The Electrochemistry of Fe3O4/Polypyrrole Composite Electrodes in Lithium-Ion Cells: The Role of Polypyrrole in Capacity Retention

Two series of magnetite (Fe3O4) composite electrodes, one group with and one group without added carbon, containing varying quantities of polypyrrole (PPy), and a non-conductive polyvinylidene difluoride (PVDF) binder were constructed and then analyzed using electrochemical and spectroscopic techniques. Galvanostatic cycling and alternating current (AC) impedance measurements were used in tandem to measure delivered capacity, capacity retention, and the related impedance at various stages of discharge and charge. Further, the reversibility of Fe3O4 to iron metal (Fe) conversion observed during discharge was quantitatively assessed ex-situ using X-ray Absorption Spectroscopy (XAS). The Fe3O4 composite containing the largest weight fraction of PPy (20wt%) with added carbon demonstrated reduced irreversible capacity on initial cycles and improved cycling stability over 50 cycles, attributed to decreased reaction with the electrolyte in the presence of PPy. This study illustrated the beneficial role of PPy addition to Fe3O4 based electrodes was not strongly related to improved electrical conductivity, but rather to improved ion transport related to the formation of a more favorable surface electrolyte interphase (SEI). Introduction Li-ion battery (LIB) technology has played a critical role in the widespread adoption of a variety of portable electronic devices. However, due to the applications-driven nature of batteries, especially ambitious capacity and power requirements by potentially new devices, there is an increasing need for the basic understanding of the complicated Li-ion related electrochemistries associated with new electroactive materials. For example, one strategy to increase the capacity of electroactive materials is to use materials capable of multi-electron transfer (such as conversion reactions) resulting in high energy density materials.(1) Towards this end, magnetite, Fe3O4, is a promising nanoscale electroactive material with a high theoretical capacity (926 mAh/g) upon full conversion to Fe metal.(2) Thus, over the past several years, the number of published studies of Fe3O4 has been increasing due in part to its high energy density, as well as environmental friendliness and low cost.(3) Electroactive materials are incorporated into batteries through the fabrication of composite electrodes, where several additives may be used to enhance conductivity and to mechanically bind the components. Typically, an electrically conductive carbon additive (e.g. acetylene black or graphite) and a polymer binder (e.g. polyvinylidene difluoride (PVDF) or polytetrafluoroethylene (PTFE)) are mixed with the electroactive material to form a composite electrode.(3, 4) As mentioned above, future device applications will dictate increases in battery energy density, which may be accomplished by incorporating high capacity electroactive materials into composite electrodes where all the electroactive material is electrochemically accessible and the mass and volume contributions of passive components is minimized. For example, composite electrodes containing single walled carbon nanotubes have shown good capacity retention for up to hundreds of cycles with nanoscaled Fe3O4 and LixV3O8 electroactive materials with no additional carbon or binder required.(5-7) Electrically conducting polymers are intriguing as composite electrode additives since conceptually they can fulfill the roles of both carbon and binder.(8, 9) Prior studies have shown benefit of added conducting polymers in Fe3O4 composite electrodes containing carbon and other binders.(10, 11) Chemically driven oxidative polymerization of PPy (using iron chloride, FeCl3, as promoter) on 200 nm Fe3O4 nanoparticles resulted in a significant increase in reversible capacity (544 mAh/g at 1 A/g for 300 cycles) compared to PPy-free Fe3O4 nanoparticles.(12) Delivery of 652 mAh/g at 2 A/g over 500 cycles for 500 nm nanocages Fe3O4-PPy composites prepared via templated synthesis was reported.(13) Notably, both prior studies involved electrodes comprised of at least 10 wt% electrically conductive carbon additive and included a non-conductive polymer binder. To further probe the role of the electrically conductive polymer in the mesoscale (bulk) properties of composite electrodes, this study focuses on the Fe3O4 conversion reaction and its relationship to the electrically conductive polymer binder, PPy, including a detailed study of composite electrodes with nanoscale Fe3O4 and PPy. Two series of electrodes were explored where one group contained additional binders and electrically conductive carbon and a second group that did not. The preparation of Fe3O4-PPy composites has been reported using an additional oxidative initiator with an acidification step to polymerize the pyrrole.(14, 15) In contrast, our preparations used no additional oxidative initiator and Fe ions generated from dissolution of Fe3O4 in acidic solution acted as the oxidant. Our hypothesis was that the dissolved Fe ions may reside in close proximity to the Fe3O4 particles and thus the PPy generated would also be in close proximity to the Fe3O4 particles. Herein, we present the role of an electrically conducting polypyrrole generated in situ from dissolution of magnetite in acidic conditions and its effect on the conversion reaction in the PPy-Fe3O4 composite electrodes with and without added carbon and binder. The electronic state of the PPy with the Fe3O4 nanoparticles was investigated using Raman and the overall impedance of the Fe3O4-PPy composite was analyzed using Electrochemical Impedance Spectroscopy (EIS). After galvanostatic cycling, Fe3O4-PPy composite electrodes were characterized ex situ using X-ray Absorption Spectroscopy (XAS) to elucidate the roles of the polymer and carbon additives in terms of capacity retention upon cycling. To our knowledge, this is the first study where the PPy was generated directly from acid dissolution of Fe3O4 and the first study where the role of the conducting polymer PPy was investigated comprehensively by comparing carbon and binder containing and carbon free Fe3O4-PPy composites. The results of this study indicate that ion transport facilitated by the presence of PPy in Fe3O4 based electrodes was a more critical factor in determining the electrochemistry rather than improved electrical conductivity. The findings reported here are relevant to future studies of composite electrodes using conversion type electroactive materials where the goal of attaining higher energy density and longer cycle life may be furthered by improved ion transport through mediation of the surface electrolyte interphase. Experimental Materials synthesis and characterization Magnetite nanoparticles were synthesized using a co-precipitation method developed previously.(16-18) Briefly, stoichiometric amounts of iron(III) chloride hexahydrate (FeCl3•6H2O) and iron(II) chloride tetrahydrate (FeCl2•4H2O) were dissolved in H2O under N2(g) and added dropwise to aqueous triethylamine ((CH3CH2)3N). The product was washed and dried in vacuo. PPy was prepared by dissolving p-toluenesulfonic acid (HTos) monohydrate (CH3C6H4SO3H•H2O) in H2O. Under stirring, distilled pyrrole (C4H5N) and then iron(III) chloride hexahydrate (FeCl3•6H2O) to initiate oxidative polymerization were added and stirred for 18 hours. The Ppy was filtered and isolated as a black solid. The magnetite-polypyrrole (Fe3O4-PPy) composite material was prepared as follows. As synthesized magnetite was suspended in H2O. Then p-toluenesulfonic acid (HTos) monohydrate (CH3C6H4SO3H•H2O) was dissolved in H2O and added to the suspension and stirred for 15 hours. Distilled pyrrole (C4H5N) was added to the acidified magnetite suspension, ultrasonicated and stirred for 4 or 24 hrs yielding the Fe3O4-PPy product. The product was then washed with H2O and dried in vacuo. Crystallite size and sample purity were determined using X-ray powder diffraction (XRD) on a Rigaku Ultima-IV diffractometer with Cu Kα measured in a 2θ range from 5-90o with a step width of 0.04o and a scan rate of 5o/min. The sample purity was confirmed using the PDF card for magnetite (01-071-6336) and the peaks of interest (2θ = 30.2o) were fit using PeakFit version 4.12 by using a Person VII area function to find the full width at half max then applying the Scherrer equation to calculate crystallite size. The weight percent of polypyrrole was determined using thermogravimetric analysis (TGA) with a Thermal Advantage Release Q series 5.4.0 instrument using an alumina sample cup ramping from 25-800oC at a rate of 10oC/min under air. Raman spectroscopy was used to evaluate the polypyrrole coating on a Horiba Scientific XploRA instrument with a 532 nm laser between 100-2000 cm. A JEOL JEM 1400 microscope was used for Transmission electron microscope (TEM) imaging. Images were obtained at 120 kV voltage using a Gatan ORIUS SC1000B CCD camera. Inductively coupled plasma optical emission spectroscopy (ICP-OES) on a Thermo Scientific iCAP 6000 series spectrometer was used to evaluate Fe dissolution. The material was pressed into a 13mm in diameter pellet with the thicknesses ranging from 0.35-0.43 mm and placed between two electrodes under constant pressure to measure AC impedance on a Bio-Logic – Science Instruments EC-lab using a sine wave amplitude of 10 mV over a frequency range of 175 Hz – 1 MHz. The impedance data was fit using equivalent circuits using the software ZView version 3.4c. Ex situ XANES and EXAFS measurements of the Fe K-edge (7.11 keV) were conducted at the Cornell High Energy Synchrotron Source (CHESS) on beamline F3 with a Sagittal Si(111) crystal monochromator in transmission geometry with a Fe reference foil. Fe metal nanoparticles, FeO, α-Fe2O3, γ-Fe2O3, and Fe3O4 standards were measured with pristine samples of the materials. The data collected were aligned, merged, and normalized in Athena. Electrochemical Characterization Electrodes were prepared using Fe3O