A novel approach for supercapacitors degradation characterization

doi:10.1016/j.jpowsour.2017.04.048

Journal of Power Sources

Volume 355, 1 July 2017, Pages 74-82

https://doi.org/10.1016/j.jpowsour.2017.04.048 Get rights and content

Highlights

•
ISGP analysis method is applied on supercapacitors during degradation.
•
Genetic programming serves as the cornerstone of our EIS analysis approach.
•
Operation and degradation study of supercapacitors via solely AC measurements.
•
Capacitance and relaxation time of each process can be calculated individually.
•
Detecting supercapacitors failure earlier than resistance and capacitance imply.

Abstract

A novel approach to analyze electrochemical impedance spectroscopy (EIS), based on evolutionary programming, has been utilized to characterize supercapacitors operation mechanism and degradation processes. This approach poses the ability of achieving a comprehensive study of supercapacitors via solely AC measurements. Commercial supercapacitors were examined during accelerated degradation. The microstructure of the electrode-electrolyte interface changes upon degradation; electrolyte parasitic reactions yield the formation of precipitates on the porous surface, which limit the access of the electrolyte ions to the active area and thus reduces performance. EIS analysis using Impedance Spectroscopy Genetic Programming (ISGP) technique enables identifying how the changing microstructure is affecting the operation mechanism of supercapacitors, in terms of each process effective capacitance and time constant. The most affected process is the transport of electrolyte ions at the porous electrode. Their access to the whole active area is hindered, which is shown in our analysis by the decrease of the capacitance gained in the transport and the longer time it takes to penetrate the entire pores depth. Early failure detection is also demonstrated, in a way not readily possible via conventional indicators. ISGP advanced analysis method has been verified using conventional and proven techniques: cyclic voltammetry and post mortem measurements.

Graphical abstract

Introduction

Supercapacitors (SCs), unlike other electrochemical storage devices, are able to deliver and store energy rapidly [1]. SCs ability to provide markedly high power densities (10³–10⁴ W kg⁻¹) is a result of high surface area electrodes with a narrow distance to opposing ions at the double layer interface [2], [3]. The lack of chemical reactions in the main energy storage mechanism leads to prolonged cycle-life [4]. It is common to monitor the state of the SC by measuring its equivalent series resistance (ESR) and capacitance. Once the ESR has increased by 100% from its initial state or the capacitance has declined by 20%, the cell has reached its end of life criteria [5].

The mechanisms associated with SCs performance degradation have been mainly related to decrease in the active area of the porous electrodes [6], [7]. A schematic structure of the SC active area is shown in Scheme 1. Organic electrolytes, most commonly acetone-nitrile (AN) and polycarbonate (PC), were found to react with the activated carbon (AC) electrodes to form polymeric products that precipitate onto the active surface and block the pores [8], [9]. Further electrolyte parasitic reaction leads to the formation of gaseous products, which causes an increased internal pressure in the device [10], [11]. This also contributes to depletion in the solvent content and precipitation of salts on the AC surface, which further diminishes the active area for the desired energy storage [12]. An additional factor that also advances degradation is oxidation of the current collector, causing its delamination from the electrodes and deterioration of the separator [5]. These findings were based on accelerated ageing tests, which included elevated temperatures and voltages that simulated standard ageing conditions [13].

One common method to comprehensively monitor the degradation process is electrochemical impedance spectroscopy (EIS) [15]. Preformed on a wide range of frequencies, EIS is especially beneficial for degradation characterization, since it is nondestructive and enables the assessment of the state of the SC without affecting the cell. Even though the measurement procedure is performed relatively easily, the interpretation of the results is a challenging task. Several models have been proposed in order to adequately characterize EIS measurements of SCs. All of them note that the spectrum is divided into two distinguished regions: the high frequency (HF) region where the transport of electrolyte ions towards the pores takes place, but the penetration depth is less than the pores size; the low frequency (LF) region, where the frequency is low enough to allow the ions to penetrate and reach the entire pore depth and the whole active area is approachable [16]. The De Levie model, which was introduced first, is based on the assumption that the pores are all identical, with a negligible diameter [17]. Since it does not take into account the dispersion in pore size, models containing constant phase element (CPE) were introduced [18] as alternatives. Other models, such as the multi pore (MP), suggest that the sample is comprised of different populations of pore sizes, each has its own contribution to the impedance spectrum [19]. Another model by Suss et al. also captures the difference in porosity in the electrodes structure [20].

Each of the above models highlights different aspects of the tested sample in an attempt to adequately interpret the measured data (e.g. the MP model and CPE; both fit the data well) and require some a priori knowledge regarding the sample. Moreover, oftentimes, while measuring during ageing, the impedance spectrum of the sample may change considerably; thus, the original model may not depict the data sufficiently at that stage. Our approach, laid in the following section, focuses on the analysis of relaxation times, which poses the ability to separate the different contributions to the impedance more clearly than other common techniques, such as equivalent circuit modelling (ECM) [21], [22]. Our approach introduces a novel technique to find the distribution function of relaxation times of SCs without almost any pre-knowledge regarding the system; in addition, it is flexible enough to depict plainly the changing impedance spectrum of the sample under test. In that manner, we are able to monitor the changes in the model parameters and correlate them to the known degradation phenomena. The aim of this article is to show a successful utilization of our analysis approach to monitor the degradation of SCs. Furthermore, by monitoring the change in the model parameters we are able to predict that the sample approach its end of life, before other common indicators (i.e., ESR and capacitance) suggest it.

Section snippets

Theoretical basis

Our analysis approach is based on finding the distribution function of relaxation times (DFRT), which stems from the EIS measurements, using an evolutionary programming technique [23]. The correlation between the impedance and the DFRT is presented in Eq. (1): $Z (ω) = R_{\infty} + R_{p o l} \int_{- \infty}^{\infty} \frac{Γ (\log (τ))}{1 + i ω τ} d (\log (τ))$

Where Z is the impedance, R_∞ is the series resistance, R_pol is the total polarization resistance, Г is the DFRT, τ is the relaxation time and ω is the angular frequency. Equation (1) is the log scale

Accelerated degradation procedure

The tested cells used were commercial Murata SCs, model DMF4B5R5G105M3DTA0, capacitance of 1 F and maximum ESR of 50 mΩ. The cells were placed in an ageing chamber specially designed for this study (Figs. S1a and b). The chamber enables applying specific temperatures and voltages on the tested cells. We divided the cells into 2 groups; each was subjected to a different temperature during the accelerated tests, 70 °C and 50 °C. In each temperature group, the cells were held at 3 different

Supercapacitor operation characterization

All tested cells show the same DFRT plot, prior to their ageing test, as can be seen in Fig. 2. The DFRT plot is comprised of three distinct peaks, which are numbered in roman numerals in the figure. Peak I is associated with what is referred to in the introductory section as the HF region, i.e. the transport of electrolyte ions through the larger pores and towards the deeper pores, where most of the capacitance is attained [30]. Peak II, which takes place at lower frequencies, is correlated

Conclusions

We demonstrated the use of a novel method to analyze EIS of SCs regarding the characterization of their degradation process. We found that the change in the microstructure, caused by formation of precipitates due to electrolyte parasitic reactions, is manifested in our analysis in two major fashions: a decrease in the capacitance gained during the transport of ions towards the deeper pores (in the HF region of the spectrum) and a longer time for the ions to penetrate to the entire pore depth.

Acknowledgment

The financial support of the Nancy and Stephen Grand Technion Energy Program (GTEP) and the Israel Science Foundation grant No. 2797/11 and INREP-2 are gratefully acknowledged.

References (35)

A.G. Pandolfo et al.
Carbon properties and their role in supercapacitors
J. Power Sources
(2006)
P. Thounthong et al.
Energy management of fuel cell/battery/supercapacitor hybrid power source for vehicle applications
J. Power Sources
(2009)
R. Kötz et al.
Aging and failure mode of electrochemical double layer capacitors during accelerated constant load tests
J. Power Sources
(2010)
P. Azaïs et al.
Causes of supercapacitors ageing in organic electrolyte
J. Power Sources
(2007)
P.W. Ruch et al.
A comparison of the aging of electrochemical double layer capacitors with acetonitrile and propylene carbonate-based electrolytes at elevated voltages
Electrochim. Acta
(2010)
M. Tokita et al.
Degradation characteristics of electric double-layer capacitors consisting of high surface area carbon electrodes with organic electrolyte solutions
Electrochim. Acta
(2016)
P.W. Ruch et al.
Aging of electrochemical double layer capacitors with acetonitrile-based electrolyte at elevated voltages
Electrochim. Acta
(2010)
P. Kurzweil et al.
Electrochemical stability of organic electrolytes in supercapacitors: spectroscopy and gas analysis of decomposition products
J. Power Sources
(2008)
A.M. Bittner et al.
Ageing of electrochemical double layer capacitors
J. Power Sources
(2012)
O. Bohlen et al.
Ageing behaviour of electrochemical double layer capacitors. Part I. experimental study and ageing model
J. Power Sources
(2007)

R. de Levie

On porous electrodes in electrolyte solutions

Electrochim. Acta

(1963)

R. Kötz et al.

Principles and applications of electrochemical capacitors

Electrochim. Acta

(2000)

M.E. Suss et al.

Impedance-based study of capacitive porous carbon electrodes with hierarchical and bimodal porosity

J. Power Sources

(2013)

B.A. Boukamp

Fourier transform distribution function of relaxation times; application and limitations

Electrochim. Acta

(2015)

F. Ciucci et al.

Analysis of electrochemical impedance spectroscopy data using the distribution of relaxation times: a Bayesian and hierarchical Bayesian approach

Electrochim. Acta

(2015)

S. Hershkovitz et al.

Harnessing evolutionary programming for impedance spectroscopy analysis: a case study of mixed ionic-electronic conductors

Solid State Ion.

(2011)

J. Chmiola et al.

Effect of pore size and surface area of carbide derived carbons on specific capacitance

J. Power Sources

(2006)

Cited by (51)

A scenario-based genetic algorithm for controlling supercapacitor aging and degradation in the industry 4.0 era
2024, Engineering Applications of Artificial Intelligence
Electric double layer capacitors (EDLCs) are promising energy storage solutions, yet aging and degradation issues impede reliability and lifespan. The research proposes integrated simulation, modeling, and optimization to actively control EDLC degradation during charge-discharge cycles, mathematically modeling and simulating electrical and aging dynamics. These aging simulations are coupled with a genetic algorithm (GA) optimization routine that identifies the optimal combinations of influential EDLC parameters like internal resistance and capacitance for mitigating deterioration. Equivalent circuit models quantify electrical signatures, as aging factors induce gradual drifts in capacitance and resistance during thousands of simulated operational cycles. These MATLAB simulations effectively capture aging phenomena noted in real EDLCs in terms of measurable capacitance fading and resistance growth trends over continual usage in line with experimental data. The GA optimization subsequently determines optimal charging voltage ranges and achievable reductions in charge/discharge asymmetry by over 70 % that significantly enhance lifespan trajectories through aging control under similar test conditions. The technique's efficacy is further ascertained through systematic tuning of GA parameters like mutation rates and population sizes using Taguchi experimental models. The findings showcase superior optimization outcomes for larger populations and lower mutation probabilities. The research integrates digital twin for rapid evaluations, addressing reliability challenges via computational aging control. The flexible modeling platform enables customized what-if analyses for EDLC designers, aiding in material, cycling, and duty cycle exploration. This aging mitigation approach offers simulation-driven insights and automated optimization tools to extend the operational duration of high-performance EDLCs.
Distribution function of relaxation times: An alternative to classical methods for evaluating the reaction kinetics of oxygen evolution reaction
2023, Chemical Engineering Journal
Reaction kinetics of RuO₂ is precisely evaluated by distribution function of relaxation times (DFRT) model using impedance spectroscopy analysis by genetic programming (ISGP). Effective resistances of the Faradaic processes, measured using Electrochemical impedance spectroscopy (EIS) at various overpotentials, were determined utilizing DFRT, by separating and associating three electrochemical phenomena occurring during the reaction. The effective resistances are used to generate a Tafel plot of potential as a function of $\log (\frac{1}{R_{eff}})$ . The classical method, based on linear sweep voltammetry (LSV), to evaluate the Tafel slope is associated with some considerations for accurate results. For RuO₂, a reference catalyst, LSV illustrates an average Tafel slope of 182 ± 8 mV/dec while the effective resistance method, estimated using DFRT, shows an average of 181 ± 3 mV/dec. A relative error of 0.3 % between the two methodologies, and a lower standard deviation for DFRT demonstrate the higher precision and effectiveness of ISGP in determining the reaction kinetics via Tafel slope analysis. Therefore, using DFRT with the ability to separate Faradaic from non-Faradaic processes to evaluate the relevant part of the effective resistance, reaction kinetics can be estimated, avoiding shortcomings of the classical Tafel method.
Electrochemical capacitor in aqueous electrolyte with long lifespan improved by hydrogen bond donor addition
2023, Materials Today
In the electrochemical capacitors (ECs) operating with aqueous electrolytes, corrosion of the current collector is one of critical issues that affects the cycling life, efficiency, and capacitance of the devices. So far, there are no stable metal current collectors in the water-based electrolyte because of its corrosive nature, even if the pH is close to neutral. Here, urea (CH₄N₂O) is used as an anticorrosive, green additive to 1 M lithium sulfate electrolyte for low-cost electrochemical capacitors based on carbon BP2000 electrodes. In this work, the corrosion study of stainless steel in the presence of such amide demonstrates an effective corrosion-inhibiting character owing to the urea interaction with water in the hydrogen-bonded network. Chemical stability of current collectors owing to a significant 10-fold decrease of the corrosion current and shift of the corrosion potential into negative side has been achieved when 2 M urea, i.e., hydrogen-bond donor (HBD) is added to electrolyte. Furthermore, the addition of this organic compound improves the capacitive performance of the system and prevents the fading of the device with no influence on the electrolyte conductivity. During the floating test of EC at 1.6 V operating in 1 M Li₂SO₄ with 2 M urea, the capacitance and resistance values keep the end-of-life criteria for over 350 h. Moreover, the galvanostatic investigation of such EC (at 1 A/g) displays performance of more than 60 000 cycles. Furthermore, operando experiments proved that carbon-based electrodes behave differently in lithium sulfate with urea and without additive owing to the formation of strong hydrogen bond network. The molecular structure of electrolyte has been estimated by DFT calculations. Overall, by using lithium sulfate and urea as an electrolytic system, the EC current collector has not only superior corrosion resistance against the aqueous electrolyte but also a high anti-ageing character at 1.6 V, which is favorable to improve the EC electrochemical performance. The use of urea as a stable, cheap electrolyte additive provides an innovative technique to enhance the performance of low-cost aqueous based supercapacitors.
Triethylammonium thiocyanate ionic liquid electrolyte-based supercapacitor fabricated using coconut shell-derived electronically conducting activated charcoal electrode material
2022, Journal of Energy Storage
Organic solvents are commonly used as commercial electrolytes in supercapacitors, but they have several drawbacks, such as high volatility, toxicity, and environmental issues. To overcome these problems, ionic liquids are used though their high costs hamper commercial applications. In this study, a novel and low-cost ionic liquid based on triethylammonium cation and thiocyanate anion were synthesized and used as the electrolyte in a low-cost supercapacitor fabricated using electronically conducting activated carbon derived from cleaned coconut shells as electrode material. The temperature dependence of electrolyte conductivity was investigated, in the frequency range of 1 Hz-500 kHz, using AC impedance analysis, in the temperature range from 30 °C - 90 °C in ten-degree increments and found to be in the range of 4.5 to 14.4 mS cm⁻¹. The ion transport kinetics behaves according to the Vogel-Tammann-Fulcher (VTF) model with an activation energy (E_a) and the pre-exponential factor (A_σT) of 0.0158 eV and 114.9 S m⁻¹ K^1/2, respectively. The activated carbon material developed has a high porosity and high electronic conductivity. The electrode fabrication methodology was optimized by varying the polyvinylpyrrolidone (PVP) binder amount with respect to the mass of activated carbon on a titanium sheet as a function of heat treatment temperature. The supercapacitor that gives the highest specific capacitance is based on the electrode constructed using 5 % w/w PVP in 0.50 g of activated carbon deposited on a titanium sheet by heat treating at 200 °C for 20 min. At the scan rate of 5 mV s⁻¹, two-electrode CV studies revealed a specific capacitance of 36.8 F g⁻¹ for the optimized supercapacitor assembly. The electrolyte triethylammonium thiocyanate (TAT) demonstrated good electrochemical stability and a large operating voltage window of 1.8 V and high stability, with 94.4 % capacity retention after 1000 cycles. The AC impedance studies also gave comparable results.
The manuscript describes a method to prepare a novel and low-cost ionic liquid material triethylammoniumthiocyanate starting from triethylamine and ammonium thiocyanate, and its application as the ionic liquid electrolyte in a low-cost supercapacitor. The electrode material used in the supercapacitor was fabricated using electronically conducting activated carbon derived from cleaned coconut shells, which were deposited firmly adhering to titanium sheets using polyvinylpyrrolidone binder under optimized heat treatment conditions. As such, both the ionic liquid electrolyte and the activated carbon electrode material are novel and attractive materials for the commercial production of low-cost supercapacitors.
The nonlinearities in the galvanostatic charging curves of supercapacitors provide insights into charging mechanisms
2022, Journal of Energy Storage
Supercapacitors are often charged using constant currents. The capacitance can be determined from the slope of the voltage-time curve if the measured voltage over the supercapacitor increases linearly with time. However, the resulting voltage-time curve is often nonlinear, which may lead one to interpret the capacitance as being either time or voltage dependent. In the current work, systematic experimental studies of the nonlinearity of galvanostatic charging curves as a function of applied current and temperature are undertaken for commercial supercapacitors in the range 1–1000 F. A consistent theory is developed to explain the available data. It is demonstrated that the nonlinearity in the voltage-time curve can be attributed to a constant capacitance in parallel with a resistance, the latter which is inversely proportional to the applied current. The influence of faradaic charge transfer reactions or surface charge reorganization on this parallel resistance is analyzed. The proposed theory is also used to analyze galvanostatic charging data available in the research literature, and the different types of nonlinearities observed provide new insight into the mechanisms occurring during charging of various types of supercapacitors.
Anticorrosive performance of green deep eutectic solvent for electrochemical capacitor
2022, Chemical Engineering Journal
Citation Excerpt :
Owing to its costless, biodegradability, availability and non-toxicity Reline DES presents additionally an anticorrosive character [34] which makes it a promising alternative electrolyte for ECs. Besides the interest in improving ECs performance, the ageing and the electrochemical stability aspects have been investigated by several researchers[38,39]. ECs demonstrate several advantages among them the excellent long-term performance during charging-discharging cycles [2].
Reline Deep Eutectic Solvent (DES) is widely recognized as a novel and green electrolyte for industrial applications due to its low cost and tolerance to moisture. It has received significant attention, however, its application in the energy storage system is still in infancy. Generally, aqueous-based electrochemical capacitors (ECs) present an ideal energy storage technology because of several advantages. Yet, high corrosion of the current collector limits their cyclability. Novel Reline DES electrolyte is studied and its anticorrosive electrochemical performance is evaluated in carbon-based ECs using stainless steel current collector (316L). Our findings show that Reline electrolyte and its water mixture display a strong inhibiting corrosion effect for 316L. Furthermore, a few aqueous electrolytes are studied to confirm the effectiveness of Reline DES as a green and anti-corrosive electrolyte for EC. Floating protocol is applied to monitor the stability of the EC performance in Reline and aqueous electrolytes. The devices using pure Reline and its water mixture exhibit higher electrochemical stability with good capacitive performance compared to 1 M Li₂SO₄ and 1 M KOH based ECs. The study may pave the way to practical Reline DES-based ECs for sustainable, highly stable, environmentally friendly and cost-effective energy storage systems in the future.

View all citing articles on Scopus

¹: Both authors contribute equally to this work.

View full text