Silanization induced inherent strain in graphene based filler influencing mechanical properties of polycarbonate urethane nanocomposite membranes

Abstract

Nanosheet type fillers apart from their size and surface functional groups may have numerous attributes affecting the mechanical properties of polymeric nanocomposites. To study these, silane-functionalized graphene based fillers were synthesized by chemically grafting N-[3-(trimethoxysilyl)propyl]diethylenetriamine (TMPT) onto graphene oxide (GO) and carboxylated GO (GOCO) using different chemistries. Their respective silanization yielded nano-fillers with amine (GOSAM) and alkoxy (GOCSAL) groups. Further, hydroiodic acid (HI) treatment led to synthesis of their reduced counterparts GRSAM and GRCSAL. The resulting TMPT-functionalized nanosheets were characterized by Fourier transform infrared spectroscopy (FT-IR) confirmed silane functionalization. A blue shift in Raman spectra indicated that during silanization with different terminal groups an inherent compressive strain has developed, while reduction with HI caused a red shift indicating a tensile strain, in these nanosheets. Polycarbonate urethane nanocomposite electrospun membranes (PCU) incorporated with these respective fillers at different loadings were analyzed. Morphology of the nanocomposite membranes was observed under SEM and membranes were characterized by static and dynamic mechanical analysis. The study indicated that the exfoliation and dispersion of graphene sheets in PCU has significantly improved due to surface functionalization while it also exhibited a novel aspect, variations in their mechanical properties in respect to the type of strain present in incorporated nanosheet fillers. The nanosheet fillers with compressive strain contributed more to the mechanical property enhancement of nanocomposite membranes, than the fillers with tensile strain. A spring and molecule model was thus proposed as possible explanation to relate inherent strain in filler to that of nanocomposite membrane mechanical properties. In vitro non-cytotoxic and hemocompatible nature of these fibroporous nanocomposite membranes provided their potential in biomedical applications.

---

Introduction

Nano-dimensions impart remarkable attributes in comparison to conventional macro sized and bulk materials. One of the extraordinary properties exhibited by nano structures consists of an extremely high surface area to volume ratio which can be successfully exploited as filler for improved matrix properties. Graphene due to its excellent mechanical, electrical, thermal, optical and biological properties is a wondrous filler material.^1,2 Graphene as filler can impart excellent properties to polymer nanocomposites.³ But, graphene surface without any functional groups has a tendency to agglomerate and cannot be dispersed in polymer matrix. So, graphene based materials like functionalized graphene (fG), graphene oxide (GO) etc. with chemically functionalized groups are now widely used in polymer nanocomposite fabrication.^4,5 These nanosheets as a filler impart superior mechanical properties to polymeric biomaterials giving rise to bionanocomposite membranes. A variety of polymeric biomaterials are being used for biomedical applications, out of which elastomeric polymers stand apart because of their tissue like mechanical properties.⁶ Polyurethane hence being a thermoplastic elastomer has wide biomedical applications like cardiac assist pumps, blood bags, heart valves, vascular grafts⁷ and also considered ideal for fabricating soft tissue scaffolds. As a scaffold, it can facilitate the distribution of loads to surrounding tissues and allow the regeneration site to be mechanically stable soon after implantation.⁸ Polyurethanes when utilized in chronic implants are exposed to hostile in vivo conditions and mechanical stress that can lead to their failure and hence, mechanical enhancement by fabricating bionanocomposite membranes using graphene based fillers would be a viable option. Functionalized graphene based fillers has been explored previously to attain such improved mechanical properties.^9,10

The fabrication of robust graphene-based polymer nanocomposites will result in commercial products if their interphases and reactivity are carefully controlled. Silanization for improvement of interphase has been established since long, wherein silane coupling agents enhance the wettability of filler to polymer matrix.¹¹ Thus, increasing the compatibility of filler to matrix, it can improve the physico-mechanical properties and chemical bonding between the two components that can provide greater resilience and protection against damage by water at the interface. Suzuki et al. in his study indicated that based on silane interactions proposed by Plueddemann, in case of thermoplastic polymers the silane coupling agent and the matrix inter-diffuses into one another and they form a structure having graded interphase. In this case, the bond strength strongly depends on the miscibility between the organofunctional group of the silane coupling agent and the polymer matrix.¹² Role of surface functional groups is to help in improving filler–matrix interaction and for providing better interphase with filler; each polymer may have different compatible functional groups. Successful interaction lies in finding such a suitable functional group and hence, may lead to better stress transfer at filler and matrix interphase. As the silanization proceeds, based upon the chemistry used, terminal functional groups of silanizing agents having vinyl, alkoxy, amine, epoxy, etc., gets retained on the nanosheet surface. These functional groups have been promising in providing a compatible interaction of graphene with the polymer matrix, thus promoting enhancement of properties.¹³

Contact area at interphase differs with filler shape, which can be fibrous like in natural fibers or carbon fibers; spherical like in nanoparticles or nanosheet like structures as in graphene flakes.¹⁴ Many aspects in this regards has not been well studied to design a polymer nanocomposite with desired mechanical properties. In case of graphene, other than that established and mostly studied factors like the size and functionalization of the nanoplatelets, nano-sheet like structures may have wrinkles and deformations, which may also have an independent or mutual affect on the nanocomposite tensile properties.¹⁵ These deformations may be directly linked to the stress or strain factors inherent to the nanosheet or brought about by different chemical treatments or functionalization. A study of these factors by simple surface characterization method can be of great benefit, not only for designing, but also for predicting mechanical behavior of such functionalized filler incorporated nanocomposite membrane.

Herein, we've made an attempt to study such properties; polycarbonate urethane a well renowned hydrolytically stable form of thermoplastic polyurethane used in biomedical implant application has been selected. Functionalized graphene based filler has been formulated using a dual functional organosilane having hydrolyzable alkoxy group and non-hydrolyzable amine group. Different chemistries were applied on GO and GOCO in such a manner that filler has amine and alkoxy groups exposed on their respective surfaces. Further, surface modified fillers have been formulated by reducing the prior functionalized filler counterparts using hydroiodic acid. Thus obtained different types of fillers, have been incorporated into PCU matrix using a versatile electrospinning method. Raman shift was used as a tool for analyzing the change in filler strain properties. Further, based on these strain properties, effect of individual fillers on mechanical properties of nanocomposite membrane has been analyzed under static and dynamic conditions.

Experimental section

Materials

Polycarbonate urethane (trade name of Carbothane), was obtained from Lubrizol Corp. Graphite flakes, N-[3-(trimethoxysilyl)propyl]diethylenetriamine (TMPT) and chloroacetic acid were procured from Sigma-Aldrich, USA. Di-methyl formamide (DMF), tetrahydrofuran (THF) and di-ethyl ether of analytical grade were bought from Spectrochem, India. H₂SO₄, H₃PO₄, KMnO₄, HCl, NaOH and 30% H₂O₂ were purchased from Merck (India). A 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay kit was purchased from Hi Media (India). De-ionized (DI) water was used throughout the study.

Synthesis of graphene oxide and carboxylation

Graphene oxide (GO) was prepared as given in previous reported work.¹⁶ 1 g (1 eqv.) graphite flake was taken with 6 g (6 equivalent) KMnO₄, mixed well followed by addition of 150 mL of 9 : 1 solution of H₂SO₄ and H₃PO₄. Stirring was done for 18 h at 50 °C and then cooled to room temperature. An ice solution of DI water with 5 mL H₂O₂ was made and the above mixture was added slowly while stirring. A mixture of shining yellow colour solution obtained was centrifuged at 6000 rpm for 30 min. Supernatant was discarded and washed thrice with DI water. Washing was carried out in similar manner using 30% HCl, absolute ethanol and di-ethyl ether. The residue was dried at 50 °C in a vacuum oven.

Carboxylation of hydroxyl groups on GO (2 mg mL⁻¹) surface was carried out using 1 M chloroacetic acid in the presence of 3 M sodium hydroxide and the mixture in DI water was sonicated for 3 h. Reaction was stopped using 4 g mL⁻¹ di-sodium hydrogen phosphate solution.¹⁷ The products were neutralized using 6 N HCl, centrifuged and washed thrice with DI water, methanol and di-ethyl ether and dried at 50 °C in vacuum oven.

Silane functionalization

Functionalization of fillers with silane amine was carried out using TMPT, via two different chemistries. First by hydrolysis of alkoxy groups on TMPT to yield hydroxyl groups followed by reaction with OH groups on GO surface to yield amine terminated graphene oxide (GOSAM).¹⁸ Ethanol/water (9 : 1) mixture was prepared and separated into 2 parts, 1% TMPT was stirred in one part of it for 3 h to allow hydrolysis of alkoxy terminals. 2 mg mL⁻¹ GO was dispersed in other part of solution via sonication for 30 min. Thus obtained GO solution was then mixed with TMPT and left stirring for 3 h at 60 °C.

In second case, activation of –COOH group on GOCO (2 mg mL⁻¹) was done using EDC/NHS (20 mM/10 mM) via EDC coupling chemistry and formation of amide bonds.¹⁹ The reaction mixture having GOCO (2 mg mL⁻¹), EDC and NHS in ethanol/water (9 : 1) was sonicated for 30 min. Temperature of reaction mixture was raised to 60 °C followed by addition of TMPT (1%) and left stirring for 24 h. NH₂ terminated TMPT bonded to EDC/NHS activated –COOH, yielded alkoxy terminated GOCO (GOCSAL).

Products after reaction in both reactions were obtained by centrifugation at 6000 rpm, washed thrice with DI water, methanol, di-ethyl ether and dried at 60 °C in a vacuum oven.

Reduction using HI

Reduction of fillers GOSAM and GOCSAL were carried out using HI lead to formation of GRSAM and GRCSAL respectively. 5 mg mL⁻¹ filler to be reduced were suspended in HI and stirred at 100 °C for 1 h. The reaction mixture was neutralized using 5% NaHCO₃, centrifuged at 6000 rpm and washed thrice with DI water and ethanol. The product was obtained after drying the residue overnight at 80 °C in vacuum oven.

A schematic representing all the reactions has been given in Fig. 1.

Fig. 1 Schematic representation of synthesis of GO, GOCO, GOSAM, GOSCSAL, GRSAM and GRCSAL.

Electrospinning of polycarbonate urethane and nanocomposites

A syringe with 12% polycarbonate urethane solution in DMF/THF (50/50 v/v) was loaded onto the electrospinning unit. The syringe was capped with a 21-gauge blunt-end needle (spinneret), and a positive potential was applied at the needle end with an external power supply (Gamma High Voltage), whereas, the target terminal was grounded (Fig. S1†). The PCU solution was delivered by the syringe pump at a flow rate of 1 mL h⁻¹ to the rotating mandrel (3000 rotations per min) kept at a distance of 16 cm and electrospun at 9–10 kV potential to obtain polycarbonate urethane fibrous mat (PCU). The process was carried out at ambient temperature (28 ± 2 °C).

The nanocomposite solutions were prepared by tip sonication of GOSAM at weight percentages of 0.5 wt% (0.5 GOSAM/PCU), 1.5 wt% (1.5% GOSAM/PCU); GRSAM at weight percentages of 0.5 wt% (PCU/1% GO), 1.5 wt% (PCU/1.5% GO); GOCSAL at weight percentages of 0.25 wt% (0.25 GOCSAL/PCU), 0.5 wt% (0.5 GOCSAL/PCU), 1 wt% (1 GOCSAL/PCU) and GRCSAL at weight percentages of 0.5 wt% (0.5 GRCSAL/PCU) of PCU. Sonication was done for 45 min to exfoliate and disperse the filler in DMF and THF (50/50) mixtures followed by the overnight stirring with polymer pellets. Electrospinning was done onto a rotating mandrel under similar conditions at varying voltages of 10–12 kV for different filler loadings. The electrospun sheets were retrieved from the mandrel and air-dried to remove any residual solvent.

FT-IR spectroscopy

Fourier transform infrared spectroscopy (FT-IR) spectra were obtained at room temperature by KBr pellet method using a FT-IR spectrometer (Jasco FT-IR 6300, Japan). The scanned wave number range was 4000–500 cm⁻¹, at 4 cm⁻¹ of spectral resolution.

Raman spectroscopy

Raman spectroscopy experiments were performed with the confocal Raman microscope (Witec Inc. alpha300R). For the Raman spectral measurements a frequency-doubled NdYAG laser: l = 532 nm was used.

Morphological analysis (scanning electron microscopy (SEM))

SEM (Hitachi model S-2400) was used to study the morphology and average diameter of the fibers. Sample preparation was done by the sputtering of thin, flat sections of the materials with gold–palladium. The reported data on the average fiber diameter are the means of 50 measurements at different areas with Image J software.

Mechanical analysis

The dumbbell specimens were cut along the direction of electrospun fiber for tensile testing and rectangular pieces were cut for dynamic mechanical analysis (DMA) testing. Tensile testing was performed with an Instron 3345 model equipped with a load cell of 100 N, following ASTM D 638. Standard dumbbell-shaped specimens were cut with an ISO 527-2 type 5A die and tested at 25 ± 2 °C at a crosshead speed of 100.0 mm min⁻¹, with sample thicknesses that varied from 0.2 to 0.3 mm (sample size n = 5). The tensile properties were derived from the stress–strain curve. DMA was done in tensile mode with a Tritec 2000B DMA instrument (Triton Technology, Ltd., United Kingdom) on a 4 mm wide, 20 mm long and 0.2–0.3 mm thick rectangular sample. A frequency of 1 Hz with 50 μm amplitude of dynamic deformation was applied in the temperature-ramp experiments. Thermal cycling was monitored within −60 °C to 60 °C at a heating rate of 1 °C min⁻¹.

X-ray diffraction

XRD data were collected with a Bruker D8 Advance diffractometer with Cu Kα radiation at a generator voltage of 40 kV and a generator current of 40 mA with a scan speed of 4 min⁻¹ over 5–60°. The filler sample was analyzed in powder form, whereas for electrospun PCU nanocomposite membranes, a 20 mm diameter discs were used.

In vitro cytocompatibility

The MTT assay was performed to measure the succinic dehydrogenase (SDH) activity of the cells to reduce yellow-colored MTT to purple-colored formazan. The material extract was prepared by the incubation of disc shaped test material with a diameter of 20 mm in 1 mL high-glucose Dulbecco's modified Eagle medium supplemented with 10% fetal bovine serum, 1% antibiotic–antimycotic solution, and 0.37% sodium bicarbonate at 37 ± 1 °C for 24 ± 2 h. The extract (100%) was diluted to 50% (1 : 1) and 25% (1 : 3) with culture medium. Cells cultured in normal medium without any extract was considered as a cell control. The control and test sample extracts were added to a sub-confluent monolayer of L-929 cells in triplicate and incubated at 37 ± 1 °C for 24 ± 2 h. Later, a stock MTT solution (20 μL of 5 mg mL⁻¹ PBS) was added to the extract and control medium; incubated with protection from light for 4 h. After 4 h of incubation, 200 μL of dimethyl sulfoxide was added to the wells and the plates were put in an orbital shaker (GeNei, SLM-INC-OS-250, India) to dissolve the formazan crystals. The plate was quantified by the measurement of the absorbance at 570 nm with an ultraviolet-visible microplate reader and the percentage viability of the cells was calculated. The formazan content of each well was computed as a percentage of the mean of the readings from cell control as reference, which was taken to represent 100% biocompatibility. The reported data are the means and standard deviations (SDs) of three parallel runs for each sample.

In vitro hemocompatibility

The hemolytic character of the material was assessed by means of an in vitro hemolysis test. These tests were done using human blood samples in accord to the ethical guidelines approved by the Institutional Ethics Committee of Sree Chitra Tirunal Institute for Medical Sciences and Technology (approval no. SCT/IEC/594/2014 dated 21/04/2014). Fresh venous blood was collected (via Blood Bank, Sree Chitra Tirunal Institute for Medical Sciences and Technology) after obtaining written consent from the volunteer in anticoagulant citrate phosphate dextrose adenine (CPD-A) tubes and care was taken to prevent mechanical hemolysis. Total hemoglobin (Hb) was analyzed with a hematology analyzer (Sysmex-K4500). Disc shape test material of 6 mm diameter in triplicate was immersed in PBS for 5 min before exposure to blood. The samples were kept in a polystyrene plate; 2 mL of blood was added to each well, and a well without any sample was treated as reference (Ref). A volume of 1 mL of blood was taken immediately for initial analysis and the remaining 1 mL of blood was incubated with a sample for 30 min under agitation at 70 ± 5 rpm at 35 ± 2 °C.

The blood samples were centrifuged at 4000 rpm for 15 min and platelet poor plasma was aspirated to prepare platelet poor plasma. The free haemoglobin (fHb) liberated into the plasma after exposure to samples was measured using UV spectrophotometer as per the equation (eqn (1)).²⁰

fHb = 1.65 × A₄₁₅ − 0.93 × A₃₈₀ − 0.73 × A₄₅₀ (1)

wherein A₄₁₅, A₃₈₀ and A₄₅₀ are absorptions at 415, 380 and 450 nm wavelengths respectively. Percentage hemolysis was calculated using the formula (free Hb/total Hb) × 100. Hemolysis was thus calculated and results were expressed as mean ± standard deviation (SD) (n = 3).

Statistical analysis

All quantitative data are expressed as the mean ± standard deviation. GraphPad Prism 6 was used to check the significance of data by calculating p-value value and a value less than 0.05 was considered as significant. The tensile properties and fiber diameter for the nanocomposite membranes were analyzed statistically with an analysis of variance (ANOVA). Tukey analysis was also carried out to compare all nanocomposite membranes with PCU and 0.5 GOCSAL/PCU with 1.5 GOSAM/PCU.

Results and discussion

Characterization of chemical linkages after silane functionalization and reduction by HI

FT-IR has been a tool to analyze chemical linkages to indicate successful functionalization after the material has undergone chemical reactions. Graphene oxide and its carboxylated form (GOCO) spectra depict their characteristic peaks (Fig. 2(a)). The carboxyl group is clearly identified through reduced intensity of the broad 3426 cm⁻¹ O–H stretching vibration and appearance of more intense band at 1602 cm⁻¹ a C=O stretching vibration which is due to the oxidation of –OH bonds²¹ during the chemical reaction involving chloroacetic acid. Appearance of an out of plane deformation band at 956 cm⁻¹ and CO₂ vibration at 778 cm⁻¹ also indicate formation of carboxyl group. Shift in peak from 1731 cm⁻¹ to 1602 cm⁻¹ may be due to the asymmetric stretching vibration of CO₂⁻. Chloroacetic acid reaction under basic condition has also led to the disappearance of epoxy bands in GO to convert into carboxyl group.²²

Fig. 2 FT-IR spectra of graphene based fillers (a) GO and GOCO and (b) silanized GO, GOCO and their HI reduced counterparts.

In the second FT-IR spectra (Fig. 2(b)) silanized GO and GOCO spectra shows presence of silane group with a peak around wavenumber 1100 cm⁻¹, which is maintained even after HI reduction. The peak at 1648 in GOSAM can be assigned to NH₂ bending mode²³ of free NH₂ while peak at 1579 corresponds to –NH bending vibration and it presents the same at peak 1571 in GOCSAL.²⁴ A decrement in peak intensity at 1648 cm⁻¹ in GOCSAL may be the indication of its consumption during reaction with activated –COOH group in presence of EDC/NHS.

Raman spectra

Raman spectroscopy is a vital part of graphene research which is used to determine numerous characteristics of layers, its quality, and the effects of different treatments and exposures, such as electric and magnetic fields, strain, functional groups etc.²⁵

Raman spectra of GO shows characteristic peaks for D band and G band at 1339 and 1578 cm⁻¹, where D band depicts defects during oxidation, while G shows sp² configuration of carbon in its structural plane. Carboxylation has brought slight red shift in GO spectra, the D peak is red shifted by ∼5 cm⁻¹ and the G peak is red shifted by ∼12 cm⁻¹. This may be due to tensile strain experienced²⁶ by GO via increased negative charge during carboxylation. A rise in D peak can be noticed which also provides evidence of increased defects on GOCO plane and successful carboxylation by chloroacetic acid (Fig. 3(a)).

Fig. 3 Raman spectra (a) GOCO and GO and (b) silane functionalized GOCO (GOCSAL), GO (GOSAM) and respective HI reduced counterparts GRCSAL and GRSAM.

Silanization process seems to have an opposite effect on GO and GOCO planes, a blue spectral shift due to compressive strain can be observed, which is more prominent in GOCSAL with blue shift of ∼28 cm⁻¹ than GOSAM with ∼3 cm⁻¹ shift in both D and G band. HI reduction has caused a red shift of 12 cm⁻¹ and 4 cm⁻¹ in peak position of respective D bands of GRCSAL and GRSAM, indicating a reversal of strain experienced, which may be due to remaining surface functional groups (Fig. 3(b)). While G band of GRCSAL follows the same trend, but in GRSAM blue shift of ∼8 cm⁻¹ was noticed which indicates that it might have gained some compressive strain hence it may be difficult to conclude which strain may prevail. In Raman spectrum we noticed that functionalized graphene based materials has shown different strain behavior, this may be due to compression or expansion of the material lattice and which in turn may be due to interaction among different functional groups on its surface. At the same time Raman spectra of electrospun nanocomposite membranes (Fig. S2†) didn't show any shift in PCU characteristic peaks as it was observed in fillers.

Fiber morphology

Investigation by SEM micrographs of electrospun membranes shows fiber morphology and architecture (Fig. 4). Morphology of fibers with no beads shows good dispersion of silane functionalized filler in PCU matrix. 1.5 GOSAM/PCU and 0.5 GOCSAL/PCU also shows well aligned arrangement of fibers and could be suitable to be used as tissue engineering scaffold material where cell growth is required to be in aligned manner.

Fig. 4 SEM micrographs of blank PCU, and the nanocomposite membranes with different wt% loadings of GOSAM, GOCSAL, GRSAM and GRCSAL.

Fiber diameter of electrospun membrane is as provided in Fig. 5(a). Plot indicates PCU blank membrane has a fiber diameter of 285 ± 87 nm and it increased with incorporation of nanosheet fillers. GOSAM, filler with amine groups functionalized on GO has caused the highest fiber diameter increment of ∼108%, 0.5 GOSAM/PCU having 593 ± 105 nm and 1.5 GOSAM/PCU having 548 ± 140 nm. Incorporation of GOCSAL with alkoxy groups over GOCO, into PCU matrix has also showed increment in fiber diameter of ∼44% at 1 wt% loading, with average value of nanocomposite membranes falling between 300 and 500 nm. A larger fiber diameter of GOSAM/PCU nanocomposite membrane may be due to GOSAM derived from GO, having a larger sheet size than GOCSAL derived from GOCO, which has undergone longer duration of sonication during synthesis.

Fig. 5 Plots showing fiber diameter (n = 50) (a) and tensile properties of electrospun nanocomposite membranes (b), (c) and (d), along with significance (*P value < 0.05, n = 6).

The membranes incorporated with their HI reduced counterparts GRSAM/GRCSAL didn't contribute much to the fiber diameter and showed lesser values compared to PCU incorporated with GOSAM/GOCSAL. This may be ascribed to improvement in electrical conductivity of fillers while undergoing reduction with HI.²⁷ An increase in electrical conductivity of filler might have increased conductivity of solution to be electrospun, leading to higher stretching of fiber in the electric field applied during electrospinning.²⁸

Tensile properties

Tensile testing was carried out using Universal Testing Machine (UTM) and dumbbell shaped test samples cut from membranes under study. Surface functional groups of fillers play a critical role in enhancing mechanical properties of a polymer matrix, a functional group having better interfacial interaction with polymer may improve the interphase and provide better stress transfer between filler and matrix.²⁹ Graphene based filler being in thin nano-sheet form may also influence the matrix due to tensile or compressive strain present in it. Presence of functional groups or charges can influence these strain values as derived from Raman spectral analysis and these strain values may indeed be transferred to polymer matrix. A comparative study was done on PCU nanocomposite membrane with improved mechanical properties fabricated by incorporating different wt% of differently functionalized GOSAM, GRSAM, GOCSAL or GRCSAL fillers and tensile properties are as shown in Fig. 5(b)–(d). In the study done at 0.5 wt% filler, only 0.5 GOCSAL/PCU containing 0.5 wt% GOCSAL was found to have improved the tensile strength by ∼27%, Young's modulus by ∼26% and elongation at break by 11%. An increase of ∼41% in tensile strength, ∼36% increase in Young's modulus and elongation at break by 25% was noticed in the class of GOSAM at higher filler loading i.e. at 1.5 wt% which is 3 times more than GOCSAL. Another loading of still lower value fabricated by incorporation of 0.25 wt% GOCSAL, 0.25 GOCSAL/PCU membrane showed increment only in elongation at break of ∼28% but didn't influence tensile strength or modulus. Hence GOCSAL as filler with alkoxy functional group on its surface seems to have better interface than GOSAM with amine functionalized surface, with the PCU matrix and thus contributing more towards improvement in tensile properties at low loadings. But, increased loading of GOCSAL and higher dispersion reduced the tensile properties which may be due to increased filler to filler interaction, while for GOSAM filler to matrix interaction improved at an increased loading. Silanization has thus provided a better interphase and adhesion of filler with the matrix.

But along with this, the compressional strain that was noticed in Raman spectra may also influence nanocomposite membrane. These thin sheets having compressional strain may resist the tensile pull at interphase and may lead to an increase in tensile properties. Variation in amount of filler incorporated, interaction of filler surface functional groups, along with inherent strain properties via silanization has shown an increment in mechanical properties. Among the HI reduced forms as fillers, none has shown improvement in tensile strength while 0.5 GRSAM/PCU has shown ∼12% increment in elongation at break and 0.5 GRCSAL exhibited a ∼38% increment in elastic modulus values. For 1.5 GRSAM/PCU membrane a 29% increment in modulus value was noticed while other tensile properties under study have been lower. This puts some light on the importance of supporting functional groups on the surface before reduction and their influence on tensile properties. Raman spectral analysis have shown tensile strain to be more dominating after reduction and along with the reduced interaction of functional groups, may have resulted in observed negative effects on tensile properties. 0.5% loading of GRCSAL has only left with increased elongation at break while other properties have shown reduction in comparison to GOCSAL. GRSAM has shown an increase in modulus values which may be due to some compressional strain left in it, as indicated in blue shift of G band in Raman spectra.

Thus along with improvement in interphase, the inherent strains of thin nanosheet fillers indicated in Raman spectra may also contribute towards mechanical properties of nanocomposite membranes and hence may be one of the important factor to be considered. These inherent strains can be changed to a great extent by the type of surface functionalization used and hence indeed play a greater role.

Viscoelastic properties

Dynamic mechanical analysis was carried out to study viscoelastic properties of these nanocomposite membranes carrying functionalized fillers.

tan δ peak indicates the glass transition temperature (T_g) of polymer. T_g value has not been much affected except for 0.5 GRSAM/PCU membrane with lowest value of 14.4 °C (Table 1, Fig. S3†). 0.5 GRSAM/PCU with lowest T_g also had lowest tensile strength (Fig. 5(b)) value in the static analysis and this may be due to reduced interaction of filler and matrix. For blank PCU membrane it was 20.7 °C while 0.5 GOSAM/PCU showed highest value at 22.2 °C. 0.5 GOCSAL/PCU and 1.5 GOSAM/PCU membranes exhibiting highest tensile properties had T_g values of 21.6 °C and 21 °C, respectively. Hence, fillers in the present study had not considerably impacted the thermal properties of PCU membrane.

Table 1 Viscoelastic properties of electrospun membranes

Sample	tan delta peak (Tg) (°C)	E′ (storage modulus) at 37 °C (MPa)	E′′ at 37 °C (MPa)
PCU	20.7	11.1	2.6
0.25 GOCSAL/PCU	21.2	10.8	2.2
0.5 GOCSAL/PCU	21.6	13.6	2.9
0.5 GRCSAL/PCU	22	9.4	1.8
0.5 GOSAM/PCU	22.2	11.3	2.0
0.5 GRSAM/PCU	14.4	9.5	1.6
1 GOCSAL/PCU	18.9	10.2	1.9
1.5 GOSAM/PCU	21	16.1	3.1
1.5 GRSAM/PCU	18.9	8.0	1.6

---

Storage (E′) (Fig. S4†) and loss modulus (E′′) (Fig. S5†) values respectively show the elastic and viscous nature of the polymeric membranes.³⁰ Both values were recorded over a temperature range of −60 °C to 60 °C wherein, highest E′ and E′′ values were exhibited by 1.5 GOSAM/PCU and lowest by 0.5 GOSAM/PCU membrane (Table 1). To have better understanding of the behavior of these membranes as bionanocomposite membranes, study was concentrated at 37 °C i.e. at normal body temperature and values obtained are as given in Table 1.

At 37 °C, PCU blank membrane has E′ 11.1 MPa and E′′ 2.6 MPa; 1.5 GOSAM/PCU show highest modulus values with improvement of 45% for E′ and 19% for E′′ while, 0.5 GOCSAL/PCU had 23% increase in E′ 12% increase in E′′ values. Other nanocomposite membranes in the study didn't show considerable improvement in modulus values. Thus improved interaction at the interphase and inherent compressional strain of GOSAM and GOCSAL at different loadings has shown improved E′ and E′′. Higher compressional strain present in GOCSAL due to higher blue shift in Raman spectra may be the reason that it shows improved modulus at thrice the lower loading of GOSAM, which may have better interaction with matrix as reflected by higher modulus values. At the same time, presence of fillers with tensile strain i.e. the reduced counterparts at equal loadings had no commendable effect on dynamic modulus values. Hence nanocomposite membranes with better filler–matrix interphase along with compressional strain in fillers may contribute highly for improved viscoelastic properties.

Spring and molecule model

The spring and molecule model is being proposed as possible explanation to improvement of mechanical properties in the current study. Filler nanosheets with wrinkles and folds have been represented as a spring and polymer as molecules (Fig. 6). Compressional strain in filler has been depicted as a strain acting on stretched spring whereas tensile strain as a strain acting on compressed spring. Incorporation of filler in a polymer matrix can raise 3 types of interactions (1) matrix to filler, (2) matrix to matrix and (3) filler to filler. It may act for the mechanical advantage of nanocomposite matrix if interaction falls in this order i.e. 1 > 2 > 3.

Fig. 6 Spring and molecule model representing strains acting on fillers and polymer–filler interactions.

Silane functionalization may facilitate this interaction and role of terminal functional group may be to increase this interaction. A compressional strain may resist a tensile pulling force while a tensile strain may be acting in the direction of that force. Higher the interaction of matrix with filler more will be transfer of this effect to the nanocomposite matrix but, for a lesser magnitude of interaction with higher compressional strain may still improve the mechanical properties. A similar condition was noticed in case of 1.5 GOSAM/PCU and 0.5 GOCSAL/PCU nanocomposite membranes. At higher loadings of GOCSAL, matrix to matrix or filler to filler interaction might have exceeded the filler to matrix interaction leading to decrease in mechanical properties.

XRD diffractograms

XRD patterns of fillers and nanocomposite membranes are shown in Fig. 6(a) and (b) respectively. GOSAM and GOCSAL may have better crystallinity attributed to 2 peaks at 2θ values of ∼21° and ∼42° with interlayer distance of 4.2 Å and 2.1 Å respectively, in comparison to GRSAM and GOCSAL having single 2θ peak at ∼22° with interlayer distance of 4.0 Å. Specific peak at 42° in GOSAM and GOCSAL depicts turbostratic band of disordered carbon materials.³¹

XRD diffractogram of PCU membrane shows characteristic 2θ peaks at 13.7°, 20° and 23° (Fig. 7(a)) wherein broad peak at 13.7° may be assigned to the internal structure of polycarbonate chains, namely, to the average distance of the carbonate groups. Disappearance of peak at 13.7° can be attributed to intercalation of fillers to the polymer and distortion of its orderly internal structure. For nanocomposite membranes, no new 2θ peak has appeared indicating that there might not be considerable change in membrane crystallinity (Fig. 7(b)). Absence of any peak corresponding to fillers also indicate that there was no filler aggregation detected by XRD which usually happens if % wt of fillers increase beyond dispersibility or lack of dispersion in PCU matrix.¹⁶

Fig. 7 XRD diffractogram of (a) PCU matrix, GOCSAL and GOSAM and (b) filler incorporated PCU nanocomposite membrane.

MTT assay

An MTT assay based upon activity of SDH, a mitochondrial enzyme complex bound to inner mitochondrial membrane, using L929 mammalian fibroblast cells was carried out to quantify the percentage cell viability of the electrospun membranes. The nanocomposite membranes with superior mechanical properties under study behaved very similar to reference and PCU membrane. Comparative cell viability percentage for each of the electrospun polymer and the polymer nanocomposite matrix is graphically represented in Fig. 8. Nanocomposite membranes incorporated with fillers were hence found to be non cytotoxic with good cell viability percentage as the parent material.

Fig. 8 MTT assay plot of electrospun PCU and nanocomposite membranes.

Hemolysis assay

A hemolysis assay quantifies the damage to the red blood cell (RBC) membrane for the period of blood contact with implanted material. The disruption of RBC membrane can cause its content to leak into the blood stream which is not desired in case of an implantable biomaterial. Fig. 9 shows the hemolysis percentage for all of the superior electrospun nanocomposite membranes. The results obtained supported the fact that the hemolysis percentage for the fibroporous membranes developed was less than 5% as recommended by ISO10993-4. This indicated non-hemolytic characteristics of the PCU and its superior bionanocomposite membranes for blood contacting applications.

Fig. 9 Percentage hemolysis plot of electrospun PCU and nanocomposite membranes.

Conclusions

Silane modified graphene based nanosheet fillers with different functional groups on surface were synthesized using a dual-functional organosilane by applying different chemistries have proven to offer good reinforcing component for PCU. Raman spectral analysis indicated that fillers had different spectral signature based upon the functionalized terminal group exposed over filler surface and after reduction by HI acid. Spectral shifts indicated presence of compressive and tensile strains on varying scales depending upon magnitude of shift. The fillers with compressive strain at particular loadings showed better performance. Incorporation of individual fillers into PCU matrix affected the mechanical properties differently, GOCSAL with highest compressive strain contributed best at 0.5 wt% loading while GOSAM with comparatively lesser compressive strain showed better properties at 1.5 wt%. Thus, we can say that a filler having matrix compatible functional group along with high compressional strain can contribute maximum towards superior mechanical properties of nanocomposite membrane. A spring and molecule model has been proposed to state the same. Based on this study a rapid and simple tool like Raman spectral study of graphene fillers functionalized with different functional groups can provide us insight into predicting their physicomechanical performance inside a polymer matrix. Cytotoxicity and hemocompatibility of PCU bionanocomposite fibroporous membranes were found to be intact indicating its potential application in biomedical field. We hope that further research into these filler matrix interactions of nanosheet type fillers based upon interphase improvement and type of strain present in filler could further extend our knowledge to new horizons.

Acknowledgements

This study was performed at the Sree Chitra Tirunal Institute for Medical Sciences and Technology (SCTIMST) in collaboration with the Indian Institute of Technology Madras (IITM). One of the authors (S. T.) acknowledges the IITM for fellowship and the SCTIMST for facility provided to perform this study.

References

1. A. K. Geim and K. S. Novoselov, Nat. Mater., 2007, 6, 183–191.
2. K. Kostarelos and K. S. Novoselov, Science, 2014, 344, 261–263.
3. R. Verdejo, M. M. Bernal, L. J. Romasanta and M. A. Lopez-Manchado, J. Mater. Chem., 2011, 21, 3301–3310.
4. J. Liang, Y. Huang, L. Zhang, Y. Wang, Y. Ma, T. Guo and Y. Chen, Adv. Funct. Mater., 2009, 19, 2297–2302.
5. M. A. Rafiee, J. Rafiee, I. Srivastava, Z. Wang, H. Song, Z. Z. Yu and N. Koratkar, Small, 2010, 6, 179–183.
6. R. J. Zdrahala and I. J. Zdrahala, J. Biomater. Appl., 1999, 14, 67–90.
7. V. Kanyanta and A. Ivankovic, J. Mech. Behav. Biomed. Mater., 2010, 3, 51–62.
8. T. Courtney, M. S. Sacks, J. Stankus, J. Guan and W. R. Wagner, Biomaterials, 2006, 27, 3631–3638.
9. T. Ramanathan, A. Abdala, S. Stankovich, D. Dikin, M. Herrera-Alonso, R. Piner, D. Adamson, H. Schniepp, X. Chen and R. Ruoff, Nat. Nanotechnol., 2008, 3, 327–331.
10. Y. R. Lee, A. V. Raghu, H. M. Jeong and B. K. Kim, Macromol. Chem. Phys., 2009, 210, 1247–1254.
11. Y. Xie, C. A. Hill, Z. Xiao, H. Militz and C. Mai, Composites, Part A, 2010, 41, 806–819.
12. N. Suzuki and H. Ishida, 1996.
13. Y.-J. Wan, L.-X. Gong, L.-C. Tang, L.-B. Wu and J.-X. Jiang, Composites, Part A, 2014, 64, 79–89.
14. G. A. Buxton and A. C. Balazs, J. Chem. Phys., 2002, 117, 7649–7658.
15. M. Terrones, O. Martín, M. González, J. Pozuelo, B. Serrano, J. C. Cabanelas, S. M. Vega-Díaz and J. Baselga, Adv. Mater., 2011, 23, 5302–5310.
16. S. Thampi, V. Muthuvijayan and R. Parameswaran, J. Appl. Polym. Sci., 2015, 132, 41809.
17. X. Sun, Z. Liu, K. Welsher, J. T. Robinson, A. Goodwin, S. Zaric and H. Dai, Nano Res., 2008, 1, 203–212.
18. C. C. Chan, N. R. Choudhury and P. Majewski, Colloids Surf., A, 2011, 377, 20–27.
19. W. Zhang, Z. Guo, D. Huang, Z. Liu, X. Guo and H. Zhong, Biomaterials, 2011, 32, 8555–8561.
20. D. S. Holt, M. Botto, A. E. Bygrave, S. M. Hanna, M. J. Walport and B. P. Morgan, Blood, 2001, 98, 442–449.
21. Z.-A. Qiao, Y. Wang, Y. Gao, H. Li, T. Dai, Y. Liu and Q. Huo, Chem. Commun., 2010, 46, 8812–8814.
22. R. Imani, S. H. Emami and S. Faghihi, J. Nanopart. Res., 2015, 17, 1–15.
23. S. Rana, S. Maddila, K. Yalagala, S. Maddila and S. B. Jonnalagadda, ChemistryOpen, 2015, 4, 703–707.
24. G. Fischer, S. Braun, R. Thissen and W. Dott, J. Microbiol. Methods, 2006, 64, 63–77.
25. A. C. Ferrari and D. M. Basko, Nat. Nanotechnol., 2013, 8, 235–246.
26. X. Zheng, W. Chen, G. Wang, Y. Yu, S. Qin, J. Fang, F. Wang and X.-A. Zhang, AIP Adv., 2015, 5, 057133.
27. S. Pei, J. Zhao, J. Du, W. Ren and H.-M. Cheng, Carbon, 2010, 48, 4466–4474.
28. D. H. Reneker and I. Chun, Nanotechnology, 1996, 7, 216.
29. M. Fang, Z. Zhang, J. Li, H. Zhang, H. Lu and Y. Yang, J. Mater. Chem., 2010, 20, 9635–9643.
30. D. S. Jones, Int. J. Pharm., 1999, 179, 167–178.
31. L. Chen, Z. Xu, J. Li, C. Min, L. Liu, X. Song, G. Chen and X. Meng, Mater. Lett., 2011, 65, 1229–1230.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra21436c

---