LETTER pubs.acs.org/NanoLett
Toward Interaction of Sensitizer and Functional Moieties in Hole-Transporting Materials for Efficient Semiconductor-Sensitized Solar Cells Sang Hyuk Im,†,§ Choong-Sun Lim,†,§ Jeong Ah Chang,†,§ Yong Hui Lee,†,§ Nilkamal Maiti,† Hi-Jung Kim,† Md. K. Nazeeruddin,‡ Michael Gr€atzel,‡ and Sang Il Seok*,† †
KRICT-EPFL Global Research Laboratory, Advanced Materials Division, Korea Research Institute of Chemical Technology, 141 Gajeong-ro, Yuseong-gu, Daejeon 305-600, Republic of Korea ‡ Laboratory of Photonics and Interfaces, Institute of Chemical Sciences and Engineering, Ecole Polytechnique Federale de Lausanne (EPFL), Station 6, CH-1050 Lausanne, Switzerland
bS Supporting Information ABSTRACT: Sb2S3-sensitized mesoporous-TiO2 solar cells using several conjugated polymers as hole-transporting materials (HTMs) are fabricated. We found that the cell performance was strongly correlated with the chemical interaction at the interface of Sb2S3 as sensitizer and the HTMs through the thiophene moieties, which led to a higher fill factor (FF), open-circuit voltage (Voc), and short-circuit current density (Jsc). With the application of PCPDTBT (poly(2,6-(4,4-bis-(2-ethylhexyl)4H-cyclopenta[2,1-b;3,4-b0 ]dithiophene)-alt-4,7(2,1,3-benzothiadiazole)) as a HTM in a Sb2S3-sensitized solar cell, overall power conversion efficiencies of 6.18, 6.57, and 6.53% at 100, 50, and 10% solar irradiation, respectively, were achieved with a metal mask. KEYWORDS: Sensitized solar cells, heterojunction solar cells, functional moieties, Sb2S3, mesoporous TiO2, hole conducting materials
S
olar cells are promising candidates for carbon-free energy sources to maintain our standard of living. Unfortunately, until now, only a very small percentage of energy has been produced from sunlight via solar cells, mainly because of their relatively high cost as compared to other energy sources. The performances of solar cells that can be inexpensively fabricated in mass production still remain lower than those of conventional Sibased solar cells. As well-known, the power conversion efficiency (PCE) of a solar cell is a product of its open-circuit voltage (Voc), short-circuit current density (Jsc), and fill factor (FF). Therefore, the only way to improve a PCE is to increase its Voc, Jsc, and the FF, simultaneously. An increase in Voc can be achieved by controlling many factors such as defects, traps, the rates of electron injection, a reduction in dark current, and the contact layers at the interface. One way to improve Jsc is to absorb a greater fraction of the incident light within the solar spectrum. Another way to increase the PCE is to increase the FF. Dye-sensitized solar cells (DSSCs) consist of three-components; a dye or inorganic sensitizer is deposited in thin layers or nanoparticles onto mesoporous n-type materials such as TiO2, and the rest of the pores are filled with p-type hole-transporting materials (HTMs). Accordingly, the FF in sensitized solar cells is attenuated by the total series resistance of the cell, including the resistance of the substrate and counter electrode, and by the charge transport resistance through the photoanode and HTM. The use of inorganic semiconductors as sensitizers in place of dyes in dye-sensitized solar cells (DSSCs)1,2 may offer benefits as r 2011 American Chemical Society
results of their excellent optical properties,3 high extinction coefficients,4 and large intrinsic dipole moments.5 Much progress has been made thus far in improving the PCE of fully solid-state semiconductor-sensitized solar cells.611 However, these cells still suffer from a low FF because of a charge transfer problem between the sensitizer and the solid HTMs, as compared to that of solid-state dye-sensitized solar cells,12 and this results in a low PCE. For example, it has been shown that mesoporous (mp)TiO2/PbS/spiro-OMeTAD (2,20 ,7,70 -tetrakis(N,N-di-p-methoxyphen ylamine)-9,90 -spirobifluorene) and mp-TiO2/CdSe/ spiro-OMeTAD have a FF of 57% and 55%, respectively, under full-sun illumination, yielding an overall PCE of less than 2%.8,9 Other studies have also revealed that Sb2S3/CuSCN shows a FF of 48.8% while Sb2S3/spiro-OMeTAD shows a FF of 48% under full-sun illumination.6,10 Recently, we reported that overall efficiencies of over 5% with a FF of around 70% were obtained for a Sb2S3-deposited mp-TiO2 electrode, using poly(3-hexylthiophene) (P3HT) as the HTM for sensitizer regeneration.7 Stimulated by these results, we fabricated inorganicorganic heterojunction solar cells based on mesoscopic TiO2, using Sb2S3 as a sensitizer in contact with PCPDTBT, PCDTBT (poly(N-900 -heptadecanyl-2,7carbazole-alt-5,5-(40 ,70 -di-2-thienyl-20 ,10 ,30 -benzothiadiazole)), and PTAA (poly(triarylamine)). In inorganicorganic heterojunction Received: July 30, 2011 Revised: September 26, 2011 Published: September 30, 2011 4789
dx.doi.org/10.1021/nl2026184 | Nano Lett. 2011, 11, 4789–4793
Nano Letters
LETTER
Scheme 1. Energy Level Diagram of the Corresponding Materials Used in Our Devices
semiconductor-sensitized solar cells, fast charge transport from the inorganic sensitizers to the neighboring n-type TiO2 and p-type organic hole conductors is crucial to the improvement of cell performance. To date, only a few reports can be found in the literature that mention the formation of an intimate contact at the inorganic semiconductor/organic hole conductor interface for interfacial charge transfer.11 Herein we demonstrate the important role played by the functional group in HTMs, which can be strongly anchored with semiconductor sensitizers, in the improvement of the FF. The interaction between Sb2S3 and organic HTMs has been systematically investigated using Raman measurements because resonance Raman spectroscopy has proved to be a powerful tool of probing the coordinative interfacial reactions in Sb2S3/HTM hybrid systems.13 Our results suggest that thiophene moieties, which are capable of exhibiting a strong bond between HTM and Sb2S3 sensitizers, represent a key factor in the fabrication of highefficiency semiconductor-sensitized inorganicorganic heterojunction solar cells. Among the available metal chalcogenide semiconductors, Sb2S3 was an attractive candidate as a light absorber in a solar cell because it typically exhibits a high extinction coefficient (α = 105 cm1 in the visible region) and appropriate band gap (Eg = ∼1.7 eV).14 In a typical device fabrication procedure, the Sb2S3 sensitizer was deposited by direct growth onto a mp-TiO2 surface through chemical bath deposition (CBD), which provides high surface coverage and strong anchoring to the electrode. As-prepared Sb2S3 was thermally annealed at 300 °C for 20 min in a nitrogen atmosphere to convert it into the crystalline stibnite with the formation of well-developed and isolated particles.7 Scheme 1 shows a schematic of the energy level diagram of the corresponding device components, depicting the conduction band of TiO2, the conduction band and valence band of Sb2S3, and the position of the highest occupied molecular orbital (HOMO) of the HTMs. In this scheme, all the HTMs are energetically conducive to hole injection, so that the holes separated from electronhole pairs, which are created by light absorption of the Sb2S3 sensitizer, are transferred to the HTM phase and are collected at opposing electrodes; the electrons are then transported to the TiO2. The external quantum efficiency (EQE) of these cells is defined as the ratio of the number of electronhole pairs collected at the electrodes to the number of incident photons. Figure 1a shows the spectral dependence of EQE with different HTMs for the mp-TiO2/Sb2S3/HTM devices. P3HT, PCPDTBT, PCDTBT, and PTAA as the HTM in the mp-TiO2/ Sb2S3/HTM configuration are denoted as device 1, 2, 3, and 4,
Figure 1. (a) External quantum efficiency (EQE) curves for TiO2/ Sb2S3/HTM/Au fabricated with P3HT, PCPDTBT, PCDTBT, and PTAA as HTMs. The Sb2S3 layers were formed by chemical bath deposition for 1.5 h and then annealed at 300 °C for 30 min in N2. (b) Current densityvoltage (JV) corresponding to (a), i.e., for mpTiO2/Sb2S3/HTMs/Au.
respectively. In these experiments, a relatively low level of Sb2S3 was deposited onto mp-TiO2 to exclude the possible difference among HTMs in terms of pore-filling, and the performances of the four different HTMs were preliminarily evaluated. It was generally observed that when there was insufficient penetration of HTMs into the pores of the mesoporous TiO2 electrode, the EQE values in the short-wavelength region below 500 nm were greatly reduced (see Figure S1 in the Supporting Information). As can be seen in Figure 1a, the four devices had similar photoresponses, except for PTAA. Devices 1, 2, and 4 had relatively higher EQEs than device 3. In the case of P3HT as the HTM, the EQE spectrum in Figure 1a begins to degrade in the wavelength range 450650 nm, which matches the maximum absorption 4790
dx.doi.org/10.1021/nl2026184 |Nano Lett. 2011, 11, 4789–4793
Nano Letters
LETTER
Figure 2. (a) Schematic illustration of the crystal structure of Sb2S3. (b) Schematic representation of the formation of pentacoordination with the neighboring thiophene moieties. (c) Raman spectra of Sb2S3 and monothiophene (MT)-, bithiophene (BT)-, and P3HT-treated Sb2S3.
range of P3HT. From analysis of the absorption spectra (see Figure S2a in the Supporting Information), we infer that the depression in the EQE spectrum is the result of absorption loss by P3HT. In other words, this loss in EQE is caused by the incomplete charge carrier transport that took place because the charge carriers generated in P3HT by the absorption of light could not be efficiently transported into Sb2S3, and consequently, the carriers could not be prevented from recombining. The absence of absorption loss by P3HT below a wavelength of 450 nm in the EQE spectrum might be the result of complete absorption by Sb2S3 or weak absorption by P3HT below a wavelength of 450 nm because the mp-TiO2/Sb2S3 film fully absorbs light below 450 nm. Moreover, the photoluminescence (PL) spectra in Figure S2b (Supporting Information) confirm that the generated electrons in P3HT could be transported into Sb2S3 but be incompletely transported. Figure 1b shows the JV curves corresponding to those in Figure 1a under full-sun illumination. Devices 1 and 2 exhibit a Voc of 525.2 and 555.5 mV, a Jsc of 9.5 and 11.9 mA/cm, a FF of 61.6 and 67.0%, and a PCE of 3.1 and 4.4%, respectively. These two cells display a respectable FF. On the other hand, devices 3 and 4 give a Voc of 545.4 and 454.6 mV, a Jsc of 8.0 and 11.1 mA/cm, a FF of 32.8 and 34.9%, and a PCE of 1.4 and 1.8%, respectively. Here, the highest photocurrent expected from EQE for device 4 is not obtained, suggesting a poor charge transfer between sensitizer and PTAA, especially under full sun, because basically the integration of EQE should be matched to Jsc of JV curves. Hence, the overall performances of devices 3 and 4 are quite low compared to those of devices 1 and 2. In sensitized solar cells consisting of an n-type semiconductor, a light absorber, and a p-type semiconductor, as mentioned in the previous section, the electron hole pairs generated by the light absorption of the absorber
(Sb2S3) are injected separately into the TiO2 and HTM. As can be seen in Scheme 1, from the viewpoint of the energy structure diagram, all the devices demonstrate similar charge separation and transfer. Moreover, the four different HTMs show comparable values for hole mobility (see Table S1 in the Supporting Information). Therefore, the high performances of devices 1 and 2 may be ascribed to the intimate interfacial contact that arises as a result of their strong chemical interactions. To better understand why P3HT and PCPDTBT show a higher fill factor than PCDTBT and PTAA, and why they operate more stable than solar cells composed of CuSCN or spiroOMeTAD, regardless of the intensity of sun illumination, we focused on the thiophene moieties in the HTMs as one possible cause, and did so by characterizing the chemical bonding between Sb2S3 and the HTMs. It should be noted that unlike other metal chalcogenides, Sb2S3 has unique properties such as a low melting temperature and a parallel chainlike structure. The structure of Sb2S3 is shown schematically in Figure 2a. The parallel chainlike stibnite Sb2S3 can consist of two types of Sb atoms and three types of S atoms.15,16 One Sb atom is formally tricoordinated to the three S atoms in a plane by strong covalent bonds, and the other Sb atom is pentacoordinated to the five S atoms, where three S atoms are connected by strong covalent bonds within the same chain, and two S atoms are connected by weak van der Waals bonds to the next parallel chain, as shown in Figure 2a, b.15,16 Here, we should consider that a material that is reactive to Sb2S3 could preferentially attack the relatively weaker long SbS bonds, and a bifunctional material could chelate to the Sb2S3, instead of the two weak SbS bonds, thus forming a more stable complex. Examining P3HT from this perspective, we can then consider that P3HT is composed of thiophene moieties, which can be easily combined with loosely bonded SbS layers. We can 4791
dx.doi.org/10.1021/nl2026184 |Nano Lett. 2011, 11, 4789–4793
Nano Letters
LETTER
Table 1. Summary of Short-Circuit Current Density (Jsc), Open-Circuit Voltage (Voc), Fill Factor (FF), and Overall Conversion Efficiencies Obtained from the Device Shown in Figure 3aa light power (mW/cm2)
Jsc (mA/cm2)
Voc (mV)
100
15.3
616
65.7
6.18
50
8.3
585
68.0
6.57
10
1.7
535
70.5
6.53
FF (%)
efficiency (%)
a
Masks (0.096 cm2) made of thin metal were attached to each cell before measurement.
Figure 3. (a) Current densityvoltage (JV) curves for mp-TiO2/ Sb2S3/PCPDTBT/Au. The Sb2S3 layers were formed by chemical bath deposition for 2.2 h (active area = 0.16 cm2, mask size = 0.096 cm2). Inset: Summary of short-circuit current density (Jsc), open-circuit voltage (Voc), fill factor (FF), and overall conversion efficiencies. (b) Spectral dependence of external quantum efficiency (EQE) of mp-TiO2/ Sb2S3/PCPDTBT solar cell.
then speculate that the bifunctional P3HT could form stable complexes by replacing the weak long SbS bonds. To examine whether the thiophene moieties in P3HT form stable complexes by replacing the long SbS bonds, as we have speculated, we compared the Raman spectra of monothiophene-treated Sb2S3 and bithiophene-treated Sb2S3 as a model system (Figure 2c). The characteristic Raman peaks of the long SbS vibration modes corresponded to 180 cm1 (the SSbS antisymmetric bending vibration) and 277 cm1 (the SbS stretching vibration of the longer bond).17,18 These Raman spectra clearly show that the bithiophene-treated Sb2S3 (Sb2S3/BT) significantly depressed the characteristic Raman peaks of the long SbS bonds, while the monothiophene-treated Sb2S3 (Sb2S3/MT) did not. This result clearly confirms that the bifunctional thiophene group is able to replace the weak long SbS bonds. As we predicted, the P3HT-treated Sb2S3 also strongly depressed the characteristic Raman peaks of the long SbS bond, as did bithiophene. We may then consider that the P3HT has neighboring thiophene moieties which are capable of bidentately chelating with Sb2S3 because the thiophene moieties directed toward the Sb atom may result in the formation of a thermodynamically favorable fivemembered ring structure (see Figure 2b). It should be noted that the pentacoordination by the thiophene moieties in P3HT is more thermodynamically stable than the tetracoordination by the monothiophene moiety; the monothiophene does not seem to bond efficiently to Sb atoms. Moreover, the bidentate chelation of P3HT to Sb2S3 could passivate the surface of Sb2S3 and, as a consequence, retard the degradation of P3HT that is in contact with Sb2S3 through oxygen. Here, the first neighboring P3HT layer and the other P3HT layer need to be distinguished, because the conformation of the first neighboring P3HT layer leads to form bidentate-chelating with Sb atom in Sb2S3 through the rearrangement of thiophene moieties. It should be noted that the bidentately chelated thiophene moieties with Sb atom cannot be
identified as in-plane or twist. Our hypothesis was further confirmed by a comparison of PCPDTBT and PCDTBT having the same thiophene moieties. PCPDTBT has continuous and neighboring thiophene moieties, which can form pentacoordinated complexes with Sb2S3, but PCDTBT possesses only nonneighboring thiophene moieties. In addition, the unique interaction between Sb2S3 and the thiophene moieties in HTM might explain why the mp-TiO2/Sb2S3/P3HT inorganicorganic heterojunction solar cell functions stably in an air atmosphere than do organic solar cells fabricated using P3HT and PCPDTBT as donor materials. Figure 3a shows the currentvoltage curves of optimized device 2 (TiO2/Sb2S3/PCPDTBT), as measured under simulated AM 1.5G illuminations over a range of light intensities. The Voc, Jsc, FF, and PCE are listed in Table 1. At low light levels of 10 and 50 mW/cm2, we observe PCE values of 6.53 and 6.57%. The PCE under full-sun illumination (100 mW/cm2) appears to be slightly reduced to 6.18%, though this value is the highest that was recorded among the reported semiconductor-sensitized cells. The EQE spectrum (Figure 3b) shows a plateau of over 70% from 370 to 590 nm, with the maximum of 80.3% at 410 nm. Unlike the case of P3HT shown in Figure 1a, this device does not exhibit a depression at a wavelength of 650 nm because the band gap of PCPDTBT is positioned at approximately 1.4 eV.16 Accordingly, the device fabricated from PCPDTBT can give a performance that is superior to that of P3HT because of the higher current density that results from the decreased absorption loss. Here, we can estimate that this device should deliver over 15 mA/cm2 with integration over the solar spectrum of the EQE shown in Figure 3b. This means that the lowering of the PCE under full sunlight can mainly be attributed to a slight loss of photocurrent. On the other hand, the Sb2S3-sensitized cell fabricated from P3HT as the HTM does not display nonlinearity with increasing illumination intensity. One possible reason for the slight dependence of the photocurrent on the light intensity might be such that the bonding strength between Sb2S3 and PCPDTBT is weaker than that between Sb2S3 and P3HT. In summary, we have fabricated Sb2S3-sensitized inorganic organic heterojunction solar cells with various HTMs. The strong interfacial interactions between Sb2S3 and the HTMs, that is the chelation of thiophene moieties in HTMs toward Sb2S3, represent a crucial factor that makes it possible to fabricate inorganicorganic heterojunction solar cells that have a high fill factor. Hence, the attachment of HTM with semiconductor is one of key factors to maximize the charge transfer resistance between the electrons in electron transporting material and holes in HTM and the hole injection into the HTM. These findings provide us 4792
dx.doi.org/10.1021/nl2026184 |Nano Lett. 2011, 11, 4789–4793
Nano Letters with a method for designing a new HTM in an inorganic organic heterojunction semiconductor-sensitized solar cell and to obtain highly efficient devices. By using PCPDTBT as the HTM, we were able to achieve a conversion efficiency exceeding 6% under full sunlight (100 mW/cm2) with AM 1.5 G radiation, the highest value thus far reported.
LETTER
(18) Kharbish, S.; Libowitzky, E.; Beran, A. Eur. J. Mineral. 2009, 21, 325–333.
’ ASSOCIATED CONTENT
bS
Supporting Information. Details of experiments and additional supplementary table and figures. This material is available free of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION Corresponding Author
*E-mail:
[email protected]. Author Contributions §
These authors contributed equally to this work.
’ ACKNOWLEDGMENT This study was by the Global Research Laboratory (GRL) Program supported through the National Research Foundation of Korea funded by the Ministry of Education, Science and Technology and KRICT 2020 program for future Technology of Korea Research Institute of Chemical Technology (KRICT), Republic of Korea. ’ REFERENCES (1) Gr€atzel, M. Acc. Chem. Res. 2009, 42, 1788–1798. (2) Bessho, T.; Zakeeruddin, S. M.; Yeh, C.-Y.; Diau, E. W.-G.; Gr€atzel, M. Angew. Chem., Int. Ed. 2010, 49, 6646–6649. (3) Kamat, P. V. J. Phys. Chem. C 2008, 112, 18737–18753. (4) Sun, J.; Goldys, E. M. J. Phys. Chem. C 2008, 112, 9261–9266. (5) Hanewinkel, B.; Knorr, A.; Thomas, P.; Koch, S. W. Phys. Rev. B 1997, 55, 13715–13725. (6) Yafit, I.; Olivia, N.; Miles, P.; Gary, H. J. Phys. Chem. C 2009, 113, 4254–4256. (7) Chang, J. A.; Rhee, J. H.; Im, S. H.; Lee, Y. H.; Kim, H.-j.; Seok, S. I.; Nazeeruddin, M. K.; Gr€atzel, M. Nano Lett. 2010, 10, 2609–2612. (8) Lee, H. J.; Leventis, H. C.; Moon, S.-J.; Chen, P.; Ito, S.; Haque, S. A.; Torres, T.; N€uesch, F.; Geiger, T.; Zakeeruddin, S. M.; Gr€atzel, M.; Nazeeruddin, Md. K. Adv. Funct. Mater. 2009, 19, 2735–2742. (9) Lee, H. J.; Wang, M.; Chen, P.; Gamelin, D. R.; Zakeeruddin, S. M.; Gr€atzel, M.; Nazeeruddin, Md. K. Nano Lett. 2009, 9, 4221–4227. (10) Moon, S.-J.; Itzhaik, Y.; Yum, J.-H.; Zakeeruddin, S. M.; Hodes, G.; Gr€atzel, M. J. Phys. Chem. Lett. 2010, 1, 1524–1527. (11) Leventis, H. C.; O’Mahony, F.; Akhtar, J.; Afzaal, M.; O’Brien, P.; Haque, S. A. J. Am. Chem. Soc. 2010, 132, 2743–2750. (12) Cai, N.; Moon, S.-J.; Cevey-Ha, L.; Moehl, T.; Humphry-Baker, R.; Wang, P.; Zakeeruddin, S. M.; Gr€atzel, M. Nano Lett. 2011, 11, 1452–1456. (13) Stergiopoulos, T.; Bernard, M.-C.; Goff, A. H.-L.; Falaras, P. Coord. Chem. Rev. 2004, 248, 1407–1420. (14) Versavel, M. Y.; Haber, J. A. Thin Solid Films 2007, 515, 7171–7176. (15) Herzog, V. P. Z.; Harmer, S. L.; Nesbitt, H. W.; Bancroft, G. M.; Flemming, R.; Pratt, A. R. Sulf. Sci. 2006, 600, 348–356. (16) Wang, X.; Liebau, X. F. Acta Crystallogr., Sect. B 1996, 52, 7–15. (17) Minceva-Sukarava, B.; Jovanovski, G.; Makreski, P.; Soptrajanov, B.; Griffith, W.; Willis, R.; Grzetic, I. J. Mol. Struct. 2003, 651653, 181–189. 4793
dx.doi.org/10.1021/nl2026184 |Nano Lett. 2011, 11, 4789–4793
