Energy harvesting from shadow-effect

Qian Zhang a, Qijie Liangb, Dilip Krishna Nandakumara, Sai Kishore Ravia, Hao Qua, Lakshmi Suresha, Xueping Zhanga, Yaoxin Zhanga, Lin Yanga, Andrew Thye Shen Weeb and Swee Ching Tan*a
aDepartment of Materials Science and Engineering, National University of Singapore, 9 Engineering Drive 1, Singapore 117574, Singapore. E-mail:
bDepartment of Physics, National University of Singapore, 2 Science Drive 3, Singapore 117551, Singapore

Received 15th March 2020 , Accepted 15th April 2020

First published on 15th April 2020

Shadows are everywhere. Not much engineering use has been found for shadows, and this ubiquitous effect is strenuously avoided in optoelectronic applications. In this work, we present a shadow-effect energy generator (SEG) that scavenges the illumination contrast that arises on the device from shadow castings, and generates a direct current, simply by placing a part of the generator in shadow. The shadow-effect mechanism is experimentally validated by Kelvin Probe Force Microscopy (KPFM). The SEG is capable of harvesting energy from illumination contrasts arising under weak ambient light. Without any optimization, our generator has a power density of 0.14 μW cm−2 under indoor conditions 0.001 sun, where shadows are persistent. Our SEG performs 200% better than that of commercial silicon solar cells under the effects of shadows. The harvested energy from our generator in the presence of shadows arising at a very low intensity (0.0025 sun) can drive an electronic watch (1.2 V). In addition, the SEG can serve as a self-powered sensor for monitoring moving objects by tracking the movement of shadows. With its cost-efficiency, simplicity and stability, our SEG offers a promising architecture to generate green energy from ambient conditions to power electronics, and as a part of a smart sensor systems, especially in buildings.

Broader context

Shadows are ubiquitous and are often considered undesirable in photovoltaic and optoelectronic applications. Contrary to this, we herein develop a novel method to harness useful energy from this redundant phenomenon. Our device, the shadow-effect energy generator, works on the principle of optical manipulation of the work function in metal thin film-semiconductor structures. Work function contrasts are introduced onto a metal thin film-semiconductor structure by partially blocking the light falling on the device thereby casting a shadow. Exposing the device completely to light does not result in power generation, thereby making this work unprecedented. With a primary focus on harnessing the illumination contrast that arises due to shadows, the developed device has a high resolution for capturing illumination contrasts, even under weak ambient light. This method brings a new perspective to sustainable energy generation and could potentially pave the way for the development of self-powered sensors for common and niche applications.


A shadow is a three-dimensional zone of relative darkness formed when the passage of light is blocked by an opaque object. Shadows are everywhere around us, such as shadows cast by the furniture in homes, shadows of people entering/exiting through doors, shadows of passengers on board a bus and even shadows of our moving arms or legs. However, not much engineering use has been derived from this ubiquitous optical effect. Photovoltaic devices generate electrical power when illuminated,1–3 and shadows cast on the active area of these devices are often deleterious to their performance.4,5 When photovoltaics are installed in huge arrays, the shading of one cell by another is often a problem and, to date, the most common solution is to provide sufficient spacing between them. This creates the problem of a greater wastage of area for solar irradiance that could have otherwise contributed to power generation. In this work, we present a simple device structure that makes of use shadows for power-generation and photosensing applications.

The concept of using the illumination contrast caused by shadows as an indirect source of power is unprecedented. Using just a steady source of light and an opaque object, shadows generated can be of immense use in optoelectronic applications. With the development of electronic devices such as smart phones, google glasses, and e-watches, efficient and continuous power supply is essential, which has made research on the development of wearable power sources that could harness ambient indoor light attractive.6–11 Batteries are currently the most widely used power supply despite the limitations of low power density, possible environmental-hazards, and fire hazards.12,13 Photovoltaic devices converting ambient light energy from the environment into electrical energy could be a sustainable alternative, although existing examples are not without limitations.14 The efficiency of many commercially available solar cells degrade significantly under indoor conditions where shadows are persistent.15–19 It is therefore highly desirable to develop a new platform to scavenge energy from both illumination and shadows associated with low light intensities to maximize the efficiency of energy harvesting.

Integrating the function of sensors with energy devices is the basis for the development of self-powered sensors.20,21 For example, sensors to monitor passing objects (such as people, animals, vehicles etc.) have applications in areas such as security surveillance, smart public transport systems, and for the Internet of Things.22–24 Conventional motion sensors invariably rely on cameras or RFID systems, which are either complex, bulky or expensive, and often need an external power source.25,26 To reduce energy usage and subsequent loss, high-performance self-powered sensors for monitoring passing objects are undoubtedly the better choice.27–29 Herein, a simple self-powered sensor that makes use of the electrical output triggered by illumination contrast is presented.30–32

In this work, we have developed a low cost, easy-to-fabricate shadow-effect energy generator (SEG). The SEG can not only convert illumination contrast by partial shadow castings to electricity, but also can serve as a self-powered proximity sensor to monitor passing objects. The operational principle of this new type of energy harvester which we call the ‘shadow-effect’ is investigated both theoretically and experimentally. The SEG performance under different conditions is systematically investigated. As an energy harvester, arrays of SEGs adhered to a flexible substrate can be attached to clothing and can power wearable electronic devices by continuously converting the shadows cast on the SEG into electricity. As a self-powered sensor, the SEG can be used to detect the movement of objects that pass by. By bringing a new perspective to sustainable energy conversion, the SEG will find numerous applications in future optoelectronics and smart sensors.

Results and discussion

Design and structure of the SEG

The schematic diagram of the SEG is depicted in Fig. 1a. Several SEG cells of the same size are arranged on a flexible and transparent polyethylene terephthalate (PET) film as the substrate as shown. The structure of the SEG cell is simple. It comprises of a Au thin film deposited on an n-type Si wafer. The surface morphology of the SEG cell shows the thickness of the Au film to be 15 nm (Fig. S1, ESI). The fabrication cost is lower than commercial silicon solar cells (C-Si cell) (Table S1, ESI).33,34 A picture of an SEG cell partly in shadow and partly illuminated by light is shown in the inset of Fig. 1a. Fig. 1b shows photos of 4 pieces of SEG cells affixed in parallel under different bending angles (30°, 60°, 90°, 120°, 150°, 180°, 210°, 240°). This shows the SEG array with the above-mentioned structure design has flexibility, which makes it practical for wearable applications.35,36
image file: d0ee00825g-f1.tif
Fig. 1 Structure design and indoor performance of the SEG. (a) The schematic illustration of the SEG (inset is the photographic image of an SEG cell partly in shadow.) (b) Photographic image of the flexible SEG under different bending angles. The output (c) Voc and (d) jsc of the shadow-effect generator (SEG) and commercial solar cells (C-Si cells) of the same size (4 × 2 cm2) under low light intensity (0.001 sun). The photographic images inserted in (c) show the SEG and C-Si cells under full illumination and half-in-shadow conditions. The table inserted in (d) shows the power density of SEG and C-Si cells under different conditions.

Effective indoor performance of SEG cell surpasses that of C-Si cell

The electrical output from the bent SEG under 1 sun intensity (1 kW m−2) is shown in Fig. S2 (ESI). An, unbent SEG (180°) achieved a short-circuit current (Isc) and open-circuit voltage (Voc) of about 1.2 mA and 500 mV, respectively. For different applications, external loads with variable resistance were connected to the generator.

The low light intensity performances of the SEG cell and C-Si cell (details in ESI Fig. S3 and Table S2) with a similar dimension of 4 cm × 2 cm in an indoor environment (0.001 sun) are also compared. The Voc and short-circuit current density (jsc) of the C-Si cell decreased significantly while that of the SEG cell increased when shifted from being fully illuminated to a half-in-shadow condition (Fig. 1c and d). There was however a negligible current flow through the SEG when the full device was completely exposed to shadows or light owing to non-uniformities of the Au coating on the n-Si as shown in Fig. S4 (ESI). In fact, there is more than a twelve-fold increase in Voc of the SEG cell after being put half-in shadow. The power density (P) can be expressed as:

image file: d0ee00825g-t1.tif(1)
where I is the Isc, V is the Voc, S is the total area of the cell (both the illuminated part and that in shadow). Using eqn (1), the power density of fully illuminated (Pfi) and half-in-shadow (Phi) of the SEG cell and C-Si cell are calculated and shown in the table inset in Fig. 1d. According to eqn (1), the Phi of the SEG cell under 0.001 sun is calculated to be 0.14 μW cm−2, which is higher than the C-Si cell (0.074 μW cm−2). This shows that the resolution of the SEG is much higher i.e., it can capture illumination contrasts even from low light intensities. The characteristics of SEG were also compared with other energy harvesting technologies as shown in the ESI, Table S3.

Working principle: shadow-effect

Fig. 2 shows the working principle of an SEG cell. The work function of Au (φm) is higher than that of the n-type Si (φS) before contact.37 Fig. 2a shows the energy band diagrams of Au and n-Si. In the shadow-contact regions of Au and n-Si, free electrons flow from n-Si to Au due to the difference in their work function. The electron transfer will continue until the Fermi levels of Au and n-Si are aligned. Under equilibrium, a depletion region will be established, leading to band bending. As φm > φS, there is a Schottky barrier (φb) formed at the Au/n-Si interface.
φb = φmφS (2)
Upon illumination, the absorption of photons with energy higher than that of the bandgap create electron–hole pairs via a band to band transition. The electron–hole pairs will separate as the free charge carriers move in different directions due to the internal electric field (electrons to n-Si, while holes go to metal). With continuous accumulation, the Schottky potential barrier is overcome and the electrons are injected into the Au. The work function of the Au changes to φm*. Illuminating one-half of the Au/n-Si, the potential between the illuminated part and shaded part is attributed to a work function contrast. Fig. 2b and Movie S1 (ESI) show the power generation process of the SEG cell when partly in shadow. In the initial state, all the SEG cell is illuminated, resulting in no charge transfer (Fig. 2b1). As previously mentioned, the work function of the illuminated Au (φm*) is lower than that in shadowed Au (φm), driving free electrons to flow from the illuminated part to the shadowed part of the Au film when the SEG cell is connected with an external circuit (Fig. 2b2). The illumination contrast is converted into electric power owing to the constant flow of carriers through the external circuit. When the illuminated part and the shadowed part are reversed, the direction of current in the external circuit is also reversed (Fig. 2b3). The front view photograph of one SEG cell with an electrode is shown in Fig. S5 (ESI). As shown in Fig. 2c, work function measurements were performed during two cycles of light ON and light OFF. With an illumination intensity of about 0.2 sun, the work function shift was recorded. The value of work function shift is defined as Δφ:
Δφ = φm* − φm (3)
image file: d0ee00825g-t2.tif(4)
UCPD is the contact potential difference, φtip is the work function of the Pt/Ir tip (5.5 eV), φsample is the work function of the sample.38 To confirm the value of the work function shift, the surface potential of the SEG cell is measured by KPFM when both fully illuminated and in shadow (Fig. S6, ESI). As calculated from eqn (3) and (4), the difference in work function between the illuminated and shadowed parts of the SEG cell is 0.13 eV (Fig. 2d). The current–voltage characteristics of the Au/n-Si cell and pure n-Si in shadow and illuminated conditions are studied. Fig. S7 (ESI) shows that the contact between Au and n-Si is a typical Schottky contact, which is consistent with previous analysis. Tuning work function by illumination contrast is potentially an interesting candidate for optical and optoelectronic applications.

image file: d0ee00825g-f2.tif
Fig. 2 Electricity-generation mechanism of the SEG cell. (a) Band structures of the Au and n-Si junction in shadow, under illumination and half-in-shadow. (b) Working mechanism of the SEG cell. (c) Surface potential map and (d) work function shift of the SEG cell with 15 nm Au film in shadow and under illumination.

Performance characterization of the SEG cell with different parameters

To comprehensively investigate the most effective metal thickness for the SEG cell, we analyzed the dependence of Isc and Voc on the thicknesses of the Au film. As shown in Fig. 3a (details in ESI, Fig. S8), SEG cells with different Au thicknesses work well under standard 1 sun illumination. The SEG cell with 15 nm Au film produces Isc and Voc of about 134 μA and 371 mV, respectively, which are larger than those of SEG cells with 30 nm, 60 nm, 120 nm and 240 nm Au coating. The reason attributed to this phenomenon is that the increase of the thickness of the Au film results in a decrease in transparency of the Au film (inset in Fig. 3a and details in Fig. S9, ESI). The effect of film thickness on Isc is mainly related to the intensity of light reaching the underlying Si surface because Si is the main source of photocarrier generation. As the thickness increases, the transparency of the film drops, thereby limiting the light reaching the Si and thus reducing the number of photocarriers generated. A similar trend is also observed in incident photon to electron conversion efficiency (IPCE) spectra of the SEG cell with 15 nm and 30 nm Au films (Fig. 3b). The maximum IPCE (at a wavelength ∼550 nm) of the 15 nm Au SEG cell is 42% higher than the 30 nm one. The details of the light transmittance efficiency with wavelength changes of Au films with 15 nm and 30 nm are also shown in the same figure.
image file: d0ee00825g-f3.tif
Fig. 3 Factors that influence the performance of the SEG cell. (a) Comparison of Voc and Isc generated with different Au film thicknesses (15 nm, 30 nm, 60 nm, 120 nm, 240 nm). Inset is the transmittance of light at different Au film thicknesses (the wavelength is ∼500 nm). (b) IPCE and transmittance curves of the SEG cell (15 nm and 30 nm). (c) Effect of the ratio of the shadowed area on the electric performance of the SEG cell. Inset is the photograph images of the SEG cell under different ratios of area in shadow (0%, 20%, 40%, 50%, 60%, 80%, 100%). (d) The schematic illustration and output performance of the SEG cell (8 × 2 cm2) half-in-shadow with electrodes connecting at different places. Ds and Di are the distances between connection points of the part in shadow and illuminated part from edges “s” and “i” of the SEG cell, respectively. (e) Comparison of Voc and Isc generated with different metals. Insets show the surface potential maps of the SEG cells with 15 nm Cu film and 15 nm Al film.

The ratio of the shadowed area (i.e., the shadowed area divided by total area) also affects the output of the SEG cell. As shown in Fig. 3c, when the whole SEG cell was under illumination (0%) or in shadow (100%), the Isc and Voc were significantly low. In particular, there is no electron flow on the surface when the SEG cell was in shadow (100%) while a small output was observed when the SEG cell was under illumination (0%). The small output is caused by illumination contrast of the SEG cell surface which is too hard to eliminate in measurements. The outputs (Isc or Voc) of the half-in-shadow (50%) SEG cell are all larger than those of 20%, 40%, 60% and 80% SEG cell in shadow, since the half-illuminated condition of the SEG cell provides optimum surface area for electron generation and electron collection. It was also found that the contact points from where the excited electrons are collected also plays a role in determining the device performance. We believe that this arises from the fact that the electric field that is formed at the interface drives the electrons from the illuminated side to the dark side and the strength of the field decreases as the distance increases from the interface as shown in Fig. 3d. The Isc of the SEG cell increased from 104 μA to 200 μA when the connection points in the shade and the illuminated parts became closer. There was not much change in the voltage of the device with respect to area of illumination or the position of the connection point as this is influenced only by the intensity of the illuminated light. The effect of sample size was studied as shown in the ESI, Fig. S10.

To investigate the performance dependency of the SEG cells on metal content and to reduce the material cost for future applications, SEG cells with Cu or Al films were also fabricated. As shown in Fig. 3e, the SEG cell with 15 nm Cu film generates a huge Voc (about 1000 mV) while the SEG cell with 15 nm Al film only generates 0.26 mV. However, the Isc generated by the SEG cell with Cu film is about 0.25 μA, which is much lower than that of the SEG cell with 15 nm Au film. Due to the poor conductivity and instability of the Al film, the Isc and Voc of the SEG cell with the 15 nm Al film could not reach as high as the SEG cell with 15 nm Au film when the light source was turned on or off (Fig. S11, ESI). The work function of the SEG cells with 15 nm Cu and 15 nm Al in shadow and under illumination were also studied by KPFM (Fig. 3e1 and e2). Under illumination, the shift in the work function of the SEG cells with 15 nm Cu and 15 nm Al calculated from eqn (3) and (4) are 0.11 eV and 0.03 eV (Fig. S12, ESI), respectively. Therefore, compared to the SEG cell with 15 nm Cu and Al, better performance was obtained from the SEG cell with 15 nm Au since the Au film has better stability and conductivity. To improve the output, the planar n-Si was replaced with textured n-Si in the SEG cell. As shown in Fig. S13 (ESI), the Isc of the SEG cell with textured n-Si was 2.5-fold larger than that obtained with planar n-Si, because a textured Si surface minimizes reflection losses.39 However, each side of the pyramid of textured n-Si receives light at different illumination intensities, which resulted in a lower Voc.

The light intensity of a shadow, which is related to the shape, position, and transmittance of the shadow-throwing object, is an important factor that influences the illumination contrast between the shaded part and illuminated part of the SEG cell. The first two factors mentioned, shape and position, were fixed in the experiments by keeping a rectangular piece of black paper right below the light source and right above the SEG cell (the black paper was held perpendicular to the light source and parallel to the SEG cell). The relationship between light intensity of the shadow and average transmittance of the shadow-throwing object is shown in Fig. 4a. A4 paper with a different number of layers was used to make shadows with different light intensity. With an increase in the number of A4 paper layers, light that passes through the paper is absorbed and scattered more (inset in Fig. 4a). The region shaded by increased layers of A4 paper becomes darker and the light intensity in the shadow region decreases. The light intensity of the shadow caused by the black paper, and the average transmittance, are 0 W m−2 and 0%, respectively.

image file: d0ee00825g-f4.tif
Fig. 4 Output characteristics and stability of the SEG cell. (a) Dependence of light intensity of the shadow on the average transmittance of the shadow-throwing object. Inset shows the transmittance spectra according to different shadow-throwing objects. (b) The electrical performance of the SEG cell when the illumination contrast (τ) between the shaded part and illuminated part of the SEG cell is changed. (c) Equivalent circuit model when an SEG cell is connected with a resistive load (Zload) and an amperemeter. (d) Output peak current (Iop) and output peak power of an SEG cell half-in-shadow with various external loads at 1 sun. (e) Dependence of the electrical performance of the SEG cell on light intensity. Inset is the Phi of the SEG cell at different light intensity. (f) Output performance of the SEG cell half-in-shadow under impulsive illumination (1 sun) over 1000 cycles. Insets show the detailed output curves over 50 cycles.

The illumination contrast between the shaded part and illuminated part of the SEG cell was calculated with two metrics, namely light intensity of the light source and light intensity of the shadow, proposed as follows:

image file: d0ee00825g-t3.tif(5)
where τ is the illumination contrast, Ll is the light intensity of the light source, Li is the light intensity of the illuminated part of the SEG cell, and Ls is the light intensity of the shadowed part of the SEG cell. Briefly, τ is a measure of Ls but an inverse measure of Li. In Fig. 4b, the Ll and Li are 1000 W m−2 while Ls is equal to the light intensity of the shadow. Improving τ from 96% to 100% leads to an improvement in output of the SEG cell (Voc is from 332 mV to 371 mV, Isc is from 131 μA to 134 μA). This result is consistent with Fig. 3c, in which an increase of τ from 0.1% to 100% leads to the increasing output of the SEG cell. In the case of Fig. 3c, the Ll and Ls were respectively fixed at 1000 W m−2 and 0 W m−2, while Li increased from 1 W m−2 to 1000 W m−2.

For different applications, external loads with variable resistance are connected to the generator. Considering this point, resistors were connected with the SEG to systematically study the optimum matching impedance. When the SEG is connected to a resistive load (Zload) and an amperemeter, the equivalent circuit model of the whole system is shown in Fig. 4c. The SEG is an open-circuit voltage source connected with an inherent impedance (ZSEG). To obtain a maximum power transfer point and estimate the ZSEG of SEG, we varied the external load. As shown in Fig. 4d, the load peak current decreases step by step as the Zload increases from 30 Ω to 6 MΩ under 1 sun. As a result, the maximum power transfer is 4.4 μW with the Zload of 120 Ω. The maximum power transfer happens when the Zload is the same as ZSEG. So we can conclude that the inherent impedance of SEG is 120 Ω.

To explore the ability of the SEG cell to capture illumination contrasts under low intensity light, we compared the Isc and Voc generated by the SEG cell under different light intensities (ranging from 1 W m−2 to 1 kW m−2). As shown in Fig. 4e, the Isc and Voc of the SEG cell are strongly dependent on the light intensity and show a decrease with decreasing light intensity. However, the Phi is less affected by the change in light intensity for this range from 1 W m−2 to 20 W m−2. After 20 W m−2, the Phi shifted faster (inset in Fig. 4e).

Because in real applications the SEG cells may be subject to impulsive light over a long period, we also investigated the stability of the device. The SEG cell when half-in-shadow maintained stable signals of Isc and Voc over 1000 cycles under the impulsive illumination of 1 sun. As shown in Fig. 4f, only slight shifts was observed for the Isc and Voc. The standard deviations of Isc and Voc are about 1.4 μA and 1.6 mV, respectively. From the details of the Isc and Voc values in the last 50 cycles (insets in Fig. 4f), there was no degradation in the performance of the SEG cell. This proved that the SEG cell could satisfy the cycling demands of practical applications.

Multifunctional applications

Other than harvesting energy from shadows present in the environment, the SEG cell can also serve as a self-powered sensor to detect movement of an object. As shown in Fig. 5a and Fig. S14 (ESI), a robot and remote-controlled car were used as objects passing by the SEG cells to generate shadows under an indoor light (0.001 sun). The self-powered sensor system can monitor the movement and record the number of times the robot has passed by, irrespective of its speed (Movie S2, ESI). Movie S3 (ESI) shows the self-powered sensor system tracking the motion of a remote-controlled car. As the distance between the SEG cells is fixed, speed and acceleration of the remote-controlled car can be computed. It is worthy to note that the size of the SEG cell has no impact on the detection. The voltage response to the robot passing by is shown in Fig. 5b. Owing to the shadow-effect, when an object passes by the top of the SEG cell, part of the SEG cell is in shadow while the other part is under illumination. Thus, the different work functions of the two parts of the SEG cell leads to a voltage generation, which drives the free electron transfer and thereby powers the LCD screen or causes the LED to glow. The response time of the SEG cell as a self-powered sensor is calculated to be 91 ms (which is limited by the sampling rate of our measurement system). The circuit diagram for the electronic interfaces of these two self-powered sensor systems are shown in the ESI, Fig. S15.
image file: d0ee00825g-f5.tif
Fig. 5 Multifunctional applications of the SEG. (a) Photographs of the individual SEG cells as a number of counter sensors for the robot and a position detector for the remote-controlled car. (b) The output voltage of the SEG cell as a sensor when the robot passes by. (c) Photographs of the electronic watch connected to 4 SEG cells. (d) Circuit diagram of the electronic watch powered by 4 SEG cells connected in series. (e) The output Voc and Isc of the 4 SEG cells. (f) Photographs of the SEG cells connected in parallel to detect the movement of a foot. Inset shows the light direction. (g) Photographs of a swinging arm on top of the SEG cells attached to clothing. Inset shows the light direction. (h) Electric output of the SEG cells connected in parallel under different conditions (left is for foot movement, right is for arm swinging).

The powering of an electronic watch (1.2 V) was demonstrated to establish the applicability of SEGs in wearable electronics. Fig. 5c shows 4 SEG cells connected in series powering an electronic watch. The circuit diagram is shown in Fig. 5d. When the light was turned on (1 sun intensity), the electronic watch was powered (Movie S4, ESI). Fig. 5e shows an SEG with 4 cells in series generating an Isc and Voc of 55 μA and 1.5 V, respectively. The ability of the SEG array to capture the illumination contrast arising from low intensity ambient light was also attempted as shown in the ESI, Movie S5. An electronic watch can be powered using 8 SEG cells in series which were exposed to partial shadows caused by a low intensity light (0.0025 sun), thereby conclusively proving that our device has better resolution for capturing light illumination contrasts.

Further, the SEG with 4 cells in parallel can also be used as a self-powered sensor, which increases output over a single SEG cell at the same time. Fig. 5f and h depict a situation where stepping on top of an SEG built into glass (0.844 sun) can generate an Isc and Voc of 630 μA and 260 mV for the first pulse, respectively. These peak Isc and Voc pulses triggered by the shadow-effect are much larger than that triggered by uneven light from a lamp (average current about 64 μA, average voltage about 135 mV). Comparing the output of an individual SEG cell with that from SEG cells connected in series and in parallel, we conclude that the output of the SEG can be multiplied by increasing its working dimension. When the foot steps away from the SEG glass, the Isc and Voc is lowered. Movie S6 (ESI) shows the recording of the entire procedure. The number of pulses of Voc and Isc generated during energy conversion counts the number of times people are passing by. As it has good stability performance (Fig. S16, ESI), the SEG could be deployed on the platform of bus stops, entrances to a playground or a mall, where a self-powered sensor could estimate the number of people entering or exiting the place. The SEG can also convert shadows arising from low intensity light energy into electric energy for powering wearable electronics. Fig. 5g and Movie S7 (ESI) demonstrate the swinging action performed by a person creating a shadow-effect onto the device attached to the person's clothing (0.672 sun), producing Isc and Voc of 420 μA and 160 mV, respectively, for the first pulse. Moreover, the SEG generated an alternating current (AC) when it is triggered by the swinging arm, whereas the foot step-up generated direct current (DC). The reason is the foot always creates a shadow on the same half of the SEG while the swinging arm shades both halves alternately.


In summary, we developed an SEG for harvesting energy from the illumination contrast caused by shadows that are cast on the device. This SEG also serves as a self-powered sensor for detecting passing objects. The difference in the work function between illuminated and shadowed zones of the SEG creates in-plane charge transport and induces remarkable electrical outputs. Notably, our half-in-shadow SEG cell experimentally achieved better performance than the C-Si cell under weak light intensity. Moreover, the SEG can be integrated with clothing for scavenging low intensity light energy to power wearable electronics. The SEG can find applications in many scenarios where shadows are created by humans or other moving objects. We successfully demonstrated the generated electrical energy from shadows arising from low light intensity, as low as 0.0025 sun, is capable of stably powering an electronic watch. Furthermore, our SEG is also a self-powered motion sensor with excellent stability for monitoring humans or other objects in motion. The response time of this kind of self-powered sensor is only 0.091 s. Our study presents a promising solution for harvesting energy caused by the illumination contrast from shadows, and may be a viable option to meet the increasing demands of electronic smart sensor systems, all of which could improve our daily lives.


Fabrication of the metal/n-Si cell

The silicon wafers are n-type with thickness of 530 μm. The n-Si wafer was cut into small pieces of designated size and then rinsed with acetone, isopropanol and water consecutively. After that, the silicon substrates were blow-dried with nitrogen gas, followed by Au deposition performed by thermal evaporation method. The typical evaporation rates are 0.5 A s−1 for films with a thickness of 15 and 30 nm, 1 A s−1 for 60, 120 and 240 nm.

Design of the integrated SEG

The size of SEG cells used for performance characterization are 4 cm × 2 cm. The SEG which is used to power the electronic watch is obtained with multi-SEG cells (4 cm × 2 cm, 15 nm Au film) connected in series. The voltage rating of the electronic watch is 1.2 V. The SEG with 4 cells in parallel was fabricated by arranging 4 pieces of Au/n-Si cells (the upper side of the length is 6 cm, the length of the ladder is 9 cm, the height is 2 cm, 15 nm Au film) on a flexible and transparent PET substrate. For measuring the output of the SEG, copper tapes were used as the electrodes. The whole SEG was protected by a transparent glass frame.


The voltage and current were all measured by a Keithley K2400. The measurements were repeated three times for each sample, and the experimental error was found to be within ca. 5%. The illumination is applied by a solar simulator (Newport, Oriel). In the performance characterization part, black paper was used to cover one half of the SEG to make a shadow. The thickness of this kind of black paper is 2 mm. The A4 paper (PaperOne, A4 70GSM) is commercially available. The paper shadows are cast by keeping the paper right below the light source, while right above the SEG cell. Furthermore, the paper is also held perpendicular to the light source and parallel to the SEG cell. The intensity of light is measured by a solar power meter IM-750. The transparency spectra were measured by a Varian CARY 100 Bio spectrometer. The morphology of the device was observed by field emission scanning electron microscopy (Zeiss GeminiSEM 500). The KPFM measurements were performed using a Dimension Icon (Bruker Nano Surfaces). Amplitude modulation KPFM was used to obtain a high signal-to-noise ratio as opposed to that of frequency modulation. All KPFM measurements were performed in dual pass mode. Work function measurements were performed with a Pt/Ir tip (5.5 eV) on the Au/n-Si, Cu/n-Si, and Al/n-Si surface. The IPCE spectra was recorded using a Zolix (SCS10-X150-DSSC), where a 250 W quartz tungsten halogen (QTH) lamp was used as the light source. The IPCE was defined as follows:
image file: d0ee00825g-t4.tif(6)
In the above formula, Pin and λ are the light energy and wavelength of the incident monochromatic light, respectively.

Conflicts of interest

The authors declare that they have no competing interests.


S. C. T. acknowledges the financial support from the Ministry of Education Academic Research Fund Tier 1 (R-284-000-161-114).

Notes and references

  1. B. O'Regan and M. Grätzel, Nature, 1991, 353, 737–740 CrossRef.
  2. S. Tembhurne, F. Nandjou and S. Haussener, Nat. Energy, 2019, 4, 399–407 CrossRef CAS.
  3. N. S. Lewis, Science, 2007, 315, 798–801 CrossRef CAS PubMed.
  4. J. Appelbaum and J. Bany, Sol. Energy, 1979, 23, 497–507 CrossRef.
  5. D. König and G. Ebest, Sol. Energy Mater. Sol. Cells, 2003, 75, 381–386 CrossRef.
  6. J. Wang, S. Li, F. Yi, Y. Zi, J. Lin, X. Wang, Y. Xu and Z. L. Wang, Nat. Commun., 2016, 7, 12744 CrossRef CAS PubMed.
  7. K. Song, J. H. Han, T. Lim, N. Kim, S. Shin, J. Kim, H. Choo, S. Jeong, Y. C. Kim, Z. L. Wang and J. Lee, Adv. Healthcare Mater., 2016, 5, 1572–1580 CrossRef CAS PubMed.
  8. G. M. Zhou, F. Li and H. M. Cheng, Energy Environ. Sci., 2014, 7, 1307–1338 RSC.
  9. J. Russo, W. Ray and M. S. Litz, Appl. Energy, 2017, 191, 10–21 CrossRef CAS.
  10. D. J. Garfield, N. J. Borys, S. M. Hamed, N. A. Torquato, C. A. Tajon, B. Tian, B. Shevitski, E. S. Barnard, Y. D. Suh, S. Aloni, J. B. Neaton, E. M. Chan, B. E. Cohen and P. J. Schuck, Nat. Photonics, 2018, 12, 402–407 CrossRef CAS.
  11. F. I. Alzakia, W. Sun, S. J. Pennycook and S. C. Tan, Introducing Normalized Centrifugation for a More Accurate Thermodynamic Analysis of Molybdenum Disulfide Dispersions: A Study on Mixed Solvents of Alcohols and Amines with Water, ACS Appl. Mater. Interfaces, 2019, 12(2), 3096–3103 CrossRef PubMed.
  12. S. Comello and S. Reichelstein, Nat. Commun., 2019, 10, 2038 CrossRef PubMed.
  13. B. Kang and G. Ceder, Nature, 2009, 458, 190 CrossRef CAS PubMed.
  14. C. Li, M. M. Islam, J. Moore, J. Sleppy, C. Morrison, K. Konstantinov, S. X. Dou, C. Renduchintala and J. Thomas, Nat. Commun., 2016, 7, 13319 CrossRef CAS PubMed.
  15. N. H. Reich, W. G. J. M. Van Sark, E. A. Alsema, R. W. Lof, R. E. I. Schropp, W. C. Sinke and W. C. Turkenburg, Sol. Energy Mater. Sol. Cells, 2009, 93, 1471–1481 CrossRef CAS.
  16. K. Shen, Q. Li, D. Wang, R. Yang, Y. Deng, M. Jeng and D. Wang, Sol. Energy Mater. Sol. Cells, 2016, 144, 472–480 CrossRef CAS.
  17. J. Burschka, N. Pellet, S. Moon, R. H. Baker, P. Gao, M. K. Nazeeruddin and M. Grätzel, Nature, 2013, 499, 316–319 CrossRef CAS PubMed.
  18. G. Li, V. Shrotriya, J. Huang, Y. Yao, T. Moriarty, K. Emery and Y. Yang, Nat. Mater., 2005, 4, 864–868 CrossRef CAS.
  19. X. Ma, Y. Mi, F. Zhang, Q. An, M. Zhang, Z. Hu, X. Lin, J. Zhang and W. Tang, Adv. Energy Mater., 2018, 8, 1702854 CrossRef.
  20. N. Horiuchi, Nat. Photonics, 2018, 12, 644 CrossRef CAS.
  21. Z. L. Wang, J. Chen and L. Lin, Energy Environ. Sci., 2015, 8, 2250–2282 RSC.
  22. G. Zhu, W. Q. Yang, T. J. Zhang, Q. S. Jing, J. Chen, Y. S. Zhou, P. Bai and Z. L. Wang, Nano Lett., 2014, 14, 3208–3213 CrossRef CAS PubMed.
  23. M. A. C. Stuart, W. T. S. Huck, J. Genzer, M. Muller, C. Ober, M. Stamm, G. B. Sukhorukov, I. Szleifer, V. V. Tsukruk, M. Urban, F. Winnik, S. Zauscher, I. Luzinov and S. Minko, Nat. Mater., 2010, 9, 101–113 CrossRef PubMed.
  24. Z. L. Wang, Mater. Today, 2017, 20, 74–82 CrossRef.
  25. Y. M. Song, Y. Xie, V. Malyarchuk, J. Xiao, I. Jung, K. J. Choi, Z. Liu, H. Park, C. Lu, R. H. Kim, R. Li, K. B. Crozier, Y. Huang and J. A. Rogers, Nature, 2013, 497, 95–99 CrossRef CAS PubMed.
  26. F. Meyer, T. Kropfreiter, J. L. Williams, R. A. Lau, F. Hlawatsch, P. Braca and M. Z. Win, Proc. IEEE, 2018, 106, 221–259 Search PubMed.
  27. Q. Liang, Z. Zhanga, X. Yan, Y. Gu, Y. Zhao, G. Zhang, S. Lu, Q. Liao and Y. Zhang, Nano Energy, 2015, 14, 209–216 CrossRef CAS.
  28. S. F. Leung, K. T. Ho, P. K. Kung, V. K. S. Hsiao, H. N. Alshareef, Z. L. Wang and J. H. He, Adv. Mater., 2018, 30, 8 CrossRef PubMed.
  29. R. Y. Liu, X. Kuang, J. N. Deng, Y. C. Wang, A. C. Wang, W. B. Ding, Y. C. Lai, J. Chen, P. H. Wang, Z. Q. Lin, H. J. Qi, B. Q. Sun and Z. L. Wang, Adv. Mater., 2018, 30, 1705195 CrossRef PubMed.
  30. Q. Zhang, Q. Liang, Q. Liao, M. Ma, F. Gao, X. Zhao, Y. Song, L. Song, X. Xun and Y. Zhang, Adv. Funct. Mater., 2018, 29, 1803117 CrossRef.
  31. Q. Zhang, Q. J. Liang, Q. L. Liao, F. Yi, X. Zheng, M. Y. Ma, F. F. Gao and Y. Zhang, Adv. Mater., 2017, 29, 1606703 CrossRef PubMed.
  32. Q. Liang, Q. Zhang, X. Yan, X. Liao, L. H. Han, F. Yi, M. Y. Ma and Y. Zhang, Adv. Mater., 2017, 29, 7 Search PubMed.
  33. C. L. Crago and E. Koegler, Energy Policy, 2018, 120, 481–491 CrossRef.
  34. F. Fertig, S. Nold, N. Wöhrle, J. Greulich, I. Hädrich, K. Krauß, M. Mittag, D. Biro, S. Rein and R. Preu, Prog. Photovoltaics, 2016, 24, 800–817 Search PubMed.
  35. J. A. Rogers, T. Someya and Y. Huang, Science, 2010, 327, 1603 CrossRef CAS PubMed.
  36. S. Xu, Y. Zhang, J. Cho, J. Lee, X. Huang, L. Jia, J. A. Fan, Y. Su, H. Zhang, H. Cheng, B. Lu, C. Yu, C. Chuang, T. Kim, T. Song, K. Shigeta, S. Kang, C. Dagdeviren, I. Petrov, P. V. Braun, Y. Huang, U. Paik and J. A. Rogers, Nat. Commun., 2013, 4, 1543 CrossRef PubMed.
  37. L. B. Luo, C. Xie, X. H. Wang, Y. Q. Yu, C. Y. Wu, H. Hu, K. Y. Zhou, X. W. Zhang and J. S. Jie, Nano Energy, 2014, 9, 112–120 CrossRef CAS.
  38. S. K. Ravi, W. Sun, D. K. Nandakumar, Y. Zhang and S. C. Tan, Sci. Adv., 2018, 4, eaao6050 CrossRef PubMed.
  39. V. K. Singh, S. K. Ravi, J. W. Ho, J. K. C. Wong, M. R. Jones and S. C. Tan, Adv. Funct. Mater., 2018, 28, 1703689 CrossRef.


Electronic supplementary information (ESI) available. See DOI: 10.1039/d0ee00825g
These authors contributed equally to this work.

This journal is © The Royal Society of Chemistry 2020