Percolated network formation in biocidal 3D porous PCL/clay nanocomposite scaffolds: effect of organic modifier on interfacial and water sorption properties

Abstract

The influence of chemical interaction between poly(ε-caprolactone) (PCL) and Cloisite 10A on rheology, water permeability and antibacterial properties were subjected to detailed investigation. Mats of PCL with varying amounts of Cloisite 10A were prepared by electrospinning technique. The hydrogen bonding interaction between PCL and the organic modifier present in Cloisite 10A encourages the exfoliation/intercalation of Cloisite 10A resulting in a strong immobilized polymeric zone which was confirmed by small angle oscillatory shear experiments (SAOS). Unimpeded permeation of water through a PCL–Cloisite 10A porous nanocomposites scaffold was confirmed by different diffusion models. This strong immobilized zone or percolated network formation, aids in the elution of the organic modifier present in the nanoclay, which leads to the rupture of the cell wall of the bacteria. The antibacterial properties were tested using Gram positive bacteria and compared with the results obtained for Gram negative bacteria to test the use of our nanocomposites for wound healing applications.

---

Introduction

Healing is a systematic process, conventionally explained in terms of overlapping phases such as hemostasis, inflammation, proliferation, and maturation.¹ An ideal wound-healing scaffold should have suitable physical and mechanical properties to prevent secondary infection along with an excellent physiological environment to facilitate cell adhesion, proliferation and differentiation. Conventional dressings greatly improve the quality of healing and decrease healing time. On the other hand they still possess certain disadvantages like poor bio stability, low mechanical properties, high cost, low shelf-life and risks of immunological rejection which limit their application.² These kind of dressings are inferior in controlling secondary infection such as osteomyelitis (bacterial infection of bone) caused by local spread from an adjacent infection after trauma, bone surgery, diabetic foot infection or joint replacement.³ Staphylococcus aureus (S. aureus), a pathogenic microorganism which is responsible for ∼80% of the infections causing osteomyelitis mainly colonizes on the skin and mucous membrane. Situations of trauma or inoculation cause the entry of S. aureues into the host infecting a variety of tissues and eventually leading to osteomyelitis.^4,5 Synthetic dressings with an inherent anti-bacterial activity can prevent these secondary infections. There is an urgent need for new antimicrobials due to the increasing number of drug-resistant bacterial infections worldwide. To that end, synthetic polymers have been widely investigated as a new molecular platform to create antimicrobial agents that are active against drug-resistant bacteria with polyurethane, poly(ε-caprolactone) (PCL), poly(acrylonitrile) (PAN), polyvinyl alcohol (PVA) and chitosan.^6–10

Electrospun polymer nanofiber membranes are smart candidate material for wound dressing applications because of its three dimensional porous structure, comparable mechanical properties and biocompatibility.¹¹ We have earlier examined the super hydrophilic nature possessed by the electrospun PCL–Cloisite 10A nanocomposites as a function of clay loading. In order to explore the extended possibilities of PCL–Cloisite 10A three dimensional porous scaffolds as smart wound dressing material/hydrophilic drug releasing membrane, here in this communication, we have investigated the rheological, mechanical, antibacterial and diffusion properties of electrospun PCL–Cloisite 10A nanocomposite scaffolds.

Materials

Polycaprolactone (M_n = 80 000 Da) was obtained from Aldrich and dichloromethane from Fluka. The organically modified montmorillonite clay (Cloisite 10A, Cloisite 15A) was purchased from southern clay products, India. All materials were used as received.

Preparation of PCL/Cloisite 10A nanocomposite scaffolds

Preparation of neat PCL as well as PCL/Cloisite 10A nanocomposites scaffold using electrospinning technique (along with solution parameters and operation parameters) has been discussed in detail elsewhere.¹²

Characterization

Rheology. Rheological analysis was carried out in stress controlled rheometer DHR 3 (Discovery Hybrid Rheometer 3) from TA instruments. The measurements were performed in 25 mm parallel plate geometry with a plate gap of 1 mm, at 80 °C. The samples having 2 mm of thickness and 25 mm of diameter were prepared by electrospinning followed by, pressing the electrospun mat. The frequency varied from 0.1 to 100 rad s⁻¹. With the ultimate aim being the study of interfacial properties, rheology experiments were conducted using a shear amplitude of 0.01.

Dynamic mechanical analysis (DMA). Rectangular test specimens of dimensions (35 × 8 × 3.5 mm) were subjected to DMA using a Perkin Elmer DMA8000 in tension mode at a frequency of 1 Hz and oscillation amplitude of 0.05 mm. The nanocomposites were studied between temperature ranges of −80 °C to 20 °C.

Wide angle X-ray diffraction studies (WAXD). WAXD analysis was carried out on neat PCL, and PCL–Cloisite 10A nanocomposites using Bruker AXS D8 X-ray diffractometer equipped with a high temperature XRD cell (HTK2000), having CuKα₁ radiation (λ = 0.154 nm), between 2θ ranges of 5–60°.

Polarizing optical microscopy (POM). Neat PCL and PCL nanocomposites were imaged with Nikon microscope (Yokohama, Japan) equipped with a Linkam heating/cooling unit (Linkam TM 600/s) (Surrey, UK) with images collected using Leica Q Win software. The nanocomposites were first melted at 80 °C followed by cooling in air atmosphere to room temperature and the final image taken was used for the study.

Transmission electron microscopy (TEM). The dispersion of nanofillers in polymer nanocomposites was investigated using TEM. The micrographs of the nanocomposites were taken in JEOL JEM transmission electron microscope with an accelerating voltage of 200 keV. Here one single fiber was separated from the electrospun mat and used for the analysis. The images were subjected to detailed study using the imageJ software.

Water sorption analysis. Moisture uptake was determined by measuring the weight periodically by soaking the specimen in water. Samples were prepared as per ASTM D5229. The specimens for water absorption test were cut into 25.4 mm length and 12.7 mm diameter cylinder. Prior to absorption experiments, the samples were dried until the weight stabilized. The specimens were then immersed in distilled water at room temperature and the weight change monitored as a function of time with absorption behavior weighed at different intervals up to equilibrium.

Antibacterial activity of electrospun scaffolds. Antibacterial activity of electrospun scaffolds was performed using disc-diffusion method (CLSI, 2009). Mueller-Hinton Agar (MHA) (Himedia, India) plates were prepared with a depth of 4 mm. S. aureus MTCC 1430 and E. coli MTCC 1610 were used as the test organisms. Growth method (log phase method) was used for inoculum preparation. Three to five well-isolated colonies of the same morphologic type were selected from an agar plate culture. These colonies were transferred with the help of a loop into a tube containing 5 ml of Mueller-Hinton broth and incubated at 35 °C to achieve a turbidity of 0.5 McFarland standard (corresponds to approximately 1 to 2 × 10⁸ CFU ml⁻¹). A BaSO₄ turbidity standard, equivalent to a 0.5 McFarland standard was used to standardize the inoculum density. The bacterial suspension was inoculated on the dried surface of the MHA plates with a swab and kept for 5 minutes. Discs (6 mm-diameter) prepared from electrospun scaffolds (neat as well as nanocomposites) were placed onto the surface of the inoculated agar plates. Each disc was pressed gently with a sterile forceps to ensure complete contact with agar surface. Erythromycin (15 μg per disc) and ampicillin (10 μg per disc) were used as a positive control for S. aureus MTCC 1430 and E. coli MTCC 1610 respectively. The plates were inverted and incubated at 35 °C for 18 hours. After incubation, zone of inhibition was measured in millimeters and recorded. The tests were performed in triplicates.

Results and discussion

Rheological analysis are sensitive to clay dispersions and even for the polymer clay interactions.^13,14 Here, we study the influence of Cloisite 10A loading on the rheological behavior of PCL–Cloisite 10A nanocomposites. Fig. 1(a and b) shows the dependence of dynamic storage modulus (G′) and loss modulus (G′′) for the neat PCL and PCL–Cloisite 10A nanocomposites at 80 °C. The elastic moduli (G′) of PCL and various nanocomposites at different frequencies were plotted in Fig. 1(a). Nguyen et al.¹⁵ reported that the plateau could be a result of networks formed due to strong interaction between modifiers and matrix or large agglomerates of clays formed due to high clay loading. Fig. 1(b) portrays the loss modulus as a function of frequency. A very distinct plateau region observed in the low frequency region (≥5 wt% nanocomposites) indicates the formation of a strong interface between PCL and Cloisite 10A. This interface can be formed due to interaction between the organic modifier of Cloisite 10A and PCL. The above said interaction restricts mobility of PCL chains which are close to the Cloisite 10A surface. These strong immobilized zone formation has earlier been reported for other nanocomposites.^16,17 Wagener et al.¹⁸ has suggested that the flow curves at low amplitude measurements can be fitted to the following power law expression (1),

η = Aωⁿ (1)

where η is the viscosity, ω is the frequency, A is the sample specific exponential factor and n is the shear thinning exponent. A plot of log η against log ω (Fig. 3) will give the shear thinning exponent value “n”. The n value decides the extent of clay orientation in nanocomposites. The extent of non-Newtonian behavior of the system can be rated from the “n” values. The pseudoplastic, dilatant and Newtonian behavior of polymers are characterized by n which have values n < 1, n > 1 and n = 1 respectively. The increase in pseudoplasticity shows a higher extent of silicate exfoliation associated with network build-up of layers. From Fig. 3, it can be noticed that n value is getting increased with the incorporation of Cloisite 10A. It can be seen that for pure PCL, n value was calculated as 0.45. With the incorporation of 5 wt% Cloisite 10A, n value increases to 0.65 and for 10 wt% nanocomposites it almost increases up to 0.93. Therefore shear thinning effect on polymer nanocomposites will be higher with the increase in shear strain compared to the neat polymer. It clearly indicates the superior orientation of the clay platelets in the direction of shear. It can be noticed that the neat PCL exhibits a Newtonian plateau at lower frequencies which gets disappeared with the addition of Cloisite 10A. It can be attributed to the formation of solid-like behavior induced by PCL–Cloisite 10A interactions.¹⁹ But at higher frequencies, shear thinning phenomena is observed due to the breakage of such kind of network with the increase in shear rate. The nature of the Cole–Cole plots is reported to be indicative of the homogeneity of the system.^20,21 Gramespacher et al. used Cole–Cole plots to study the interfacial properties of immiscible polymer blends and nanocomposites.²² A semicircular shape of the plotted curves indicates good compatibility or phase homogeneity in the melt. Any deviation from the so called semicircular shape indicates non homogeneous dispersion in the polymer melt. Here, it can be noticed that addition of nanoclay (above 2.5 wt%) leads to a deviation from the semicircular shape for Cole–Cole plot, which clearly indicates non homogeneous dispersion in the polymer melt with the addition of higher clay loading. Fig. 2 shows the Cole–Cole plot for PCL–Cloisite 10A nanocomposites and the tail region highlighted by dotted circle points to a strong interface formed as result of the interaction between PCL and organic modifier of Cloisite 10A. Different kinds of theoretical models were proposed to explain the change in Newtonian plateau at lower shear rates in polymer nanocomposites. Many authors^23,24 argued that the equations like Carreau–Yasuda²⁵ (CY) and Cross model²⁶ can explain the rheological behavior of nanofiller loaded composites. Recently, Ajesh et al. described that the CY model can be used to explain the shear thinning behavior of NR nanocomposites at lower frequency region.²⁷ This kind of models were found successful in many other systems like PP and PP-grafted maleic anhydride blend systems. Charman et al. applied the Cross model to CNT incorporated EVA nanocomposites.²³ In Cross model, the change of the rheological behavior in the low-frequency region were described by calculating the yield stress values. The Cross equation is given by model;²⁶where σ₀ is the yield stress, η₀ is the zero shear viscosity, m the power law index and λ is the mean relaxation time. These parameters were adjusted to obtain the best fit curve. The optimized values are given in Table 1. From Table 1, it is clear that the yield stress values are increasing with the increase in the nanofiller concentration in PCL matrix. Here, for PCL nanocomposites, the yield stress value increased from 233.15 Pa to 1533 Pa upon the addition of 5 wt% Cloisite 10A. There is a large variation in the yield stress values on further loading of Cloisite 10A. The yield stress value changes to 31 052.08 Pa for 10 wt% of Cloisite 10A nanocomposite system. For Cloisite 15A loaded system, the yield stress value increases to 346.70 Pa for 5 wt% of Cloisite 15A. The yield stress value reaches to 13 370.28 Pa upon the addition of 10 wt% of Cloisite 15A i.e., the extent of increase in yield stress value upon the addition of Cloisite 15A is not prominent as observed in Cloisite 10A loaded systems. Many studies pointed out that the yield stress value variation is a direct indication of the extent of exfoliation of nanoclay platelets in polymer matrices. Thus, as observed from previous reports,¹² it is clear that exfoliation mechanism is more preferred for Cloisite 10A loaded nanocomposites as compared to Cloisite 15A. The zero shear viscosity values obtained from the model fitting were also presented in Table 1. The zero shear viscosity values are also increases for Cloisite 10A loaded system and also shows the same trend as for that of yield stress. The relaxation time, (λ) shows irregular variation at 5 wt% loading for Cloisite 10A and 5, 7.5 wt% for Cloisite 15A loading. But, at higher filler loadings (i.e., at 7.5 wt% and 10 wt% of Cloisite 10A loading and 10 wt% of Cloisite 15A loading) this variations are matching with the yield stress values. This is due to the improper dispersion of nanofiller in PCL matrix at lower filler loadings. Fig. 4 shows the WAXD results for PCL and PCL–Cloisite 10A nanocomposites. The slight shift in 2θ values observed for the Cloisite 10A confirms the interaction between PCL and organic modifier of Cloisite 10A.²⁸ This interaction results in a change in spherulite morphology.²⁹ Therefore the results obtained from WAXD analysis was used to calculate the crystallite size (l) using Scherer eqn (3),³⁰where l is the crystallite size or fractal dimension, λ is the wavelength of X-ray (0.154 nm), and β is the full width at half maximum (FWHM). Similarly the d spacing was calculated using the Bragg eqn (4)

Nλ = 2d sin θ (4)

where n is the order, and d is the d spacing. Fig. 5 shows the l and d values obtained for PCL nanocomposites. Clearly the crystallite size (l) and d spacing (d) values shows an initial increase (2.5 wt% nanocomposite) and thereafter shows a gradual decrease. This can be due to the fact that increasing amount of Cloisite 10A hinders the segmental motion of PCL chains thereby obstructing the crystallization process^28,29,31 This is evident form the polarized optical microscopy images shown in Fig. 6. Mechanical properties of the nanocomposites are extremely important when considering their application in wound dressing and other biomedical fields. DMA of PCL and PCL based nanocomposites provide the information in that relation. Fig. 7 shows the variation of dynamic storage modulus (E′) and tan δ as a function of temperature. It can be observed that the E′ value increases as the amount of Cloisite 10A is increased to 5 wt% which is due to the fact the interaction between PCL and organic modifier of Cloisite 10A aids the dispersion of Cloisite 10A in the matrix. Further increase in the amount of Cloisite 10A causes a decrease in the E′ which can be due to the agglomeration as observed in Fig. 7(a). Fig. 7(b) shows the variation of tan δ as a function of temperature. Width of the tan δ curve gives an idea about the dispersion of Cloisite 10A. It can be observed that the width of tan δ curve shows an increase up to 5 wt% of Cloisite 10A and thereafter it shows a decrease. The height of tan δ peak is a measure of the damping properties and the height decreases from 0.0521 (for neat PCL) to 0.0358 (for 5 wt% nanocomposite) followed by an increase to 0.044 (for 10 wt% nanocomposites). The extent of dispersion of Cloisite 10A in PCL–Cloisite 10A nanocomposites was investigated using transmission electron microscopy (TEM) and Fig. 8 shows the TEM image of PCL–Cloisite 10A nanocomposites. From the Fig. 8(a), it can be seen that 2.5 wt% nanocomposite has maximum dispersion compared to other clay loadings and having partially exfoliated structure in the PCL matrix. As the clay platelets gets aligned over themselves, stacking of clay platelets (tactoid formation) occur at higher clay loadings (≥5 wt%) which can be clearly seen in Fig. 8(b). Fig. 9 shows a schematic representation for the formation of the immobilized polymeric zone formation. Water sorption behavior of the composite is a concern in biological applications of the electrospun fibrous mats. The water sorption property depends on the percentage content of the fiber, fiber orientation, temperature, area of the exposed surface, permeability of fibers, void content, hydrophilicity of the individual components etc. In case of polymer clay nanocomposites, amount of nanofiller loading and the extent of polymer clay interaction were also get accounted towards the estimation of water sorptivity. Fig. 10 shows the moisture up take of neat PCL and nanocomposites scaffolds when it is immersed in distilled water. ​Fig. 11 shows the modelling of water absorption using Peppas–Sahlin and Haguchi models.

Fig. 1 Rheological properties of PCL and PCL–Cloisite 10A nanocomposites where (a) represents the storage modulus, (b) loss modulus as a function of frequency at 80 °C.

Fig. 2 Cole–Cole plot for PCL and PCL–Cloisite 10A nanocomposites.

Fig. 3 Complex viscosity against frequency curves of (a) Cloisite 10A and (b) Cloisite 15A loaded PCL nanocomposites (the dashed line indicates the cross model fitting curves).

Table 1 Results obtained for fitting Cross model with experimental results obtained for PCL–Cloisite 10A and Cloisite 15A nanocomposites

Sample	Nanoclay	Yield stress (Pa), σ0	Zero shear viscosity, η0 (Pa s)	Relaxation time (λ) (s)	Power law index (m)
Neat PCL		233.15	9914.44	0.04	0.69
5 wt%	Cloisite 10A	1466.82	33 426.04	5.03	0.27
	Cloisite 15A	346.70	19 337.81	1.35	0.27
7.5 wt%	Cloisite 10A	36 669.90	54 817.20	0.33	0.58
	Cloisite 15A	1242.86	35 086.98	3.54	0.28
10 wt%	Cloisite 10A	31 052.087	50 879.87	0.25	0.62
	Cloisite 15A	13 370.28	40 602.54	0.92	0.43

---

Fig. 4 WAXD analysis of PCL–Cloisite 10A nanocomposites.

Fig. 5 Crystallite size and d spacing for PCL–Cloisite 10A nanocomposites as a function of the amount of Cloisite 10A.

Fig. 6 Polarized optical microscopy image for (a) neat PCL and, (b) 2.5 wt% nanocomposites.

Fig. 7 Plot of (a) storage modulus (E′) and (b) tan δ as a function of temperature for PCL–Cloisite 10A nanocomposites.

Fig. 8 Tem image for PCL–Cloisite 10A nanocomposite with 2.5 and 5 wt% of Cloisite 10A.

Fig. 9 Schematic representation of the immobilized zone of PCL chains (percolate network) formed as a result of interaction between PCL and organic modifier of Cloisite 10A.

Fig. 10 Water uptake as a function of time for PCL–Cloisite 10A nanocomposites.

Fig. 11 Model fitting of the water sorption through the electrospun (a) Cloisite 10A loaded and (b) Cloisite 15A loaded PCL nanocomposite using Peppas–Sahlin and Haguchi models.

Weight gain rate (%) was calculated by (5)

where M (%) is the moisture content in percentage; M_w is the weight of the wet sample at the time (t) and M₀ is the initial weight of the sample. From Fig. 10 it follows that the moisture up take property of the scaffolds increases with time. At the beginning of the curve, the weight increases sharply demonstrating the rapid moisture penetration into the composite materials. From the Fig. 10, it can be clearly seen that the amount of water uptake by fibrous mat increases with increase in Cloisite 10A loading. From SEM analysis, it was confirmed that the fiber diameter gets reduced with the incorporation of Cloisite 10A loading. This has been discussed in detail in our earlier manuscript.¹² Reduction in fiber diameter causes higher surface area which also facilitates higher water sorption efficiency of the electrospun membrane. There are three known mechanisms for water transport in polymer nanocomposites which are: Fickian diffusion, relaxation controlled, and non-Fickian or anomalous. The dominant mechanism depends on factors such as chemical structure of the polymer and morphology of the PCL–Cloisite 10A nanocomposites and their interfacial adhesion. These cases can be distinguished theoretically by the shape of the sorption curve represented by the following eqn (6).^32,33where M_t, M_∞, k, and q are the water absorption at time t, the water absorption at the saturation point, and constants, respectively. The value of q is different for the cases as follows: in Fickian diffusion q = 0.5, where the rate of diffusion of permeant molecules is much less than the polymer segment mobility, non-Fickian q = 1, where permeant diffusion rates are much faster than polymer relaxation process, and anomalous transport 0.5 < q < 1, where the permeant mobility and polymer segment relaxation rates are similar. The transport coefficients (q and k) were obtained from regression analysis of the linear portion of the curve log M_t/M_∞ versus log t. The values of q and k are furnished in Table 2. From the table, it is clear that the both Cloisite 10A and Cloisite 15A loaded 5 wt% PCL nanocomposite shows non-Fickian diffusion process. PCL nanocomposite with 10 wt% of Cloisite 15A shows anomalous transport behavior. All other nanocomposite shows Fickian transport. The variation in the k values with the addition of nanofillers in PCL matrix is quite interesting. According to previous reports, k values represent the interaction between polymer and nanofillers. Stephen et al. pointed out that higher the k value the higher is the interaction between the polymer and the solvent.³⁴ So the nanoclay platelets can reduce the interaction between the polymer matrix and the solvent used. Here, 5 wt% nanocomposite with Cloisite 10A and Cloisite 15A, and 10 wt% nanocomposite of Cloisite 15A showed lower k value compared to the neat PCL sample. This indicates that these systems have lesser affinity towards water as we have reported in our previous article.¹² Different mathematical models are available to elucidate exact mechanism during the diffusion process in polymeric systems.³⁵ Here used Korsmeyer–Peppas³⁶ and Peppas–Sahlin³⁷ models for determining the water sorption kinetics. These models were actually developed to study the drug release kinetics. Recently, Maria et al. argued that since these models are based on the process of penetrant migration from the initial position in the polymeric system to the polymer's outer surface, the modelling using these mathematical formulations are valid.³⁸ These models are based on the Fick's basic law of diffusion. The Korsmeyer–Peppas³⁶ equation is given by (7),where k₁ is the constant characteristic of the filler–polymer system, d is the diffusion exponent and Q_t/Q_∞ is the fraction of solvent released at time t and the values are given in Table 3. According to Hasçiçek et al.³⁹ the diffusion exponent d is the major parameter that decides the diffusion through polymeric matrices. Hasçiçek and coworkers³⁹ pointed out that when d takes the value of 0.45, it indicates the system follows diffusion-controlled mechanism. From our model fitting, it was observed that neat PCL and 10 wt% 15A loaded nanocomposites shows n value close to 0.45. This indicates that these two systems follows a diffusion controlled mechanism. This kind of assumption is not possible for other systems since the rate of diffusion is high for these systems and hence the d values. So a proper interpretation of the diffusion mechanism cannot be made using the Korsemeyer–Peppas equation.

Table 2 The diffusion transport kinetics values obtained from the eqn (6)

Sample	Nanoclay	k value	q value
Neat PCL		0.024	0.552
5 wt%	Cloisite 10A	0.003	1.373
	Cloisite 15A	0.006	1.367
7.5 wt%	Cloisite 10A	0.742	0.048
	Cloisite 15A	0.388	0.235
10 wt%	Cloisite 10A	0.698	0.061
	Cloisite 15A	0.012	0.607

---

Table 3 Values of k₁, k₂ and d obtained using Peppas–Sahlin and Korsmeyer–Peppas models

Peppas–Sahlin model
	Nanoclay	k1	k2	d	R^2
Neat PCL	—	−1.06	0.91	0.08	0.97
7.5 wt%	Cloisite 10A	0.93	−0.22	0.09	0.99
	Cloisite 15A	0.64	−0.10	0.18	0.98
10 wt%	Cloisite 10A	0.88	−0.19	0.10	0.99
	Cloisite 15A	0.03	−1.2 × 10^−6	1.07	0.99

---

Korsmeyer–Peppas model
		k1	d	R^2
Neat PCL		0.05	0.43	0.96
7.5 wt%	Cloisite 10A	0.77	0.04	0.99
	Cloisite 15A	0.69	0.06	0.96
10 wt%	Cloisite 10A	0.75	0.04	0.99
	Cloisite 15A	0.07	0.39	0.73

---

So the Peppas–Sahlin equation is applied on the diffusion behavior of PCL/clay nanocomposites. This model predicts that the diffusion mechanism in polymer matrices are due to two processes; diffusion into the swollen polymer and matrix relaxation. The Peppas–Sahlin equation can be given by eqn (8),³⁷

where k₁ is the diffusion Fickian contribution coefficient, k₂ is the relaxation contribution coefficient and d is Fickian diffusion exponent. When k₁ > k₂, the system follows a diffusion controlled mechanism, when k₁ < k₂, the system follows matrix controlled swelling mechanism. Here, neat PCL follows a matrix controlled swelling mechanism while all others follows diffusion controlled mechanism i.e., from the Peppas–Sahlin fitting, we can conclude that the electrospun PCL nanocomposites show diffusion controlled water uptake due to the high affinity of the water molecules and the nanocomposite matrix and this effect is more pronounced in Cloisite 10A loaded system. The values obtained for Cloisite 10A and Cloisite 15A nanocomposites are shown in Table 3. The antibacterial activity of PCL and PCL nanocomposite fiber mats were assessed by observing their activity (based on the disc diffusion method) against both Gram-negative (E. coli) and Gram-positive (S. aureus) bacteria. The activity of the neat PCL membranes against these bacteria was used as a control. The results of the antimicrobial activity analysis are shown in Fig. 12. From Fig. 12, it is clear that the fabricated membrane had good antimicrobial activity against S. aureus, but not with E. Coli. According to the results obtained, neat PCL membranes has showed no activity against both of the tested bacteria. The measured inhibition zone diameter is reported in Table 4. Electrospun scaffolds with varying Cloisite 10A loading shows good antibacterial activity against S. aureus and it can also be seen that nanocomposites having 5 wt% Cloisite 10A loading showed optimum activity on comparison with other loadings. Bacterial cell wall has a negative charge due to the presence of teichoic acids linked to either the peptidoglycan or to the underlying plasma membrane. These teichoic acids are negatively charged because of presence of phosphate in their structure. When the composite scaffold was placed in the medium, quaternary ammonium moiety present in the organic modifier of clay platelets releases their ammonium cations which are attracted to the negatively charged cell membrane of bacteria by means of electrostatic interactions. The mode of action of cationic biocides has been suggested to progress as follows: (1) adsorption onto the bacterial cell surface, (2) diffusion through the cell wall, (3) binding to the cytoplasmic membrane, (4) disruption of the cytoplasmic membrane, (5) release of cell cytoplasmic constituents, and (6) cell death. Quaternary ammonium cation got penetrated into the cell wall and have a destructive interaction with the cytoplasmic membrane, followed by the leakage of intracellular components and consequent cell death.^40,41 At higher concentrations of nanofillers, the interaction between the polymer matrix and the filler will be apparently low due to higher filler–filler interactions. At higher clay loading, clay platelets are stacked each other so that the amount of ammonium cations provided by the organic modifier got reduced significantly which causes comparable reduction in its antibacterial activity. The antibacterial activity is very less towards the growth of E. Coli bacteria in comparison with S. aureus. The reason for such an observation could be explained in terms of the difference in the cell wall structure of these bacteria. The outer cell membrane of Gram negative bacteria contains lipopolysaccharide layer in its outer leaflet whereas Gram-positive bacteria lack such kind of layer so that antibacterial agent can easily invade the cell wall of Gram positive bacteria.^42,43

Fig. 12 Antibacterial activity of PCL–Cloisite 10A nanocomposites towards (a) E. coli, and (b) S. aureus.

Table 4 Diameter of the inhibition zone for PCL–Cloisite 10A nanocomposites against S. aureus

Bacteria	Inhibition zone diameter (mm)
	Erythromycin (15 μg per disc) (a0)	Neat PCL (a1)	2.5 wt% nanocomposite (a2)	5 wt% nanocomposite (a3)	7.5 wt% nanocomposite (a4)	10 wt% nanocomposite (a5)
S. aureus	22.00 ± 0.00	—	9.33 ± 0.58	11.67 ± 0.58	11.00 ± 0.00	10.67 ± 0.58

---

Conclusion

Detailed investigation of mechanical, water permeability and antibacterial properties are essential for analyzing the use of a material in biomedical applications. The chemical modifier used in Cloisite 10A interact with PCL resulting in the formation of an immobilized region was confirmed by detailed rheological analysis. This interface formed has a pronounced effect on the crystallization, water transport, mechanical and antibacterial properties. Flow behavior of shear thinning fluids were well described by using Cross model and parameters such as yield stress and zero shear viscosity were also evaluated. Similarly the water transport shows an increasing trend with increasing amount of Cloisite 10A which can be due to its hydrophilic nature. Diffusion studies showed that the PCL–Cloisite 10A nanocomposites followed Fickian mode of diffusion. Peppas–Sahlin fitting gives an idea that the electrospun PCL nanocomposites shows diffusion controlled water uptake due to the high affinity of the water molecules towards the nanocomposite matrix. The strong inhibition zone observed for S. aureus bacteria proves their robust antibacterial properties. Thus PCL–Cloisite 10A nanocomposites can be considered as to be strong candidates for wound healing applications. Super hydrophilic nature and water intake properties of the 3D porous nanocomposites scaffold can make it as a future candidate for controlled drug delivery applications.

Acknowledgements

The authors are thankful to the Department of Biotechnology and DST Nanomission, Government of India, for the financial support.

References

1. P. K. Ghosh and A. Gaba, J. Pharm. Pharm. Sci., 2013, 16, 760–820.
2. A. Song, A. A. Rane and K. L. Christman, Acta Biomater., 2012, 8, 41–50.
3. D. P. Lew and F. A. Waldvogel, Lancet, 2004, 364, 369–379.
4. F. D. Lowy, N. Engl. J. Med., 1998, 339, 520–532.
5. W. Zimmerli, P. D. Lew and F. A. Waldvogel, J. Clin. Invest., 1984, 73, 1191–1200.
6. D. Liang, Z. Lu, H. Yang, J. Gao and R. Chen, ACS Appl. Mater. Interfaces, 2016, 8, 3958–3968.
7. R. Augustine, E. A. Dominic, I. Reju, B. Kaimal, N. Kalarikkal and S. Thomas, RSC Adv., 2014, 4, 24777–24785.
8. R. Gharibi, H. Yeganeh, A. Rezapour-Lactoee and Z. M. Hassan, ACS Appl. Mater. Interfaces, 2015, 7, 24296–24311.
9. K. T. Shalumon, K. H. Anulekha, K. P. Chennazhi, H. Tamura, S. V. Nair and R. Jayakumar, Int. J. Biol. Macromol., 2011, 48, 571–576.
10. Y. Zhou, D. Yang, X. Chen, Q. Xu, F. Lu and J. Nie, Biomacromolecules, 2008, 9, 349–354.
11. R. Sahay, P. S. Kumar, R. Sridhar, J. Sundaramurthy, J. Venugopal, S. G. Mhaisalkar and S. Ramakrishna, J. Mater. Chem., 2012, 22, 12953–12971.
12. E. Elias, S. Chandran C, N. Chandran, F. G. Souza and S. Thomas, RSC Adv., 2016, 6, 21376–21386.
13. N. Chandran, S. Chandran, H. J. Maria and S. Thomas, RSC Adv., 2015, 5, 86265–86273.
14. K. M. Lee and C. D. Han, Macromolecules, 2003, 36, 7165–7178.
15. Q. Dzuy Nguyen, C. T. B. Jensen and P. G. Kristensen, Chem. Eng. J., 1998, 70, 165–171.
16. T. Domenech, R. Zouari, B. Vergnes and E. Peuvrel-Disdier, Macromolecules, 2014, 47, 3417–3427.
17. M. Segad, B. Jönsson and B. Cabane, J. Phys. Chem. C, 2012, 116, 25425–25433.
18. R. Wagener and T. J. G. Reisinger, Polymer, 2003, 44, 7513–7518.
19. B. Hoffmann, J. Kressler, G. Stöppelmann, C. Friedrich and G.-M. Kim, Colloid Polym. Sci., 2000, 278, 629–636.
20. J. S. Borah and T. K. Chaki, Appl. Clay Sci., 2012, 59–60, 42–49.
21. S. Sinha Ray, Clay-Containing Polymer Nanocomposites, Elsevier, Amsterdam, 2013, pp. 305–350.
22. H. Gramespacher and J. Meissner, J. Rheol., 1997, 41, 27–44.
23. M. Charman, F. Léonardi, S. Dominguez, C. Bissuel and C. Derail, J. Polym. Sci., Part B: Polym. Phys., 2011, 49, 1597–1604.
24. G. Naderi, M. Razavi-Nouri, E. Taghizadeh, P. G. Lafleur and C. Dubois, Polym. Eng. Sci., 2011, 51, 278–284.
25. P. J. Carreau, D. D. Kee and R. P. Chhabra, Rheology of Polymeric Systems: Principles and Applications, Hanser Publishers, 1997.
26. M. M. Cross, J. Colloid Sci., 1965, 20, 417–437.
27. A. K. Zachariah, V. G. Geethamma, A. K. Chandra, P. K. Mohammed and S. Thomas, RSC Adv., 2014, 4, 58047–58058.
28. V. Romeo, G. Gorrasi, Vittoria and I. S. Chronakis, Biomacromolecules, 2007, 8, 3147–3152.
29. N. K. Singh, B. D. Purkayastha, J. K. Roy, R. M. Banik, M. Yashpal, G. Singh, S. Malik and P. Maiti, ACS Appl. Mater. Interfaces, 2010, 2, 69–81.
30. A. L. Patterson, Phys. Rev., 1939, 56, 978–982.
31. P. Maiti, Langmuir, 2003, 19, 5502–5510.
32. J. S. Chiou and D. R. Paul, Polym. Eng. Sci., 1986, 26, 1218–1227.
33. R. Wilson, S. M. George, H. J. Maria, T. S. Plivelic, S. A. Kumar and S. Thomas, J. Phys. Chem. C, 2012, 116, 20002–20014.
34. R. Stephen, C. Ranganathaiah, S. Varghese, K. Joseph and S. Thomas, Polymer, 2006, 47, 858–870.
35. J. Crank, The Mathematics of Diffusion, Clarendon Press, 1979.
36. R. W. Korsmeyer, R. Gurny, E. Doelker, P. Buri and N. A. Peppas, Int. J. Pharm., 1983, 15, 25–35.
37. N. A. Peppas and J. J. Sahlin, Int. J. Pharm., 1989, 57, 169–172.
38. H. J. Maria, N. Lyczko, A. Nzihou, C. Mathew, S. C. George, K. Joseph and S. Thomas, J. Mater. Sci., 2013, 48, 5373–5386.
39. C. Hasçiçek, G. Yüksel-Tilkan, B. Türkmen and N. Ozdemir, Acta Pharm., 2011, 61, 303–312.
40. Y. Ohta, Y. Kondo, K. Kawada, T. Teranaka and N. Yoshino, J. Oleo Sci., 2008, 57, 445–452.
41. Y. Xue, H. Xiao and Y. Zhang, Int. J. Mol. Sci., 2015, 16, 3626–3655.
42. L. H. Muschel and L. J. Larsen, J. Bacteriol., 1969, 98, 840–841.
43. K. M. Reddy, K. Feris, J. Bell, D. G. Wingett, C. Hanley and A. Punnoose, Appl. Phys. Lett., 2007, 90, 213902.

---