In Operando Calorimetric Measurements for Activated Carbon Electrodes in Ionic Liquid Electrolytes under Large Potential Windows

Abstract

This study aims to investigate the effect of the potential window on heat generation in carbon‐based electrical double layer capacitors (EDLCs) with ionic‐liquid (IL)‐based electrolytes using in operando calorimetry. The EDLCs consisted of two identical activated‐carbon electrodes with either neat 1‐butyl‐1‐methylpyrrolidinium bis(trifluoromethane‐sulfonyl)imide ([Pyr14][TFSI]) electrolyte or 1.0 m [Pyr14][TFSI] in propylene carbonate (PC) as electrolyte. The instantaneous heat generation rate at each electrode was measured under galvanostatic cycling for different potential windows ranging from 1 to 4 V. First, the heat generation rates at the positive and negative electrodes differed significantly in neat IL owing to the differences in the ion sizes and diffusion coefficients. However, these differences were minimized when the IL was diluted in PC. Second, for EDLC in neat [Pyr14][TFSI] at high potential window (4 V), a pronounced endothermic peak was observed at the beginning of the charging step at the positive electrode owing to TFSI− intercalation in the activated carbon. On the other hand, for EDLC in 1.0 m [Pyr14][TFSI] in PC at potential window above 3 V, an endothermic peak was observed only at the negative electrode owing to the decomposition of PC. Third, for both neat and diluted [Pyr14][TFSI] electrolytes, the irreversible heat generation rate increased with increasing potential window and exceeded Joule heating. This was attributed to the effect of potential‐dependent charge redistribution resistance. A further increase in the irreversible heat generation rate was observed for the largest potential windows owing to the degradation of the PC solvent. Finally, for both types of electrolyte, the reversible heat generation rate increased with increasing potential window because of the increase in the amount of ion adsorbed/desorbed at the electrode/electrolyte interface.

Heat generation in electric double layer capacitors with neat and diluted ionic liquid electrolytes under large potential window between 5 and 80 °C

“…Furthermore, the instantaneous heat generation rateQ + (t) at the positive electrode in both devices was strictly (i) exothermic during charging due to ion adsorption and (ii) endothermic during discharging due to ion desorption. This was consistent with results obtained from numerical simulations [55,56] and from previous experimental studies on AC- [54,61]. Similarly, the fact that the endothermic dip inQ − (t) during charging was larger in Device 2 than in Device 1 could be due not only to the overscreening effect but also to the complete or partial endothermic desolvation of Pyr + 14 cations of their PC solvation shell as they enter the AC nanopores to form an EDL [11,34,102,103].…”

“…The ratioQ irr,T /Q J increased with increasing temperature and decreasing current in both Devices 1 and 2. Similar results were also observed in similar devices cycled at 20 • C under large potential windows ∆ψ s up to 4 V[61]…”

“…Finally, the endothermic dip at the negative electrode of Device 2 became noticeably larger with increasing temperature as a result of increasing PC decomposition caused by endothermic PC reduction reaction [61]. Note that, this dip dominated at the beginning of the charging step due to the competing and sharply increasing exothermic heat generation rate associated with cation adsorption throughout the charging step [61].…”

“…Here, m is the total mass loading of AC in both electrodes (m = 4.0 mg) while I(ψ s ) is the current measured at cell potential ψ s . A cell potential window between ψ s,min = 0 V and ψ s,max = 2.5 V was selected because it fell within the ESW of both electrolytes investigated at room temperature [61].…”

“…In fact, the ratio of the diffusion coefficients of Pyr . This can be attributed to a combination of phenomena namely (i) the charge redistribution in porous electrodes caused by higher ions desorption rate with increasing temperature and modeled as the leakage current flowing through the pore resistance [14,16,61,71,74] and (ii) the longer charging-discharging time t cd at low current (see Figure S3 in Supplementary Materials). In fact, the pore resistance is responsible for EDLC self-discharge and increased with increasing potential window [14,61,107,108].…”