Effect of Different Lead Precursors in Perovskite Solar Cells Performance and Stability

We present the use of halide (PbCl2) and non-halide lead precursors (Pb(OAc)2 (OAc=CH3CH2COO ), Pb(NO3)2, Pb(acac)2 (acac=(CH3COCHCOCH3) ) and PbCO3) for the preparation of perovskite solar cells. We have confirmed by X-ray diffraction the growth of CH3NH3PbI3 in all the analyzed cases, except for PbCO3, independently of the lead precursor used for the synthesis of the perovskite. In addition, different cell configurations, thin film and mesoporous scaffolds, TiO2 or Al2O3, have also been prepared. We have observed that the lead precursor influences strongly on the structural properties of perovskite (grain size), as well as on the solar cell performance. Photovoltaic conversion efficiencies comparable to those achieved when using the commonly employed PbCl2 have been obtained with Pb(OAc)2 as lead source. Stability studies of the perovskite films and devices have also been carried out; demonstrating that the lead precursor also influences this aspect. Stability is strongly affected by atmosphere and illumination conditions, but also by the lead precursor employed for the perovskite synthesis. These results highlight that other lead sources, different to the commonly used PbCl2 and PbI2, are also suitable for the development of PSCs, opening a new way for device performance optimization.


Introduction
Nowadays, hybrid halide perovskites with general formula ABX 3 , where X=Cl, Br or I, can be considered, without any doubts, an ideal candidate for the preparation of photovoltaic devices.2][3] The first example of the use of hybrid halide perovskites in solar cells was reported by Miyasaka in 2009, 4 and from that moment the field of Perovskite Solar Cells (PSCs) experienced the best ever efficiency enhancement within an outstandingly short period of time. 5In 2012, the use of perovskite for photovoltaics took a step forward with the preparation of all-solid devices with spiro-OMeTAD, as selective contact and Hole Transporting Material (HTM), producing 10-11% efficiency. 6,78][19][20] In addition, high efficiency samples have been obtained from a broad variety of deposition techniques, including spin-coating 6,7 two step procedures involving dip-coating 21 or drop casting methods 22 and solvent engineering strategies. 1 Besides, evaporation procedures have also been demonstrated to be excellent processes for the preparation of perovskite films. 235][26] However, most of these perovskites, synthesized from non-halide precursors, have not been analyzed in complete photovoltaic devices yet.Therefore, in this article we study the performance and stability of PSCs prepared from non-halide lead precursors.These cells were prepared in mesoporous heterojunction configuration glass/FTO/compactTiO 2 /mesoporous scaffold /CH Hole transport layer (HTM) deposition.A ∼300-400 nm-thick of HTM was deposited on top of the perovskite substrates by spin coating at 4000 r.p.m for 30 s under air conditions, using 100 µL of spiro-OMeTAD solution.The spiro-OMeTAD solution was prepared by dissolving 72.3 mg of (2,2′,7,7′-tetrakis(N,N′-di-pmethoxyphenylamine)-9,9′-spirobifluorene), 28.8 µL of 4-tert-butylpyridine and 17.5 µL of a stock solution of 520 mg/mL of lithium bis(trifluoromethylsulphonyl)imide in acetonitrile, in 1 mL of chlorobenzene.
Gold electrodes deposition.The deposition of 60 nm of gold was performed by thermal evaporation under ultrahigh vacuum conditions, using a commercial MBraun vacuum chamber.Before beginning the evaporation the chamber was evacuated until pressure of 1•10 -6 mbar.
Perovskite films characterization.The morphology and structural properties of the films were analyzed using a JEOL 7001F scanning electron microscope with a film emission gun (SEMFEG) and a Bruker AXS-D4 Endeaver Advance X-ray diffractometer (XRD) using Cu Kα radiation.The PL spectra of the films were registered by using a spectrophotometer based on a CCD (Andor i-DUS DV420A-OE) coupled with a spectrograph as a diffraction grating (Newport 77400).Commercial laser diode (532 nm, 52 mW・cm -2 ) was used as excitation source, whose power intensity was adjusted by means of neutral density filters.

Solar cells characterization.
The Incident Photon to Current Eficiency (IPCE) were performed using a Xenon lamp power source coupled with a monochromator controlled by a computer; the photocurrent was measured using an optical power meter 70310 from Oriel Instruments and a Si photodiode to calibrate the system.Current densityvoltage (J-V) curves were performed under 1 sun illumination (100 mW•cm -2 ) using a xenon arc lamp simulator (Sun 2000, ABET Technologies) with an AM 1.5 G spectral filter and a Keithley 2400, previously calibrated with an NREL-calibrated Si solar cell.
All the measurements were performed with an opaque mask of 0.08 cm 2 and without encapsulation.The electrochemical Impedance spectroscopy measurements were carried out by means of a FRA equipped PGSTAT-30 from Autolab under 1 sun illumination conditions at different applied voltages and applying a 30 mV A/C voltage perturbation over the constant applied bias with a frequency ranging.

Results and discussion
With the aim to evaluate the possibility of using non-halide lead sources for the synthesis of perovskite films, we have chosen different lead non-halide reagents: Pb(OAc) 2 , Pb(NO3) 2 , Pb(acac) 2 and PbCO 3 .For comparative purposes, we decided to contrast the results obtained with these lead precursor with those achieved in our laboratory with PbCl 2 precursor as reference.Perovskite films were synthesized as explained in the experimental section.These thin films, obtained using the different lead XRD measurements of the different perovskite films allowed us to confirm that precursors Pb(OAc) 2 , Pb(NO 3 ) 2 and Pb(acac) 2 are suitable candidates for the preparation of CH 3 NH 3 PbI 3 (see Figure 1).Failure of PbCO 3 in the formation of perovskite was attributed to its low solubility in DMF, even at 80ºC (not shown in Figure 1).In all the cases analyzed in Figure 1, diffraction peaks of CH 3 NH 3 PbI 3 are clearly recognized indicating the formation of this perovskite when using the four different lead precursors.PbCl 2 , Pb(OAc) 2 and Pb(NO 3 ) 2 produce CH 3 NH 3 PbI 3 films with preferential (110) orientation, especially the former two, as can be deduced by the relative intensities between diffraction peaks in comparison with a powder samples.As previously observed, the presence of a minor peak at 2Ɵ ~12.75ºC indicates that traces of PbI 2 remain in the film. 17This peak is sensible reduced when Pb(NO 3 ) 2 is used as lead precursor.Presence of PbI 2 phase does not imply necessarily a deleterious effect for device performance, as it has been reported a passivation effect of CH 3 NH 3 PbI 3 by PbI 2 . 29The intensity of this peak increases for Pb(acac) 2 , suggesting that in this case the reaction is not completed.Note that different precursors are compared using the same experimental procedures, probably film and subsequent photovoltaic device will require a different optimized preparation procedure for each precursor for an optimum behavior.
Moreover, the presence of additional peaks in the XRD spectrum of the perovskite film, corresponding to the substrate (FTO and TiO 2 ), for Pb(acac) 2 and to a lesser extent in Pb(NO 3 ) 2 precursors, suggests that further optimization of the reaction process is needed.Absorbance data also proved that perovskite CH 3 NH 3 PbI 3 can be synthesized from the non-halide lead reagents Pb(OAc) 2 , Pb(NO 3 ) 2 and Pb(acac) 2, by comparison with the spectra obtained when using PbCl 2 .As can be seen in Figure 2a (normalized absorption spectra), all non-halide precursors and reference formed perovskite films that present similar absorption spectra features.However, the absorption signal for Pb(NO 3 ) 2 is deformed in comparison with the reference sample.This effect arises from the light scattering of the sample mainly due to the non homogeneous film deposition when using this lead source.Besides, the intensity of the absorption spectrum is significantly lower (Figure S1, Supporting Information), thus suggesting that although perovskite CH 3 NH 3 PbI 3 is formed, further improvement is needed in order to obtain an optimal amount and quality of the perovskite deposited.Additionally, photoluminescence data (Figure 2b) confirmed that all the samples showed the same emission pattern regardless of the lead precursor employed for the perovskite preparation.Curiously, a small shift in the maximum for each emission is observed depending on the lead source, which could be attributed to the different morphology of the perovskite crystal of each film sample.
In fact, previous studies revealed a direct influence of the crystal size on the optical properties of the perovskite films. 30,31 articularly, a slight blue-shift of the absorption and photoluminescence signals is observed for small crystallites.In any case, further studies are required for elucidating the fundamentals of this phenomenon; nevertheless these issues are out of the scope of this work.The morphology of the perovskite films, using Pb(OAc) 2 and PbCl 2 precursors, was analyzed by Scanning Electron Microscope. Figure 3 shows the SEM images taken from samples with and without TiO 2 mesoporous scaffold configuration.Samples prepared with TiO 2 mesoporous scaffold presents an overlayer of perovskite that does not cover completely the mesoporous layer, see Figure 3c and 3d.Interestingly, smaller crystal sizes of perovskite were obtained for the case of Pb(OAc) 2 (in both configurations flat and with mesoporous scaffold), probably due to the fast crystallization, in comparison with PbCl 2 precursor, detected when using this reagent.
These results reflect how the type of lead precursor utilized for the growth of CH 3 NH 3 PbI 3 affects dramatically the morphological properties of perovskite layers.XRD analysis clearly establishes the preparation of perovskite CH 3 NH 3 PbI 3 films from non-halide lead precursors, but different morphological properties have been revealed by SEM micrograph analysis.In order to study the influence on the final photovoltaic performance, PSCs have been prepared using the different lead precursors under study.All devices were prepared using the same experimental procedure (details in Experimental section).Figure 4 and Table 1 present the performance of the best PSCs obtained from each lead precursor.As can be seen, only PSC prepared from Pb(OAc) 2 gave satisfactory results in terms of efficiency, comparable to those obtained with reference PbCl 2 precursor (entry 2 and 1, respectively).On the other hand, Pb(NO3) 2 and Pb(acac) 2 gave PSCs with low efficiencies (entries 3 and 4, respectively).These results demonstrate that perovskite for solar cells devices can be prepared from lead sources different to the conventional halide precursor PbCl 2 or PbI 2 .In addition PSC performance depends dramatically on the lead precursor employed, and optimum performance should require a different optimization of growth conditions and preparation.An important dependence of solar cell efficiency and lead precursor was previously reported for quantum dot sensitized solar cell using PbS as light absorbing material. 32In that case a clear influence of the lead precursor on the trap states was established.or without any scaffold (details in the Experimental section).From the data shown in Table 2, it can be concluded that the best results in terms of efficiency were obtained when using a configuration with a mesoporous TiO 2 layer of 200nm.In all the configurations studied, Pb(OAc) 2 provided lower efficiency than samples prepared with PbCl 2 , however it is important to note that the deposition technique employed for both precursors was the same, being this procedures only optimized for PbCl 2 .All these results are promising regarding the preparation of high efficiency PSCs from non-halide lead precursor, though additional research is needed to find the optimum deposition conditions, presumably different to the ones employed for PbCl 2 .It is noteworthy that the use of Pb(OAc) 2 allowed the preparation of flat perovskite solar cells (without scaffold), whereas this configuration is not straightforward when using PbI 2 instead of PbCl 2 . 33This fact adds an additional degree of versatility in the use of Pb(OAc) 2 as lead source.applied bias samples with TiO 2 scaffold present lower recombination resistance, R rec , than flat samples, indicating a lower recombination rate for planar cells.Nevertheless, few differences are detected regarding the lead precursors when samples using the same configuration are compared.R rec has been obtained by fitting the impedance spectra with the equivalent circuits previously described. 10,34 hese results are in good agreement with the lower recombination rate, and consequently higher V oc , observed for CH 3 NH 3 PbI 3 perovskite in a flat device 35 and in samples with Al 2 O 3 scaffold, 10 in comparison with samples using TiO 2 scaffold, where just halide lead precursors were used for perovskite formation.These studies were performed for perovskites obtained from PbCl 2 and Pb(OAc) 2 precursors.Figure 6 shows the cell parameters of the devices prepared and stored under conditions A and B for several days.It can be clearly observed that independently of the precursor, cells stored under ambient lab illumination (conditions A) present a continuous decrease of their efficiency, showing an efficiency just ~20% of the initial one after 15 days of exposure to conditions A. This efficiency decrease can be attributed mainly to a decrease of photocurrent, and additionally to a reduction of the FF.
Noteworthy, the decrease in cell performance is slightly higher in the case of devices prepared from Pb(OAc) 2 than for the analogues from PbCl 2 .Moreover, under dark store conditions instead of lab illumination (conditions B), performance of reference cells, using PbCl 2 precursor, present a good stability after 32 days.However, during this period the efficiency of cells prepared from Pb(OAc) 2 was reduced to ~40% of the initial value.Under conditions C, stability increases even further and efficiencies around 97% and 84% of the initial values are preserved after two months for samples prepared with PbCl 2 and Pb(OAc) 2 , respectively (see Supporting Information, Table S1 and Figure S4).This study reflects that lead precursors not only affect the film morphology or cell performance, but also the device stability.These aspects have to be considered for further optimization of cell performance and stability using non-conventional precursors. of studies focused on the physical properties of these materials have established that the use of PbCl 2 affects dramatically the properties of perovskite film, such as a significant increase of the diffusion length of free carriers (exceeding 1µm) and photoluminescence lifetime, 40,41 properties that are crucial for the development of high efficient devices.
Nevertheless, the exact amount of Cl incorporated in the crystalline structure was very small (below 3−4%) 42 or even below the detection limit of the techniques employed in the compositional analysis. 43Therefore, the combination of iodide and chloride lead precursors reflects a strong influence of morphology, thus affecting the final cell performance. 43,44 esults in this work point in the same direction, extending the influence of the lead precursor on perovskite properties to non-halide precursors.

Conclusions
We have prepared CH 3 NH 3 PbI 3 perovskite films and PSCs using different lead precursors.We have obtained devices with significant efficiency with the non-halide sources Pb(NO 3 ) 2 , Pb(acac) 2 and Pb(OAc) 2 , reaching efficiencies close to the reference cells prepared with PbCl 2 for the case of precursor Pb(OAc) 2 .We propose a similar reaction process when a lead precursor different to PbI 2 is used.In that case, a 3:1 molar ratio between CH 3 NH 3 I and PbX 2 is needed to ensure the final stoichiometry of perovskite CH 3 NH 3 PbI 3 .By using this experimental condition, an excess of methylammonium salt remains unreacted, in contrast when using PbI

28 Scheme 1 :
Scheme 1: General reactions proposed for the synthesis of CH 3 NH 3 PbI 3 perovskite from different lead precursors.

Figure 1 :
Figure 1: XRD of the films prepared with different lead precursors, with the exception of PbCO 3 that does not produce CH 3 NH 3 PbI 3 with the common synthesis procedure employed.

Figure 2 :
Figure 2: (a) Normalized light absorption at wavelength 750 nm and (b) normalized photoluminescence of the films prepared with different lead precursors.

Figure 4 :
Figure 4: J-V curve for PSCs prepared with different lead precursors using 200 nm of TiO 2 mesoporous scaffold layer.Champion cells obtained for each lead precursor are represented.J-V curves were scanned from positive voltage to zero, at a scan rate of 50 mV•s -1 .

Figure 6 :
Figure 6: Normalized cell parameters short circuit current, J sc , open circuit potential, V oc , fill factor, FF, and photoconversion efficiency, η, to the values obtained after device preparation (day 0).Parameters measured under 1 sun illumination at different times after preparation.Values represented are average values of 10 samples and error bars indicates the standard deviation.Samples were stored at Conditions A (air with light) and Condition B (air without light).The parameters values were obtained from the J-V curves scanned from positive voltage to zero, at a scan rate of 50 mV•s -1 .
2 and CH 3 NH 3 I in 1:1 molar ratio.However, the use of non-iodine lead precursors allows acting in the final properties of the CH 3 NH 3 PbI 3 perovskite films, provoking important implications in the final solar cell performance.We have observed that the morphology of perovskite films, the photoconversion efficiencies and the stability of the devices are strongly influenced by the lead precursor used.The use of Pb(OAc) 2 precursor has produced cells with efficiencies very close to the reference samples prepared with PbCl 2 , being a promising precursors for further improvement of PSCs performance, but presenting lower long term stability than reference.The preparation of PSCs from non-halide lead precursors, different to the normally used PbI 2 and PbCl 2 , clarifies some former questions on perovskite synthesis and properties.The use of non-halide sources opens a new research line in concordance with the use of non-passive reactants for the optimization of this kind of solar cells.These non-passive reactants might affect the final film properties, not only in terms of efficiency but also in stability.
3 NH 3 PbI 3 /spiro-OMeTAD/Au, with FTO/compactTiO 2 /CH 3 NH 3 PbI 3 /spiro-OMeTAD/Au.Performance and stability have been compared to those obtained with the solar cells prepared with lead chloride precursor, CH 3 NH 3 PbI 3-x Cl x as reference.We have observed that the lead precursor employed affects dramatically the perovskite morphology and crystal size, even though the same experimental procedures were followed for the synthesis.Furthermore, preliminary studies of the stability of both perovskite films and PSCs have also been performed and compared with CH 3 NH 3 PbI 3-x Cl x .
2of active electrode area.The substrates were cleaned with soap (Hellmanex) and rinsed with milliQ water and ethanol.Then, the sheets were sonicated for 15 minutes in a solution of acetone: isopropanol (1:1 v/v), rinsed with ethanol and dried with compressed air.After that, a UV/ozone treatment was performed for 15 minutes.Then, a TiO 2 blocking layer was deposited onto the substrates by spray pyrolysis at 450ºC, using a titanium diisopropoxide bis(acetylacetonate) (75% in isopropanol, Sigma-Aldrich) solution diluted in ethanol (1:39, v/v), with oxygen as carrier gas.After the spraying process the films were kept at 450 ºC for 5 minutes.

Table 1 .
Solar cell parameters under 1 sun illumination: short circuit current, J sc , open circuit potential, V oc , fill factor, FF, and photoconversion efficiency, η, as obtained from best PSCs performances by using PbCl 2 , Pb(OAc) 2 , Pb(NO3) 2 and Pb(acac) 2 as lead sources for the synthesis of the perovskite.Cells were prepared using 200 nm of TiO 2 mesoporous scaffold.Cells using the two precursors presenting the highest efficiencies, PbCl 2 and Pb(OAc) 2 , have been prepared with different configurations, i.e TiO 2 or Al 2 O 3 scaffolds

Table 2 .
Different configuration for the PSCs using Pb(OAc) 2 and PbCl 2 as lead sources.IPCE of each cell was measured and integrated, achieving a good agreement with the measured photocurrent obtained in most of the cases.Moreover, it is also interesting to highlight that an important cell parameter, open circuit potential, V oc , depends more strongly on PSCs configuration than in the lead precursor employed.Lower V oc is observed when TiO 2 scaffold is used in comparison with samples prepared with Al 2 O 3 scaffold or no-scaffold, see Table2.From Impedance spectroscopy characterization, see Figure5, it can be clearly observed that at high