Enhancing Hydrovoltaic Power Generation through Coupled Heat and Light-Driven Surface Charge Dynamics

TL;DR

This work introduces a spatially decoupled hydrovoltaic platform that separates the evaporating surface from the ion-transported bottom nanopillar array to independently control evaporation, ion transport, and solid–liquid interfacial chemistry. A transfer-capacitance–based equivalent circuit links three key regions—the evaporating top, the electrolyte, and the photovoltage-active bottom—revealing that capacitive photocharging and thermally modulated surface equilibria dominate energy conversion when interfaces are engineered. The device achieves Voc ≈ 1 V and Pmax ≈ 0.25 W/m^2 at 0.1 M salt, with performance further enhanced by silicon-doping and dielectric-shell choice (e.g., Al2O3 vs TiO2). These findings offer a predictive, design-centered framework for optimizing EDHV devices to harvest ambient heat and sunlight across varying salinities and environmental conditions.

Abstract

Harnessing natural evaporation offers a sustainable and untapped pathway for next-generation energy technologies. Here, we present a unified physical and experimental framework for evaporation-driven hydrovoltaic (EDHV) systems that decouples and systematically controls the key interfacial processes underlying electricity generation from ambient heat and sunlight. By introducing an intermediate ion-conducting layer, we spatially and functionally separate the evaporative top interface from the silicon-dielectric nanopillar array at the bottom, enabling independent modulation of evaporation, ion transport, and interfacial chemical equilibrium. This decoupling strategy enhances device performance, facilitating the study of thermal and photo-induced charge generation, and improving ion migration and electricity generation. We develop a predictive equivalent electrical circuit model that captures the coupling between these processes through a transfer capacitance term, which we derive analytically as a function of geometric and material parameters. Our study reveals that capacitive photocharging and thermally modulated surface equilibria, rather than Faradaic or photothermal effects, are the dominant drivers of energy conversion when interfacial environments are adequately engineered. The device achieves a state-of-the-art open-circuit voltage of 1 V and a peak power density of 0.25 W/m2 at a 0.1 M salt concentration. Strategic variation of doping reveals that increasing silicon doping enhances voltage by 28% and power by 1.6 times, while switching the dielectric shell from TiO2 to Al2O3 boosts voltage (power) by up to 1.9 times (3.6 times). These findings offer insights for enhancing EDHV devices and suggest strategies that consider environmental conditions, water salinity, and material engineering to better harness waste heat and sunlight.

Figures (7)

• Figure 1: Evaporation-driven Hydrovoltaic Device Architecture, Mechanisms, and Materials.A) Schematic representation of the hydrovoltaic device featuring a top evaporating electrode surface and a bottom array of SiNPs immersed in water. The top and bottom components do not physically contact but are electrochemically connected through the water. The inset displays the three effects contributing to the device's performance. i) A side view of the evaporating surface with the liquid meniscus and the thermal gradient across the liquid layer. ii) An intermediate electrolyte layer and thermally tuned chemical equilibrium at the bottom nanostructures, resulting in a higher surface charge at increased temperatures. iii) photoactive nanostructure-electrolyte interface depicting the enhanced surface charge under irradiation due to electron-hole pair generation. B) Scanning electron microscopy (SEM) image of the SiNPs array. (left) Top view, and (right) cross-sectional view.
• Figure 2: (continued)C) Scanning transmission electron microscopy (STEM) image of a single NP. The cross-sectional cut of a single NP reveals the presence of a silicon core and Al_2O_3 shell. Intensity mapping was performed in the rectangular region. TEM-EDX image of the NP displaying the elemental maps of aluminum, silicon, and oxygen. E) (Top) A detailed view of the interface potential and free space charge in the silicon and electrolyte mediated by the oxide layer. The total potential is the sum of the potential in the space charge layer of silicon, the oxide layer, and the double layer of the electrolyte (bottom), as well as the free charge profile in the respective regions and the corresponding equivalent capacitance.
• Figure 3: Role of Temperature for different coatings and salinity levels.A) The time trace of the measured open circuit voltage at ambient temperature is presented when the silicon surface temperature is increased and then allowed to cool down. The inset illustrates a qualitative increase in the chemical potential difference for a single NP wetted with electrolyte. B) The time trace of the silicon surface temperature when the heater is turned on and turned off once the maximum is reached. The right axis shows the corresponding temperature difference between the bottom silicon surface and the top electrode (at the end of the liquid meniscus). C) The voltage-temperature profile of the two devices was measured using the same top electrode at varying salinities. Each device consists of an identical silicon core and a dielectric shell made of Al_2O_3 and TiO_2: 1 mM (solid lines) and 100 mM (dashed lines). D) (Top)Schematic representation of the core-shell nanopillars with Al_2O_3 (pink) and TiO_2 (green) shells. (bottom) The slope of the $\mathrm{V_{oc}}$ -temperature curves for the linear regime (up to a 20 K increase in temperature) and the corresponding estimated values of $\Delta_H$.
• Figure 4: Role of Light for different coatings and salinity levels.A) Time trace of the measured open-circuit voltage and capacitance at 1 mM KCl for two devices with different dielectric shells and low N-doping ($\mathrm{1-20~ \Omega.cm}$). The test was conducted under ambient conditions and $\mathrm{100~ mWcm^{-2}}$ solar illumination (yellow-shaded region). B) Measured photovoltage under $\mathrm{100~ mWcm^{-2}}$ intensity for samples with Al_2O_3 and TiO_2 shells (same low N-doped silicon core) at different concentrations of KCl. C) Measured open circuit voltage in the dark and the corresponding photovoltage measured as the device is heated and then allowed to cool down. D) Steady-state capacitance (blue bars) values for two devices with Al_2O_3 shell, but with different doping of silicon core (low N-doped: ($\mathrm{1-20~ \Omega.cm}$) and high N-doped: ($\mathrm{< 0.05~ \Omega .cm}$)) at 1 mM KCl. The blue and red shaded region is measured at a surface temperature of $T_{\text{s}}= T_{\text{ambient}} = 25 ^\circ \text{C}$ and $T_{\text{s}} = 70 ^\circ \text{C}$, respectively.
• Figure 5: Effect of wavelength and intensity of the irradiation.A) Band diagram of the semiconductor-metal oxide-electrolyte interface. (left) The liquid-solid interface is formed, but equilibrium has not yet been reached. (middle) Equilibrium has been established, resulting in Fermi-level alignment across the interface. The red-white gradient represents the filling of surface states leading to the Fermi level pinning. (right) Under illumination, the Fermi-level splitting occurs. The difference in the quasi-Fermi levels of electrons and holes equals the measured photovoltage. B) Measured photovoltage for different monochromatic incidence as a function of light intensity, plotted in logarithmic scale. I_0 is equal to $\mathrm{100~ mWcm^{-2}}$. The points corresponding to solar are the measured photovoltage for the full solar spectrum. The experimental data points are shown in circles, while the line is the linear fit. C) The black curve is the measured photovoltage for different at $\mathrm{7~ mWcm^{-2}}$. The red curve is the estimated $\beta_{\lambda}$ from the linear fit. The blue curve is the value of $\beta_{\lambda}$, the normalized absorptance $\alpha_{\lambda}$ of the sample. The subscripts $\lambda$ signify wavelength-dependent physical quantities.