Cite this Article
Zhao Yao, Zhao Yi-Cheng, Zhou Wen-Ke, Fu Rui, Li Qi, Yu Da-Peng, Zhao Qing. Importance of ligands on TiO2 nanocrystals for perovskite solar cells. Chinese Physics B, 2018, 27(1): 018401  
Permissions
Importance of ligands on TiO2 nanocrystals for perovskite solar cells
Zhao Yao, Zhao Yi-Cheng, Zhou Wen-Ke, Fu Rui, Li Qi, Yu Da-Peng, Zhao Qing †
State Key Laboratory for Mesoscopic Physics and Electron Microscopy Laboratory, School of Physics, Peking University, Beijing 100871, China

 

† Corresponding author. E-mail: zhaoqing@pku.edu.cn

Abstract

The fabrication of high-quality electron-selective layers at low temperature is a prerequisite to realizing efficient flexible and tandem perovskite solar cells (PSCs). A colloidal-quantum-dot ink that contains TiO2 nanocrystals enables the deposition of a flat film with matched energy level for PSCs; however, the selection of ligands on the TiO2 surface is still unexplored. Here, we systematically studied the effect of the titanium diisopropoxide bis(acetylacetonate) (TiAc2) ligand on the performance of PSCs with a planar n–i–p architecture. We prepared TiO2 nanocrystals from TiCl4 and ethyl alcohol with Cl− ligands attached on its surface and we found that a tiny amount of TiAc2 treatment of as-prepared TiO2 nanocrystals in a mixed solution of chloroform and methyl alcohol can enhance PSC power conversion efficiency (PCE) from 14.7% to 18.3%. To investigate the effect of TiAc2 ligand on PSCs, TiO2 samples with different TiAc2 content were prepared by adding TiAc2 into the as-obtained TiO2 nanocrystal solution. We use x-ray photoelectron spectroscopy to identify the content of Cl so as to reveal that Cl ligands can be substituted by TiAc2. We speculate that the improvement in PCE originates from amorphous TiO2 formation on the TiO2 nanocrystal surface, whereby a single-molecule layer of amorphous TiO2 facilitates charge transfer between the perovskite film and the TiO2 electronic transport layer, but excessive TiAc2 lowers the PSC performance dramatically. We further prove our hypothesis by x-ray diffraction measurements. We believe the PCE of PSCs can be further improved by carefully choosing the type and changing the content of surface ligands on TiO2 nanocrystal.

PACS: 84.60.Jt;88.40.H-;84.60.Bk
Keyword:TiAc2;TiO2 nanocrystal;surface ligand;perovskite solar cells
1. Introduction

The rapid progress of hybrid organic–inorganic metal halide perovskite solar cells PSCs represents a breakthrough for next-generation photovoltaic devices. Perovskite films possess many distinct properties such as tunable and suitable bandgap energy,[1–4] large dielectric constant,[5,6] long carrier diffusion lengths,[6–8] large absorption coefficient,[9–12] high carrier mobility,[12] and long carrier lifetime.[13] These special properties enable PSCs to be an ideal material as active layer of solar cells. Since Miyasaka et al.[14] reported PSCs for the first time, impressive progress has been made to promote their power conversion efficiency (PCE) from 3.8% to 22.1%.[15] To further optimize the PCE and stability of PSCs, many effective methods have been used such as changing electronic transport layer (ETL) materials, introducing additives, and using inverted device architecture.[16–27] Among all the ETLs in PSCs, TiO2-based solar cells are always prone to achieving the highest PCE.[15] However, traditional perovskite TiO2 ETL requires high-temperature (450 °C) processing, which is not only time-consuming and complex but also excludes the use of other substrates such as indium tin oxide (ITO) and flexible organic substrates.[28–30] Therefore, low-temperature TiO2 nanocrystals are especially welcome, because they provide an opportunity to use flexible substrates to adopt roll-to-roll manufacturing of PSCs in the future.[31,32]

Yang et al. first invented low-temperature TiO2 ETL by modifying TiO2 nanocrystals dispersed in ethanol with YCl3 and they reached then-record efficiency of 19.3%.[10] Tan et al. came up with a new strategy to modify TiO2 nanocrystals with Cl− ligand by dispersing nanocrystals into a mixed solution of chloroform and methyl alcohol (volume ratio 1:1). PSCs comprising Cl− capped TiO2 nanocrystal ETL reach efficiency greater than 20% and retain 90% of their initial performance after 500 h of continuous room-temperature operation at their maximum power point under one-sun illumination.[33] Kim et al. reported that an amorphous TiO2 compact layer with a well-crystallized surface PSCs reached 9.97% efficiency, whereas pure amorphous TiO2 ETL PSCs only have 1.73% PCE.[34] This discovery reminds us that it is the ETL surface rather than the bulk material that really affects the electron extraction. Dong et al. modified TiO2 with WO3 and they found that the photocurrent density increased dramatically.[35]

TiAc2 is conventionally used as a dispersing agent to disperse TiO2 nanocrystals in alcohol, but we think it is also a functional ligand that can enhance the electron extraction efficiency and thus lead to a great improvement in PSC performance. We systematically study the effect of TiAc2 content on PSCs for the first time and reveal the mechanism behind the change in PSC performance.

2. Experiment details
2.1. Materials

Methylammonium iodide, methylammonium bromine, and formamidine iodide were purchased from Dyesol. Lead iodide, lead bromide, N,N-dimethylformamide, hydrochloric acid, chlorobenzene, bis(trifluoromethane)sulfonamide lithium, 4-tert-butylpyridine, titanium tetraisopropoxide, and cesium iodide were purchased from Sigma-Aldrich. Spiro-OMeTAD was purchased from Lumtec. ITO was purchased from Thin Film Device Inc. All solid chemicals used in this experiment were 99.999% pure.

2.2. Synthesis of TiO2 nanocrystals

The following experiments were all carried out in ambient environment. First, 2 mL TiCl4 (99% Alfa-Aesar) was injected very slowly into 8 mL ethanol alcohol with 3000 rpm stirring speed to avoid local overheating of ethanol. The reaction vial was placed into an ice–water mixture. After 30 min, 40 mL of anhydrous benzyl alcohol was added to the previous solution and stirred for 10 min. The resulting solution was sealed in a vial and heated in an 80 °C water bath for approximately 10 h. The as-prepared TiO2 nanocrystals were then precipitated from the as-obtained solution by the addition of 200 ml diethyl ether and isolated by centrifugation at 5000 rpm for 2 min. The solid was subsequently washed by adding anhydrous ethanol and diethyl ether (volume ratio 1:5), followed by an identical centrifugation process. This washing procedure was repeated twice. To obtain the Cl-ligand-capped TiO2 colloidal solution, the washed TiO2 nanocrystals were dispersed into anhydrous chloroform and anhydrous methanol (1:1 volume ratio). Finally, , , , , and TiAc2 were separately added into 6 mg/ml TiO2 nanocrystal solution.

2.3. Device fabrication

ITO substrates were cleaned with acetone and isopropanol in an ultrasonic bath, sequentially. These substrates were then spin-coated with as-obtained TiO2 nanocrystals in solution at 3000 rpm for 30 s and annealed at 150 °C for 30 min on a hotplate. For the perovskite layer, a Cs0.05FA0.81MA0.14PbI2.55Br0.45 precursor solution (1.4 M) was prepared with the molar ratios of PbI2/PbBr2 and FAI/MABr both fixed at 0.85:0.15, molar ratio of CsI/(FAI+MABr) = 0.05:0.95, and molar ratio of (FAI+MABr+CsI)/(PbI2+PbBr2) fixed at 1:1. The perovskite films were deposited onto the TiO2 substrates with a two-step spin-coating procedure, i.e., 2000 rpm for 10 s with an acceleration of 200 rpm/s, followed by 5000 rpm for 40 s with a ramp-up of 1000 rpm/s. Chlorobenzene () was dropped on the spinning substrate during the second spin-coating step, 10 s before the end of the procedure. To form a thick but still smooth perovskite film, chlorobenzene was slowly dropped on the precursor film within 3 s to allow sufficient extraction of extra DMSO through the entire precursor film. The substrate was then immediately transferred to a hotplate and held at 100 °C for 10 min. After cooling down to room temperature, the hole transport layer (HTL) was then coated on the sample. The precursor of the HTL was prepared by using 1 mL spiro-OMeTAD/chlorobenzene (90 mg/mL) solution with the addition of 15 mL Li-TFSI/acetonitrile (210 mg/mL), and 40 mL TBP. Finally, 100 nm gold was deposited as a bottom electrode by thermal evaporation under a pressure of 8 × 10−6 Pa.

2.4. Photovoltaic characterization

The current density–voltage (J–V) curves were measured using an Agilent B2912 source–measure unit under AM1.5G illumination at 100 mW/cm2 provided by a Zolix simulator in an ambient environment. The active area was 0.049 cm2.

3. Results and discussion

Figures 1(a)–1(d) show top-view scanning electron microscopy (SEM) images of the TiO2 nanocrystals ETL with different TiAc2 ligand content. We spin-coated TiO2 nanocrystal ink three times to ensure good coverage. All the TiO2 ETL samples cover the ITO perfectly, suggesting ligand content would not have identifiable effect on the coverage of the TiO2 layer, and excluding the possibility that PCE variation stems from ETL coverage. It is found that the granular sensation of TiO2 nanocrystals decays with the increase in TiAc2 ligand content. We speculate that some amorphous TiO2 is formed because the low annealing temperature (150 °C) is unable to transform TiAc2 into TiO2 crystal. The morphologies of the perovskite films are the same on different TiO2 ETLs (Fig. 1(f)).

Fig. 1. Top-view SEM images of TiO2 ETL with different TiAc2 addition (a) , (b) , (c) , (d) , (e) , and (f) perovskite film.

As shown in Fig. 2(a), the cross-sectional SEM image indicates the presence of the TiO2 ELT (50 nm), perovskite active layer (600 nm), spiro-OMeTAD HTL (120 nm), and Au electrode. Figures 2(a) and 1(f) suggest that a dense-grained uniform morphology with grain sizes around 500 nm was formed in all samples. It can be seen that the photovoltaic performance was dramatically enhanced for the solar cells with TiAc2 addition (Fig. 2(b)). When TiAc2 is added, the device short-circuit current density () increases from 21.43 mA/cm2 to 22.45 mA/cm2; the device PCE increases from 14.52% to 17.74%; then, when the TiAc2 additive increases to , increases to 22.6 mA/cm2, the open-circuit voltage () increases markedly from 0.972 V to 1.08 V, the fill factor (FF) increases to 75.22%, and PCE increases to 18.36%. The device PCE as a function of TiAc2 content is shown in Fig. 2(c), and the detailed photovoltaic performance parameters are summarized in Table 1. A detailed discussion of these parameters will be presented later. Excessive TiAc2 addition results in significant drops in FF and , indicating there is a problem at the perovskite/TiO2 ETL interface.

Fig. 2. (color online) (a) Cross-section SEM image of perovskite solar cells. (b) I–V curves of perovskite solar cells with different TiAc2 content. (c) PCE changes with TiAc2 addition variation. (d) Steady-state PL spectra of perovskite film on glass substrate. (e) Time-resolved PL decay spectra.
Table 1.

PSCs performance parameters with different TiAc2 addition.

.

Steady-state photoluminescence (PL) spectra (Fig. 2(d)) indicate that the central wavelength is 780 nm, corresponding to a bandgap of 1.59 eV, which proves the perovskite active layer composition is FA0.85MA0.15PbI2.55Br0.45. We use time-resolved PL decay to further identify the film quality on glass substrate (Fig. 2(e)). The lifetime of the as-prepared perovskite film is approximately , suggesting its high quality.

X-ray photoelectron spectroscopy (XPS) measurements reveal the composition of the films prepared by TiO2 nanocrystals (Fig. 3(a)). Figure 3(d) indicates that the C 1s binding energy is centered at 285.3 eV, which deviates from the standard 284.6 eV, originated from electrons accumulated on the TiO2 surface, resulting in a general shift in the XPS spectra. Taking this shift into account, we obtain the Ti 2p1/2 central at 458.8 eV and Ti 2p3/2 at 464.5 eV, respectively, which confirms the existence of TiO2 (Fig. 3(c)). Now, we turn to Cl− and obtain the XPS spectra for different TiAc2 addition samples. The XPS results presented in Fig. 3(b) were normalized to the Ti counts in order to exclude the effect of different samples with different Ti signals. It was found that the control sample without TiAc2 addition shows a strong signal of Cl 2p1/2 at 198.7 eV and 2p3/2 198.9 eV after correction for the general shift. However, after adding TiAc2 into TiO2 chloroform and methyl alcohol mixed solution (6 mg/mL), the Cl signal suddenly disappears, to say nothing of the , , , and TiAc2 additions. The interaction between TiAc2 and TiO2 nanocrystals is so strong that almost all the Cl− ligands desorbed from the TiAc2 surface after addition.

Fig. 3. (color online) (a) General XPS spectra of TiO2 nanocrystals. (b) Cl 2p XPS spectra of TiO2 nanocrystals with different TiAc2 content (). (c) Ti 2p XPS spectra. (d) C1s XPS spectra.

We surmise that the annealing temperature of 150 °C is insufficient to transfer TiAc2 into TiO2 crystals, because TiAc2 is the raw material for high-temperature pyrolysis of TiO2 at 450 °C. It is probable that under 150 °C, TiAc2 transforms into amorphous TiO2 with a bandgap between TiO2 nanocrystals and perovskite film. TiO2 nanocrystals are approximately 5 nm in diameter, and TiO2 lattice length is 0.2 nm. Assuming a cubic shape of TiO2 nanocrystals, we calculate that each nanocrystal contains 15625 TiO2 lattice points and 3125 lattice surfaces. Then, we can roughly estimate the coverage of TiAc2. The TiAc2 molecular density is approximately ; 1 mL spin-coated TiO2 solution has lattice surface; considering the strong interaction between TiO2 and TiAc2, the TiO2 nanocrystal will be covered by a single-molecule layer of TiAc2. Once these modified TiO2 nanocrystals experience 150 °C annealing, amorphous TiO2 forms on the surface. The single-molecule layer of amorphous TiO2 with a bandgap between perovskite and TiO2 ETL facilitates electron transfer by lowering the barrier between the perovskite film and ETL, which explains why the PCE and FF increase when adding TiAc2 into TiO2 solution. will increase as well owing to better carrier transportation and reduced carrier recombination. We performed an ultraviolet photoelectron spectroscopy (UPS) measurement to prove our hypothesis. As shown in Fig. 4, the work function of TiO2 increases from 5.65 eV to 5.86 eV after TiAc2 addition, indicating that the conduction band of TiAc2-modified TiO2 nanocrystals becomes 0.21 eV lower than the pristine TiO2 nanocrystals, and thus lowers the barrier between the perovskite and ETL. However, the newly formed TiO2 is essentially amorphous with relatively high series resistance and defect density; excessive amorphous TiO2 will inevitably impact the PSC performance.

Fig. 4. (color online) UPS measurement of TiO2 and TiAc2-modified TiO2 film.

We use x-ray diffraction (XRD) measurements to prove our scenario; is ascribed to the anatase phase of TiO2 (Fig. 5). The intensity of the XRD signal decays with TiAc2 addition; when the TiAc2 addition reached , the signal of anatase TiO2 disappeared. This strongly suggests that amorphous TiO2 formed during the 150 °C annealing process. The XRD intensity of TiO2 film is still obvious with TiAc2 addition, because the single-molecule layer amorphous TiO2 is insufficient to screen the signal of the well crystalized TiO2. However, when excessive TiAc2 is added, the XRD signal intensity decays precipitously.

Fig. 5. (color online) XRD pattern of TiO2 with different TiAc2 addition (unit: ).
4. Conclusion and perspectives

In summary, we confirm that TiAc2 as an additive to TiO2 nanocrystals in a mixed solution of chloroform and methyl alcohol (6 mg/mL) enhances the PCE of PSCs from 14.52% to 18.36%. We propose a mechanism behind the enhanced PSC performance: when adding TiAc2 into as-prepared TiO2 solution, Cl− will be replaced by TiAc2 and the attached TiAc2 will transform to a single-molecule layer of amorphous TiO2, the bandgap of which is between perovskite and TiO2 nanocrystals. This lowers the barrier between the perovskite and ETL, and facilitates electron transport. However, when excessive () TiAc2 is added, the FF and drop dramatically because the thickness and defect density of the newly formed amorphous TiO2 are so high that the electron transport between the perovskite and ETL is blocked. These results indicate that the PSC performance can be further optimized by carefully choosing other ligands and controlling their content.

Reference
[1] Papavassiliou G C 1996 Molecular Crystals and Liquid Crystals Science and Technology. Section A. Molecular Crystals and Liquid Crystals 286 231
[2] Mei A Li X Liu L Ku Z Liu T Rong Y Xu M Hu M Chen J Yang Y Grätzel M Han H 2014 Science 345 295
[3] Zhang W Anaya M Lozano G Calvo M E Johnston M B Míguez H Snaith H J 2015 Nano Lett. 15 1698
[4] Hao F Stoumpos C C Cao D H Chang R P H Kanatzidis M G 2014 Nat. Photon. 8 489
[5] Stranks S D Eperon G E Grancini G Menelaou C Alcocer M J P Leijtens T Herz L M Petrozza A Snaith H J 2013 Science 342 341
[6] Snaith H J 2013 J. Phys. Chem. Lett. 4 3623
[7] Heo J H Im S H Noh J H Mandal T N Lim C-S Chang J A Lee Y H Kim H J Sarkar A NazeeruddinMd K Gratzel M Seok S I 2013 Nat. Photon. 7 486
[8] Chen W Wu Y Yue Y Liu J Zhang W Yang X Chen H Bi E Ashraful I Gratzel M Han L 2015 Science 350 944
[9] Yang W S Noh J H Jeon N J Kim Y C Ryu S Seo J Seok S I 2015 Science 348 1234
[10] Zhou H Chen Q Li G Luo S Song T B Duan H S Hong Z You J Liu Y Yang Y 2014 Science 345 542
[11] Li Y Cooper J K Liu W Sutter-Fella C M Amani M Beeman J W Javey A Ager J W Liu Y Toma F M Sharp I D 2016 Nat. Commun. 7 12446
[12] Shi D Adinolfi V Comin R Yuan M Alarousu E Buin A Chen Y Hoogland S Rothenberger A Katsiev K Losovyj Y Zhang X Dowben P A Mohammed O F Sargent E H Bakr O M 2015 Science 347 519
[13] Kim H S Lee C R Im J H Lee K B Moehl T Marchioro A Moon S J Humphry-Baker R Yum J H EMoser J Grätzel M Park N G 2012 Sci. Rep. 2 591
[14] Kojima A Teshima K Shirai Y Miyasaka T 2009 J. Am. Chem. Soc. 131 6050
[15] Yang W S Park B W Jung E H Jeon N J Kim Y C Lee D U Shin S S Seo J Kim E K Noh J H Seok S I 2017 Science 356 1376
[16] Cao J Yin J Yuan S Zhao Y Li J Zheng N 2015 Nanoscale 7 9443
[17] Lee J W Kim H S Park N G 2016 Acc. Chem. Res. 49 311
[18] Wang Z Xia Y Chen Y Huang W 2017 J. Mater. Chem. 5 12602
[19] Li X Dar M I Yi C Luo J Tschumi M Zakeeruddin S M Nazeeruddin M K Han H Gratzel M 2015 Nat. Chem. 7 703
[20] Qin C Matsushima T Fujihara T Adachi C 2017 Adv. Mater. 29 1603808
[21] Quan L N Yuan M Comin R Voznyy O Beauregard E M Hoogland S Buin A Kirmani A R Zhao K Amassian A Kim D H Sargent E H 2016 J. Am. Chem. Soc. 138 2649
[22] Rong Y Hou X Hu Y Mei A Liu L Wang P Han H 2017 Nat. Commun. 8 14555
[23] Si H Liao Q Kang Z Ou Y Meng J Liu Y Zhang Z Zhang Y 2017 Adv. Function. Mater. 27 1701804
[24] Wu Y Xie F Chen H Yang X Su H Cai M Zhou Z Noda T Han L 2017 Adv. Mater. 29 1701073
[25] Zhang J Hu Z Huang L Yue G Liu J Lu X Hu Z Shang M Han L Zhu Y 2015 Chem. Commun. 51 7047
[26] Zuo C Ding L 2014 Nanoscale 6 9935
[27] Xu L G Qiu W Chen R F Zhang H M Huang W 2018 Acta Phys. Chim. Sin. 34 36
[28] You J Hong Z Yang Y Chen Q Cai M Song T B Chen C C Lu S Liu Y Zhou H Yang Y 2014 ACS Nano 8 1674
[29] Shin S S Yang W S Noh J H Suk J H Jeon N J Park J H Kim J S Seong W M Seok S I 2015 Nat. Commun. 6 7410
[30] Chhowalla M Tretiak S Alam M A Wang H L Mohite A D 2015 Science 347 522
[31] Nie W Tsai H Asadpour R Blancon J C Neukirch A J Gupta G Crochet J J 2003 J. Am. Chem. Soc. 125 14539
[32] Docampo P Ball J M Darwich M Eperon G E Snaith H J 2013 Nat. Commun. 4 2761
[33] Tan H Jain A Voznyy O Lan X García de Arquer F P Fan J Z Quintero-Bermudez R Yuan M Zhang B Zhao Y Fan F Li P Quan L N Zhao Y Lu Z H Yang Z Hoogland S Sargent E H 2017 Science 355 722
[34] Kim I S Haasch R T Cao D H Farha O K Hupp J T Kanatzidis M G Martinson A B F 2016 ACS Appl. Mater. Interf. 8 24310
[35] Wang W G Zhu L Weng Y Y Dong W 2017 Chin. Phys. 34 028201