, aspect ratio R and shape parameter ξ should be calculated by the average length and thickness, as given by

1

3 Composite modulus Fibres Platelets E11=Ep E11=Ep E 22=Et E22=Ep E 33=Et E33=Et

Shape parameter Fibres Platelets ξ =2(l/d) ξ =2(l/t) ξ =2 ξ =2(w/t) ξ =2 ξ =2

Figure 11. Illustration of the fibre and platelet filler and shape parameters for each case for Halpin composite theory. E p and E t are the moduli in directions parallel (longitudinal direction) and perpendicular (traverse direction) to the major axis of the fillers, respectively. l is the length of the fibre or platelet, d is the diameter of the fibre, and w and t are the width and thickness of the clay platelet.

where E f is the elastic modulus of the filler. The tensile modulus of an individual MMT platelet is 178 GPa [25] and the measured elastic modulus of agarose polymer was 3.5 GPa. ξ is the shape parameter depending on filler geometry and loading direction, as shown in figure 11. Ashton et al [34] found that ξ = 2(l/d) can be used to calculate the longitudinal modulus E 11 , where l and d are the characteristic length and diameter of the fibre, respectively. When ξ = 2, on the other hand, the same equation can be applied for evaluating the transverse modulus E 22 [34]. Halpin–Tsai composite theory is consistent with the rules of composites in specific cases such as isostrain and isostress. For example, when ξ → ∞, equation (5) reduces to the rule of mixtures for the isostrain case, i.e. E = Vf E f + Vm E m

(7)

where Vm is the volume fraction of the matrix and Vf + Vm = 1. Conversely, when ξ → 0, equation (5) reduces to the inverse rule of mixtures for the isostress case, i.e. Vf Vm 1 + . = E Ef Em

(8)

ξ = 2R = 2 ·

 ¯ l t¯

where l¯ and t¯ are the average length and thickness of clay fillers, respectively. Both l¯ and t¯ can be evaluated by statistically analysing the TEM images of the composites. In general, TEM images with magnifications in the range of 60 000× to 80 000× can provide data accurate enough to represent the length and thickness of the clay particles within the matrix [25, 41]. The evaluation could be done manually or with the help of some specific software. In either way, the clay platelets or particles shown in the TEM images are numbered and with all their information such as length, orientation, number of platelets per particle or thickness collected. Thus, the histogram plots of particle length and platelet numbers per particle can be plotted, as shown in figure 4. Alternatively, the averaged clay particle thickness can also be measured by x-ray diffraction (XRD) [42–44]. Nevertheless, it is difficult to measure the full widths at half-maximum of the 001 reflection accurately from those wide-angle XRD curves in this study. Thus, the characteristic length, thickness and aspect ratio of the clay particles were measured by TEM observation. In addition to the aspect ratio and shape parameter, orientation of the clay particles has a significant effect on modulus. Practically, almost all composites have some extent of filler misalignment. Previous studies have revealed that the deviation from unidirectional reinforcement would result in sizable reductions in composite modulus. van Es et al [45] studied the nanocomposites reinforced by both fully exfoliated fibre and platelet fillers, and established the following equations to quantitatively calculate the elastic moduli: fibres = 0.184E p + 0.816E t (10) E composites platelets

More conveniently, the Halpin equations retain the same form for discontinuous cylindrical fibres and platelet fillers. The shape parameter, ξ , however, is required to change from 2(l/d) to 2(l/t) for the platelet filler, where l is the length of the fibre or the platelet, d is the diameter of the fibre and t the thickness of the platelet. According to the characteristic differences between fibre and platelet, it is evident that the fibre reinforces only in the longitudinal direction while the platelet can reinforce in directions that are parallel to the platelet surface. Figure 11 illustrates the fibre and platelet filler and shape parameters for each case when the Halpin– Tsai composite theory is applied. The clay fillers, which may be in the forms of platelets or platelet stacks (particles), were often simulated by lamellar squares or circular discs with varying thickness [25, 37–39]. In fact, clay platelets cannot be simply represented by any shape, as shown in the AFM image in figure 1. Nevertheless, it is required to represent clay platelets by a proper shape in order

(9)

E composites = 0.49E p + 0.51E t

(11)

where E p and E t are the composite moduli in the directions parallel and perpendicular to the major axis of the filler, respectively. These two equations show clearly that for completely exfoliated nanocomposites the platelet filler provides higher reinforcement than the fibre filler. As mentioned above, the aspect ratio of clay particles can be measured by either TEM observation or XRD. For the fully exfoliated polymer–clay nanocomposites, the thickness of the clay fillers is the thickness of the MMT platelet. However, the complete exfoliation of clay particles is hard to achieve as with the present study, in which almost all clay particles exist as intercalated platelet stacks within the matrix. The number of clay platelets and measured particle thickness tc can be calculated by n=

tc − tp +1 d001

(12) 2027

X Li et al

Aspect ratio

30

(a)

20

10

0

0

20 40 60 80 Clay concentration (wt.%)

Elastic modulus (GPa)

40

(b)

30

100

Tensile tests Bending tests Simulation

20 10 0

0

20 40 60 80 Clay concentration ( wt.%)

100

Figure 12. (a) Plot of clay aspect ratio as a function of clay concentration, and (b) elastic modulus as a function of clay concentration obtained by tensile tests, three-point bending tests, and theoretical modelling results.

where n is the number of platelets per particle and d001 is the interlayer spacing of the clay particle, which may be evaluated from high resolution TEM images. Otherwise, wide-angle XRD analysis can accurately measure the interlayer spacing of intercalated clay within nanocomposites. tp = 0.94 nm is the MMT platelet thickness [25]. According to the data from TEM observation, the aspect ratios of clay particles within the nanocomposites were evaluated to be 16.27, 15.08, and 8.83 for the nanocomposites with 30, 60, and 80% clay loadings, respectively. Since the aspect ratio decreases linearly with increasing clay concentration in the range from 30 to 60%, the linear interpolation was applied to estimate the aspect ratio of clay particles for the nanocomposites with 5–30% clay loadings. Figure 12(a) shows the curve of the aspect ratio as a function of clay concentration. Note that the aspect ratio decreases sharply when the clay loading is higher than 60%, which indicates the deterioration of clay enhancement and is consistent with the sharp decrease in mechanical properties. Based on structural characterization, the composites studied can be modelled by a three-phase composite system, i.e. the clay platelet stacks, the agarose intercalated into these platelet stacks and agarose matrix. For different clay loadings, only the density of the MMT platelet stacks is changed. This model can be solved in two steps. The platelet stacks and agarose intercalated within these stacks can be first regarded as localized two-phase composites or as equivalent reinforcements, whose elastic modulus is given by equation (7). Then, the equivalent reinforcements and agarose 2028

matrix could compose of the resulting two-phase composites. Halpin–Tsai composite theory and equation (11) can be applied to the second two-phase composite model. It should be noted that the simulated elastic moduli are available only for the nanocomposites with clay loadings up to 60%. When clay loading is beyond 60%, the interface between the nanoclay and agarose matrix will be degraded and hence cannot be simulated by any existing model. This approach should be more reasonable because all phases within nanocomposites were represented properly and each sub-model can be evaluated accurately. In the second step, the Halpin–Tsai equations were applied by substituting all the data corresponding to the equivalent reinforcement, such as η , ξ  aspect ratio R  , volume fraction Vf , and resulting E f of the equivalent reinforcement. However, further exploration of this approach revealed that the final modelling results can only predict the elastic modulus for the nanocomposite with clay weight concentration up to 30%. For higher clay loading nanocomposites, the simulation results were increasingly higher compared with the tensile and bending test results, which are in good agreement with each other. Therefore, the final simulation results for those nanocomposites with clay loading higher than 30% need to be corrected according to the experimental results. The correction factor was fitted to be a function of the clay volume fraction, i.e. f = (1 − Vf ) for all nanocomposites with clay weight concentration higher than 30%, in this study. The correction factor is interpreted to compensate the incompleteness of intercalation caused by the combination of high surface-to-volume ratio of clay platelets and the high clay concentration. Finally, the elastic moduli simulated, as well as those measured by tensile and bending tests, are plotted in figure 12(b). If the clay weight concentration is less than about 5%, it is possible that the nanoclays exist as fully exfoliated platelets or as the combination of exfoliated platelets and intercalated platelet stacks. It is easy to model the completely exfoliated nanocomposites. For the polymer–clay nanocomposites with mixed morphology, i.e. the combination of exfoliation and intercalation, theoretically, it can be analysed as multiphase composites. Firstly, the composite system is required to evaluate the degree of exfoliation by TEM and XRD analysis. Secondly, the exfoliated clay platelets and matrix can form a two-phase composite as the equivalent matrix phase. Thirdly, the intercalated platelet stacks and the polymer intercalated within these stacks are treated as previously for the equivalent reinforcing phase. Finally, the resulting two phases are composed of the final composites.

5. Conclusion Nanoclay-reinforced agarose nanocomposite films with varying weight concentration ranging from 0 to 80% of clay were prepared and characterized. The pre-exfoliated clay platelets were re-aggregated into clay particles during the composite preparation process. Each particle consists of approximately 11 clay platelets stacked together. The nanoclay particles were largely homogeneously dispersed within the agarose matrix. The clay particles were oriented with a slight preference of the stacked platelets being

Structural and mechanical characterization of nanoclay-reinforced agarose nanocomposites

parallel to the composite film surfaces within the low loading composite films. It is concluded that the nanoclay gives a remarkable enhancement to the elastic modulus in the direction parallel to the nanocomposite film surfaces. A sixfold increase in elastic modulus has been achieved for the nanocomposite with 60% clay concentration. The tensile strength increased by as much as 50% over that of the agarose matrix for the nanocomposite with 40–60% clay concentration. Nevertheless, the ductility decreases as clay concentration is increased. The methodology for modelling polymer–clay nanocomposites with different morphologies such as the combination of exfoliation and intercalation of nanoclays has been developed. The developed nanoclay-reinforced agarose nanocomposites should find more applications.

Acknowledgments Financial support for this study was provided by the National Aeronautical and Space Administration through grants NASA NCC5-174 and NASA NAG-1-03017, the University of South Carolina NanoCenter, NASA/EPSCoR grant, and the National Science Foundation (grant number EPS-0296165 and CMS0201345). The content of this information does not necessary reflect the position or policy of the Government and no official endorsement should be inferred.

References [1] Kojima Y, Usuki A, Kawasumi M, Okada A, Kurauchi T and Kamigaito O 1993 J. Appl. Polym. Sci. 49 1259–64 [2] Kojima Y, Usuki A, Kawasumi M, Okada A, Fukushima Y, Kurauchi T and Kamigaito O 1993 J. Mater. Res. 8 1185–9 [3] Okada A and Usuki A 1995 Mater. Sci. Eng. C 3 109–15 [4] Wang Z and Pinnavaia T J 1998 Chem. Mater. 10 1820–6 [5] Kornmann X, Lindberg H and Berglund L A 2001 Polymer 42 1303–10 [6] Agag T, Koga T and Takeichi T 2001 Polymer 42 3399–408 [7] Wang Z and Pinnavaia T J 1998 Chem. Mater. 10 3769–71 [8] Chen T K, Tien Y I and Wei K H 2000 Polymer 41 1345–53 [9] Lan T and Pinnavaia T J 1998 Chem. Mater. 6 2216–9 [10] Svoboda P, Zeng C C, Wang H, Lee L J and Tomasko D L 2002 J. Appl. Polym. Sci. 85 1562–70 [11] Yang F and Nelson G L 2004 J. Appl. Polym. Sci. 91 3844–50 [12] Lan T, Kaviratna P D and Pinnavaia T J 1995 Chem. Mater. 7 2144–50 [13] Gilman J W, Jackson C L, Morgan A B, Harris R, Manias E, Giannelis E P, Wuthenow M, Hilton D and Phillips S H 2000 Chem. Mater. 12 1866–73

[14] Kojima Y, Fukumori K, Usuki A, Okada A and Kurauchi T 1993 J. Mater. Sci. Lett. 12 889–90 [15] Manias E, Touny A, Wu L, Strawhecker K, Lu B and Chung T C 2001 Chem. Mater. 13 3516–23 [16] Muzny C D, Butler B D, Hanley H J M, Tsvetkov F and Peiffer D G 1996 Mater. Lett. 28 379–84 [17] Usuki A, Kojima Y, Kawasumi M, Okada A, Fukushima Y, Kurauchi T and Kamigaito O 1993 J. Mater. Res. 8 1179–84 [18] Upcroft P and Upcroft J A 1993 J. Chromatogr. Biomed. Appl. 618 79–93 [19] Ray S S and Okamoto M 2003 Prog. Poly. Sci. 28 1539–41 [20] Li X and Bhushan B 2002 Mater. Charact. 48 11–36 [21] Bhushan B and Li X 2003 Int. Mater. Rev. 48 125–64 [22] Oliver W C and Pharr G M 1992 J. Mater. Res. 7 1564–83 [23] Sneddon I N 1965 Int. J. Eng. Sci. 3 47–56 [24] Piner R D, Xu T T, Fisher F T, Qiao Y and Ruoff R S 2003 Langmuir 19 7995–8001 [25] Fornes T D and Paul D R 2003 Polymer 44 4993–5013 [26] Luo J J and Daniel I M 2003 Compos. Sci. Technol. 63 1607–16 [27] Rho J Y, Ashman R B and Turner C H 1993 J. Biomech. 26 111–9 [28] Rho J Y, Kuhn-Spearing L and Zioupos P 1998 Med. Eng. Phys. 20 92–102 [29] Zysset P K, Guo X E, Hoffler C E, Moore K E and Goldstein S A 1999 J. Biomech. 32 1005–12 [30] Tang Z Y, Kotov N A, Magonov S and Ozturk B 2003 Nat. Mater. 2 413–8 [31] Li X and Bhushan B 2000 Scr. Mater. 42 929–35 [32] Halpin J C 1969 J. Compos. Mater. 3 732–4 [33] Halpin J C and Kardos J L 1976 Polymer Eng. Sci. 16 344–52 [34] Ashton J E, Halpin J C and Petit P H 1969 Primer on Composite Materials: Analysis (Standford, CT: Technomic) [35] Mori T and Tanaka K 1973 Acta Metall. 21 571–4 [36] Tandon G P and Weng G J 1984 Polym. Compos. 5 327–33 [37] Sheng N, Boyce M C, Parks D M, Rutledge G C, Abes J I and Cohen R E 2004 Polymer 45 487–506 [38] Shia D, Hui C Y, Burnside S D and Giannelis E P 1998 Polym. Compos. 19 608–17 [39] Hui C Y and Shia D 1998 Polym. Eng. Sci. 38 774–82 [40] Wang J and Pyrz R 2004 Compos. Sci. Technol. 64 925–34 [41] Nam P H, Maiti P, Okamoto M, Kotaka T, Hasegawa N and Usuki A 2001 Polymer 42 9633–40 [42] Ren J X, Silva A S and Krishnamoorti R 2000 Macromolecules 33 3739–46 [43] Krishnamoorti R, Ren J X and Silva A S 2001 J. Chem. Phys. 114 4968–73 [44] Nakatani A I, Hielm R P, Gerspacher M and Krishnamoorti R 2002 Mater. Res. Soc. Symp. vol 661 (New York: Materials Research Society) [45] van Es M, Feng X Q, van Turnhout J and van der Giessen E 2001 Specialty Polymer Additives—Principles and Applications ed S Al-Malaika, A Golovoy and C A Wilkie (Oxford: Blackwell Science) pp 391–413

2029