Electrochemical Study of Microporous Polymer Electrolytes Based on Poly(vinylidene fluoride) Copolymer

• Main Authors: Jin-Yiing Song, 宋金穎
• Other Authors: Yung-Yun Wang
• Format: Others
• Language: zh-TW
• Published: 2000
• Online Access: http://ndltd.ncl.edu.tw/handle/83004252533597681090

博士 === 國立清華大學 === 化學工程學系 === 88 === This dissertation presents the ionic transport property, microstruc-ture, impact of different preparation schemes, and interfacial property of microporous gel electrolytes based on poly(vinylidene fluo-ride-co-hexafluoropropylene) (PVDF/HFP) for lithium-ion batteries by scanning electron microscopy (SEM) ,nitrogen adsorption/desorption method (BET), differential scanning calorimetry (DSC), ac-impedance spectroscopy, and certain standard electrochemical techniques. The dif-ferences between the PVDF/HFP-based electrolytes and conventional battery separators were clearly pointed out. In addition, all complex im-pedance spectra pertaining to the measurement of ionic conductivity and interfacial property were systematically introduced and discussed as well. Depending upon the absorbed liquid electrolytes, the ambi-ent-temperature conductivities of PVDF/HFP-based electrolytes ranged from 1.4 to 2.1 mS/cm, approximately half to one-third of those corre-sponding liquid electrolytes, whereas the transference numbers remained at around 0.75. The conductivities of PVDF/HFP-based electrolytes were even, in most cases, superior to those of Celgard® 2400 separators. In fact, the latter often suffers from poor wetting with many liquid electrolytes. Besides, weak interaction between polyolefin and polar electrolytes also facilitated the escape of electrolyte components from separators at ele-vated temperature. In contrary, PVDF/HFP-based electrolytes were free from these deficiencies, possibly due to the swelling between PVDF/HFP and liquid electrolytes. Nonetheless, the average activation energy of con-duction for PVDF/HFP-based electrolytes (ca. 17 kJ/mol) was larger than those of liquid electrolytes (£ 8 kJ/mol) and battery separators (ca. 11 kJ/mol) but smaller than that of polyacrylonitrile (PAN)-based gel electrolytes, implying that the porous nature of the membranes indeed enhanced the ionic mobility whereas the polymer still hindered the ionic transport to certain extent. PVDF/HFP-based porous membranes were covered on one side with silica particles that were brought to the surface of the membrane by the casting solvent during preparation. Basically, the porous structure of the membranes were very uniform and the pore shape was kind of cylin-drical. The average pore diameter is about 20 nm and most pores are no larger than 40 nm. On the other hand, the maximum pore dimension at surface of Celgard® 2400 is as large as 60 ´ 240 nm. Except for the ex-pense of slightly larger pore size, the membranes would have better uni-formity, higher porosity, more electrolyte uptake, and hence higher ionic conductivity if plasticizers with higher boiling point and viscosity were used during preparation. Both the plasticization/extraction method (i.e. the Bellcore process) and controlled-evaporation are capable of forming microporous mem-branes. The membranes prepared by controlled evaporation, which had an average pore diameter of ca. 30 nm, could absorb almost twice as that of the original weight of a dry membrane. As a result, controlled evaporation method is qualified to replace the original plasticization/extraction method. Simple solvent evaporation method, on the contrary, often leads to non-porous membranes and thus having inferior electrolyte-uptake ca-pability. Interfacial impedance between PVDF/HFP-based electrolytes and lithium metal, although still growing, could attain delicate kinetic equi-librium upon prolonged storage. Graphite, on the other hand, was found to remain inert toward PVDF/HFP membranes. Its interfacial impedance did not vary with increasing storage time and with different lithium salts. However, it was found that the lithium-side impedance of a Li/C half cell would often prevent us from obtaining practical information of the carbon electrode and thus a three-electrode impedance study was called for under this condition. We found, in particular, that an inductive loop would ap-pear in the low-frequency region of the impedance spectrum of a carbon electrode once after the first lithium-intercalation step, probably implying that an adsorption/desorption phenomenon might exist at the interface. Moreover, another inductive effect arising from the connecting leads would also appear at high-frequency region. Finally, the positive elec-trode was found to be the major source of cell impedance and it would increase with increasing cycle number.