Search

The extraordinary stability and cycle life performance of today’s electrochemical double-layer capacitors (EDLCs) are generally ascribed to the fact that charge storage in activated carbon (AC) is based on pure double-layer charging. In contrast, Faradaic charge-transfer reactions like those occurring in batteries are often connected with dimensional changes, which can affect the cycle life of these storage devices. Here we report the charge-induced height change of an AC electrode in an aprotic electrolyte solution, 1 mol/l (C2H5)4NBF4 (TEABF4) in acetonitrile. The results are compared with those obtained for a graphite electrode in the same electrolyte. For both electrodes, we observe an expansion/contraction of several percent for a potential window of ±2 V vs. the immersion potential (ip). For the EDLC electrode, significant expansion starts at about 1 V remote from the ip and hence is well within the normal EDLC operation range. For the graphite electrode, the height changes are unambiguously caused by intercalation/deintercalation of both anions and cations. The close analogies between the graphite and the EDLC electrode suggest that ion intercalation or insertion processes might play a major role for charge storage, self discharge, cyclability, and the voltage limitation of EDLCs.

Gas evolution and associated pressure build up in supercapacitor devices was monitored using commercial electrode rolls and 1 mol/l solutions of TEABF4 in acetonitrile (AN), propylene carbonate (PC), and γ-butyrolactone (GBL). During subsequent constant voltage periods (24 h) at 2.5 V, 2.75 V, 3.0 V and 3.25 V, significant gas evolution was observed in GBL already at 2.5 V. Under the same conditions, severe gas evolution in PC started only at 3.0 V, while in AN rate of gas evolution remained low even at 3.25 V. Keywords: Electrochemical double layer capacitor, Aprotic solvent, Electrolyte decomposition, Gas evolution

International audience; The main objective of this study is to investigate heat transfer inside and around a supercapacitor stack. The crossflow staggered arrangement with five rows was also studied for different transverse and longitudinal pitches. The mathematical model governing the problem is numerically solved in a three-dimensional computational domain using the commercial computational fluid dynamic package Fluent 6.3. The model is elliptic and takes into account orthotropic thermal conductivity in the active zone. The average Nusselt number and friction factor results of the present model are in good agreement with previously published experimental data of different authors for flow across tube banks. The results include evolution of the maximum temperature of each supercapacitor and contours of temperature distribution. For a definite ventilation power, the arrangement with the lowest spacing provides maximum cooling performance. Inside the supercapacitor the temperature distribution is not concentric around the axis and the maximum temperature is downstream. In the axial direction the temperature variation can be neglected.

For the first time gas evolution in a double-layer capacitor cell employing activated carbon electrodes and a solution of 1 M (C2H5)4NBF4 in propylene carbonate (PC) has been monitored by means of on-line mass spectrometry (DEMS). During slow scan voltammetry, CO2, propene, and H2 were detected as the major gaseous decomposition products. Probably propene and H2 are formed by solvent reduction at the negative electrode, while CO2 is formed by solvent oxidation at the positive electrode. A small amount of propene is already detected at a cell voltage of below 1 V. For all species the onset of significant gas evolution coincides roughly with the rise of faradaic currents. The increase of coulombic efficiency during subsequent cycling is attributed to passivation processes taking place at both electrodes. Keywords: On-line mass spectrometry, Head space analysis, Electrochemical double layer capacitor, Activated carbon, Propylene carbonate decomposition

A new method of supercapacitor online health diagnosis based on an electric model and on experimental results obtained by calendar ageing tests is proposed. The calendar ageing of supercapacitor is mainly related to thermal and voltage constraints from the application. This ageing causes the degradation of supercapacitor performances which can finally lead to failures. In vehicle applications, the goal is to avoid these failures which may have serious consequences and high non-quality costs. The device survey and ageing state detection are a priority for anticipating the problems. This paper presents the case of calendar ageing tests on supercapacitors made of activated carbon and organic electrolyte. The ageing is performed at constant temperature (65 °C) under a 2.9 Vdc bias voltage. These tests permit to highlight online and in real-time the measurable external indicators of ageing.

This paper presents supercapacitor ageing according to the voltage, the temperature and thermal shock tests. To investigate this effect, a test bench of accelerated supercapacitor calendar ageing was carried out. Experimental tests are realized at constant temperature when the supercapacitors are polarized at the maximum voltage. To quantify the supercapacitor ageing, the equivalent series resistance (ESR) and the equivalent capacitance (C) are measured using the DC and AC characterization. To lead to the determination of the supercapacitor lifetime, Arrhenius law, that describes the effect of temperature on the velocity of a chemical reaction, is considered. Finally, experimental results of supercapacitor thermal shock are presented.

International audience; A simple electrical model has been established to describe supercapacitor behaviour as a function of frequency, voltage and temperature for hybrid vehicle applications. The electrical model consists of 14 RLC elements, which have been determined from experimental data using electrochemical impedance spectroscopy (EIS) applied on a commercial supercapacitor.The frequency analysis has been extended for the first time to the millihertz range to take into account the leakage current and the charge redistribution on the electrode. Simulation and experimental results of supercapacitor charge and discharge have been compared and analysed. A good correlation between the model and the EIS results has been demonstrated from 1 mHz to 1 kHz, from −20 to 60 °C and from 0 to 2.5 V.