(a) Cross section of the inverted planar perovskite solar cell. Scale bar represents 250 nm. The di ff erent layers have been tinted with the color scheme of the device schematic shown in (b). (c) Approximate energy band diagram of the fabricated inverted structure taken from ref. 30. The fi gure was taken from ref. 29 with permission. 

Context in source publication

Context 1

... also a ff ects the photovoltaic performance. Edri and Cahen et al. used electron beam-induced current (EBIC) imaging to study the cell mechanism of CH 3 NH 3 PbI 3 and CH 3 NH 3 PbI 3 À x Cl x . They refer its structure to p – i – n devices. The authors extracted the di ff usion length of electrons and holes from the EBIC contrast near the contacts (Fig. 6), and reveal that the di ff usion length for holes is longer than the di ff usion length for electrons with the former being at least 1 m m. This  nding is in disagreement with di ff usion length of ca. 100 nm as deduced earlier. 35,36 (Additional discussion about the di ff usion length can be found in the next section.) As a result, the CH 3 NH 3 PbI 3 requires an electron conductor (mesoporous TiO 2 ) because of the short di ff usion length of the electrons. In contrast, due to the long di ff usion length of the holes, hole-transporting material is not essential. An interesting and notable observation regarding perovskite- based solar cells is the ability to use them in planar architecture, which means that sensitized architecture isn't necessarily the only option for the perovskite to function in a solar cell. The planar architecture eliminates in  ltration problems of the perovskite and the hole transport material into the porous, which results in less recombination and better reproducibility. Planar architecture opens the way to implementation of other deposition techniques, as will described below. Fig. 7 show a schematic illustration of the layers involved in the planar architecture. Lee and Snaith et al. 24 reported for the  rst time that the electron collector (TiO 2 metal oxide) isn't necessary for the operation of the perovskite solar cells. In their study, they used photo induced absorption (PIA) spectroscopy to examine the charge separation at the mesoporous TiO 2 and Al 2 O 3 coated with perovskite with and without HTM. In the TiO 2 coated with perovskite, the PIA spectrum con  rmed e ff ective sensitization of the TiO 2 by the perovskite. Alternatively, Al 2 O 3 coated with perovskite revealed no PIA signals, which con  rmed the role of alumina as an insulator. A  er adding the HTM, the oxidized species of the HTM (Spiro-OMETAD) created a  er photoexcitation of the perovskite were monitored. The research demonstrated that from the photoexcited perovskite to HTM the hole conductor is highly e ff ective, and the hole transport material is necessary to enable long-lived charged species within the perovskite coated on Al 2 O 3 . Moreover, small perturbation transient photocurrent decay measurements were performed to learn about the e ff ectiveness of the perovskite as electronic charge transport. Charge collection in Al 2 O 3 based devices was faster by a factor of >10 compared with TiO 2 based devices, indicating that the electron di ff usion through the perovskite is faster in Al 2 O 3 devices. The conclusion is that Al 2 O 3 is simply acting as a sca ff old for the device, which means that the perovskite solar cell demonstrated in this work is not a sensitized solar cell. The authors refer the Al 2 O 3 cells as mesosuperstructured. Low temperature planar perovskite solar cell was demonstrated using graphene  akes and presynthesized TiO 2 nanoparticles. 25 The reported e ffi ciency was 15.6%, comparable to high temperature sintering process of TiO 2 based devices. The graphene + TiO 2 electrodes showed lower series resistance than the TiO 2 electrodes; moreover, the recombination resistance was higher in the case of the graphene + TiO 2 electrodes. The cross section of the device and its corresponding energy level diagram are presented in Fig. 8. Additional work on room temperature planar perovskite heterojunction solar cells was reported by using ZnO nanoparticles 27 as an alternative to both mesoporous TiO 2 and Al 2 O 3 sca ff olds. ZnO nanoparticles have a fabrication advantage due to the fact that they only need a spin coating step and do not require a heating and sintering step. Nanoparticles in the size range of 5 nm were made by hydrolysis of zinc acetate in methanol. Optimal  lm thickness was found to be around 25 nm and further increases in thickness did not improve PCE. The best PCE was found to be 15.7% with a 25 nm  lm thickness which was achieved by doing three spin coating layers. The cells were prepared similarly to most planar heterojunction perovskite cells, where the perovskite was deposited on the ZnO layer; then the HTM layer was deposited on top of the perovskite. PCE of 10.2% was found when this fabrication technique was employed on a  exible substrate made of ITO/PET. A recent work on low temperature (sub 150 C) with meso-superstructured perovskite solar cells demonstrate slightly improved e ffi ciency of 15.9%. 28 The use of the planar architecture in  exible device was demonstrated 29 using CH 3 NH 3 PbI 3 À x Cl x , which was sand- wiched between two organic contacts. The device structure employed (Fig. 9) was deemed “ inverted ” from regular device structure. The FTO was  rst coated with p-type material on top perovskite  lm, followed by n-type layer,  nally aluminum was coated on top for the anode. Several p and n type contacts were studied. The best combination was PEDOT:PSS as the p-type layer and a bilayer of PCBM with a compact layer of TiO x as the n-type layer. The titanium was employed only to create a stable electrical contact with the anode. The power conversion achieved for this “ inverted ” structure was 9.8%, with optimal perovskite  lm thickness of 300 – 400 nm. The e ffi ciency of this design was slightly lower than that of the regular cells structure due to a slightly lower FF. An interesting example for planar heterojunction perovskite solar cells is the ability to form semitransparent planar heterojunction solar cells with high e ffi ciency. Semitransparent solar cells can be integrated, for instance, into windows and automotive applications. In order to make semitransparent perovskite solar cells, it is essential to control the morphology of the perovskite thin  lm. One way to do that is to make perovskite “ islands ” which are small enough to appear continuous to the eye, but large enough to absorb the light to deliver high power conversion e ffi ciency. The authors indicate that despite the observed voids in the  lms, high open circuit voltage and approximately 8% e ffi ciency with 10% average of visible transmittance of the full device was observed. As mentioned earlier one of the attractive properties of the perovskite is the ability to tune its optical properties by chemical modi  cations. One possible way to do it is to substitute the Pb with Sn, which results in an additional bene  t, the reduction of the toxicity. Perovskite of the structure CH 3 NH 3 Sn x Pb (1 À x ) I 3 was formed, 32 in which both the conduction and valence bands are shi  ed, and are shallower than those of titania by $À 0.4 eV. Power conversion e ffi ciency of 4.18% was achieved for the perovskite having the chemical formula of CH 3 NH 3 Sn 0.5 Pb 0.5 I 3 . In this con  guration, Spiro-OMETAD could not be success- fully employed as an HTM due to having a HOMO level that was deeper than that of the perovskite. This made hole injection into the HTM di ffi cult; therefore, P3HT was employed as an HTM instead. The IPCE curve shi  ed to 1060 nm in the best performing Sn-halide perovskite, compared to CH 3 NH 3 PbI 3 , which has an IPCE edge of 800 nm. Fig. 10 show the absorbance spectra of the several perovskite compositions. Two recent reports showed complete substitution of the Pb by Sn which results with a lead free perovskite solar cell. In these reports the power conversion e ffi ciency achieved was around 6%. 33,34 The di ff usion length of the electrons and holes was calculated 35,36 to be more than 1 m m for CH 3 NH 3 PbI 3 À x Cl x , justifying the high e ffi ciency of the perovskite solar cell. As mentioned earlier di ff erent measurements techniques provide a variation in the calculated di ff usion length, moreover the deposition technique also in  uence the di ff usion length as indicated by Yang et al. 43 and Snaith et al. 42 It is clear that long di ff usion length of both carriers enables the use of the planar architecture. The ability of perovskite to conduct both electrons and holes is an exceptional, distinctive property. This ability has been supported by the measurements of the long di ff usion length of both carriers. 20,35,36 The ability to conduct both electrons and holes simpli  es the device structure, enhancing the stability and eliminating the need to deal with in  ltration issues of the HTM. The structure of the hole conductor free perovskite solar cell is presented in Fig. 11. Several research groups have already demonstrated enhanced PV performance of these hole conductor free perovskite solar cells, the highest PCE observed was 10.85%. 37 – 40 The current – voltage curve of the best PV performance is presented in Fig. 11B. Additionally, it appears that changing the halides in the perovskite (which in  uence its band gap) does not reduce its ability to conduct the electrons and holes as proved by its implementation in the hole conductor free devices. 41 Vapor deposition technique was investigated for planar heterojunction perovskite solar cells achieving high e ffi ciency. 42 A comparison between the vapor deposition technique and the solution-processed technique was reported. For the vapor deposition technique, the  lm was stacked and uniform with a thickness of $ 330 nm. While for the solution-processed  lm, the thickness ranged between 50 – 410 nm and non-uniform. This is the explanation for the di ff erence in performance; in the case of solution processed  lm, the voids in the perovskite layer lead to direct contact between the polymer and the TiO 2 causing shunt path, that is, in part, responsible for lowering cell performance (Fig. 12). Except the ...