⇦Chinese Physics Letters, 2020, Vol. 37, No. 8, Article code 087802 Effect of Dopant Concentration in a Base Layer on Photocurrent–Voltage Characteristics of Photovoltaic Power Converters Wen-Xue Huo (霍雯雪)1,2,3, Ming-Long Zhao (赵明龙)1,2,3, Xian-Sheng Tang (唐先胜)1,2,3, Li-Li Han (韩丽丽)1,2,3, Zhen Deng (邓震)1,3,5, Yang Jiang (江洋)1,3, Wen-Xin Wang (王文新)1,3,4, Hong Chen (陈弘)1,3,4, Chun-Hua Du (杜春花)1,3,5*, and Hai-Qiang Jia (贾海强)1,3,4* Affiliations 1Key Laboratory for Renewable Energy, Beijing Key Laboratory for New Energy Materials and Devices, Beijing National Laboratory for Condensed Matter Physics, Institute of Physics, Chinese Academy of Sciences, Beijing 100190, China 2University of Chinese Academy of Sciences, Beijing 100049, China 3Center of Materials and Optoelectronics Engineering, University of Chinese Academy of Sciences, Beijing 100049, China 4Songshan Lake Materials Laboratory, Dongguan 523808, China 5The Yangtze River Delta Physics Research Center, Liyang 213300, China Received 8 April 2020; accepted 22 May 2020; published online 28 July 2020 Supported by the National Natural Science Foundation of China (Grant Nos. 61704008 and 11574362) and the Strategic Priority Research Program of Chinese Academy of Sciences (Grant No. XDB33000000).
*Corresponding authors. Email: duchunhua@iphy.ac.cn; mbe2@iphy.ac.cn Citation Text: Huo W X, Zhao M L, Tang X S, Han L L and Deng Z et al. 2020 Chin. Phys. Lett. 37 087802    Abstract It is known that the p–n junction of an absorption region is a crucial part for power conversion efficiency of photovoltaic power converters. We fabricate four samples with different dopant concentrations in base layers. The dependences of power conversion efficiency and fill factor on input power are displayed by photocurrent–voltage measurement. Photoluminescence characteristics under open circuit and connected circuit conditions are also studied. It is found that the status of p–n junction matching is the critical factor in affecting the power conversion efficiency. In addition, series resistance of photovoltaic power converters impairs the efficiency especially at high input powers. Both the key factors need to be considered to obtain high efficiency, and this work provides promising guidance on designing photovoltaic power converters. DOI:10.1088/0256-307X/37/8/087802 PACS:78.55.Cr, 78.55.-m, 73.40.Lq, 73.50.Pz © 2020 Chinese Physics Society Article Text As an effective route to transfer power to remote locations without any metal wires, laser power beaming has attracted attention for many years, especially in space missions.[1–11] Better than solar power generation, the laser power transfer can generate power any time, avoiding the influence of weather changing. Moreover, a laser power beaming system could show a higher power conversion efficiency compared to the systems with broad-solar spectra[12,13] since the laser can be exactly matched to an appropriate photovoltaic power converter.[14,15] Most photovoltaic power converters are GaAs-based material systems,[16–18] exhibiting high power conversion efficiency up to 60%.[19,20] It is noticed that superior photovoltaic power converters with high power conversion efficiency commonly benefit from appropriate device structures. Therefore, the epilayer design plays an important role in promoting the efficiency of photovoltaic power converters. In diode devices, the most significant part is the p–n junction of the absorption region, including an emitter layer and a base layer. In previous studies, the thickness and dopant concentration of emitter layers have been already discussed. It is reported that in p/n-GaAs devices, the sample owning a thicker p-type emitter layer with lower dopant concentration has a higher efficiency; in contrast, a thinner n-type emitter layer with a higher dopant concentration is more efficient in n/p-GaAs devices.[2,5,21] In this letter, by changing dopant concentration in the base layer, deep research is conducted to explore the function of p–n junction matching for performance of photovoltaic power converters by photoluminescence measurement. In addition, series resistance is also testified and calculated by the global method to investigate the changes of performance for different samples at different input powers. The multilayers of the GaAs photovoltaic power converter are grown epitaxially on a 2-inch-diameter n-type (1 0 0) GaAs substrate misoriented 15$^{\circ}$ toward [111] by using a Veeco E450 metal-organic chemical vapor deposition (MOCVD) system. The carrier gas is high-purity H$_{2}$. Element III precursors for Ga, Al and In are trimethylgallium (TMGa), trimethylaluminum (TMAl) and trimethylindium (TMIn), respectively, and element V sources for As and P come from arsine (AsH$_3$) and phosphine (PH$_3$) separately. The growth temperature of arsenide and phosphide-based materials is about 620 ℃ and the reactor pressure is kept at 42 Torr, while a lower temperature of 600 ℃ is adopted for the tunnel junction growth. The epilayers start with a 500 nm GaAs buffer layer on a 350 µm n-type GaAs substrate. Then the single-junction photovoltaic power converter is grown uprightly as follows: a 44-nm-thick n$^{++}$-GaAs/p$^{++}$-GaAs reverse tunneling junction, a 55-nm-thick p-type Al$_{0.3}$Ga$_{0.7}$As back surface field layer with dopant concentration of $3 \times 10^{17}$ cm$^{-3}$, a 2.7 µm-thick p-type GaAs base layer with the dopant concentration ranging from $2.5 \times 10^{17}$ cm$^{-3}$ to $5.2 \times 10^{18}$ cm$^{-3}$, a 100-nm-thick n-type GaAs emitter layer with the concentration of $2 \times 10^{18}$ cm$^{-3}$, a 30-nm-thick n-type InAlP window layer with the concentration of $3 \times 10^{18}$ cm$^{-3}$, and a 400-nm-thick n-type GaAs ohmic contact layer with the dopant concentration of $5 \times 10^{18}$ cm$^{-3}$ sequentially. To be more specific, the different dopant concentrations in the p-type GaAs base layers are obtained by changing the ratio of TMGa/AsH$_{3}$/DMZn. The detailed dopant concentrations in the p-type base layers of samples A, B, C and D are listed in Table 1. The two-dimensional schematic diagram of the epilayer is shown in Fig. 1(a), and Figs. 1(b)–1(d) are the SEM images of different parts of the epilayers.
cpl-37-8-087802-fig1.png
Fig. 1. (a) The schematic structure of the GaAs-based photovoltaic power converter; (b)–(d) the SEM image of epitaxy structure; (e) the schematic diagram of the fabricated GaAs-based photovoltaic power converter device.
Table 1. The dopant concentration $C_{\rm d}$ of the base layer for different samples.
Samples A B C D
$C_{\rm d}$ (cm$^{-3}$) $5.2 \times 10^{18}$ $3.2 \times 10^{18}$ $3.6 \times 10^{17}$ $2.5 \times 10^{17}$
The fabrication process of photovoltaic power converter devices is as follows. Initially, the top-grid n-type electrodes were fabricated by depositing Ni/Au/Ge/Ni/Au metal layers sequentially through electron beam evaporation and were patterned through photolithographic processing. Then, Ni/Au/Ge/Ni/Au metal layers were deposited on the n$^{+}$-GaAs substrate as the bottom electrodes, followed by the rapid thermal annealing process at 380 ℃ in N$_{2}$ ambient to improve ohmic contact property for both top and bottom electrodes. After that, via building isolation trenches by selective chemical etching, the wafer was divided into discrete devices with the effective mesa of 5 mm $\times 6$ mm. Lastly, bilayer anti-reflective coating (ARC) of SiO$_{2}$/TiO$_{2}$ was covered on the top of n-type GaAs ohmic contact layer region, with the top-grid n-type electrodes exposed. The two-dimensional schematic diagram of the fabricated GaAs photovoltaic power converter device is shown in Fig. 1(e). The two-dimensional schematic diagram of epilayers was obtained by a scanning electron microscope (SEM, Hitachi S-4800). The photocurrent–voltage ($I$–$V$) characteristics were measured with a digital semiconductor device (2400 Source-Meter, Keithley Instrument) under the 808 nm laser at variable powers. Utilizing a PID temperature controller (XMT7100), the experiment temperature was kept at 25 ℃. Photoluminescence spectra were obtained by a photoluminescence measurement system including an 808 nm laser with the incident power of 7.5 mW, a beam chopper, two condenser lens locating in front of the sample and detector separately, a sample rack, an analog lock-in amplifier (Stanford Research Systems model 830), an InGaAs detector, and a computer with the Spectra Sense software (Acton Research, Ver. 4.3.2). Figure 2 depicts the input power dependence of power conversion efficiency (PCE) and fill factor (FF) of four samples. The PCE and FF are obtained by $I$–$V$ measurements under seven different input powers. According to Fig. 2(a), the PCEs of four samples increase with increasing incident power at low input powers, whereas they decrease with increasing incident power at high input powers. Samples C and D with low dopant concentration in the base layer show higher PCE than samples A and B at powers lower than 1600 mW/cm$^{2}$. However, the PCE curves of samples C and D decay faster than that of samples A and B after 1600 mW/cm$^{2}$. As a result, PCE of sample B, which owns high dopant concentration, gradually surpasses that of samples D and C with the increasing input power. PCE of sample A, which is the smallest one at low input powers, exceeds that of sample D at power of 6400 mW/cm$^{2}$. In Fig. 2(b), the FFs of four samples display different relationships with input powers. The inset shows the curves at powers before reaching 800 mW/cm$^{2}$. Samples C and D have better FF at first, but FFs of samples A and B become higher than those of samples C and D after 400 mW/cm$^{2}$ and keep ahead thereafter. Above all, owing to the different dopant concentrations in the base layer, four samples possess different behaviors with the input power increasing. The deep reason will be further discussed in the following results.
cpl-37-8-087802-fig2.png
Fig. 2. (a) PCE and (b) FF distributions of four samples under different input powers.
Different dopant concentrations in a base layer usually lead to different statuses of p–n junction matching. Photoluminescence measurement is an effective method to examine the status of p–n junction matching. As is known, under the open circuit condition, the carriers generated by the incoming laser have no place to go but recombine radiatively, causing external luminescence. Under the connected circuit condition, most photo-generated carriers will recombine through the external circuit. However, because of bad p–n junction matching, the drift velocities of electrons and holes in the built-in region are not matching well, leading to the accumulation of carriers at the boundary of the built-in region.[22,23] As a result, part of carriers cannot transport throughout the external circuit and may recombine to emit externally. Thus, the intensity ratio of external luminescence between the connected circuit condition and the open circuit condition can be thought as an accordance to assess the status of p–n junction matching. In detail, the intensity ratio decreases as the p–n junction matches better. Figure 3 demonstrates the normalized photoluminescence curves of four samples under the open and connected circuit conditions. With the increasing dopant concentration, the peak shifts toward long wavelength and the full wavelength at half maximum (FWHM) becomes larger, which agrees with the results previously reported.[24,25] From Fig. 3, we can conclude that sample D shows the best p–n junction matching status and sample C displays comparably good p–n junction matching status. However, the statuses of p–n junction matching of samples A and B are unsatisfied in Fig. 3, where sample A performs the worst. The photoluminescence spectra show that the samples with lower base-layer dopant concentration have a better status of p–n junction matching. Combining the results of Fig. 2 and Fig. 3, we can see that the PCE behaviors of the four samples partially agree with their statuses of p–n junction matching. At input powers lower than 1600 mW/cm$^{2}$, the samples with better status of p–n junction matching are seen to have higher PCE. Thus, PCEs of samples C and D are remarkably higher than those of samples A and B. However, it is worth noting that the PCE of sample D, which owns the best status of p–n junction matching, is slightly lower than that of sample C and falls behind obviously after 400 mW/cm$^{2}$. Furthermore, the p–n junction matching data does not accord with the PCE results of the four samples after 1600 mW/cm$^{2}$. According to these abnormal results, it is supposed that the status of p–n junction matching may not be the only key factor to affect the performance of photovoltaic power converters.
cpl-37-8-087802-fig3.png
Fig. 3. The normalized photoluminescence curves of (a) sample A, (b) sample B, (c) sample C and (d) sample D under the open-circuit and connected-circuit conditions.
Figure 4 shows the $J$–$V$ curves of the four samples at four different input powers. We can see that $J_{\rm sc}$ and $V_{\rm oc}$ of the four samples are different, which may be caused by their different statuses of p–n junction matching. In addition, the difference of $V_{\rm oc}$ is also owing to their different bandgaps, because the wider bandgaps will lead to higher $V_{\rm oc}$. Thus, the samples with lower dopant concentration will have a wider bandgap and own higher $V_{\rm oc}$. Moreover, it is also worth noting that the slopes of curves for four samples are different. This curve slope is related to a series resistance of devices. The series resistance will decrease the voltage, which will further lower PCE and FF. Therefore, the curve becomes more inclined when the series resistance increases[26,27] or the current grows. To further testify series resistance of the four samples, the global method from IEC[28,29] based on the $J$–$V$ curves in Fig. 4 is used. The $J$–$V$ curves of 400 mW/cm$^{2}$, 1600 mW/cm$^{2}$, and 6400 mW/cm$^{2}$ are used to calculate their series resistances. The calculation results are listed in Table 2. It is found that samples A and B have small series resistances, which leads to low decay rates of PCE and FF at powers higher than 1600 mW/cm$^{2}$. Sample B with the smallest series resistance has the highest PCE and FF at 6400 mW/cm$^{2}$. Moreover, samples C and D own high series resistance, resulting in the fast decay of PCE and FF at powers higher than 1600 mW/cm$^{2}$. Especially, PCE and FF of sample D decay faster and are the lowest at 6400 mW/cm$^{2}$ compared to the other three samples.
cpl-37-8-087802-fig4.png
Fig. 4. $J$–$V$ curves of the four samples at input powers of (a) 100 mW/cm$^{2}$, (b) 400 mW/cm$^{2}$, (c) 1600 mW/cm$^{2}$ and (d) 6400 mW/cm$^{2}$.
Table 2. The series resistance of four samples.
Samples A B C D
Series resistance $R_{\rm s}$ ($\Omega$) 0.13643 0.04399 0.18337 0.3512
Combining the two measurement results shown in Figs. 3 and 4, statuses of p–n junction matching and series resistance are two major factors that affect the performance of photovoltaic power converters. At low input powers, the function of series resistance on $I$–$V$ characteristics of photovoltaic power converters is weak and the status of p–n junction matching plays an important role to affect their performances. Therefore, samples C and D with better statuses of p–n junction matching possess higher PCEs than samples A and B. Furthermore, when input power is lower than 1600 mW/cm$^{2}$, the PCEs of all the samples increase with the growing input powers, while that of sample D starts to decay after 400 mW/cm$^{2}$. In addition, sample D exhibits the best matching of p–n junction but has a lower PCE than sample C. The abnormal behavior of sample D should be ascribed to its big series resistance. With the increasing input power, the effect of series resistance becomes more and more significant. Series resistance will impair the PCE and FF of photovoltaic power converters, which results in attenuation of them at high input powers. Thus, after 1600 mW/cm$^{2}$, PCE and FF of samples C and D with larger series resistances have a faster rate of decay than samples A and B. Herein, sample D exhibits the fastest rate of PCE and FF decay and has the lowest PCE at 6400 mW/cm$^{2}$ because of its largest series resistance. Moreover, owing to the smallest series resistance, sample B has little decay of PCE and FF after 1600 mW/cm$^{2}$ and exhibits the highest PCE at 6400 mW/cm$^{2}$. In summary, by changing the dopant concentration in the base layer, four samples with the similar structure are fabricated. PCE and FF of four devices are obtained by $I$–$V$ measurement at various input powers. Through a photoluminescence test, the status of p–n junction matching is evaluated. In addition, series resistance can be calculated by the global method from IEC based on the $J$–$V$ curves. As a result, p–n junction matching and series resistance are two key factors to determine performance of photovoltaic power converters. At low input powers, series resistance plays a less important role on $I$–$V$ characteristics, and the function of p–n junction matching stands out. Thus, samples C and D with better status of p–n junction matching have higher PCE. However, with the increasing input powers, the effect of series resistance becomes dominant. Samples with high series resistance own high decay rate and have serious attenuation of PCE at high input powers. This work analyzes the functions of those two key factors for affecting PCE of photovoltaic power converters by changing the dopant concentration in the base layer and provides a promising and meaningful guidance for designing photovoltaic power converters working at different input powers. References GaAs converter for high power laser diodeGaAs converters for high power densities of laser illuminationLong-wavelength laser power converters for optical fibersLaser Power Converters for optical power supplyEfficiency limits of laser power converters for optical power transfer applicationsConversion efficiencies of single-junction III–V solar cells based on InGaP, GaAs, InGaAsP, and InGaAs for laser wireless power transmissionDevelopment and characterisation of laser power converters for optical power transfer applicationsHigh-Power High-Efficiency Laser Power Transmission at 100 m Using Optimized Multi-Cell GaAs ConverterHigh-Voltage GaAs Photovoltaic Laser Power ConvertersSimulation of the ohmic loss in photovoltaic laser-power converters for wavelengths of 809 and 1064 nmEnhancing solar cell efficiency: the search for luminescent materials as spectral convertersEnhancing the performance of silicon solar cells via the application of passive luminescence conversion layersInAlGaAs/InP-Based Laser Photovoltaic Converter at $\sim 1070$ nmTowards Efficient Spectral Converters through Materials Design for Luminescent Solar DevicesPower over Fibre: Material Properties of Homojunction Photovoltaic Micro-CellsNitric Oxide (NO) Mediates the Inhibition of Form-Deprivation Myopia by Atropine in ChicksDesign and optimization of GaAs photovoltaic converter for laser power beamingAIP Conference ProceedingsSPIE ProceedingsHigh efficiency monochromatic GaAs solar cellsPhotoluminescence characteristics of InGaP/GaAs dual-junction solar cells under current mismatch conditionsStrong Internal and External Luminescence as Solar Cells Approach the Shockley–Queisser LimitBandgap shrinkage of degenerate p-type GaAs by photoluminescence spectroscopyPhotoluminescence of Heavily p-Type-Doped GaAs: Temperature and Concentration DependencesPhotovoltaic cell characteristics for high-intensity laser lightSeries resistance effects on solar cell measurementsTheoretical analysis of the series resistance of a solar cell
[1] Fave A, Kaminski A, Gavand M et al. 1996 The Conference Record of the 25th IEEE Photovoltaic Specialists Conference (Washington DC 13–17 May 1996) p 101
[2] Oliva E, Dimroth F and Bett A W 2008 Prog. Photovoltaics 16 289
[3] Wojtczuk S J 1997 The Conference Record of the 26th IEEE Photovoltaic Specialists Conference (Anaheim CA 29 September–3 October 1997) p 971
[4] Sohr S, Rieske R, Nieweglowski K et al. 2011 Proceedings of the 34th International Spring Seminar on Electronics Technology (Slovakia 11–15 May 2011) p 122
[5]Andreev V, Khvostikov V, Kalinovsky V et al. 2003 Proceedings of the 3rd World Conference on Photovoltaic Energy Conversion (Oszka 11–18 May 2003) p 761
[6] Mukherjee J, Jarvis S, Perren M et al. 2013 J. Phys. D 46 264006
[7] Jomen R, Tanaka F, Akiba T et al. 2018 Jpn. J. Appl. Phys. 57 08RD12
[8] Jarvis S D, Sweeney S J, Perren M et al. 2014 IET Optoelectron. 8 64
[9] He T, Yang S H, Zhang H Y et al. 2014 Chin. Phys. Lett. 31 104203
[10] Schubert J, Oliva E, Dimroth F et al. 2009 IEEE Trans. Electron Devices 56 170
[11] Emelyanov V M, Mintairov S A, Sorokina S V et al. 2016 Semiconductors 50 125
[12] Huang X, Han S, Huang W et al. 2013 Chem. Soc. Rev. 42 173
[13] Richards B S 2006 Sol. Energy Mater. Sol. Cells 90 2329
[14] Singh N, Kin Fai Ho C , Nelvin Leong Y et al. 2016 IEEE Electron Device Lett. 37 1154
[15] McKenna B and Evans R C 2017 Adv. Mater. 29 1606491
[16] Allwood G, Wild G and Hinckley S 2011 Sixth IEEE International Symposium on Electronic Design, Test and Application (Queenstown, New Zealand 17–19 January 2011) p 78
[17] Zhao Y M, Sun Y R, He Y et al. 2016 Sci. Rep. 6 9
[18] Shan T and Qi X 2015 Infrared Phys. & Technol. 71 144
[19] Khvostikov V, Kalyuzhnyy N, Mintairov S et al. 2014 AIP Conf. Proc. 1616 21
[20] Valdivia C E, Wilkins M M, Bouzazi B et al. 2015 Proc. SPIE 9358 93580E
[21] Olsen L C, Huber D A, Dunham G et al. 1991 The Conference Record of the Twenty-Second IEEE Photovoltaic Specialists Conference (Las Vegas, USA 7–11 October 1991) p 419
[22] Huo W X, Zhao M L, Tang X S et al. 2019 Appl. Phys. Express 12 115507
[23] Miller O D, Yablonovitch E and Kurtz S R 2012 IEEE J. Photovoltaics 2 303
[24] Feng M, Fang C A and Chen H 1995 Mater. Chem. Phys. 42 143
[25] Chen H D, Feng M S, Chen P A et al. 1994 Jpn. J. Appl. Phys. 33 1920
[26] Miyakawa H, Tanaka Y and Kurokawa T 2005 Sol. Energy Mater. Sol. Cells 86 253
[27] Wolf M and Rauschenbach H 1963 Adv. Energy Convers. 3 455
[28] Handy R 1967 Solid-State Electron. 10 765
[29]International Electrotechnical Commission 1987 IEC 60891