Nanometer-Scale Operando Imaging on Ionic and Electronic Transport through Polycrystalline Ceramic Solid Electrolytes

Abstract

Solid electrolyte (SE) is a promising candidate for enabling next generation high-density energy storage devices such as Li/chalcogenide or Li/air batteries. However, one of the major challenges in current SE is either low ion-conductivity or long-term stability: on one hand, SE such as organic polymer or glass ceramic has generally lower ion-conductivity than required by normal charging rate; on another hand, SE such as polycrystalline ceramic has instable problem encountering numerous cycling. Besides chemical and structural perspectives, the instability can also be initiated by highly inhomogeneous ionic conduction and electronic insulation through the SEs, where a point defect or weak point can develop to catastrophic device shunting by locally-enhanced ionic or electronic currents. Ionic and electronic conductivity in polycrystalline ceramic SE can be highly inhomogeneous, due to its structural nonuniformity in both micrometer and nanometer scales. Inexpensive processing such as pressing or liquid-casting can form micrometer-sized defects such as pin-holes. The structural and chemical nature of polycrystalline ceramic has the nm-sized inhomogeneity. All these structural and chemical inhomogeneity can cause significant nonuniform ionic and electronic transport during battery cycling, and further initiate device degradation and failure. However, characterization of the charge transport in SE has been limited in macroscopic scales such as electrochemical voltammetry or impedance spectroscopy. We here present on our recent development of nm-scale ionic and electronic transport imaging technique and the characterization results on two polycrystalline ceramic SEs of Li7P3S11(LPS) and Li1.4Ti1.6Al0.4(PO4)3(LTAP). The imaging technique was implemented by developing AFM-based scanning spreading microscopy (SSRM), which is essentially a half-cell with Li or Li-In alloy as one electrode and the SSRM probe as the other. This setup enables an operando measurement of charge transports, and has the following functions in order to meet the various SE characterization needs: (1) extremely wide range (Up to 1013orders of magnitude) of current sensitivity from 10-16 A (sub-fA) to 10-3 A (mA) for the low and high ionic/electronic conductivity, by using a logarithm current amplifier; (2) variable bias voltage polarity and amplitude to separate ion and electronic transport; (3) accurately position the probe at a point and hold stably for many hours by implementing a close-loop scanner control; (4) controllable probe/electrolyte contact area for balancing resolution and electrical signal strength; (5) controllable contact force (sub-nN to μN) for different hardness of the electrolytes from polymer to inorganic materials. The capability to separate the ionic and electronic conductions is based on the asymmetrical SSRM setup of probe/SE/Li-In. The electrical current under a negative sample bias (Vs) is the electronic conduction through the SE or an electronic leaking current, since there is no Li supply from the probe side (probe is highly-doped diamond-coated Si) and thus no Li ionic current flowing through the SE. Under a positive Vs, the electrical current is sum of electronic leaking and Li ionic currents flowing from the Li-In to the probe sides through the SE. The results on LPS shows huge asymmetrical current as large as 105orders of magnitude (several pA and 10-5 pA under positive and negative Vs), demonstrating the working SE with highly ionic conductivity and highly electronic resistivity (~1011 Ωcm). The asymmetry of LTAP is also large (103 orders) but less than LPS, with 10-1 PA ionic current and 10-4 pA electronic current. The ionic conduction on both the SEs is highly inhomogeneous in sub-μm scale with one to two orders of magnitude fluctuations, reflecting the complexity of SE ionic conduction by the local structural and chemical nonuniformities. For example, grain boundaries, grain orientation, local phase separation, and local chemical non-stoichiometry and aggregations can all have effects on the local ionic conduction in nm-μm scales. The local ionic fluctuation also changes with Vs, indicating the local energy levels of ion transport associated with the local structures. These local nm-scale imaging of ionic and electronic transports is expected to have broad impact on understanding and development of the SEs.