Behaviors of functional interfaces are crucial factors in the performance and safety of energy storage and conversion devices. Indeed, solid electrode-solid electrolyte interfacial impedance is now considered the main limiting factor in all-solid-state batteries rather than low ionic conductivity of the solid electrolyte. Here, we present a new approach to conducting in situ scanning transmission electron microscopy (STEM) coupled with electron energy loss spectroscopy (EELS) in order to uncover the unique interfacial phenomena related to lithium ion transport and its corresponding charge transfer. Our approach allowed quantitative spectroscopic characterization of a galvanostatically biased electrochemical system under in situ conditions. Using a LiCoO2/LiPON/Si thin film battery, an unexpected structurally disordered interfacial layer between LiCoO2 cathode and LiPON electrolyte was discovered to be inherent to this interface without cycling. During in situ charging, spectroscopic characterization revealed that this interfacial layer evolved to form highly oxidized Co ions species along with lithium oxide and lithium peroxide species. These findings suggest that the mechanism of interfacial impedance at the LiCoO2/LiPON interface is caused by chemical changes rather than space charge effects. Insights gained from this technique will shed light on important challenges of interfaces in all-solid-state energy storage and conversion systems and facilitate improved engineering of devices operated far from equilibrium.

Understanding the role of interfaces is important for improving the performance of all-solid-state lithium ion batteries. To study these interfaces, we present a novel approach for fabrication of electrochemically active nanobatteries using focused ion beams and their characterization by analytical electron microscopy. Morphological changes by scanning transmission electron microscopy imaging and correlated elemental concentration changes by electron energy loss spectroscopy mapping are presented. We provide first evidence of lithium accumulation at the anode/current collector (Si/Cu) and cathode/electrolyte (LixCoO2/LiPON) interfaces, which can be accounted for the irreversible capacity losses. Interdiffusion of elements at the Si/LiPON interface was also witnessed with a distinct contrast layer. These results highlight that the interfaces may limit the lithium transport significantly in solid-state batteries. Fabrication of electrochemically active nanobatteries also enables in situ electron microscopy observation of electrochemical phenomena in a variety of solid-state battery chemistries.

The effects of laser wavelength (355 nm and 532 nm) and laser pulse energy on the quantitative analysis of LiFePO₄ by atom probe tomography are considered. A systematic investigation of ultraviolet (UV, 355 nm) and green (532 nm) laser assisted field evaporation has revealed distinctly different behaviors. With the use of a UV laser, the major issue was identified as the preferential loss of oxygen (up to 10 at%) while other elements (Li, Fe and P) were observed to be close to nominal ratios. Lowering the laser energy per pulse to 1 pJ/pulse from 50 pJ/pulse increased the observed oxygen concentration to nearer its correct stoichiometry, which was also well correlated with systematically higher concentrations of (16)O₂(+) ions. Green laser assisted field evaporation led to the selective loss of Li (~33% deficiency) and a relatively minor O deficiency. The loss of Li is likely a result of selective dc evaporation of Li between or after laser pulses. Comparison of the UV and green laser data suggests that the green wavelength energy was absorbed less efficiently than the UV wavelength because of differences in absorption at 355 and 532 nm for LiFePO₄. Plotting of multihit events on Saxey plots also revealed a strong neutral O2 loss from molecular dissociation, but quantification of this loss was insufficient to account for the observed oxygen deficiency.

Improved methods for the flux growth of single crystals of the important battery material LiFePO 4 have been developed, allowing the facile preparation of single crystals up to 1 cm across with well-developed facets at relatively low temperatures. The structural characterization of these samples by both powder X-ray diffraction and single crystal diffraction (X-ray and neutron) indicates that the samples are typically stoichiometric with a very low concentration of Fe defects on the Li site, though crystals with larger concentrations of defects can be specifically grown using Fe-rich fluxes. These defects occur through the formation of a Fe-rich (Li 1−2x Fe x)FePO 4 partial solid solution, in contrast to the antisite defects more commonly discussed in the literature which would preserve the ideal LiFePO 4 stoichiometry. The LiFePO 4 defects are shown to be sarcopside-like (2 Li + → Fe 2+ + vacancy) based on compositions refined from single crystal diffraction data, the observed dependence of unit cell parameters on defect concentration, and their observed phase behavior (defects only appear in growths from fluxes which are Ferich relative to stoichiometric LiFePO 4). The distribution of defects has been studied by aberration corrected scanning transmission electron microscopy and was found to be highly inhomogenous, suggesting that defect-containing crystals may consist of endotaxial intergrowths of olivine LiFePO 4 and sarcopside Fe 3 (PO 4) 2 in a manner that minimizes the detrimental influence of Fe Li defects on the rate of Li-ion transport within crystallites.