Performance Enhancement of Modified 3D SWCNT/RVC Electrodes Using Microwave-Irradiated Graphene Oxide
- 153 Downloads
The goal of this article is to increase the electrode performance of 3D CNT/RVC electrodes by improving the ease of ion adsorption to and ion desorption from the electrode surfaces. This achievement was done by preparing different composites of synthesized microwave-irradiated graphene oxide (mwGO) with CNT and coated on RVC. The morphology of GO was examined by field emission scanning electron microscopy (FESEM) and X-ray diffraction (XRD) study. Its surface property was checked by energy-dispersive X-ray spectra (EDX), and Fourier transform infrared spectra; whereas, for mwGO by XRD, Raman spectra, and X-ray photoelectron spectra (XPS), which revealed some structural changes of GO after irradiation, where CNTs, being sandwiched between graphene layers, built 3D highly porous architecture inside the electrodes. The electrochemical test of composite electrodes showed increased electrodes conductivity and afforded rapid ions diffusion. It is observed that the 9-CNT/mwGO/RVC composite electrode performed as the best electrode, which showed 29% increment in specific capacitance value compared to the normal CNT/RVC electrode. This best electrode also showed very high cyclic stability in its cyclic voltammetry test that maintained 97% current stability after 2000 cycles, indicating that the electrode can be an effective material for water purification technology.
KeywordsComposite electrode Surface chemistry Morphology Thermal stability Electrochemical property
Energy-dispersive X-ray spectra
Electrostatic impedance spectra
Field emission scanning electron microscopy
Microwave-irradiated graphene oxide
Reticulated vitreous carbon
Single-walled carbon nanotubes
X-ray photoelectron spectra
Various carbons like graphene-like carbon flake, multi-walled carbon nanotubes, electrospun fibres, activated carbon cloth, carbon aerogels, reticulated vitreous carbon (RVC), and ordered mesoporous carbon were used as electrodes in the past [2, 4, 47, 51]. These materials have been deemed good electrodes because of their easy accessibility, low cost, resistance to acidic and basic environments, low density, diverse porous structure, and other properties . The use of intrinsically conductive polymer polypyrrole (PPy) and polyethylene dioxylthiophene (PEDOT) as electrode materials is also reported [3, 5, 28].
Graphene material has recently become a material of interest in various applications such as capacitive deionization (CDI), supercapacitor, optical detection, electrochemical biosensing, nanoelectronic circuits, energy, battery, and analytical applications [8, 19, 21, 22, 38, 48, 61, 63]. This is because graphene has an impressive theoretical surface area (2630 m2/g) [9, 44], high theoretical specific capacitance (550 F/g) [63, 66], a theoretical electrical conductivity as high as 7200 S/m , mean pore diameter and total pore volume reaching to 4.38 nm and 0.44 cm3/g, respectively , and other interesting properties. These intriguing properties enable graphene to be considered for significant enhancement of carbon nanotube-coated RVC electrodes by acting as conductive bridges for the pores between the carbon nanotube (CNT) particles. Moreover, according to Zhang et al., graphene sheets decreased the aggregation of the CNT bundle and thereby increase the actual accessible area . In addition, the CNT with graphene will build 3D multilayer architecture pores inside the film which are responsible for providing many more available spaces to accommodate ions during electrosorption [30, 71].
CNTs and graphene combinations were used to prepare electrodes for CDI systems [30, 63, 72]. Significant property enhancement was observed in these materials where graphene sheets are intercalated with CNTs to improve the electrons’ transfer amongst the graphene sheets as the salt solutions are inherently resistive. In addition, the spacing between the graphene sheets is increased which improves solution flow and facilitates access of ions to these electrode materials. When these composite electrodes are compared with CNT or graphene electrodes alone, the electrical conductivity, accessible surface area, specific capacitance volumes, electrochemical stability, and electrosorption capacity are improved resulting in low energy consumption for the CDI system. All of them used the Hummer’s method to prepare GO from graphite and were exfoliated by low-power sonication or thermal treatment. These electrodes were made to be used as two-dimensional (2D) sheet electrodes with no significant electrode thickness (0.02 cm).
The CNT/RVC electrode, described in our previous publications [2, 4], was prepared successfully for using as an electrode. This is because it had increased stability at high flow rate pressure, increased the possibility of ions to reach the entire electrode surface in a short time for electrosorption, and shortened the time of ion desorption from the electrode surface. Moreover, we aimed to further increase the electrosorption capacity of CNT/RVC electrode. This can be achieved by firstly improving the pore structure and surface area. To perform that, the electrodes were made as novel three-dimensional (3D) electrodes using RVC electrode as a substrate. In addition, GO nanosheets were synthesized through a modified Hummer’s method and irradiated by microwave to reduce and exfoliate (GO) nanosheets. After that, the microwave-irradiated graphene oxide (mwGO) was dispersed with functionalized CNT in amide solvent (DMF) at various ratios and then used to coat RVC electrodes by a dip-coating method. The sonication time was optimized by monitoring with visible spectra. In addition, the energy of sonication of each milligram of the composite material was selected. The composite electrodes were characterized to check their surface chemistry, morphology, thermal stability, and electrochemical properties. The performance of these electrodes in the system was investigated under different loadings and scan rates.
Chemicals and Materials
Single-walled carbon nanotubes (SWCNTs) (Hipco-CCNI /Lot # p1001) were procured from Carbon Nanotechnologies, Inc. (Houston, TX), and graphite powder was supplied by Bay Carbon, Inc. N, N-dimethylformamide (DMF) (AR grade), concentrated nitric acid (HNO3 70%) (AR grade), potassium permanganate (KMNO4) (AR grade), ethanol (C2H6O) (AR grade), and NaCl (AR grade) were procured from Sigma-Aldrich. Moreover, the concentrated sulphuric acid (H2SO4, 98% w/v) (AR grade), hydrogen peroxide (30% aqueous H2O2) (AR grade), and concentrated (36% w/v) hydrochloric acid (HCl) (AR grade) were purchased from Univar. The RVC (compressed 60 ppi (pores per inch)) was procured from ERG Materials and Aerospace Engineering. All the above materials were used as received. Membrane filters (0.2 μm pore size GTTP) was supplied by MILLIPORE and cleaned. Milli-Q water (resistivity = 18.2 mΩ cm−1) has been used as per requirements.
Functionalization of CNTs
The functionalization of SWCNT was carried out by refluxing into a round-bottom flask containing 20 mg of it in 40 ml of nitric acid (6 M HNO3). The round-bottom was equipped with a magnetic stirrer and immersed in an oil bath at 120 °C for 6 h. The refluxed mixture was then neutralized by washing with water and subsequently with 10 ml of methanol and DMF, separately. Finally, it was dried in an oven at 105 °C for 48 h. The acid-treated SWCNT was designated as a-SWCNT.
Synthesis of GO
GO was synthesized by modified Hummer’s method, whose outline was given by Marcano et al. . In this typical method, graphite powder (1 g) was thoroughly mixed with conc H2SO4 (60 ml) for few minutes. Then, 3.5 g of KMnO4 was added in small aliquots so the temperature will not exceed 100 °C. The mixture was continuously stirred for 18 h and then hydrolysed by adding 300 to 500 ml distilled H2O (ice bath condition) to form graphite oxide. Thereafter, 30% aqueous H2O2 (~ 3 ml) was added drop-wise into the mixture till its complete colour changed. Finally, the vacuum-filtrated mixture was washed with HCl solution, distilled H2O, and C2H5OH. The slurried mass was dried under vacuum-oven overnight at 50 °C.
Exfoliation and Reduction of GO Using Microwave Irradiation
Exfoliated and reduced GO (mwGO) were formed using a conventional microwave oven (1200 W). After irradiation, the GO glowed red hot accompanied by fuming and sparking, leading to a remarkable volume expansion caused by the violent expulsion of the volatile materials present at the interlayer space of graphene-intercalated compounds [17, 58].
Dispersion of mwGO and a-SWCNT
A dispersion of 5 ml containing mwGO 0.1% w/v in DMF, a good solvent for GO dispersions because it can be coupled to an amphiphilic oligoester to produce amphiphilic graphene oxide that is dispersible [13, 17], was prepared using a homogenizer (Branson, Sonifier model S-450D), which was attached with a 13-mm step disruptor horn and a 3-mm tapered microtip. The sonicator was operated by setting the frequency at 20 kHz, amplitude at 25% (means power output 100 W), pulse on-off for 2 and 1 s, and placed on a water-ice bath to prevent overheating of dispersion. Furthermore, functionalized SWCNT 0.2% w/v was dispersed in DMF and optimized. The optimization was done by dispersing a-SWCNT 0.2% w/v in 15 ml DMF using the same sonicator and conditions.
Preparation of a-SWCNT/mwGO Composite Coating Solution
The preparation of a-SWCNT/mwGO composite material-coated RVC electrode involves the steps shown in Additional file 1: Figure S1. The dispersion, containing a-SWCNT and mwGO solution, was mixed via probe sonication for 30 min at 30% amplitude (1 s on/2 s off pulse) and bath sonication for 30 min  and was optimized for the individual components. The a-SWCNT/mwGO composites were prepared using the weight ratios 90% a-SWCNT to 10% mwGO, 80% a-SWCNT to 20% mwGO, and 70% a-SWCNT to 30% mwGO which were the weight ratios optimized in our laboratory, one of which (90% a-SWCNT to 10% mwGO) afforded the highest specific capacitance for the composite material in an electrode and that also gave the highest capacity for electrosorption. This has been discussed later on in the “Results and Discussion” section.
Pre-treatment of the RVC Electrode
The best pores per inch (ppi) RVC electrode, used as a substrate for loading carbon material, was optimized (see Additional file 1: Figure S2). It optimized one was 60 ppi RVC electrode because of its increased conductivity that readily allows the movement of electrical charge and had comparatively higher capacitance and surface area. Therefore, this optimized electrode was chosen as a substrate for loading a-SWCNT/mwGO composite material to be used as an electrode in a CDI system. Definite sizes of electrodes (length 4 cm × width 1.8 cm × thickness 0.3 cm) (2.16 cm3) were taken by cutting from a block of RVC and were soaked in 2 M HNO3 for 24 h to remove any impurity. The electrodes were washed several times with distilled water to remove the acid. The effluent’s pH was checked regularly until it became neutral. The removal of organic impurity from the electrode was carried out by soaking in CH3OH for 2 h. The electrode was then dried under N2 flow and followed by heating in an oven overnight at 110 °C. The dried electrodes were weighed properly through a suitable balance.
a-SWCNT, mwGO, and a-SWCNT/mwGO Composite Dip-Coated RVC Electrodes
All electrodes which are a-SWCNT, mwGO, or a-SWCNT/mwGO composites with RVC were made by the dip-coating method by slowly immersing into the composite solution. Additional file 1: Figure S3 schematically shows the dip-coating process of RVC in a composite solution of 9 CNT: 1 mwGO. The dip-coated substrates were initially placed in an oven for 2 h for drying at 100 °C and later on in a vacuum oven for 2 h at 50 °C for removing all organic solvents that remained in the micropore of the electrodes.
Details of carbon materials coated on RVC (2.16 cm3) electrodes
Ratio in coating solution
Material weight (mg)
In future discussions, these composite electrodes, depending on the ratio of a-SWCNT in the sample, will be identified as 9-CNT/mwGO/RVC (9 a-SWCNT: 1 mwGO-coated RVC electrode), 8-CNT/mwGO/RVC (8 a-SWCNT: 2 mwGO-coated RVC electrode), 7-CNT/mwGO/RVC (7 a-SWCNT: 3 mwGO-coated RVC electrode), CNT /RVC (10 a-SWCNT: 0 mwGO-coated RVC electrode), and mwGO/RVC (0 a-SWCNT: 10 mwGO-coated RVC electrode).
The capacitance of a-SWCNT, mwGO, and a-SWCNT/mwGO composite electrodes was measured using a cyclic voltammetry (CV). In this measurement, a-SWCNT/RVC, mwGO/RVC, or a-SWCNT/mw rGO/RVC electrode was taken as the working electrode (WE). The test was carried out in 1 M NaCl aqueous solution by scanning the voltage range − 0.2–1.0 V using a three-electrode system; RVC was taken as the counter electrode (CE) whereas Ag/AgCl (3 M NaCl) as the reference electrode (RE). Pt wire was used for making contact between WE and CE. The scanning rate for the measurement was taken as 5, 10, 20, 50, 100, and 200 mV/s. The electrochemical impedance spectra (EIS) for three electrodes namely mwGO/RVC, CNT/RVC, and 9-CNT/mwGO/RVC were measured over the frequency range 101–107 Hz. Galvanostatic charge-discharge (GCD) process was tested within voltage range 0–1 V.
Homogeneity and birefringence of the dispersion were checked by Olympus BH-2 microscope. The morphological test of the a-SWCNT-, mwGO-, and a-SWCNT/mwGO composite-coated RVC electrodes was done using a FESEM at voltage 1.0 kV. Furthermore, a-SWCNT and GO functionalization characterization was carried out by (1) visible absorption spectra, (2) Raman spectra (measured by a Raman spectrometer attached with a Raman microscope and CCD detector at the excitation wavelength of 632.81 nm where spectra were recorded over 30 s at 1.0 cm−1 resolution, and (3) XPS for determining the functional groups in carbon binding to carbon, and hydrogen and oxygen in GO and mwGO. The XPS binding energy spectra were obtained by passing energy 20 eV in the fixed analyser transmission mode. The XRD test was performed to determine the crystallinity and the interlayer distance between nanosheets. The XRD measurement was carried out using CuKα radiation (λ = 0.154) source that operated at 40 keV with a cathode current 20 mA. Movement, the thermal stability of the graphite, GO, and mwGO materials were determined by thermogravimetric analysis (TGA). The experiment was performed using a Q500 (TA Instruments) apparatus at a ramp rate of 5 °C/min in air, with a combined gas flow of 10 ml min−1 nitrogen (N2) and 90 ml min−1 air from 25 to 800 °C.
Results and Discussion
Chemical Conversion of Graphite to Graphene Oxide
The intercalated compound was hydrolyzed by adding water to form graphite oxide. The decomposition of permanganate ions into manganese (IV) ions was carried out by adding H2O2, and thereafter, the mixture was vacuum filtrated to remove manganese (IV) ions. The resultant was washed with HCl solution to ensure that no undesirable Mn(OH)2 was formed . The inset in Additional file 1: Figure S4 (b) shows an optical image of graphene oxide after washing and vacuum drying, at the bottom of a glass vial.
Surface Chemistry of GO
EDX spectra, FTIR spectra, and XRD patterns were used to confirm the complete oxidation of graphite flakes into GO using strong acid.
Energy-Dispersive X-Ray Spectroscopy
Additional file 1: Figure S4 (a and b) shows the FESEM images and corresponding EDX spectra of graphite flake powder and as-prepared GO, respectively. The smooth leaf structure of graphite flakes was completely changed into the uniform rough and porous structure containing valleys and elevated regions which reflect vast amounts of sheet stacking after the chemical oxidation process, which is clearly viewed from the SEM images of graphite and as-prepared GO given as insets in Additional file 1: Figure S4 (a) and (b), respectively. The crystal structures of graphite flake powder and as-prepared GO are identified using XRD patterns (Additional file 1: Figure S4 (c)). A strong and sharp peak was observed for graphite flakes at 26.6° corresponding to a basal spacing d002 = 0.343 nm. This indicated a highly ordered structure of graphite. The XRD pattern of GO, on the contrary, exhibited at low-angle region (10.8°, 001 reflection) corresponded to a basal spacing of d001 = 0.820 nm, whereas the broader peak at 2θ = 20.7° corresponded to the (d002 = 0.435 nm) crystal plane of graphite and amorphous carbon [6, 12]. The observed results are in good agreement with the other GO precursors [6, 52]. This is because, during oxidation, the oxygen-containing functional groups are introduced into the carbon lattice . Hence, the results indicated that the interlayer spacing of GO (d002 = 0.435 nm) was increased because of the oxidation of graphite flakes (d spacing = 0.343 nm). In addition to SEM images and XRD, the corresponding EDX spectra in Additional file 1: Figure S4 (b, for GO) reveal the formation of a new structure with a higher oxygen content of 30.1 wt% O (the oxygen content of graphite is too low to be detected (Additional file 1: Figure S4 (a)) and lower C/O ratio of 2.3; indicating the oxidation of graphite flakes and formation of GO. The higher oxygen content of GO compared to the graphite is because of the formation of epoxide, carboxylic acid and –OH groups on the edges and the basal planes of the carbon structure, respectively . Besides C and O atoms in Additional file 1: Figure S4 (b), no peaks existed in the EDX spectra of GO; which indicates that no impurities, presence of sulphur and manganese, have been left as residuals of H2SO4 and KMnO4 used in the preparation procedure.
Fourier Transform Infrared Spectroscopy
The presence of oxygen-containing functional groups was evident from the Fourier transform infrared (FTIR) spectrum as shown in Additional file 1: Figure S4 (d). This confirmed that the graphite flakes were successfully oxidized to the graphene oxide. The broad and wide peak observed at 3386 cm−1 correspond to the –OH stretching vibrations of the hydroxyl (C–OH) groups and water, and the epoxy (C–O–C) group peak at 1228 cm−1 [33, 41, 74]. These groups usually are present throughout the basal planes of GO . In addition, the sharp intense peak at 1587 cm−1 is ascribed to (C=C) benzene rings, and the absorption bands at 1720 and 1047 cm−1 can be ascribed to the stretching vibration of carboxyl (C=O) and alkoxy (C–O) groups [37, 41, 54, 69], respectively, which usually were distributed at the edges of these 2D nanostructures .
Exfoliation and Reduction of GO Through Microwave Irradiation
Microwave irradiation technique was used for preparing exfoliated graphite (EG) [17, 58] and GO . In addition, microwave-assisted reduction of GO can be achieved in less than 1 min by the treatment of GO precursor within a microwave oven [6, 78]. Therefore, the microwave irradiation method was used to exfoliate and reduce graphene oxide. The GO powders were irradiated within a microwave oven for 10 s that largely expanded its volume and fumed violently. The SEM image of GO and mwGO films are shown in Additional file 1: Figure S5 (a) and 2 (b), respectively. The image in Additional file 1: Figure S5 (b) shows that the graphene sheets expanded, developed an accordion-type structure of highly 3D macropores folded with each other, and formed an interconnected network with minimal re-stacking compared with GO before microwave irradiation in Additional file 1: Figure S5 (a) [6, 58] which were transparent to the electron beam and ion salts. In addition, the SWCNT can be sandwiched into the layers more easily . Furthermore, the XRD spectra, Additional file 1: Figure S5 (c), shows one distinct broad peak for mwGO at 2θ = 20.7° corresponding to the (d002 = 0.435 nm) crystal plane of graphite, and the sharp peak at 2θ = 10.8° is heavily suppressed because of the rapid expansion of GO layer. This result is in good agreement with other mwGO precursors  and corresponding to the removal of some of the functional groups of GO and successful exfoliation of graphene [18, 64, 72]. Besides XRD and SEM images, the corresponding EDX spectra, presented in Additional file 1: Figure S5 (d) and S5 (e), revealed the formation of a new structure with the lower oxygen content of 9.9 wt% O and higher C/O ratio of 9.1. This indicated that reduced graphene oxide (rGO) was formed after irradiation. The lower oxygen content of rGO compared to GO was because of releasing epoxide, carboxylic acid, and –OH groups on the edges and the basal planes of the carbon structure, respectively. It can be viewed from the inset in Additional file 1: Figure S5 (e) that there is a dramatic expansion of the GO powder (placed inside the glass vial (inset Additional file 1: Figure S5 (d)) that yielded a black and fluffy mwGO powder. Zhu et al. demonstrated that further treating mwGO powder with microwave irradiation-induced sparking and even burnt the sample .
Surface Chemistry of GO After Microwave Irradiation
The chemical changes that were observed in GO because of microwave irradiation were investigated by Raman spectra, TGA, XPS, and XRD spectra to confirm its reduction and exfoliation.
Raman spectroscopy is a powerful diagnostic tool used for the structural and qualitative characterization of carbons [14, 15], particularly for investigating the presence of any defect and ordered or disordered structure of graphite and graphene. Many scientific papers are mentioning that there are three typical peaks always attributed to D, G, and 2D bands of graphite, GO, and rGO that are observed at about 1350, 1580, and 2680 cm−1, respectively [42, 55]. The 2D band is always a strong band in graphene even when there is no D band present, and it does not represent defects. Additional file 1: Figure S6 exhibits the Raman spectral curves of natural graphite flakes, as-prepared GO, and GO after microwave irradiation. It is clearly viewed from the Raman spectra that graphite exhibits two peaks at G and 2D bands. The chemical oxidation on graphite to produce graphene oxide (GO) resulted in the appearance of a D band at a lower wave number compared to the G band coupled with the disappearance of the 2D band peak. The D band peak of GO is associated with the breathing modes of rings or κ-point phonons of A1g symmetry of graphene oxide occurring at 1321 cm−1, while the G band peak, associated with first-order scattering of the stretching vibration mode E2g observed for carbon sp2 atoms domains, occurs at 1587 cm−1 . The integrated intensity ratio (ID/IG) of these two Raman bands has been demonstrated to depend upon the physical state of the graphitic carbon and can reveal the ordered and disordered crystal structures of graphene. As expected, after microwave irradiation, both D and G band peaks were slightly shifted to 1330 cm−1 and 1590 cm−1, respectively, and the ratio of the ID/IG band decreased from 1.26 to 1.08. This decrease suggests a slight reduction of the GO [29, 65]. Moreover, this microwave-assisted reduction of GO has a good advantage over chemical reduction by ultrasonic dispersion or rapid thermal expansion followed by chemical reduction because chemical reduction often involves highly toxic chemicals, require long reduction time, or require high-temperature treatment [20, 53].
XPS Spectra Analysis
XPS was employed for investigating the surface chemical states of GO and microwave-irradiated graphene oxide (mwGO) powder by observing any change in the binding energy of C1s peaks. Additional file 1: Figure S7 depicts the XPS spectra of GO and mwGO within the binding energies scanned range from 280 to 294 eV, which presents information on the main chemical components in the samples from the detected peaks (C1s spectra) of both samples. It can be seen that in Additional file 1: Figure S7 (a), the C1s XPS spectrum of GO clearly indicates a considerable degree of oxidation with three components corresponding to carbon atoms in different functional groups : the non-oxygenated C at 284.4 eV corresponds to C1s of sp2 (C=C), the carbonyl carbon (C=O) at 287.2 eV, and the carboxylate carbon (O–C=O) at 288.4 eV. Based on the XPS analysis, the as-prepared GO had an atomic ratio for carbon/oxygen of 2.91. After microwave irradiation of GO (Additional file 1: Figure S7 (b)), the number of C=C band increased from 34.21 to 56.47%, revealing that the graphitic structure of rGO is remarkably restored, while the C/O atomic ratio of mwGO was found to be 5.32. The XPS results show significant removal of the oxygen-containing groups because of exposure to the microwave radiation as reported for graphite oxide and graphene oxide [6, 33, 78]. It can also be mentioned that the slight shifting of the de-convoluted peaks for mwGO is because of the enhanced regular structure compared to the GO .
Thermogravimetric analysis (TGA) was used for studying the thermal stability of natural graphite flakes, GO, and microwave-irradiated graphene oxide (Additional file 1: Figure S8). The TGA curves represented the mass loss during the heating in air. It is clear that pristine graphite is thermally very stable, only 4.8 wt% mass loss occurred upon heating to 600 °C as in agreement with the Song et al.’s report . For graphene oxide (GO), three stages of weight loss are apparent, as shown in Additional file 1: Figure S8. The first one is below 120 °C, corresponding to the loss of intercalated water that was not removed from drying. The second one, between 120 and 300 °C, corresponding to the removal of oxygen-containing functional groups, such as C–O, C=O, and O–C=O [27, 40, 64]. This major weight loss is related to degradation and decomposition, whereas the third weight loss that occurs at approximately 530 °C is because of the bulk pyrolysis of the carbon skeleton [62, 64]. This pyrolysis is also observed in mwGO at 565 °C and in graphite at 680 °C. The TGA analysis indicates the incomplete reduction of the functional groups by microwave because roughly 15 wt% of oxygen-containing groups did not undergo reduction.
Dispersion of mwGO in DMF
Recently, sonication has been extensively used as an exfoliation and dispersion strategy for producing graphene nanosheets in a liquid medium . The first parameter to be investigated was the impact of sonication time for dispersing mwGO in DMF. The prepared dispersion contained mwGO 0.1% w/v in 15 ml DMF; this is one of the best solvents reported for GO dispersions because it gave the highest spacing between GO sheets of up to 1.05 nm and the smallest GO sheet thickness, down to 0.83 nm . It is necessary to determine the optimum sonication time for dispersing mwGO in DMF because excessive sonication may result in the shortening or creating a defect in the sheets. The dispersion and stabilization of mwGO were performed using UV-VIS spectroscopy. The visible spectra (between 300 and 1100 nm) of mwGO dispersion with respect to sonication time are presented in Additional file 1: Figure S9 (a). All mwGO solutions were diluted five times by DMF before measuring their visible spectra, using 1-cm path length quartz cuvettes and considering the baseline for DMF. The absorbance of the supernatant at 660 nm was considered for plotting with respect to sonication time and represented in Additional file 1: Figure S9 (b). Peak intensities were read directly from the as-measured absorption spectrum. It is observed that, with the increase in sonication time, the mwGO absorption bands became more pronounced, indicating that the good dispersion of mwGO happened with respect to time. Moreover, there is very little change in the absorption spectra of the dispersion after further sonicating beyond 35 min is observed. This implied that the solvent was saturated and the minimum time for effective dispersion of the mwGO was 35 min. The photographs of mwGO dispersions before starting the sonication and after 35 min of sonication are shown in the inset photographs in Additional file 1: Figure S9 (a). It is observed that with the increase in sonication time, the colour of the dispersion is becoming gradually black, indicating that the mwGO was homogenously dispersed within DMF. It is well known that the complete dispersion is affected by sonication energy. The visible absorbance of mwGO dispersion with respect to sonication energy is presented in Additional file 1: Figure S9 (c), which shows the sonication energy required to disperse 15 mg of mwGO in 15 ml DMF. It is obvious that the energy increased with sonication time. The energy was 12 kJ at 2 min and was 90 kJ at 15 min. The effective dispersion of mwGO was achieved at the minimum energy of 210 kJ at 30 min because the dispersion’s absorbance remained constant beyond the sonication energy of 210 kJ. This energy, as represented in Additional file 1: Figure S9 (c), was for 35 mg in DMF (0.1% w/v). This indicates that 14 kJ energy was required per milligram of mwGO in a 0.1% w/v solution.
Electrodes Surface Morphology
The top view of 8-CNT/mwGO composite materials (Fig. 1g) obviously shows that nanotube bundles are still dominant and highly tangled, wrapped inside the composite. The strands of CNTs adhere to form bigger aggregation bundles compared with 9-CNT/mwGO composite materials. The cross-section of this composite electrode shown in Fig. 1h indicates a layered structure and the CNTs are uniformly sandwiched between the GO sheets making a multilayered 3D network structure with good porous architecture inside it. Obviously, the graphene nanosheets restacking and gaps interfacing between graphene sheets and CNTs to form a favourable open pore system was decreased. This results in difficult electrolyte ion access to the surface and minimizes surface area compared to 9-CNT/mwGO composite material. This phenomenon could possibly because of the minimization of surface area and specific capacitance. It is evident that, in Fig. 1j, when the mwGO ratio reaches to 3 in the composite, the graphene nanosheets dominate in the composite and the cross-section surface was completely changed into the 3D uniform rough multilayer and porous structure containing voids with CNTs distributed in them but is difficult to clearly image. This structure results in the reduction of electrolyte accessibility to the surface area. In contrast, Fig. 1k exhibits the top surface of mwGO/RVC electrode. The top surface of mwGO has rippled and crumpled structures and wrinkled structures with clear aggregation, and the stacked structure of the nanosheets in the cross-section image of mwGO (Fig. 1l) can be clearly viewed. The cross-section SEM reveals a curled morphology consisting of rippled structures. Consequently, graphene sheets with curled or wrinkled morphologies will present the greatest difficulty for accessing electrolyte to the surface area. Therefore, the specific capacitance in mwGO/RVC electrode is expected to be the smallest, comparing amongst all composite electrodes.
Electrochemical Behaviour of Electrodes Studied Using CV
The electrochemical properties of all the coated RVC electrodes were measured using CV. In this article, CV was used to measure the variation of capacitance of the electrodes with respect to loading level and scan rate and to check the electrode cyclic stability using 1 M NaCl solution within the voltage range between − 0.2 and 1.0 V vs Ag/AgCl in a three-electrode system. A RVC was taken as the counter electrode whereas mwGO/RVC, CNT/RVC, 7-CNT/mwGO/RVC, 8-CNT/mwGO/RVC, and 9-CNT/mwGO/RVC as the working electrodes.
Capacitive Behaviours of mwGO/RVC, CNT/RVC, and Various CNT/mwGO/RVC Composite Electrodes
We reported the CV curve of the CNT/RVC electrode, where the value of specific capacitance measured at 5 mV/s scan rate was 139.65 F/g . This capacitance is increasing when the mwGO material is combined with CNT material at the approximate ratio of 2:10, as presented in Fig. 2. The inset in Fig. 2 exhibits that the 9-CNT/mwGO/RVC electrode possesses the highest specific capacitance value (179.39 F/g), which is 29% higher compared to the CNT/RVC electrode. This enhancement is due to the synergistic effect of CNT and mwGO. The incorporation of mwGO into CNT forms an advantageous network structure that conductively bridges the pores between the CNT particles by facilitating the rapid electrolytic ions’ transport inside the electrode materials, and thereby, the electrode’s surface area is increased. This surface area increased due to the large external active surface area of mwGO sheets and the diameter of bundling of CNTs in CNT/mwGO/RVC composite electrode being smaller than 25 nm. This is smaller compared to the diameter of bundling of CNTs alone, reported elsewhere . The specific capacitance value, measured at 5 mV/s scan rate for 9-CNT/mwGO electrode, was 167.56 F/g, less compared to 9-CNT/mwGO/RVC composite electrode (179.39 F/g). This high value of specific capacitance for 9-CNT/mwGO/RVC electrode may be due to amazing property of RVC, which reduce the resistance of solution flow through the electrode thereby increasing the possibility of ions to reach all electrode surfaces.
The inset in Fig. 2 exhibits a decreasing trend of the specific capacitance value with further increase in mwGO content in the CNT/mwGO/RVC composite electrode. For example, the specific capacitance values of 8-CNT/mwGO/RVC and 7-CNT/mwGO/RVC electrodes are found to be 156.47 and 132.60 F/g, respectively. This decrement in specific capacitance value is because of the low specific surface area of mwGO. Furthermore, this decrement in specific capacitance value can be ascribed to the difficulty of electrolytic ions’ transport into the porous CNT structure because the excessive mwGO mostly covered its surface. Interestingly, the CNT/RVC electrode shows a higher value of specific capacitance compared to 7-CNT/mwGO/RVC electrodes. From the above-going discussion of specific capacitance, it can be suggested that the electrosorption capacity performance of 9-CNT/mwGO/RVC electrode will be better compared to all other electrodes. Moreover, the specific capacitance result of mwGO/RVC electrode, being lower compared to other electrodes, suggests not to applying this electrode in the CDI system selected for experiments because the electrosorption capacity will be very small compared with CNT/RVC and CNT/mwGO/RVC electrodes.
The Effect of Increasing Loading Level on Capacitive Behaviours
In the previous section, it is reported that the 9-CNT/mwGO/RVC composite electrode has shown the highest specific capacitance value. Hence, this electrode has been selected for studying the effect of increasing loading level of composite materials in geometric area terms and geometric volume terms on its capacitive behaviours. The capacitive behaviours of all 9-CNT/mwGO/RVC electrodes, which had similar geometric volume (2.16 cm3) of RVC electrodes with various amounts of composite materials coated, 10, 30, and 50 mg, were studied by CV. The maximum loading level was set at 50 mg, because beyond this level, the pores of the RVC electrode was completely filled, which is evident from the top view of SEM image (Fig. 1a).
Effect of Different Scan Rates on the Electrode Capacitance
Specific capacitance value of different electrodes measured with respect to mass (F/g), area (F/cm2), and volume (F/cm3)
Scan rate (mV/s)
It is important to note that the capacitance of the RVC electrode per geometric volume under the same conditions at a scan rate of 20 mV/s was discussed in the literature . The RVC was 0.002 F/cm3, and the capacitance had increased by a factor of 1060 when the same electrode was filled completely by functionalized CNT material. It can be observed in Table 2 that the capacitance of the RVC electrode had increased by a factor of 1775 when filled completely with 9-CNT/mwGO material. This is related to the large surface area of 9-CNT/mwGO/RVC compared to a RVC electrode according to the Randles-Sevcik relationship .
Electrostatic Impedance Spectra
In Fig. 5b, the real part of impedance (Z′) has been plotted against the angular frequency (ω) for mwGO/RVC, CNT/RVC, and 9-CNT/mwGO/RVC electrodes. This type of plotting is also known as Bode plot. It is observed that the impedance value of mwGO/RVC electrode is high over the whole frequency region compared to both CNT/RVC and 9-CNT/mwGO/RVC electrodes. This low value of impedance for CNT/RVC and 9-CNT/mwGO/RVC electrodes is due to the formation of preferential interconnected conductive networks on their surfaces that facilitate both the ionic and electronic transports. It has been shown diagrammatically in the inset of Fig. 5b how different parameters are correlated with the different regions of curve. Zw, RCT, CDL, and Re are identified over the low-frequency region inclined approximately 45°, over the middle frequency region inclined nearly 0° means horizontally, over the high-frequency region inclined nearly 90°, and finally over the extreme high-frequency region horizontally at 0°, respectively. Hence, it can be inferred that the result and diagram of both Nyquist and Bode plots are in good agreement with each other.
Galvanostatic Charge-Discharge Process
Capacitance with Respect to Number of Cycle
Graphene oxide nanosheets were successfully synthesized through a modified Hummer’s method and were reduced and exfoliated using microwave irradiation. The SEM images show significantly altered structure that is highly porous, forming an interconnected network with minimal restacking of the graphene. The XRD confirms the exfoliation, by microwave irradiation, of the GO due to the significant suppression of the sharp peak at 2θ = 10.8° and the increase in distance between graphene nanosheet layers in the individual stacks, while XPS spectra confirm that reduction of the GO is apparent due to the much sharper sp2 (C=C) at 284.4 eV peak. The D/G band ratio of the Raman spectra decreased from 1.26 to 1.08 signifying that the GO has undergone removal of some of its functional groups. In the novel CNT/mwGO/RVC electrodes, the CNT materials were found to be sandwiched amongst the graphene sheets that built a 3D highly porous architecture to facilitate rapid ion diffusion. These lead to enhanced van der Waals forces and friction. Moreover, the best electrochemical response with respect to capacitance and kinetic behaviour amongst all target electrodes concluded that the composite 9 a-SWCNT: 1 mwGO-coated RVC electrode (9-CNT/mwGO/RVC) performed as the best electrode. It showed specific capacitance value 179.39 F/g that means 29% increment compared to the CNT/RVC electrode. This optimal electrode also had a very high cyclic voltammogram (CV) curve stability that maintained an electrochemical cycling stability of 97% after 2000 cycles. The results indicate that the electrodes can be efficiently used as the electrode materials in capacitive deionization cell for purifying water.
The authors extend their appreciation to the Deanship of Scientific Research at King Saud University for funding this work through Research Group No. RG-1436-005.
AA, MR, and MA have designed and performed the experiments and analysed the data. All authors have contributed in writing the manuscript. All authors have read and approved the final manuscript.
The authors declare that they have no competing interests.
- 1.Abdulhakeem B, Farshad B, Damilola M, Fatemeh T, Mopeli F, Julien D, Ncholu M (2014) Morphological characterization and impedance spectroscopy study of porous 3D carbons based on graphene foam-PVA/phenol-formaldehyde resin composite as an electrode material for supercapacitors. RSC Adv 4:39066–39072CrossRefGoogle Scholar
- 6.Antiohos D, Pingmuang K, Romano MS, Beirne S, Romeo T, Aitchison P, Minett A, Wallace G, Phanichphant S, Chen J (2013) Manganosite–microwave exfoliated graphene oxide composites for asymmetric supercapacitor device applications. Electrochimica Acta 101:99–108. https://doi.org/10.1016/j.electacta.2012.10.007 CrossRefGoogle Scholar
- 7.Bard A (2001) In: Bard AJ, Faulkner LR (eds) Electrochemical methods: fundamentals and applications, 2nd edn. John Wiley, New YorkGoogle Scholar
- 14.DasA PS, Chakraborty B, PiscanecS SSK, Waghmare UV, Novoselov KS, Krishnamurthy HR, Geim AK, Ferrari AC, Sood AK (2008) Monitoring dopants by Raman scattering in an electrochemically top-gated graphene transistor. Nat Nano 3:210–215 http://www.nature.com/nnano/journal/v3/n4/suppinfo/nnano.2008.67_S1.html CrossRefGoogle Scholar
- 21.Ghosh S, Calizo I, Teweldebrhan D, Pokatilov EP, Nika DL, Balandin AA, Bao W, Miao F, Lau CN (2008) Extremely high thermal conductivity of graphene: prospects for thermal management applications in nanoelectronic circuits. Appl Phys Lett 92:151911Google Scholar
- 24.Hernandez Y, Nicolosi V, Lotya M, Blighe FM, Sun Z, De S, McGovern IT, Holland B, Byrne M, Gun’ko YK, Boland JJ, Niraj P, Duesberg G, Krishnamurthy S, Goodhue R, Hutchison J, Scardaci V, Ferrari AC, Coleman JN (2008) High-yield production of graphene by liquid-phase exfoliation of graphite. Nat Nanotechnol 3:563–568CrossRefGoogle Scholar
- 25.Hirata (2003) Thin-film-like particles having skeleton constructed by carbons and isolated films, US patent 6596396Google Scholar
- 26.Jalili R, Aboutalebi SH, Esrafilzadeh D, Konstantinov K, Moulton SE, Razal JM, Wallace GG (2013) Organic solvent-based graphene oxide liquid crystals: a facile route toward the next generation of self-assembled layer-by-layer multifunctional 3D architectures. ACS Nano 7:3981–3990. https://doi.org/10.1021/nn305906z CrossRefGoogle Scholar
- 39.Lu T, Pan L, Li H, Nie C, Zhu M, Sun Z (2011) Reduced graphene oxide–carbon nanotubes composite films by electrophoretic deposition method for supercapacitors. Journal of Electroanalytical Chemistry 661: 270-273 doi: https://doi.org/10.1016/j.jelechem.2011.07.042
- 42.Metin Ö, Kayhan E, Özkar S, Schneider JJ (2012) Palladium nanoparticles supported on chemically derived graphene: an efficient and reusable catalyst for the dehydrogenation of ammonia borane. International Journal of Hydrogen Energy 37:8161–8169. https://doi.org/10.1016/j.ijhydene.2012.02.128 CrossRefGoogle Scholar
- 43.Mitali S, Soma D, Monica D (2011) A study of effect of electrolytes on the capacitive properties of mustard soot containing multiwalled carbon nanotubes. Res J Chem Sci 1:109–113Google Scholar
- 45.Naoi K, Simon P (2008) New materials and new configurations for advanced electrochemical capacitors. J Electrochem Soc 17:34–37Google Scholar
- 50.Romano MS, Li N, Antiohos D, Razal JM, Nattestad A, Beirne S, Fang S, Chen Y, Jalili R, Wallace GG, Baughman R, Chen J (2013) Carbon nanotube – reduced graphene oxide composites for thermal energy harvesting applications. Adv Mater 25:6602–6606. https://doi.org/10.1002/adma.201303295 CrossRefGoogle Scholar
- 51.Semiat R (2000) Desalination: present and future. Intern Water Resour Assoc 25:54–65Google Scholar
- 53.Shin H-J, Kim KK, Benayad A, Yoon S-M, Park HK, Jung I-S, Jin MH, Jeong H-K, Kim JM, Choi J-Y, Lee YH (2009) Efficient reduction of graphite oxide by sodium borohydride and its effect on electrical conductance. Adv Funct Mater 19:1987–1992. https://doi.org/10.1002/adfm.200900167 CrossRefGoogle Scholar
- 60.Wang Y-S, Yang S-Y, Li S-M, Tien H-W, Hsiao S-T, Liao W-H, Liu C-H, Chang K-H, Ma C-CM HC-C (2013) Three-dimensionally porous graphene–carbon nanotube composite-supported PtRu catalysts with an ultrahigh electrocatalytic activity for methanol oxidation. Electrochimica Acta 87:261–269. https://doi.org/10.1016/j.electacta.2012.09.013 CrossRefGoogle Scholar
- 62.Wilson NR, Pandey PA, Beanland R, Young RJ, Kinloch IA, Gong L, Liu Z, Suenaga K, Rourke JP, York SJ, Sloan J (2009) Graphene oxide: structural analysis and application as a highly transparent support for electron microscopy. ACS Nano 3:2547–2556. https://doi.org/10.1021/nn900694t CrossRefGoogle Scholar
- 64.Wojtoniszak M, Chen X, Kalenczuk RJ, Wajda A, Łapczuk J, Kurzewski M, Drozdzik M, Chu PK, Borowiak-Palen E (2012) Synthesis, dispersion, and cytocompatibility of graphene oxide and reduced graphene oxide. Colloids Surfaces B: Biointerfaces 89:79–85. https://doi.org/10.1016/j.colsurfb.2011.08.026 CrossRefGoogle Scholar
- 66.Xia J, Chen F, Li J, Tao N (2009) Measurement of the quantum capacitance of graphene. Nat Nano 4:505–509 http://www.nature.com/nnano/journal/v4/n8/suppinfo/nnano.2009.177_S1.html CrossRefGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.