Photo-rechargeable batteries, which can efficiently convert and store solar energy into chemical energy within single devices, are essential to use sunlight effectively. Photo-extraction is the key point for the photo-rechargeable battery[1,2]. Photo-extraction has been attempted using various cations and electrode materials. However, the details of the reaction mechanism have not been understood well. In this study, we tried to analyze photo-extraction using an all-solid-state thin-film battery. We attempted to demonstrate that Li⁺ was extracted by light using operando synchrotron radiation X-ray diffractometry.

Anatase-type TiO₂:Nb(a-TiO₂) was selected as the cathode. a-TiO₂ is an n-type semiconductor that can progress the photo-extraction of lithium[3]. It is necessary for photo-charge that the irradiating light must reach the electrode. We applied LaSrAlO₄(001) as a substrate, whose band gap was wider than that of a-TiO₂. a-TiO₂:Nb(001)/CaRuO₃(001) films were epitaxially stacked on a LaSrAlO₄(001) substrate by pulsed laser deposition. Li₃PO₄ as a solid electrolyte and metallic Li of an anode were additionally stacked on a-TiO₂(001)/CaRuO₃(001) film by radio frequency magnetron sputtering and vacuum vapor deposition, respectively. In this way, We fabricated an all-solid-state thin-film battery composed of Li/Li₃PO₄/a-TiO₂:Nb/CaRuO₃/LaSrAlO₄(001).

The charging was conducted under constant-current + constant-voltage + constant-voltage with light irradiation (CC–CV–CV_photo) modes. The light was only applied during CV_photo charging using an LED with a central wavelength of 365 nm as the light source. Light irradiation significantly enhanced the charge capacities when the applied voltage during CV_photo was 3.0 V or higher. The capacities obtained during CV_photo charging closely matched the discharge capacities following CV_photo charging. Moreover, the discharge capacities within the plateau region around 1.75 V were also increased by light irradiation. Therefore, reversible photocharge-discharge was successfully achieved. Additionally, in the discharge curves following CC– CV–CV_photo charging, new discharge capacities were observed between 2.8 V and 1.9 V. Consequently, the charge/discharge behavior appears well-founded.

Lithium extraction was investigated using operando XRD. In these operando XRD measurements, the 004 peak of a-TiO₂:Nb was continuously monitored during charge/discharge cycles. Under dark, the 004 peak gradually shifted to a higher L-value with increasing intensity as the charging process advanced. These results indicate that Li⁺ was indeed extracted from a-Li_xTiO₂:Nb. Upon discharging, the 004 peak reverted to its original L-value and intensity, signifying the reversible lithium extraction in the absence of light. Under light irradiation, the 004 peak similarly shifted to a higher L-value than its pristine state. Furthermore, the full-width half maximum of the peak under light irradiation was narrower than that observed under dark conditions. These findings suggest that a-Li_xTiO₂:Nb at the charged state more closely resembled pure a-TiO₂:Nb under light irradiation compared to dark conditions. Consequently, Li⁺ was observed to extract more significantly from a-Li_xTiO₂:Nb under light irradiation than in the absence of light. After charging under light irradiation, the 004 peak reverted to its pristine state during discharging. This observation indicates that Li⁺, extracted in response to light, was reversibly inserted back into a-TiO₂:Nb. These results indicate that the extracted Li⁺ under light irradiation can be inserted into a-TiO₂:Nb reversibly.

We have successfully demonstrated the processes of lithium photo-extraction using a-TiO₂:Nb/ Li₃PO₄/Li in an all-solid-state thin-film battery and operando XRD. The use of solid electrolytes with a wide potential window is expected to accelerate research on photo-rechargeable batteries.