Optical Wireless Power Transmission under Deep Seawater Using GaInP Solar Cells

Abstract

Optical wireless power transmission (OWPT) attracts attention because it enables wireless power transfer over longer distances than current wireless power transfer methods, irradiating laser light to a light-receiving element. In this study, an OWPT system was investigated under water and deep seawater using visible lasers with low optical absorption loss in water. Three laser beams (450 nm, 532 nm, and 635 nm) were transmitted through 30 cm, 60 cm, and 90 cm long tanks filled with tap water and deep seawater and were irradiated to 1.0 × 1.0 cm² GaInP solar cells. The light reaching rate (η_op) of laser light and the system efficiency (η_sys) of the system (excluding the laser efficiency) were investigated. GaInP solar cells showed photo-electric conversion efficiencies of 30.6%, 40.3%, and 39.3% for 450 nm, 532 nm, and 635 nm irradiations, respectively. As a result, a 532 nm laser through a 90 cm water tank in tap water showed a 78.4% η_op and a 30.8% η_sys. Under deep seawater, a 532 nm laser through a 90 cm tank exhibited a 58.3% η_op and a 23.5% η_sys. A 532 nm green laser showed a higher efficiency than the other 450 nm and 635 nm lasers in this underwater system using GaInP solar cells.

1. Introduction

With the development of IoT technology, various devices and equipment are widely connected to the internet. Therefore, wireless power supply is also expected to become available. The electromagnetic induction method and microwave methods have been put to practical use in wireless power transmission (WPT) [1,2]. The electromagnetic induction method is highly efficient, but the power supply distance is only a few centimeters or less. The microwave method can supply power over longer distances than the electromagnetic induction method and can be used over several meters. However, the microwave system presents the risk of electromagnetic noise interference because of the use of high-power electromagnetic waves.

Optical wireless power transfer (OWPT) has attracted attention [3,4], and it can convert light energy into electric power by irradiating a laser beam to a light receiver located at a long distance away. Therefore, OWPT is being considered for application as a new energy supply method for moving objects such as automobiles and transport robots. In recent years, high photoelectric conversion efficiency exceeding 40% has been reported for irradiating III-V solar cells with near-infrared light of 800–1600 nm [5,6,7,8,9]; 74.7% for GaAs 3-junction cells at 150 K irradiated with 808 nm laser lights [5]; 68.9% for GaAs/AlGaAs 1J cells under 11.4 Wcm⁻²/858 nm laser lights [6]; 66.5% from GaAs 12-J cells under about 65 Wcm⁻²/841 nm laser lights [7]; 52.8% for InGaAsP 1-J cells under 2 Wcm⁻²/1310 nm [8]; and 49% for AlGaAsSb 1-J cells under 1680 nm [9].

With the recent confirmation of the existence of marine resources on the seafloor, the OWPT system has also been researched in anticipation of under water and seawater exploration [10,11,12,13]. Miyamoto et al. reported that a GaAs solar cell was irradiated with 450 nm laser light at 6 W through a fly-eye lens in a 90 cm long water tank, and the power was converted to 0.755 W [10]. Kim reported an attenuation of 3.0 dB/m in OWPT in seawater and showed that power can be transmitted wirelessly with a transmission efficiency of 2% up to an underwater optical path length of 1 m [11]. Putra and Maruyama calculated a total system efficiency (including the laser efficiency) of 13.1% at a wavelength 570 nm for a 1 m depth of water, 10.2% at 480 nm for a 10 m depth of water, and 7.3% and 5.5% at around 400 nm wavelength for 50 m and 100 m depths of water [12]. Jin et al. have demonstrated that the on-board batteries of a commercial AUV were charged by transmitting 225 W optical output power from blue LEDs under seawater at a 10 m distance [13]. These were researched using visible light due to the low attenuation coefficient under seawater [14].

The current underwater probes are mainly powered by a wired connection, which limits their activity time and range. The use of OWPT solves these problems and enables efficient exploration. However, OWPT under seawater requires overcoming some challenges, such as low optical transparency and high light scattering loss in seawater. Attenuation coefficients have been reported for lake water [15,16,17,18,19], pure water [20], and seawater in the Pacific Ocean and Mediterranean Sea [21].

Table 1. Attenuation coefficients for various sea areas.

Ref.		$\mathbf{Attenuation}\mathbf{Coefficient}({\mathbf{m}}^{-1}$)
		450 nm	530 nm	635 nm
[20]	Pure water	0.00922	0.0434	0.3012
[21]	Pacific Ocean	0.023	0.052	0.33
[21]	Mediterranean Sea	0.032	0.056	0.319

shows representative examples of the reported attenuation coefficients.

The attenuation coefficient varies by sea area location due to plankton and suspended solids. As shown in Table 1, under water and seawater, the attenuation coefficient is lowest in the order of 450 nm, 532 nm, and 635 nm. However, as shown in Figure 1, the attenuation rate of the 532 nm green laser was the lowest, although there were some differences depending on the seawater sampling area [22,23].

It has also been reported that the light reaching rate (η_op) varies with the season in gulf waters, as shown in Figure 2 [24]. During winter in Tokyo Bay, the light reaching rate η_op of the 532 nm laser was the highest; however, the η_op of the 635 nm laser in summer was the highest due to the increase in the proliferation of phytoplankton because of the temperature rise in seawater and the increased sunshine hours.

As explained above, the light reaching rate varies depending on location [8,9,10,11,12,13,14], season [24], and depth of seawater collection [12]. In this study, deep seawater and tap water were prepared to reproduce the seafloor exploration environment.

Since visible light has a high reaching rate in water [14], it is necessary to use solar cells that absorb visible light. GaInP solar cells have achieved high conversion efficiencies of over 40% in the visible wavelength range [3,25,26,27].

Steinsiek reported a photo-electric conversion efficiency of 40% in an OWPT system using a 532 nm laser and GaInP solar cells (a length of 25 mm) under 0.27 W/cm² irradiation [3]. Komuro et al. irradiated GaInP photovoltaic devices (2.4 mm × 2.4 mm) with a 638 nm laser at 17 W/cm², and they reported a photoelectric conversion efficiency of 43.0% [25]. Kurooka et al. irradiated GaInP solar cells (2.4 mm × 2.4 mm) with a 638 nm laser at 1.5 W/cm², and they reported a photoelectric conversion efficiency of 46.7% [27]. Wang et al. irradiated GaInP solar cells (2.4 mm × 2.4 mm) with a 635 nm laser at 53.5 W/cm⁻², and they reported a photoelectric conversion efficiency of 37.2% [26].

This study prepared relatively large GaInP solar cells (1.0 cm × 1.0 cm) compared with those cells [25,26,27] to bring the experiment closer to practical application. The photoelectric conversion efficiency, laser light reaching rate, and system efficiency (excluding the laser efficiency) in the optical wireless power supply system were evaluated through deep seawater using blue, green, and red lasers.

2. Materials and Methods

Firstly, the light reaching rate of the laser beam under tap water and seawater was investigated in the experimental configuration, as shown in Figure 3. Figure 4 shows an experimental view of 450 nm irradiation. In this experiment, laser power was measured in air. The fiber laser beam was transmitted through a water tank filled with tap water or seawater. The light reaching rate was calculated by measuring the laser beam intensity P₁ before entering the tank and the laser beam intensity P₂ after transmission through the tank using a power meter. After light transmission through the water tank, the laser beam was irradiated to the GaInP solar cells, and the photoelectric conversion efficiency was measured to calculate the system efficiency.

Three fiber lasers with 450 nm, 532 nm, and 635 nm were used as light sources. Because these three lasers are mass produced for use in displays, their costs are decreasing. The visible light range is 380 nm to 780 nm, and three different wavelengths, blue (450 nm), green (532 nm), and red (635 nm), were used to investigate a wide range of wavelengths.

To realize the OWPT system underwater, it was first necessary to investigate whether sufficient system efficiency could be obtained within 1 m and how the system efficiency varied with wavelength and optical path length. Therefore, optical path lengths L in the water or seawater were set to 30 cm, 60 cm, and 90 cm to match the easily available aquarium lengths. The deep seawater collected in June at a depth of 800 m in Ito City, Shizuoka Prefecture, was prepared. The laser source and solar cell were set up outside the tank. As shown in Figure 3, the water tank was placed 10 cm before the laser source, and the GaInP solar cell was placed 20 cm away from the water tank. A plano-convex lens was placed in front of the GaInP solar cell to adjust the laser beam diameter. The water tank was tilted at a 10° angle to the incident laser beam to prevent the light reflected from the tank surface from interfering with the incident light and changing the irradiation intensity.

The GaInP solar cells had a photosensitive area of 1.0 × 1.0 cm², and were mounted on a copper heatsink. The temperature of the GaInP cells was kept at room temperature using a water-cooled circulating chiller. The position of the plano-convex lens was adjusted so that each beam shape on the cell had a 6 mmϕ circular diameter. The irradiated area to the cell was approximately 28.27 mm². The output power P_out was measured using a source measure (B2901A, KEYSIGHT, Santa Rosa, CA, USA) with a 1 μV and 100 pA measurement resolution.

When we measured the reflectance by incident of the 532 nm laser beam into the glass aquarium without water with an incident angle of 10 degrees, the reflectance was 7.9%. However, the transmittance was 88%, indicating that there was a scattering loss of 4.1% at the glass interfaces. In the case of an aquarium filled with water, the reflectance was actually measured to be 4.6% (3.5% by calculation); therefore, since it is difficult to measure the reflectance from water to glass, we can roughly estimate that the transmittance is 80.4%, including the scattering losses between the left and right glasses and the attenuation loss by water (L = 30 cm).

The light reaching rate ${\eta }_{\mathrm{o}\mathrm{p}}$, photoelectric conversion efficiency ${\eta }_{\mathrm{p}\mathrm{v}}$, and system efficiency ${\eta }_{\mathrm{s}\mathrm{y}\mathrm{s}}$ were calculated by the following equations.

$${\eta }_{\mathrm{o}\mathrm{p}}=P_2/P_1$$

(1)

$${\eta }_{\mathrm{p}\mathrm{v}}=P_{\mathrm{o}\mathrm{u}\mathrm{t}}/P_2=I_{\mathrm{s}\mathrm{c}}\times V_{\mathrm{o}\mathrm{c}}\times FF/P_2$$

(2)

$${\eta }_{\mathrm{s}\mathrm{y}\mathrm{s}}={\eta }_{\mathrm{o}\mathrm{p}}\times {\eta }_{\mathrm{p}\mathrm{v}}$$

(3)

where I_sc is the short-circuit current, V_oc is the open-circuit voltage, and FF is the fill factor.

3. Results and Discussion

Table 2. Photoelectric conversion efficiency of GaInP solar cell at P₁ = 50 mW.

Wavelength (nm)	450	532	635
${\eta }_{\mathrm{p}\mathrm{v}}$(%)	30.6	40.3	39.3

shows the results of the photoelectric conversion efficiencies calculated for the GaInP solar cells. The photoelectric conversion efficiencies ${\eta }_{\mathrm{p}\mathrm{v}}$ of 30.6% at 450 nm, 40.3% at 532 nm, and 39.3% at 635 nm were evaluated in air under 50 mW incident laser power with the same beam size of 6 mmϕ, respectively.

Figure 5 shows that the EQE values of 60.5%, 70.0%, and 60.4% were obtained for the 450 nm, 530 nm, and 630 nm laser irradiations, respectively. The EQE values at 450 nm and 635 nm were almost the same, but the ${\eta }_{\mathrm{p}\mathrm{v}}$ under the 635 nm laser irradiation was higher than that of 450 nm. This is because the photon energy (2.76 eV) of the 450 nm light was larger than the band gap energy (1.88 eV) of the GaInP, which meant that the over-excited energy became thermal loss.

Figure 6 shows the dependence of I_sc, V_oc, and FF on P₂ and the I-V curve. I-V curve was shown for a wavelength of 532 nm with P₁ = 50 mW. Figure 6a,b shows that I_sc increased in proportion to P₂ and V_oc increased with increasing current. However, FF decreased with increasing P₂ as shown in Figure 6c. This is thought to be due to Joule heat loss (I²R) caused by the increase in I_sc.

Figure 7 shows the results of the dependence of the light reaching rate ${\eta }_{\mathrm{o}\mathrm{p}}$ on the optical path length L under water. The incident laser wavelength of 532 nm had the highest ${\eta }_{\mathrm{o}\mathrm{p}}$ in all the waters and seawaters sampled in this experiment.

In tap water, the ${\eta }_{\mathrm{o}\mathrm{p}}$ at 532 nm was 81.2% at L = 30 cm, 79.6% at L = 60 cm, and 78.4% at L = 90 cm, showing no significant change, although it slightly decreased as the underwater optical path length increased. Similarly, the ${\eta }_{\mathrm{o}\mathrm{p}}$ at 450 nm did not decrease significantly as the underwater optical path length L increased. On the other hand, the ${\eta }_{\mathrm{o}\mathrm{p}}$ at 635 nm was 71.3% at L = 30 cm and it decreased linearly with increasing L (red dot line).

In deep water, the ${\eta }_{\mathrm{o}\mathrm{p}}$ at 532 nm was 78.2% at L = 30 cm, 64.4% at L = 60 cm, and 58.3% at L = 90 cm. The ${\eta }_{\mathrm{o}\mathrm{p}}$ at 450 nm and 532 nm decreased significantly from L = 30 cm to L = 60 cm and then decreased slowly from L = 60 cm to L = 90 cm. However, the ${\eta }_{\mathrm{o}\mathrm{p}}$ at 635 nm decreased linearly from L = 30 cm to L = 90 cm, as was the case with the tap water.

From the above results, blue and green lasers are suitable for highly transparent water, such as tap water and deep seawater. On the other hand, red lasers are suitable for the low-transparency water in the bays of Okinawa Prefecture and Tokyo Bay. In the Okinawa Bay seawater, the ${\eta }_{\mathrm{o}\mathrm{p}}$ at 635 nm exceeded 450 nm in the case of L = 90 cm. When L = 30 cm and L = 90 cm, the reaching rate of 635 nm was higher than that of 450 nm. Even in the Tokyo Bay seawater, the ${\eta }_{\mathrm{o}\mathrm{p}}$ at 635 nm exceeded that at 450 nm at all optical path lengths L. It is suggested that the effect of scattering by suspended matter and plankton is more pronounced for shorter-wavelength light.

A micrograph of the seawater used in the measurements is shown in Figure 8. These photos were taken with the seawater sandwiched between two glass slides. As shown in Figure 8a,b, suspended matters were observed in the gulf seawater in Okinawa Bay and Tokyo Bay. Live plankton with sizes of 150–200 μm were also observed in the seawater of Tokyo Bay. It is speculated that in seawater, laser beams could be scattered by such suspended matters, leading to a decrease in the light reaching rate. The gulf seawater had a mixture of suspended solids with a relatively large particle size of about 5–60 μm, as well as microscopic suspended solids. The Tokyo Bay seawater was less transparent than the other seawaters used in this study and is considered to have been strongly affected by scattering due to the larger particle size of the suspended matter.

On the other hand, in the deep seawater, there were many microscopic suspended solids with a maximum diameter of about 20 μm, and plankton with sizes of 100–125 μm were also observed. Therefore, the deep seawater had less scattering loss due to the smaller suspended matter.

Finally, the results of the dependence of the system efficiency on the optical path length L are shown in Figure 9.

In tap water, the highest system efficiency was obtained at 31.5% with L = 30 cm and 30.8% at L = 90 cm at the incident light wavelength of 532 nm. In the deep seawater, ${\eta }_{\mathrm{s}\mathrm{y}\mathrm{s}}$ of 31.3% and 23.5% were obtained at 532 nm with L = 30 cm, and L = 90 cm, respectively. The system efficiency of the blue and green lasers also decreased significantly due to the effect of the lower light reaching rates from the 30 cm to 60 cm underwater path lengths. In the Okinawa seawater, the ${\eta }_{\mathrm{s}\mathrm{y}\mathrm{s}}$ at 532 nm was 25.6% at L = 30 cm and 16.7% at L = 90 cm [15].

For the tap water and deep seawater, the ${\eta }_{\mathrm{o}\mathrm{p}}$ was higher at 532 nm, 450 nm, and 635 nm, while the ${\eta }_{\mathrm{s}\mathrm{y}\mathrm{s}}$ was higher at 532 nm, 635 nm, and 450 nm. This was because the photoelectric conversion efficiency of the GaInP solar cells used was higher at 635 nm than at 450 nm.

GaInP solar cells are more suitable for using red light than blue light regarding photoelectric conversion efficiency; however, in highly transparent water, InGaN solar cells with higher photoelectric conversion efficiency for blue light will be advantageous if realized.

4. Conclusions

In this study, the efficiencies of the OWPT system under deep seawater and tap water were investigated using GaInP solar cells and three lasers with 450 nm, 532 nm, and 635 nm wavelengths. The GaInP solar cells showed photo-electric conversion efficiencies of 30.6%, 40.3%, and 39.3% for the 450 nm, 532 nm, and 635 nm irradiations, respectively. In deep seawater and tap water, a 532 nm laser through a 90 cm water tank showed light reaching rates of 78.4% and 58.3% and entire system efficiencies, excluding the efficiency of the laser itself, of 30.8% and 23.5%, respectively. The 532 nm green laser showed higher efficiency than the other 450 nm and 635 nm lasers in the underwater system using GaInP solar cells. These results suggest that it is essential to select a laser wavelength with high transmittance suitable for the actual sea area and to develop highly efficient solar cells to realize an OWPT system under seawater.