Carbon nanotube reinforced flexible multifunctional regenerated cellulose films for nonlinear optical application

Abstract

Flexible regenerated cellulose/multiwalled carbon nanotube (RC–CNT) composites were successfully fabricated using a simple solution blending technique. The influence of functionalized multiwalled carbon nanotubes (f-CNTs) on the mechanical, electrical, thermal and optical properties of the regenerated cellulose was evaluated. Field emission scanning electron microscopy (FESEM) confirmed the uniform integration of f-CNT into the regenerated cellulose (RC) matrix. The tensile results indicated that the mechanical properties of the RC film increased considerably with f-CNT loading. The composites with 1.0 wt% f-CNT content (RC–CNT1.0) showed 60% increase in Young's modulus and 40% increase in elongation at break in combination with 50% increase in toughness, compared with RC film. The measured electrical conductivity of the composite film (RC–CNT1.0) was 6.2 × 10⁻⁴ S cm⁻¹, which is significantly higher than that of the RC film (6.8 × 10⁻¹² S cm⁻¹). Improvement in conductivity as well as dielectric permittivity values of the RC–CNT composites can be attributed to the unique network structure of f-CNT in the RC matrix. The RC–CNT composites showed high dielectric constant in combination with low dielectric loss, confirming their suitability for capacitor applications. All the composites showed a good optical limiting effect at 532 nm. The third-order nonlinear optical properties of composite films were investigated by Z-scan technique at a wavelength of 532 nm with pulse duration of 7 ns. The conductivity, thermal, dielectric, mechanical and nonlinear optical (NLO) properties were found to increase with f-CNT loading. Thus, the RC–CNT composites have turned out to be excellent next generation materials for energy storage and photonic devices.

---

Introduction

With the advent of new technologies, the need for materials possessing good mechanical strength and desirable opto-electronic properties is increasing day by day. The concerns over environmental pollution and the energy crisis have shifted the focus from traditional petrochemical-based polymeric materials to more eco-friendly alternatives. Flexible electronic devices based on ‘Green composite materials’ has attracted world-wide attention.^1,2 Hence, specific interest has been shown in the replacement of synthetic polymers with natural renewable resource based biopolymers such as chitosan, chitin, pectin, starch and cellulose.^3–8 Among the biodegradable materials, cellulose, the most abundant and inexhaustible biomass in nature, has received significant attention in various fields as an ideal candidate to replace petroleum based materials owing to its many fascinating properties such as biocompatibility, high mechanical properties and good thermal as well as chemical stability.^3,4 Nevertheless, pristine regenerated cellulose film suffers from low toughness and low optical as well as electrical performances, which restrict its application in certain fields. An effective strategy to overcome this limitation is the fabrication of high performance and multifunctional hybrid materials by the incorporation of nanofillers like carbon nanotubes (CNTs). As they possess unique and superior mechanical, thermal, electrical and optical properties, on account of their long range π-conjugation and large length to diameter ratio, CNTs have great potential to be utilized in the preparation of reinforced and functionalized cellulose nanocomposites,^9–14 realizing good mechanical as well as conducting properties at low percolation threshold. CNTs have wide applications in the field of antistatic packing, transistor, bioelectrode, electromagnetic shielding, sensors and actuators.^15–19

In this paper, we have attempted to blend f-CNT in RC matrix, so as to improve the electrical and nonlinear optical performance of cellulose to attain a good multifunctional material. The performance of CNTs/polymer nanocomposites depends on several factors such as, the synthesis process used to produce CNTs, purification process of CNTs,²⁰ the polymer system used, homogeneous distribution of CNTs in the polymer system and compatibility between the polymer and CNTs to ensure efficient stress transfer from the matrix to the CNTs.²¹ The major drawback of these composites is the poor dispersion of CNTs in the polymer matrix because of its large aggregation due to intrinsic van der Waals and π–π interactions. Hence, allowing proper dispersion of CNTs in the polymer matrix is of vital importance in the preparation of CNT/polymer nanocomposites. It is very difficult to disperse CNTs using ordinary fabrication methods like simple mechanical mixing. Common strategies for improving dispersion, like surface modification with suitable modifier and addition of surfactant, may adversely affect its electrical and optical properties. Here, we employ simple solution blending technique for the fabrication of RC–CNT composites and investigated the effect of f-CNT in cellulose matrix in terms of its mechanical, electrical, thermal and optical properties. Even though multiwalled carbon nanotubes (MWCNT) is known to have good NLO properties, till now the NLO properties of regenerated cellulose-multiwalled carbon nanotube composites remain unexplored. This fact prompted the NLO studies of the RC–CNT composites. In the present work, NLO and optical limiting performance of RC–CNT composites are investigated by Z-scan technique. Besides, we report the dielectric properties of RC–CNT composites and also explore its feasibility as a flexible NLO material.

Experimental

Materials

The cellulose microfibers, extracted from coconut leaf midrib were used for the composite fabrication. MWCNT Baytube^R 150P was supplied by Bayer Materials Science AG (Leverkusen, Germany). Dimethyl acetamide (DMAc), lithium chloride (LiCl), isopropanol and concentrated nitric acid were provided by Spectrochem Pvt. Ltd. and used as received.

Purification of MWCNT

Purification along with a small extent of functionalization was achieved by simple nitric acid treatment. The MWCNT was purified using the following method. 0.3 g of the as received MWCNT was dispersed in 25 mL of HNO₃ (65%) in a 100 mL round bottom flask equipped with a condenser and the dispersion was refluxed under magnetic stirring for 48 h. The resultant dispersion was diluted in water and filtered. The residue was washed with distilled water to neutral pH and dried overnight at 40 °C in vacuum. Nitric acid treatment removed impurities such as catalyst residues, carbon black and graphite nanoparticles.

Preparation of cellulose solution

In the present work, lithium chloride/N,N-dimethyl acetamide (LiCl/DMAc) solvent system was used for cellulose dissolution.²² The isolated cellulose microfibers and LiCl were heated under reduced pressure at 110 °C for 3 h to eliminate absorbed water. After the dissolution of LiCl in DMAc, the powdered cellulose fiber was treated with LiCl/DMAc solution in 2/8/90 proportions of cellulose/LiCl/DMAc. The mixture was heated at 110 °C followed by cooling to 40 °C and keeping for 30 min, which resulted in a clear cellulose solution.

Fabrication of RC–CNT composites

For the preparation of f-CNT dispersion, DMAc was used as the solvent. The surface energy of DMAc solvent matches very well with that of the carbon nanotubes. This result in minimum energy cost to overcome the van der Waals interactions between the f-CNT and leads to effective debundling.²³ Initially, a definite amount of f-CNT (15 mg) was well dispersed in LiCl/DMAc solvent system (40 mL) by ultra-sonication for one hour at a constant output power of 70 W. The concentration of the colloidal dispersion of f-CNT in LiCl/DMAc obtained was 0.37 mg mL⁻¹.

The RC–CNT composites were prepared by mixing the two solutions together according to the mass ratios of cellulose to f-CNT as 100/0 (RC), 99.8/0.2 (RC–CNT0.2), 99.6/0.4 (RC–CNT0.4), 99.2/0.8 (RC–CNT0.8), and 99.0/1.0 (RC–CNT1.0). The mixed solutions were stirred vigorously at 85 °C for 0.5 h; the use of same solvent system allows adequate contacts between cellulose and f-CNT, resulting in the uniform distribution of f-CNT and strong interfacial adhesion between f-CNT and cellulose matrix. To fabricate the composite films, the cellulose/f-CNT dispersions were cast on a glass tray (15 × 15 cm²) at room temperature and allowed to cure slowly. The slow curing may inhibit aggregation of f-CNT and remove residual LiCl and DMAc completely.²⁴ The resultant composite films were washed thoroughly with isopropanol–water mixture and then with running water until there were no salts. Finally, the films were dried in vacuum to obtain RC–CNT composite films. The flexible films had thickness in the range of 0.02–0.04 mm which could be controlled by adjusting the volume of cellulose/f-CNT dispersion during the casting process.

Characterization techniques

Infra-red absorption spectra were taken by a FTIR spectrometer (Perkin Elmer Spectrum 100) to investigate the presence of hydroxyl functional group on the surface of f-CNT and its interaction with regenerated cellulose matrix. f-CNT and composites were characterized by X-ray Photoelectron Spectroscopy (XPS) measurements with Kratos Analytical XPS, which was carried out using a monochromatic Al anode at energy of 40 eV. XPS peak 41 software was used to perform curve fittings. The Raman spectra have been recorded using HORIBA/JOBIN YVON HR800 UV Raman and photoluminescence spectrophotometer with a 200 mW argon-ion laser at an excitation wavelength of 633 nm and power density of 25.46 W cm⁻². The samples were analyzed with a 50× objective and an integration time of 0.5 s. For a better comparison, each spectrum has been normalized to the highest peak in the range of 800 to 2000 cm⁻¹. To determine the crystal structure and crystallinity, wide-angle X-ray diffraction (WXRD) patterns of the RC and RC–CNT composites were obtained from Rigaku D-Max X-ray diffractometer. Before testing, RC and RC–CNT composites samples were dried in a vacuum oven at 60 °C for 24 h to remove moisture. The diffraction experiments were carried out with Cu-Kα radiation (λ = 1.542 Å) at 30 kV and 30 mA over the range of 2θ = 5–40°, a size step of 0.05°. The crystallinity index (CI) for RC and RC–CNT composites was calculated using eqn (1):²⁵where, I₍₂₀₀₎ is the intensity at (200) peak (2θ = 22°) and I_(Am) is the intensity at the minimum between (110) and (200) peaks. It is assumed that intensity of (200) peak represents both crystalline and amorphous parts, while the minimum intensity at the mentioned location is for amorphous part only. Thermal degradation behavior of RC film and its f-CNT composites was studied using thermogravimetric analysis (TA Instrument Q50). The thermograms were run under a nitrogen atmosphere at a heating rate of 10 °C min over a temperature range of 40–700 °C. The surface morphology of the samples was observed using field emission scanning electron microscopy (FESEM, JEOL 7400-FS). The FESEM images were obtained by placing the samples on an aluminum SEM disk, which was then coated with gold. The acceleration voltage was 5 kV. Surface and internal morphologies of the samples were investigated using a transmission electron microscope (TEM, JEOL JEM-2100), operated at an acceleration voltage of 160 kV. Transmittance spectra were recorded with Thermo Scientific, Evolution 201, Ultraviolet-visible (UV-visible) spectrophotometer.

According to ASTM standard D882-97, the mechanical performances of the composite films were evaluated on a Universal testing machine (10 kN), Shimadzu Autograph AG-I series with a gauge length of 50 mm at a testing speed of 5 mm min⁻¹. The films were cut into rectangular strips with dimensions of 70 × 10 mm. The tensile toughness of the samples was obtained by integrating the area under the stress–strain curves.^26,27 Mechanical properties like tensile strength, Young's modulus, elongation at break and tensile toughness were evaluated by taking average values of at least five specimens. The DC electrical conductivity of RC–CNT composite films was measured by a Keithley 2400 measure unit; copper sheets were attached to the surface of each sample to ensure good contact of the sample surface with the electrode. The average value of at least five measurements was calculated for every set of films. Dielectric measurements were carried out at frequencies ranging from 40 Hz to 40 MHz using an Impedance analyzer, Agilent 4294A Precision LCR meter. Dielectric constant (ε′), dielectric loss (ε′′) and the AC conductivity (σ_AC) were obtained from the instrument. The results were directly read on the monitor and recorded on a computer data sheet file.

NLO measurements

NLO measurements were carried out by the single beam Z-scan technique with nanosecond laser performed with a Q-switched Nd:YAG laser system (Spectra Physics LAB-1760) with pulse width of 7 ns at 10 Hz repetition rate and 532 nm wavelength. The sample was moved in the direction of light propagation near the focal spot of the lens with a focal length of 200 mm. The radius of the beam waist was calculated to be 42.56 μm. The Rayleigh length, Z_o = πω_o²/λ, was calculated to be 10.69 mm, which was greater than the thickness of the sample, an essential requirement for Z-scan experiments. The Z-scan system was calibrated using CS₂ as the standard. The transmitted beam energy, reference beam energy and their ratios were measured simultaneously by an energy ratiometer (REj7620, Laser Probe Corp.) having two identical pyroelectric detector heads (Rjp 735). The effect of fluctuations of laser beam was eliminated by dividing the transmitted power by the power obtained at the reference detector; both being measured using identical photo detectors. The data were analyzed according to the procedure described by M. S. Bahae et al.,²⁸ and the nonlinear optical coefficients were obtained by fitting the experimental Z-scan plot with the theoretical plots.

Results and discussion

The RC–CNT composites with different nanotube loading of 0.2, 0.4, 0.6, 0.8 and 1.0 wt% were fabricated by solution blending technique and followed by simple regeneration. At first, the MWCNT were purified by nitric acid treatment and hence MWCNT were decorated with a little amount of polar groups such as carboxyl and hydroxyl groups, which is believed to promote its affinity towards cellulose chain by physical cross-linking.

The functionalized MWCNT (f-CNT) was uniformly dispersed in LiCl/DMAc solution without adding any surfactant. The addition of f-CNT dispersions does not deteriorate the stability of cellulose/LiCl/DMAc solution. In this way, good dispersion of f-CNT in the RC matrix could be effected which allowed the facile fabrication of RC–CNT composite films. Fig. 1a illustrates the preparation of RC–CNT composite films. The optical photographs of RC and RC–CNT1.0 composite films are shown in Fig. 1b and c. As seen in Fig. 1b, the films are transparent with a clear view of the background, indicating that the f-CNT was well dispersed in the RC matrix. As far as the device performance is concerned, the bending stability of a film is an important parameter. In Fig. 1c, RC–CNT1.0 composite films exhibited smooth surfaces along with good bending stability. The RC–CNT1.0 composite film can be bend upto 180° without breaking.

Fig. 1 (a) Fabrication of RC–CNT composites, (b) optical photographs of RC and RC–CNT1.0 and (c) bending stability of RC–CNT1.0.

Fourier transform infrared spectroscopy (FTIR)

Fig. 2a shows FTIR spectra of the RC, MWCNT, f-CNT and RC–CNT1.0 composite film. The characteristic peak observed at 1745 cm⁻¹ indicates that the surface of f-CNT was decorated with carboxyl groups. The peaks at 2854 and 2921 cm⁻¹ were ascribed to –CH stretching and the peak at 3436 cm⁻¹ was attributed to the –OH stretching, confirming the oxidation of f-CNT by concentrated nitric acid.²⁹ Pristine MWCNT is devoid of any characteristic peaks in this region. The RC spectrum shows the characteristic vibrations at 1158, 1052 and 3348 cm⁻¹, which are assigned to the C–O–C vibration between the glucose units, coupled C–O stretching and –OH stretching of cellulose units, respectively.³⁰ Compared with the RC film, the RC–CNT1.0 composite film showed an additional doublet peak around 1730 cm⁻¹ owing to the incorporation of f-CNT. The FTIR spectrum of RC–CNT1.0 composite film exhibits the combination of f-CNT and RC spectra, which confirms the integration of f-CNT in the RC matrix. Nevertheless, the –OH stretching band of the RC–CNT1.0 composite film was shifted to a lower frequency of 3321 cm⁻¹, which indicates the presence of intermolecular interactions such as hydrogen bonding between f-CNT and cellulose units.³¹

Fig. 2 FTIR (a) and Raman (b) spectra of f-CNT, RC and RC–CNT1.0.

Raman spectroscopy

To study the interaction between f-CNT and RC matrix, Raman spectra of f-CNT, RC and RC–CNT1.0 composite were recorded. As seen in Fig. 2b, the Raman spectrum of RC reveals two bands centred at 900 cm⁻¹ and 1100 cm⁻¹, which are the characteristic bands from cellulose II form.³²

Raman spectrum of f-CNT shows D-band (1325 cm⁻¹) and G-band (1562 cm⁻¹) corresponding to disorder induced by defects and in-plane vibration of the C–C bonds, respectively.³³ RC–CNT1.0 composite film displays four bands: two from f-CNT (1331 and 1575 cm⁻¹) and two from cellulose moieties (917 and 1124 cm⁻¹), which confirms the integration of f-CNT in the RC matrix. To evaluate the extent of carbon containing defects of the RC–CNT1.0 composite film, the integral area ratio of the D-band and G-band was calculated³⁴ and is obtained to be in the range of 1.4 which is obviously close to I_D/I_G of f-CNT (1.3). It indicates that the f-CNT is not destroyed during the composite preparation. However in the RC–CNT1.0 composite film, up-shifts in both D-band and G-band were observed. This confirms the presence of strong non-covalent interactions such as hydrogen bonds between f-CNT and RC matrix.^34–36

X-ray photoelectron spectroscopy (XPS)

The presence of functional groups on f-CNT was further investigated using XPS spectra. Fig. 3a shows the wide scan spectra of f-CNT, RC and RC–CNT1.0. As seen in Fig. 3a, f-CNT, RC and RC–CNT1.0 samples showed C1s and O1s signals at 286 and 531 eV, respectively. From the survey spectrum of f-CNT, it is clear that a considerable amount of oxygen functional groups are present within f-CNT, which were introduced by the purification process. No notable peaks are observed at binding energy of 55.5 eV, confirming the complete removal of Li⁺ ions from the composite film. Fig. 3b and c show deconvoluted C1s peak of f-CNT and RC–CNT1.0 samples, respectively. The main peak centered at 284.1 eV confirms the graphitic structure of f-CNT.³⁷ Peak centered at 285.2 eV shows the presence of defects on the carbon nanotube surface, while the peak centered at 286.5 eV represents carbon atoms bonded to different oxygen containing functional groups,³⁸ which are responsible for hydrogen bonding between cellulose and f-CNT moieties. RC–CNT1.0 composite film shows a notable peak centered at 284.6 eV attributed to the graphitic structure of f-CNT. The results clearly indicate that f-CNT was successfully inserted in to the RC matrix by the composite preparation.

Fig. 3 (a) XPS spectra of RC, f-CNT and RC–CNT1.0, C1s high resolution spectra of (b) f-CNT and (c) RC–CNT1.0, and (d) XRD spectra of RC, f-CNT and RC–CNT1.0.

X-ray diffraction (XRD)

Fig. 3d shows X-ray diffraction patterns of the RC–CNT composites in comparison with f-CNT and RC film. The XRD pattern of f-CNT shows a broad peak centered at 26° which is a characteristic peak of MWCNT.³⁹ RC film shows diffraction peaks at 12° and 22° which are obtained from the (110) and (200) planes respectively, based on the cellulose II crystal structure.⁴⁰ Like RC film, all the RC–CNT composites reveal the typical cellulose II crystalline structure. This indicates that inclusion of f-CNT does not affect the crystalline structure of RC film even if f-CNT was well dispersed. The intensity of the characteristic diffraction peak of f-CNT is too weak to be recognized in the diffraction patterns of composite films due to the very low loading of f-CNT. The diffraction intensity of the RC–CNT composites was slightly lower than that of RC film, implying a lower crystallinity of the former. This could be due to the interaction between f-CNT and RC matrix which hinder the formation of crystalline regions among cellulose chains.³⁶ The crystallinity index values of the composites are calculated and summarized in Table 1. Low crystallinity index (CI) values of composite films indicate the lowering of intermolecular interactions between the cellulose moieties and formation of intermolecular interactions between f-CNT and RC matrix.

Table 1 Mechanical and electrical properties of RC and RC–CNT composites

Sample	CNT loading (%)	CI (%)	Tensile strength (MPa)	Elongation at break (%)	Young's modulus (GPa)	Toughness (MPa)	Conductivity (S cm^−1)
RC	0.0	54	94 ± 0.5	6.5 ± 0.3	3.8 ± 0.2	4.68 ± 0.2	6.8 × 10^−12
RC–CNT0.2	0.2	53	101 ± 0.2	6.7 ± 0.4	4.6 ± 0.3	6.7 ± 0.3	1.2 × 10^−6
RC–CNT0.4	0.4	51	112 ± 0.2	7.2 ± 0.4	5.2 ± 0.3	7.3 ± 0.1	8.2 × 10^−6
RC–CNT0.6	0.6	50	120 ± 0.5	7.5 ± 0.3	5.5 ± 0.2	8.8 ± 0.4	5.2 × 10^−5
RC–CNT0.8	0.8	49	126 ± 0.6	8.6 ± 0.6	5.7 ± 0.1	9.0 ± 0.3	12.8 × 10^−5
RC–CNT1.0	1.0	47	134 ± 1.0	9.1 ± 0.5	6.1 ± 0.3	9.4 ± 0.3	6.2 × 10^−4

---

Thermogravimetric (TG) analysis

To investigate the thermal stability of RC–CNT composites, TG analysis for MWCNT, f-CNT, RC and RC–CNT composite films was carried out and is depicted in Fig. 4. As seen in Fig. 4 (insert figure), pristine MWCNT shows no weight loss below 600 °C. The thermal degradation begins above 600 °C, which is due to the occurrence of disordered and amorphous carbon on the MWCNT.⁴¹ On the other hand, f-CNT shows a weight loss of 25% before 600 °C, owing to the decomposition of carboxyl and hydroxyl groups on the nanotube surface. In Fig. 4, RC film begins to lose weight at 247 °C. For composite containing 1.0 wt% of f-CNT, the onset of degradation is above 257 °C, which is 10 °C higher than that of RC film. Decomposition of RC and RC–CNT composite films consist of two steps. The minor weight loss between 50–200 °C and major weight loss between 250–400 °C are attributed to the evaporation of absorbed water and decomposition of cellulose backbone, respectively.⁴²

Fig. 4 Thermal stability of f-CNT (insert figure) and RC–CNT composites.

All the RC–CNT composite films show higher onset of degradation and peak degradation temperature than that of RC film indicating the improved thermal stability of RC–CNT composites. This confirms the presence of non-covalent interaction like intermolecular hydrogen bonding between f-CNT and RC matrix.

Mechanical properties

The mechanical properties of RC–CNT composite films were investigated by tensile test method. Fig. 5 shows the stress–strain characteristics of RC–CNT composite films with varying degree of f-CNT content, in comparison with RC. The RC film has high tensile strength (94 MPa) as well as Young's modulus (3.8 GPa) and low strain to failure (6.5%), which is in accordance with a previous report.⁴³ The mechanical properties of the RC and RC–CNT composite films are summarized in Table 1.

Fig. 5 Stress–strain curves of RC and RC–CNT composites.

Incorporation of f-CNT improves the tensile strength and Young's modulus of RC film which increases with the increase of f-CNT loading. The RC–CNT composite containing 1.0 wt% f-CNT exhibited a 42.5% increment in tensile strength as compared with RC sample. Young's modulus of RC–CNT1.0 composite is obtained to be 6.1 GPa, which is approximately 60% higher than that of RC film. Moreover, elongation at break of the RC–CNT1.0 composite is obtained to be 9.1%, which is very much higher than that of RC film (6.5%), indicating high flexibility of the composite films. Thus, the incorporation of f-CNT into the RC matrix results in an increase in strength of the composite with simultaneous increase of toughness and stretchability. The improved mechanical properties like tensile strength, Young's modulus and toughness could be attributed to the homogeneous distribution of f-CNT together with good intermolecular interaction between f-CNT and cellulose moieties, resulting in a high strength material. A modulus enhancement of polymers resulting from the incorporation of either single walled carbon nanotubes (SWCNTs) or MWCNT has been found in many studies.⁴⁴ The nanostructure characteristics of f-CNT together with high interfacial bonding, delay material damage during deformation and explain the observed ductility as well as high toughness of the films. The solution blended RC–CNT composites exhibited a comparable mechanical performance to covalently grafted polymer–CNT composites.⁴⁵ The former possess obvious advantages owing to the simple preparation process. The compatibility and strong interactions between the f-CNT and RC matrix improved the dispersion and thus significantly enhanced mechanical properties of the RC matrix.

Morphological analysis

To observe the morphological changes during composite preparation, FESEM investigation was carried out. The surface morphologies of RC and RC–CNT1.0 composite film are illustrated in Fig. 6a and b. The surface of RC film is highly uniform and smooth due to strong hydrogen bonding network. However, the surface of composite film is relatively rougher than RC film. The rough surfaces of the composite film might be associated with multiple bonding between cellulose and f-CNT. Fig. 6c and d represents the FESEM cross-section images of RC and RC–CNT1.0 composite samples, respectively. The cross-section image of RC film reveals it has layer-by-layer structure owing to inter and intra-molecular hydrogen bond network.^46,47 From the fracture images of composite film, it can be seen that the f-CNT was uniformly distributed in cellulose matrix without disrupting the layer-by-layer structure, indicating good adhesion between the f-CNT and RC matrix. A well ordered layer-by-layer structure is preserved while the spacing between the layers is increased. Fig. 7 shows the TEM image of regenerated cellulose composite with 1.0 wt% f-CNT. It can be seen that the f-CNT were well dispersed in RC matrix with some agglomeration.

Fig. 6 Surface (a & b) and cross-section (c & d) images of RC and RC–CNT1.0.

Fig. 7 TEM image of RC–CNT1.0 film.

Transparency

The variation in transmittance of RC and RC–CNT composite films are illustrated in Fig. 8. The transmittance of RC film is ∼80% at 750 nm while the transmittance of RC–CNT0.2 is obtained to be ∼78%. On increasing the concentration of f-CNT, the transparency of RC–CNT composites decreased, as expected. When the concentration of f-CNT was increased to 1.0 wt%, the transparency of RC film decreased from 80% to 65%. However, all the composite films showed reasonable transmittance, which is inevitable in optical applications.

Fig. 8 Transmittance spectra of RC and RC–CNT composites.

Electrical conductivity

To study the electrical properties of RC–CNT composite films, I–V measurements were carried out. Variation of DC conductivity of RC–CNT composites with f-CNT content are summarized in Table 1. The electrical conductivity of RC film (6.8 × 10⁻¹² S cm⁻¹) increased rapidly with the addition of 0.2 wt% f-CNT to 1.2 × 10⁻⁶ S cm⁻¹. The sudden increase in conductivity indicates that rod shaped f-CNT was homogeneously inserted and tunnels within the RC matrix, as revealed from SEM and TEM observations. Conductivity of RC–CNT composites increased with f-CNT loading. RC–CNT composite film with 1.0 wt% f-CNT has conductivity of about 6.2 × 10⁻⁴ S cm⁻¹, which is very much higher than that of RC film. The electrical conductivity of these composite films are also highly uniform. This indicates that the dispersion of f-CNT in the cellulose matrix by solution blending technique is more efficient and uniform than by using direct mechanical mixing methods.

Dynamic mechanical analysis (DMA)

To investigate the viscoelastic properties of RC–CNT composites in comparison with RC film, DMA was carried out. The variation of storage modulus (G′) with temperature of RC–CNT composites of different f-CNT loading is shown in Fig. 9.

Fig. 9 Variation of storage modulus of RC–CNT composites with temperature.

From the figure it can be seen that the storage moduli of all the RC–CNT composites are higher than that of pristine RC film and this enhancement is particularly significant at lower temperatures. The RC–CNT1.0 composite exhibited a 1.3 fold increase in the storage modulus at 50 °C and 1.24 fold increase at 150 °C, compared to RC film. This enabled the cellulose matrix to sustain high modulus upto higher temperatures. This high G′ values are attributed to the formation of strong hydrogen bonding between the f-CNT and RC matrix. To measure the reduction of the mechanical properties of the materials with increasing temperature, retention ratio was calculated. The retention ratio is defined as the ratio of the storage modulus at 150 °C to that at 50 °C and for RC film it was only 0.85. But for RC–CNT1.0 composite, retention ratio is obtained to be 0.9, indicating the higher retention of storage modulus with increasing content of f-CNT. This reveals that the RC–CNT composite retains mechanical properties at higher temperatures, and hence can be used for high temperature applications.

Dielectric properties

To analyze the energy storage capability of RC–CNT composites, the dielectric responses of the samples, dielectric constant (ε′) and dielectric loss (ε′′) are measured in the frequency range of 40 Hz to 40 MHz. Fig. 10a depicts the variation of dielectric constant of RC–CNT composites as a function of frequency at different f-CNT loading. For comparison, dependence of dielectric constant and dielectric loss of RC film on frequency is also measured. The RC film showed a low dielectric constant of 14 at 100 Hz, which is in accordance with a previous report.⁴⁸ The dielectric constant of a material mainly depends on the amount of polarizable electrical charges in it and its degree of mobility. As observed in Fig. 10a, the dielectric permittivity of RC–CNT composites increases with increasing f-CNT loading over the entire frequency range. The CNTs are large domain of free charge carriers and can cause large increase in dielectric properties of RC–CNT composites owing to the interfacial and electronic polarization mechanisms,⁴⁹ i.e., the integration of f-CNT would lead to the formation of more micro capacitor networks within the RC matrix. The dielectric permittivity increases from 14 to 1580 as the f-CNT loading increases from 0 to 1.0 wt% at 100 Hz.

Fig. 10 Variation of (a) dielectric constant, (b) dielectric loss and (c) AC conductivity of RC and RC–CNT composites with frequency.

The dielectric constant values of RC–CNT composites are high at lower frequencies and then decrease exponentially with increase in frequency. At lower frequencies, all polarization mechanisms including dipolar, interfacial (space charge), electronic and ionic can follow the applied electric field, resulting in the higher value of dielectric constant, ε′. Among these, both dipolar and interfacial polarizations are more significant at lower frequencies, because it can greatly influence the capacitive properties.

At higher frequencies, the electric field change too fast, hence the polarization becomes out of phase with the applied field resulting in significant reduction in ε′. From the Fig. 10a, it is revealed that the RC–CNT1.0 composite showed a 97 fold increment in ε′ than that of RC at 500 Hz. This effect can be explained by Maxwell–Wagner (MW) interfacial polarization mechanism.⁵⁰ In heterogeneous systems, interfacial polarization occurs due to the accumulation of charges at the interfaces between various regions that may differ in its DC conductivity. This leads to the formation of micro-capacitor networks in these composites.

Fig. 10b represents the variation of dielectric loss of RC–CNT composites with frequency. Owing to the high conductivity of f-CNT, the dielectric loss of RC–CNT composites is increased with f-CNT loading. The dielectric loss (ε′′) of a composite is attributed to its DC conductance and interfacial polarization which mainly depends upon the type of matrix as well as the conductivity, particle size, crystal structure and more importantly the dispersion of filler used.^51,52 In the case of RC–CNT composites, the dielectric loss initially decreases and then remains almost constant. Higher values of ε′′ at lower frequencies is attributed to the interfacial polarization mechanism of the heterogeneous system. At higher frequencies, the ε′′ has minimum value because of the failure of the induced charges to follow the reversing field, thus leading to reduced electronic oscillations and the vibration of ions may be the only source of dielectric loss.⁵³

Fig. 10c depicts the variation of AC conductivity of RC–CNT composites with different f-CNT loadings in the frequency range of 40 Hz to 40 MHz. AC conductivity of RC–CNT composites is found to be increasing with increase in f-CNT loading and also with increase in frequency as evident from Fig. 10c. The increase of AC conductivity with frequency can be attributed to the increase in charge carriers and the increase in the hopping of these charge carriers. AC conductivity is varying from 3.16 × 10⁻⁸ to 2 × 10⁻⁶ S m⁻¹ depending on the amount of f-CNT. This indicates that conductivity of the composites can be controlled by varying the amount of f-CNT in the RC matrix. The number of free charge carriers per unit volume increases with increase in f-CNT loading. Another reason for increase in conductivity with increased loading of f-CNT could be due to the increase in MW interfacial polarization.⁵⁴ The conductivity of composite mainly depends upon aspect ratio, conductivity nature, doping level and dispersion of filler. The significant increase in conductivity of the RC–CNT composites could be due to the tunnelling effect of the electrons from one nanotube to another because of the high conductivity nature and better dispersion.⁵⁵ The tremendous increase in conductivity also reveals that f-CNT is networked quite well within the RC matrix, providing continuous path for charge carriers.

Nonlinear optical (NLO) properties

In recent times, applications of lasers have become very common in the fields of medicine, cosmetics and even in toys. Nevertheless, it has been recognized that intense laser beam can damage optical devices and especially human eye. Thus, the protection from laser induced damages is not only a scientific interest but also a potential public safety issue. This has led to extensive research in the field of optical limiting. An ideal optical limiter should strongly attenuate intense and potentially dangerous laser beams, while displaying high transmittance for low intensity light. The demand for flexible eco-friendly NLO materials is ever increasing because of the environmental impacts and depletion in petroleum resources. Cellulose is an attractive candidate as a suitable substrate for carbon nanotube loading owing to its light weight, mechanical stability, compatibility and environmental friendly characteristics. Hence, the development of RC–CNT composites would be highly noteworthy in various fields of modern optics.

To assess the potential application of RC–CNT composites with different f-CNT loading, third-order NLO measurements were done at 532 nm using Z-scan technique. To evaluate the NLO coefficients, both open and closed aperture scans were performed on ∼0.02 mm thin RC–CNT composite films at 73 μJ. As expected, RC film (Fig. 11a) exhibits slight decrease in optical transmission, indicating that there is observable nonlinear absorption effect. Fig. 11b displays the open aperture (OA) traces of RC–CNT composites at different f-CNT content. Here, Z-scan traces exhibit obviously different characteristics, i.e., a well-defined transmission valley at the focus, revealing the existence of reverse saturation effects with positive nonlinear absorption coefficient at 532 nm. This could be due to the molecules present in ground state and excited states may absorb the incident photons of identical wavelengths and more excited state absorption. The resulting OA Z-scan traces were fitted to eqn (2) to yield the magnitude of nonlinear absorption coefficient, β by employing standard Z-scan theory (two photon absorption theory):

where, q_0(z,r,t) = βI₀(t)L_eff and L_eff =  (1 − e^−αl)/α is the effective thickness with linear absorption coefficient α, and ‘I₀’ is the irradiance at focus.

Fig. 11 Open and closed aperture curves of (a & c) RC and (b & d) RC–CNT composites.

The solid curves in the Fig. 11 are the theoretical fit to the experimental Z-scan data (scattered dots). Analysis of OA transmittance data obtained from the RC–CNT composite films result in β in the order of 10⁻⁸ m W⁻¹, which is comparable to reported values of some MWCNT non-biodegradable composites.⁵⁶ As seen in Fig. 11a and b, experimental curve is fitted well with two photon absorption (TPA) theoretical model. This infers that TPA is the basic mechanism involved in nonlinear absorption process.

The nonlinear refractive (NLR) property of the composite films was investigated by closed aperture (CA) Z-scan configuration in comparison with pristine RC film. The NLR Z-scan curves after excluding nonlinear absorption effects can be attained by dividing the CA Z-scan data by the corresponding open aperture data (Fig. 11c and d). Here, the normalized transmittance T(z) is given by eqn (3):

As seen in Fig. 11c, the RC film displays a peak-valley pattern, indicating a negative nonlinear refractive (NLR) index due to self-defocusing effect. The CA traces of RC–CNT composites also show a post-focal peak followed by a pre-focal valley, indicative of a negative nonlinear refraction due to self-defocusing effect in the sample. From Fig. 11d, it can be seen that with the increase of concentration of f-CNT, the peak height and the valley depth increase synchronously, suggesting that increased nonlinear refraction at higher f-CNT contents. The nonlinear refractive index (n₂), the real parts of third-order nonlinear susceptibility (Reχ⁽³⁾), imaginary part of third-order nonlinear susceptibility (Imχ⁽³⁾) and third-order nonlinear susceptibility (χ⁽³⁾) of the composites are calculated⁵⁷ and listed in Table 2.

Table 2 Calculated values of third-order nonlinear absorption, refraction and nonlinear susceptibility of RC and RC–CNT composite films

Sample	β, m W^−1	n2, esu	Imχ^(3), esu	Reχ^(3), esu	χ^(3), esu
RC	7.29 × 10^−10	−1.27 × 10^−9	7.8 × 10^−11	2.56 × 10^−11	1.83 × 10^−11
RC–CNT0.2	2.84 × 10^−8	−1.82 × 10^−8	9.03 × 10^−10	3.32 × 10^−9	3.43 × 10^−9
RC–CNT0.6	3.05 × 10^−8	—	3.2 × 10^−9	—	—
RC–CNT1.0	3.35 × 10^−8	−1.88 × 10^−8	4.1 × 10^−9	3.58 × 10^−9	4.79 × 10^−9

---

The RC–CNT composite films hold strong NLO refractive effect with n₂ value of −1.27 × 10⁻⁹, −1.82 × 10⁻⁸ and −1.88 × 10⁻⁸ esu for RC, RC–CNT0.2, and RC–CNT1.0, respectively. Analysis of the resulting Z-scan traces yielded χ⁽³⁾ values of 1.83 × 10⁻¹¹, 3.43 × 10⁻⁹ and 4.79 × 10⁻⁹ esu for RC, RC–CNT0.2 and RC–CNT1.0, respectively. The values obtained from RC–CNT composites appear to be consistent with many previous determinations of the third-order nonlinear susceptibilities of CNT based systems that typically range from 10⁻⁸ to 10⁻¹⁰ esu, depending upon the particular polymer composite studied and experimental method employed.^58,59

Optical limiting properties

Better mechanical properties with good optical limiting performances are appropriate for the development of laser protection devices. To study the optical limiting properties of the samples, OA Z-scan technique at 532 nm was used.

Fig. 12 presents the variation of normalized transmittance as a function of input fluence. The data for pristine RC was also included for comparison. The optical limiting threshold of composite films are obtained to be 0.27, 0.19 and 0.12 GW cm⁻², for RC–CNT0.2, RC–CNT0.6 and RC–CNT1.0, respectively, indicating that all the films show an excellent optical limiting effect at 532 nm. The lower the optical limiting threshold better is the optical limiting material. Thus, the RC–CNT composites exhibit superior optical limiting performances than that of pristine RC film. The coiled polymer confirmations permit it to wrap the f-CNT, allowing sufficiently close intermolecular proximity for π–π interaction to occur. The strong hydrogen bonding interaction between the polymer chains and f-CNT also support enhancement of NLO properties. From the results, it is revealed that, optical limiting effect is dependent on the concentration of f-CNT. Highest amount of f-CNT results in lowest limiting threshold and best optical limiting efficiency. The results reveal that the RC–CNT composites are effective OL candidates of 532 nm nanosecond laser pulses in the field of opto-electronics.

Fig. 12 Optical limiting curves of RC (insert figure) and RC–CNT composites.

Conclusion

Multifunctional regenerated cellulose/multiwalled carbon nanotube (RC–CNT) composites were successfully fabricated by simple solution blending technique using LiCl/DMAc solvent system. X-ray diffraction patterns indicated that the inclusion of f-CNT does not affect the crystalline structure of RC matrix. The presence of non-covalent interaction of f-CNT to cellulose matrix was confirmed by FTIR, Raman, XPS and TG analysis. The microstructure of RC–CNT composites showed good dispersion of f-CNT within the polymer matrix. Tensile results revealed that tensile strength, elongation at break, Young's modulus and toughness increased by 42.5%, 40%, 60% and 100.8%, respectively with 1.0 wt% mixing of f-CNT in the cellulose matrix. On increasing the f-CNT loading from 0 to 1.0 wt%, the thermal stability was improved by 10 °C compared to RC film. Improved thermal stability and mechanical properties of the RC–CNT composite films arises from the uniform dispersion of f-CNT in the cellulose matrix as well as good interfacial interaction between cellulose and f-CNT. From the electrical studies, it is revealed that both conductivity and dielectric permittivity increases with the increase in f-CNT loading. Impedance analysis showed a high dielectric constant of 1580 and low dielectric loss of 12 at 100 Hz for RC–CNT1.0 composite film. The composite films showed lower optical limiting threshold of 0.27, 0.19 and 0.12 GW cm⁻² for RC–CNT0.2, RC–CNT0.6 and RC–CNT1.0, respectively, which is lower than the optical limiting value of RC film (0.46 GW cm⁻²), suggesting the good optical limiting efficiency at 532 nm. Thus the RC–CNT composites showed superior nonlinear optical and optical limiting properties than that of regenerated cellulose without deteriorating its crystalline structure. The results indicate that the prepared composites are potential candidates for environment friendly optical limiting device application.

Acknowledgements

The authors gratefully acknowledge financial support from Council of Scientific and Industrial Research (CSIR) and University Grants Commission Research Fellowships in Sciences for Meritorious Students (UGC-RFSMS), Department of Science and Technology (DST), Government of India.

References

1. S. Liu, T. Yu, Y. Wu, W. Li and B. Li, RSC Adv., 2014, 4, 34134.
2. A. Kafy, K. K. Sadasivuni, H. C. Kim, A. Akether and J. Kim, Phys. Chem. Chem. Phys., 2015, 17, 5923.
3. C. L. de Dicastillo, F. Rodríguez, A. Guarda and M. J. Galotto, Carbohydr. Polym., 2016, 136, 1052.
4. M. D. Harun-or-Rashid, M. D. S. Rahaman, S. E. Kabir and M. A. Khan, J. Appl. Polym. Sci., 2016, 133, 42870.
5. M. K. Gök, S. Özgümüş, K. Demir, Ü. Cirit, S. Pabuccuoğlu, E. Cevher, Y. Özsoy and S. Bacınoğlu, Carbohydr. Polym., 2016, 136, 63.
6. M. V. Lorevice, C. G. Otoni, M. R. de Moura and L. H. C. Mattoso, Food Hydrocolloids, 2016, 52, 732.
7. R. Fernández, C. Ocando, S. C. M. Fernandes, A. Eceiza and A. Tercjak, J. Renewable Mater., 2015, 3, 49.
8. C. Chen, C. Yang, S. Li and D. Li, Carbohydr. Polym., 2015, 134, 309.
9. J. H. Kim, S. Yun, H. U. Ko and J. Kim, Curr. Appl. Phys., 2013, 13, 897.
10. S. Yun and J. Kim, Smart Mater. Struct., 2007, 16, 1471.
11. R. E. Anderson, J. Guan, M. Ricard, G. Dubey, J. Su, G. Lopinski, G. Dorris, O. Bourne and B. Simard, J. Mater. Chem., 2010, 20, 2400.
12. L. R. Shobin and S. Manivannan, Sens. Actuators, B, 2015, 220, 1178.
13. S. Han, Q. Sun, H. Zheng, J. Li and C. Jin, Carbohydr. Polym., 2016, 136, 95.
14. S. Yun and J. Kim, Carbon, 2011, 49, 518.
15. F. H. Gojny, M. H. G. Wichmann, U. Köpke, B. Fiedler and K. Schulte, Compos. Sci. Technol., 2004, 64, 2363.
16. D.-J. Yun, Y. J. Jeong, H. Ra, J.-M. Kim, T. K. An, M. Seol, J. Jang, C. E. Park, S.-W. Rhee and D. S. Chung, J. Mater. Chem. C, 2015, 3, 7325.
17. E. Dilonardo, M. Penza, M. Alvisi, C. D. Franco, R. Rossi, F. Palmisano, L. Torsi and N. Cioffi, Sens. Actuators, B, 2016, 223, 417.
18. C. H. Phan, M. Mariatti and Y. H. Koh, J. Magn. Magn. Mater., 2016, 401, 472.
19. K. Czaniková, N. Torras, J. Esteve, I. Krupa, P. Kasák, E. Pavlova, D. Račko, I. Chodák and M. Omastová, Sens. Actuators, B, 2013, 186, 701.
20. M. Moniruzzaman and K. I. Winey, Macromolecules, 2006, 39, 5194.
21. A. A. Kovalchuk, A. N. Shchegolikhin, V. G. Shevchenko, P. M. Nedorezova, A. N. Klyamkina and A. M. Aladyshev, Macromolecules, 2008, 41, 3149.
22. X. Zhang, J. Feng, X. Liu and J. Zhu, Carbohydr. Polym., 2012, 89, 67.
23. S. D. Bergin, V. Nicolosi, P. V. Streich, S. Giordani, Z. Y. Sun, A. H. Windle, P. Ryan, N. P. Niraj, Z. T. Wang, L. Carpenter, W. J. Blau, J. Boland, J. P. Hamilton and J. N. Coleman, Adv. Mater., 2008, 20, 1876.
24. S. Yun, Y. Chen, J. N. Nayak and J. Kim, Sens. Actuators, B, 2008, 129, 652.
25. L. Segal, J. J. Creely, A. E. Martin and C. M. Conrad, Text. Res. J., 1959, 29, 786.
26. H. Xu, G. J. Zhong, Q. Fu, J. Lei, W. Jiang, B. S. Hsiao and Z.-M. Li, ACS Appl. Mater. Interfaces, 2012, 4, 6774.
27. P. A. Song, Z. Xu and Q. Guo, ACS Macro Lett., 2013, 2, 1100.
28. M. S. Bahae, A. A. Said, T.-H. Wei, D. J. Hagan and E. W. Van Stryland, IEEE J. Quantum Electron., 1990, 26, 760.
29. C. C. Chen, C. F. Chen, C. M. Chen and F. T. Chuang, Electrochem. Commun., 2007, 9, 159.
30. M. L. Nelson and R. T. Oconnor, Appl. Polym. Sci., 1964, 8, 1311.
31. M. Gopiraman, K. Fujimori, K. Zeeshan, B. S. Kim and I. S. Kim, eXPRESS Polym. Lett., 2013, 7, 554.
32. R. G. Zhbankov, S. P. Firsov, D. K. Buslov, N. A. Nikonenko, M. K. Marchewka and H. Ratajczak, J. Mol. Struct., 2002, 614, 117.
33. A. Jorio, M. A. Pimenta, A. G. S. Filho, R. Saito, G. Dresselhaus and M. S. Dresselhaus, New J. Phys., 2003, 5, 139.
34. L. Li, L. Meng, X. Zhang, C. Fu and Q. Lu, J. Mater. Chem., 2009, 19, 3612.
35. D. Baskaran, J. W. Mays and M. S. Bratcher, Chem. Mater., 2005, 17, 3389.
36. H. Qi, J. Liu, S. Gao and E. Mader, J. Mater. Chem. A, 2013, 1, 2161.
37. H. Ago, T. Kugler, F. Cacialli, W. Salaneck, M. Shaffer and A. Windle, J. Phys. Chem. B, 1999, 103, 8116.
38. G. Beamson and D. Briggs, The scienta ESCA A300 database, Wiley, Chichester, 1992.
39. J. Lu, H. Zhang, Y. Jian, H. Shao and X. Hu, J. Appl. Polym. Sci., 2012, 123, 956.
40. P. Langan, N. Sukumar, Y. Nishiyama and H. Chanzy, Cellulose, 2005, 12, 551.
41. S.-Y. Yang, C.-C. M. Ma, C.-C. Teng, Y.-W. Huang, S.-H. Liao, Y.-L. Huang, H.-W. Tien, T. M. Lee and K. C. Chiou, Carbon, 2010, 48, 592.
42. H. Dogan and N. D. Hilmioglu, Vacuum, 2010, 84, 1123.
43. X. Zhang, X. Liu, W. Zheng and J. Zhu, Carbohydr. Polym., 2012, 88, 26.
44. N. George, C. S. J. Chandra, A. Mathiazhagan and R. Joseph, Compos. Sci. Technol., 2015, 116, 33.
45. S. Yun and J. Kim, Carbon, 2011, 49, 518.
46. T. Barroso, M. Temtem, A. Hussain, A. Aguiar-Ricardo and A. C. A. Roque, J. Membr. Sci., 2010, 348, 224.
47. M. Soheilmoghaddam, M. U. Wahit, A. A. Yussuf, M. A. Al-Saleh and W. T. Whye, Polym. Test., 2014, 33, 121.
48. L. Zhai, S. Mun, X. Gao, J. W. Kim and J. Kim, Cellulose, 2015, 22, 927.
49. Q. Li, Q. Z. Xue, X. L. Gao and Q. B. Zheng, eXPRESS Polym. Lett., 2009, 3, 769.
50. A. A. Zaky and R. Hawley, Dielectric solids, Routledge & Kegan Paul Ltd. Great Britain, 1970.
51. C. G. A. Lima, R. S. Oliveira, S. D. Figueiro, C. F. Wehmann, J. C. Góes and A. S. B. Sombra, Mater. Chem. Phys., 2006, 99, 284.
52. S. Ameen, V. Ali, M. Zulfeguar, M. M. Haq and M. Husain, Curr. Appl. Phys., 2007, 7, 215.
53. S. Bhadra, N. K. Singha and D. Khastgir, Curr. Appl. Phys., 2009, 9, 396.
54. P. Hedvig, Dielectric Spectroscopy of Polymers, Hilger, Bristol, 1977, p. 282.
55. N. Joseph, C. Janardhanan and M. T. Sebastian, Compos. Sci. Technol., 2014, 101, 139.
56. Z. Jin, X. Sun, G. Xu, S. H. Goh and W. Ji, Chem. Phys. Lett., 2000, 318, 505.
57. S. Narayanan, A. Abbas, S. P. Raghunathan, K. Sreekumar, C. S. Kartha and R. Joseph, RSC Adv., 2015, 5, 8657.
58. Y. Zhu, H. I. Elim, Y. L. Foo, T. Yu, Y. Liu, W. Ji, J. Y. Lee, Z. Shen, A. T. S. Wee, J. L. Thong and C. H. Sow, Adv. Mater., 2006, 18, 587.
59. E. M. N. Mhuircheartaigh, S. Giordani and W. J. Blau, J. Phys. Chem. B, 2006, 110, 23136.

---