Crystal Engineering for Low Defect Density and High Efficiency Hybrid

Research Article www.acsami.org

Crystal Engineering for Low Defect Density and High Efficiency Hybrid Chemical Vapor Deposition Grown Perovskite Solar Cells Annie Ng,† Zhiwei Ren,† Qian Shen,† Sin Hang Cheung,‡ Huseyin Cem Gokkaya,† Shu Kong So,‡ Aleksandra B. Djurišić,§ Yangyang Wan,⊥ Xiaojun Wu,⊥ and Charles Surya*,† †

Department of Electronic and Information Engineering, The Hong Kong Polytechnic University, Hong Kong, P.R. China Department of Physics, Hong Kong Baptist University, Hong Kong, P.R. China § Department of Physics, University of Hong Kong, Hong Kong, P.R. China ⊥ Hefei National Laboratory of Physical Sciences at the Microscale, Synergetic Innovation Center of Quantum Information and Quantum Physics, CAS Key Laboratory of Materials for Energy Conversion, and Department of Materials Science and Engineering, University of Science and Technology of China, Hefei, Anhui 230000, P.R. China ‡

S Supporting Information *

ABSTRACT: Synthesis of high quality perovskite absorber is a key factor in determining the performance of the solar cells. We demonstrate that hybrid chemical vapor deposition (HCVD) growth technique can provide high level of versatility and repeatability to ensure the optimal conditions for the growth of the perovskite films as well as potential for batch processing. It is found that the growth ambient and degree of crystallization of CH3NH3PbI3 (MAPI) have strong impact on the defect density of MAPI. We demonstrate that HCVD process with slow postdeposition cooling rate can significantly reduce the density of shallow and deep traps in the MAPI due to enhanced material crystallization, while a mixed O2/N2 carrier gas is effective in passivating both shallow and deep traps. By careful control of the perovskite growth process, a champion device with power conversion efficiency of 17.6% is achieved. Our work complements the existing theoretical studies on different types of trap states in MAPI and fills the gap on the theoretical analysis of the interaction between deep levels and oxygen. The experimental results are consistent with the theoretical predictions. KEYWORDS: perovskites, solar cells, defects, passivation, growth ambient, cooling rates, crystallization, hybrid chemical vapor deposition

1. INTRODUCTION

A survey of the existing literature indicates that three key factors responsible for the dramatic rise in the PCEs of the devices are (i) the improvement in the material quality,6−8 (ii) the optimization of the device structures,9−11 and (iii) development of different growth techniques.6−8,12,13 The growth processes have strong impact on the material quality, and existing growth processes for the perovskite materials can be roughly classified into thermal deposition,7 vapor assisted,6 and the solution techniques.8,13 To date, there is little progress on the fabrication of high-efficiency PSCs by thermal evaporation presumably due to the higher cost for the thermal evaporation process compared to the solution technique. Furthermore, thermally evaporated films exhibit smaller grain size compared to those processed from solution and consequently lower PCE.14,15 Attempts have been made to combine solution and vapor processes to prepare high quality perovskite films with large grain size by Yang and colleagues resulting in high quality MAPI films.6,16 With the optimization of the carrier conduction pathways17 and further enhancement

The development of cost-effective and highly efficient renewable energy sources is one of the biggest challenges for the 21st Century. At present, the total power consumption worldwide is ∼18 TW and is projected to grow by as much as one-third by 2035.1 Continued dependence on fossil fuels will result in irreversible damages to the environment with devastating consequences. According to the International Energy Agency, nearly half of the net increase in electricity generation will come from renewables.2 Hence, the development of high efficiency photovoltaic cells will be critical for meeting the future global energy demands. Among the various photovoltaic materials, organic−inorganic perovskite thin films have drawn significant attention in recent years due to its tremendous progress in the development of high efficiency solar cells. The first perovskitebased photoelectrochemical cell, with a power conversion efficiency (PCE) of 3.8%, was demonstrated by Kojima et al. utilizing CH3NH3PbI3 (MAPI) nanocrystalline particles as sensitizers.3 The first all solid-state perovskite-based solar cell (PSC) was reported by Kim et al. with an efficiency of 9.7%.4 Today, only a few years from its first report, the PSCs have already surpassed the 22% PCE mark.5 © 2016 American Chemical Society

Received: August 23, 2016 Accepted: November 7, 2016 Published: November 7, 2016 32805

DOI: 10.1021/acsami.6b07513 ACS Appl. Mater. Interfaces 2016, 8, 32805−32814

Research Article

ACS Applied Materials & Interfaces

et al.28 Such deep traps may serve as efficient recombination centers, which are detrimental to solar cell performance.28,29 We have designed a highly versatile HCVD process for the growth of high quality MAPI layers, which were crystallized in a well-controlled ambient consisting of N2/O2 mixture resulting in high crystallinity, large grain size, good uniformity, and low defect density films. Our work complements the existing theoretical studies through detailed experimental investigations. The obtained results show that oxygen is effective in passivating both shallow traps and deep levels. Furthermore, our results indicate that the postdeposition cooling rates of the HCVD process are also important parameters for the reduction of deep levels and shallow states. The effects of the growth parameters on the material and device properties will be discussed in detail, and the comparison on the performance for both the planar and mesoporous device architectures will be provided.

in the crystallization through the introduction of a controlled amount of moisture during growth, substantial increase in the PCE was achieved.18 However, moisture is widely known to be an agent for device degradation in PSCs, and hence, it may adversely affect the stability of the devices. It is clear that the film morphology, such as the coverage and the grain size, has significant impact on the performance of the device. It is believed that perovskites with larger grain size are desired for photovoltaic applications as they exhibit lower trapstate density, higher charge-carrier mobility, and longer carrier diffusion length. Several reports have verified these remarkable properties for the single crystal perovskites.19−22 On the basis of these studies, it is known that growing high quality perovskite film is critical for achieving high efficiency solar cell. In this paper, we present systematic investigations on the optoelectronic properties of MAPI thin films and devices deposited by the HCVD technique for both planar and mesoporous devices. Our process was carried out in a standard quartz tube commonly used for Si processing. The proposed technique takes advantage of the highly developed Siprocessing equipment that provides excellent control and repeatability of the process parameters and potential for batch processing. Some initial studies focused on perovskite materials grown by chemical vapor deposition (CVD) technique have been reported, and the results are significant for future development of vapor-based growth of PSCs.12,23−26 Leyden et al.23,24 carried out the CVD process in the ambient of nitrogen and achieved the highest PCE of 11.8% for CH3NH3PbIxCl3−x based solar cell and 14.2% for formamidinium based perovskite solar cell. Luo et al.25,26 have reported the utilization of low pressure chemical vapor deposition (LPCVD) and in situ tubular chemical vapor deposition (ITCVD) for growing MAPI films, and the device efficiencies of 12.73% and 12.2% have been achieved using these techniques, respectively. The vapor-based grown perovskite films reported so far have shown excellent uniformity and coverage, while the PCEs of the devices are still below the predication. More research efforts should be devoted to investigate the impact of HCVD growth conditions on the perovskite film quality and their device performance. One of the important results reported in this paper is the investigation of the impact of the carrier gas composition on the crystallization of the perovskite material. In a recent publication, the authors reported significant enhancements in the device performance by postdeposition annealing of perovskite film in dry oxygen.15 Initial experimental results indicate substantial reduction in the bandgap states by oxygen postdeposition annealing. Further investigation is conducted by introducing oxygen during the perovskite formation stage to enhance the incorporation of oxygen in the perovskite material. Previous experimental results by our group15 are consistent with recent theoretical studies by Yin et al.27 on localized states at the grain boundaries of MAPI thin films, which indicated that such material defects mainly consist of shallow states. It was also shown that O and Cl spontaneously segregate into the grain boundaries and passivate the shallow states.27 Their work provides the theoretical basis for our experimental observations on the passivation of shallow states by oxygen annealing. So far, there is no reported work on the effectiveness of Cl and O species on the passivation of deep levels, which are believed to be instrumental in nonradiative recombination processes. This is a significant issue in light of the first-principles calculation on the formation of deep levels in MAPI performed by Agiorgousis

2. RESULTS AND DISCUSSION The experimental setup for the HCVD process is shown in Figure 1, panel a. The technique provides a high level of

Figure 1. (a) Experimental setup for HCVD growth of perovskite thin films. (b) Device architecture of the MAPI solar cells.

versatility and repeatability to ensure the optimal conditions for the growth of the perovskite films. A layer of PbI2, dissolved in DFM solution, is first spin-coated onto a glass/FTO/TiO2 substrate. The sample is then placed in the quartz tube as illustrated in Figure 1, panel a. The MAI powder is maintained at TMAI = 180 °C. The substrate is placed downstream to the MAI, and a carrier gas is used to transport the sublimated MAI to the sample maintained at temperature TS for the crystallization of the MAPI thin films. 32806


Research Article


Figure 2. SEM pictures of perovskite films grown by (a) HCVD in N2 at a postdeposition cooling rate of 8 °C/min; (b) HCVD in N2/O2 (85%:15%) at a postdeposition cooling rate of 8 °C/min; (c) HCVD in N2 at a postdeposition cooling rate of 4 °C/min; (d) HCVD in N2/O2 (85%:15%) at a postdeposition cooling rate of 4 °C/min; (e) HCVD in N2 at a postdeposition cooling rate of 0.7 °C/min; (f) HCVD in N2/O2 (85%:15%) at a postdeposition cooling rate of 0.7 °C/min; (g) cross-section SEM picture of a HCVD-grown perovskite layer, at a cooling rate of 8 °C/min; and (h) cross-section SEM picture of a HCVD-grown perovskite layer, at a cooling rate of 0.7 °C/min, deposited on FTO/TiO2 substrate.

°C.30,31 It is believed that a higher growth temperature is beneficial to atomic motion during the crystallization process and thereby enhances the grain size of the MAPI layer. However, it is observed that using reaction temperatures >100 °C, for solution processed perovskites, will lead to the decomposition of the MAPI due to the sublimation of MAI from the film.32 In our case, the carrier gas is saturated with MAI, which suppresses the sublimation of the MAI from the film, and thus, a higher crystallization temperature can be used without suffering any film degradation.

We followed a three-step process to optimize the growth parameters. First, we optimized the growth temperature for the material. By using pure N2 as the carrier gas, MAPI thin films were grown at different substrate temperatures. Standard devices, with a structure as given in Figure 1, panel b, were fabricated, and the optoelectronic properties of the cells are summarized in Table S1. The data show that the optimal growth temperature is ∼165 °C. This is substantially higher than the typical reaction temperatures used in the solution growth techniques, which are commonly limited to 90−100 32807


Research Article

ACS Applied Materials & Interfaces To optimize the oxygen content in the perovskite film, we have systematically varied the ratio of N2/O2 in the carrier gas from 100%:0% to 75%:25%. Devices, with planar structure, were fabricated using the MAPI films grown under different carrier gas compositions, and the optoelectronic properties of the devices are summarized in Table S2. Strong dependencies of the device PCEs on the carrier gas composition are observed. Our experimental results indicate that using an optimal carrier gas composition of N2/O2 (85%:15%) leads to substantial enhancement in the PCE of the devices. Typical I−V characteristics of the devices for different carrier gas compositions are shown in Figure S1a. This is corroborated by the experimental results on EQE measurements of the devices as shown in Figure S1b. The results clearly illustrate highest EQE for the device grown in N2/O2 (85%:15%) carrier gas. From the experimental data, it is apparent that the incorporation of a small amount of oxygen during HCVD process is effective in the passivation of trap states in the MAPI. However, incorporation of excessive oxygen may lead to the creation of defect states, which may be responsible for the slight reduction in the photovoltaic performance of the devices with MAPI prepared in the ambient of N2/O2 (80%:20%) and N2/ O2 (75%:25%) compared to the devices with MAPI grown in N2/O2 (85%:15%). Photothermal deflection spectroscopy was conducted to investigate this observation, and the results will be discussed in the latter text. Previous work by Sriram et al.33 on perovskite-oriented (Pb0.92Sr0.08) (Zr0.65Ti0.35)O3 thin films demonstrated that a lower postdeposition cooling rate resulted in the reduction of stress and an increase in the degree of orientation for the film. In this work, we have also examined the impact of the postdeposition cooling rate on the film morphology. SEM pictures of MAPI films grown by HCVD technique under different conditions are shown in Figure 2. It is worth noting that HCVD-grown films with fast cooling rate, at 8 °C/min (Figure 2a,b), demonstrate high concentration of pinholes, which is substantially reduced when the cooling rate decreases to 4 °C/min (Figure 2c,d). Films deposited with a slow cooling rate at 0.7 °C/min (Figure 2e,f), regardless of the ambient, demonstrate highly compact films with substantial enlargement of crystal size ranging from ∼300 nm to > ∼1.5 μm and the lowest pinhole concentration compared to the films with the fast and the intermediate cooling rates. Such pinholes may have significant impact on the optoelectronic properties of the devices due to the presence of high concentration of defect states that may potentially function as recombination centers. Furthermore, shorts between the electrodes may develop through the pinholes. When the SEM images shown in Figures 2 and S2 are compared, it is noteworthy that different carrier gas compositions do not cause obvious differences in the morphology of the MAPI layers. Figure 2, panels g and h show the cross-sectional SEM images for the MAPI films grown on the FTO/compact-TiO2 (c-TiO2) substrates using different growth parameters. The results clearly demonstrate the superior crystal quality of the slow-cooled HCVD grown film, in which the entire thickness of the perovskite layer is composed of a single crystal compared to the fast-cooled perovskite film, which demonstrates much smaller crystallites. The larger grain size with better crystal quality for the slowcooled HCVD-grown MAPI layer is expected to result in promising optoelectronic properties for the PSCs. To demonstrate the effects of the growth ambient and postdeposition cooling rate on the optoelectronic properties of

the MAPI materials and the PSCs, we have performed systematic characterizations on the lifetimes of the minority carriers by analyzing the time-resolved photoluminescence (TRPL) signals from the perovskite samples. The experimental data are shown in Figure 3, panel a, in which six different types of MAPI films, with thickness 420 nm ±10 nm, were investigated: (i) N2 carrier gas and cooling rate of ∼8 °C/ min (open blue triangles) (type I); (ii) N2 carrier gas and cooling rate of ∼4 °C/min (open green squares) (type II); (iii) N2 carrier gas and cooling rate of ∼0.7 °C/min (open red

Figure 3. (a) Time-resolved photoluminescence data. (b) Photothermal deflection spectroscopies and (c) normalized current noise power spectral densities of the devices for HCVD-grown MAPI films using different carrier gas compositions and postdeposition cooling rates. 32808


Research Article


reported by Yin et al.27 Furthermore, it is noteworthy that the effectiveness of the oxygen passivation process also depends on the initial MAPI crystal quality. By comparing the fast-cooled HCVD type I and IV samples with high concentration of pinholes and grain boundaries, the presence of oxygen in the growth ambient reduces the Urbach energy from 22.2 to 21.9 meV. The reduction in Urbach energy becomes more significant (from 24 to 23.2 meV) for the case of solution processed MAPI films (Figure S4), for which the MAPI average grain size is smaller than the HCVD films, while the Urbach energy remains practically unchanged when the slow-cooled HCVD samples (type III and type VI) are compared, for which both samples exhibit excellent crystallization with large grain size and good coverage. We next consider the impact on the density of the deep traps due to the variations in the growth conditions. Here, type I film basically serves as the control sample. Comparing between type I and type III films, we observe that utilizing slow cooling rate alone leads to substantial reduction in the trap density between 1.24 and 1.45 eV. Comparing between type I film and type IV film, we observe significant reduction in the deep traps between 0.8 and 1.08 eV. It is interesting to point out that when both the optimal carrier gas and slow cooling rate are used, deep traps over the entire energy range from 0.8 to 1.45 eV exhibit substantial reduction as observed in type VI film. Thus, the results clearly show that both carrier gas ambient and the cooling rates have strong influence in the density of the deep traps. Furthermore, the technique of PDS has been performed on another set of samples grown with different gas compositions. The results are summarized in Figure S5. It is found that the effectiveness of the passivation depends on the amount of oxygen introduced during the HCVD process. Use of the ambient of N2/O2 (90%:10%) yields the MAPI film with obvious reduction in shallow and deep trap density and that is further reduced when the oxygen content is increased to 15%. However, continuous increase of oxygen content to 25% in the growth ambient causes a slight increase in density of states from the band edge to 0.8 eV. The PDS results agree with the photovoltaic performance of the devices with the MAPI grown in different oxygen contents as shown in Table S2. The use of the optimized ambient N2/O2 (85%:15%) in HCVD process yields the highest PCE, while introducing less and more oxygen content in the ambient can improve the device performance, but the extent of enhancement is not as much as the optimized N2/O2 (85%:15%) ambient when comparing them to the control devices (N2/O2 (100%:0%)). We believe that there are two competing processes for incorporation of oxygen during the HCVD growth. In addition to the effect of oxygen passivation, oxygen can be also regarded as a p-type dopant in MAPI.27 Incorporation of excessive oxygen into the material will create additional traps in MAPI, which is analogous to the case of excess dopants in semiconductors.42 To provide further evidence that oxygen is effective in the passivation of deep traps in MAPI, we have conducted lowfrequency noise measurements on all six types of devices with the MAPI layers grown under the conditions as listed above. Low-frequency noise measurement is a nondestructive characterization technique performed directly on the complete device structure. It has been shown that the low-frequency noise in a semiconductor device arises from the modulation of the device conductance due to the random capture and emission of carriers by localized states in the device.43−45 Under

circles) (type III); (iv) optimal carrier gas and cooling rate of ∼8 °C/min (solid blue triangles) (type IV); (v) optimal carrier gas and cooling rate of ∼4 °C/min (solid green squares) (type V); and (vi) optimal carrier gas and cooling rate of ∼0.7 °C/ min (solid red circles) (type VI). The absorbance data of the corresponding samples are shown in Figure S3, for which MAPI films prepared by different growth parameters exhibit small variations. It is believed that the result of TRPL can provide the insight into the trap density among different types of films. The TRPL data are well fitted to a biexponential decay function, in which two distinct lifetimes τ1 and τ2 can be determined from the data. It is clearly observed that there are significant improvements for the lifetimes of MAPI samples grown in the N2/O2 ambient when compared to the samples grown in pure N2. It is believed that oxygen in the carrier gas may passivate the defects in MAPI and reduce the chance for carrier recombination at the defect sites, which results in the enhancement in the carrier lifetime. Also, use of a slow postdeposition cooling rate for the materials appears to further enhance the carrier lifetime, which indicates strong correlation between the cooling rate and the concentration of material defects. On the basis of the lifetime determined from the TRPL, it is estimated that the carrier diffusion length for our optimized sample can be higher than 3 μm.34,35Table S3 shows the comparison of the estimated carrier diffusion lengths of different types of samples. To characterize the density of the bandgap states, we have examined the photothermal deflection spectroscopy (PDS) of the MAPI films deposited under different growth conditions. PDS is an absorption characterization technique with high sensitivity of the order 10−4. It can be utilized to detect changes in the thermal state of the samples due to the nonradiative relaxation of photoexcited carriers. Therefore, it is commonly applied to characterize the energetic disorder such as the exponential decay of the absorption below the bandgap with a characteristic Urbach energy.36 PDS has been used to analyze the electronic defects in amorphous and organic semiconductors.37,38 In this work, PDS can facilitate our understanding on both shallow and deep traps in the MAPI film. All samples were grown on quartz to avoid absorption by the substrates. Here we choose four different representative samples types I, III, IV, and VI with their growth conditions as described above, and the results are presented in Figure 3, panel b. From the experimental results, we observe that the conduction band edge is approximately 1.59 eV from the valence band edge, which is in excellent agreement with existing reports. The shallow traps are classified in the range between EC and EC − 15 meV, while the deep levels refer to the bandgap states, which are at least ∼0.15 eV from the band edge.28,29,39,40 The shallow traps are most conveniently characterized by the magnitude of the Urbach energy, which reflects the steepness of the band tail states located at the conduction band edge.41 A small Urbach energy stipulates a lower density of shallow traps and is, thus, a useful figure-of-merit for the film. The Urbach energies for samples I, III, IV, and VI are 22.2 meV, 20.7 meV, 21.9 meV, and 20.8 meV, respectively. The shallow traps appear to be most strongly affected by the postdeposition cooling rate. The Urbach energy shows significant reduction by 1.5 and 1.1 meV for slow-cooled samples (type III and type VI) compared to their fast-cooled counterparts (type I and type IV, respectively). Recent calculations by Yin et al.27 have suggested that O can passivate the shallow traps at the MAPI grain boundaries. In our PDS results, we observe consistent trends as 32809


Research Article

ACS Applied Materials & Interfaces Table 1. Effects of Postdeposition Cooling Rates on the Performance of the Planar and Mesoscopic Devicesa cooling rate planar 8 °C/min planar 4 °C/min planar 0.7 °C/min mesoscopic 8 °C/min mesoscopic 4 °C/min mesoscopic 0.7 °C/min a

VOC (V) 0.97 1.01 1.00 0.99 1.00 0.99

± ± ± ± ± ±

0.02 0.03 0.02 0.02 0.02 0.01

JSC (mA/cm2) 22.8 22.7 23.3 23.2 23.2 23.1

± ± ± ± ± ±

0.6 1.0 0.4 0.6 0.4 0.4

FF 0.66 0.67 0.72 0.68 0.69 0.75

± ± ± ± ± ±

PCE (%) 0.02 0.02 0.01 0.02 0.01 0.02

14.6 15.5 16.7 15.5 16.1 17.2

± ± ± ± ± ±

0.6 0.3 0.4 0.3 0.4 0.2

RS (Ω cm2) 7 7 7 13 14 12

± ± ± ± ± ±

3 3 1 6 5 5

RSH (kΩ cm2) 15 18 24 15 19 35

± ± ± ± ± ±

5 9 1 2 6 5

Jo (mA/cm2) 1 2 1 5 2 1

× × × × × ×

10−8 10−10 10−10 10−8 10−8 10−12

The values were averaged from seven devices.

Figure 4. (a) HCVD-grown MAPI with fast cooling rate (8 °C/min) on mp-TiO2 spun at 3500 rpm; (b) HCVD-grown MAPI with medium cooling (4 °C/min) on mp-TiO2 spun at 3500 rpm; (c) HCVD-grown MAPI with slow cooling (0.7 °C/min) on mp-TiO2 spun at 3500 rpm; and (d) crosssection of HCVD-grown MAPI with postdeposition cooling rate 0.7 °C/min on mp-TiO2 spun at 3500 rpm.

a constant voltage bias, the current fluctuation due to a single trap gives rise to a random telegraph noise with a power spectral density in the form of a Lorentzian, τ 2 2 . Since the

and VI) compared to those grown in pure N2 (types I, II, and III); and (ii) SI(f)/I2 is lowered more gradually with decreasing cooling rate in the same growth ambient. The data show that the concentration of the deep traps can be lowered by enhancement in the crystallization through utilizing a slow postdeposition cooling rate and particularly by defect passivation with the use of N2/O2 carrier gas. Standard planar devices were also fabricated using HCVDgrown perovskite films with different postdeposition cooling rates in the optimized ambient. The corresponding optoelectronic properties are summarized in Table 1. It is found that the average PCE of the devices increased significantly by 2.1%, from 14.6% to 16.7%, for the MAPI layers grown in N2/O2 (85%:15%) with reducing the cooling rate from 8 °C/min to 0.7 °C/min. This observation is highly consistent with the results obtained from the above characterizations. Furthermore, we have analyzed the dark currents (Figure S6) of the devices with the MAPI layers grown in the optimized ambient with different cooling rates.46 It is found that the shunt resistance of the planar devices as indicated in Table 1 increases when the postdeposition cooling rate is reduced from 8 °C/min to 0.7 °C/min, which indicates fewer shunting paths in the slowcooled MAPI devices. It is expected as lowering the cooling rate is desired for MAPI crystallization with reduction in the concentration of pinholes. The reverse saturation current density (Jo) obtained from fitting the dark current of MAPI

1+ω τ

individual trapping and detrapping events are statistically independent, the total current noise power spectral density, SI( f), of the complete device is given by SI (f ) = 4(ΔI0)2

∫E ∫z ∫y ∫x NT(E , x , y , z) 1 + 4τπ 2f 2 τ 2 dx dy dz dE (1)

in which ΔI0 is the current fluctuation arising from capture of a single carrier under constant voltage bias. The Lorentzian is a sharply peaked function in energy at Ep = −kBT ln(ωτ0), and 4Cf the trap density can be expressed as NT(Ep) ≈ k T SI (f , T ) B

where C is a proportionality constant. Figure 3, panel c illustrates the normalized room temperature current noise power spectral density for all six types of devices. All the devices are identical in their fabrication process except for the growth conditions of the MAPI layers. Using a typical inverse phonon frequency (10−14 s) for τ0, we have estimated 0.89 eV < Ep < 0.95 eV. Thus, low-frequency noise measurement enables one to monitor the dependence of the deep levels on the MAPI growth parameters. We observe two important trends for the data: (i) significant reduction in SI(f)/I2 is observed for devices with the MAPI layers grown in N2/O2 ambient (types IV, V, 32810


Research Article


Figure 5. I−V characteristics in forward (open symbol) and reverse scans (close symbol) for the best devices for: (a) HCVD-grown MAPI layer in N2/O2 (85%:15%) ambient with a planar device structure and (b) HCVD-grown MAPI layer in N2/O2 (85%:15%) ambient with the incorporation of an mp-scaffold.

and b, albeit with substantial reduction in the density compared to Figure 2, panels a−d. The slow-cooled sample exhibits highly compact films with excellent coverage. From Figure 4, panel c, the grain size of the slow-cooled MAPI layer ranges from ∼300 nm to >1 μm, which is similar to the MAPI layers deposited on the c-TiO2 as shown in Figure 2, panels e and f. Furthermore, Figure 4, panel d shows the cross-section of the perovskite deposited on the FTO/c-TiO2/mp-TiO2 substrate, by HCVD process, which demonstrates a single crystal spanning across the entire thickness of the perovskite layer. Photovoltaic devices utilizing an mp-TiO2 scaffold were fabricated based on the device structure as illustrated in Figure 1, panel b. Detailed investigations on the characteristics of the mp-PSCs as a function of the postdeposition cooling rate were performed. Table 1 summarizes the photovoltaic parameters of the mp-PSCs. By comparing the experimental results obtained for the planar devices, the mp-PSCs demonstrate similar average VOC and JSC, while the average FF values are significantly higher than that of the planar devices. In general, improved FF of the device is an indication of better charge transport properties within the devices.54 It was noted that the presence of the mp-TiO2 scaffold may facilitate effective electron extraction from the perovskite and rapid injection of the carriers into the c-TiO2 layer.49 It is also observed that the cooling rate affects the performance of the mp-PSCs. Similar to planar devices, significant enhancement in shunt resistance and reduction in magnitude of Jo is observed from mp-PSCs when the postdeposition cooling rates of the HCVD process are lowered, which is in good agreement with the observed trend for the performance of the mp-PSCs shown in Table 1. To compare the MAPI films deposited on different substrates, we have conducted X-ray diffraction (XRD) characterizations of the perovskite films deposited on the glass/FTO/c-TiO2 and glass/FTO/c-TiO2/mp-TiO2 substrates, which are similar to the configurations of the real devices. The diffraction peaks, as indicated in Figure S7, which are attributed to the perovskite planes (110), (220), and (310), can be clearly resolved for the samples. The degree of crystallinity can be evaluated from the XRD data by computing the ratio between the integrated intensity under the crystalline peaks and the integrated intensity under the complete XRD spectrum.55 It is found that perovskites grown by HCVD process on mp-TiO2 substrates exhibit improvement in the crystallinity compared to the samples without the mp-TiO2 scaffold. Figure 5 shows the I−V characteristics of the best

devices with different postdeposition cooling rate is also summarized in Table 1, which is reduced in magnitude with lowering the postdeposition cooling rate of the HCVD process, indicating improved performance of MAPI devices. The characterizations performed on the MAPI material and the corresponding devices show that the optimal growth ambient and slow postdeposition cooling rate are the critical factors for the HCVD process to yield high quality MAPI films with low defect density. The best planar device fabricated utilizing the optimal fabrication parameters as presented above is shown to have a PCE of 17.2%. It is clear from the experimental data presented above that the incorporation of oxygen in the perovskite leads to the passivation of both deep traps and shallow states in the material. This is attributed to be the main reason underlying the observed enhancements in the optoelectronic properties of the films. Our results complement the theoretical analyses presented by Yin et al.27 who demonstrated the passivation of shallow traps at the grain boundaries by oxygen. However, so far there is little theoretical analysis on oxygen passivation of deep traps in MAPI. To evaluate the interaction between oxygen and the potential intrinsic deep traps28,29 in MAPI, firstprinciple analyses were conducted. The deep traps under consideration are in the same energy range as the experimental data presented previously. The detailed calculations and results are referred to in the Supporting Information. Device performance can also be enhanced through the optimization of the device structure. So far, the bulk of our work is performed on planar structure, and it is important to investigate the performance of mesoscopic PSCs (mp-PSCs) using HCVD technique. Existing reports pointed out that the mp-TiO2 layer has strong impacts in perovskite crystallization47,48 and device operation49−52 as well as the device stability.53 We first performed systematic characterizations on the effects of the mp-TiO2 layer on properties of HCVD-grown MAPI films. The mp-TiO2 layer was prepared by spin coating the diluted titania paste in ethanol on FTO/c-TiO2 substrates. The MAPI layer was then grown on the top of the mp-TiO2 layer using the optimized growth temperature and ambient as described above. The morphology of the MAPI layer and the device performance were examined in detail as a function of the postdeposition cooling rate. The plan-view images of the HCVD-grown samples with the cooling rates 8 °C/min, 4 °C/ min, and 0.7 °C/min are shown in Figure 4, panels a, b, and c, respectively. Pinholes can still be observed in Figure 4, panels a 32811


Research Article


substrates over a long period of time. Further investigations on the devices stability will be performed in the future work.

devices for: (a) HCVD-grown device under optimal processing conditions and with a planar device structure; and (b) HCVDgrown device under optimal processing conditions with an mpscaffold. It is clearly shown that the HCVD devices exhibit high FF values with little hysteresis, which is believed to arise from effective control of the defect density and enhancement in crystallization by using optimized perovskite growth methods. Similar observation in the elimination of the hysteresis due to the reduction of density of traps states in MAPI was reported by Shao et al.40 By optimizing the growth conditions and utilizing the mp-layer in the device architecture, the champion device with VOC = 1.0 V, JSC = 23.0 mA/cm2, FF = 0.77, and a high PCE of 17.6% was successfully fabricated (Figure 5b). The distribution of the photovoltaic parameters shown in Table 1 is summarized in Figure S8. Furthermore, we have compared the stability of the devices with the MAPI grown in pure nitrogen or N2/O2 (85%:15%) on c-TiO2 and mp-TiO2 coated FTO substrate. Figures 6 and

3. CONCLUSION We demonstrate that the high-temperature HCVD process conducted in an ambient of N2/O2 (85%/15%) with slow postdeposition cooling rate can successfully reduce the density of both the shallow traps and the deep levels in MAPI material. Incorporation of an mp-TiO2 scaffold in the device can further improve the device performance due to further improvement in the crystallinity of the MAPI and enhancement of the carrier transport properties of the devices, leading to higher FF. With the optimized fabrication conditions and device architecture, the champion device with VOC = 1.0 V, JSC = 23.0 mA/cm2, FF = 0.77 giving a PCE of 17.6% was obtained. 4. EXPERIMENTAL SECTION The devices were fabricated in the configuration shown in Figure 1b. Detailed preparation of the substrates and the materials used can be referred to the Supporting Information. The formation of the perovskite absorber layer of the device was achieved by the HCVD process. First, PbI2 was dissolved in DMF (462 mg/mL) and was spincoated on the substrate at 2000 rpm, and then the sample was loaded into the setup as illustrated in Figure 1, panel a, in which 2 g of MAI powder was maintained at 180 °C for the sublimation of the material. A carrier gas, composed of high purity grade (>99.9%) nitrogen/ oxygen mixture, was passed into the quartz tube. Five different N2/O2 ratios were used (100%:0%, 90%:10%, 85%:15%, 80%:20%, and 75%:25%) in our experiment for the optimization of the growth ambient. The substrate was placed downstream to the MAI vapor, and the temperature was systematically varied at 150, 165, or 175 °C. After 2 h, the setup was allowed to cool at a rate of 8 °C/min, 4 °C/min, or 0.7 °C/min. The hole transport layer was prepared by dissolving spiroMeOTAD (80 mg/mL) in chlorobenzene with the additives of LiTFSI (17.5 μL from a stock solution of 520 mg/mL in acetonitrile) and 29 μL of tBP and was deposited by spin coating on top of the perovskite at 4500 rpm. The prepared samples with spiro-MeOTAD on top were stored in O2 ambient for 12 h at room temperature. Finally, gold electrodes with a thickness of 60 nm were deposited by thermal evaporation through a shadow mask, and the device area was 0.06 cm2. The devices were encapsulated in an N2-filled glovebox before characterizations. The details for the characterizations can be referred to in the Supporting Information.

Figure 6. Stability of the HCVD-grown devices stored in nitrogen.

S9 show the evolution of the PCEs of the three types of devices as a function of the storage time in N2 and in air: (i) Type A devices, mp-MAPI layer grown by HCVD technique in N2/O2 (85%:15%) ambient; (ii) Type B devices, MAPI layer grown by HCVD technique in N2/O2 (85%:15%) ambient with a planar device structure; and (iii) Type C devices, MAPI layer grown by HCVD technique in pure N2 ambient. For the results shown in Figures 6 and S9, it is found type C devices exhibit the fastest degradation in the PCE, while Type B devices exhibit intermediate rate of degradation and type A devices demonstrate the best device stability among the three types of devices under test regardless of storage ambient. The results are expected as relatively higher trap density is found for the MAPI films grown in N2 ambient, as confirmed by the TRPL, PDS, and low frequency noise measurements. Perovskites with high trap density will likely lead to the accumulation of photoexcited carriers within the layer, which not only affects the photovoltaic performance of the devices, but also causes the devices to be unstable as the accumulated charges will readily react with the external species from the environment as well as the additives in the HTM.56,57 Since the MAPI films for type A and type B devices are grown in the same conditions, the enhanced stability of the type A devices is mainly due to the presence of mp-TiO2 layer. In fact, mp-TiO2 acts as an efficient electron acceptor, which can facilitate the interfacial charge transfer between the perovskite and the photoanode so that charge accumulation within the devices can be avoided. We also believe that mp-TiO2 can behave as a buffer layer, enabling an intimate contact between the large perovskite grain and the

■

ASSOCIATED CONTENT

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsami.6b07513. Experimental details; photovoltaic performance of perovskite solar cells; experimental setup for HCVD growth and device architecture; SEM images; absorbance; PDS results; dark I−V characteristics; XRD results; distribution of photovoltaic parameters; stability of the HCVDgrown devices stored in air; theoretical results for the effect of oxygen postdeposition annealing on MAPI films; estimated carrier diffusion lengths (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Charles Surya: 0000-0002-2990-7402 Notes

The authors declare no competing financial interest. 32812


Research Article


■

in Perovskite Photovoltaic Cells Prepared Using Evaporated PbI2/ CH3NH3I Multilayers. J. Mater. Chem. A 2015, 3, 9223−9231. (15) Ren, Z.; Ng, A.; Shen, Q.; Gokkaya, H. C.; Wang, J.; Yang, L.; Yiu, W. K.; Bai, G.; Djurišić, A. B.; Leung, W. W.; Hao, J.; Chan, W. K.; Surya, C. Thermal Assisted Oxygen Annealing for High Efficiency Planar CH3NH3PbI3 Perovskite Solar Cells. Sci. Rep. 2014, 4, 6752. (16) Liu, C.; Fan, J.; Zhang, X.; Shen, Y.; Yang, L.; Mai, Y. Hysteretic Behavior upon Light Soaking in Perovskite Solar Cells Prepared via Modified Vapor-Assisted Solution Process. ACS Appl. Mater. Interfaces 2015, 7, 9066−9071. (17) Zhou, H.; Chen, Q.; Li, G.; Luo, S.; Song, T.-B.; Duan, H. − S.; Hong, Z.; You, J.; Liu, Y.; Yang, Y. Interface Engineering of Highly Efficient Perovskite Solar Cells. Science 2014, 345, 542−546. (18) You, J.; Yang, Y.; Hong, Z.; Song, T.-B.; Meng, L.; Liu, Y.; Jiang, C.; Zhou, H.; Chang, W. − H.; Li, G.; Yang, Y. Moisture Assisted Perovskite Film Growth for High Performance Solar Cells. Appl. Phys. Lett. 2014, 105, 183902. (19) Zhumekenov, A. A.; Saidaminov, M. I.; Haque, M. A.; Alarousu, E.; Sarmah, S. P.; Murali, B.; Dursun, I.; Miao, X. H.; Abdelhady, A. L.; Wu, T.; Mohammed, O. F.; Bakr, O. M. Formamidinium Lead Halide Perovskite Crystals with Unprecedented Long Carrier Dynamics and Diffusion Length. ACS Energy Lett. 2016, 1, 32−37. (20) Bi, Y.; Hutter, E. M.; Fang, Y.; Dong, Q.; Huang, J.; Savenije, T. J. Charge Carrier Lifetimes Exceeding 15 μs in Methylammonium Lead Iodide Single Crystals. J. Phys. Chem. Lett. 2016, 7, 923−928. (21) Dong, Q.; Fang, F.; Shao, Y.; Mulligan, P.; Qiu, J.; Cao, L.; Huang, J. Electron-Hole Diffusion Lengths > 175 mm in SolutionGrown CH3NH3PbI3 Single Crystals. Science 2015, 347, 967−970. (22) 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. Low Trap-State Density and Long Carrier Diffusion in Organolead Trihalide Perovskite Single Crystals. Science 2015, 347, 519−522. (23) Leyden, M. R.; Ono, L. K.; Raga, S. R.; Kato, Y.; Wang, S.; Qi, Y. High Performance Perovskite Solar Cells by Hybrid Chemical Vapor Deposition. J. Mater. Chem. A 2014, 2, 18742−18745. (24) Leyden, M. R.; Lee, M. V.; Raga, S. R.; Qi, Y. Large Formamidinium Lead Trihalide Perovskite Solar Cells Using Chemical Vapor Deposition with High Reproducibility and Tunable Chlorine Concentrations. J. Mater. Chem. A 2015, 3, 16097−16103. (25) Luo, P.; Liu, Z.; Xia, W.; Yuan, C.; Cheng, J.; Lu, Y. A Simple In Situ Tubular Chemical Vapor Deposition Processing of Large-Scale Efficient Perovskite Solar Cells and the Research on Their Novel RollOver Phenomenon in J−V Curves. J. Mater. Chem. A 2015, 3, 12443− 12451. (26) Luo, P.; Liu, Z.; Xia, W.; Yuan, C.; Cheng, J.; Lu, Y. Uniform, Stable, and Efficient Planar-Heterojunction Perovskite Solar Cells by Facile Low-Pressure Chemical Vapor Deposition under Fully OpenAir Conditions. ACS Appl. Mater. Interfaces 2015, 7, 2708−2714. (27) Yin, W. J.; Chen, H.; Shi, T.; Wei, S. H.; Yan, Y. Origin of High Electronic Quality in Structurally Disordered CH3NH3PbI3 and the Passivation Effect of Cl and O at Grain Boundaries. Adv. Electron. Mater. 2015, 1, 1500044. (28) Agiorgousis, M. L.; Sun, Y. Y.; Zeng, H.; Zhang, S. Strong Covalency-Induced Recombination Centers in Perovskite Solar Cell Material CH3NH3PbI3. J. Am. Chem. Soc. 2014, 136, 14570−14575. (29) Yin, W. J.; Shi, T.; Yan, Y. Unusual Defect Physics in CH3NH3PbI3 Perovskite Solar Cell Absorber. Appl. Phys. Lett. 2014, 104, 063903. (30) Sun, S.; Salim, T.; Mathews, N.; Duchamp, M.; Boothroyd, C.; Xing, G.; Sum, T. C.; Lam, Y. M. The Origin of High Efficiency in Low-Temperature Solution-Processable Bilayer Organometal Halide Hybrid Solar Cells. Energy Environ. Sci. 2014, 7, 399−407. (31) Noel, N. K.; Abate, A.; Stranks, S. D.; Parrott, E. S.; Burlakov, V. M.; Goriely, A.; Snaith, H. J. Enhanced Photoluminescence and Solar Cell Performance via Lewis Base Passivation of Organic−Inorganic Lead Halide Perovskites. ACS Nano 2014, 8, 9815−9821. (32) Conings, B.; Drijkoningen, J.; Gauquelin, N.; Babayigit, A.; D’Haen, J.; D’Olieslaeger, L.; Ethirajan, A.; Verbeeck, J.; Manca, J.;

ACKNOWLEDGMENTS This work was supported by a GRF grant (Grant No. PolyU 152045/15E) and RGC Theme-based Research Scheme (Grant No. HKU T23-713/11). S.K.S. would like to acknowledge the RGC for the support of the PDS experiments through the GRF scheme (Grant No. HKBU211913). X.W. would like to acknowledge the following grants for the support of the section on theoretical analysis: National Key Basic Research Program (2011CB921404, 2012CB922001) and USTCSCC, Tianjin, and Shanghai Supercomputer Centers. The authors thank Prof. W. K. Chan for providing equipment for the EQE measurement. Special thanks are given to Dr. Jianhua Hao and Gongxun Bai for professional assistance in TRPL measurement.

■

REFERENCES

(1) BP. Statistical Review of World Energy 2015, 2016. http://www.bp. com/en/global/corporate/about-bp/energy-economics/statisticalreview-of-world-energy.html (accessed September, 2015). (2) International Energy Agency. 2014 Key World Energy Statistics, 2015. http://www.bp.com/content/dam/bp/en/corporate/pdf/bpstatistical-review-of-world-energy-2015-full-report.pdf (accessed September, 2015). (3) Kojima, A.; Teshima, K.; Shirai, Y.; Miyasaka, T. Organometal Halide Perovskites as Visible-Light Sensitizers for Photovoltaic Cells. J. Am. Chem. Soc. 2009, 131, 6050−6051. (4) 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.; Moser, J. E.; Grätzel, M.; Park, N.-G. Lead Iodide Perovskite Sensitized AllSolid-State Submicron Thin Film Mesoscopic Solar Cell with Efficiency Exceeding 9%. Sci. Rep. 2012, 2, 591. (5) Green, M. A.; Emery, K.; Hishikawa, Y.; Warta, W.; Dunlop, E. D. Solar Cell Efficiency Tables (version 48). Prog. Photovoltaics 2016, 24, 905−913. (6) Chen, Q.; Zhou, H.; Hong, Z.; Luo, S.; Duan, H.-S.; Wang, H.H.; Liu, Y.; Li, G.; Yang, Y. Planar Heterojunction Perovskite Solar Cells via Vapor-Assisted Solution Process. J. Am. Chem. Soc. 2014, 136, 622−625. (7) Liu, M.; Johnston, M. B.; Snaith, H. J. Efficient Planar Heterojunction Perovskite Solar Cells by Vapour Deposition. Nature 2013, 501, 395−398. (8) Burschka, J.; Pellet, N.; Moon, S.-J.; Humphry-Baker, R.; Gao, P.; Nazeeruddin, M. K.; Grätzel, M. Sequential Deposition as a Route to High-Performance Perovskite-Sensitized Solar Cells. Nature 2013, 499, 316−319. (9) Luo, Q.; Ma, H.; Zhang, Y.; Yin, X.; Yao, Z.; Wang, N.; Li, J.; Fan, S.; Jiang, K.; Lin, H. Cross-Stacked Superaligned Carbon Nanotube Electrodes for Efficient Hole Conductor-Free Perovskite Solar Cells. J. Mater. Chem. A 2016, 4, 5569−5577. (10) Liu, T.; Wang, C.; Hou, J.; Zhang, C.; Chen, H.; He, H.; Wang, N.; Wu, H.; Cao, G. Enhanced Electron Collection in Perovskite Solar Cells Employing Thermoelectric NaCo2O4/TiO2 Coaxial Nanofibers. Small 2016, 12, 5146. (11) Luo, Q.; Zhang, Y.; Liu, C.; Li, J.; Wang, N.; Lin, H. IodideReduced Graphene Oxide with Dopant-Free Spiro-OMeTAD for Ambient Stable and High- Efficiency Perovskite Solar Cells. J. Mater. Chem. A 2015, 3, 15996−16004. (12) Ono, L. K.; Leyden, M. R.; Wang, S.; Qi, Y. Organometal Halide Perovskite Thin Films and Solar Cells By Vapor Deposition. J. Mater. Chem. A 2016, 4, 6693−6713. (13) Li, E.; Guo, Y.; Liu, T.; Hu, W.; Wang, N.; He, H.; Lin, H. Preheating-Assisted Deposition of Solution- Processed Perovskite Layer for an Efficiency- Improved Inverted Planar Composite Heterojunction Solar Cell. RSC Adv. 2016, 6, 30978−30985. (14) Ng, A.; Ren, Z.; Shen, Q.; Cheung, S. H.; Gokkaya, H. C.; Bai, G.; Wang, J.; Yang, L.; So, S. K.; Djurišić, A. B.; Leung, W. W.; Hao, J.; Chan, W. K.; Surya, C. Efficiency Enhancement by Defect Engineering 32813


Research Article


(50) Edri, E.; Kirmayer, S.; Henning, A.; Mukhopadhyay, S.; Gartsman, K.; Rosenwaks, Y.; Hodes, G.; Cahen, D. Why Lead Methylammonium Tri-Iodide Perovskite-Based Solar Cells Require a Mesoporous Electron Transporting Scaffold (But Not Necessarily a Hole Conductor). Nano Lett. 2014, 14, 1000−1004. (51) O’Regan, B. C.; Barnes, P. R. F.; Li, X.; Law, C.; Palomares, E.; Marin-Beloqui, J. M. Optoelectronic Studies of Methylammonium Lead Iodide Perovskite Solar Cells with Mesoporous TiO2: Separation of Electronic and Chemical Charge Storage, Understanding Two Recombination Lifetimes, and the Evolution of Band Offsets During J−V Hysteresis. J. Am. Chem. Soc. 2015, 137, 5087−5099. (52) Kim, H.-S.; Park, N.-G. Parameters Affecting I−V Hysteresis of CH3NH3PbI3 Perovskite Solar Cells: Effects of Perovskite Crystal Size and Mesoporous TiO2 Layer. J. Phys. Chem. Lett. 2014, 5, 2927−2934. (53) Yin, J.; Cao, J.; He, X.; Yuan, S.; Sun, S.; Li, J.; Zheng, N.; Lin, L. Improved Stability of Perovskite Solar Cells in Ambient Air by Controlling the Mesoporous Layer. J. Mater. Chem. A 2015, 3, 16860− 16866. (54) Qi, B.; Wang, J. Fill Factor in Organic Solar Cells. Phys. Chem. Chem. Phys. 2013, 15, 8972−8982. (55) Rabiej, S. A Comparison of Two X-Ray Diffraction Procedures for Crystallinity Determination. Eur. Polym. J. 1991, 27, 947−954. (56) Kohle, O.; Grätzel, M.; Meyer, A. F.; Meyer, T. B. The Photovoltaic Stability of, Bis(isothiocyanato)rlutheniurn(II)-bis-2,2′ bipyridine-4, 4′ -dicarboxylic Acid and Related Sensitizers. Adv. Mater. 1997, 9, 904−906. (57) Niu, G.; Guo, X.; Wang, L. Review of Recent Progress in Chemical Stability of Perovskite Solar Cells. J. Mater. Chem. A 2015, 3, 8970−8980.

Mosconi, E.; Angelis, F. D.; Boyen, H.-G. Intrinsic Thermal Instability of Methylammonium Lead Trihalide Perovskite. Adv. Energy Mater. 2015, 5, 1500477. (33) Sriram, S.; Bhaskaran, M.; Holland, A. S. The Effect of PostDeposition Cooling Rate on the Orientation of Piezoelectric (Pb0.92Sr0.08) (Zr0.65Ti0.35)O3 Thin Films Deposited by RF Magnetron Sputtering. Semicond. Sci. Technol. 2006, 21, 1236−1243. (34) Oga, H.; Saeki, A.; Ogomi, Y.; Hayase, S.; Seki, S. Improved Understanding of the Electronic and Energetic Landscapes of Perovskite Solar Cells: High Local Charge Carrier Mobility, Reduced Recombination, and Extremely Shallow Traps. J. Am. Chem. Soc. 2014, 136, 13818−13825. (35) Ponseca, C. S., Jr.; Savenije, T. J.; Abdellah, M.; Zheng, K.; Yartsev, A.; Pascher, T.; Harlang, T.; Chabera, P.; Pullerits, T.; Stepanov, A.; Wolf, J.; Sundström, V. Organometal Halide Perovskite Solar Cell Materials Rationalized: Ultrafast Charge Generation, High and Microsecond-Long Balanced Mobilities, and Slow Recombination. J. Am. Chem. Soc. 2014, 136, 5189−5192. (36) De Wolf, S.; Holovsky, J.; Moon, S. − J.; Löper, P.; Niesen, B.; Ledinsky, M.; Haug, F. − J.; Yum, J. − H.; ballif, C. Organometallic Halide Perovskites: Sharp Optical Absorption Edge and its Relation to Photovoltaic Performance. J. Phys. Chem. Lett. 2014, 5, 1035−1039. (37) Thèye, M. L.; Chahed, L.; Cabarrocas, P. R. I.; Zellama, K. Studies by Photothermal Deflection Spectroscopy of Defect Formation in A-Si:H. Philos. Mag. B 1991, 63, 143−150. (38) Goris, L.; Poruba, A.; Purkrt, A.; Vandewal, K.; Swinnen, A.; Haeldermans, I.; Haenen, K.; Manca, J. V.; Vaněcě k, M. Optical Absorption by Defect States in Organic Solar Cells. J. Non-Cryst. Solids 2006, 352, 1656−1659. (39) Oga, H.; Saeki, A.; Ogomi, Y.; Hayase, S.; Seki, S. Improved Understanding of the Electronic and Energetic Landscapes of Perovskite Solar Cells: High Local Charge Carrier Mobility, Reduced Recombination, and Extremely Shallow Traps. J. Am. Chem. Soc. 2014, 136, 13818−13825. (40) Shao, Y. C.; Xiao, Z. G.; Bi, C.; Yuan, Y. B.; Huang, J. S. Origin and Elimination of Photocurrent Hysteresis by Fullerene Passivation in CH3NH3PbI3 Planar Heterojunction Solar Cells. Nat. Commun. 2014, 5, 5784. (41) Carr, J. A.; Chaudhary, S. The Identification, Characterization and Mitigation of Defect States in Organic Photovoltaic Devices: A Review and Outlook. Energy Environ. Sci. 2013, 6, 3414−3438. (42) Kaufmann, U.; Schlotter, P.; Obloh, H.; Köhler, K.; Maier, M. Hole Conductivity and Compensation in Epitaxial GaN:Mg Layers. Phys. Rev. B: Condens. Matter Mater. Phys. 2000, 62, 10867−10872. (43) Fong, W. K.; Ng, S. W.; Leung, B. H.; Surya, C. Characterization of Low-Frequency Noise in Molecular Beam Epitaxy-Grown Gan Epilayers Deposited on Double Buffer Layers. J. Appl. Phys. 2003, 94, 387−391. (44) Wong, H. Low-Frequency Noise Study in Electron Devices: Review and Update. Microelectron. Reliab. 2003, 43, 585−599. (45) Surya, C.; Hsiang, T. Y. A Thermal Activation Model for 1/f y Noise in Si-MOSFETs. Solid-State Electron. 1988, 31, 959−964. (46) Lee, J. H.; Cho, S.; Roy, A.; Jung, H. T.; Heeger, A. J. Enhanced Diode Characteristics of Organic Solar Cells Using Titanium Suboxide Electron Transport Layer. Appl. Phys. Lett. 2010, 96, 163303. (47) Park, B.-W.; Philippe, B.; Gustafsson, T.; Sveinbjörnsson, K.; Hagfeldt, A.; Johansson, E. M. J.; Boschloo, G. Enhanced Crystallinity in Organic−Inorganic Lead Halide Perovskites on Mesoporous TiO2 via Disorder−Order Phase Transition. Chem. Mater. 2014, 26, 4466− 4471. (48) Choi, J. J.; Yang, X.; Norman, Z. M.; Billinge, S. J. L.; Owen, J. S. Structure of Methylammonium Lead Iodide within Mesoporous Titanium Dioxide: Active Material in High-Performance Perovskite Solar Cells. Nano Lett. 2014, 14, 127−133. (49) Moehl, T.; Im, J. H.; Lee, Y. H.; Domanski, K.; Giordano, F.; Zakeeruddin, S. M.; Dar, M. I.; Heiniger, L.-P.; Nazeeruddin, M. K.; Park, N.-G.; Grätzel, M. Strong Photocurrent Amplification in Perovskite Solar Cells with a Porous TiO2 Blocking Layer under Reverse Bias. J. Phys. Chem. Lett. 2014, 5, 3931−3936. 32814


Crystal Engineering for Low Defect Density and High Efficiency Hybrid

Recommend Documents