Correlation between the Effectiveness of the Electron-Selective

Subscriber access provided by TULANE UNIVERSITY

Energy Conversion and Storage; Plasmonics and Optoelectronics

Correlation between the Effectiveness of the Electron-Selective Contact and Photovoltaic Performance of Perovskite Solar Cells Karen Valadez-Villalobos, Jesus Idigoras, Lilian P. Delgado, David Meneses-Rodríguez, Juan Antonio Anta, and Gerko Oskam J. Phys. Chem. Lett., Just Accepted Manuscript • DOI: 10.1021/acs.jpclett.8b03834 • Publication Date (Web): 08 Feb 2019 Downloaded from http://pubs.acs.org on February 8, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry Letters

TOC graphic 338x190mm (54 x 54 DPI)

ACS Paragon Plus Environment

The Journal of Physical Chemistry Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 22

Correlation between the Effectiveness of the Electron-selective Contact and Photovoltaic Performance of Perovskite Solar Cells

Karen Valadez-Villalobosa, Jesús Idígorasb, Lilian P. Delgadoa, David MenesesRodrígueza, Juan A. Antab,*, Gerko Oskama,*

aDepartment of Applied Physics, CINVESTAV-IPN, Antigua Carretera a Progreso km 6, Mérida, Yucatán 97310, México bÁrea de Química Física, Universidad Pablo de Olavide, Seville, Spain.

AUTHOR INFORMATION

Corresponding Author E-mail: [email protected] (J.A.A.) E-mail: [email protected] (G.O.)


1

Page 3 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


ABSTRACT Metal halide perovskites (MHPs) are mixed electronic-ionic semiconductors with a remarkable photovoltaic potential that has led to a current world record efficiency surpassing 23%. This good performance stems from the combination of excellent light harvesting and relatively slow non-radiative recombination, which are characteristic of MHPs. However, taking advantage of these properties requires electron and hole transport materials that can efficiently extract charge with minimal photovoltage losses and recombination. It is well-known that n-type anatase TiO2 is a good electron selective contact (ESC), although the fundamental reasons for its functioning are not completely clear so far. In this letter, we investigate this issue by preparing perovskite-based solar cells with various n-type metal-oxide electron selective contacts of different chemical nature and crystal structure. Our main finding is that the open-circuit photovoltage remains essentially independent of the nature of the contact for highly selective electron contacts, a fact that we attribute to a recombination rate that is mainly governed by the bulk of the MHPs. In contrast, replacement of the “standard” TiO2 contact by alternative contacts leads to lower short-circuit photocurrents and more pronounced hysteresis, related to enhanced surface recombination at less effective electron selective contacts. TOC GRAPHIC


2


Page 4 of 22

Emerging technologies in photovoltaics such as dye-sensitized solar cells, organic solar cells and the more recent concept of perovskite solar cells (PSCs) have been the object of intense research in the last few decades.1 These new types of solar cell combine competitive efficiencies with lower fabrication costs.2 PSCs are a paramount example related to the impressive progress accomplished in less than ten years, with a current certified record efficiency of 23.3%.2,3 PSCs are based on hybrid metal halide perovskites (MHPs), which are mixed ionicelectronic semiconductors extensively studied in the last decade.3–6 However, making a PSC device with a state-of-the-art efficiency requires good electron and hole selective contacts. In a seminal work by Lee et al.7, it was demonstrated that even a mesoporous Al2O3 insulating film on top of compact TiO2 layer is sufficient to extract the full potential of a MHP, however, in this case the compact TiO2 layer is the electron-selective contact. Good contacts guarantee that the large number of electron-hole pairs photogenerated within the active layer are not lost via surface or bulk recombination processes, related to bad transport or inadequate band offset. For instance in the work by Juárez-Pérez et al.8 it is nicely shown that a “stand-alone” MHP film produces no rectifying behavior, and that adding a hole or electron (or both) selective layer is necessary to increase the short-circuit photocurrent (Jsc) and the open-circuit photovoltage (Voc). In addition, it is well-known that using a “bad” hole selective layer leads to a substantially lower Voc.9–12 It is also a common finding in the field that different contacts produce a quite distinct hysteresis behavior.13 Recently Ravishankar and coworkers14 studied high-efficiency PSCs prepared with FTO, compact TiO2 and mesoporous TiO2, and found that the Voc was determined by quasi-Fermi level splitting within the MHP rather than by the work function of the electron selective


3

Page 5 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


layer. Very recently15 we reported that different morphologies and crystallinities of mesoporous TiO2 contacts do not lead to substantial differences in Voc, but produce a high impact on the Jsc and the hysteresis behavior instead. Detailed numerical modelling showed that this feature is a consequence of increased surface recombination for contacts where electron extraction is not sufficiently efficient. In this work, we have studied the effect of different n-type electron selective contacts (ESC) in the performance of mesoporous methylammonium lead iodide (CH3NH3PbI3 = MAPI) perovskite solar cells. In addition to the typically employed ESC, Dyesol TiO2 paste, this study includes the three crystalline phases of TiO2 (anatase, rutile and brookite) and zinc stannate (ZTO) synthesized in our lab, as well as commercial alumina (Al2O3) and zirconia (ZrO2). ESC/MAPI films were characterized by optical absorption and steady state photoluminescence (PL) measurements. Full devices were studied by current-voltage voltammetry and impedance spectroscopy (IS) to investigate trends in the photovoltaic parameters with respect to the nature of the contact and their link to the recombination kinetics. The results suggest that the different charge separation properties of the ESC/PS interface have a strong impact on the hysteresis and the Jsc, but that the effect on the Voc is small, except for less efficient electron contacts. By charge separation we mean the combined effect of the charge extraction (electron injection) and the surface recombination, which determines the “selectivity” of the contact but which are difficult to distinguish in the experiments. We attribute our observations to the predominance of the bulk recombination under open-circuit conditions and strong surface recombination at short-circuit for poor electron-extracting contacts. Perovskite

solar

cells

were

fabricated

following

the

structure

FTO/cTiO2/ESC(m)/MAPI/Spiro-OMeTAD/Au. Of the seven different materials studied as


4


Page 6 of 22

electron extraction layers, in addition to commercial TiO2 paste (Dyesol TiO2 18NR-T), two of these materials were obtained from commercial sources, ZrO2 paste (Solaronix Zrnanoxide ZT-SP) and Al2O3 nanoparticles (Sigma Aldrich < 50 nm), while the other five materials were synthesized: ZnO, Zn2SnO4 (ZTO) and the three crystalline phases of TiO2 (rutile, anatase and brookite). The details of the synthesis of these materials, the MAPI and the preparation of the paste for spin-coating deposition, as well as structural data can be found in Experimental Section below and in the Supporting Information (Figures S1 and S2). In Figure 1, Figure S3 and Table 1 the photovoltaic performance of the fabricated devices is summarized.

Figure 1. Current-voltage curves for champion MAPI devices (left) and result of their normalization with respect the short-circuit photocurrent (right). Measurements were carried out under 1-sun AM1.5G standard conditions with a scan rate of 50 mV/s in reverse scan.

Table 1. Photovoltaic parameters obtained for the freshly prepared PSCs (with CH3NH3PbI3 = MAPI) fabricated with the different electron selective contacts. The measurements were performed under standard


5

Page 7 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


illumination conditions with a scan rate of 50 mV/s in reverse scan direction. Samples marked in red correspond to the configurations chosen for further studies. Average and standard deviation values correspond to results obtained from 4 to 6 solar cells per material.

The best performance was achieved in the solar cells fabricated with Dyesol paste, with the best cell reaching an efficiency of 15.2%. The average performance of the devices fabricated with microwave-assisted sol-gel synthesized anatase was actually very close to the average values obtained for the Dyesol paste. This shows that in analogy to dyesensitized solar cells,16 where anatase is by far the best performing photoanode, the choice of the right crystalline polymorph of TiO2 is also critical for good electron selectivity in PSCs. Zinc stannate (ZTO)-based devices showed an average efficiency of around 10%, with the best performance reaching 12.3%. The rest of the materials showed a significantly lower performance. With the exception of alumina, it is a general result that all contacts produce a high Voc close to 1 V. In fact, when the photocurrent is normalized with respect to its short-circuit value, all devices yield the same Voc (with the alumina exception). Hence, the loss of performance is mainly due to a reduced Jsc and fill factor. In order to gain more insight into the origin of these differences in the photovoltaic parameters, further characterization was carried out using optical and IS measurements. To simplify the analysis and work with reliable data, we focus these studies on the four


6


Page 8 of 22

configurations that yield the best performance (marked in red in Table 1). Brookite, rutile and alumina configurations showed significantly smaller efficiencies in addition to limited stability and reproducibility, which complicated the IS analysis. Figure S4 in the Supporting Information displays the absorbance of MHP films deposited on the four chosen ESC materials. The obtained curves show unambiguously that the nature of the contact does not affect the light harvesting capabilities of the perovskite active layer, with a predicted optical band gap of 1.6 eV for the four configurations. PL data for the same ESC/MHP layers is reported in Figure 2.

Figure 2. (Left) Normalized photoluminescence spectra of the MAPI films deposited on the different electron selective materials. (Right) The Jsc values obtained from the same MAPI films turned into devices is represented against the maximum intensity of the corresponding PL spectrum. Note that, to show the reproducibility of the experiment, data from 2-3 samples of the same configuration are included in the graphs.

The PL spectra feature the typical peak of band-to-band radiative recombination of MHPs with a maximum close to 800 nm.5,17 The intensity of the peak is a direct indication of the quantum yield for electron injection into the ESC: the lower the peak the faster the injection kinetics. The results obtained show that Dyesol and anatase layers are much more


7

Page 9 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


efficient in extracting electrons out of the perovskite layer than their ZTO and zirconia counterparts. In fact, it is possible to find an inverse correlation (Figure 2 right) between the normalized PL intensity at the maximum and the Jsc values obtained in the full devices. This correlation, and the independence of the absorbance data with respect to the type of contact, strongly suggests that the lower photocurrent obtained for contacts other than anatase TiO2 is a direct consequence of poor charge separation due to slow electron injection at the ESC/MAPI interface. As mentioned above, the nature of the ESC also influences the hysteresis behavior of PSCs. In Figure 3, cyclic voltagramms for the four configurations studied at different scan rates are collected. All devices show “normal” hysteresis, that is, higher fill factor values are always obtained in the reverse scan, and the difference between reverse and forward scans (hysteresis index) is always larger for faster scan rates. Dyesol and anatase exhibit a quite similar hysteresis behavior. In contrast, ZTO and especially zirconia present strong hysteresis. In the latter case, a clear bump is observed at intermediate voltages.


8


1

1 DYESOL

Normalized Photocurrent

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 22

ANATASE

500 mV/s 100 mV/s 10 mV/s 1 mV/s

0

0.0

0.5

1.0

0.0

0.5

1.0

0.5

1.0

1

1

ZIRCONIA ZINC TIN OXIDE

0

0.0

0.5

1.0

0.0

Voltage (V) Figure 3. Current-voltage curves for devices marked in red in Table 1 at different scan rates. Current has been normalized to its short-circuit value.

Although a precise and unambiguous description of hysteretic phenomena in PSC is still lacking, there is growing evidence that this phenomenon is produced by the combination of the relatively slow motion of the ions18,19 and interfacial phenomena.20,21 In the present study, the perovskite layer is common to all devices and only the electron selective contact is modified. Although a fundamental description of the hysteresis is out of the scope of this work, our results clearly show that the nature of the ESC can provoke very different types of hysteresis, including “inverted” hysteresis for ZTO at the slowest scan rate. However, the most eye-catching observation is that for a poor electron-extracting material such as ZrO2 the hysteresis is very large. The J-V curves obtained for zirconia display a striking


9

Page 11 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


resemblance to those obtained before for PSCs with an amorphous TiO2 contact15, with substantially lower Jsc values and a prominent bump at intermediate voltages. As in our previous publication, we attribute this effect to the enhanced contribution of surface recombination, induced by poor electron extraction at the ESC/MAPI interface. Interestingly, Wojciechowski and coworkers17 also found that modifying the TiO2 surface to enhance charge separation resulted in devices with higher photocurrents and lower hysteresis. The short-circuit current density is thus explained to be a consequence of the different extraction capabilities of the devices studied. However, as mentioned before, the photovoltage remains essentially unaltered by the extraction features of the different ESCs. Following previous work,14,20,22 we have investigated the device performance under opencircuit conditions. In Figure 4, the Voc is shown as a function of the temperature of the device. A linear dependence is observed for all four configurations.


10


1050

DYESOL

1050

EA= 1.68 eV

1000

ANATASE EA= 1.59 eV

1000

m= -2.3

Open-circuit voltage (V)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 22

m= -2.4

950

950

900

900

260

280

300

320

340

1050

ZTO

1000

EA= 1.63 eV

260

280

300

EA= 1.59 eV

1000

m=-2.3

950

950

900

900

280

300

320

340

ZIRCONIA

1050

m= -2.1

260

320

340

260

280

300

320

340

Temperature (K) Figure 4. Open-circuit voltage as a function of the absolute temperature for the configurations marked in red in Table 1. In the graph the results of the fittings to Eq. (3) are shown.

In a previous publication,22 we provided a general description of the recombination kinetics under open-circuit conditions. Following this description, we assume that at high frequencies, for a “frozen” ionic distribution within the active layer, the recombination rate in units of volume-1 and time-1 is given by a rate law of the type 𝑑𝑛

𝑈𝑟𝑒𝑐 ≈ ― 𝑑𝑡 = 𝑘𝑇𝑝0𝑛𝛾

(1)

where n is the minority carrier concentration in the active layer (electrons for a p-type semiconductor), p0 is the majority carrier concentration, kT is a rate constant and  is the


11

Page 13 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


recombination reaction order. It is possible to relate the carrier concentration with the photovoltage, V and the band gap, Eg22,23

(

𝐸𝑔 ― 𝑞𝑉

𝑛 = (𝑁𝐶𝑁𝑉)1/2𝑒𝑥𝑝 ―𝛼

𝑘𝐵𝑇

)

(2)

where Nc and Nv represent the density of states of the conduction and the valence band, respectively. Combining Eqs. (1) and (2) it is possible to find the recombination rate as a function of open-circuit potential at high injection conditions (n = p). Imposing the opencircuit condition (generation rate = recombination rate) one finds from (1) and (2)22 𝑉𝑂𝐶 =

𝐸𝑔 𝑞

―

𝑚𝑘𝐵𝑇 𝑞

𝐽00

( )

𝑙𝑛

𝑞𝑑𝐺

(3)

where m, the ideality factor, is given by m = 1/(γ), with  = 0.5 for high-injection conditions. Equation (3) predicts a linear relationship between Voc and temperature. Figure 4 confirms this prediction and also provides an estimation of the band gap of the perovskite. As a matter of fact, all devices yield values very close to the optical band gap of MAPI of 1.61 eV. This outcome indicates that the dominant recombination process quantified by Eq. (1) takes place in the bulk of the perovskite or it is determined by the perovskite only.24 To confirm this interpretation, we have carried out impedance spectroscopy measurements under open-circuit conditions. These studies were performed at two optical excitation wavelengths, blue and red light respectively, in order to detect possible spatial inhomogeneities in the impedance response. This is especially crucial if we want to assess the impact of the ESC, which is also the front contact. Thus, using blue light (λblue = 465 nm), most carriers are photogenerated close to and within the mesoporous ESC layer, whereas red light is absorbed throughout the MAPI capping layer. Differences between red and blue light can therefore be a signature of interfacial processes.15,22


12


Page 14 of 22

In Figure S5 in the Supporting information impedance spectra for the four selected ESCs are presented. The spectra of the Dyesol devices exhibit the typical response22,25–27 of a PSC with two main signals (peaks in the frequency plot or arcs in the Nyquist plot) at low (0.110 Hz) and high (104-105 Hz) frequencies. However, a third signal at mid frequencies shows up for the anatase, ZTO and zirconia devices. Comparison of the “red” and “blue” data reveals that both excitations give in essence the same results for Dyesol and ZTO contacts. In contrast, anatase and zirconia reveal clear differences at low and mid frequencies. Given the strong hysteresis observed in devices with anatase and, especially, zirconia contact, this feature and the presence of the mid-frequency signal in the spectra confirms the presence of additional processes in the vicinity of the ESC/MAPI interface. Zirconia gives a larger capacitance with blue light at low frequencies, and the low frequency limit of the apparent capacitance (Figure S6) is also larger for zirconia in comparison with the rest of the ESCs. As in Ref.15, we interpret this behavior and the presence of the mid-frequency signal as a signature of a surface recombination process provoked by the poor electron injection efficiency at the zirconia contact. The fact that this shows up a low frequencies combined with the results of the modeling in Ref.15, points to the important role of ionic migration in this process. To quantify and describe this complex interplay between ion migration and surface recombination and the corresponding signature in a small-perturbation measurement is a challenge that falls outside the scope of this work. However, it is important to stress that the different low-frequency features of the devices do not affect the recombination kinetics that determine the open-circuit photovoltage, which is the main message of our work. The recombination resistance is the inverse of the voltage derivative of the recombination current derived from Eqs. (1) and (2)


13

Page 15 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


𝑅𝑟𝑒𝑐 =

∂𝐽𝑟𝑒𝑐 ―1

( ) ∂𝑉

(

= 𝑅00𝑒𝑥𝑝 ―

)

𝛽𝑞𝑉𝑂𝐶 𝑘𝐵𝑇

(4)

with R00 = (kBT J00/qγ)exp(γEg/kBT) and β = γ. From this result we also get m = 1/β. This equation predicts an exponential dependence for recombination resistance with respect to open-circuit potential.

Figure 5. Recombination (high frequency) resistance as a function of the open-circuit potential Voc for the selected configurations marked in red in Table 1, where the VOC is varied by changing the illumination light intensity. In the graphs the results of the fittings to Eq. (4) are shown.


14


Page 16 of 22

It is important to establish how the recombination resistance can be estimated from the impedance spectrum. To do so, the spectra can be fitted28 to simple equivalent circuits of the tipe -Rs-[-(RiCPEi)-]i=1-3, that is, as a series connection of RC elements whose number depend on the number of peaks that show up in the spectrum. However, given the complexity of the spectra, in this work we have focused on the high frequency region only. In Figure 5 results for the high frequency resistance as extracted from a simple RC fit of the high frequency peak of the impedance data is presented. The data area plotted versus the VOC for both red and blue light illumination. As predicted by Eq. (4) an exponential behavior is clearly reproduced, with a slope that is consistent with   0.5 (m  2), signature of Shockley-Read-Hall (trap-limited) recombination in the bulk of the perovskite.22,24,29,30 It is important to remark that only the high frequency component of the impedance spectra reproduces the ideality factor of the Voc measurements, and the similar fits of the mid frequency and low frequency signal do not lead to the correct ideality factor. We explain this observation based on the fact that only at high frequencies, where the ionic distribution is frozen, pure electronic effects (Eq. (1)) are taking place. It is illustrative to compare the recombination resistance of the devices with different ESC materials (Figure S7). In general, a higher 1-sun Voc value (Table 1) is correlated with a higher recombination resistance at the same potential in Figure 5 (corresponding to a lower recombination rate at given Fermi-level splitting). The extreme case is zirconia for which RHF is the lowest. This observation indicates that surface recombination is also having an effect at open circuit if the ESC is not performing well, as it is the case of zirconia. Kirchatz et al.30 recently discussed the relative impact of surface recombination velocities on the open-circuit photovoltage using up-to-date literature parameters. Although


15

Page 17 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


marginal with respect to bulk recombination, surface recombination can account for a reduction of up to 50 mV in the Voc. Experimental results also confirm that improving charge separation at the electron-selective contact, for instance by passivating with a thin layer of a fullerene derivative,17,31 results in lower hysteresis and improved photovoltaic parameters, in line with what we observe here. In summary, we have measured and analyzed the performance of perovskite solar cells based on electron selective contacts of different chemical nature and crystal structure. From current-voltage curves, the temperature dependence of the Voc, and impedance experiments under open circuit conditions, we conclude that all studied contacts except alumina the open-circuit photovoltage is essentially the same because it is governed by recombination events that take place in the bulk of the perovskite active layer. The impact of the electron selective contact is mainly seen in the short-circuit photocurrent. Thus, a correlation is found between the charge extracting capability of each material, as derived from photoluminescence data, and the value actually achieved for the photocurrent. A material with low performance in this respect, such as zirconia, presents the lowest value of the short-circuit photocurrent and a strong hysteresis. Impedance results confirm the bulk nature of the recombination at open circuit in all devices and reveal a quite characteristic behavior at low frequencies for the zirconia-based devices, indicative of additional surface recombination, which leads to stronger hysteresis and lower photocurrent. Device fabrication The devices were prepared on FTO coated glass (25 × 25 mm, Pilkington TEC15) with a laser patterned etching. The substrates were cleaned with soap (Hellmanex), rinsed with Milli-Q water and subsequently washed with ethanol, isopropanol and acetone in an


16


Page 18 of 22

ultrasonic bath. The blocking layer used in all devices was deposited by spray pyrolysis at 450 °C, using a titanium diisopropoxide bis(acetylacetonate) (75% in isopropanol, SigmaAldrich) solution diluted in ethanol (1:19 v/v), with oxygen as the carrier gas. The precursor solution was sprayed in 3 steps, waiting 30 seconds between each step. After the spraying process, the films were kept at 450 °C for 30 min. In some of the experiments the blocking layer deposited by spray pyrolysis was substituted by a blocking layer deposited by atomic layer deposition (ALD), using tetrakis (dimethylamido) titanium(IV) (TDMAT) and deionized water as precursors. The dosing time was: 0.1 s for TDMAT and 0.015 s for water. The precursors were transported to the chamber using nitrogen gas at a flow rate of 20 sccm; nitrogen was also used to purge the chamber for 10 s between the dosing pulses. Deposition was performed at a substrate temperature of 100 °C. No variation of the performance was observed with respect to the blocking layers deposited by spray pyrolysis (see Supporting Information, Table S1 and Figure S8). The MAPI films were deposited in a one-step process, in air, at a relative humidity between 30 and 40% using diethyl ether as antisolvent. The precursor solution was prepared mixing a solution of PbI2 and MAI with a mixture of DMSO and DMF. The ratio of the solvents was in function of the relative humidity in the environment following the outlines of a previous report32. A volume of 50 µL of the perovskite precursor solution was deposited on the mesoporous substrate and spin-coated at 4000 rpm for 50 s. The diethyl ether was dispensed during the third second of the spinning cycle. After the deposition step, the substrate was sintered at 100 °C for 3 minutes. The perovskite films were covered with the hole-transporting material by spin coating at 4000 rpm for 30 seconds using 50 μL of the following spiro-OMeTAD solution: 1 mL chlorobenzene, 72.3 mg (2,2′,7,7′tetrakis(N,N′-di-p-methoxyphenylamine)-9,9′-

spirobifluorene),


28.8

μL

4-tert17

Page 19 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


butylpyridine, and 17.5 μL of a stock solution of 520 mg/mL lithium bis(trifluoromethylsulfonyl)imide in acetonitrile. Finally, a 60 nm gold film was thermally evaporated on top of the device to form the electrode top contact. Electrical characterization Perovskite films were deposited on the different electron extraction layers for their optical characterization, including UV-Vis and steady-state photoluminescence (PL) spectrophotometry, the latter at an excitation wavelength of 532 nm. The performance of the devices was determined from the current density-voltage curves measured under 1 sun illumination conditions, with a masked area of 0.16 cm2 at a scan rate of 50 mV/s. Additionally, cyclic voltammetry was performed at different scan rates to study the hysteresis of the devices. The effect of the different ESC materials on the recombination dynamics of the devices was probed using impedance spectroscopy (IS). The illumination for the IS measurements was provided by a red and a blue LED (λblue = 465 nm and λred = 635 nm) over a range of five DC light intensities. The impedance spectra were obtained applying a 20 mV perturbation in the 106 -10-2 Hz frequency range.

Supporting Information

Experimental details, SEM cross-sectional images, XRD patterns, UV/Vis absorption spectra, Nyquist and Bode plots from impedance experiments. Comparison between ALD and spray-pyrolysis. Statistics of short-circuit photocurrent data for the selected electron selective materials.

ACKNOWLEDGMENTS We thank Junta de Andalucía for financial support via grant FQM 1851 and FQM 2310. We thank Ministerio de Economía y Competitividad of Spain and Agencia Estatal de


18


Page 20 of 22

Investigación (AEI) and EU (FEDER) under grants MAT2013-47192-C3-3-R, MAT201676892-C3-2-R, MAT2016-79866-R and Red de Excelencia “Emerging photovoltaic Technologies” for financial support. We also thank Servicio de Microscopía Electrónica (Universidad Pablo de Olavide). CONACyT is gratefully acknowledged for funding under the Fronteras de la Ciencia program, grant number FDC-2015-110.

REFERENCES (1) Karazhanov, S.; Kharton, V. Novel Materials for Solar Cells and Related Technologies. Mater. Lett. 2018, 228, 450. https://doi.org/10.1016/j.matlet.2018.06.078. (2) NREL. NREL: National Center for Photovoltaics Home Page http://www.nrel.gov/ncpv/ (accessed Mar 5, 2015). (3) Snaith, H. J. Perovskites: The Emergence of a New Era for Low-Cost, HighEfficiency Solar Cells. J. Phys. Chem. Lett. 2013, 4 (21), 3623–3630. https://doi.org/10.1021/jz4020162. (4) Park, N.-G. Organometal Perovskite Light Absorbers Toward a 20% Efficiency Low-Cost Solid-State Mesoscopic Solar Cell. J. Phys. Chem. Lett. 2013, 4 (15), 2423– 2429. https://doi.org/10.1021/jz400892a. (5) Manser, J. S.; Christians, J. A.; Kamat, P. V. Intriguing Optoelectronic Properties of Metal Halide Perovskites. Chem. Rev. 2016. https://doi.org/10.1021/acs.chemrev.6b00136. (6) Tress, W. Perovskite Solar Cells on the Way to Their Radiative Efficiency Limit – Insights Into a Success Story of High Open‐Circuit Voltage and Low Recombination. Adv. Energy Mater. 2017, 7 (14), 1602358. https://doi.org/10.1002/aenm.201602358. (7) Lee, M. M.; Teuscher, J.; Miyasaka, T.; Murakami, T. N.; Snaith, H. J. Efficient Hybrid Solar Cells Based on Meso-Superstructured Organometal Halide Perovskites. Science 2012, 338 (6107), 643–647. https://doi.org/10.1126/science.1228604. (8) Juarez-Perez, E. J.; Wuβler, M.; Fabregat-Santiago, F.; Lakus-Wollny, K.; Mankel, E.; Mayer, T.; Jaegermann, W.; Mora-Sero, I. Role of the Selective Contacts in the Performance of Lead Halide Perovskite Solar Cells. J. Phys. Chem. Lett. 2014, 5 (4), 680– 685. https://doi.org/10.1021/jz500059v. (9) Bi, D.; Yang, L.; Boschloo, G.; Hagfeldt, A.; Johansson, E. M. J. Effect of Different Hole Transport Materials on Recombination in CH3NH3PbI3 Perovskite-Sensitized Mesoscopic Solar Cells. J. Phys. Chem. Lett. 2013, 4 (9), 1532–1536. https://doi.org/10.1021/jz400638x. (10) Christians, J. A.; Fung, R. C. M.; Kamat, P. V. An Inorganic Hole Conductor for Organo-Lead Halide Perovskite Solar Cells. Improved Hole Conductivity with Copper Iodide. J. Am. Chem. Soc. 2014, 136 (2), 758–764. https://doi.org/10.1021/ja411014k. (11) Pydzińska, K.; Karolczak, J.; Kosta, I.; Tena-Zaera, R.; Todinova, A.; Idígoras, J.; Anta, J. A.; Ziółek, M. Determination of Interfacial Charge-Transfer Rate Constants in


19

Page 21 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Perovskite Solar Cells. ChemSusChem 2016, 9 (13), 1647–1659. https://doi.org/10.1002/cssc.201600210. (12) Salado, M.; Idigoras, J.; Calio, L.; Kazim, S.; Nazeeruddin, M. K.; Anta, J. A.; Ahmad, S. Interface Play between Perovskite and Hole Selective Layer on the Performance and Stability of Perovskite Solar Cells. ACS Appl. Mater. Interfaces 2016, 8 (50), 34414– 34421. https://doi.org/10.1021/acsami.6b12236. (13) 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 (17), 2927–2934. https://doi.org/10.1021/jz501392m. (14) Ravishankar, S.; Gharibzadeh, S.; Roldán-Carmona, C.; Grancini, G.; Lee, Y.; Ralaiarisoa, M.; Asiri, A. M.; Koch, N.; Bisquert, J.; Nazeeruddin, M. K. Influence of Charge Transport Layers on Open-Circuit Voltage and Hysteresis in Perovskite Solar Cells. Joule 2018, 2 (4), 788–798. https://doi.org/10.1016/j.joule.2018.02.013. (15) Idígoras, J.; Contreras‐Bernal, L.; Cave, J. M.; Courtier, N. E.; Barranco, Á.; Borras, A.; Sánchez‐Valencia, J. R.; Anta, J. A.; Walker, A. B. The Role of Surface Recombination on the Performance of Perovskite Solar Cells: Effect of Morphology and Crystalline Phase of TiO2 Contact. Adv. Mater. Interfaces 2018, 5 (21), 1801076. https://doi.org/10.1002/admi.201801076. (16) Park, N. G.; Lagemaat, J. van de; Frank, A. J. Comparison of Dye-Sensitized Rutileand Anatase-Based TiO2 Solar Cells. J. Phys. Chem. B 2000, 104 (38), 8989–8994. (17) Wojciechowski, K.; Stranks, S. D.; Abate, A.; Sadoughi, G.; Sadhanala, A.; Kopidakis, N.; Rumbles, G.; Li, C.-Z.; Friend, R. H.; Jen, A. K.-Y.; et al. Heterojunction Modification for Highly Efficient Organic–Inorganic Perovskite Solar Cells. ACS Nano 2014, 8 (12), 12701–12709. https://doi.org/10.1021/nn505723h. (18) Chen, B.; Yang, M.; Priya, S.; Zhu, K. Origin of J–V Hysteresis in Perovskite Solar Cells. J. Phys. Chem. Lett. 2016, 7 (5), 905–917. https://doi.org/10.1021/acs.jpclett.6b00215. (19) Richardson, G.; O’Kane, S. E. J.; Niemann, R. G.; Peltola, T. A.; Foster, J. M.; Cameron, P. J.; Walker, A. B. Can Slow-Moving Ions Explain Hysteresis in the Current– Voltage Curves of Perovskite Solar Cells? Energy Environ. Sci. 2016, 9, 1476–1485. https://doi.org/10.1039/C5EE02740C. (20) Ravishankar, S.; Almora, O.; Echeverría-Arrondo, C.; Ghahremanirad, E.; Aranda, C.; Guerrero, A.; Fabregat-Santiago, F.; Zaban, A.; Garcia-Belmonte, G.; Bisquert, J. Surface Polarization Model for the Dynamic Hysteresis of Perovskite Solar Cells. J. Phys. Chem. Lett. 2017, 8 (5), 915–921. https://doi.org/10.1021/acs.jpclett.7b00045. (21) Contreras, L.; Idígoras, J.; Todinova, A.; Salado, M.; Kazim, S.; Ahmad, S.; Anta, J. A. Specific Cation Interactions as the Cause of Slow Dynamics and Hysteresis in Dye and Perovskite Solar Cells: A Small-Perturbation Study. Phys. Chem. Chem. Phys. 2016, 18 (45), 31033–31042. https://doi.org/10.1039/C6CP05851E. (22) Contreras-Bernal, L.; Salado, M.; Todinova, A.; Calio, L.; Ahmad, S.; Idígoras, J.; Anta, J. A. Origin and Whereabouts of Recombination in Perovskite Solar Cells. J. Phys. Chem. C 2017, 121 (18), 9705–9713. https://doi.org/10.1021/acs.jpcc.7b01206. (23) Sze, S. M.; Ng, K. K. Physics of Semiconductor Devices; John Wiley & Sons, 2006. (24) Leong, W. L.; Ooi, Z.-E.; Sabba, D.; Yi, C.; Zakeeruddin, S. M.; Graetzel, M.; Gordon, J. M.; Katz, E. A.; Mathews, N. Identifying Fundamental Limitations in Halide Perovskite Solar Cells. Adv. Mater. 2016, 28 (12), 2439–2445. https://doi.org/10.1002/adma.201505480.


20


Page 22 of 22

(25) Guerrero, A.; Garcia-Belmonte, G.; Mora-Sero, I.; Bisquert, J.; Kang, Y. S.; Jacobsson, T. J.; Correa-Baena, J.-P.; Hagfeldt, A. Properties of Contact and Bulk Impedances in Hybrid Lead Halide Perovskite Solar Cells Including Inductive Loop Elements. J. Phys. Chem. C 2016, 120 (15), 8023–8032. https://doi.org/10.1021/acs.jpcc.6b01728. (26) Pockett, A.; Eperon, G. E.; Sakai, N.; Snaith, H. J.; Peter, L. M.; Cameron, P. J. Microseconds, Milliseconds and Seconds: Deconvoluting the Dynamic Behaviour of Planar Perovskite Solar Cells. Phys. Chem. Chem. Phys. 2017, 19 (8), 5959–5970. https://doi.org/10.1039/C6CP08424A. (27) Zarazua, I.; Han, G.; Boix, P. P.; Mhaisalkar, S.; Fabregat-Santiago, F.; Mora-Seró, I.; Bisquert, J.; Garcia-Belmonte, G. Surface Recombination and Collection Efficiency in Perovskite Solar Cells from Impedance Analysis. J. Phys. Chem. Lett. 2016, 7 (24), 5105– 5113. https://doi.org/10.1021/acs.jpclett.6b02193. (28) Todinova, A.; Contreras-Bernal, L.; Salado, M.; Ahmad, S.; Morillo, N.; Idígoras, J.; Anta, J. A. Towards Universal Approach to the Analysis of Impedance Spectra of Perovskite Solar Cells. Equivalent Circuits and Empirical Analysis. ChemElectroChem. https://doi.org/10.1002/celc.201700498. (29) Tress, W.; Yavari, M.; Domanski, K.; Yadav, P.; Niesen, B.; Baena, J. P. C.; Hagfeldt, A.; Graetzel, M. Interpretation and Evolution of Open-Circuit Voltage, Recombination, Ideality Factor and Subgap Defect States during Reversible Light-Soaking and Irreversible Degradation of Perovskite Solar Cells. Energy Environ. Sci. 2018, 11, 155–165. https://doi.org/10.1039/C7EE02415K. (30) Kirchartz, T.; Krückemeier, L.; Unger, E. L. Research Update: Recombination and Open-Circuit Voltage in Lead-Halide Perovskites. APL Mater. 2018, 6 (10), 100702. https://doi.org/10.1063/1.5052164. (31) Li, Y.; Zhao, Y.; Chen, Q.; Yang, Y. (Michael); Liu, Y.; Hong, Z.; Liu, Z.; Hsieh, Y.-T.; Meng, L.; Li, Y.; et al. Multifunctional Fullerene Derivative for Interface Engineering in Perovskite Solar Cells. J. Am. Chem. Soc. 2015, 137 (49), 15540–15547. https://doi.org/10.1021/jacs.5b10614. (32) Aranda, C.; Cristobal, C.; Shooshtari, L.; Li, C.; Huettner, S.; Guerrero, A. Formation Criteria of High Efficiency Perovskite Solar Cells under Ambient Conditions. Sustain. Energy Fuels 2017, 1 (3), 540–547. https://doi.org/10.1039/C6SE00077K.


21

Correlation between the Effectiveness of the Electron-Selective

Recommend Documents