Lead tin sulfide (Pb1−xSnxS) nanocrystals: A potential solar absorber material

Journal of Colloid and Interface Science 488 (2017) 246–250

Contents lists available at ScienceDirect

Journal of Colloid and Interface Science journal homepage: www.elsevier.com/locate/jcis

Regular Article

Lead tin sulfide (Pb1
xSnxS) nanocrystals: A potential solar absorber material Yen-Chen Zeng a, Sheng-Fong Sie a, Nipapon Suriyawong a, Belete Asefa Aragaw a, Jen-Bin Shi b, Ming-Way Lee a,⇑ a b

Institute of Nanoscience and Department of Physics, National Chung Hsing University, Taichung 402, Taiwan Department of Electronic Engineering, Feng Chia University, Taichung 40724, Taiwan

g r a p h i c a l a b s t r a c t

a r t i c l e

i n f o

Article history: Received 24 August 2016 Revised 31 October 2016 Accepted 2 November 2016 Available online 2 November 2016 Keywords: Lead tin sulfide Solar cell Quantum dot Successive ionic layer adsorption and reaction Liquid junction

a b s t r a c t We present a new ternary semiconductor absorber material – Pb1
xSnxS – for solar cells. Pb1
xSnxS nanocrystals (NCs) were synthesized using the successive ionic layer adsorption reaction (SILAR) process. Energy-dispersive X-ray spectroscopy revealed the Sn ratio for a sample prepared with five SILAR cycles to be x = 0.55 (i.e. non-stoichiometric formula Pb0.45Sn0.55S). The optical spectra revealed that the energy gap Eg of the Pb1
xSnxS NCs decreased with an increasing number of SILAR cycles n, with Eg = 1.67 eV for the sample with n = 5. Liquid-junction Pb1
xSnxS quantum dot-sensitized solar cells were fabricated using the polysulfide electrolyte. The best cell yielded a short-circuit current density Jsc of 10.1 mA/cm2, an open circuit voltage of 0.43 V, a fill factor FF of 50% and an efficiency of 2.17% under 1 sun. The external quantum efficiency spectrum (EQE) covered a spectral range of 300–800 nm with a maximum EQE of 67% at k = 650 nm. At the reduced light of 0.1 sun, the efficiency increased to 3.31% (with a normalized Jsc = 17.7 mA/cm2) – a respectable efficiency for a new sensitizer. This work demonstrates that Pb1
xSnxS shows potential as a solar cell absorber. Ó 2016 Elsevier Inc. All rights reserved.

1. Introduction

⇑ Corresponding author. E-mail address: [email protected] (M.-W. Lee). http://dx.doi.org/10.1016/j.jcis.2016.11.005 0021-9797/Ó 2016 Elsevier Inc. All rights reserved.

Quantum dot-sensitized solar cells (QDSCs) are a promising lowcost alternative to Si-based photovoltaic devices. The central component of a QDSC is a mesoporous TiO2 photoelectrode (mp-TiO2)

Y.-C. Zeng et al. / Journal of Colloid and Interface Science 488 (2017) 246–250

coated with a layer of semiconductor nanocrystals (or quantum dots) serving as the light-harvesting material. The most commonly used absorber materials for QDSCs have been the binary metal chalcogenides such as CdS, CdSe, PbS, Ag2S, and Sb2S3 so far [1–5]. Ternary materials, in contrast, have been relatively less studied (the widely investigated Cu-In-Se system being an exception). Ternary semiconductor absorbers have several advantages over binary materials: (1) a large absorption coefficient, (2) an energy gap Eg near the optimal Eg of 1.4 eV [6,7], and (3) they possess a tunable Eg which can be accomplished by varying the ratios among the three elements. Recently, the photovoltaic properties of some QDSCs fabricated using ternary materials such as AgSbS2, CuSbS2, CuSbSe2, and CuBiS2, have been reported [8–12]. However, progress in adopting ternary QDSCs has been slow due to the difficulty of material synthesis. Here we present a new ternary absorber material – lead tin sulfide (Pb1
xSnxS). Binary IV – VI metal chalcogenides – PbS and SnS – have been widely studied and have wide applications in infrared and thermoelectric devices [13,14]. Ternary sulfide Pb1
xSnxS, a relatively less explored material, can be formed by alloying SnS with PbS. The Eg of Pb1
xSnxS can be tuned by varying the Sn content x, which is an advantageous property that allows for variable sunlight absorption regions. Wei et al. observed that there is a linear decrease in the Eg of Pb1
xSnxS, with Eg varying from 2.0 (x = 0) to 1.5 eV (x = 0.5) [15]. This Eg of 1.5 eV is close to the optimal Eg (1.4 eV) of a solar absorber [7]. In addition, Pb1
xSnxS has a large optical absorption coefficient a > 104 cm
1 [16,17]. Moreover, the three elements contained in Pb1
xSnxS have the advantages of being earth-abundant, lowcost, and relatively nontoxic. These favorable features – a tunable Eg, a near optimal Eg, a large a and low cost – make Pb1
xSnxS a potential candidate for solar absorber materials. However, to date, there has been no report on the fabrication of solar cells based on Pb1
xSnxS. The only relevant study has been the observation of the photoresponse in Pb1
xSnxS thin film devices [18]. Pb1
xSnxS nanocrystals (NCs) have been grown using hydrothermal, solution-processing and hot-injection methods [15,18,19]. None of these methods, however, are suitable for the fabrication of QDSCs because they do not allow the deposition of ternary NCs into the porous space of the mp-TiO2 electrode in a QDSC. Here we demonstrate the growth of Pb1
xSnxS NCs on an mp-TiO2 electrode using the sequential ionic layer adsorption reaction (SILAR) method. The Sn elemental ratio x was tuned by varying the number of SILAR cycles. Liquid-junction QDSCs were fabricated from the synthesized NCs. We then investigate the dependence of the photovoltaic performance of Pb1
xSnxS QDSCs on the SILAR cycle n. The best cell exhibits an efficiency of 3.3%, a respectable efficiency for the first demonstration of a new solar material. 2. Experimental 2.1. Preparation of TiO2 electrodes Mesoporous TiO2 photoelectrodes (thickness 10 lm) were prepared by spreading transparent anatase TiO2 paste (Dyesol, DSL 30 NR-D, particle size 30 nm) onto a fluorine-doped tin oxide glass substrate (FTO, Pilkington, sheet resistance 7 X/h), followed by heating at 125 °C for 5 min. Prior to the mp-TiO2 coating, a TiO2 compact layer (thickness 60 nm) was coated on the FTO glass by the hot spray method. Finally, a TiO2 scattering layer (Dyesol WER2-O, particle size 250 nm, thickness 6–8 lm) was coated onto the electrode, finished by heating in air in a tube furnace at 500 °C for 15 min. 2.2. Synthesis of Pb-Sn-S nanocrystals Pb-Sn-S NCs were synthesized using a two-stage SILAR process. The growth involved three steps: (1) SILAR growth of Pb-S QDs, (2)

247

SILAR growth of Sn-S QDs on top of the Pb-S QDs, and (3) post annealing to produce the ternary phase. In brief, each Pb-S SILAR cycle comprised dipping an mp-TiO2 electrode into a 25 °C 0.05 M Pb(NO3)2 aqueous solution for 30 s, rinsing in DI water, then letting it dry in air at 65 °C. The Pb2+-coated electrode was then dipped into a 25 °C, 0.1 M Na2S methanol solution for 30 s, rinsed and dried as above. This completed a Pb-S SILAR cycle. The process was repeated n times to produce a Pb-S(n) film. For the Sn-S SILAR, the Pb-S coated electrode was dipped into a 25 °C, 0.1 M SnSO4 aqueous solution for 20 s, rinsed and dried as above, then dipped into a 0.1 M Na2S methanol solution for 30 s. The number of Sn-S SILAR cycles was also equal to n. After completing the SILAR process, the Pb-S/Sn-S coated TiO2 film was annealed in a N2-flowing tube furnace at 190 °C for 4 min. Ternary Pb-Sn-S QDs formed after annealing. 2.3. Solar cell assembly The Pb-Sn-S QD-coated TiO2 electrode was assembled with a counterelectrode into a sandwich configuration using a Parafilm spacer (thickness 190 lm). Two types of counterelectrodes – Pt and Au – were used. The polysulfide electrolyte consisted of 0.25 M Na2S, 1 M S, 0.2 M KCl and 0.1 M KI in ethanol/water (7:3, vol.) solution. 2.4. Structural and photovoltaic characterization The structure of the grown material was characterized using a transmission electron microscope (TEM, JEOL-JEM 2010), a fieldemission scanning electron microscope (SEM, ZEISS Ultra plus), and an X-ray diffractometer (XRD, Bruker, D8 SSS). Optical spectra were recorded using a Hitachi U-2800A UV–vis spectrophotometer. Photocurrent-voltage (I-V) curves were recorded with a Keithley 2400 source meter connected to a 150 W Oriel Xe lamp with a band-pass AM1.5 simulating filter. The external quantum efficiency (EQE) spectra were recorded with an Acton monochromator coupled to a 250 W tungsten halogen lamp. A metal mask defined the active area of the cell to be 3 mm  3 mm. 3. Results and discussion 3.1. TEM Fig. 1(a) shows a TEM image of a bare TiO2 film. The TiO2 particles are rectangular in shape with round corners and an average length of 30–40 nm. Fig. 1(b) shows a TEM image of Pb-Sn-S QDs coated on a TiO2 electrode. It can be seen that many QDs are distributed randomly over the TiO2 surface without observable aggregation with a wide distribution of particle size ranging from 5 to 17 nm and an average diameter of 11 nm. 3.2. XRD Fig. 2(a) shows an XRD pattern of the Pb-Sb-S QDs for SILAR cycle n = 5. For comparison, the peaks of the PbS are shown in the bottom panel. The Pb-Sb-S QDs have a cubic crystal structure with a lattice constant of a = 5.866 Å. The structure is identical to that of PbS (a = 5.938 Å) but the lattice constant is smaller. In addition, the Pb-Sb-S XRD peaks shifted to higher angles relative to PbS, e.g., the (2 0 0) plane shifted from 30.08° (PbS) to 30.44° (Pb-Sb-S). Since Sn2+ has a smaller atomic radius than that of Pb2+, the incorporation of Sn into the PbS host matrix leads to shrinkage of the crystallographic unit cell, resulting in a reduced lattice constant. The XRD spectrum was used to calculate the QD particle size using Scherrer’s formula. Analyzing the strongest (2 0 0) peak yielded an

248

Y.-C. Zeng et al. / Journal of Colloid and Interface Science 488 (2017) 246–250

Fig. 1. TEM images of (a) a bare TiO2 electrode and (b) a Pb-Sn-S QD-coated TiO2 electrode.

QDs, which is a favorable property for a solar material. The intercepts of the Tauc plot, (Ahm)2 vs. hm, yield Eg = 1.77, 1.67, 1.52 eV for n = 4, 5, 6, respectively (see Fig. 3(c)). The decreases in Eg with increasing SILAR cycle n is attributed to the quantum-size effect, in that a smaller particle exhibits a larger Eg. 3.4. Photovoltaic performance

Fig. 2. (a) XRD spectrum and (b) EDS data for Pb-Sn-S QDs prepared with SILAR cycles n = 5.

average QD size of 13 nm, which is in good agreement with the TEM results shown in Fig. 1(b). The elemental ratios between Pb, Sn and S were analyzed using energy-dispersive X-ray spectroscopy (EDS). Fig. 2(b) shows the EDS data averaged over ten different sample positions, yielding the atomic ratio: Pb:Sn: S = 0.220:0.271:0.509. This implies that the non-stoichiometric formula is Pb0.45Sn0.55S for the SILAR cycle n = 5 sample. It also reveals a slightly higher Sn composition than that of Pb. 3.3. Optical spectra Fig. 3(a) and (b) shows the transmittance T(m) and absorbance A (m) of three Pb-Sn-S films with n = 4, 5, 6. It can be seen that A(m) increased with increasing n, indicating increasing optical absorption with the deposition of an increase in the amount of QDs with increasing number of SILAR cycles. In addition, T(m) is extremely low (1% for n = 6) over the short-wavelength spectral range of 400–600 nm, indicating the strong light-absorbing ability of the

The photovoltaic performance of SILAR-prepared QDSCs is highly sensitive to the number of SILAR cycles n. Fig. 4(a) shows the I-V curves of Pb-Sn-S QDSCs fabricated with various number of SILAR cycles n. The photovoltaic parameters are listed in Table 1. For the three samples using Pt counterelectrodes (sample Nos. 1– 3), the efficiency g increased with n, reaching a maximum g = 1.20% at n = 5. When n > 6, g started to decrease again. The initial increase in g with n is due to an increasing amount of QDs being deposited on the electrode, resulting in increased light absorption. After reaching the optimal n, further loading of QDs could impede the electrolyte flow, leading to a decreased g. When Au replaced Pt as the counterelectrode, the g increased to 1.68% (sample No. 4, see Fig. 4(b)), a 40% improvement over the best Pt cell (sample No. 3). The QDs were further coated with a ZnS layer in order to reduce carrier recombination. The ZnS coating increased the g to 2.17% (sample No. 5), an improvement of 29% over the uncoated sample (No. 4). The positive effect of ZnS coating indicates strong carrier recombination in the Pb-Sn-S solar cells, which is commonly observed in NCs synthesized using the SILAR method, which inherently contain a large number of defects acting as recombination centers. The best cell yielded a short-circuit current density Jsc of 10.1 mA/cm2, an open-circuit voltage Voc of 0.43 V, a fill factor FF of 50%, and an efficiency g of 2.17% under 1 sun. Fig. 5 shows the I-V curves under various reduced sun intensities. Table 2 lists the photovoltaic data. The performance improved with a reduction in light intensity. The Jsc data in the bracket in Table 2 represent the Jsc normalized to 1 sun. The photovoltaic parameters in Table 2 exhibit the following features: (a) the normalized Jsc increased from 10.1 to 17.7 mA/cm2, (b) FF increased from 50.0 to 53.5%, and (c) g increased from 2.17 to 3.31% as the sun intensity decreased from 1 to 0.1 sun. Meanwhile, Voc decreased from 0.43 to 0.35 V. The g improved by 53% with reduced sun intensity. The most pronounced increase was with the normalized Jsc, which increased by 75% as the sun intensity was reduced. The increased efficiency under reduced light intensities is attributed to the reduction in carrier recombination. Photoelectron-hole recombination could occur at the interfaces between QD/electrolyte, TiO2 particles/electrolyte and FTO/

249

Y.-C. Zeng et al. / Journal of Colloid and Interface Science 488 (2017) 246–250

Fig. 3. Optical spectra: (a) transmission, (b) absorbance and (c) (Ahm)2 vs. hm plots of Pb-Sn-S QDs with SILAR cycle n = 4,5,6.

Fig. 4. I-V curves of Pb-Sn-S QDSCs: (a) for various SILAR cycles and (b) for Pt and Au counterelectrodes.

Table 1 Photovoltaic performance of Pb-Sn-S QDSCs with various numbers of SILAR cycles and counter electrodes. Electrolyte: polysulfide. Sample No.

SILAR cycles

Counter electrode

Jsc (mA/cm2)

Voc (V)

FF (%)

g (%)

1 2 3 4 5

4 5 6 6 6

Pt Pt Pt Au Au/ZnS

3.37 7.17 3.89 8.73 10.10

0.36 0.38 0.38 0.39 0.43

47.9 44.0 41.2 49.2 50.0

0.58 1.20 0.61 1.68 2.17

cles prepared using solution processing methods such as SILAR or chemical bath deposition, because they inherently contain a large number of surface defects acting as recombination centers. At a reduced light intensity, the number of photocarriers generated by the sunlight is reduced, which leads to a reduction in carrier recombination, and, hence, improves photovoltaic performance. In other words, the g measured under a reduced light intensity represents the maximal g that can be achieved by the solar cell. The resultant reduction in Voc with reduced light intensity can also be explained by the reduction in carrier number. The theoretical upper limit of Voc for a QDSC is Voc = EF
Eredox, where EF is the quasi Fermi level of TiO2, and Eredox is the redox level of the liquid electrolyte. The Fermi level EF of TiO2 is a function of the electron density nCB in the conduction band (CB) according to [20]: Fig. 5. I-V curves under various reduced sun intensities.

electrolyte of a QDSC, resulting in a loss of the photocurrent and cell efficiency. This phenomenon is especially severe for nanoparti-

EF ¼ kB T  lnðnCB Þ:

ð1Þ

A reduced light intensity generates a smaller carrier density nCB, leading to a lower EF and hence, a lower Voc, as revealed in Table 2. Fig. 6 displays the EQE spectrum. The spectrum covers the spectral range of 300–800 nm with a maximum EQE = 67% at

250

Y.-C. Zeng et al. / Journal of Colloid and Interface Science 488 (2017) 246–250

Table 2 Photovoltaic performance of Pb-Sn-S QDSCs under various sun intensities. The Jsc in the brackets represents the Jsc normalized to 1 sun. Sun intensity

Jsc (mA/cm2)

Voc (V)

FF (%)

g (%)

100% 50% 23% 10%

10.10 5.93 (11.86) 3.39 (14.7) 1.77 (17.7)

0.43 0.41 0.37 0.35

50.00 50.53 51.52 53.49

2.17 2.46 2.81 3.31

as the light intensity was decreased due to a reduction in carrier recombination. Potential higher efficiencies are expected in QDSCs using solid-state electrolytes (such as spiro-OMeTAD). The present work shows that Pb-Sn-S is an attractive candidate for solar absorber materials. Acknowledgments The authors wish to express our gratitude for financial support from the Ministry of Science and Technology of the Republic of China under grant No. MOST 103-2112-M-005-004-MY3. References

Fig. 6. EQE spectrum of the best Pb-Sn-S QDSC.

k = 650 nm. The EQE onset corresponds to the energy where photon-to-electron conversion occurs; namely, energy gap Eg. The onset in Fig. 6 yields an Eg = 1.59 eV (k = 780 nm), which is in good agreement with the optical Eg (1.67 eV) shown in Fig. 3. The area under the EQE curve represents the photocurrent produced by the solar cell according to:

Z

J ph ¼ e

UðkÞEQEðkÞdk;

ð2Þ

where e is the elementary charge and UðkÞ is the solar photon flux, which can be found in the literature. Integrating the EQE curve yields Jsc = 15.6 mA/cm2, which is in agreement with the normalized Jsc under reduced sun intensities, as indicated in Table 2. We now summarize the notable results obtained in this work. Pb-Sn-S QDs were synthesized using a low-cost chemical method. The material exhibited large optical absorption, a near-optimal Eg, and, probably the most important feature, a tunable Eg. The cells exhibited an g of 2.17% under 1 sun. Under a reduced 0.1 sun, the g increased to 3.31%, a respectable g, considering that this is the first work on Pb-Sn-S solar cells. Since work on Pb-Sn-S is still in the beginning stages, higher efficiencies are expected with more research carried out in the future. 4. Conclusions We demonstrated the functioning of liquid-junction Pb-Sn-S QDSCs fabricated on mp-TiO2 electrodes using the SILAR method. The optimal samples were prepared using six SILAR cycles with Au counter electrodes and a ZnS coating. The best cell yielded a respectable g of 3.31% under 0.1 sun. The efficiencies increased

[1] G. Larramona, C. Chone, A. Jacob, D. Sakakura, B. Delatouche, D. Pere, X. Cieren, M. Nagino, R. Bayon, Nanostructured photovoltaic cell of the type titanium dioxide, cadmium sulfide thin coating, and copper thiocyanate showing high quantum efficiency, Chem. Mater. 18 (2006) 1688–1696. [2] N. Guijarro, T. Lana-Villarreal, I. Mora-Seró, J. Bisquert, R.J. Gómez, CdSe quantum dot-sensitized TiO2 electrodes: effect of quantum dot coverage and mode of attachment, J. Phys. Chem. C 113 (2009) 4208–4214. [3] V. González-Pedro, C. Sima, G. Marzari, P.P. Boix, S. Giménez, Q. Shen, T. Dittrich, I. Mora-Seró, High performance PbS Quantum Dot Sensitized Solar Cells exceeding 4% efficiency: the role of metal precursors in the electron injection and charge separation, Phys. Chem. Chem. Phys. 15 (2013) 13835– 13843. [4] A. Tubtimtae, K.L. Wu, H.Y. Tung, M.W. Lee, G.J. Wang, Ag2S quantum dotsensitized solar cells, Electrochem. Commun. 12 (2010) 1158–1160. [5] Y. Itzhaik, O. Niitsoo, M. Page, G. Hodes, Sb2S3-sensitized nanoporous TiO2 solar cells, J. Phys. Chem. C 113 (2009) 4254–4256. [6] S.I. Boldish, W.B. White, Optical band gaps of selected ternary sulfide minerals, Am. Mineral. 83 (1998) 865–871. [7] T. Zdanowicza, T. Rodziewiczb, M. Zabkowska-Waclawek, Theoretical analysis of the optimum energy band gap of semiconductors for fabrication of solar cells for applications in higher latitudes locations, Sol. Energy Mater. Sol. Cells 87 (2005) 757–769. [8] Y.R. Ho, M.W. Lee, AgSbS2 semiconductor-sensitized solar cells, Electrochem. Commun. 26 (2013) 48–51. [9] Y.C. Choi, E.J. Yeom, T.K. Ahn, S.I. Seok, CuSbS2-sensitized inorganic-organic heterojunction solar cells fabricated using a metal-thiourea complex solution, Angew. Chem. Int. Ed. 54 (2015) 4005–4009. [10] A.W. Welch, L.L. Baranowski, P. Zawadzki, S. Lany, C.A. Wolden, A. Zakutayev, CuSbSe2 photovoltaic devices with 3% efficiency, Appl. Phys. Express 8 (2015) 082301. [11] N. Suriyawong, B. Aragaw, J.B. Shi, M.W. Lee, Ternary CuBiS2 nanoparticles as a sensitizer for quantum dot solar cells, J. Colloid Interface Sci. 473 (2016) 60– 65. [12] D. Colombara, L.M. Peter, K.D. Rogers, J.D. Painter, S. Roncallo, Formation of CuSbS2 and CuSbSe2 thin films via chalcogenisation of Sb–Cu metal precursors, Thin Solid Films 519 (2011) 7438–7443. [13] A.L. Rogach, A. Eychmuller, S.G. Hickey, S.V. Kershaw, Infrared-emitting colloidal nanocrystals: synthesis, Assembly Spectrosc. Appl. Small 3 (2007) 536–557. [14] S. Johnsen, J. He, J. Androulakis, V.P. Dravid, I. Todorov, D.Y. Chung, M.G. Kanatzidis, Nanostructures boost the thermoelectric performance of PbS, J. Am. Chem. Soc. 133 (2011) 3460–3470. [15] H. Wei, Y.J. Su, S.Z. Chen, Y. Lin, Z. Yang, H. Sun, Y.F. Zhang, Synthesis of ternary PbxSn1
xS nanocrystals with tunable band gap, CrystEngComm 13 (2011) 6628–6631. [16] B. Thangaraju, P. Kaliannan, Polycrystalline lead tin chalcogenide thin film grown by spray pyrolysis, Cryst. Res. Technol. 35 (2000) 71–75. [17] D.M. Unuchak, K. Bente, G. Kloess, W. Schmitz, V.F. Gremenok, V.A. Ivanov, V. Ukhov, Structure and optical properties of PbS-SnS mixed crystal thin films, Phys. Status Solidi C 6 (2009) 1191–1194. [18] X. Liu, Y. Li, B. Zhou, D.Q. Wang, A.N. Cartwright, M.T. Swihart, Formation of IVVI alloy nanocrystals for application in solution-processed optoelectronic devices: the case of Pb1
xSnxS, Chem. Mater. 25 (2013) 4409–4415. [19] R.B. Soriano, C.D. Malliakas, J.S. Wu, M.G. Kanatzidis, Cubic form of Pb2
xSnxS2 stabilized through size reduction to the nanoscale, J. Am. Chem. Soc. 134 (2012) 3228–3233. [20] J. Nelson, The Physics of Solar Cells, Imperial College Press, London, 2003. p. 59.