Elsevier

Journal of Power Sources

Volume 282, 15 May 2015, Pages 385-393
Journal of Power Sources

The importance of the electrode mass ratio in a Li-ion capacitor based on activated carbon and Li4Ti5O12

https://doi.org/10.1016/j.jpowsour.2015.02.079Get rights and content

Highlights

  • The performance of an AC//LTO asymmetric supercapacitor has been improved by modulating the electrode masses.

  • An AC/LTO mass ratio of 0.72 leads to higher energy density at high power density.

  • Lower AC/LTO mass ratio gives lower diffusion and charge transfer resistances.

Abstract

This study shows how the simple modulation of the cathode/anode mass ratio, in a Li-ion capacitor based on activated carbon (AC) and Li4Ti5O12 (LTO), results in a drastic increase in performance. Starting with a device balanced in the classical way (with an AC/LTO mass ratio of 4.17), the cathode/anode mass ratio has been reduced to 1.54 and then to 0.72. At a high power density, the device with a cathode/anode mass ratio of 0.72 shows the highest energy density. In fact, at 2.3 kW L−1, it delivers an energy density of 31 Wh L−1, which is almost 10 times greater than the energy obtained with a capacitor balanced with an AC/LTO ratio of 4.17 (3.68 Wh L−1). Moreover, the reduction in the cathode/anode mass ratio from 4.17 to 0.72 improves the cycling stability with a factor of 4.8 after 1000 cycles at 10C. Electrochemical impedance spectroscopy reveals that the better power performance is due to reduced diffusion and charge transfer resistances. In addition, the anode polarization is less pronounced for the system with a lower AC/LTO mass ratio, leading to a minimization in electrolyte decomposition on the anode surface and therefore limiting the increase in the electrode resistance during cycles.

Introduction

Li-ion capacitors (also called “hybrid” or “asymmetric” supercapacitors) have been introduced for the first time by Amatucci et al. [1]. They developed a non-aqueous, asymmetric-hybrid configuration with Activated Carbon (AC) as a positive electrode and a Li4Ti5O12 (LTO) intercalation compound as a negative electrode. The reaction mechanism involves lithium intercalation in the LTO structure and anion adsorption on the AC surface. Cations and anions are reversibly consumed from the electrolyte, which is typically based on LiPF6 or LiClO4 salt dissolved in alkyl carbonates, such as propylene carbonate (PC), ethylene carbonate (EC), dimethyl carbonate (DMC) or others. The reactions at the cathode and at the anode in LiPF6 can be described by the following equations:AC+PF6AC++PF6+e(cathode)Li4Ti5O12+3Li++3eLi7Ti5O12(anode)

The idea of Amatucci et al. prompted the scientific community to develop and improve such asymmetric systems by employing various types of Li-insertion materials, such as graphite [2], [3], LiMn2O4 [4], Li3V2(PO4)3 [5], LiFePO4 [6] and others. The main purpose of this hybridization is to bridge the gap existing between the electrochemical double layer capacitors (EDLCs) and the Li-ion batteries. In fact, on one side, EDLCs are able to deliver high power (above 10 kW kg−1), but they are very limited in energy (below 10 Wh kg−1). On the contrary, Li-ion batteries are high energy devices (150–250 Wh kg−1) but limited in power (below 1 kW kg−1) [7]. The hybrid approach can double (or even triple) the energy density in comparison with a conventional electrochemical capacitor. However, the kinetics of Li-intercalation are much lower in comparison to the double layer formation at the activated carbon electrode surface. This situation results in a certain power limitation of the asymmetric device in comparison to the classical EDLC. Recently, Cericola et al. [4] evaluated the behaviour of two types of asymmetric devices. The first device was made with an LTO anode and an AC cathode, the second with AC as anode and LiMn2O4 as the cathode. In addition, they also explored the possibility of using bi-material composite electrodes obtained by the combination of Li-insertion and double layer materials mixed together in the same electrode. In his work, Cericola demonstrated that the two types of asymmetric systems are competitive with neither EDLCs nor Li-ion batteries. The main reason is that the battery material represents a serious power limitation for the whole system. One possible solution to overcome this issue is to increase the rate capability of the lithium insertion material by properly controlling the particles size and morphology or by introducing very high conductive carbons materials. Naoi et al. presented this concept in a very elegant way by synthesizing ultrafast LTO anchored to carbon nanofibers [8] or to carbon nanotubes [9]. However, the production of such types of materials can be rather expensive at the industrial scale.

If one looks closer at the design of a classical asymmetric system, it is possible to make some considerations. A mass balance between the cathode and anode is normally made to take full advantage of the performances of both materials in their optimal working potential range. For this purpose, the charge passing through the positive and the negative electrodes must be balanced such that QAC = QLTO. The theoretical capacity of LTO is 175 mAhg−1 (when it is intercalated to Li7Ti5O12), and its reversible capacity at a low C-rate can range from 150 to 170 mAhg−1. Instead, the capacity of the activated carbon depends on the specific capacitance and on the potential range used (QC·V). Therefore, depending on the specific properties of the activated carbon (i.e., surface area and pore size distribution), the cathode capacity can range from approximately 30 to 45 mAhg−1 in a voltage range between 3 V and 4.3 V vs. Li/Li+. Therefore, the optimal mass ratio between the AC cathode and LTO anode should be approximately four or five, as reported in many publications [1], [4], [10], [11]. However, it is important to note that all of these considerations are correct only if the electrodes are cycled at relatively low C-rates (e.g., C/5 or 1C). Instead, a supercapacitor is a device that is meant to work at high currents, where the above calculations are not valid anymore. This problem has already been discussed by Li et al. [12]. They have provided a mathematical analysis of asymmetric capacitors in relation to the mass ratio and other parameters, such as the working voltage and specific capacitance as well as the energy and power densities.

In another report [13], Zheng discussed a model for predicting the maximum energy density achievable in the asymmetric systems. He illustrated that the specific capacitance and energy density are functions of the electrode mass ratio, volume ratio and ion concentration in the electrolyte. Both studies are based on theoretical assumptions. They assume that the voltage swing of the device only occurs at the EDLC electrode, while the faradaic electrode potential does not change during charge or discharge. Another of their assumptions is that the polarization of the battery-type electrode is so mild that its charge plateau practically overlaps with the discharge plateau. Furthermore, in their study, Li et al. assume that the resistance of the capacitor-type electrode can be completely neglected (iR = 0). All of these assumptions can be valid only at a low current. In a real situation, at high currents, all of these phenomena have to be considered. At the state of the art, there are no scientific reports that empirically analyse the influence of the AC/LTO mass ratio on the energy and power of an asymmetric supercapacitor. In this work, we evaluate the electrochemical performances of an asymmetric AC//LTO device balanced on the classical method (at a rate of 1C) and we compare it to other AC//LTO systems with different cathode/anode mass ratios. The asymmetric devices have been analysed in terms of galvanostatic charge/discharge, in terms of anode–cathode potential swing at various currents and also in terms of Electrochemical Impedance Spectroscopy (EIS). The separate contributions of the anode and cathode have also been evaluated. In addition, it is important to notice that the devices have been constructed using only commercial materials that are available at large scale, which allows for easy up-scaling and commercialization.

Section snippets

Electrode preparation

The activated carbon-based electrodes were prepared as described in our previous work [14]. In detail, 85 wt% of activated carbon (HDLC-20BST-UW, Haycarb, Sri Lanka), 5 wt% of carbon black (SuperP, Timcal, Switzerland) and 10 wt% of polytetrafluroethylene (PTFE) as a binder (TF 2025 Z, 3M Dyneon, Germany) were mixed dry. The obtained mixture was compacted with a mortar, and the resulting dough was then calendered to obtain a free-standing film. The thickness was modulated to obtain the desired

Analysis of single electrodes

The anodic and cathodic materials were first electrochemically analysed in a half-cell configuration. Fig. 1a shows the potential window opening experiment on the activated carbon electrode. The goal of the experiment was to obtain information about the electrochemical window inside which the activated carbon can be used without decomposition. The quantification of the stability limit can be calculated by determining the faradaic fraction (represented by R) using the following formula:R=QaQc1

Conclusions

With this work, we demonstrated that it is possible to overcome the limitations of AC//LTO asymmetric supercapacitors by controlling the mass ratio between the cathode and anode. A device balanced using the classical method (by fully utilizing the capacity of the anode and cathode) is not competitive with symmetric AC//AC systems at a high power level.

It was found that the increase in performance is due to: (1) the thickness of the AC electrode – a lower AC/LTO mass ratio leads to a lower AC

Acknowledgements

Financial support from the German Federal Ministry of Education and Research (BMBF) under the Grant FKZ 03EK3021 (project Novacap) is gratefully acknowledged. The authors also wish to thank Clariant for providing the Li4Ti5O12 and HayCarb for providing the activated carbon.

References (23)

  • S.R. Sivakkumar et al.

    Electrochim. Acta

    (2012)
  • V. Khomenko et al.

    J. Power Sources

    (2008)
  • D. Cericola et al.

    J. Power Sources

    (2011)
  • N. Böckenfeld et al.

    J. Power Sources

    (2013)
  • N. Böckenfeld et al.

    J. Power Sources

    (2011)
  • K. Naoi et al.

    J. Power Sources

    (2010)
  • H-G. Jung et al.

    J. Power Sources

    (2013)
  • C.M. Ionica-Bousquet et al.

    J. Power Sources

    (2011)
  • J. Li et al.

    J. Power Sources

    (2009)
  • S. Dsoke et al.

    J. Power Sources

    (2013)
  • K. Xu et al.

    Electrochim. Acta

    (2001)
  • Cited by (0)

    View full text