Energy fuels the human civilization. However, primary energy, including coal, nuclear fuel, oil, and natural gas, is limited. Among the renewable energy, wind and solar energy account for the largest proportion, but they are intermittent and contribute only about 3% of the global energy at present^[1]. Water, which covers about 71% of the Earth's surface, plays a key role in the energy cycle of the Earth. It absorbs nearly 70% of the solar radiation reaching the Earth's surface, which is around 60 petawatts (10¹⁵ W), three orders of magnitude higher than the global average energy consumption (about 18 terawatts·year in 2016). Nearly half of the energy adsorbed by water drives evaporation, which powers water cycle and evolves into raindrop energy, flow energy, falling water energy, wave energy, and energy of other various forms. Moreover, around 38 petawatts of the Sun radiation are adsorbed by atmosphere before reaching the Earth's surface, becoming thermal energy, which can also be utilized through water evaporation.

It has always been people's pursuit to extract useful energy from the water cycle, which can be traced back to the Spring and Autumn Period of China when water wheels were developed. With the birth of electromagnetism, hydropower technology has been developed since the late 19^th century. However, traditional hydro-technology solely utilizes the kinetic energy of water flow in a large amount, and is incompetent to harvest other forms of water energy. In the last few decades, various novel effects to harvest water energy emerged following the blooming development of nanomaterials and structures, supporting the emergence of hydrovoltaics^[2]. Hydrovoltaic effects generate electricity from the interaction of nanomaterials with flowing, condensing, waving, dropping, and even evaporating water, providing more versatile and broad ways to harvest water energy. It becomes exciting that both output power and duration of hydrovoltaic generators are significantly improved over the past few years with distinct perspectives and developments contributed by scientists from many fields such as materials science, mechanics, chemistry, biology, and so on. The hydrovoltaics is definitely boosted, and on the edge of giving rise to disruptive technologies for harvesting water energy. In this article, representative developments are highlighted that support the progress of hydrovoltaics during the last few years, and the future directions are envisioned.

Fig.1

Fig.1 Overview of energy source on the Earth and tremendous energy from water

Most of the hydrovoltaic effects convert mechanical kinetic energy of water, such as flowing or waving water, rain droplets, and feeding moisture, into electricity. Conventional streaming potential converts pressure gradient to electricity by ionic charge accumulation across insulating nano-channels with typical width narrower than the Debye length of the electrical double layer (EDL). The streaming potential can easily exceed 1 V at a pressure difference around 1 bar, but with currents only at the order of picoampere and an optimized efficiency around 3%^[3]. Another issue hindering its practical applications is that to use or measure the streaming current, a pair of nonpolarized electrodes (e.g., Ag/AgCl electrodes) has to be used to form a closed ion-electron circuit by including the Faradaic reaction.

The obstacles faced by streaming potential was resolved by utilizing the dynamic boundaries of moving EDL^[4-5]. Moving water droplets on graphene surface could give rise to an open-circuit voltage (OCV) of tens of millivolt, referred to as drawing potential, and a short-circuit current (SCC) at the order of microampere^[5]. This is attributed to the simultaneously charging and discharging of pseudocapacitance at the front and rear of the droplets. The drawing potential can be used to harvest raindrop energy. Similarly, waving energy can be harvested by placing a graphene sheet across waving water surface, thus referred to as waving potential^[4]. The OCV of waving potential can be as high as 0.1 V with an SCC more than 10 μ A. Since the structure and intensity of ion adsorption at the EDL is significantly dependent on surface charge, drawing and waving potential can be greatly enhanced by introducing polarized or charged substrates to the volt scale^[6-8]. Integrating drawing potential effect, solar cells working in rainy day are constructed^[9-10]. Quite recently, combining contact electrification and electrostatic induction effect, the dynamic formation and vanishing of EDL could give rise to OCV up to hundreds of volt and SCC close to milliampere. One dropping droplet is enough to power one hundred commercial LEDs, although the working time is limited to several miliseconds^[11]. Despite notable progress made in hydrovoltaic generators for harvesting mechanical energy, most of these provide pulsed or short term output voltage, thus hindering them from serving as sustainable energy source.

Fig.2

Fig.2 Overview of various technologies to harvest water energy and their corresponding fundamental theories, which supports the development of hydrovoltaics and a vision of harvesting hydrovoltaic energy from the whole water cycle

Hydrovoltaic technology is not limited to harvest mechanical energy but also environmental energy, such as osmotic energy and thermal energy. Conventional osmotic energy, i.e., blue energy, is available from water with different salinity, while classical technologies to harvest blue energy are rather inefficient. Two-dimensional materials with supra-nanometer pores has significantly improved its energy conversion efficiency due to its atomic thickness^[12]. Another kind of osmotic energy can be harvested from chemiosmosis process. It is well known that the concentration gradient of ions across membrane gives rise to an electrochemical gradient, manifesting in biological systems as membrane potential. The directional diffusion of ions down this electrochemical gradient, i.e., chemiosmosis process, would be harnessed for the generation of electricity in artificial systems with asymmetric structure or functional groups cyclically exposed to moisture. Exposing directionally reduced graphene oxides membrane in ambient air could provide a membrane potential around 0.2 V ~ 0.45 V spontaneously at a relative humidity from 25% to 85%^[13]. It is impressive that the device can work stably for at least 100 hours. Asymmetrically feeding moisture to polymer electrolyte membranes could even offer an OCV of up to 0.8 V and a remarkable high current density up to 0.1 mA·cm^-2[14].

The finding of water evaporation induced electricity sets a milestone to continuously harvest thermal ambient energy through water evaporation without any interruption^[15]. It was first reported in 2017 that evaporation process in ambient environment could persistently produce an OCV higher than 1 V in porous carbon black film due to the evaporation induced capillary flow. Coating carbon black film with charged molecular could facilely increase the OCV by two times and can even modify its polarity^[16]. Given its promising potential with regard to ubiquitous ambient thermal energy and abundant water distribution, intensive attention was paid in this direction over the last three years. Materials showing capability of harvesting water-evaporation energy have expanded from carbon materials to oxides and even wood and textile^[17-19], with output power exceeding 10 μ W for a single device of centimeter size^[20-21]. Introducing the concept of deliquescent chemical, devices can operate continuously without the supplementation of water, making it a candidate for portable energy source^[22].

Recently, it is demonstrated that simply exposing a small piece of thin protein nanowire film sheared from microbes to ambient humidity can generate a sustained voltage around 0.5 V^[23]. The protein nanowire device can maintain a stable DC voltage around 0.5 V for more than 2 months in ambient humidity around 60% without any external stimuli, and even can generate electricity in low humidity ~20%, comparable to a desert environment. The humidity generator based on protein nanowires does not require any external stimuli, thus is less restricted by location or environmental conditions than other approaches. The authors attributed the output voltage to a self-maintained moisture gradient in the film and the chemiosmosis of deionized protons, similar to that adapted by aforementioned moisture generator. However, harvesting energy from the ambient environment is much more difficult than maintaining a membrane potential. It is not clear at this stage what kind of energy the humidity generator harvests. Unlike the water evaporation through porous nanomaterials, where ambient thermal energy is consumed, the equilibrium adsorption-desorption exchange of water molecules at the air/solid interface does not require net environment energy input. The clarification of the energy source and underlying mechanism for these humidity generators requires further efforts.

Hydrovoltaics is a blooming field, and has shown promising potential to bring disruption for key societal question in energy. To achieve its potential, one has to go beyond the 'simple investigations' —for example, try different materials with similar device structures—and goes deeper into the fundamental physics. One of the keys lies in the understanding of solid-liquid interface and nano-fluid transport. The behavior of water, ions, and at the nanoscales departs in many aspects from the framework of classical electron theory^[24]. Although quantum mechanics can effectively describe the physical properties of solids at nanoscale, the behavior of liquids and solid-liquid interaction is beyond its ability. Reliable and effective theory methods are urgently desired. Recent theoretical work along this line reported enhanced solid-liquid interfacial evaporation through nanochannels, providing a clue to reveal a comprehensive mechanism of evaporation induced electricity^[25].

Another key is learning to make use of charge. As mentioned before, drawing potential, waving potential, and water-evaporation induced electricity all can be enhanced by applying charged surface^{[6, 11, 16]}. By electrostatic gating, streaming current in molecular sized slit-like channels can be increased by up to 20 times^[26]. The outstanding performance of protein film in ambient humidity might also be a manifestation of charged amino acid residues in protein. From the fundamental aspect, it is not surprising that hydrovoltaic effects show tight relationship with charge, because the structure of EDL, the solid-liquid interaction, and even the structure of water all can be tuned by charging. Now, the question is where and how to place the charge, and what is the guiding principles.

The hydrovoltaic effects have shed light on an unprecedented avenue for the whole chain energy capture from the Earth's water cycle. But its capability is not limited to harvesting energy. Another benefit from hydrovoltaic technology could be the hydrovoltaic ecology. In contrast to primary energy, the generation of hydrovoltaic energy does not introduce any thermal and carbon emission. On the contrary, hydrovoltaic technology, such as water-evaporation induced electricity, could consume low quality latent heat in ambient environment and convert it to high quality electricity. In Fig. 3(a), a tiny hydrovoltaic ecology system is illustrated. Water-evaporation through porous materials converts heat to electricity to light up a bulb, and simultaneously lowers the temperature. Besides, water condensation on the roof offers us purified water. With further efforts devoted into this area, hydrovoltaic energy can become our response to climate change in this century without scarifying our demand on energy.

Fig.3

Fig.3 Schematic illustration of hydrovoltaic ecology and hydrovoltaic intelligence. (a) Hydrovoltaic ecology. Water-evaporation through porous materials provides electricity to light up a bulb, and adsorbs heat from ambient to lower the temperature. Besides, water condensation on the roof offers purified water.(b) Hydrovoltaic intelligence. Hydrovoltaics reveals the interactions between ions, water, and electron in solid materials, which will guide our understanding on neurons, their connection, and the way our brain giving birth to intelligence.

Moreover, hydrovoltics could fuel brain intelligence. Water accounts for about 70% of the human body's weight, and 80% of the brain. There are abundant neurons, biological channels, and neuro transmitters in the human brain, which work based on ion channels to achieve electrical signal transmission, at least according to traditional neuroscience theory. The study of hydrovoltaic effects would definitely boosts our understanding on the ion transport in liquid-solid system, and further inspire people to understand the way our brains work, helping to develop brain-like artificial intelligence technology.