High energy supercapattery with an ionic liquid solution of LiClO

10 Supercapattery combining an ideally polarized capacitor-like electrode and a battery-like 11 electrode is demonstrated theoretically and practically using an ionic liquid electrolyte 12 containing 1-butyl-1-methylpyrrolidinium tri(pentafluoroethyl)trifluorophosphate (BMPyrrFAP), 13 gamma-butyrolactone (γ-GBL) and LiClO4. The electrochemical deposition and dissolution of 14 lithium metal on a platinum and glass carbon electrode were investigated in this ionic liquid 15 solution. The CV data shows the fresh electrochemical deposited lithium metal is stable in the 16 electrolyte, which encourages the investigation of this ionic liquid solution in supercapattery 17 with a lithium battery negative electrode. The active material counted specific energy of the 18 supercapattery based on a lithium negative electrode and an activated carbon (Act-C) positive 19 electrode can reach 230 Wh kg -1 under the Galvanostatic charge-discharge current density of 1 20 mA cm -2 . The positive electrode material, Act-C was also investigated by CV, AC impedance, 21 SEM and BET. The non-uniform particle size and micropore porous structure of the Act-C 22 enable its electric double layer capacitor (EDLC) behavior in the ionic liquid solution. The 23 This is the accepted manuscript which differs slightly from the finally published version. http://dx.doi.org/10.1039/C5FD00232J calculated specific capacitance of the Act-C in this ionic liquid solution is higher than same Act1 C in aqueous solution, which indicates the pseudocapacitance behaviour of Act-C with the 2 species in the ionic liquid electrolyte. 3 4 Introduction 5 Supercapattery (=supercapacitor + battery) takes the advantages of both supercapacitor 6 (also known as electric double layer capacitor, or EDLC) and battery by combining an ideally 7 polarized capacitor-like electrode and a battery-like electrode. 1 Although lithium ion capacitor 2,3 8 is also comprised of this hybrid configuration, supercapattery is the more general term of the 9 particular design. 10 In theory, supercapattery can possess higher energy density than both battery and 11 supercapacitor and can supply this energy at a power output almost as high as supercapacitor. 12 The high power output of supercapattery is mainly a result of sharing the same electrochemical 13 active materials with supercapacitor, where the nanostructured carbons, like activated carbon 14 (Act-C), carbon nanotubes (CNTs) and graphene are the best choice for the ideally polarized 15 electrode because of their large surface area, porosity, stability over a wide potential window, 16 and intrinsically low electrical resistance. In addition, pseudo-capacitive materials including 17 MnO2, RuO2 and conducting polymers can also be used as the capacitor-like electrode materials 18 providing the high electrode capacitance, but limiting the potential windows. 19 As to the battery-like electrode, various electrode materials from commercial battery 20 systems can be the candidate, from lead acid batteries to metal/air systems, but in practice, metal 21 compounds 3-5 like SnO2, MnO2 and LiFePO4 are more common and commercially available. 22 Theoretically, the hypothetical battery comprising lithium metal and fluorine gas (Li-F battery) 23 would output a cell voltage about 6.1 V and offer a specific energy content of 6304 Wh kg -1 . Any 1 battery cannot go beyond the specific energy content of the hypothetical Li-F battery. As to Li2 ion battery (LixC6 | Li1-xCoO2), the theoretical specific energy is 552 Wh kg -1 at 3.5 V. A 3 hypothetical supercapattery comprising a lithium metal negative electrode and a supercapacitor 4 positive electrode (assumed 400 F g -1 ) is evaluated and analyzed here. Considering the specific 5 charge capacity of lithium is much larger than that of the supercapacitor electrode, the mass of 6 the lithium metal is negligible. The theoretical specific energy value would be 625 Wh kg -1 for 7 the cell voltage vary from 3.5 V to 1.0 V. This theoretical value is even higher than the one of 8 the Li-ion battery. 9 The above calculation is based on the equation of the capacitor energy, Eq. 1, 10 2 max max 2 1 CU E  (1) 11 where C is the specific capacitance of a capacitor, and Emax is the maximum energy capacity of a 12 capacitor correlated to its maximum tolerable voltage, Umax. It should be mentioned that a hybrid 13 cell of a battery electrode and a supercapacitor electrode does not always show typical 14 supercapacitor behaviour, where the cell voltage (U) is always proportional to the time (t) during 15 a constant current charging or discharging test. If the cell presents battery like features, the term 16 supercabattery is recommended, 6 but not discussed in this paper. In the case of supercapattery, 17 the Umax is a key factor for the energy capacity of the devices. Because the behaviour of 18 supercapattery is close to that of supercapacitor, 1 several strategies that have been applied in 19 supercapacitor can be utilized in supercapattery to improve the practical energy capacity, such as 20 using the design of asymmetric supercapacitor cell, 7 controlling the capacitance ratio of the 21 positive and negative electrode made from the same material, 8 and serially stacking the cells 22 through the bipolar electrodes. 9 Apart from these efforts based on the aqueous electrolytes, there 23 is a strong desire for changing the aqueous to organic electrolytes to achieve a high working 1 voltage. 2 Recent studies have revealed both aqueous and non-aqueous supercapatteries using lithium 3 metal 10 and Li-ion battery material 3 as the electrode. An aqueous supercapattery consisting of a 4 MnO2 positive electrode and a Li/LISICON/PEO-LiTFSI/Li + negative electrode had achieved a 5 specific energy capacity of 114 Wh kg -1 with a 4.3 V cell voltage. 10 The Li/LISICON/PEO6 LiTFSI/Li + electrode is a multi-layered Li electrode, which consists of lithium metal, a 7 LISICON-type solid glass ceramic as the water-stable solid electrolyte, and a buffer layer 8 consisting of polyethylene oxide with Li(CF3SO2)2N polymer electrolyte (PEO-LiTFSI) between 9 the lithium metal and the solid electrolyte. If MnO2 is replaced by RuO2 as the positive electrode, 10 the specific energy capacity of the device comes to 520 Wh kg -1 with a 3.8 V cell voltage. 10 11 However, the current density of the aforementioned devices is only 0.255 mA cm -2 , which is 12 limited by the solid/liquid interphase. This disadvantage prevents these high energy capacity 13 devices in any high power application. Another non-aqueous supercapattery using LiFePO4 as 14 the positive electrode and Cabot carbon black as the negative electrode possesses good cycling 15 stability at high current density in a Li-ion contained propylene carbonate (PC) electrolyte. 3 16 However, the potential range of the Cabot carbon black can only keep its capacitor-like 17 behaviour from 2.80 to 1.25 V vs. Li/Li + in the electrolyte, which has limited the cell voltage. 18 Ionic liquids are specially featured by their zero or negligible volatility, but still able to offer 19 the highly ionic environment, and wide temperature and potential windows. 11 They have brought 20 about unique opportunities, including synthesis, 12 trace analysis, 13,14 thermochromic/cryochromic 21 materials, 15-17 and electrochemical energy storage. 18-20 Consequently, ionic liquid solutions can 22 be chosen as competitive candidates for the electrolytes of supercapattery when the battery1 behaviour electrode is a lithium-ion or lithium battery negative electrode. 2 Here, we report the recent supercapattery work based on an activated carbon positive 3 electrode and a Li/Li + negative electrode using an ionic liquid electrolyte, 1-butyl-14 methylpyrrolidinium tri(pentafluoroethyl)trifluorophosphate (BMPyrrFAP) containing gamma5 butyrolactone (γ-GBL) and LiClO4. The characterization of the activated carbon has been done 6 and presented here, including the data of cyclic voltammogram (CV), AC impedance, SEM and 7 BET. The demonstrated supercapattery cell shows a clear capacitor-like behaviour in the CV and 8 Galvanostatic charge and discharge (GCD) tests. This particular hybrid design ensures the full 9 usage of the electrochemical window of the ionic liquid solution and the capacitance of the 10 activated carbon, and basically maintains the high power of the supercapacitor. 11 12 Experimental 13 In this work, a supercapacitor grade commercial product of activated carbon (Act-C, YP50F, 14 Kuraray Chemical Co.) was used. The other chemicals, 1-butyl-1-methylpyrrolidinium 15 tri(pentafluoroethyl)trifluorophosphate (BMPyrrFAP, Merck), 1-ethyl-3-methylimidazolium 16 tetracyanoborate (EMIM[B(CN)4], Merck), gamma-butyrolactone (γ-GBL, Sigma Aldrich) and 17 LiClO4 (Sigma Aldrich), were commercially available and used without further purification. 18 Lithium metal (foil, Sigma Aldrich) was kept and handled in an argon filled glove box. The 19 homemade electrochemical cells (2-electrode/3-electrode cells and sandwich type cell) were 20 fabricated in an argon-filled glove box (O2 < 10 ppm, H2O < 10 ppm) and transferred outside the 21 glove box for the electrochemical experiments by an AUTOLAB 302N potentiostat. Membrane 22 from Celgard was used as the separator in the sandwich type cell. 23 For the electrochemical tests, the Act-C powder was made into pellet type and casted 1 electrodes with PTFE and PVDF, respectively. The details of fabricating the Act-C/PTFE pellet 2 electrode can be found in previous publication from this laboratory. 21 The Act-C/PVDF electrode 3 was fabricated by the following process. 10 μL of an Act-C suspension (0.950 g Kuraray Act-C 4 and 0.050 g PVDF powder in 20 mL DMSO suspension) was dropped on a 5 mm diameter 5 graphite disc electrode. The electrode loaded with 0.5 mg of Act-C composite (95 % w. Act-C 6 and 5 % w. PVDF) was dried in a vacuum oven at 75 ○ C overnight, and then was ready for the 7 electrochemi


Introduction
Supercapattery (=supercapacitor + battery) takes the advantages of both supercapacitor 6 (also known as electric double layer capacitor, or EDLC) and battery by combining an ideally 7 polarized capacitor-like electrode and a battery-like electrode. 1 Although lithium ion capacitor 2,3 8 is also comprised of this hybrid configuration, supercapattery is the more general term of the 9 particular design. 10 In theory, supercapattery can possess higher energy density than both battery and 11 supercapacitor and can supply this energy at a power output almost as high as supercapacitor. 12 The high power output of supercapattery is mainly a result of sharing the same electrochemical 13 active materials with supercapacitor, where the nanostructured carbons, like activated carbon 14 (Act-C), carbon nanotubes (CNTs) and graphene are the best choice for the ideally polarized 15 electrode because of their large surface area, porosity, stability over a wide potential window, 16 and intrinsically low electrical resistance. In addition, pseudo-capacitive materials including 17 MnO 2 , RuO 2 and conducting polymers can also be used as the capacitor-like electrode materials 18 providing the high electrode capacitance, but limiting the potential windows. 19 As to the battery-like electrode, various electrode materials from commercial battery 20 systems can be the candidate, from lead acid batteries to metal/air systems, but in practice, metal 21 compounds 3-5 like SnO 2 , MnO 2 and LiFePO 4 are more common and commercially available. 22 Theoretically, the hypothetical battery comprising lithium metal and fluorine gas (Li-F battery) 23 would output a cell voltage about 6.1 V and offer a specific energy content of 6304 Wh kg -1 . Any 1 battery cannot go beyond the specific energy content of the hypothetical Li-F battery. As to Li-2 ion battery (Li x C 6 | Li 1-x CoO 2 ), the theoretical specific energy is 552 Wh kg -1 at 3.5 V. A 3 hypothetical supercapattery comprising a lithium metal negative electrode and a supercapacitor 4 positive electrode (assumed 400 F g -1 ) is evaluated and analyzed here. Considering the specific 5 charge capacity of lithium is much larger than that of the supercapacitor electrode, the mass of 6 the lithium metal is negligible. The theoretical specific energy value would be 625 Wh kg -1 for 7 the cell voltage vary from 3.5 V to 1.0 V. This theoretical value is even higher than the one of 8 the Li-ion battery.

9
The above calculation is based on the equation of the capacitor energy, Eq. 1, where C is the specific capacitance of a capacitor, and E max is the maximum energy capacity of a supercabattery is recommended, 6 but not discussed in this paper. In the case of supercapattery, 17 the U max is a key factor for the energy capacity of the devices. Because the behaviour of 18 supercapattery is close to that of supercapacitor, 1 several strategies that have been applied in 19 supercapacitor can be utilized in supercapattery to improve the practical energy capacity, such as from Celgard was used as the separator in the sandwich type cell.

23
For the electrochemical tests, the Act-C powder was made into pellet type and casted 1 electrodes with PTFE and PVDF, respectively. The details of fabricating the Act-C/PTFE pellet 2 electrode can be found in previous publication from this laboratory. 21 The Act-C/PVDF electrode 3 was fabricated by the following process. 10 µL of an Act-C suspension (0.950 g Kuraray Act-C 4 and 0.050 g PVDF powder in 20 mL DMSO suspension) was dropped on a 5 mm diameter 5 graphite disc electrode. The electrode loaded with 0.5 mg of Act-C composite (95 % w. Act-C 6 and 5 % w. PVDF) was dried in a vacuum oven at 75 ○ C overnight, and then was ready for the 7 electrochemical tests.

8
All the experiment is operated at room temperature. More experimental details are specified 9 in the following sections.

11
Results and discussion 12 The pore size and volume distribution of an Act-C are the most important physical 13 characteristics. Fig. 1 presents the SEM images of an Act-C pellet containing 5 % w. PTFE. The

14
Act-C powder is non-uniform in particle size, and the PTFE is binding the Act-C particles like a 15 spider web to maintain the mechanical strength of the Act-C pellet. The surface area of the 16 Kuraray Act-C sample is 1724 m 2 g -1 determined from the nitrogen absorption/desorption 17 isotherms at 77 K (ASAP 2420) by applying the density function theory (DFT). In previous 18 work, 21 the dominant range of pore widths of the same Kuraray Act-C sample is between 1.5 and 19 2.5 nm. For an aqueous electrolyte, its wetting ability with the Act-C pellet will affect the contact 20 between the electrolyte and Act-C particles, and vary the charge storage performance of the Act-21 C sample. As to the organic electrolyte, the wetting is not a problem anymore, while the 22 interaction between electrolyte and carbon materials inside the pore will play an important role.

23
A theory was proposed for the traditional organic electrolyte, that the longer the pore inside the 1 particle is, the poorer performance of the capacitance of the porous carbon materials is. 3 This is 2 mainly due to the ion block of the pores by the non-ionic solvent molecular. Similar phenomenon 3 was also observed in the case of aqueous electrolyte with organic additive. 21  . This reference can be prepared following the method described in reference. 22 It should be mentioned that the reference electrode is changed to Li/Li + electrode in the latter 1 discussion except in Fig. 2. Any potential in this paper will be suffixed with the reference 2 electrode used in the experiment.  while the resistance of the cell will increased compared to the cell using an aqueous electrolyte.

8
In this case, changing electrolyte from an aqueous to a non-aqueous electrolyte is not economic.

9
A new strategy should be made to improve the energy capacity of the cell. shows an electrochemical window more than 4 V using Pt and GC electrodes. The ionic liquid 4 solution could be a potential candidate for the supercapattery electrolyte.  The galvanostatic charge-discharge test was run under a current density of 1.02 mA cm -2 .

7
Because the negative electrode is a lithium foil, the voltage of the cell decays smoothly and 8 linearly during the discharging process from 4.3 to 1.7 V. Because the charging and discharging 9 curve is almost symmetric, we still can calculate the specific capacitance and energy capacity by