High-efficiency InGaAsP Laser Power Converters for Optical Wireless Power Transmission

2.1 Device design for high-intensity 1000-nm-class laser illumination

For long-distance OWPT using a 1000-nm-class laser, the receiver must convert high-power monochromatic light into electrical power with minimal loss. Under high-intensity laser illumination, the photocurrent density increases substantially, and even a small series resistance can cause a large voltage drop and output-power loss. Therefore, an LPC for OWPT must be designed not only to absorb the target wavelength efficiently but also suppress resistive loss during high-current operation.

Figure 1 shows the device structure of the single-junction InGaAsP LPC designed for 1064-nm laser illumination. The LPC was fabricated on an InP substrate, and lattice-matched InGaAsP was used as the light-absorbing layer. The bandgap of the InGaAsP absorber was adjusted so that photons at 1064 nm could be efficiently converted into electrical energy. An InP window layer and InGaAsP emitter layer were introduced on the incident-light side, while an InGaAsP back-surface-field layer was placed on the rear side. This layer design is intended to promote carrier collection while reducing recombination loss in the absorber. An antireflection coating optimized for 1064 nm was also formed on the device surface to reduce optical reflection at the laser wavelength.

Fig. 1. Schematic device structure of InGaAsP LPC [9].

A key feature of this LPC is the use of a thick surface electrode. Under high-intensity laser illumination, the large photocurrent increases resistive loss in the surface electrode, which can reduce the fill factor (FF) and conversion efficiency (η). Increasing the electrode thickness suppresses this loss, although the electrode pattern must still balance low resistance with low optical shading. The LPC was also designed as a 1-cm²-class LPC, rather than a very small mm²-scale device, to evaluate current collection and electrode-resistance effects under conditions closer to those required for practical OWPT receiver panels.

2.2 Photovoltaic performance under high laser intensity

The high-illumination tolerance of the InGaAsP LPC was examined by measuring its device characteristics under 1064-nm laser illumination at different laser intensities (LIs). Figures 2(a)–(d) show the photovoltaic performance of the 1-cm² InGaAsP LPC under 1064-nm laser illumination as a function of LI. The short-circuit current density (J_SC) increased almost proportionally with LI. The external quantum efficiency (EQE) estimated from the J_SC–LI relationship was 0.77–0.78 for LI = 0.5–5.0 W·cm⁻². The EQE was calculated as EQE = (J_SC/LI) × (hc/qλ), where h is the Planck constant, c is the speed of light, q is the elementary charge, and λ is the laser wavelength. These results indicate that the incident laser power was effectively converted into photocurrent over the measured intensity range. The open-circuit voltage (V_OC) also increased with LI because of the increase in photogenerated carrier density. These trends confirm that the InGaAsP absorber and optical design are suitable for 1064-nm laser conversion. The FF is a particularly important parameter for evaluating high-intensity operation. Under strong laser illumination, the generated current becomes large, and the voltage loss caused by series resistance increases. If this loss is not sufficiently suppressed, the FF decreases and the η is limited even when the optical absorption and carrier collection are high. In this InGaAsP LPC, the thick surface electrode effectively suppressed the decrease in FF until LI = 3.0 W·cm⁻². Therefore, the LPC maintained stable output under high-current operation and achieved a maximum η of 42% at a LI of 4.0 W·cm⁻². These results indicate that the reduction in resistive loss is essential for improving the illumination tolerance of LPCs for OWPT.

Fig. 2. (a) J_SC, (b) V_OC, (c) FF, and (d) η of the single-junction InGaAsP LPC with a device area of 1 cm² as a function of LI. This figure is reprinted from the original paper [8] (CC BY-NC-ND 4.0).

To clarify the origin of the remaining efficiency loss, the contribution of reflectance (R), carrier lifetime and transmittance (τ and T), shunt resistance (R_SH), and series resistance (R_S) was evaluated from the measured current density versus voltage (J–V) characteristics at LI = 4.0 W·cm⁻², as shown in Figs. 3(a)–(d). In this analysis, we adopted a one-diode equivalent circuit model:

where J_ph denotes photocurrent density, J₀ reverse saturated current density, n ideal factor, k_B the Boltzmann constant, and T_LPC the LPC temperature. The theoretical limit at LI = 4.0 W·cm⁻² was calculated on the basis of the Shockley-Queisser limit. The measured voltage includes not only the intrinsic voltage of the pn junction but also the voltage drops caused by current flow through the electrode, semiconductor layers, and contacts. This voltage drop becomes more significant as the current density increases. Therefore, the difference between the measured J–V curve and series-resistance-free behavior provides a direct indication of how much output power is lost through resistive components. The analysis showed that the ratio of R_S loss in FF was 35% in total FF deviation from the theoretical limit (Fig. 3(c)); however, the ratio of R_S loss in η was only 7% (Fig. 3(d)). These results suggest that the R_S loss was greatly reduced by the thick surface electrode even in a 1-cm²-area device. To further enhance η, recombination of photogenerated carriers should be suppressed according to the one-diode-model-based analysis.

Fig. 3. Loss factors limiting the performance of the InGaAsP LPC at LI = 4.0 W·cm⁻². The decreases in (a) J_SC, (b) V_OC, (c) FF, and (d) η from the theoretical limit are separated into the effects of R, τ, T, R_SH, and R_S. This figure is reprinted from the original paper [8] (CC BY-NC-ND 4.0).

These results provide an important design guideline for practical receiver panels. When the incident laser power is increased to obtain higher electrical output, the photocurrent density also increases, and the allowable series of resistance becomes smaller. Therefore, electrode design, contact resistance, current-spreading layers, and interconnection structures must be optimized together. This 1-cm² InGaAsP LPC shows that high efficiency can be achieved under high-intensity 1064-nm laser illumination when the device structure is designed to suppress resistive loss. This approach is directly relevant to long-distance OWPT systems, where the receiver must convert high optical power into electrical power efficiently and stably.

2.3 Benchmarking against other LPC technologies

The performance of LPCs has been widely investigated using various semiconductor materials, laser wavelengths, device areas, and junction configurations. Therefore, benchmarking should be interpreted carefully. A direct comparison based only on the maximum η can be misleading because the output characteristics strongly depend on the laser wavelength, illumination intensity, device area, junction number, temperature, and current-collection design. Small-area LPCs tend to particularly show lower resistive loss, whereas larger-area LPCs are more relevant to practical OWPT receivers but are more strongly affected by current spreading and electrode resistance. For this reason, we focused on single-junction LPCs designed for 1064-nm laser illumination with a device area of 1 cm² or larger. Under these conditions, Si and InGaAs have been mainly evaluated as receiver materials.

A key difference among previously reported 1-cm²-class single-junction LPCs for 1064-nm lasers is the LI at which the maximum η is obtained. In conventional 1064-nm LPCs with a device area of 1 cm² or larger, η typically reached maximum at approximately 1–2 W·cm⁻² (Fig. 4). At higher LIs, the generated photocurrent increased, but the FF decreased because of enhanced R_S loss. As a result, η no longer increased with LI, instead became limited by current extraction. In this InGaAsP LPC, the high-illumination tolerance was improved by suppressing the R_S loss with a thick surface electrode. This design shifted the optimum operating point to a higher LI and enabled efficient operation up to LI = 4.0 W·cm⁻². The LPC achieved a η of 42.0% and FF of 76.7% at LI = 4.0 W·cm⁻². To the best of our knowledge, these values represent the highest reported performance for a single-junction LPC designed for 1064-nm laser illumination with an active area of 1 cm² or larger. The important point is not only the absolute efficiency but also the fact that the maximum efficiency was obtained at a higher LI than in previous 1-cm²-class single-junction LPCs. These results indicate that reducing resistive loss is essential for extending the practical operating range of LPCs toward higher optical input power.

Fig. 4. Benchmarking of 1-cm²-class single-junction LPCs for 1064-nm lasers. The plot compares the reported η at different LIs. This figure is reprinted from the original paper [8] (CC BY-NC-ND 4.0).

2.4 Potential impact on long-distance OWPT system efficiency

The improvement in high-illumination tolerance has a direct implication for long-distance OWPT systems. In a practical OWPT link, increasing the incident optical power is effective in increasing the electrical output. However, this also increases the photocurrent generated in the receiver, making resistive loss more severe. Therefore, an LPC optimized only for low-intensity operation may not be suitable for high-power OWPT systems, even if it exhibits high η under moderate illumination. The InGaAsP LPC addresses this issue by maintaining high η under stronger 1064-nm laser illumination.

The potential impact can be estimated by considering the recent 1-km OWPT demonstration using a Si-based photovoltaic receiver. The transmitted laser power was 1035 W, and the optical power received by the 60 cm × 60 cm receiver panel was 773 W [5]. The optical power used to evaluate the intrinsic photovoltaic conversion, after excluding losses associated with a module-level factor such as series/parallel interconnection, was 626 W [5]. If the receiver-side photovoltaic conversion were replaced with an InGaAsP LPC operating at 42% efficiency, the electrical output expected from the same 626 W of effective optical power would be approximately 263 W. This value is 1.73 times higher than the reported electrical output of 152 W. This simple estimate indicates that improving the receiver η can directly increase the usable electrical power without increasing the transmitted laser power.

The high-illumination tolerance of the InGaAsP LPC can also reduce the required receiver area. In the 1-km demonstration, 773 W of optical power was received by a 60 cm × 60 cm panel, corresponding to an average optical intensity of approximately 0.21 W·cm⁻² [5]. In contrast, the InGaAsP LPC maintained high efficiency at LI = 4.0 W·cm⁻². If the same received optical power of 773 W could be accepted at this intensity, the required receiver area would be approximately 193 cm², corresponding to a square receiver with a side length of approximately 14 cm. This represents an approximately 95% reduction in receiver area compared with a 60 cm × 60 cm panel.

These estimates are intended to show the potential system-level benefit of high-illumination-tolerant LPCs, rather than to claim direct replacement of the receiver without additional design considerations. In practice, the receiver area, beam size, heat dissipation, optical alignment tolerance, series/parallel interconnection, and safety requirements must be designed together. In particular, concentrating the same optical power onto a smaller receiver increases the thermal load density, and the receiver panel must be designed to dissipate heat while maintaining stable electrical operation. Nevertheless, the estimate clearly shows that a high-efficiency InGaAsP LPC can contribute to both higher electrical output and smaller receiver size in long-distance OWPT systems. Therefore, the significance of the InGaAsP LPC is not limited to the device-level η. By maintaining high efficiency at a higher LI, the LPC can relax the tradeoff between output power and receiver size. This feature is especially important for mobile and space-related OWPT systems, where receiver area, weight, and thermal management capacity are strongly constrained.

3.1 Operation-relevant radiation evaluation

The previous section discussed how a high-efficiency InGaAsP LPC can improve receiver-side conversion in long-distance OWPT. For space-related OWPT, however, the receiver must also maintain this performance after exposure to energetic particles. We therefore evaluated the electron-irradiation tolerance of InGaAsP LPCs designed for 1064-nm laser illumination under operation-relevant conditions. The purpose of this evaluation was not only to quantify radiation-induced degradation but also to clarify how it appears under high-intensity monochromatic illumination, which is the actual operating condition of LPCs in OWPT systems.

The InGaAsP LPCs with a device area of 0.06 cm² were irradiated with 1-MeV electrons at a total electron fluence of 1 × 10¹⁵ cm⁻², which is often used as an accelerated irradiation level corresponding to long-duration space exposure. Their photovoltaic characteristics were measured under 1064-nm laser illumination and under air mass 1.5 global (AM1.5G) illumination. By comparing these illumination conditions, we examined whether the radiation-induced performance loss evaluated under conventional solar-cell-like illumination differs from that observed under OWPT-relevant laser operation. In the evaluation, the LI was also varied to examine whether high-intensity laser illumination changes the apparent degradation behavior. This point is important because practical OWPT receivers are exposed to strong monochromatic light during operation, and the high carrier injection caused by laser illumination can impact recombination processes in the irradiated device. The LPCs were therefore characterized not only immediately after electron irradiation but also during continuous 1064-nm laser illumination, enabling the evaluation of both irradiation-induced degradation and laser-illumination-induced recovery.

3.2 Electron-irradiation degradation and laser-induced recovery

The electron-irradiation evaluation showed that the performance of InGaAsP LPCs degraded after exposure to 1-MeV electrons. Figure 5 shows how the photovoltaic parameters changed after electron irradiation during continuous illumination. The values are expressed as remaining factors, and t_ill = 0 min represents the initial measurement immediately after the onset of illumination. Electron irradiation mainly reduced J_SC and η, indicating that the degradation was primarily caused by reduced carrier collection rather than by increased resistive loss. At t_ill = 0 min, the decreases in J_SC and η were larger under 1064-nm laser illumination than under AM1.5G illumination. This difference is explained by the EQE degradation in the long-wavelength region, where photons are absorbed deeper in the InGaAsP absorber. These results suggest that irradiation-induced defects increased recombination in the absorber and reduced the collection of carriers generated by 1064-nm light.

Fig. 5. Comparison of post-irradiation performance recovery under different illumination conditions. After 1-MeV electron irradiation at a fluence of 1 × 10¹⁵ cm⁻², the remaining factors of J_SC, V_OC, FF, and η were monitored for 30 min under 1064-nm laser illumination at LI = 6.8 W·cm⁻² and 0.1 W·cm⁻², and under AM1.5G illumination at 0.1 W·cm⁻² [9].

An important feature of LPC operation is that the LPC is continuously illuminated by high-intensity monochromatic laser light during power transmission. We therefore examined whether the irradiated InGaAsP LPC simply remained degraded or whether its performance changed during subsequent illumination. Under 1064-nm laser illumination at LI = 6.8 W·cm⁻², the remaining factor of η at t_ill = 0 min was 0.574, which was substantially lower than that under AM1.5G illumination. However, during continuous laser illumination, η gradually recovered and reached a remaining factor of 0.721 at t_ill = 30 min. This value became comparable to the remaining factor under AM1.5G illumination after 30 min, which was approximately 0.71 and showed no meaningful recovery. The recovery of η followed the recovery of J_SC: under high-intensity 1064-nm laser illumination, the remaining factor of J_SC increased from 0.637 at t_ill = 0 min to 0.791 at t_ill = 30 min, whereas the J_SC remaining factor under AM1.5G illumination changed only slightly from 0.791 to 0.796. These results indicate that the apparent radiation degradation of an LPC depends strongly on the illumination condition during operation, and that high-intensity 1064-nm laser illumination can partially recover electron-irradiation-induced performance loss.

These results highlight the importance of operation-relevant radiation evaluation for OWPT receivers. If an LPC is evaluated only under conventional solar-cell illumination, the degradation and recovery behavior under actual laser-powered operation may be overlooked. For space-related OWPT systems, this means that radiation qualification should account for not only the accumulated particle fluence but also the wavelength, intensity, and duration of laser illumination during operation. The results also suggest that InGaAsP LPCs are promising for space-related OWPT because they combine high-efficiency 1064-nm conversion with the possibility of partial in-situ recovery under high-intensity laser operation.