A research team from Tohoku University and the University of Cambridge has obtained a photovoltage of about 1 V under simulated sunlight from a "photo-rechargeable battery" that integrates photovoltaic power generation with energy storage. By stacking materials in layers to separate the pathways for charge, the design suppresses the recombination of photogenerated electrons and holes, allowing energy that was lost in conventional mixed electrodes to be extracted as voltage. The peer-reviewed paper, led by Assistant Professor Kohei Shimokawa of Tohoku University and colleagues, was published in Nano Letters on October 13, and the university announced the results the following day, October 14. However, the experiments that confirmed the storage performance used "photo-assisted charging," in which an external voltage is applied. Realizing a self-powered source that runs on sunlight alone will require linking the generated photovoltage to sustained charging.
What Causes Battery Capacity to Increase Under Light?
Even if a battery delivers more charge under illumination, that increase was not necessarily stored directly from light. Light can simply warm the battery, speeding up internal chemical reactions and making it possible to recover charge that was previously hard to extract. Evaluating a photo-rechargeable battery therefore requires distinguishing between a simple heating effect and the effect of light separating charges and driving the charging reaction.
This issue was examined in a 2023 study by Arvind Pujari and colleagues, who also took part in the present research. Using vanadium pentoxide and lithium manganese oxide as photoelectrodes and measuring under controlled temperature, they found that most of the capacity increase under illumination simulating standard sunlight could be explained by thermal effects. This finding was obtained for a specific material system, but it shows that a capacity increase after illumination alone cannot establish whether light actually charged the battery.
In a 2024 study, Pujari and colleagues combined a dye-based photoelectrode with negative electrodes of different potentials to examine the energy conditions required for electron transfer. For a photo-rechargeable battery to work, it is not enough to absorb light; the excited electrons must also sit at energy levels that can drive the battery's charging reaction. If the combination of potentials is wrong, side reactions can accelerate after full charge. In other words, between absorbing light and storing electricity there are several steps, including charge generation and separation and transfer to the electrodes at suitable potentials, and a loss at any of them degrades storage performance. The present study aims to reduce the loss from charge recombination by rearranging the materials.
A Layered Structure That Separates Electron and Hole Pathways
Using an aqueous zinc-ion battery as a model, the team compared three types of electrodes that differ in how the light-absorbing and storage materials are arranged. According to the detailed release published by Tohoku University, the representative photovoltages obtained for each structure are as follows.
| Electrode structure | Roles and arrangement of materials | Reported photovoltage (approx.) |
|---|---|---|
| Mixed | Dye-loaded titanium oxide, zinc–manganese oxide, and carbon blended together | ~0.01 V |
| Bifunctional | A cuprous oxide thin film handles both light absorption and storage | ~0.1 V |
| Layered | A charge-transport layer and a storage layer built into a dye-sensitized solar cell structure | ~1 V |
All three were evaluated with the same electrolyte and solar simulator, but the measured terminals and electrode states were not perfectly uniform. The bifunctional type also uses different materials, so the figures in the table cannot simply be used to rank the structures. By contrast, the mixed and layered types use the same storage active material and dye, so comparing them is important for isolating how material arrangement affects performance. The team explains the characteristics of each electrode, with these measurement conditions and material differences in mind, in Figure 1 and the main text of the original paper.
In the mixed type, the electrons excited by light and the holes, the states left behind when electrons leave, readily recombine before they can be used in the storage reaction. When this recombination occurs, the energy the charges carried is lost as heat. Making an electrode by blending material particles is easy to manufacture, but it is hard to secure paths that efficiently carry electrons and holes in different directions.
The layered type, by contrast, builds a storage function into a solid-state dye-sensitized solar cell and is designed so that electrons and holes travel along separate routes. Electrons generated when the dye absorbs light are extracted to a transparent conductive film via an electron-transport layer of titanium oxide. Holes, meanwhile, move through the conductive polymer PEDOT to the zinc–manganese oxide storage layer, where they drive the oxidation reaction. Separating the migration routes of electrons and holes reduces the chances of the two recombining.
Stacking materials in layers does not by itself suppress recombination, however. If the energy levels of the dye, the charge-transport layers, and the storage material do not match appropriately, charge movement is blocked along the way. The layered structure here takes into account not only the interface where photogenerated charges are separated, but also the path that carries them smoothly on to the storage reaction. Designing the entire flow from photoconversion to storage as a unit, through the choice and arrangement of materials, led to the photovoltage of about 1 V.
About 1 V of Photovoltage and the Conditions for Photo-Assisted Charging
The roughly 1 V obtained with the layered type is the open-circuit voltage, measured with no current flowing externally. The team evaluated a device with an active area of about 1.0 cm² under 1-sun illumination, which simulates standard sunlight. Before the electrolyte was introduced, an open-circuit voltage of about 1.0 V was obtained, but once electrolyte was added, the voltage in solar-cell mode fell to about 0.7 V. After pre-cycling the device in the dark, the photovoltage was about 0.6 V, confirming that the voltage obtained also depends on the state of the electrode.
The device uses three terminals, a photoelectrode, a zinc negative electrode, and an intermediate back electrode, connected differently depending on the measurement. To examine solar-cell performance, the photoelectrode and back electrode are connected; for photo-assisted charging, the photoelectrode and zinc negative electrode; and for discharge in the dark, the back electrode and zinc negative electrode. This configuration makes it possible to measure separately the voltage produced in the power-generating part and the voltage involved in driving the storage reaction.
| Item measured for the layered type | Reported value | Measurement conditions |
|---|---|---|
| Open-circuit voltage in solar-cell mode | ~1.0 V | 1-sun illumination, before electrolyte introduction |
| Photovoltage after electrolyte introduction | ~0.7 V | Solar-cell mode |
| Photovoltage after pre-cycling in the dark | ~0.6 V | Under illumination after charge–discharge |
| Charge stored by photo-assisted charging | ~60 mAh/g | 1.2 V applied externally, 5 hours of illumination |
It is important to note that the photovoltage of about 1 V and the storage capacity of about 60 mAh/g were obtained under different conditions. The former is the open-circuit voltage measured in solar-cell mode, while the latter is the capacity obtained through "photo-assisted charging," in which light is applied for 5 hours while 1.2 V is applied externally. The applied voltage is referenced to the Zn²⁺/Zn potential, corresponding to zinc deposition and dissolution, and capacity is expressed per unit mass of the storage active material. The figure of about 60 mAh/g therefore does not mean that the whole device stored 60 mAh of electricity.
The team also confirmed that electricity stored by photo-assisted charging can be extracted in the dark. In a separate measurement, after about 40 mAh/g was stored in roughly one hour of photo-assisted charging, about 34 mAh/g was recovered in the dark. Discharge was carried out at a constant current of 20 mA/g until the potential reached 1.2 V versus Zn²⁺/Zn. The coulombic efficiency, the ratio of discharged to charged electricity, was about 85%, and the authors also regard the starting discharge potential of about 1.6 V, above the externally applied voltage, as evidence that the photovoltage contributed to storage.
How much of the incident light energy was stored as chemical energy, on the other hand, must be assessed with a different metric from coulombic efficiency. Based on the output energy stored beyond the externally applied voltage, the authors calculated a solar-to-chemical energy conversion efficiency of about 0.015%. In other words, the roughly 85% is the ratio of charge exchanged in charging and discharging, whereas the roughly 0.015% is the efficiency of converting light energy into stored energy. The two differ in what they measure and what they mean, and a high coulombic efficiency cannot be read as the energy efficiency of the photo-rechargeable battery as a whole.
Challenges Remaining for Self-Powered Operation
In this layered photo-rechargeable battery, the photo-assisted charging current was observed to fall from about 1.0 mA immediately after illumination began to roughly half within a few minutes. The team explains that as charging proceeds, the potential of the storage layer rises and the energy difference for injecting holes shrinks. The photocurrent was also confirmed to recover upon discharge. This is a challenge that arises from integrating the part that absorbs light and generates charge with a storage part whose potential changes with its state of charge. Even if a photovoltage of about 1 V is obtained, a charging current that drops within a short time means that sufficient energy cannot be stored even under prolonged illumination. The original paper also cites this drop in current during photocharging as the main reason for the low conversion efficiency of about 0.015%, calling for materials and structures that can hand off charge efficiently even as charging advances.
The scope of the durability confirmed in the experiments should also be understood precisely. Over several weeks, the team carried out minimal current and voltage measurements and tracked the battery's open-circuit voltage to check whether the contact between the stacked power-generating and storage parts was maintained. This, however, was mainly a test of the static stability of the interface and did not demonstrate lifetime over repeated charge–discharge cycles. Zinc electrodes can also develop dendritic metal growth that causes short circuits, so how stably the device operates under real operating conditions remains a subject for future verification.
Tohoku University expects the technology to lead to fewer wires and intermediate components, with applications in small sensors and wearable devices. Practical use, however, will require showing how much energy can be stored by light alone, without an externally applied voltage, and confirming performance and durability over repeated charge–discharge cycles. The significance of this result is that it shows a layered structure that separates charge pathways can reduce voltage losses in a photo-rechargeable battery. If materials and structures that maintain current as charging proceeds are developed, integrated power sources that store the electricity they need while receiving light will come closer to practical use.