Physical Delithiation of Epitaxial LiCoO2 Battery Cathodes as a Platform for Surface Electronic Structure Investigation

Abstract

We report a novel delithiation process for epitaxial thin films of LiCoO₂(001) cathodes using only physical methods, based on ion sputtering and annealing cycles. Preferential Li sputtering followed by annealing produces a surface layer with a Li molar fraction in the range 0.5 < x < 1, characterized by good crystalline quality. This delithiation procedure allows the unambiguous identification of the effects of Li extraction without chemical byproducts and experimental complications caused by electrolyte interaction with the LiCoO₂ surface. An analysis by X-ray photoelectron spectroscopy (XPS) and X-ray absorption spectroscopy (XAS) provides a detailed description of the delithiation process and the role of O and Co atoms in charge compensation. We observe the simultaneous formation of Co⁴⁺ ions and of holes localized near O atoms upon Li removal, while the surface shows a (2 × 1) reconstruction. The delithiation method described here can be applied to other crystalline battery elements and provide information on their properties that is otherwise difficult to obtain.

Keywords: absorption spectroscopy; lithium cobalt oxide; lithium ion batteries; photoemission spectroscopy; sputtering.

Figure 1

(a–d) Schemes of the delithiation process using Ne⁺ sputtering and annealing. Ne⁺ sputtering (a) induces preferential Li removal (b) in the ion penetration layer; some damage to the atomic structure is also produced due to the removal of Co and O atoms. (c) After annealing in O₂, the penetration layer recovers its crystalline quality, but it remains depleted in Li (d).

Figure 2

(a) Co 3p and Li 1s XPS for hv = 150 eV, corresponding to the pristine/stoichiometric (blue), Ne⁺ sputtered (black) and sputtered and annealed in O₂ (red) samples. The spectra are normalized to the intensity of Co 3p. The top row corresponds to samples of (001) orientation, and the bottom row corresponds to (104) orientation. LEED patterns are shown in (b) for the pristine and sputtered-annealed samples. All LEED patterns are taken at E = 95 eV. Note that both surfaces are reconstructed. (c) Scheme of the LEED patterns highlighting the (1 × 1) reciprocal cell (white) and the (2 × 1) domains (colored), together with simple schemes of the real space.

Figure 3

XPS spectra of Co 3p (a) and Li 1s (b) for a pristine sample (x = 1) and several delithiated samples (Li molar fractions x = 0.8, 0.7, and 0.3) highlighting the deconvolution of the peaks, hv = 150 eV. In both graphs, the intensity is normalized to the Co 3p peak to observe the decrease in Li ratio. Li 1s has been scaled by 0.08 for comparison with Co 3p.

Figure 4

XPS spectra of O 1s (hv = 670 eV) for different preparation conditions and core-level deconvolution as a function of Li molar fraction (x). Oxygen components are O1 (main LCO component), O2 and O4 (LCO surface), and O3 (surface contamination).

Figure 5

(a) XAS spectra of Co L_3,2-edge for different Li molar fractions taken at an incidence angle of 30°, with horizontal linear polarization (no changes for vertical polarization observed or after changing the incidence angle). (b) Close-up view of the L₃-edge. The line color denotes the Li molar fraction: black x = 1, green x = 0.6, red x = 0.5, blue x = 0.3.

Figure 6

(a) XAS spectra of O K-edge corresponding to different Li molar fractions taken at an incidence angle of 30° with horizontal polarization (no change for vertical polarization observed). (b) Close-up view of the main peak at low energy. The line color denotes the Li molar fraction: black x = 1, green x = 0.6, red x = 0.5, blue x = 0.3.