⇦Chinese Physics Letters, 2016, Vol. 33, No. 5, Article code 058801⇨ Excess Carrier Lifetime Improvement in c-Si Solar Cells by
YAG:Ce$^{3+}$-Yb$^{3+}$ * Talib Hussain1,2, Hui-Qi Ye(叶慧琪)1, Dong Xiao(肖东)1** Affiliations 1National Astronomical Observatories, Nanjing Institute of Astronomical Optics & Technology, Key Laboratory of Astronomical Optics & Technology, Nanjing Institute of Astronomical Optics & Technology, Chinese Academy of Sciences, Nanjing 210042 2University of Chinese Academy of Sciences, Beijing 100049 Received 28 January 2016 ^*Supported by the Natural Science Foundation of Jiangsu Province under Grant No BK2011033.
^**Corresponding author. Email: dxiao@niaot.ac.cn Citation Text: Hussain T, Ye H Q and Xiao D 2016 Chin. Phys. Lett. 33 058801 Abstract Ce$^{3+}$-Yb$^{3+}$ doped Y$_{3}$A$_{l5}$O$_{12}$ (YAG) is a luminescent down-conversion material which could convert visible photons to near infrared photons. In this work, YAG:Ce$^{3+}$-Yb$^{3+}$ is applied on the front surface of mass-produced mono crystalline Si solar cells. For the coated cells, the external quantum efficiency from the visible to the near infrared is improved, and the energy conversion efficiency enhances from 11.70% to 12.2% under AM1.5G. Furthermore, the phosphor down-conversion effect on the solar cell is characterized by the microwave detected photoconductivity technique on the n-type silicon wafer under the 977 nm excitation. The down-conversion materials improve the average excess carrier lifetime from 22.5 μs to 24.2 μs and the average surface recombination velocity reduces from 424.5 cm/s to 371.6 cm/s, which reveal the significant reduction in excess carrier recombination by the phosphors. DOI:10.1088/0256-307X/33/5/058801 PACS:88.40.-j, 88.40.jj, 84.60.Jt © 2016 Chinese Physics Society Article Text The maximum renewable energy available on earth is the solar energy and the silicon (Si) photovoltaic is a promising, as well as dominating, market technology to harness it. However, its efficiency is low due to the mismatch between the Si band gap and the solar spectrum. The efficiency of Si photovoltaic is wavelength dependent and sensitive at long wavelength where band-to-band transitions happen with good penetration depth, while in conversion of short-wavelength photon energy into electron-hole pairs, the excess energy is wasted through parasitic absorption,[1] which is due to the low absorption depth.[2] The Si solar cell excess carrier's (electron–hole) short lifetime and high thermal generation by short wavelength photons limit the solar cell efficiency. The rare-earth doped phosphors have been attracting attention to modify the photon at short-to-long wavelengths[3] (down-conversion) as per need of photovoltaic efficiency improvement. The one of main efficiency loss in Si solar cells is thermalization of charge carriers at short wavelength photons, resulting in short excess carrier lifetime.[3] This can be overcome by employing the down-converting phosphors, also known as one of third generation approach. Theoretically predicted efficiency improvement in Si solar cells by down-conversion is 38.6% by using un-concentrated sun light. The down-conversion in Ce$^{3+}$-Yb$^{3+}$ doped with different hosts has been suggested to be a suitable solution for the thermalization of charge carriers in Si solar cells, owing to their unique electronic structure properties.[4,5] The ground and excited state energy difference in the Yb$^{3+}$ energy level of two manifolds is 1000 nm, which corresponds to the maximum response in Si solar cells.[6] Further, Yb$^{3+}$ is a good acceptor and emitter with sharp emission line. Ce$^{3+}$ is only donor in all lanthanides, which has allowed electric dipole transitions of 4$f$–5$d$ levels resulting in broad band emission as well as high absorption cross section. Ce$^{3+}$ has the absorption range of 200–500 nm,[7] therefore, Ce$^{3+}$ can be used as an excellent sensitizer to absorb the UV-visible part of the sun spectrum and good donor for excitation energy transfer to Yb$^{3+}$. Having strong ligand field and high luminescence efficiency, Ce$^{3+}$ ion in yttrium aluminum garnet (YAG) host is a suitable donor for Yb$^{3+}$ acceptor in down-conversion materials of Si solar cells.[4] Meanwhile, Ce$^{3+}$ has prominent absorption peaks at 250 nm, 340 nm and 460 nm. It is reported as the UV protector and down-shifter resulting in efficiency stability and long lifetime of devices.[8] The excess carrier lifetime in semiconductors is an average time of carriers at excited state before recombination or trapping.[9] The excess carrier lifetime is an essential evaluating part for analyzing the photovoltaic effects in the semiconductor inline processing, due to the fact that the solar cell efficiency strongly depends upon the recombination rate of excess carriers.[10-12] Where the carriers stay at excited state for a longer time, the better solar cell could be made from that wafer. If the lifetime of carriers is longer than the transit time of the wafer, the photo generated current extraction will be maintained at higher voltages.[13] This substantial characteristic is a key factor for design and development of solar cells at any stage.[14] The YAG:Ce$^{+3}$-Yb$^{+3}$ (Ce$_{0.01}$Yb$_{0.15}$Y$_{0.84}$Al$_{5}$O$_{12}$) phosphor is prepared by the conventional high temperature solid state reaction method and photoluminescence (PL) is measured by an Ideaoptics spectrometer (PG2000-Pro-EX). Mono-crystalline silicon solar cell of size 2.25 cm$^{2}$ was purchased from Sheng De Technology, Weifang New Energy Technology Co., Sidley. The 12 μl PMMA (Poly Methyl Methacrylate, German Tech. Co. Ltd.) and 24 ml acetone are mixed ultrasonically for 30 min at room temperature with 20 mg of phosphor, then the mixture was spin coated at 200 RPM (Smart Coater 100) on the c-Si solar cell. Finally the phosphor-coated c-Si solar cell is dried at 150$^\circ\!$C for 15 min to remove the organic material.[7] Solar cell characteristics are checked by using a source meter (Keithley 2612) and a solar simulator (Newport 91160) under AM1.5G condition at 100 mW/cm$^{2}$ calibrated with standard Si solar cell (Newport 91150). Newport monochromator (74125) integrated with lamp (Newport 74125), the power meter 1918 and Si detector 918D were used to measure the external quantum efficiency (EQE). Charge carrier lifetime and surface recombination velocity response with phosphor is observed on commercial single sided polished n-type Si (100) wafer with 300 μm thickness and 1–3 $\Omega$-cm resistance. The wafer is hydrogenated in a mixture of DI water and hydrofluoric acids in ratio of 3:1 for 10 min, then rinsed by DI water and dried. Phosphor is coated on Si wafer following the same procedure as on the c-Si solar cell. The carrier lifetime and surface recombination velocity is measured by MDPmap (Freiberg Instruments Inline Metrology) by using laser 977 nm, spot size 0.5 mm at the sample distance of 1 mm.

Fig. 1. (a) SEM image of silicon solar cells coated with YAG:Ce$^{3+}$-Yb$^{3+}$. (b) Current density–voltage curve of solar cells under AM1.5 G values.

We observed the photoluminescence spectra of YAG:Ce$^{3+}$-Yb$^{3+}$ under 457 nm excitation (MSL-FN-457) with emission peaks 550 nm and 1030 nm, which can be assigned to the parity allowed transitions from the excited state 5$d$ to the ground state 4$f$ of Ce$^{3+}$ and from the excited state $^{2}\!F_{5/2}$ to the ground state $^{2}\!F_{7/2}$ of Yb$^{3+}$, respectively. The excitation at 457 nm and the emission spectra from 470 nm to 1050 nm prove the energy transfer from Ce$^{3+}$ to Yb$^{3+}$, hence confirming the down-conversion from UV-visible to infrared wavelength. The particle size of the synthesized YAG:Ce$^{3+}$-Yb$^{3+}$ is normally in the 5–15 μm range that can block the incoming light, and finally would reduce the solar cell efficiency.[15] To overcome this deficiency with the phosphors, we manually grind them by mortar and pestle; finally obtained sub-micron particle, the size distribution profile with average size 455 nm (Malvern Instrument, ZEN3690). To probe out the phosphor effects on the Si solar cell performance, the phosphors are spin coated on the front surface (Fig. 1(a)). The surface morphology (SEM, Carel Zeiss Microscopy) of the solar cell depict the non-uniform distribution of phosphors, however most of the phosphor particles are located in the bottom of grooves. Figure 1(b) shows the ($J$–$V$) characteristics of current density of the solar cell versus voltage. The electrical output parameters with and without phosphors are summarized in Table 1. Under the YAG:Ce$^{3+}$-Yb$^{3+}$ effect, the solar cell power conversion efficiency (PCE) is improved from 11.7% to 12.2%. To investigate the YAG:Ce$^{3+}$-Yb$^{3+}$ effects, the excess carrier properties in Si wafer are measured by microwave detected photoconductivity (MDP) techniques. The absorption spectroscopy shows the one of the main absorption peak in Yb appearing at 977 nm[16] and MDPmap built in the 977 nm laser is the most suitable source of Yb excitation. The transient MDP signal at laser power 75 mW is shown in Fig. 2(a). The transient signal shows that without phosphor, the charge carriers start to recombine at 257 mV.[17] while under the phosphor effect, the recombination starts at a potential of 284 mV. The excess carriers are generated by pulse excitation, and the potential depends on the generation rate of charge carriers and the recombination rate.[18] The transient signal is higher with phosphor, indicating that charge carrier generations are higher with low recombination by applying phosphor.[19] The transient signal also gives the evolution of the carrier generation and recombination corresponding to laser light on and off as well as under the phosphor effect. By excitation of 977 nm pulse with pulse duration 110 μs, in the photo response time, the charge carriers reach a steady state position. After the laser pulse is off, the charge carriers recombine and the transient signal returns to the original position as in the dark. The measured recombination time for the bare n-type Si wafer is 22.6 μs, while for the wafer coated with phosphor it is 23.8 μs; meanwhile the carrier densities are $1.008\times10^{17}$ cm$^{-3}$ and $1.013\times10^{17}$ cm$^{-3}$, respectively. With the phosphors, the high carrier generations combined with a long excess carrier lifetime contribute to the high amplitude of the transient signal.

**Table 1.** Electrical output characteristics of the device with and without YAG:Ce$^{3+}$-Yb$^{3+}$ under AM1.5G values.
Device	$V_{\rm oc}$	$J_{\rm sc}$	Fill factor	PCE
Device	(mV)	(mA/cm$^{2}$)	(%)	(%)
Bare	580	30.01	67.20	11.7
Ce-Yb coated	580	31.05	67.67	12.2

However, the above recombination lifetime data is at a single point of the wafer and does not represent the whole wafer. To check the effective carrier lifetime and surface recombination velocity, the whole wafer is spatially resolved by point mapping by using the same laser power with 0.2 mm resolution, and consequently we obtain the average carrier lifetime (Fig. 3). A clear difference between the two mapping images is observable by contrast, as both are being excited homogeneously at the same laser power. In Fig. 3(a), the image area has less darkness than Fig. 3(b). For point comparison, the color at point B in Fig. 3(b) is darker than that at point A in Fig. 3(a), thus point B has longer carrier lifetime and deeper absorption depth than those of point A. The absorption depth increasing in the phosphor-coated Si wafer at point B will increase the carrier's generation, resulting in the carrier lifetime enhancement[20] and lowering in surface recombination velocity. By mapping carrier lifetime of the bare Si wafer, we observe the average carrier lifetime 22.5 μs and average surface recombination velocity 424.5 cm/s, while for the Si wafer coated with phosphor they are 24.2 μs and 371.6 cm/s, respectively. The mapping average carrier lifetime with the phosphor effect is slightly far from a single point that can be due to the non-uniform distribution of phosphor.

Fig. 2. (a) Transient signal of n-type Si wafer with and without phosphor during the turning-on and -off of laser light. (b) Injection intensity dependence carrier lifetime.

The excess carrier lifetime in solar cells strongly depends upon the excess carrier density and, in this regard, it is essential to measure the carrier recombination lifetime at different carrier densities to know the nature of the recombination process. Therefore, the injection intensity dependence carrier recombination is measured (Fig. 2(b)) with laser power from 10 mW to 88 mW. For the bare Si wafer, we observe the effective carrier lifetime from 20.57 μs to 23.21 μs, while the carrier density spans from $1.32\times10^{16}$ cm$^{-3}$ to $1.31\times10^{17}$ cm$^{-3}$. In contrast, the effective carrier lifetime of Si wafer coated with the phosphors varies from 21.36 μs to 24.63 μs and the carrier density ranges from $1.37\times10^{16}$ cm$^{-3}$ to $1.39\times10^{17}$ cm$^{-3}$. The initial carrier lifetime difference at 10 mW laser power between bare Si wafer and wafer coated with phosphor is 0.79 μs, while at 88 mW laser power, the difference is 1.42 μs. Such a difference is attributed to the higher excess carrier generation after the wavelength shifting by phosphors, plus the linear dependence of the down-conversion phosphor PL with the exciting power.[21] The investigated ranges of carrier lifetime with respect to carrier densities lie in the range vital for solar cell production. The highly efficient Si solar cell works in the carrier density range from $1\times10^{15}$ cm$^{-3}$ to $1\times10^{16}$ cm$^{-3}$. The solar cell working under concentrated sun light has carrier density up to $1\times10^{17}$ cm$^{-3}$.[22] The excess carrier lifetime with specific carrier density defines the performance of the photovoltaic device and long carrier lifetime with higher carrier density, which would benefit for high power devices.

Fig. 3. Spatially resolved effective lifetime mapping of n-type Si wafer at area 1 cm$^{2}$ (a) without and (b) with phosphor.

The wafer experiments help us to understand what will happen inside the solar cells. As the short wavelength photons hit the Si solar cell surface, the carrier generation depth is shallow; the recombination rate is high, also the surface recombination velocity is fast. With shallow carrier generation depth, the excess energy by hot carriers will contribute more to the decay of the excited carriers, consequently excess carriers would have short lifetimes. The decrease of the excess carrier lifetime results in low device efficiency. When the solar cell is covered with phosphors, the incoming high energy photons shift to long wavelength photons by phosphors and the excess energy is minimized to generate hot carriers. The low energy photons induce excess carriers in deep generation depth, with low recombination rate and slow surface recombination velocity. The long excess carrier lifetime improves the efficiency of the device. Further, we also measure the EQE of the device with and without phosphor (Fig. 4(a)). In this connection, the measured surface reflectance spectra (Shimadzu UV-3600) of Si solar cells coated with YAG:Ce$^{3+}$-Yb$^{3+}$ (Fig. 4(b)) is clearly reduced, which corresponds to the whole spectra of Ce-Yb, while steep reflectance rises after 1050 nm, attributed to the Si band gap.[23] The reduction in surface reflectance would enhance the light absorption[24,25] in the whole range of solar spectrum and would contribute the power conversion efficiency enhancement in the devices. In fact, YAG:Ce$^{3+}$-Yb$^{3+}$ phosphor also functions as the anti-reflector due to the lower refractive index value than Si or Si$_{3}$N$_{4}$.[26]

Fig. 4. (a) External quantum efficiencies of a bare Si solar cell (black line) and YAG:Ce$^{3+}$-Yb$^{3+}$ coated Si solar cell (red line). (b) Reflectance spectra of bare Si solar cell (black line) and coated with YAG:Ce$^{3+}$-Yb$^{3+}$ (red line).

In conclusion, the conversion efficiency improvement of c-Si solar cells by down-conversion YAG:Ce$^{3+}$-Yb$^{3+}$ phosphor has been demonstrated. The short wavelength photons are converted to long wavelength photons toning with higher spectral sensitivity of c-Si solar cells. The EQE and $I$–$V$ curve of the c-Si solar cell are improved after coating of phosphor, which can be due to the down-conversion of short-wavelength photons to long-wavelength photons. Effective average carrier lifetime mapping and injection intensity dependent carrier lifetime confirm the phosphor down-conversion effects on the c-Si solar cell. The reduction in surface recombination velocity and improvement of carrier lifetime are directly affected by the down-converted phosphor, resulting in the enhancement of device PCE. Our work not only demonstrates the PCE improvement under phosphor effect, but its potential for the UV protection and other fruitful utilization. References Space-separated quantum cutting with silicon nanocrystals for photovoltaic applicationsOptical properties of intrinsic silicon at 300 KAbsolute quantum cutting efficiency of Tb3+-Yb3+ co-doped glassVisible to near infrared conversion in Ce[sup 3+]–Yb[sup 3+] Co-doped YAG ceramicsLuminescent layers for enhanced silicon solar cell performance: Down-conversionPlasmon-enhanced luminescence in Yb[sup 3+]:Y[sub 2]O[sub 3] thin film and the potential for solar cell photon harvestingEnhancing the efficiency of solar cells by down shifting YAG: Ce3+ phosphorsDegradations of silicon photovoltaic modules: A literature reviewLocal Mapping of Generation and Recombination Lifetime in BiFeO ₃ Single Crystals by Scanning Probe Photoinduced Transient SpectroscopyQuantitative carrier lifetime measurement with micron resolutionPersistent Photoconductivity in Strained Epitaxial BiFeO ₃ Thin FilmsConversion Efficiency Enhancement of Multi-crystalline Si Solar Cells by Using a Micro-structured JunctionMinority-Carrier Lifetime and Surface Recombination Velocity in Single-Crystal CdTeSimultaneous phase and size control of upconversion nanocrystals through lanthanide dopingAbsorption and emission spectroscopy in Er3+–Yb3+ doped aluminum oxide waveguidesBiasMDP: Carrier lifetime characterization technique with applied bias voltageContactless electrical defect characterization in semiconductors by microwave detected photo induced current transient spectroscopy (MD-PICTS) and microwave detected photoconductivity (MDP)Suppressing Charge Recombination in ZnO-Nanorod-Based Perovskite Solar Cells with Atomic-Layer-Deposition TiO ₂Theoretical and experimental comparison of contactless lifetime measurement methods for thick silicon samplesDown-conversion process in Tb3+–Yb3+ co-doped Calibo glassesPrepare dispersed CIS nano-scale particles and spray coating CIS absorber layers using nano-scale precursorsMoO ₃ /Ag/Al/ZnO intermediate layer for inverted tandem polymer solar cellsApplication of TiO ₂ with different structures in solar cellsEnhancing the performance of photovoltaic cells by using down-converting KCaGd(PO4)2:Eu3+ phosphors

[1]	TimmermanD, IzeddinI, StallingaP, Yassievich I N and GregorkiewiczT 2008 Nat. Photon. 2 105
[2]	Green M A and Keevers M J 1995 Prog. Photovolt.: Res. Appl. 3 189
[3]	Duan Q, Qin F, Zhao H, Zhang Z and Cao W 2013 J. Appl. Phys. 114 213513
[4]	Ueda J and Tanabe S 2009 J. Appl. Phys. 106 043101
[5]	Richards B 2006 Sol. Energy Mater. Sol. Cells 90 1189
[6]	Lu Y and Chen X 2009 Appl. Phys. Lett. 94 193110
[7]	Shao G, Lou C and Xiao D 2015 J. Lumin. 157 344
[8]	Ndiaye A, Charki A, Kobi A, Kébé C M, Ndiaye P A and Sambou V 2013 Sol. Energy 96 140
[9]	Alexe M 2012 Nano Lett. 12 2193
[10]	GundeP l, Heinz F D, Schubert M C, Giesecke J A and Warta W 2010 J. Appl. Phys. 108 033705
[11]	Bhatnagar A, Kim Y H, Hesse D and Alexe M 2014 Nano Lett. 14 5224
[12]	Li L, Yu D, Wu S L, Wang W, Liu W C, Wu X S and Zhang F M 2015 Chin. Phys. Lett. 32 118401
[13]	Kuciauskas D, Kanevce A, Dippo P, Seyedmohammadi S and Malik R 2015 IEEE J. Photovoltaics 5 366
[14]	Schüler N, Mittelstrass D, Dornich K, Niklas J and Neuhaus H 2010 Photovolt. Spec. Conf. (PVSC) 2010 35th IEEE p 000852
[15]	Wang F, Han Y, Lim C S, Lu Y, Wang J, Xu J, Chen H, Zhang C, Hong M and Liu X 2010 Nature 463 1061
[16]	Strohhöfer C and Polman A 2003 Opt. Mater. 21 705
[17]	Jordan P M, Simon D K, Mikolajick T and Dirnstorfer I 2015 Appl. Phys. Lett. 106 061602
[18]	Berger B, Schüler N, Anger S, Gründig-Wendrock B, Niklas J R and Dornich K 2011 Phys. Status Solidi A 208 769
[19]	Dong J et al 2015 Chin. Phys. Lett. 32 078401
[20]	Schüler N, Hahn T, Dornich K, Niklas J and Gründig-Wendrock B 2010 Sol. Energy Mater. Sol. Cells 94 1076
[21]	Terra I, Borrero-González L, Figueredo T, Almeida J, Hernandes A, Nunes L and Malta O 2012 J. Lumin. 132 1678
[22]	Sinton R and Mankad T 2009 SEMI Draft Document 4772 1
[23]	Sun H, Wei J, Jia Y, Cui X, Wang K and Wu D 2014 Nanoscale Res. Lett. 9 1
[24]	Qing J, Zhong Z F, Liu Y, Li B J and Zhou X 2014 Chin. Phys. B 23 038802
[25]	Zhang T H et al 2012 Chin. Phys. B 21 118401
[26]	Chen Y C, Huang W Y and Chen T M 2011 J. Rare Earths 29 907