CHAPTER 5. EXPERIMENTAL STUDIES ON COCONUT SHELL BASED EDLCs

Transcription

1 CHAPTER 5 EXPERIMENTAL STUDIES ON COCONUT SHELL BASED EDLCs 5.1 Introduction Electrochemical double layer capacitors (EDLCs), also known as supercapacitors or ultracapacitors are energy-storage devices which deliver 100 times the power of batteries and are able to store 10,000 times more energy as compared to conventional capacitors [Morimoto et al. (1996), Kötz and Carlen (2000), Conway (1991), Béguin and Frąckowiak (2013)]. Supercapacitor stores energy by forming a double layer of electrolyte ions on the surface of conductive electrodes. It is not limited by the electrochemical charge transfer kinetics of batteries and hence it can operate at very high charge and discharge rates and have a lifetime of over a millions of cycles [Hantel et al. (2011), Miller and Simon (2008)]. At present, the energy stored in supercapacitors is lower than that of batteries, which restricts their adoption to the applications that require high life cycle and power density. Therefore ample of efforts have been made to explore electrode materials for supercapacitor applications [Yang et al. (2008), He et al. (2013), Hu et al. (2006)]. Among the diverse options, activated carbons are commercially used as an electrode material for the EDLC application because of their high surface area, favorable pore size distribution and good electrical conductivity. Activated carbon is an attractive option because of its abundance, cost effective and environment friendly in nature [Frackowiak and Beguin (2002)]. The high specific surface area and porosity is the key advantage of activated carbon to form an effective double layer, which is the characteristic of supercapacitors possessing high power density. It may be noted that the activated carbon from biomass is a natural and renewable one and its synthesis process is completely eco-friendly. Coconut shell can be used for fuel and it is a source of charcoal also. It may be noted that activated carbon synthesized from coconut shell is considered better in comparison to those 102

2 obtained from other sources because of its mesoporous structure which makes it suitable for its application in EDLCs as electrode material [Zhou et al. (2012), Xiong et al. (2011), Kuratani et al. (2011)]. Further, after carbonization of coconut shell, the three dimensional coconut shell based carbon has a hierarchical porous structure which can provide a large surface area, high conductivity, well connected and highly ordered microstructure [Jiang et al. (2010)], which helps in ion transport by providing low resistance pathways, hence expecting improved performance as an EDLC electrode. In the present chapter, the synthesis technique of chemically treated and activated coconut shell based activated charcoal (CST) and its application as an electrode material in EDLCs has been reported. The synthesized activated charcoal was characterized using N 2 -adsorptiondesorption isotherms, scanning electron microscopy (SEM), X-ray diffraction (XRD) and thermogravimetric analysis (TGA). EDLC cells have also been fabricated using optimized composition of blend polymer electrolytes comprising of polyvinylidene fluoride-co-hexa-fluoro propylene [PVdF(HFP)]- poly methyl methacrylate [PMMA]- sodium thiocynate [NaSCN]. Galvanostatic charge-discharge, a.c impedance spectroscopy and cyclic voltammetry (CV) was used to characterize the EDLC cell. 5.2 Experimental Details Preparation of Blend Polymer Electrolytes PVdF(HFP) with an average molecular weight of 1.3x10 5 (Sigma Aldrich), PMMA with an average molecular weight of 1.2x10 5 (Sigma Aldrich) and inorganic salt NaSCN (Loba chemie) were used in the present study. All the chemicals were dried before use. A solution-cast method has been used to synthesize polymer blend electrolyte films. Different composition of NaSCN, PVdF(HFP), PMMA were dissolved in DMF (Merck) and mixed together. The resultant solution was stirred and heated continuously at 60 0 C for 10 hours until the mixture gets homogenous and becomes gelly in nature. Finally the mixture was poured in glass petri dishes to get a mechanically stable film having thickness of μm. Thereafter, film was optimized in terms of its electrical conductivity. The optimized composition of polymer blend electrolyte used in the present studies was found to be [PVdF(HFP)(80wt%) - PMMA(20wt%)](80wt%) 103

3 [NaSCN (1.0 M)] (20 wt%) having electrical conductivity of the order of ~ 10-2 S cm -1. The details of the characterization and optimization of electrolyte material is mentioned elsewhere [Gupta A. (2013)] Preparation of Coconut Shell Based Activated Charcoal (CST) Coconut shell based chemically treated activated charcoal (CST) has been synthesized by using chemical activation techniques. The detail of the synthesis technique is given in section KOH as an activating agent is one of the most effective agents employed for organic materials. KOH is mostly preferred over other activating agents because it causes more localized reactions with the carbon precursor and is more effective for highly ordered materials [Lillo- Rodenas et al. (2003)]. The chemical reaction between KOH and carbon during activation proceeds as: 6 KOH + 2 C 2 K + 3 H K 2 CO 3 The reaction is then further continued by decomposition of K 2 CO 3 and/or reaction of K/ K 2 CO 3 /CO 2 with carbon Construction of Electrodes used in the Experiments The electrodes were prepared by making slurry of prepared coconut based activated charcoal powder and PVdF-HFP in the ratio 90:10 (w/w) in a common solvent acetone by thorough mixing. Fine films of electrodes were coated by spraying the slurry on carbon cloth (Ballard, USA) and kept in oven at 70 C for h Characterization of Electrodes SEM micrograph was taken with the help of JEOL Model-JSM 6380LA. Nitrogen adsorption-desorption isotherms were measured at 77 K using Autosorb-1 Quantachrome Instruments, USA. X-ray diffraction (XRD) measurements were carried out using X Pert PRO, Panalytical, Netherland using an operational voltage of 40 kv and current of 40 ma respectively. In 104

4 order to have better understanding in the carbonization process, a thermogravimetric analyzer (Diamond TGA/ DTA, Perkin Elmer Instruments, USA) was employed to monitor the volatile evolution behavior of CST in N 2 atmosphere Electrochemical measurements The EDLCs were fabricated using CST electrodes with polymer blend electrolyte PVdF(HFP)-PMMA-NaSCN which was sandwiched between two symmetrical electrodes. The performance characteristics of the EDLCs were characterized by using a.c impedance analysis, galvanostatic charge-discharge technique, cyclic voltammetry and prolonged cyclic tests. The a.c impedance measurements for the characterization of EDLCs were carried out by using computer controlled LCR HI TESTER (Model , Hioki, Japan) in the frequency range from 10 mhz to 100 khz. The overall capacitance C of the capacitor cells were evaluated by using the relation C = (5.1) where, ω is the angular frequency and Z is the imaginary part of the total complex impedance. The single electrode specific capacitance of the capacitor cells were evaluated by multiplying the overall capacitance by a factor of two and dividing by the mass of the single electrode material. The cyclic voltammetry was carried out by using computer controlled CHI 608C, CH Instruments, USA. The capacitance values from this technique were evaluated by using the relation C = (5.2) where, I represents the current and s is the scan rate. The charge-discharge characteristics of the capacitor cells were performed at constant current and its discharge capacitance C d was calculated from the linear part of the discharge curves using the relation 105

5 C d = (5.3) where, i is the constant current and t is the time interval for the voltage change V. The specific capacitance (C s ) of the single electrode can be given as: C s = (5.4) where, m is the mass of active electrode material of single electrode, the factor of 2 comes from the fact that the total capacitance (C) measured from the EDLC cell is the addition of two equivalent single electrode capacitor connected in series. The impedance data as a function of frequency and complex power analysis of the EDLC cells were also made by using mathematical equations ( ). 5.3 Results and Discussion Activated carbons are those carbons which have high specific surface area and porosity. These materials are synthesized by either thermal or chemical activation, where small hexagonal carbon rings are formed and are called as graphene sheets. Methods of preparation of activated carbon mainly decides the size, orientation and stacking of graphene sheets. In general, there exists a very little order between the graphene sheets without having any three dimensional order of long range. Dahn et al. has given the model of falling cards to explain the complex structure of activated carbon [Liu et al. (1996)]. During the activation process of activated carbon, the thermal energy is enough to break the links between adjacent graphene sheets and allows some of them to rotate in parallel orientation. Thus, at that moment, an activated carbon can be treated as a combination of many small domains which consist of some graphene sheets aligned in parallel. In order to understand the structural surface and thermal details of coconut shell based activated carbon in the present case, different studies (SEM, XRD, TGA and BET) has been carried out. 106

6 5.3.1 Scanning Electron Microscopy (SEM) Fig. 5.1 shows the SEM image of coconut shell based chemically treated activated charcoal (CST) which clearly shows its porous structure and complete opening of cell pores on the surface. As can be seen from the micrograph, there is a formation of pores which may be due to release of some volatiles and activating agent from the sample. The external surface of CST was full of cavities which are due to the removal of KOH thereby leaving the space which was previously occupied by the activating agent. In this way it provides a larger surface area which helps in the rapid access of the electrolyte into the bulk phase of the electrode material, which further enhances the utilization ratio of active substances. Figure 5.1: SEM micrograph of CST electrode. Micrograph also reveals that CST has a high specific surface area and a shorter diffusion path, which provides a structural foundation for a high specific capacitance [Pell and Conway (2001)]. 107

7 5.3.2 X-Ray Diffraction (XRD) Studies Fig. 5.2 displays the XRD pattern of the activated carbon prepared from coconut shell which was activated at 600 C using KOH as an activating agent. The activated carbon shows two peaks at Figure 5.2: XRD pattern of CST around 2θ = 26 and 44 which corresponds to the peak of graphite [Lua and Yang (2004)]. Further it can be seen that XRD signals of powder sample contains large amount of noises in it. This behavior confirms the amorphous structure of carbon. The present result shows that the pyrolytic reaction of organic compounds comprises of; the breaking of chemical bonds with temperature and re-polymerization of radicals which further reduces to active compounds. These compounds form typical graphitic layers and stacking of various planes during carbonization [Kurosaki et al. (2003)]. As can be inferred from XRD pattern that the compounds formed were probably K 2 O and K 2 CO 3 corresponding to the peak of 2θ = 28, 31.5 and 42. K 2 CO 3 is formed by the reaction between KOH and CO 2. Finally at the end of the reaction K 2 O is left as such in the carbon [Hsu and Teng (2000)]. The broad peak at 2θ = 24 corresponds to the presence of silica and thereby confirming the existence of amorphous SiO 2 in activated charcoal. 108

8 5.3.3 Surface Characterization of Coconut Shell based Activated Carbon The N 2 adsorption-desorption isotherm of CST is shown in Fig According to IUPAC classification the isotherm of CST can be classified as type IV isotherm. The isotherm clearly shows that at comparatively lower pressure, there is a sudden increase in the adsorbed volume and the knee of curve become rounded. Clearly, CST has wide pore size distribution that ranges Figure 5.3: Nitrogen adsorption-desorption isotherm and pore size distribution (inset) of CST at 77 K. from micropores to mesopores. Moreover, at relatively higher pressures (P/P 0 0.1), the data shows a linear change in the adsorption capacity with increasing pressure, suggesting the presence of much more widely distributed and heterogeneous mesoporosity. The inset of Fig. 5.3 represents the pore size distribution curve of CST. The average pore size of CST is found to be 3.0 nm which are characteristics of mesopores and was calculated from the desorption branch of the isotherm using BJH (Barrett-Joyner-Halenda) model. The data extracted from this isotherm demonstrate a well-developed porosity. The BET surface area, micropore volume and average pore diameter are enlisted in Table

9 Table 5.1 Sample BET Surface Area Micropore Volume Pore Size (m 2 g -1 ) (cm 3 g -1 ) (nm) CST * CST * = Coconut shell based chemically treated activated charcoal Thermogravimetric Analysis (TGA) As we know that biomass may be converted into carbon by controlled thermal decomposition. The steps involved in the conversion of biomass to carbon are: (1) up to 150 C desorption of adsorbed water takes place, (2) splitting up of matter structure and water between 150 C and 260 C, (3) chain scissions or depolymerization and breaking of C O and C C bonds within ring units releasing water, CO and CO 2 between 260 C and 400 C, (4) aromatization forming graphitic layers above 400 C and (5) above 800 C, thermal induced decomposition and the rearrangement reaction are almost terminated leaving behind a carbon template structure. The systematic order of breaking down of main components of biomass is: C (hemicellulose), C (cellulose) and C (lignin). 80% of the total weight loss takes place between 260 C and 400 C, which may vary between 40% lignin to about 80% cellulose because of the evolution of H 2 O, CO 2 and volatile hydrocarbon species from fragmentation reaction of the polyaromatic constituents [Qian et al. (2004)]. Fig. 5.4 shows the TGA curve of CST under N 2 atmosphere. It can be seen from the plot that during thermogravimetric analysis, three weight loss stages were observed in the present case. In the first stage i.e. <150 C a moisture content of 11.4% in CST has been recorded. In the second stage, i.e. from 150 to 500 C devolatilization of 36.4% is observed, it may be because of the decomposition of remaining amount of hemicellulose which is further followed by cellulose degradation. A considerable amount of decomposition in the third stage i.e. from 500 to 750 C with a decomposition rate of 54.4% is observed in CST. It is very much clear from TGA results, that higher lignin content and higher char yield were obtained from CST. Hence, in the present studies lignin is found to be the main contributor to the final char weight [Cagnon et al. (2009)]. 110

10 Figure 5.4: Thermogravimetric curve (TGA) of CST Characteristics of EDLC cell EDLC cell of the following configuration have been constructed using CST based chemically treated activated charcoal powder electrodes with optimized blend polymer gel electrolyte: Cell A: CST PVdF(HFP)-PMMA-NaSCN CST where, CST is coconut shell based chemically treated activated charcoal powder electrode on carbon cloth. In order to characterize fabricated EDLCs, different physical techniques like a.c impedance spectroscopy, cyclic voltammetry, galvanostatic charge-discharge tests and prolonged cyclic test have been adapted. Impedance Spectroscopy Impedance spectroscopy is one of the most commonly used analytical tools for electrochemical characterization of EDLCs. In this technique, impedance is generally measured by applying an a.c potential to an electrochemical cell and thereby measuring the current passing through it. The typical impedance spectrum of EDLC cell has been carried out at room temperature and it is shown in Fig It should be noted that pure capacitor cell shows the impedance behavior, which is a straight line parallel to imaginary axis. In the present studies, capacitor cell A shows a small semicircle in the higher frequency range followed by a steep rising portion towards lower frequency region. 111

11 Figure 5.5: Nyquist plot of an EDLC cell CST PVdF(HFP)-PMMA-NaSCN CST at room temperature It may be noted that higher frequency semicircular spur shows the bulk properties of electrolyte and charge transfer processes at electrode-electrolyte interfaces, whereas the steep rising behavior in the lower frequency range shows the capacitive nature of the cell. The values of bulk resistance R b and charge transfer resistance R ct as calculated from the intercepts of real axis of the complex impedance response, along-with the total resistance R and capacitance C measured at a frequency of 10 mhz and 100 mhz are summarized in Table 5.2. Table mhz 10 mhz Cell R ct R b R C R C ( cm 2 ) ( cm 2 ) ( cm 2 ) (mf cm -2 ) a (F g -1 ) b ( cm 2 ) (mf cm -2 ) a (F g -1 ) b A a Overall capacitance of the cell b Single electrodes specific capacitance of the cell 112

12 A hypothetical electrical circuit which consists of parameters with well defined electrical properties have been used to describe the impedance response of the coconut shell based EDLC. Figure 5.6(a) gives the impedance measurement and simulation circuit response of an EDLC cell with a good fit having minimum chi-squared value of which has been evaluated by using SAI ZPLOT impedance analyzer. Fig. 5.6(b) shows equivalent circuit model in which R b is the bulk resistance, R ct is the charge transfer resistance, C is the double layer capacitance and Z W represents the Warburg diffusion. As can be seen from the plot, EDLC cell shows capacitive behavior in the region of high to medium frequency as shown by depressed semicircle in the impedance pattern. The semicircle has two intersection points on the real axis; the first point at the left end which represents the ohmic resistance of the bulk electrolyte solution R b that is connected in series in the equivalent circuit model as shown in Fig. 5.6(b). Figure 5.6: (a) Experimental plot (red) and simulation plot (green) of the complex impedance spectrum of EDLC Cell A (b) equivalent circuit model 113

13 The diameter of the semicircle represents the charge transfer resistance (R ct ), modeled by inparallel resistance R ct in the equivalent circuit model. In the present case, the value of R ct is appreciably low which indicates the better electrolyte pore accessibility. On the other hand, towards the right side of semicircle, the equivalent circuit model in Fig. 5.6(b) includes Warburg diffusion constant (Z W ) which represents the diffusion of ion into the bulk of the electrode through pores of varying size. From impedance measurements, not only the ionic resistance but capacitance dependence with frequency inside the porous structure of the carbon can also be measured. It is one of the most important characteristic of the supercapacitor cell, as they are well known to deliver a high power during short duration [Taberna et al. (2003)]. Fig. 5.7(a) represents the change of real part of the capacitance C as a function of frequency for CST based electrodes using blend polymer electrolytes. Real part of capacitance (C ) is calculated according to the model given by Taberna et al. [Taberna et al. (2003)]. The model suggests that the supercapacitor behaves like a series combination of resistance and a capacitance which depends on frequency as per the relation: C (ω) = "( ) ( ) (5.5) where Z is the imaginary part of the impedance Z, ω is the frequency, C corresponds to the capacitance of the cell measured under direct current or low frequency a.c conditions. The plot helps in studying the change of the capacitance depending on the frequency from a qualitative point of view. Three regions can be distinguished from C versus frequency plot [Taberna et al. (2003)]. At a high frequency, system behaves like a pure resistance and capacitance tends to zero because the ions can only access the outer part of the carbon grains without entering the porosity [Taberna et al. (2003)]. Whereas on the other hand, at very low frequency, the capacitance is maximal since the ion can reach the entire surface area in the depth of the carbon grain [[Taberna et al. (2003)]]. Between these two frequency boundaries, a transition region is observed where the capacitance varies linearly with the frequency. 114

14 (a) (b) Figure 5.7: (a) Real capacitance C and (b) imaginary capacitance C as a function of frequency It can be explained using electrical circuit model by a series of R, C parallel network which describes the penetration of electrolyte into the grains of carbon. These features can be clearly observed in Fig. 5.7(a). Figure 5.7 (c): Plots of normalized active power P / S and reactive power Q / S as a function of frequency for EDLC cell 115

15 The CST based EDLCs present a capacitive plateaus at low frequency regime confirming that maximum capacitance has been achieved and it may be due to the good pore accessibility of the electrolyte. Fig. 5.7(b) shows the imaginary part of the capacitance (C ) as a function of frequency. The relaxation time computed from the frequency corresponding to the maximum of the curve is 2 seconds. Fig. 5.7(c) shows the plot of P / S and Q / S of the complex power as a function of frequency for the CST based EDLC cell. From the crossing of two plots at a frequency, f 0 the value of τ 0 has been calculated which gives the figure of merit for any supercapacitor cell. In the present study it is found to be of the order of 2 seconds. Cyclic Voltammetry Studies Cyclic voltammetry is an effective tool to reveal the capacitive behavior of given material. This technique gives us various kinds of information including nature of charge storage at the individual interfaces in the anodic and cathodic regions alongwith the overall behavior of the capacitor cell. Fig. 5.8 represents the cyclic voltammetric curves of capacitor cell A at different Figure 5.8: Cyclic voltammograms of an EDLC cell CST PVdF(HFP)-PMMA-NaSCN CST at different scan rates 116

16 voltage scan rates. It can be seen from a plot that the response for capacitor cell A strongly depends on the scan rate which is expected for a capacitor cell [Shen et al. (2013)]. There is no evidence of any redox current on both positive or negative voltammetric sweeps and current is almost constant over most of the potential range. As the scan rate increases, CV curves for CST electrode is tilted to some extent due to the presence of equivalent series resistance (ESR) which is practically present in the real capacitor cell, but still it shows a rectangular-like shape which clearly shows that CST as electrode material has high power capability with blend polymer electrolyte [Zhou et al. (2012), Xiong et al. (2011), Kuratani et al. (2011), Jain et al. (2013)]. The capacitance value of cell A, as calculated using equation (5.2) in the present studies is found to be of the order of 540 mf cm -2 which is equivalent to single electrode specific capacitance of 360 F g -1 and are almost in good agreement with the values obtained from impedance analysis and charge-discharge measurements. This larger value of capacitance for CST based EDLC is due to its larger surface area and high porosity. From application point of view, cyclic efficiency is an important and essential parameter of EDLC over rechargeable batteries [Béguin and Frąckowiak (2013)]. Figure 5.9: Variation of capacitance as a function of voltammetric cycles at a scan rate of 100 mv s

17 Therefore the electrochemical stability of CST based EDLC was evaluated by repeating the cyclic voltammetry test at a scan rate of 100 mv s -1 for 600 cycles. The specific capacitance of the electrode material as a function of cycle number is shown in Fig It has been observed that CST electrode exhibit excellent cycle life in the entire cycle numbers under preliminary investigations and it shows almost stable and constant values of capacitance for Cell A upto 600 cycles and even more. A slight fluctuation in capacitance values has been observed for initial few cycles which can be explained by irreversible charge consumption due to some faradic reactions associated with the possible oxidation and reduction of loosely bound surface groups, specifically hydroxyl group at the electrode/electrolyte interfaces [Béguin and Frąckowiak (2013)]. After initial few cycles the interaction force between electrode and electrolyte remain unchanged which shows that the transfer ability of charges would remain fairly constant. Fig represents the specific capacitance versus scan rate for CST based EDLC cell using blend polymer electrolytes. As can be seen from the plot that specific capacitance decreases with increase in the scan rate. This decrease is attributed to diffusion limitations of electrolyte in the pores of CST and low ion mobility within the pores. As the scan rate was increased to 500 mv s -1 practically no capacitive behavior was observed. Figure 5.10: Variation of capacitance of an EDLC cell as a function of scan rate 118

18 Galvanostatic Charge Discharge Studies Galvanostatic charge discharge technique is carried out by keeping current constant until the device reaches a set potential limit. Then the current is reversed, and the process is repeated. The resulting potential is plotted as a function of time. This technique is most commonly used because of its simplicity and straight forward parameter extraction, like the ohmic drop V is the product of the switching current I and the series resistance R ; and the charge-discharge rate is equal to the ratio of the charge/discharge current to the ideal capacitance C. The galvanostatic charge-discharge profile of capacitor cell A is shown in Fig The charge-discharge curves clearly indicate that CST possesses a good capacitive behavior as can be inferred from the linear region of discharge curve. It is also observed that even at current density of 5.0 ma cm -2, there is no sudden voltage drop during the switching of current which shows quite low resistance of the electrode material. Figure 5.11: Charge discharge curve of EDLC cell CST PVdF(HFP)-PMMA-NaSCN CST at a current density of 5.0 ma cm

19 The specific capacitance, power and energy density values as obtained from charge-discharge curves are given in Table 5.3. Table 5.3 R i Discharge capacitance, C d Working voltage Energy density Power density Cells ( cm 2 ) (mf cm -2 ) a (F g -1 ) b (V) (Wh kg -1 ) (kw kg -1 ) A a Overall capacitance of the cell. b Single electrodes specific capacitance of the cell. Impact of Porosity The specific double layer capacitance measured from different technique tells us that CST based EDLC gives the satisfactory performance having capacitance value of 534 mf cm -2 which is much higher as compared to the result reported in literature for activated carbon with BET surface area close to 3000 m 2 g -1 [Keirzek et al. (2004)]. So it is clear that some additional factors apart from surface area is playing role to increase the capacitance values. It may be explained with ion sieving effect due to the small pore size of CST powder [Lozano-Castello et al. (2000), Gryglewicz et al. (2005), Kim et al. (2008)]. The achieved capacitance is maximum because the porosity of CST is well developed to remove the ion sieving effect. Hence it may be inferred from the studies that porosity plays a key role in the enhancement of capacitance. 5.4 Conclusions Combining all the results, it can be concluded that: High performance activated charcoal from natural coconut shell has been successfully synthesized by impregnation method with BET specific surface area of 1640 m 2 g -1 having mesopores of average pore size of 3.0 nm. 120

20 XRD studies reveals that the main constituents of CST based activated charcoal was predominantly graphite and amorphous carbon. The electrochemical studies suggest that CST electrode shows excellent capacitive behavior having specific capacitance of 534 mf cm -2, which is equivalent to single electrode specific capacitance of F g -1. The energy density of 88.8 Wh kg -1 and power density of 1.63 kw kg -1 has been achieved. The cell shows almost stable capacitance values upto 600 cycles and even more. Combining all the above studies, it may be concluded that CST based electrode can be considered as potential candidate for high performance supercapacitor. 121